diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,31901 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 4553,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.812648680225853,
+      "learning_rate": 2.9197080291970804e-07,
+      "loss": 2.184,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.477363513420425,
+      "learning_rate": 5.839416058394161e-07,
+      "loss": 2.3547,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.667187992946625,
+      "learning_rate": 8.759124087591242e-07,
+      "loss": 2.3479,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.975762686028327,
+      "learning_rate": 1.1678832116788322e-06,
+      "loss": 2.5346,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.085661236932879,
+      "learning_rate": 1.4598540145985402e-06,
+      "loss": 2.4729,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.885329238897419,
+      "learning_rate": 1.7518248175182485e-06,
+      "loss": 2.2296,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.312140664960022,
+      "learning_rate": 2.0437956204379563e-06,
+      "loss": 2.3955,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.9862669851230255,
+      "learning_rate": 2.3357664233576643e-06,
+      "loss": 2.3144,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.660452935107435,
+      "learning_rate": 2.627737226277373e-06,
+      "loss": 2.2619,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.839865306189313,
+      "learning_rate": 2.9197080291970804e-06,
+      "loss": 2.0876,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.218399728082995,
+      "learning_rate": 3.2116788321167884e-06,
+      "loss": 2.1238,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.2177468696706075,
+      "learning_rate": 3.503649635036497e-06,
+      "loss": 1.9576,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.546161818531478,
+      "learning_rate": 3.7956204379562045e-06,
+      "loss": 1.8759,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.6162250465985224,
+      "learning_rate": 4.0875912408759126e-06,
+      "loss": 1.7203,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.134376192710703,
+      "learning_rate": 4.379562043795621e-06,
+      "loss": 1.6647,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.625610462231106,
+      "learning_rate": 4.671532846715329e-06,
+      "loss": 1.6871,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.90084477103016,
+      "learning_rate": 4.963503649635037e-06,
+      "loss": 1.5972,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.074734287101781,
+      "learning_rate": 5.255474452554746e-06,
+      "loss": 1.5691,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.3580179922502973,
+      "learning_rate": 5.547445255474453e-06,
+      "loss": 1.4562,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.1643910072609365,
+      "learning_rate": 5.839416058394161e-06,
+      "loss": 1.4711,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9961717444782574,
+      "learning_rate": 6.13138686131387e-06,
+      "loss": 1.5118,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9070903614201673,
+      "learning_rate": 6.423357664233577e-06,
+      "loss": 1.5549,
+      "step": 22
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1352623364264467,
+      "learning_rate": 6.715328467153285e-06,
+      "loss": 1.5622,
+      "step": 23
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1105178562720255,
+      "learning_rate": 7.007299270072994e-06,
+      "loss": 1.526,
+      "step": 24
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.943169696695018,
+      "learning_rate": 7.299270072992701e-06,
+      "loss": 1.308,
+      "step": 25
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9445026964211596,
+      "learning_rate": 7.591240875912409e-06,
+      "loss": 1.4118,
+      "step": 26
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6960437094301213,
+      "learning_rate": 7.883211678832117e-06,
+      "loss": 1.3774,
+      "step": 27
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.806297937832825,
+      "learning_rate": 8.175182481751825e-06,
+      "loss": 1.3251,
+      "step": 28
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.925963480221896,
+      "learning_rate": 8.467153284671533e-06,
+      "loss": 1.2795,
+      "step": 29
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.761128734910925,
+      "learning_rate": 8.759124087591241e-06,
+      "loss": 1.3643,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6845485873825556,
+      "learning_rate": 9.05109489051095e-06,
+      "loss": 1.1851,
+      "step": 31
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6619942577047109,
+      "learning_rate": 9.343065693430657e-06,
+      "loss": 1.3317,
+      "step": 32
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.684565641977221,
+      "learning_rate": 9.635036496350367e-06,
+      "loss": 1.4255,
+      "step": 33
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7809624690835513,
+      "learning_rate": 9.927007299270073e-06,
+      "loss": 1.24,
+      "step": 34
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5927230667953016,
+      "learning_rate": 1.0218978102189783e-05,
+      "loss": 1.1678,
+      "step": 35
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.708251796110156,
+      "learning_rate": 1.0510948905109491e-05,
+      "loss": 1.2783,
+      "step": 36
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.61533474408533,
+      "learning_rate": 1.0802919708029198e-05,
+      "loss": 1.2122,
+      "step": 37
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.534946029285862,
+      "learning_rate": 1.1094890510948906e-05,
+      "loss": 1.086,
+      "step": 38
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7288084999717537,
+      "learning_rate": 1.1386861313868614e-05,
+      "loss": 1.2476,
+      "step": 39
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6147373823283309,
+      "learning_rate": 1.1678832116788322e-05,
+      "loss": 1.2814,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6600660591309946,
+      "learning_rate": 1.1970802919708031e-05,
+      "loss": 1.3103,
+      "step": 41
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5377864291720114,
+      "learning_rate": 1.226277372262774e-05,
+      "loss": 1.2692,
+      "step": 42
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5523788716951157,
+      "learning_rate": 1.2554744525547446e-05,
+      "loss": 1.1435,
+      "step": 43
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5435394390617185,
+      "learning_rate": 1.2846715328467154e-05,
+      "loss": 1.1664,
+      "step": 44
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6764266187454582,
+      "learning_rate": 1.3138686131386862e-05,
+      "loss": 1.1319,
+      "step": 45
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5873994149609632,
+      "learning_rate": 1.343065693430657e-05,
+      "loss": 1.2192,
+      "step": 46
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8001990049344387,
+      "learning_rate": 1.372262773722628e-05,
+      "loss": 1.1245,
+      "step": 47
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8115756867886097,
+      "learning_rate": 1.4014598540145988e-05,
+      "loss": 1.2252,
+      "step": 48
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.577246544210627,
+      "learning_rate": 1.4306569343065696e-05,
+      "loss": 1.1797,
+      "step": 49
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5385036325197972,
+      "learning_rate": 1.4598540145985402e-05,
+      "loss": 1.0706,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5936588934494904,
+      "learning_rate": 1.489051094890511e-05,
+      "loss": 1.1354,
+      "step": 51
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5968142951430568,
+      "learning_rate": 1.5182481751824818e-05,
+      "loss": 1.1386,
+      "step": 52
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5492790883712821,
+      "learning_rate": 1.5474452554744528e-05,
+      "loss": 1.1177,
+      "step": 53
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6938133819094698,
+      "learning_rate": 1.5766423357664234e-05,
+      "loss": 1.178,
+      "step": 54
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4934152874568871,
+      "learning_rate": 1.6058394160583944e-05,
+      "loss": 1.1386,
+      "step": 55
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5257065527660285,
+      "learning_rate": 1.635036496350365e-05,
+      "loss": 0.9883,
+      "step": 56
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3479382206148167,
+      "learning_rate": 1.664233576642336e-05,
+      "loss": 0.9845,
+      "step": 57
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4843433577527922,
+      "learning_rate": 1.6934306569343066e-05,
+      "loss": 1.034,
+      "step": 58
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5578311593823568,
+      "learning_rate": 1.7226277372262773e-05,
+      "loss": 1.0024,
+      "step": 59
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6841397207836093,
+      "learning_rate": 1.7518248175182482e-05,
+      "loss": 1.1098,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5967023842041272,
+      "learning_rate": 1.7810218978102192e-05,
+      "loss": 1.0532,
+      "step": 61
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.458201329751063,
+      "learning_rate": 1.81021897810219e-05,
+      "loss": 1.0601,
+      "step": 62
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4896668719717416,
+      "learning_rate": 1.8394160583941608e-05,
+      "loss": 0.9523,
+      "step": 63
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.594477543540855,
+      "learning_rate": 1.8686131386861315e-05,
+      "loss": 1.0128,
+      "step": 64
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5053984138120584,
+      "learning_rate": 1.897810218978102e-05,
+      "loss": 1.0969,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4337321118516833,
+      "learning_rate": 1.9270072992700734e-05,
+      "loss": 0.978,
+      "step": 66
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5514572553484482,
+      "learning_rate": 1.956204379562044e-05,
+      "loss": 1.0519,
+      "step": 67
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5646238007711162,
+      "learning_rate": 1.9854014598540147e-05,
+      "loss": 1.0664,
+      "step": 68
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5916594463537463,
+      "learning_rate": 2.0145985401459857e-05,
+      "loss": 1.0785,
+      "step": 69
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3923766430469164,
+      "learning_rate": 2.0437956204379566e-05,
+      "loss": 0.9919,
+      "step": 70
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6016619395577614,
+      "learning_rate": 2.0729927007299273e-05,
+      "loss": 1.0618,
+      "step": 71
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5406397758843127,
+      "learning_rate": 2.1021897810218982e-05,
+      "loss": 1.0485,
+      "step": 72
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6912492146176519,
+      "learning_rate": 2.131386861313869e-05,
+      "loss": 1.1137,
+      "step": 73
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4428980274558936,
+      "learning_rate": 2.1605839416058395e-05,
+      "loss": 1.0067,
+      "step": 74
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4853785284781014,
+      "learning_rate": 2.1897810218978105e-05,
+      "loss": 1.0771,
+      "step": 75
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5161668261227093,
+      "learning_rate": 2.218978102189781e-05,
+      "loss": 1.0348,
+      "step": 76
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4922789700419026,
+      "learning_rate": 2.248175182481752e-05,
+      "loss": 1.0129,
+      "step": 77
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3967521784164405,
+      "learning_rate": 2.2773722627737227e-05,
+      "loss": 0.9375,
+      "step": 78
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4611514337742868,
+      "learning_rate": 2.3065693430656934e-05,
+      "loss": 1.0035,
+      "step": 79
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4726560071545307,
+      "learning_rate": 2.3357664233576643e-05,
+      "loss": 0.9118,
+      "step": 80
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.509851608690166,
+      "learning_rate": 2.3649635036496353e-05,
+      "loss": 0.9736,
+      "step": 81
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4854110314617655,
+      "learning_rate": 2.3941605839416063e-05,
+      "loss": 1.0382,
+      "step": 82
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4438095939193825,
+      "learning_rate": 2.423357664233577e-05,
+      "loss": 0.9426,
+      "step": 83
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.588853832486513,
+      "learning_rate": 2.452554744525548e-05,
+      "loss": 1.0912,
+      "step": 84
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4629616000754389,
+      "learning_rate": 2.4817518248175185e-05,
+      "loss": 0.9363,
+      "step": 85
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5126486586979293,
+      "learning_rate": 2.510948905109489e-05,
+      "loss": 0.8875,
+      "step": 86
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6442965886013419,
+      "learning_rate": 2.54014598540146e-05,
+      "loss": 1.0676,
+      "step": 87
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5786077686311497,
+      "learning_rate": 2.5693430656934308e-05,
+      "loss": 1.0544,
+      "step": 88
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5424329382663844,
+      "learning_rate": 2.5985401459854017e-05,
+      "loss": 1.0603,
+      "step": 89
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3763163074731595,
+      "learning_rate": 2.6277372262773724e-05,
+      "loss": 0.9341,
+      "step": 90
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.36645217739804,
+      "learning_rate": 2.656934306569343e-05,
+      "loss": 0.9277,
+      "step": 91
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.476560239436927,
+      "learning_rate": 2.686131386861314e-05,
+      "loss": 0.9549,
+      "step": 92
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.651790661601902,
+      "learning_rate": 2.7153284671532846e-05,
+      "loss": 1.0434,
+      "step": 93
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.564711569632748,
+      "learning_rate": 2.744525547445256e-05,
+      "loss": 0.9528,
+      "step": 94
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3892575516439725,
+      "learning_rate": 2.7737226277372266e-05,
+      "loss": 0.9055,
+      "step": 95
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.47338517258931,
+      "learning_rate": 2.8029197080291975e-05,
+      "loss": 0.9045,
+      "step": 96
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4283014468117288,
+      "learning_rate": 2.832116788321168e-05,
+      "loss": 0.9418,
+      "step": 97
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4984556035926158,
+      "learning_rate": 2.861313868613139e-05,
+      "loss": 1.0147,
+      "step": 98
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5189781457158873,
+      "learning_rate": 2.8905109489051098e-05,
+      "loss": 0.9159,
+      "step": 99
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4628925807313362,
+      "learning_rate": 2.9197080291970804e-05,
+      "loss": 0.9209,
+      "step": 100
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5929870275147635,
+      "learning_rate": 2.9489051094890514e-05,
+      "loss": 0.9557,
+      "step": 101
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4503689132203692,
+      "learning_rate": 2.978102189781022e-05,
+      "loss": 0.8777,
+      "step": 102
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.387439894877902,
+      "learning_rate": 3.007299270072993e-05,
+      "loss": 0.8537,
+      "step": 103
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4809151733801909,
+      "learning_rate": 3.0364963503649636e-05,
+      "loss": 0.8067,
+      "step": 104
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4872114687064437,
+      "learning_rate": 3.0656934306569346e-05,
+      "loss": 0.9561,
+      "step": 105
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3704334201331871,
+      "learning_rate": 3.0948905109489056e-05,
+      "loss": 0.832,
+      "step": 106
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2970664101409684,
+      "learning_rate": 3.1240875912408765e-05,
+      "loss": 0.8222,
+      "step": 107
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4920176464295187,
+      "learning_rate": 3.153284671532847e-05,
+      "loss": 0.9368,
+      "step": 108
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3703280318629414,
+      "learning_rate": 3.182481751824818e-05,
+      "loss": 0.9345,
+      "step": 109
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3550291640320893,
+      "learning_rate": 3.211678832116789e-05,
+      "loss": 0.8403,
+      "step": 110
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4486058010850487,
+      "learning_rate": 3.24087591240876e-05,
+      "loss": 0.8456,
+      "step": 111
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5455373509810715,
+      "learning_rate": 3.27007299270073e-05,
+      "loss": 0.9237,
+      "step": 112
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.562125351812911,
+      "learning_rate": 3.299270072992701e-05,
+      "loss": 0.8496,
+      "step": 113
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.409157185100568,
+      "learning_rate": 3.328467153284672e-05,
+      "loss": 0.8066,
+      "step": 114
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5594940262435777,
+      "learning_rate": 3.357664233576642e-05,
+      "loss": 0.8461,
+      "step": 115
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.587613727971374,
+      "learning_rate": 3.386861313868613e-05,
+      "loss": 0.8593,
+      "step": 116
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4364812807164014,
+      "learning_rate": 3.416058394160584e-05,
+      "loss": 0.9084,
+      "step": 117
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4354830980730897,
+      "learning_rate": 3.4452554744525545e-05,
+      "loss": 0.8356,
+      "step": 118
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.370928545179788,
+      "learning_rate": 3.474452554744526e-05,
+      "loss": 0.8,
+      "step": 119
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4805453237896125,
+      "learning_rate": 3.5036496350364965e-05,
+      "loss": 0.8707,
+      "step": 120
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3529579147537623,
+      "learning_rate": 3.5328467153284675e-05,
+      "loss": 0.7764,
+      "step": 121
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.547681039749328,
+      "learning_rate": 3.5620437956204384e-05,
+      "loss": 0.8775,
+      "step": 122
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5332301823846584,
+      "learning_rate": 3.5912408759124094e-05,
+      "loss": 0.9876,
+      "step": 123
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.512126463485808,
+      "learning_rate": 3.62043795620438e-05,
+      "loss": 0.9528,
+      "step": 124
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4554233218365884,
+      "learning_rate": 3.649635036496351e-05,
+      "loss": 0.8889,
+      "step": 125
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3808554901543213,
+      "learning_rate": 3.6788321167883217e-05,
+      "loss": 0.7677,
+      "step": 126
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4277159223916567,
+      "learning_rate": 3.708029197080292e-05,
+      "loss": 0.7742,
+      "step": 127
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.49818015966614,
+      "learning_rate": 3.737226277372263e-05,
+      "loss": 0.7219,
+      "step": 128
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6176138053372855,
+      "learning_rate": 3.766423357664234e-05,
+      "loss": 0.9805,
+      "step": 129
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6421738352692505,
+      "learning_rate": 3.795620437956204e-05,
+      "loss": 0.9051,
+      "step": 130
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3299414386620134,
+      "learning_rate": 3.824817518248176e-05,
+      "loss": 0.7942,
+      "step": 131
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.390975222259264,
+      "learning_rate": 3.854014598540147e-05,
+      "loss": 0.9009,
+      "step": 132
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.33912799723986,
+      "learning_rate": 3.883211678832117e-05,
+      "loss": 0.8367,
+      "step": 133
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.302621175946519,
+      "learning_rate": 3.912408759124088e-05,
+      "loss": 0.7549,
+      "step": 134
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.405998800941922,
+      "learning_rate": 3.941605839416059e-05,
+      "loss": 0.7584,
+      "step": 135
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5325238903977623,
+      "learning_rate": 3.9708029197080294e-05,
+      "loss": 0.8071,
+      "step": 136
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4485039194756526,
+      "learning_rate": 4e-05,
+      "loss": 0.7941,
+      "step": 137
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4259171930589734,
+      "learning_rate": 3.99999949389387e-05,
+      "loss": 0.8034,
+      "step": 138
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3034281791915299,
+      "learning_rate": 3.999997975575736e-05,
+      "loss": 0.7876,
+      "step": 139
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5026318191108297,
+      "learning_rate": 3.9999954450463665e-05,
+      "loss": 0.8762,
+      "step": 140
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3526883590322558,
+      "learning_rate": 3.9999919023070414e-05,
+      "loss": 0.8062,
+      "step": 141
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.385679946587705,
+      "learning_rate": 3.999987347359555e-05,
+      "loss": 0.7728,
+      "step": 142
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4683155472393012,
+      "learning_rate": 3.999981780206212e-05,
+      "loss": 0.7596,
+      "step": 143
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3424429191942802,
+      "learning_rate": 3.99997520084983e-05,
+      "loss": 0.7176,
+      "step": 144
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4314773279087492,
+      "learning_rate": 3.999967609293739e-05,
+      "loss": 0.9188,
+      "step": 145
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3056482672181737,
+      "learning_rate": 3.99995900554178e-05,
+      "loss": 0.7416,
+      "step": 146
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4585949890098764,
+      "learning_rate": 3.999949389598309e-05,
+      "loss": 0.824,
+      "step": 147
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.337799603783892,
+      "learning_rate": 3.999938761468192e-05,
+      "loss": 0.6902,
+      "step": 148
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5249693382807843,
+      "learning_rate": 3.9999271211568084e-05,
+      "loss": 0.7957,
+      "step": 149
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5249336134421996,
+      "learning_rate": 3.999914468670048e-05,
+      "loss": 0.7885,
+      "step": 150
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3788826483849044,
+      "learning_rate": 3.999900804014317e-05,
+      "loss": 0.807,
+      "step": 151
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.408791428424769,
+      "learning_rate": 3.9998861271965285e-05,
+      "loss": 0.7954,
+      "step": 152
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.288136513782112,
+      "learning_rate": 3.999870438224111e-05,
+      "loss": 0.7176,
+      "step": 153
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3975412917477223,
+      "learning_rate": 3.999853737105007e-05,
+      "loss": 0.8284,
+      "step": 154
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5646770569437787,
+      "learning_rate": 3.9998360238476655e-05,
+      "loss": 0.8673,
+      "step": 155
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4448607692497883,
+      "learning_rate": 3.999817298461054e-05,
+      "loss": 0.8099,
+      "step": 156
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4221062105434186,
+      "learning_rate": 3.999797560954649e-05,
+      "loss": 0.8186,
+      "step": 157
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3541790105795066,
+      "learning_rate": 3.999776811338439e-05,
+      "loss": 0.7527,
+      "step": 158
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.310569615173891,
+      "learning_rate": 3.999755049622926e-05,
+      "loss": 0.6947,
+      "step": 159
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2855053459421322,
+      "learning_rate": 3.9997322758191244e-05,
+      "loss": 0.6425,
+      "step": 160
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.456957613232235,
+      "learning_rate": 3.999708489938559e-05,
+      "loss": 0.7138,
+      "step": 161
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3154442325459081,
+      "learning_rate": 3.999683691993268e-05,
+      "loss": 0.6217,
+      "step": 162
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3707120967385478,
+      "learning_rate": 3.999657881995802e-05,
+      "loss": 0.734,
+      "step": 163
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3464468131540774,
+      "learning_rate": 3.9996310599592244e-05,
+      "loss": 0.8018,
+      "step": 164
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3131685370844959,
+      "learning_rate": 3.9996032258971097e-05,
+      "loss": 0.7634,
+      "step": 165
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3014680349591794,
+      "learning_rate": 3.9995743798235445e-05,
+      "loss": 0.7904,
+      "step": 166
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3306836781749012,
+      "learning_rate": 3.999544521753128e-05,
+      "loss": 0.7516,
+      "step": 167
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3280616745156995,
+      "learning_rate": 3.999513651700971e-05,
+      "loss": 0.7404,
+      "step": 168
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.382294035015746,
+      "learning_rate": 3.999481769682699e-05,
+      "loss": 0.7423,
+      "step": 169
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.289816572248431,
+      "learning_rate": 3.9994488757144454e-05,
+      "loss": 0.6838,
+      "step": 170
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.351114849421247,
+      "learning_rate": 3.999414969812859e-05,
+      "loss": 0.7064,
+      "step": 171
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.112602374906335,
+      "learning_rate": 3.9993800519951e-05,
+      "loss": 0.5968,
+      "step": 172
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3237891836733933,
+      "learning_rate": 3.99934412227884e-05,
+      "loss": 0.7149,
+      "step": 173
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3127998282064572,
+      "learning_rate": 3.999307180682264e-05,
+      "loss": 0.6785,
+      "step": 174
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3832342075926638,
+      "learning_rate": 3.9992692272240684e-05,
+      "loss": 0.681,
+      "step": 175
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2908340765054303,
+      "learning_rate": 3.99923026192346e-05,
+      "loss": 0.6676,
+      "step": 176
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3544599313339964,
+      "learning_rate": 3.999190284800162e-05,
+      "loss": 0.7121,
+      "step": 177
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2556712245455064,
+      "learning_rate": 3.9991492958744046e-05,
+      "loss": 0.7016,
+      "step": 178
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3426939784728575,
+      "learning_rate": 3.9991072951669334e-05,
+      "loss": 0.7133,
+      "step": 179
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2358437986095847,
+      "learning_rate": 3.999064282699006e-05,
+      "loss": 0.6931,
+      "step": 180
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4329296933567661,
+      "learning_rate": 3.999020258492391e-05,
+      "loss": 0.7873,
+      "step": 181
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3413303680374504,
+      "learning_rate": 3.998975222569368e-05,
+      "loss": 0.668,
+      "step": 182
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3699727488682398,
+      "learning_rate": 3.9989291749527314e-05,
+      "loss": 0.744,
+      "step": 183
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2672579556409398,
+      "learning_rate": 3.998882115665786e-05,
+      "loss": 0.7541,
+      "step": 184
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1930919291337678,
+      "learning_rate": 3.998834044732348e-05,
+      "loss": 0.6611,
+      "step": 185
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3237952171145917,
+      "learning_rate": 3.9987849621767473e-05,
+      "loss": 0.6954,
+      "step": 186
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3491031122269155,
+      "learning_rate": 3.998734868023825e-05,
+      "loss": 0.6532,
+      "step": 187
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2474273911255802,
+      "learning_rate": 3.998683762298933e-05,
+      "loss": 0.614,
+      "step": 188
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4339242497046318,
+      "learning_rate": 3.9986316450279365e-05,
+      "loss": 0.7592,
+      "step": 189
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3155313182105586,
+      "learning_rate": 3.9985785162372135e-05,
+      "loss": 0.6463,
+      "step": 190
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2720545649011348,
+      "learning_rate": 3.998524375953651e-05,
+      "loss": 0.6925,
+      "step": 191
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3250032424887084,
+      "learning_rate": 3.998469224204652e-05,
+      "loss": 0.6736,
+      "step": 192
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3797040149914255,
+      "learning_rate": 3.998413061018126e-05,
+      "loss": 0.7955,
+      "step": 193
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2805602845669852,
+      "learning_rate": 3.9983558864225005e-05,
+      "loss": 0.7063,
+      "step": 194
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4422115567764362,
+      "learning_rate": 3.9982977004467106e-05,
+      "loss": 0.6954,
+      "step": 195
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.207992023189052,
+      "learning_rate": 3.998238503120205e-05,
+      "loss": 0.6594,
+      "step": 196
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.363506484348838,
+      "learning_rate": 3.998178294472944e-05,
+      "loss": 0.6448,
+      "step": 197
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.322121278442904,
+      "learning_rate": 3.998117074535398e-05,
+      "loss": 0.6708,
+      "step": 198
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.349313308768587,
+      "learning_rate": 3.9980548433385525e-05,
+      "loss": 0.6601,
+      "step": 199
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3259708662327934,
+      "learning_rate": 3.997991600913903e-05,
+      "loss": 0.6067,
+      "step": 200
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2868747989456146,
+      "learning_rate": 3.9979273472934556e-05,
+      "loss": 0.6959,
+      "step": 201
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2581457798279778,
+      "learning_rate": 3.9978620825097306e-05,
+      "loss": 0.6414,
+      "step": 202
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2659091865878211,
+      "learning_rate": 3.997795806595758e-05,
+      "loss": 0.6121,
+      "step": 203
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2788722607588214,
+      "learning_rate": 3.9977285195850816e-05,
+      "loss": 0.7051,
+      "step": 204
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.268616331979891,
+      "learning_rate": 3.9976602215117554e-05,
+      "loss": 0.5972,
+      "step": 205
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2646466465057298,
+      "learning_rate": 3.997590912410345e-05,
+      "loss": 0.5288,
+      "step": 206
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4566461703664726,
+      "learning_rate": 3.997520592315929e-05,
+      "loss": 0.706,
+      "step": 207
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.298420685844027,
+      "learning_rate": 3.997449261264095e-05,
+      "loss": 0.673,
+      "step": 208
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2174582849832838,
+      "learning_rate": 3.997376919290946e-05,
+      "loss": 0.5606,
+      "step": 209
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.164802930233094,
+      "learning_rate": 3.997303566433094e-05,
+      "loss": 0.5624,
+      "step": 210
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1716197435217768,
+      "learning_rate": 3.997229202727663e-05,
+      "loss": 0.5921,
+      "step": 211
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2121105296122996,
+      "learning_rate": 3.99715382821229e-05,
+      "loss": 0.6432,
+      "step": 212
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2524098531752716,
+      "learning_rate": 3.997077442925122e-05,
+      "loss": 0.6192,
+      "step": 213
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1942955627853422,
+      "learning_rate": 3.997000046904817e-05,
+      "loss": 0.6344,
+      "step": 214
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2520875665187947,
+      "learning_rate": 3.996921640190547e-05,
+      "loss": 0.6857,
+      "step": 215
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.177316721281303,
+      "learning_rate": 3.996842222821994e-05,
+      "loss": 0.5736,
+      "step": 216
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.224012311462119,
+      "learning_rate": 3.9967617948393504e-05,
+      "loss": 0.6102,
+      "step": 217
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2894347607016468,
+      "learning_rate": 3.996680356283322e-05,
+      "loss": 0.6487,
+      "step": 218
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3895495252866978,
+      "learning_rate": 3.996597907195126e-05,
+      "loss": 0.661,
+      "step": 219
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.32050710293565,
+      "learning_rate": 3.996514447616489e-05,
+      "loss": 0.6459,
+      "step": 220
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.180191235644313,
+      "learning_rate": 3.996429977589653e-05,
+      "loss": 0.6245,
+      "step": 221
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.249228191518979,
+      "learning_rate": 3.9963444971573656e-05,
+      "loss": 0.6258,
+      "step": 222
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2439700598103596,
+      "learning_rate": 3.996258006362891e-05,
+      "loss": 0.6282,
+      "step": 223
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.262995307376024,
+      "learning_rate": 3.996170505250002e-05,
+      "loss": 0.5968,
+      "step": 224
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2049038110227466,
+      "learning_rate": 3.9960819938629834e-05,
+      "loss": 0.6372,
+      "step": 225
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1953981655564385,
+      "learning_rate": 3.995992472246632e-05,
+      "loss": 0.5788,
+      "step": 226
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1900866094695106,
+      "learning_rate": 3.995901940446254e-05,
+      "loss": 0.571,
+      "step": 227
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2301740989906733,
+      "learning_rate": 3.995810398507669e-05,
+      "loss": 0.5917,
+      "step": 228
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2523850099236096,
+      "learning_rate": 3.995717846477207e-05,
+      "loss": 0.5853,
+      "step": 229
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2378971695538876,
+      "learning_rate": 3.9956242844017094e-05,
+      "loss": 0.5738,
+      "step": 230
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1288816455944937,
+      "learning_rate": 3.995529712328528e-05,
+      "loss": 0.5542,
+      "step": 231
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1148924829622244,
+      "learning_rate": 3.995434130305526e-05,
+      "loss": 0.5093,
+      "step": 232
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2799793548707197,
+      "learning_rate": 3.995337538381079e-05,
+      "loss": 0.5565,
+      "step": 233
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.231420239912531,
+      "learning_rate": 3.995239936604072e-05,
+      "loss": 0.5714,
+      "step": 234
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3992004870442785,
+      "learning_rate": 3.995141325023902e-05,
+      "loss": 0.6154,
+      "step": 235
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1949779565924818,
+      "learning_rate": 3.995041703690477e-05,
+      "loss": 0.5143,
+      "step": 236
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.198730767088812,
+      "learning_rate": 3.994941072654215e-05,
+      "loss": 0.4691,
+      "step": 237
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2731486267833902,
+      "learning_rate": 3.9948394319660485e-05,
+      "loss": 0.5808,
+      "step": 238
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3653911887989123,
+      "learning_rate": 3.994736781677416e-05,
+      "loss": 0.65,
+      "step": 239
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3105154929120355,
+      "learning_rate": 3.994633121840271e-05,
+      "loss": 0.6434,
+      "step": 240
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2381977811292837,
+      "learning_rate": 3.994528452507076e-05,
+      "loss": 0.5829,
+      "step": 241
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1402137746645646,
+      "learning_rate": 3.994422773730803e-05,
+      "loss": 0.509,
+      "step": 242
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.208429458784759,
+      "learning_rate": 3.99431608556494e-05,
+      "loss": 0.6454,
+      "step": 243
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1542773043398307,
+      "learning_rate": 3.99420838806348e-05,
+      "loss": 0.5745,
+      "step": 244
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0919978139422268,
+      "learning_rate": 3.99409968128093e-05,
+      "loss": 0.4986,
+      "step": 245
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.113345977507724,
+      "learning_rate": 3.993989965272308e-05,
+      "loss": 0.5712,
+      "step": 246
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3593819278233579,
+      "learning_rate": 3.99387924009314e-05,
+      "loss": 0.6304,
+      "step": 247
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1946094913942888,
+      "learning_rate": 3.9937675057994666e-05,
+      "loss": 0.5803,
+      "step": 248
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1889538648909175,
+      "learning_rate": 3.993654762447837e-05,
+      "loss": 0.5862,
+      "step": 249
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0987200486198738,
+      "learning_rate": 3.9935410100953105e-05,
+      "loss": 0.4049,
+      "step": 250
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2101428962694776,
+      "learning_rate": 3.993426248799458e-05,
+      "loss": 0.57,
+      "step": 251
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.234140639450306,
+      "learning_rate": 3.993310478618361e-05,
+      "loss": 0.5653,
+      "step": 252
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2708140773043917,
+      "learning_rate": 3.993193699610612e-05,
+      "loss": 0.588,
+      "step": 253
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2785513299746012,
+      "learning_rate": 3.9930759118353124e-05,
+      "loss": 0.6005,
+      "step": 254
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4274452845681305,
+      "learning_rate": 3.992957115352077e-05,
+      "loss": 0.6807,
+      "step": 255
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2004749788586693,
+      "learning_rate": 3.992837310221028e-05,
+      "loss": 0.587,
+      "step": 256
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1666004752501002,
+      "learning_rate": 3.9927164965028006e-05,
+      "loss": 0.4766,
+      "step": 257
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.18853834833233,
+      "learning_rate": 3.9925946742585385e-05,
+      "loss": 0.5422,
+      "step": 258
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2094667424537495,
+      "learning_rate": 3.9924718435498964e-05,
+      "loss": 0.559,
+      "step": 259
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2747637155785871,
+      "learning_rate": 3.9923480044390405e-05,
+      "loss": 0.5448,
+      "step": 260
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1455729564410346,
+      "learning_rate": 3.9922231569886464e-05,
+      "loss": 0.4595,
+      "step": 261
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2280623868943858,
+      "learning_rate": 3.9920973012619e-05,
+      "loss": 0.4862,
+      "step": 262
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0514020822946684,
+      "learning_rate": 3.9919704373224984e-05,
+      "loss": 0.4702,
+      "step": 263
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1276534365607926,
+      "learning_rate": 3.991842565234647e-05,
+      "loss": 0.5056,
+      "step": 264
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1465460843532795,
+      "learning_rate": 3.991713685063063e-05,
+      "loss": 0.4813,
+      "step": 265
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0890863993360724,
+      "learning_rate": 3.991583796872974e-05,
+      "loss": 0.5131,
+      "step": 266
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.291958832256156,
+      "learning_rate": 3.991452900730116e-05,
+      "loss": 0.5547,
+      "step": 267
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1784316698414166,
+      "learning_rate": 3.991320996700737e-05,
+      "loss": 0.4902,
+      "step": 268
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1724553006564273,
+      "learning_rate": 3.991188084851596e-05,
+      "loss": 0.4878,
+      "step": 269
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2264566558071297,
+      "learning_rate": 3.991054165249958e-05,
+      "loss": 0.5361,
+      "step": 270
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.204194868916881,
+      "learning_rate": 3.990919237963602e-05,
+      "loss": 0.4693,
+      "step": 271
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.165114880738373,
+      "learning_rate": 3.9907833030608153e-05,
+      "loss": 0.4728,
+      "step": 272
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2662323801183593,
+      "learning_rate": 3.990646360610395e-05,
+      "loss": 0.5497,
+      "step": 273
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3041430039539232,
+      "learning_rate": 3.9905084106816494e-05,
+      "loss": 0.5175,
+      "step": 274
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2265639578452223,
+      "learning_rate": 3.990369453344394e-05,
+      "loss": 0.5045,
+      "step": 275
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.100467868486642,
+      "learning_rate": 3.9902294886689576e-05,
+      "loss": 0.5279,
+      "step": 276
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2000185011391165,
+      "learning_rate": 3.990088516726177e-05,
+      "loss": 0.508,
+      "step": 277
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2701732714039602,
+      "learning_rate": 3.9899465375873985e-05,
+      "loss": 0.5277,
+      "step": 278
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3908324783365231,
+      "learning_rate": 3.989803551324479e-05,
+      "loss": 0.6504,
+      "step": 279
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2523779661533918,
+      "learning_rate": 3.989659558009784e-05,
+      "loss": 0.5366,
+      "step": 280
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.122515159453593,
+      "learning_rate": 3.98951455771619e-05,
+      "loss": 0.4976,
+      "step": 281
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2132206711933338,
+      "learning_rate": 3.989368550517083e-05,
+      "loss": 0.5127,
+      "step": 282
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0510171391498828,
+      "learning_rate": 3.989221536486357e-05,
+      "loss": 0.3852,
+      "step": 283
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.084079497691504,
+      "learning_rate": 3.989073515698417e-05,
+      "loss": 0.4238,
+      "step": 284
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2574323944666181,
+      "learning_rate": 3.988924488228178e-05,
+      "loss": 0.4225,
+      "step": 285
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2767442702237093,
+      "learning_rate": 3.988774454151063e-05,
+      "loss": 0.4731,
+      "step": 286
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1875667553258227,
+      "learning_rate": 3.988623413543006e-05,
+      "loss": 0.4941,
+      "step": 287
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1722931179545868,
+      "learning_rate": 3.9884713664804485e-05,
+      "loss": 0.5455,
+      "step": 288
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1571276658690817,
+      "learning_rate": 3.9883183130403424e-05,
+      "loss": 0.5155,
+      "step": 289
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.273868698414468,
+      "learning_rate": 3.98816425330015e-05,
+      "loss": 0.6212,
+      "step": 290
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1349200407927071,
+      "learning_rate": 3.9880091873378416e-05,
+      "loss": 0.5138,
+      "step": 291
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2036075862726578,
+      "learning_rate": 3.9878531152318966e-05,
+      "loss": 0.4458,
+      "step": 292
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2523328946048258,
+      "learning_rate": 3.987696037061304e-05,
+      "loss": 0.5277,
+      "step": 293
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0786026711932888,
+      "learning_rate": 3.9875379529055624e-05,
+      "loss": 0.4084,
+      "step": 294
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3213455402401926,
+      "learning_rate": 3.987378862844679e-05,
+      "loss": 0.5018,
+      "step": 295
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3459430807302186,
+      "learning_rate": 3.987218766959171e-05,
+      "loss": 0.5742,
+      "step": 296
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.16850411383607,
+      "learning_rate": 3.987057665330063e-05,
+      "loss": 0.5138,
+      "step": 297
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1359638447179328,
+      "learning_rate": 3.986895558038889e-05,
+      "loss": 0.4713,
+      "step": 298
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2118807164844807,
+      "learning_rate": 3.986732445167694e-05,
+      "loss": 0.4895,
+      "step": 299
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1783109806499985,
+      "learning_rate": 3.9865683267990295e-05,
+      "loss": 0.4614,
+      "step": 300
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2716055957300278,
+      "learning_rate": 3.986403203015957e-05,
+      "loss": 0.4847,
+      "step": 301
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.293195952307599,
+      "learning_rate": 3.9862370739020455e-05,
+      "loss": 0.5353,
+      "step": 302
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2657770607105185,
+      "learning_rate": 3.9860699395413764e-05,
+      "loss": 0.499,
+      "step": 303
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0391015640602386,
+      "learning_rate": 3.985901800018535e-05,
+      "loss": 0.4169,
+      "step": 304
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.215289852748404,
+      "learning_rate": 3.98573265541862e-05,
+      "loss": 0.5027,
+      "step": 305
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1711716130426661,
+      "learning_rate": 3.985562505827235e-05,
+      "loss": 0.4303,
+      "step": 306
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1602282774687112,
+      "learning_rate": 3.985391351330494e-05,
+      "loss": 0.4306,
+      "step": 307
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0702587587550187,
+      "learning_rate": 3.985219192015019e-05,
+      "loss": 0.4604,
+      "step": 308
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.163744991959594,
+      "learning_rate": 3.985046027967943e-05,
+      "loss": 0.5091,
+      "step": 309
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0533315348509231,
+      "learning_rate": 3.984871859276902e-05,
+      "loss": 0.4756,
+      "step": 310
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.164256649339924,
+      "learning_rate": 3.984696686030046e-05,
+      "loss": 0.4769,
+      "step": 311
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.099305458146492,
+      "learning_rate": 3.9845205083160315e-05,
+      "loss": 0.441,
+      "step": 312
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0560989542749786,
+      "learning_rate": 3.984343326224022e-05,
+      "loss": 0.4744,
+      "step": 313
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2728584236684755,
+      "learning_rate": 3.9841651398436907e-05,
+      "loss": 0.5036,
+      "step": 314
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.092996828428869,
+      "learning_rate": 3.983985949265219e-05,
+      "loss": 0.4692,
+      "step": 315
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0802089067026577,
+      "learning_rate": 3.983805754579297e-05,
+      "loss": 0.3904,
+      "step": 316
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1006761055323477,
+      "learning_rate": 3.98362455587712e-05,
+      "loss": 0.4057,
+      "step": 317
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.109374892543734,
+      "learning_rate": 3.9834423532503975e-05,
+      "loss": 0.4865,
+      "step": 318
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0895982144411165,
+      "learning_rate": 3.9832591467913405e-05,
+      "loss": 0.4018,
+      "step": 319
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1037257711790474,
+      "learning_rate": 3.9830749365926716e-05,
+      "loss": 0.4063,
+      "step": 320
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.137745763617826,
+      "learning_rate": 3.982889722747621e-05,
+      "loss": 0.4296,
+      "step": 321
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0638032101547767,
+      "learning_rate": 3.9827035053499264e-05,
+      "loss": 0.4478,
+      "step": 322
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1201766246933287,
+      "learning_rate": 3.982516284493834e-05,
+      "loss": 0.4129,
+      "step": 323
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0943915529159343,
+      "learning_rate": 3.982328060274097e-05,
+      "loss": 0.4141,
+      "step": 324
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0233495543454487,
+      "learning_rate": 3.982138832785976e-05,
+      "loss": 0.4061,
+      "step": 325
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2608735646201592,
+      "learning_rate": 3.981948602125242e-05,
+      "loss": 0.4925,
+      "step": 326
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1493183831878495,
+      "learning_rate": 3.98175736838817e-05,
+      "loss": 0.465,
+      "step": 327
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1636103321511055,
+      "learning_rate": 3.981565131671546e-05,
+      "loss": 0.5018,
+      "step": 328
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0613981592394197,
+      "learning_rate": 3.981371892072661e-05,
+      "loss": 0.4161,
+      "step": 329
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1859586148399008,
+      "learning_rate": 3.981177649689317e-05,
+      "loss": 0.4475,
+      "step": 330
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0116317888392918,
+      "learning_rate": 3.980982404619819e-05,
+      "loss": 0.42,
+      "step": 331
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1123049018464455,
+      "learning_rate": 3.9807861569629815e-05,
+      "loss": 0.4625,
+      "step": 332
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0650799226994028,
+      "learning_rate": 3.980588906818129e-05,
+      "loss": 0.3434,
+      "step": 333
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1856989001065437,
+      "learning_rate": 3.980390654285088e-05,
+      "loss": 0.4546,
+      "step": 334
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1165114271140357,
+      "learning_rate": 3.980191399464198e-05,
+      "loss": 0.4023,
+      "step": 335
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1062497456868197,
+      "learning_rate": 3.979991142456302e-05,
+      "loss": 0.3757,
+      "step": 336
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1251817132637552,
+      "learning_rate": 3.9797898833627514e-05,
+      "loss": 0.4204,
+      "step": 337
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.076857499009745,
+      "learning_rate": 3.979587622285404e-05,
+      "loss": 0.3577,
+      "step": 338
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.149526689402929,
+      "learning_rate": 3.979384359326626e-05,
+      "loss": 0.5064,
+      "step": 339
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0409359785175913,
+      "learning_rate": 3.97918009458929e-05,
+      "loss": 0.3905,
+      "step": 340
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0025633502264812,
+      "learning_rate": 3.9789748281767754e-05,
+      "loss": 0.3792,
+      "step": 341
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1626214445933571,
+      "learning_rate": 3.978768560192969e-05,
+      "loss": 0.4412,
+      "step": 342
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2902301008688999,
+      "learning_rate": 3.978561290742265e-05,
+      "loss": 0.4021,
+      "step": 343
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1143634555494666,
+      "learning_rate": 3.978353019929562e-05,
+      "loss": 0.4048,
+      "step": 344
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.195219148468686,
+      "learning_rate": 3.978143747860269e-05,
+      "loss": 0.4526,
+      "step": 345
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0725898738809163,
+      "learning_rate": 3.977933474640298e-05,
+      "loss": 0.4468,
+      "step": 346
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1025137615426255,
+      "learning_rate": 3.9777222003760714e-05,
+      "loss": 0.4239,
+      "step": 347
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0611781143806014,
+      "learning_rate": 3.977509925174515e-05,
+      "loss": 0.4622,
+      "step": 348
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1455237866366754,
+      "learning_rate": 3.977296649143064e-05,
+      "loss": 0.4485,
+      "step": 349
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0029408603777608,
+      "learning_rate": 3.9770823723896574e-05,
+      "loss": 0.3871,
+      "step": 350
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0213749026395467,
+      "learning_rate": 3.976867095022742e-05,
+      "loss": 0.4471,
+      "step": 351
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.148935430626447,
+      "learning_rate": 3.9766508171512715e-05,
+      "loss": 0.4755,
+      "step": 352
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0381439853059833,
+      "learning_rate": 3.976433538884706e-05,
+      "loss": 0.376,
+      "step": 353
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2431017791896115,
+      "learning_rate": 3.97621526033301e-05,
+      "loss": 0.478,
+      "step": 354
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1850241399471089,
+      "learning_rate": 3.9759959816066575e-05,
+      "loss": 0.4424,
+      "step": 355
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0482157192652117,
+      "learning_rate": 3.975775702816625e-05,
+      "loss": 0.3953,
+      "step": 356
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.173974001130304,
+      "learning_rate": 3.975554424074397e-05,
+      "loss": 0.4347,
+      "step": 357
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1813716361579474,
+      "learning_rate": 3.975332145491965e-05,
+      "loss": 0.4359,
+      "step": 358
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1235249173635753,
+      "learning_rate": 3.975108867181826e-05,
+      "loss": 0.3873,
+      "step": 359
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1060145348426775,
+      "learning_rate": 3.974884589256981e-05,
+      "loss": 0.3915,
+      "step": 360
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2229112353142,
+      "learning_rate": 3.97465931183094e-05,
+      "loss": 0.4774,
+      "step": 361
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0023051158568568,
+      "learning_rate": 3.9744330350177156e-05,
+      "loss": 0.3639,
+      "step": 362
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1390226422572052,
+      "learning_rate": 3.974205758931828e-05,
+      "loss": 0.3841,
+      "step": 363
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.079691315794632,
+      "learning_rate": 3.973977483688305e-05,
+      "loss": 0.3615,
+      "step": 364
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1729070569328504,
+      "learning_rate": 3.9737482094026764e-05,
+      "loss": 0.3758,
+      "step": 365
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.250347184603744,
+      "learning_rate": 3.9735179361909803e-05,
+      "loss": 0.4381,
+      "step": 366
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0533736911619336,
+      "learning_rate": 3.9732866641697586e-05,
+      "loss": 0.3279,
+      "step": 367
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.454656798533515,
+      "learning_rate": 3.9730543934560595e-05,
+      "loss": 0.3942,
+      "step": 368
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.079306576961687,
+      "learning_rate": 3.9728211241674363e-05,
+      "loss": 0.3655,
+      "step": 369
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1195722616367776,
+      "learning_rate": 3.972586856421949e-05,
+      "loss": 0.4542,
+      "step": 370
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1305671274736946,
+      "learning_rate": 3.9723515903381625e-05,
+      "loss": 0.4273,
+      "step": 371
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.050735189002656,
+      "learning_rate": 3.9721153260351446e-05,
+      "loss": 0.3854,
+      "step": 372
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0818517653131625,
+      "learning_rate": 3.971878063632471e-05,
+      "loss": 0.395,
+      "step": 373
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0001708123235014,
+      "learning_rate": 3.971639803250221e-05,
+      "loss": 0.4188,
+      "step": 374
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0316954142109276,
+      "learning_rate": 3.9714005450089815e-05,
+      "loss": 0.3553,
+      "step": 375
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0061576919391308,
+      "learning_rate": 3.971160289029841e-05,
+      "loss": 0.3474,
+      "step": 376
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.07023146942143,
+      "learning_rate": 3.9709190354343936e-05,
+      "loss": 0.382,
+      "step": 377
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9957014321050548,
+      "learning_rate": 3.9706767843447417e-05,
+      "loss": 0.3575,
+      "step": 378
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9866634592012732,
+      "learning_rate": 3.970433535883489e-05,
+      "loss": 0.3528,
+      "step": 379
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2674516282642205,
+      "learning_rate": 3.970189290173744e-05,
+      "loss": 0.4578,
+      "step": 380
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0813481346976104,
+      "learning_rate": 3.969944047339122e-05,
+      "loss": 0.3588,
+      "step": 381
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.035604185679093,
+      "learning_rate": 3.969697807503742e-05,
+      "loss": 0.3632,
+      "step": 382
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0593263688079588,
+      "learning_rate": 3.969450570792227e-05,
+      "loss": 0.3982,
+      "step": 383
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1305488858558645,
+      "learning_rate": 3.969202337329705e-05,
+      "loss": 0.4326,
+      "step": 384
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9800345078543079,
+      "learning_rate": 3.968953107241809e-05,
+      "loss": 0.3461,
+      "step": 385
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1507518570325785,
+      "learning_rate": 3.9687028806546756e-05,
+      "loss": 0.4266,
+      "step": 386
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.160973771402753,
+      "learning_rate": 3.968451657694946e-05,
+      "loss": 0.3961,
+      "step": 387
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0522505120629397,
+      "learning_rate": 3.9681994384897654e-05,
+      "loss": 0.3832,
+      "step": 388
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0262066600350739,
+      "learning_rate": 3.967946223166784e-05,
+      "loss": 0.3564,
+      "step": 389
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1021390447735178,
+      "learning_rate": 3.967692011854155e-05,
+      "loss": 0.3057,
+      "step": 390
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1459370595048486,
+      "learning_rate": 3.967436804680537e-05,
+      "loss": 0.406,
+      "step": 391
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1801973971437896,
+      "learning_rate": 3.9671806017750915e-05,
+      "loss": 0.4478,
+      "step": 392
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.142012235916646,
+      "learning_rate": 3.966923403267485e-05,
+      "loss": 0.3831,
+      "step": 393
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0224250852753491,
+      "learning_rate": 3.9666652092878856e-05,
+      "loss": 0.3376,
+      "step": 394
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0248021935123073,
+      "learning_rate": 3.966406019966968e-05,
+      "loss": 0.3892,
+      "step": 395
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0170305356568081,
+      "learning_rate": 3.9661458354359105e-05,
+      "loss": 0.3874,
+      "step": 396
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1045133957989504,
+      "learning_rate": 3.9658846558263925e-05,
+      "loss": 0.4759,
+      "step": 397
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0138047559369607,
+      "learning_rate": 3.965622481270599e-05,
+      "loss": 0.3802,
+      "step": 398
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9892519680502024,
+      "learning_rate": 3.9653593119012185e-05,
+      "loss": 0.3557,
+      "step": 399
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0388442075496735,
+      "learning_rate": 3.965095147851442e-05,
+      "loss": 0.4092,
+      "step": 400
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0368870498665264,
+      "learning_rate": 3.9648299892549654e-05,
+      "loss": 0.3768,
+      "step": 401
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0832108220495158,
+      "learning_rate": 3.964563836245987e-05,
+      "loss": 0.359,
+      "step": 402
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.03824313562385,
+      "learning_rate": 3.964296688959208e-05,
+      "loss": 0.4537,
+      "step": 403
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.054298611752923,
+      "learning_rate": 3.964028547529832e-05,
+      "loss": 0.3528,
+      "step": 404
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9577731934987448,
+      "learning_rate": 3.9637594120935697e-05,
+      "loss": 0.366,
+      "step": 405
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9435210790316217,
+      "learning_rate": 3.9634892827866306e-05,
+      "loss": 0.3064,
+      "step": 406
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9685154907886202,
+      "learning_rate": 3.9632181597457296e-05,
+      "loss": 0.3803,
+      "step": 407
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1840539421603884,
+      "learning_rate": 3.9629460431080825e-05,
+      "loss": 0.4352,
+      "step": 408
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0350869209537155,
+      "learning_rate": 3.96267293301141e-05,
+      "loss": 0.3756,
+      "step": 409
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9668692821784871,
+      "learning_rate": 3.962398829593935e-05,
+      "loss": 0.3315,
+      "step": 410
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0181116034631583,
+      "learning_rate": 3.962123732994383e-05,
+      "loss": 0.3729,
+      "step": 411
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0769434549688277,
+      "learning_rate": 3.961847643351981e-05,
+      "loss": 0.344,
+      "step": 412
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0925582933041047,
+      "learning_rate": 3.961570560806461e-05,
+      "loss": 0.4162,
+      "step": 413
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.102557118811125,
+      "learning_rate": 3.9612924854980556e-05,
+      "loss": 0.4166,
+      "step": 414
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1286347270584014,
+      "learning_rate": 3.9610134175675e-05,
+      "loss": 0.3693,
+      "step": 415
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9415423662860831,
+      "learning_rate": 3.960733357156033e-05,
+      "loss": 0.3004,
+      "step": 416
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0366728990271383,
+      "learning_rate": 3.960452304405394e-05,
+      "loss": 0.3846,
+      "step": 417
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9635313119417835,
+      "learning_rate": 3.960170259457826e-05,
+      "loss": 0.2909,
+      "step": 418
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9654734792005484,
+      "learning_rate": 3.959887222456075e-05,
+      "loss": 0.3031,
+      "step": 419
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.95233152536584,
+      "learning_rate": 3.959603193543385e-05,
+      "loss": 0.3166,
+      "step": 420
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0463975415402151,
+      "learning_rate": 3.959318172863506e-05,
+      "loss": 0.3488,
+      "step": 421
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1291429212168487,
+      "learning_rate": 3.95903216056069e-05,
+      "loss": 0.4022,
+      "step": 422
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0365884913215422,
+      "learning_rate": 3.958745156779688e-05,
+      "loss": 0.347,
+      "step": 423
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0666907074325134,
+      "learning_rate": 3.9584571616657544e-05,
+      "loss": 0.3873,
+      "step": 424
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9862395898344498,
+      "learning_rate": 3.958168175364646e-05,
+      "loss": 0.3259,
+      "step": 425
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8618518508668179,
+      "learning_rate": 3.957878198022621e-05,
+      "loss": 0.2429,
+      "step": 426
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9302577665948445,
+      "learning_rate": 3.957587229786437e-05,
+      "loss": 0.3339,
+      "step": 427
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9680432848986547,
+      "learning_rate": 3.9572952708033564e-05,
+      "loss": 0.3006,
+      "step": 428
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0056531379893139,
+      "learning_rate": 3.9570023212211405e-05,
+      "loss": 0.3533,
+      "step": 429
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.042314188290271,
+      "learning_rate": 3.956708381188054e-05,
+      "loss": 0.3704,
+      "step": 430
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9955369477688994,
+      "learning_rate": 3.95641345085286e-05,
+      "loss": 0.3224,
+      "step": 431
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1437350162712814,
+      "learning_rate": 3.956117530364826e-05,
+      "loss": 0.3429,
+      "step": 432
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1434482489093942,
+      "learning_rate": 3.955820619873719e-05,
+      "loss": 0.3979,
+      "step": 433
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0885754934405945,
+      "learning_rate": 3.955522719529807e-05,
+      "loss": 0.3778,
+      "step": 434
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.006883414746575,
+      "learning_rate": 3.9552238294838584e-05,
+      "loss": 0.3259,
+      "step": 435
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9424097245084614,
+      "learning_rate": 3.954923949887144e-05,
+      "loss": 0.2967,
+      "step": 436
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.879260419905861,
+      "learning_rate": 3.954623080891435e-05,
+      "loss": 0.2797,
+      "step": 437
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0029412763867633,
+      "learning_rate": 3.954321222649003e-05,
+      "loss": 0.3128,
+      "step": 438
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9197166521725636,
+      "learning_rate": 3.95401837531262e-05,
+      "loss": 0.281,
+      "step": 439
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1461508686625692,
+      "learning_rate": 3.953714539035558e-05,
+      "loss": 0.3273,
+      "step": 440
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0014668673416294,
+      "learning_rate": 3.9534097139715926e-05,
+      "loss": 0.3021,
+      "step": 441
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0224160491504546,
+      "learning_rate": 3.9531039002749955e-05,
+      "loss": 0.331,
+      "step": 442
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0680427709794582,
+      "learning_rate": 3.952797098100543e-05,
+      "loss": 0.3789,
+      "step": 443
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9893873455653348,
+      "learning_rate": 3.952489307603507e-05,
+      "loss": 0.309,
+      "step": 444
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0181761172525154,
+      "learning_rate": 3.9521805289396645e-05,
+      "loss": 0.3104,
+      "step": 445
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1361743895130338,
+      "learning_rate": 3.951870762265288e-05,
+      "loss": 0.3481,
+      "step": 446
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9401021448397183,
+      "learning_rate": 3.9515600077371545e-05,
+      "loss": 0.2785,
+      "step": 447
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0281058266194,
+      "learning_rate": 3.951248265512538e-05,
+      "loss": 0.3919,
+      "step": 448
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0209799810421567,
+      "learning_rate": 3.950935535749213e-05,
+      "loss": 0.3407,
+      "step": 449
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0680804960868895,
+      "learning_rate": 3.950621818605453e-05,
+      "loss": 0.3009,
+      "step": 450
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9809830673051324,
+      "learning_rate": 3.950307114240034e-05,
+      "loss": 0.2788,
+      "step": 451
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.091927999903504,
+      "learning_rate": 3.9499914228122286e-05,
+      "loss": 0.3584,
+      "step": 452
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0496342294388752,
+      "learning_rate": 3.9496747444818105e-05,
+      "loss": 0.3215,
+      "step": 453
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9472906994204717,
+      "learning_rate": 3.9493570794090524e-05,
+      "loss": 0.2863,
+      "step": 454
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.026604796399034,
+      "learning_rate": 3.9490384277547266e-05,
+      "loss": 0.3253,
+      "step": 455
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0059919251445888,
+      "learning_rate": 3.9487187896801054e-05,
+      "loss": 0.3646,
+      "step": 456
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1146466364962124,
+      "learning_rate": 3.9483981653469586e-05,
+      "loss": 0.3671,
+      "step": 457
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8839794649332503,
+      "learning_rate": 3.948076554917556e-05,
+      "loss": 0.2903,
+      "step": 458
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0456336903595724,
+      "learning_rate": 3.9477539585546676e-05,
+      "loss": 0.335,
+      "step": 459
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.052518634948309,
+      "learning_rate": 3.9474303764215606e-05,
+      "loss": 0.345,
+      "step": 460
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0089797719510034,
+      "learning_rate": 3.9471058086820024e-05,
+      "loss": 0.3258,
+      "step": 461
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9364948924689621,
+      "learning_rate": 3.9467802555002584e-05,
+      "loss": 0.2942,
+      "step": 462
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.027574984559381,
+      "learning_rate": 3.946453717041093e-05,
+      "loss": 0.3173,
+      "step": 463
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.019409640137024,
+      "learning_rate": 3.94612619346977e-05,
+      "loss": 0.3705,
+      "step": 464
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9900787652596897,
+      "learning_rate": 3.94579768495205e-05,
+      "loss": 0.338,
+      "step": 465
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9350070869462811,
+      "learning_rate": 3.9454681916541936e-05,
+      "loss": 0.2733,
+      "step": 466
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.005931903483432,
+      "learning_rate": 3.94513771374296e-05,
+      "loss": 0.2998,
+      "step": 467
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9176156521492399,
+      "learning_rate": 3.9448062513856056e-05,
+      "loss": 0.277,
+      "step": 468
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0603063040051155,
+      "learning_rate": 3.944473804749885e-05,
+      "loss": 0.3182,
+      "step": 469
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.878378035447139,
+      "learning_rate": 3.944140374004052e-05,
+      "loss": 0.2254,
+      "step": 470
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0312910360783618,
+      "learning_rate": 3.9438059593168586e-05,
+      "loss": 0.3281,
+      "step": 471
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9213199884441453,
+      "learning_rate": 3.943470560857553e-05,
+      "loss": 0.2669,
+      "step": 472
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0196847617381182,
+      "learning_rate": 3.943134178795883e-05,
+      "loss": 0.2907,
+      "step": 473
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0983891331319187,
+      "learning_rate": 3.942796813302094e-05,
+      "loss": 0.3776,
+      "step": 474
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9573627754065492,
+      "learning_rate": 3.942458464546928e-05,
+      "loss": 0.3094,
+      "step": 475
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.002048777872047,
+      "learning_rate": 3.942119132701625e-05,
+      "loss": 0.3293,
+      "step": 476
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9371921987692436,
+      "learning_rate": 3.9417788179379245e-05,
+      "loss": 0.3096,
+      "step": 477
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9228511103876207,
+      "learning_rate": 3.941437520428061e-05,
+      "loss": 0.2665,
+      "step": 478
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7805922600482893,
+      "learning_rate": 3.941095240344766e-05,
+      "loss": 0.2244,
+      "step": 479
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9002954594099436,
+      "learning_rate": 3.940751977861272e-05,
+      "loss": 0.341,
+      "step": 480
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8502514747667408,
+      "learning_rate": 3.9404077331513044e-05,
+      "loss": 0.3013,
+      "step": 481
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9184259777508003,
+      "learning_rate": 3.940062506389089e-05,
+      "loss": 0.2894,
+      "step": 482
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.973871506001312,
+      "learning_rate": 3.9397162977493455e-05,
+      "loss": 0.2658,
+      "step": 483
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9796821204711598,
+      "learning_rate": 3.939369107407293e-05,
+      "loss": 0.3231,
+      "step": 484
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9145901127233058,
+      "learning_rate": 3.939020935538647e-05,
+      "loss": 0.2403,
+      "step": 485
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9041537672139351,
+      "learning_rate": 3.938671782319619e-05,
+      "loss": 0.253,
+      "step": 486
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9090908183292864,
+      "learning_rate": 3.938321647926918e-05,
+      "loss": 0.3126,
+      "step": 487
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.939022195366846,
+      "learning_rate": 3.937970532537749e-05,
+      "loss": 0.2991,
+      "step": 488
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9406854553058501,
+      "learning_rate": 3.937618436329813e-05,
+      "loss": 0.2799,
+      "step": 489
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.028239160942009,
+      "learning_rate": 3.937265359481309e-05,
+      "loss": 0.2919,
+      "step": 490
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9503870238436994,
+      "learning_rate": 3.936911302170931e-05,
+      "loss": 0.2791,
+      "step": 491
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0751408174236268,
+      "learning_rate": 3.936556264577869e-05,
+      "loss": 0.331,
+      "step": 492
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8973227509717273,
+      "learning_rate": 3.9362002468818105e-05,
+      "loss": 0.2709,
+      "step": 493
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9372297533259436,
+      "learning_rate": 3.935843249262939e-05,
+      "loss": 0.2964,
+      "step": 494
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8989041982170147,
+      "learning_rate": 3.9354852719019306e-05,
+      "loss": 0.2787,
+      "step": 495
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8896321903424332,
+      "learning_rate": 3.935126314979962e-05,
+      "loss": 0.3095,
+      "step": 496
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8254472878283721,
+      "learning_rate": 3.934766378678704e-05,
+      "loss": 0.2711,
+      "step": 497
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9562348458080392,
+      "learning_rate": 3.93440546318032e-05,
+      "loss": 0.2691,
+      "step": 498
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9090852124946016,
+      "learning_rate": 3.934043568667473e-05,
+      "loss": 0.2988,
+      "step": 499
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.056004246219132,
+      "learning_rate": 3.933680695323321e-05,
+      "loss": 0.3244,
+      "step": 500
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.927229316161898,
+      "learning_rate": 3.9333168433315144e-05,
+      "loss": 0.2705,
+      "step": 501
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9840266504477935,
+      "learning_rate": 3.932952012876203e-05,
+      "loss": 0.2898,
+      "step": 502
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7816009496163775,
+      "learning_rate": 3.9325862041420275e-05,
+      "loss": 0.2449,
+      "step": 503
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8185233814439627,
+      "learning_rate": 3.9322194173141284e-05,
+      "loss": 0.2487,
+      "step": 504
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9029351150313263,
+      "learning_rate": 3.931851652578137e-05,
+      "loss": 0.2685,
+      "step": 505
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.827850296234817,
+      "learning_rate": 3.9314829101201814e-05,
+      "loss": 0.2831,
+      "step": 506
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7696270568383994,
+      "learning_rate": 3.9311131901268855e-05,
+      "loss": 0.2205,
+      "step": 507
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0038396909743488,
+      "learning_rate": 3.930742492785366e-05,
+      "loss": 0.3061,
+      "step": 508
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9429509639247784,
+      "learning_rate": 3.930370818283235e-05,
+      "loss": 0.2612,
+      "step": 509
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.938932373628078,
+      "learning_rate": 3.9299981668085997e-05,
+      "loss": 0.288,
+      "step": 510
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9599716170406715,
+      "learning_rate": 3.929624538550061e-05,
+      "loss": 0.3057,
+      "step": 511
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9938410999118946,
+      "learning_rate": 3.929249933696715e-05,
+      "loss": 0.3066,
+      "step": 512
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9474069706180214,
+      "learning_rate": 3.92887435243815e-05,
+      "loss": 0.2881,
+      "step": 513
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9390446335906655,
+      "learning_rate": 3.928497794964452e-05,
+      "loss": 0.3123,
+      "step": 514
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8349768564591767,
+      "learning_rate": 3.928120261466198e-05,
+      "loss": 0.2915,
+      "step": 515
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7805330038112448,
+      "learning_rate": 3.92774175213446e-05,
+      "loss": 0.2132,
+      "step": 516
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9165333015064896,
+      "learning_rate": 3.927362267160804e-05,
+      "loss": 0.2749,
+      "step": 517
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9711977889904639,
+      "learning_rate": 3.92698180673729e-05,
+      "loss": 0.2505,
+      "step": 518
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0065510862094755,
+      "learning_rate": 3.9266003710564706e-05,
+      "loss": 0.3477,
+      "step": 519
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.876275699363774,
+      "learning_rate": 3.9262179603113934e-05,
+      "loss": 0.2555,
+      "step": 520
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8644224193306435,
+      "learning_rate": 3.925834574695599e-05,
+      "loss": 0.2725,
+      "step": 521
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0584898049535252,
+      "learning_rate": 3.9254502144031204e-05,
+      "loss": 0.3504,
+      "step": 522
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.869473785667698,
+      "learning_rate": 3.925064879628485e-05,
+      "loss": 0.2564,
+      "step": 523
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9656636079150069,
+      "learning_rate": 3.924678570566714e-05,
+      "loss": 0.3155,
+      "step": 524
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8794120519948292,
+      "learning_rate": 3.9242912874133186e-05,
+      "loss": 0.2562,
+      "step": 525
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9395312239131293,
+      "learning_rate": 3.9239030303643074e-05,
+      "loss": 0.2911,
+      "step": 526
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9857349991843363,
+      "learning_rate": 3.9235137996161786e-05,
+      "loss": 0.307,
+      "step": 527
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8903683074995272,
+      "learning_rate": 3.9231235953659244e-05,
+      "loss": 0.2542,
+      "step": 528
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9085176641041314,
+      "learning_rate": 3.9227324178110295e-05,
+      "loss": 0.2314,
+      "step": 529
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7851733685045973,
+      "learning_rate": 3.922340267149472e-05,
+      "loss": 0.2062,
+      "step": 530
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0448656298749703,
+      "learning_rate": 3.9219471435797205e-05,
+      "loss": 0.3111,
+      "step": 531
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9052352485708491,
+      "learning_rate": 3.921553047300739e-05,
+      "loss": 0.2788,
+      "step": 532
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9514362508272822,
+      "learning_rate": 3.9211579785119804e-05,
+      "loss": 0.3458,
+      "step": 533
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9124283488112793,
+      "learning_rate": 3.9207619374133917e-05,
+      "loss": 0.2927,
+      "step": 534
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7971053071157986,
+      "learning_rate": 3.920364924205412e-05,
+      "loss": 0.2527,
+      "step": 535
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8210814368924492,
+      "learning_rate": 3.9199669390889725e-05,
+      "loss": 0.2371,
+      "step": 536
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8722992155557622,
+      "learning_rate": 3.919567982265495e-05,
+      "loss": 0.2327,
+      "step": 537
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0838935882765854,
+      "learning_rate": 3.9191680539368956e-05,
+      "loss": 0.2947,
+      "step": 538
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9371555967328974,
+      "learning_rate": 3.9187671543055785e-05,
+      "loss": 0.2702,
+      "step": 539
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.017192862829085,
+      "learning_rate": 3.918365283574443e-05,
+      "loss": 0.3159,
+      "step": 540
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9847778827974156,
+      "learning_rate": 3.9179624419468766e-05,
+      "loss": 0.3008,
+      "step": 541
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9191660114354011,
+      "learning_rate": 3.917558629626762e-05,
+      "loss": 0.2984,
+      "step": 542
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8005048321274897,
+      "learning_rate": 3.917153846818471e-05,
+      "loss": 0.258,
+      "step": 543
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8634314427972728,
+      "learning_rate": 3.916748093726864e-05,
+      "loss": 0.2936,
+      "step": 544
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8338975347471274,
+      "learning_rate": 3.9163413705572984e-05,
+      "loss": 0.2798,
+      "step": 545
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8007292939985748,
+      "learning_rate": 3.9159336775156165e-05,
+      "loss": 0.2621,
+      "step": 546
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7741079025905953,
+      "learning_rate": 3.9155250148081564e-05,
+      "loss": 0.2351,
+      "step": 547
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7709212940307969,
+      "learning_rate": 3.9151153826417436e-05,
+      "loss": 0.2426,
+      "step": 548
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.028936449677963,
+      "learning_rate": 3.914704781223696e-05,
+      "loss": 0.3413,
+      "step": 549
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8438898606413698,
+      "learning_rate": 3.9142932107618214e-05,
+      "loss": 0.28,
+      "step": 550
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8175362974968846,
+      "learning_rate": 3.913880671464418e-05,
+      "loss": 0.2164,
+      "step": 551
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9028316681419125,
+      "learning_rate": 3.9134671635402745e-05,
+      "loss": 0.2527,
+      "step": 552
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9539675506215372,
+      "learning_rate": 3.91305268719867e-05,
+      "loss": 0.3183,
+      "step": 553
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9654179150393969,
+      "learning_rate": 3.912637242649373e-05,
+      "loss": 0.2934,
+      "step": 554
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.997502725449283,
+      "learning_rate": 3.912220830102643e-05,
+      "loss": 0.3352,
+      "step": 555
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.903468665235631,
+      "learning_rate": 3.911803449769228e-05,
+      "loss": 0.2729,
+      "step": 556
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.866565928350931,
+      "learning_rate": 3.911385101860369e-05,
+      "loss": 0.2233,
+      "step": 557
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.940595000286064,
+      "learning_rate": 3.9109657865877924e-05,
+      "loss": 0.282,
+      "step": 558
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9123056270044905,
+      "learning_rate": 3.910545504163716e-05,
+      "loss": 0.2633,
+      "step": 559
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8137460104380139,
+      "learning_rate": 3.9101242548008496e-05,
+      "loss": 0.2456,
+      "step": 560
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9614642800546382,
+      "learning_rate": 3.9097020387123876e-05,
+      "loss": 0.2924,
+      "step": 561
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.816181425687044,
+      "learning_rate": 3.9092788561120174e-05,
+      "loss": 0.2334,
+      "step": 562
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8107074624429237,
+      "learning_rate": 3.9088547072139145e-05,
+      "loss": 0.2132,
+      "step": 563
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.000443360272963,
+      "learning_rate": 3.9084295922327414e-05,
+      "loss": 0.2928,
+      "step": 564
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8417879179369573,
+      "learning_rate": 3.908003511383654e-05,
+      "loss": 0.2363,
+      "step": 565
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8597390530824697,
+      "learning_rate": 3.907576464882294e-05,
+      "loss": 0.2598,
+      "step": 566
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8753681089747455,
+      "learning_rate": 3.90714845294479e-05,
+      "loss": 0.2594,
+      "step": 567
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8363784268898656,
+      "learning_rate": 3.9067194757877635e-05,
+      "loss": 0.2449,
+      "step": 568
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7827517000446115,
+      "learning_rate": 3.906289533628322e-05,
+      "loss": 0.2065,
+      "step": 569
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7850525441169758,
+      "learning_rate": 3.9058586266840614e-05,
+      "loss": 0.2414,
+      "step": 570
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7769308862816632,
+      "learning_rate": 3.905426755173068e-05,
+      "loss": 0.2141,
+      "step": 571
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8177360400387669,
+      "learning_rate": 3.904993919313912e-05,
+      "loss": 0.247,
+      "step": 572
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8041845064375056,
+      "learning_rate": 3.9045601193256564e-05,
+      "loss": 0.2307,
+      "step": 573
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9279095293640436,
+      "learning_rate": 3.9041253554278486e-05,
+      "loss": 0.2471,
+      "step": 574
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9726708652363452,
+      "learning_rate": 3.9036896278405264e-05,
+      "loss": 0.3427,
+      "step": 575
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9646956604271677,
+      "learning_rate": 3.9032529367842145e-05,
+      "loss": 0.2814,
+      "step": 576
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.893512462181103,
+      "learning_rate": 3.902815282479923e-05,
+      "loss": 0.2492,
+      "step": 577
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7250958066974518,
+      "learning_rate": 3.902376665149153e-05,
+      "loss": 0.2237,
+      "step": 578
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9281262022068445,
+      "learning_rate": 3.9019370850138915e-05,
+      "loss": 0.3319,
+      "step": 579
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9334790785845333,
+      "learning_rate": 3.9014965422966115e-05,
+      "loss": 0.2829,
+      "step": 580
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.718761526927567,
+      "learning_rate": 3.9010550372202756e-05,
+      "loss": 0.1868,
+      "step": 581
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.860903057648702,
+      "learning_rate": 3.900612570008331e-05,
+      "loss": 0.2548,
+      "step": 582
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8865750234134181,
+      "learning_rate": 3.900169140884715e-05,
+      "loss": 0.2613,
+      "step": 583
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8633144251460134,
+      "learning_rate": 3.899724750073848e-05,
+      "loss": 0.2258,
+      "step": 584
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7977090846094947,
+      "learning_rate": 3.899279397800639e-05,
+      "loss": 0.2384,
+      "step": 585
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8879218656965611,
+      "learning_rate": 3.8988330842904844e-05,
+      "loss": 0.2288,
+      "step": 586
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9116832056120259,
+      "learning_rate": 3.8983858097692656e-05,
+      "loss": 0.2735,
+      "step": 587
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9106254187002378,
+      "learning_rate": 3.8979375744633515e-05,
+      "loss": 0.3284,
+      "step": 588
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7889815185084196,
+      "learning_rate": 3.897488378599596e-05,
+      "loss": 0.2806,
+      "step": 589
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7409425835211813,
+      "learning_rate": 3.8970382224053414e-05,
+      "loss": 0.2252,
+      "step": 590
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8416200881445274,
+      "learning_rate": 3.8965871061084126e-05,
+      "loss": 0.2315,
+      "step": 591
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8657765637959218,
+      "learning_rate": 3.896135029937123e-05,
+      "loss": 0.2482,
+      "step": 592
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9129124767347182,
+      "learning_rate": 3.895681994120272e-05,
+      "loss": 0.2974,
+      "step": 593
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8568144364938592,
+      "learning_rate": 3.8952279988871425e-05,
+      "loss": 0.2579,
+      "step": 594
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8096790861857919,
+      "learning_rate": 3.894773044467505e-05,
+      "loss": 0.2494,
+      "step": 595
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9138980293654226,
+      "learning_rate": 3.8943171310916146e-05,
+      "loss": 0.2804,
+      "step": 596
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7697566907591287,
+      "learning_rate": 3.893860258990212e-05,
+      "loss": 0.2106,
+      "step": 597
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8346549918050894,
+      "learning_rate": 3.893402428394522e-05,
+      "loss": 0.2223,
+      "step": 598
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.923388192914631,
+      "learning_rate": 3.892943639536257e-05,
+      "loss": 0.2345,
+      "step": 599
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.88024735599371,
+      "learning_rate": 3.8924838926476114e-05,
+      "loss": 0.2564,
+      "step": 600
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7871692719872424,
+      "learning_rate": 3.892023187961268e-05,
+      "loss": 0.2286,
+      "step": 601
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9327809454662354,
+      "learning_rate": 3.891561525710389e-05,
+      "loss": 0.2585,
+      "step": 602
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7632589174906094,
+      "learning_rate": 3.891098906128628e-05,
+      "loss": 0.211,
+      "step": 603
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.020860879227997,
+      "learning_rate": 3.890635329450118e-05,
+      "loss": 0.2721,
+      "step": 604
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8419214497236249,
+      "learning_rate": 3.890170795909477e-05,
+      "loss": 0.2499,
+      "step": 605
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6970832009057898,
+      "learning_rate": 3.88970530574181e-05,
+      "loss": 0.1971,
+      "step": 606
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9467412696800499,
+      "learning_rate": 3.889238859182703e-05,
+      "loss": 0.2719,
+      "step": 607
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9102254685698031,
+      "learning_rate": 3.888771456468229e-05,
+      "loss": 0.2563,
+      "step": 608
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8363262947985433,
+      "learning_rate": 3.8883030978349416e-05,
+      "loss": 0.196,
+      "step": 609
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8584818620470888,
+      "learning_rate": 3.887833783519882e-05,
+      "loss": 0.2561,
+      "step": 610
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.839118786563085,
+      "learning_rate": 3.887363513760571e-05,
+      "loss": 0.275,
+      "step": 611
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8060011279995126,
+      "learning_rate": 3.8868922887950165e-05,
+      "loss": 0.2507,
+      "step": 612
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8452849730912027,
+      "learning_rate": 3.886420108861708e-05,
+      "loss": 0.2726,
+      "step": 613
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7714129837419096,
+      "learning_rate": 3.885946974199618e-05,
+      "loss": 0.2003,
+      "step": 614
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6995611577801152,
+      "learning_rate": 3.8854728850482034e-05,
+      "loss": 0.1937,
+      "step": 615
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9346788556596127,
+      "learning_rate": 3.884997841647404e-05,
+      "loss": 0.2539,
+      "step": 616
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9487142595839332,
+      "learning_rate": 3.8845218442376416e-05,
+      "loss": 0.2369,
+      "step": 617
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7760025415649283,
+      "learning_rate": 3.8840448930598216e-05,
+      "loss": 0.2482,
+      "step": 618
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7882415731253646,
+      "learning_rate": 3.8835669883553315e-05,
+      "loss": 0.199,
+      "step": 619
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8341878959807444,
+      "learning_rate": 3.883088130366042e-05,
+      "loss": 0.2105,
+      "step": 620
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8429857961152651,
+      "learning_rate": 3.882608319334306e-05,
+      "loss": 0.2527,
+      "step": 621
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8876945403277204,
+      "learning_rate": 3.88212755550296e-05,
+      "loss": 0.29,
+      "step": 622
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8439356988078114,
+      "learning_rate": 3.88164583911532e-05,
+      "loss": 0.2386,
+      "step": 623
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7962827164352877,
+      "learning_rate": 3.881163170415186e-05,
+      "loss": 0.1941,
+      "step": 624
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8120727149134773,
+      "learning_rate": 3.88067954964684e-05,
+      "loss": 0.2458,
+      "step": 625
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8753465238779636,
+      "learning_rate": 3.880194977055045e-05,
+      "loss": 0.2773,
+      "step": 626
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8029545020424,
+      "learning_rate": 3.8797094528850474e-05,
+      "loss": 0.2832,
+      "step": 627
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7780470996760221,
+      "learning_rate": 3.8792229773825716e-05,
+      "loss": 0.2225,
+      "step": 628
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7320854126262574,
+      "learning_rate": 3.878735550793827e-05,
+      "loss": 0.1794,
+      "step": 629
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7694790430319758,
+      "learning_rate": 3.8782471733655044e-05,
+      "loss": 0.1954,
+      "step": 630
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8376801766691149,
+      "learning_rate": 3.877757845344773e-05,
+      "loss": 0.2288,
+      "step": 631
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8453942983822659,
+      "learning_rate": 3.8772675669792855e-05,
+      "loss": 0.2335,
+      "step": 632
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8760557277124035,
+      "learning_rate": 3.876776338517174e-05,
+      "loss": 0.2096,
+      "step": 633
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0169787261523278,
+      "learning_rate": 3.876284160207053e-05,
+      "loss": 0.2896,
+      "step": 634
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9192946577722143,
+      "learning_rate": 3.875791032298017e-05,
+      "loss": 0.2784,
+      "step": 635
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9142417405933866,
+      "learning_rate": 3.875296955039641e-05,
+      "loss": 0.302,
+      "step": 636
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8178038976593689,
+      "learning_rate": 3.874801928681979e-05,
+      "loss": 0.1832,
+      "step": 637
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.760477864838044,
+      "learning_rate": 3.87430595347557e-05,
+      "loss": 0.2009,
+      "step": 638
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8154029704759584,
+      "learning_rate": 3.873809029671427e-05,
+      "loss": 0.2425,
+      "step": 639
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8286592451665006,
+      "learning_rate": 3.873311157521048e-05,
+      "loss": 0.2414,
+      "step": 640
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9469051028965093,
+      "learning_rate": 3.8728123372764085e-05,
+      "loss": 0.2503,
+      "step": 641
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.727655593449102,
+      "learning_rate": 3.8723125691899646e-05,
+      "loss": 0.2048,
+      "step": 642
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1400898762459823,
+      "learning_rate": 3.871811853514652e-05,
+      "loss": 0.1988,
+      "step": 643
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8699188520141753,
+      "learning_rate": 3.871310190503886e-05,
+      "loss": 0.2335,
+      "step": 644
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9371101840525764,
+      "learning_rate": 3.870807580411561e-05,
+      "loss": 0.2717,
+      "step": 645
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9328906871545495,
+      "learning_rate": 3.870304023492051e-05,
+      "loss": 0.2805,
+      "step": 646
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7801770662800196,
+      "learning_rate": 3.8697995200002105e-05,
+      "loss": 0.236,
+      "step": 647
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1148783153797293,
+      "learning_rate": 3.8692940701913706e-05,
+      "loss": 0.2292,
+      "step": 648
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7670811410430906,
+      "learning_rate": 3.868787674321343e-05,
+      "loss": 0.2143,
+      "step": 649
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8094460238957172,
+      "learning_rate": 3.868280332646417e-05,
+      "loss": 0.2613,
+      "step": 650
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7476612820351332,
+      "learning_rate": 3.867772045423362e-05,
+      "loss": 0.2234,
+      "step": 651
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.784818700094082,
+      "learning_rate": 3.8672628129094255e-05,
+      "loss": 0.2476,
+      "step": 652
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.793072773541811,
+      "learning_rate": 3.8667526353623326e-05,
+      "loss": 0.194,
+      "step": 653
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.892512980615161,
+      "learning_rate": 3.866241513040288e-05,
+      "loss": 0.2593,
+      "step": 654
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9285416794821778,
+      "learning_rate": 3.8657294462019735e-05,
+      "loss": 0.2425,
+      "step": 655
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8956017231761512,
+      "learning_rate": 3.865216435106549e-05,
+      "loss": 0.2275,
+      "step": 656
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7304122413360475,
+      "learning_rate": 3.8647024800136524e-05,
+      "loss": 0.1792,
+      "step": 657
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8574688318554834,
+      "learning_rate": 3.8641875811834004e-05,
+      "loss": 0.2151,
+      "step": 658
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7538143754231571,
+      "learning_rate": 3.863671738876385e-05,
+      "loss": 0.2138,
+      "step": 659
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8502167733846998,
+      "learning_rate": 3.863154953353679e-05,
+      "loss": 0.2274,
+      "step": 660
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8107420169367895,
+      "learning_rate": 3.8626372248768295e-05,
+      "loss": 0.2394,
+      "step": 661
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9014610828498489,
+      "learning_rate": 3.862118553707863e-05,
+      "loss": 0.3027,
+      "step": 662
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9112825423972205,
+      "learning_rate": 3.86159894010928e-05,
+      "loss": 0.2908,
+      "step": 663
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6628921068182834,
+      "learning_rate": 3.8610783843440626e-05,
+      "loss": 0.1389,
+      "step": 664
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7176790136393478,
+      "learning_rate": 3.8605568866756666e-05,
+      "loss": 0.1723,
+      "step": 665
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7682664032440877,
+      "learning_rate": 3.860034447368024e-05,
+      "loss": 0.2148,
+      "step": 666
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9302215965381995,
+      "learning_rate": 3.8595110666855466e-05,
+      "loss": 0.296,
+      "step": 667
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8521139476720233,
+      "learning_rate": 3.858986744893119e-05,
+      "loss": 0.23,
+      "step": 668
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7464405472134348,
+      "learning_rate": 3.858461482256103e-05,
+      "loss": 0.2147,
+      "step": 669
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8596453154852116,
+      "learning_rate": 3.8579352790403395e-05,
+      "loss": 0.2161,
+      "step": 670
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.809061992905901,
+      "learning_rate": 3.857408135512142e-05,
+      "loss": 0.2014,
+      "step": 671
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8437509889950078,
+      "learning_rate": 3.8568800519383e-05,
+      "loss": 0.2255,
+      "step": 672
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7568962614568128,
+      "learning_rate": 3.856351028586082e-05,
+      "loss": 0.1787,
+      "step": 673
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8055013226380432,
+      "learning_rate": 3.855821065723228e-05,
+      "loss": 0.2017,
+      "step": 674
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8826729993400942,
+      "learning_rate": 3.855290163617956e-05,
+      "loss": 0.2608,
+      "step": 675
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7806975127749172,
+      "learning_rate": 3.8547583225389596e-05,
+      "loss": 0.2001,
+      "step": 676
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7618422970469434,
+      "learning_rate": 3.8542255427554065e-05,
+      "loss": 0.2289,
+      "step": 677
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7862885694034616,
+      "learning_rate": 3.85369182453694e-05,
+      "loss": 0.2798,
+      "step": 678
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6671069424883396,
+      "learning_rate": 3.853157168153677e-05,
+      "loss": 0.1671,
+      "step": 679
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.739590608981885,
+      "learning_rate": 3.852621573876212e-05,
+      "loss": 0.2328,
+      "step": 680
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7928258081145453,
+      "learning_rate": 3.8520850419756104e-05,
+      "loss": 0.2329,
+      "step": 681
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8702197972343868,
+      "learning_rate": 3.851547572723416e-05,
+      "loss": 0.2651,
+      "step": 682
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8421114800562508,
+      "learning_rate": 3.851009166391646e-05,
+      "loss": 0.2441,
+      "step": 683
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8450856410310613,
+      "learning_rate": 3.850469823252789e-05,
+      "loss": 0.2525,
+      "step": 684
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7375467269448306,
+      "learning_rate": 3.849929543579812e-05,
+      "loss": 0.187,
+      "step": 685
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.807894049241486,
+      "learning_rate": 3.849388327646152e-05,
+      "loss": 0.2275,
+      "step": 686
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8135449219336123,
+      "learning_rate": 3.848846175725722e-05,
+      "loss": 0.2376,
+      "step": 687
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8856364968360584,
+      "learning_rate": 3.84830308809291e-05,
+      "loss": 0.2332,
+      "step": 688
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6883587675575743,
+      "learning_rate": 3.8477590650225735e-05,
+      "loss": 0.164,
+      "step": 689
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7784342568726977,
+      "learning_rate": 3.8472141067900485e-05,
+      "loss": 0.1987,
+      "step": 690
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7494893719948371,
+      "learning_rate": 3.84666821367114e-05,
+      "loss": 0.2167,
+      "step": 691
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.739221327572983,
+      "learning_rate": 3.846121385942128e-05,
+      "loss": 0.1996,
+      "step": 692
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8289543173872378,
+      "learning_rate": 3.845573623879766e-05,
+      "loss": 0.1907,
+      "step": 693
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8561945235943589,
+      "learning_rate": 3.845024927761279e-05,
+      "loss": 0.2405,
+      "step": 694
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7586867804796876,
+      "learning_rate": 3.844475297864366e-05,
+      "loss": 0.1756,
+      "step": 695
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.820059891816306,
+      "learning_rate": 3.843924734467199e-05,
+      "loss": 0.2047,
+      "step": 696
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8113901187168117,
+      "learning_rate": 3.843373237848419e-05,
+      "loss": 0.1931,
+      "step": 697
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8109688269530724,
+      "learning_rate": 3.842820808287144e-05,
+      "loss": 0.2202,
+      "step": 698
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8227791892675015,
+      "learning_rate": 3.842267446062962e-05,
+      "loss": 0.2283,
+      "step": 699
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7573331516312104,
+      "learning_rate": 3.841713151455931e-05,
+      "loss": 0.1784,
+      "step": 700
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7528941421925391,
+      "learning_rate": 3.8411579247465845e-05,
+      "loss": 0.2109,
+      "step": 701
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7843007603501683,
+      "learning_rate": 3.840601766215926e-05,
+      "loss": 0.21,
+      "step": 702
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8236345145931844,
+      "learning_rate": 3.840044676145431e-05,
+      "loss": 0.2033,
+      "step": 703
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.752640763991213,
+      "learning_rate": 3.839486654817045e-05,
+      "loss": 0.2086,
+      "step": 704
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8262639276251157,
+      "learning_rate": 3.838927702513187e-05,
+      "loss": 0.2025,
+      "step": 705
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7761474298353154,
+      "learning_rate": 3.838367819516746e-05,
+      "loss": 0.1692,
+      "step": 706
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8734125314600135,
+      "learning_rate": 3.837807006111082e-05,
+      "loss": 0.2258,
+      "step": 707
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7484943854071063,
+      "learning_rate": 3.837245262580027e-05,
+      "loss": 0.1844,
+      "step": 708
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8271628494538433,
+      "learning_rate": 3.836682589207882e-05,
+      "loss": 0.1746,
+      "step": 709
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8443352117983162,
+      "learning_rate": 3.836118986279419e-05,
+      "loss": 0.235,
+      "step": 710
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8112915295285071,
+      "learning_rate": 3.835554454079882e-05,
+      "loss": 0.2069,
+      "step": 711
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7067376891994592,
+      "learning_rate": 3.834988992894983e-05,
+      "loss": 0.1963,
+      "step": 712
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7762592744503582,
+      "learning_rate": 3.834422603010906e-05,
+      "loss": 0.2214,
+      "step": 713
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7241916326324417,
+      "learning_rate": 3.833855284714305e-05,
+      "loss": 0.2028,
+      "step": 714
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8083965411167431,
+      "learning_rate": 3.833287038292303e-05,
+      "loss": 0.2359,
+      "step": 715
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7589641747017328,
+      "learning_rate": 3.832717864032492e-05,
+      "loss": 0.1956,
+      "step": 716
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7667360613003436,
+      "learning_rate": 3.832147762222936e-05,
+      "loss": 0.1972,
+      "step": 717
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7712735817641144,
+      "learning_rate": 3.8315767331521655e-05,
+      "loss": 0.1818,
+      "step": 718
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9010674834916976,
+      "learning_rate": 3.831004777109183e-05,
+      "loss": 0.2536,
+      "step": 719
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8485400958054092,
+      "learning_rate": 3.8304318943834584e-05,
+      "loss": 0.2258,
+      "step": 720
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7302916613026862,
+      "learning_rate": 3.8298580852649316e-05,
+      "loss": 0.2157,
+      "step": 721
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7151270654118754,
+      "learning_rate": 3.82928335004401e-05,
+      "loss": 0.1962,
+      "step": 722
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7447894614304839,
+      "learning_rate": 3.828707689011572e-05,
+      "loss": 0.2116,
+      "step": 723
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7271553000291014,
+      "learning_rate": 3.828131102458962e-05,
+      "loss": 0.1966,
+      "step": 724
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7429093314076324,
+      "learning_rate": 3.827553590677996e-05,
+      "loss": 0.1906,
+      "step": 725
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.716800949542447,
+      "learning_rate": 3.8269751539609525e-05,
+      "loss": 0.1946,
+      "step": 726
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7666347229218324,
+      "learning_rate": 3.8263957926005855e-05,
+      "loss": 0.2524,
+      "step": 727
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8161614155859114,
+      "learning_rate": 3.825815506890111e-05,
+      "loss": 0.2168,
+      "step": 728
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6637236010874694,
+      "learning_rate": 3.825234297123216e-05,
+      "loss": 0.1729,
+      "step": 729
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8388439526076809,
+      "learning_rate": 3.824652163594056e-05,
+      "loss": 0.2296,
+      "step": 730
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8218573419682884,
+      "learning_rate": 3.8240691065972486e-05,
+      "loss": 0.2196,
+      "step": 731
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8293467452318114,
+      "learning_rate": 3.823485126427886e-05,
+      "loss": 0.2345,
+      "step": 732
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6922598919352861,
+      "learning_rate": 3.822900223381522e-05,
+      "loss": 0.1656,
+      "step": 733
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8068375981982576,
+      "learning_rate": 3.8223143977541806e-05,
+      "loss": 0.2273,
+      "step": 734
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7797871053978261,
+      "learning_rate": 3.821727649842352e-05,
+      "loss": 0.2594,
+      "step": 735
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7766407226832587,
+      "learning_rate": 3.821139979942992e-05,
+      "loss": 0.1986,
+      "step": 736
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8017956519170995,
+      "learning_rate": 3.820551388353525e-05,
+      "loss": 0.2132,
+      "step": 737
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7925431937110463,
+      "learning_rate": 3.819961875371839e-05,
+      "loss": 0.2119,
+      "step": 738
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7698235900481465,
+      "learning_rate": 3.819371441296292e-05,
+      "loss": 0.2025,
+      "step": 739
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8134358225280137,
+      "learning_rate": 3.8187800864257065e-05,
+      "loss": 0.2063,
+      "step": 740
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7621643328146783,
+      "learning_rate": 3.818187811059369e-05,
+      "loss": 0.1876,
+      "step": 741
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9041169813600348,
+      "learning_rate": 3.817594615497035e-05,
+      "loss": 0.2496,
+      "step": 742
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7800227635069699,
+      "learning_rate": 3.817000500038924e-05,
+      "loss": 0.2249,
+      "step": 743
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7763990863508718,
+      "learning_rate": 3.8164054649857206e-05,
+      "loss": 0.204,
+      "step": 744
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6792244923631663,
+      "learning_rate": 3.815809510638578e-05,
+      "loss": 0.1702,
+      "step": 745
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8416913667837724,
+      "learning_rate": 3.81521263729911e-05,
+      "loss": 0.2529,
+      "step": 746
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6874297279510242,
+      "learning_rate": 3.8146148452694e-05,
+      "loss": 0.182,
+      "step": 747
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7552469616361313,
+      "learning_rate": 3.8140161348519924e-05,
+      "loss": 0.238,
+      "step": 748
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7362886210805616,
+      "learning_rate": 3.813416506349899e-05,
+      "loss": 0.1778,
+      "step": 749
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6942190723817684,
+      "learning_rate": 3.8128159600665954e-05,
+      "loss": 0.184,
+      "step": 750
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7012048145090823,
+      "learning_rate": 3.812214496306022e-05,
+      "loss": 0.1744,
+      "step": 751
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7449986674789698,
+      "learning_rate": 3.8116121153725824e-05,
+      "loss": 0.1862,
+      "step": 752
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7682776915316732,
+      "learning_rate": 3.8110088175711456e-05,
+      "loss": 0.2081,
+      "step": 753
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7913974923405264,
+      "learning_rate": 3.810404603207045e-05,
+      "loss": 0.2172,
+      "step": 754
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9037616043612097,
+      "learning_rate": 3.809799472586077e-05,
+      "loss": 0.2525,
+      "step": 755
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9625775987134346,
+      "learning_rate": 3.809193426014501e-05,
+      "loss": 0.2522,
+      "step": 756
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7113503315961358,
+      "learning_rate": 3.808586463799042e-05,
+      "loss": 0.1654,
+      "step": 757
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.748331040872349,
+      "learning_rate": 3.807978586246887e-05,
+      "loss": 0.1693,
+      "step": 758
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6747668375753214,
+      "learning_rate": 3.8073697936656866e-05,
+      "loss": 0.1564,
+      "step": 759
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7857764152943993,
+      "learning_rate": 3.806760086363554e-05,
+      "loss": 0.2224,
+      "step": 760
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8058116800725232,
+      "learning_rate": 3.806149464649066e-05,
+      "loss": 0.2057,
+      "step": 761
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7572603869023338,
+      "learning_rate": 3.8055379288312625e-05,
+      "loss": 0.1831,
+      "step": 762
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8191501193026167,
+      "learning_rate": 3.8049254792196443e-05,
+      "loss": 0.1991,
+      "step": 763
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7519661245832501,
+      "learning_rate": 3.804312116124177e-05,
+      "loss": 0.2015,
+      "step": 764
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8708368444295916,
+      "learning_rate": 3.8036978398552876e-05,
+      "loss": 0.2625,
+      "step": 765
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6962016132327279,
+      "learning_rate": 3.803082650723864e-05,
+      "loss": 0.1752,
+      "step": 766
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7236730484580767,
+      "learning_rate": 3.802466549041258e-05,
+      "loss": 0.2141,
+      "step": 767
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7697581619865724,
+      "learning_rate": 3.8018495351192825e-05,
+      "loss": 0.2054,
+      "step": 768
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.769659506086983,
+      "learning_rate": 3.801231609270212e-05,
+      "loss": 0.203,
+      "step": 769
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6636546733887948,
+      "learning_rate": 3.800612771806781e-05,
+      "loss": 0.1446,
+      "step": 770
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6869472535775166,
+      "learning_rate": 3.79999302304219e-05,
+      "loss": 0.1652,
+      "step": 771
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7367351024507836,
+      "learning_rate": 3.799372363290095e-05,
+      "loss": 0.1737,
+      "step": 772
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7346502964947343,
+      "learning_rate": 3.798750792864617e-05,
+      "loss": 0.1551,
+      "step": 773
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8866153270215831,
+      "learning_rate": 3.798128312080336e-05,
+      "loss": 0.2897,
+      "step": 774
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7273975618740585,
+      "learning_rate": 3.7975049212522934e-05,
+      "loss": 0.1959,
+      "step": 775
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7774320772682493,
+      "learning_rate": 3.7968806206959915e-05,
+      "loss": 0.1848,
+      "step": 776
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8070108520057053,
+      "learning_rate": 3.7962554107273926e-05,
+      "loss": 0.2094,
+      "step": 777
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7002195516351798,
+      "learning_rate": 3.795629291662919e-05,
+      "loss": 0.2065,
+      "step": 778
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7795505349867997,
+      "learning_rate": 3.795002263819453e-05,
+      "loss": 0.1989,
+      "step": 779
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7985563678683656,
+      "learning_rate": 3.7943743275143384e-05,
+      "loss": 0.2017,
+      "step": 780
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8227484003895893,
+      "learning_rate": 3.793745483065377e-05,
+      "loss": 0.2073,
+      "step": 781
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7756298858710011,
+      "learning_rate": 3.7931157307908304e-05,
+      "loss": 0.1646,
+      "step": 782
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7802675554915577,
+      "learning_rate": 3.792485071009421e-05,
+      "loss": 0.2172,
+      "step": 783
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7938318675917536,
+      "learning_rate": 3.7918535040403284e-05,
+      "loss": 0.2116,
+      "step": 784
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.788083214737397,
+      "learning_rate": 3.791221030203193e-05,
+      "loss": 0.1677,
+      "step": 785
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8406960709035459,
+      "learning_rate": 3.790587649818115e-05,
+      "loss": 0.2083,
+      "step": 786
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8092112371843942,
+      "learning_rate": 3.78995336320565e-05,
+      "loss": 0.2049,
+      "step": 787
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7295052196559106,
+      "learning_rate": 3.789318170686816e-05,
+      "loss": 0.2061,
+      "step": 788
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8338696939204928,
+      "learning_rate": 3.788682072583087e-05,
+      "loss": 0.206,
+      "step": 789
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8162035530743967,
+      "learning_rate": 3.788045069216396e-05,
+      "loss": 0.2347,
+      "step": 790
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7600016092609635,
+      "learning_rate": 3.787407160909134e-05,
+      "loss": 0.2107,
+      "step": 791
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7609946305841209,
+      "learning_rate": 3.786768347984152e-05,
+      "loss": 0.1974,
+      "step": 792
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.735952873636344,
+      "learning_rate": 3.7861286307647555e-05,
+      "loss": 0.1909,
+      "step": 793
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6722596642043193,
+      "learning_rate": 3.78548800957471e-05,
+      "loss": 0.1568,
+      "step": 794
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6333915980625863,
+      "learning_rate": 3.7848464847382376e-05,
+      "loss": 0.1725,
+      "step": 795
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.796500772603936,
+      "learning_rate": 3.7842040565800184e-05,
+      "loss": 0.1909,
+      "step": 796
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7852543631975609,
+      "learning_rate": 3.783560725425188e-05,
+      "loss": 0.1969,
+      "step": 797
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7451596629183987,
+      "learning_rate": 3.782916491599341e-05,
+      "loss": 0.1965,
+      "step": 798
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6560391132655947,
+      "learning_rate": 3.782271355428529e-05,
+      "loss": 0.1756,
+      "step": 799
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.859883106182944,
+      "learning_rate": 3.781625317239258e-05,
+      "loss": 0.2243,
+      "step": 800
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7195191207343877,
+      "learning_rate": 3.780978377358493e-05,
+      "loss": 0.153,
+      "step": 801
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6798048904594878,
+      "learning_rate": 3.7803305361136534e-05,
+      "loss": 0.1773,
+      "step": 802
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7079466905190117,
+      "learning_rate": 3.7796817938326155e-05,
+      "loss": 0.1877,
+      "step": 803
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7535864591552314,
+      "learning_rate": 3.7790321508437124e-05,
+      "loss": 0.1981,
+      "step": 804
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7119576397655025,
+      "learning_rate": 3.778381607475732e-05,
+      "loss": 0.1625,
+      "step": 805
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8590587033895257,
+      "learning_rate": 3.777730164057919e-05,
+      "loss": 0.2357,
+      "step": 806
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6929512219539972,
+      "learning_rate": 3.777077820919972e-05,
+      "loss": 0.1484,
+      "step": 807
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7499874829201456,
+      "learning_rate": 3.776424578392045e-05,
+      "loss": 0.2245,
+      "step": 808
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7393534307485781,
+      "learning_rate": 3.775770436804751e-05,
+      "loss": 0.1982,
+      "step": 809
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7135289926633145,
+      "learning_rate": 3.775115396489153e-05,
+      "loss": 0.1488,
+      "step": 810
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7734600025332912,
+      "learning_rate": 3.77445945777677e-05,
+      "loss": 0.1963,
+      "step": 811
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.710812400458968,
+      "learning_rate": 3.773802620999579e-05,
+      "loss": 0.1591,
+      "step": 812
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7657793629017298,
+      "learning_rate": 3.773144886490007e-05,
+      "loss": 0.1688,
+      "step": 813
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.811511355240138,
+      "learning_rate": 3.7724862545809394e-05,
+      "loss": 0.1988,
+      "step": 814
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7990316490808798,
+      "learning_rate": 3.771826725605713e-05,
+      "loss": 0.192,
+      "step": 815
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6723617298825829,
+      "learning_rate": 3.771166299898118e-05,
+      "loss": 0.169,
+      "step": 816
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7521643403897306,
+      "learning_rate": 3.770504977792402e-05,
+      "loss": 0.1875,
+      "step": 817
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7114388930055676,
+      "learning_rate": 3.7698427596232636e-05,
+      "loss": 0.1927,
+      "step": 818
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7085657907035885,
+      "learning_rate": 3.7691796457258546e-05,
+      "loss": 0.1811,
+      "step": 819
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7836401046171007,
+      "learning_rate": 3.7685156364357825e-05,
+      "loss": 0.1845,
+      "step": 820
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7266652024397258,
+      "learning_rate": 3.767850732089105e-05,
+      "loss": 0.2094,
+      "step": 821
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8105800759742975,
+      "learning_rate": 3.7671849330223345e-05,
+      "loss": 0.199,
+      "step": 822
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7686728710156864,
+      "learning_rate": 3.766518239572437e-05,
+      "loss": 0.2106,
+      "step": 823
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7493906724832401,
+      "learning_rate": 3.76585065207683e-05,
+      "loss": 0.2183,
+      "step": 824
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6940159657288101,
+      "learning_rate": 3.765182170873383e-05,
+      "loss": 0.1491,
+      "step": 825
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6327228718145429,
+      "learning_rate": 3.7645127963004176e-05,
+      "loss": 0.129,
+      "step": 826
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7669510937458179,
+      "learning_rate": 3.76384252869671e-05,
+      "loss": 0.191,
+      "step": 827
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.757024453588316,
+      "learning_rate": 3.7631713684014866e-05,
+      "loss": 0.2127,
+      "step": 828
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7048436247771179,
+      "learning_rate": 3.7624993157544246e-05,
+      "loss": 0.1713,
+      "step": 829
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8488451798433339,
+      "learning_rate": 3.761826371095655e-05,
+      "loss": 0.2248,
+      "step": 830
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6127937468869562,
+      "learning_rate": 3.7611525347657584e-05,
+      "loss": 0.1252,
+      "step": 831
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6806426735887422,
+      "learning_rate": 3.7604778071057685e-05,
+      "loss": 0.1563,
+      "step": 832
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6743973690986225,
+      "learning_rate": 3.759802188457168e-05,
+      "loss": 0.1774,
+      "step": 833
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7603683630524931,
+      "learning_rate": 3.759125679161893e-05,
+      "loss": 0.236,
+      "step": 834
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6793108367801289,
+      "learning_rate": 3.758448279562327e-05,
+      "loss": 0.1412,
+      "step": 835
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6519421923379566,
+      "learning_rate": 3.757769990001308e-05,
+      "loss": 0.1167,
+      "step": 836
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7396227239586961,
+      "learning_rate": 3.757090810822122e-05,
+      "loss": 0.1992,
+      "step": 837
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7742992955873289,
+      "learning_rate": 3.756410742368505e-05,
+      "loss": 0.204,
+      "step": 838
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7887451364575835,
+      "learning_rate": 3.7557297849846444e-05,
+      "loss": 0.1708,
+      "step": 839
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7601953776719075,
+      "learning_rate": 3.7550479390151766e-05,
+      "loss": 0.169,
+      "step": 840
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6972835762854639,
+      "learning_rate": 3.754365204805189e-05,
+      "loss": 0.1534,
+      "step": 841
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7799931143187769,
+      "learning_rate": 3.753681582700216e-05,
+      "loss": 0.1763,
+      "step": 842
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7701780437869349,
+      "learning_rate": 3.752997073046244e-05,
+      "loss": 0.1606,
+      "step": 843
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8026366534002741,
+      "learning_rate": 3.752311676189708e-05,
+      "loss": 0.1872,
+      "step": 844
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8119168389561734,
+      "learning_rate": 3.75162539247749e-05,
+      "loss": 0.2032,
+      "step": 845
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6718226124056262,
+      "learning_rate": 3.750938222256924e-05,
+      "loss": 0.1569,
+      "step": 846
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7637510136724011,
+      "learning_rate": 3.75025016587579e-05,
+      "loss": 0.2004,
+      "step": 847
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6415349498989384,
+      "learning_rate": 3.7495612236823175e-05,
+      "loss": 0.1509,
+      "step": 848
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8286716528290092,
+      "learning_rate": 3.7488713960251845e-05,
+      "loss": 0.1616,
+      "step": 849
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6569549543464369,
+      "learning_rate": 3.748180683253518e-05,
+      "loss": 0.1562,
+      "step": 850
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7173550378920462,
+      "learning_rate": 3.747489085716891e-05,
+      "loss": 0.1558,
+      "step": 851
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7504036135246938,
+      "learning_rate": 3.746796603765325e-05,
+      "loss": 0.1769,
+      "step": 852
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6710972165250231,
+      "learning_rate": 3.7461032377492905e-05,
+      "loss": 0.1642,
+      "step": 853
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7352966450199,
+      "learning_rate": 3.745408988019703e-05,
+      "loss": 0.1819,
+      "step": 854
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7553549569396881,
+      "learning_rate": 3.744713854927928e-05,
+      "loss": 0.1826,
+      "step": 855
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8046090541490418,
+      "learning_rate": 3.7440178388257746e-05,
+      "loss": 0.2045,
+      "step": 856
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7820578784963078,
+      "learning_rate": 3.743320940065503e-05,
+      "loss": 0.1769,
+      "step": 857
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6971775715256504,
+      "learning_rate": 3.7426231589998166e-05,
+      "loss": 0.1782,
+      "step": 858
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.782002353801258,
+      "learning_rate": 3.741924495981867e-05,
+      "loss": 0.2065,
+      "step": 859
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6984916614807525,
+      "learning_rate": 3.741224951365251e-05,
+      "loss": 0.1677,
+      "step": 860
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6849913343173801,
+      "learning_rate": 3.740524525504014e-05,
+      "loss": 0.156,
+      "step": 861
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6846778338963099,
+      "learning_rate": 3.739823218752645e-05,
+      "loss": 0.1699,
+      "step": 862
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7003902445939452,
+      "learning_rate": 3.7391210314660796e-05,
+      "loss": 0.1621,
+      "step": 863
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6867990822182168,
+      "learning_rate": 3.7384179639996997e-05,
+      "loss": 0.1499,
+      "step": 864
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.644974619639601,
+      "learning_rate": 3.7377140167093316e-05,
+      "loss": 0.1218,
+      "step": 865
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.783606332695021,
+      "learning_rate": 3.7370091899512464e-05,
+      "loss": 0.1907,
+      "step": 866
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6615394006948083,
+      "learning_rate": 3.736303484082163e-05,
+      "loss": 0.1477,
+      "step": 867
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6937158662114102,
+      "learning_rate": 3.7355968994592414e-05,
+      "loss": 0.1625,
+      "step": 868
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7429604370422286,
+      "learning_rate": 3.7348894364400914e-05,
+      "loss": 0.1805,
+      "step": 869
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7475581950878672,
+      "learning_rate": 3.734181095382761e-05,
+      "loss": 0.204,
+      "step": 870
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6430529894982431,
+      "learning_rate": 3.733471876645749e-05,
+      "loss": 0.1358,
+      "step": 871
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7923755023892722,
+      "learning_rate": 3.732761780587993e-05,
+      "loss": 0.1778,
+      "step": 872
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7170539000161786,
+      "learning_rate": 3.732050807568878e-05,
+      "loss": 0.1676,
+      "step": 873
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6665991411184689,
+      "learning_rate": 3.7313389579482315e-05,
+      "loss": 0.1833,
+      "step": 874
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6874944513270407,
+      "learning_rate": 3.7306262320863245e-05,
+      "loss": 0.161,
+      "step": 875
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6105940436302919,
+      "learning_rate": 3.729912630343874e-05,
+      "loss": 0.1261,
+      "step": 876
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6619533928976752,
+      "learning_rate": 3.729198153082036e-05,
+      "loss": 0.1569,
+      "step": 877
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.569243487381569,
+      "learning_rate": 3.7284828006624125e-05,
+      "loss": 0.1229,
+      "step": 878
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7142199102812501,
+      "learning_rate": 3.7277665734470476e-05,
+      "loss": 0.1871,
+      "step": 879
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.629269678533343,
+      "learning_rate": 3.727049471798429e-05,
+      "loss": 0.1472,
+      "step": 880
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.654901322109771,
+      "learning_rate": 3.726331496079486e-05,
+      "loss": 0.149,
+      "step": 881
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6841025284739801,
+      "learning_rate": 3.7256126466535896e-05,
+      "loss": 0.1534,
+      "step": 882
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7865678235231162,
+      "learning_rate": 3.724892923884555e-05,
+      "loss": 0.2295,
+      "step": 883
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7841110484485813,
+      "learning_rate": 3.724172328136638e-05,
+      "loss": 0.2037,
+      "step": 884
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6599561399278121,
+      "learning_rate": 3.723450859774536e-05,
+      "loss": 0.1311,
+      "step": 885
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6987130451495835,
+      "learning_rate": 3.7227285191633894e-05,
+      "loss": 0.1495,
+      "step": 886
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7402887248919082,
+      "learning_rate": 3.722005306668778e-05,
+      "loss": 0.1579,
+      "step": 887
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6984442998567899,
+      "learning_rate": 3.721281222656725e-05,
+      "loss": 0.1619,
+      "step": 888
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7065217305585293,
+      "learning_rate": 3.7205562674936945e-05,
+      "loss": 0.1709,
+      "step": 889
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.732226984161801,
+      "learning_rate": 3.719830441546589e-05,
+      "loss": 0.1782,
+      "step": 890
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6858639758075553,
+      "learning_rate": 3.7191037451827545e-05,
+      "loss": 0.1719,
+      "step": 891
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7299863476978986,
+      "learning_rate": 3.718376178769976e-05,
+      "loss": 0.2216,
+      "step": 892
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8304332736342658,
+      "learning_rate": 3.71764774267648e-05,
+      "loss": 0.2096,
+      "step": 893
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7169787516591115,
+      "learning_rate": 3.716918437270932e-05,
+      "loss": 0.1684,
+      "step": 894
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7166803665478112,
+      "learning_rate": 3.7161882629224386e-05,
+      "loss": 0.1607,
+      "step": 895
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.724778723656452,
+      "learning_rate": 3.7154572200005446e-05,
+      "loss": 0.1759,
+      "step": 896
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7554750870258391,
+      "learning_rate": 3.714725308875236e-05,
+      "loss": 0.2013,
+      "step": 897
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7150835979381013,
+      "learning_rate": 3.713992529916936e-05,
+      "loss": 0.1539,
+      "step": 898
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6787440010634678,
+      "learning_rate": 3.7132588834965104e-05,
+      "loss": 0.1945,
+      "step": 899
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6542829825221722,
+      "learning_rate": 3.712524369985262e-05,
+      "loss": 0.1392,
+      "step": 900
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.665774640200682,
+      "learning_rate": 3.711788989754931e-05,
+      "loss": 0.1615,
+      "step": 901
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7030846690161123,
+      "learning_rate": 3.711052743177699e-05,
+      "loss": 0.1855,
+      "step": 902
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7936334689648732,
+      "learning_rate": 3.710315630626185e-05,
+      "loss": 0.2092,
+      "step": 903
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6753815367370155,
+      "learning_rate": 3.7095776524734464e-05,
+      "loss": 0.1527,
+      "step": 904
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7084634511777324,
+      "learning_rate": 3.7088388090929776e-05,
+      "loss": 0.1491,
+      "step": 905
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7112534208835603,
+      "learning_rate": 3.708099100858712e-05,
+      "loss": 0.1755,
+      "step": 906
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6211746451040592,
+      "learning_rate": 3.7073585281450206e-05,
+      "loss": 0.1484,
+      "step": 907
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6404982883466234,
+      "learning_rate": 3.706617091326712e-05,
+      "loss": 0.1445,
+      "step": 908
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7272965199740121,
+      "learning_rate": 3.705874790779032e-05,
+      "loss": 0.1616,
+      "step": 909
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7658933441744524,
+      "learning_rate": 3.705131626877664e-05,
+      "loss": 0.2129,
+      "step": 910
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6850082803142199,
+      "learning_rate": 3.7043875999987254e-05,
+      "loss": 0.1677,
+      "step": 911
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6362997911677422,
+      "learning_rate": 3.7036427105187754e-05,
+      "loss": 0.1608,
+      "step": 912
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6987606444597272,
+      "learning_rate": 3.7028969588148056e-05,
+      "loss": 0.1946,
+      "step": 913
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7026652210384798,
+      "learning_rate": 3.702150345264247e-05,
+      "loss": 0.1781,
+      "step": 914
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7209204369399616,
+      "learning_rate": 3.701402870244963e-05,
+      "loss": 0.1754,
+      "step": 915
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.727384123209215,
+      "learning_rate": 3.700654534135257e-05,
+      "loss": 0.1557,
+      "step": 916
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6332642391018083,
+      "learning_rate": 3.699905337313866e-05,
+      "loss": 0.1417,
+      "step": 917
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6780052852874722,
+      "learning_rate": 3.699155280159964e-05,
+      "loss": 0.1562,
+      "step": 918
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7154214327932445,
+      "learning_rate": 3.698404363053158e-05,
+      "loss": 0.1461,
+      "step": 919
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.700188076774634,
+      "learning_rate": 3.697652586373493e-05,
+      "loss": 0.1707,
+      "step": 920
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7230025827617601,
+      "learning_rate": 3.696899950501447e-05,
+      "loss": 0.1257,
+      "step": 921
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.778229060740858,
+      "learning_rate": 3.6961464558179333e-05,
+      "loss": 0.1885,
+      "step": 922
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7561932991777188,
+      "learning_rate": 3.695392102704302e-05,
+      "loss": 0.171,
+      "step": 923
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7260621922208205,
+      "learning_rate": 3.694636891542334e-05,
+      "loss": 0.1571,
+      "step": 924
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7317835353052108,
+      "learning_rate": 3.693880822714247e-05,
+      "loss": 0.1773,
+      "step": 925
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6661179092038415,
+      "learning_rate": 3.693123896602692e-05,
+      "loss": 0.1748,
+      "step": 926
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6743277867135127,
+      "learning_rate": 3.692366113590754e-05,
+      "loss": 0.1851,
+      "step": 927
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6938020824093218,
+      "learning_rate": 3.691607474061951e-05,
+      "loss": 0.164,
+      "step": 928
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6384162741338156,
+      "learning_rate": 3.690847978400236e-05,
+      "loss": 0.1469,
+      "step": 929
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6410681192219297,
+      "learning_rate": 3.690087626989994e-05,
+      "loss": 0.1861,
+      "step": 930
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5976785393840714,
+      "learning_rate": 3.689326420216044e-05,
+      "loss": 0.1303,
+      "step": 931
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6056843527575427,
+      "learning_rate": 3.6885643584636366e-05,
+      "loss": 0.1374,
+      "step": 932
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6691852000232525,
+      "learning_rate": 3.6878014421184565e-05,
+      "loss": 0.1767,
+      "step": 933
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6218312041996754,
+      "learning_rate": 3.68703767156662e-05,
+      "loss": 0.1642,
+      "step": 934
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6618795080274562,
+      "learning_rate": 3.6862730471946766e-05,
+      "loss": 0.1342,
+      "step": 935
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6863753483288044,
+      "learning_rate": 3.685507569389606e-05,
+      "loss": 0.1342,
+      "step": 936
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6551060924803995,
+      "learning_rate": 3.6847412385388236e-05,
+      "loss": 0.1366,
+      "step": 937
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7465156840052052,
+      "learning_rate": 3.683974055030172e-05,
+      "loss": 0.1627,
+      "step": 938
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6625105281209221,
+      "learning_rate": 3.6832060192519286e-05,
+      "loss": 0.1526,
+      "step": 939
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7086275742437467,
+      "learning_rate": 3.6824371315928e-05,
+      "loss": 0.1485,
+      "step": 940
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7417666798211808,
+      "learning_rate": 3.681667392441926e-05,
+      "loss": 0.1433,
+      "step": 941
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7386253806277999,
+      "learning_rate": 3.680896802188876e-05,
+      "loss": 0.1707,
+      "step": 942
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7037793611394684,
+      "learning_rate": 3.6801253612236506e-05,
+      "loss": 0.1838,
+      "step": 943
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6457543991417521,
+      "learning_rate": 3.679353069936681e-05,
+      "loss": 0.1514,
+      "step": 944
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7350562567817532,
+      "learning_rate": 3.678579928718827e-05,
+      "loss": 0.1374,
+      "step": 945
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6461300988449086,
+      "learning_rate": 3.6778059379613815e-05,
+      "loss": 0.1402,
+      "step": 946
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6394658184714149,
+      "learning_rate": 3.6770310980560654e-05,
+      "loss": 0.1477,
+      "step": 947
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7594333893105324,
+      "learning_rate": 3.676255409395031e-05,
+      "loss": 0.2031,
+      "step": 948
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7019386880161659,
+      "learning_rate": 3.675478872370858e-05,
+      "loss": 0.1665,
+      "step": 949
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6626999283463161,
+      "learning_rate": 3.674701487376557e-05,
+      "loss": 0.1439,
+      "step": 950
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6551620456940492,
+      "learning_rate": 3.673923254805566e-05,
+      "loss": 0.1425,
+      "step": 951
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.63563867244813,
+      "learning_rate": 3.6731441750517566e-05,
+      "loss": 0.1256,
+      "step": 952
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6957176828119622,
+      "learning_rate": 3.672364248509422e-05,
+      "loss": 0.147,
+      "step": 953
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7562490242565576,
+      "learning_rate": 3.67158347557329e-05,
+      "loss": 0.2017,
+      "step": 954
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6225715905123154,
+      "learning_rate": 3.670801856638514e-05,
+      "loss": 0.1394,
+      "step": 955
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.678394908191864,
+      "learning_rate": 3.6700193921006766e-05,
+      "loss": 0.1524,
+      "step": 956
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7398774587490798,
+      "learning_rate": 3.669236082355787e-05,
+      "loss": 0.1792,
+      "step": 957
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6206214596588789,
+      "learning_rate": 3.668451927800283e-05,
+      "loss": 0.1363,
+      "step": 958
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6027596192132293,
+      "learning_rate": 3.667666928831032e-05,
+      "loss": 0.1196,
+      "step": 959
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7351357598038659,
+      "learning_rate": 3.666881085845324e-05,
+      "loss": 0.2001,
+      "step": 960
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5988897173734646,
+      "learning_rate": 3.6660943992408817e-05,
+      "loss": 0.1227,
+      "step": 961
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6458349279158401,
+      "learning_rate": 3.66530686941585e-05,
+      "loss": 0.1292,
+      "step": 962
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7829938118934919,
+      "learning_rate": 3.664518496768802e-05,
+      "loss": 0.1853,
+      "step": 963
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6575365851561895,
+      "learning_rate": 3.663729281698741e-05,
+      "loss": 0.1553,
+      "step": 964
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6642437855720914,
+      "learning_rate": 3.662939224605091e-05,
+      "loss": 0.142,
+      "step": 965
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6479234151294552,
+      "learning_rate": 3.6621483258877055e-05,
+      "loss": 0.1318,
+      "step": 966
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7309571127932903,
+      "learning_rate": 3.6613565859468626e-05,
+      "loss": 0.1765,
+      "step": 967
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7002418747321221,
+      "learning_rate": 3.660564005183268e-05,
+      "loss": 0.179,
+      "step": 968
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7563400514440753,
+      "learning_rate": 3.659770583998051e-05,
+      "loss": 0.1672,
+      "step": 969
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.77837280709059,
+      "learning_rate": 3.658976322792766e-05,
+      "loss": 0.2449,
+      "step": 970
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6816422864134033,
+      "learning_rate": 3.658181221969395e-05,
+      "loss": 0.1447,
+      "step": 971
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6553011346867443,
+      "learning_rate": 3.657385281930343e-05,
+      "loss": 0.1439,
+      "step": 972
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5941559257267682,
+      "learning_rate": 3.65658850307844e-05,
+      "loss": 0.1584,
+      "step": 973
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6940833811429217,
+      "learning_rate": 3.65579088581694e-05,
+      "loss": 0.1608,
+      "step": 974
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6568072768488372,
+      "learning_rate": 3.6549924305495225e-05,
+      "loss": 0.1553,
+      "step": 975
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7325600048784592,
+      "learning_rate": 3.6541931376802906e-05,
+      "loss": 0.1493,
+      "step": 976
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5516350533636646,
+      "learning_rate": 3.653393007613771e-05,
+      "loss": 0.1195,
+      "step": 977
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7327823413100717,
+      "learning_rate": 3.652592040754917e-05,
+      "loss": 0.1737,
+      "step": 978
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6378932955160965,
+      "learning_rate": 3.651790237509098e-05,
+      "loss": 0.1621,
+      "step": 979
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6945962967680264,
+      "learning_rate": 3.650987598282116e-05,
+      "loss": 0.1377,
+      "step": 980
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6427290487016571,
+      "learning_rate": 3.6501841234801886e-05,
+      "loss": 0.1622,
+      "step": 981
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6658021914902404,
+      "learning_rate": 3.649379813509961e-05,
+      "loss": 0.1562,
+      "step": 982
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6810380080871552,
+      "learning_rate": 3.648574668778499e-05,
+      "loss": 0.1648,
+      "step": 983
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7029924267717719,
+      "learning_rate": 3.647768689693291e-05,
+      "loss": 0.1484,
+      "step": 984
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8445207819259016,
+      "learning_rate": 3.646961876662248e-05,
+      "loss": 0.2061,
+      "step": 985
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6621453829626172,
+      "learning_rate": 3.6461542300937035e-05,
+      "loss": 0.1546,
+      "step": 986
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7710667892427043,
+      "learning_rate": 3.645345750396412e-05,
+      "loss": 0.186,
+      "step": 987
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6722052228392159,
+      "learning_rate": 3.64453643797955e-05,
+      "loss": 0.1712,
+      "step": 988
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6294382345032234,
+      "learning_rate": 3.643726293252717e-05,
+      "loss": 0.1267,
+      "step": 989
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6357988604761875,
+      "learning_rate": 3.642915316625929e-05,
+      "loss": 0.14,
+      "step": 990
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7048080222855339,
+      "learning_rate": 3.642103508509629e-05,
+      "loss": 0.1741,
+      "step": 991
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5709092662827024,
+      "learning_rate": 3.641290869314676e-05,
+      "loss": 0.1459,
+      "step": 992
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7015535808582499,
+      "learning_rate": 3.640477399452354e-05,
+      "loss": 0.1733,
+      "step": 993
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6066420477109242,
+      "learning_rate": 3.639663099334363e-05,
+      "loss": 0.1353,
+      "step": 994
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5647230459909748,
+      "learning_rate": 3.6388479693728266e-05,
+      "loss": 0.1051,
+      "step": 995
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6302293400902045,
+      "learning_rate": 3.638032009980286e-05,
+      "loss": 0.1498,
+      "step": 996
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7179210692669409,
+      "learning_rate": 3.637215221569705e-05,
+      "loss": 0.1741,
+      "step": 997
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6376237580841241,
+      "learning_rate": 3.636397604554463e-05,
+      "loss": 0.1267,
+      "step": 998
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6459328410696206,
+      "learning_rate": 3.635579159348362e-05,
+      "loss": 0.1058,
+      "step": 999
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7772120071799035,
+      "learning_rate": 3.634759886365623e-05,
+      "loss": 0.1715,
+      "step": 1000
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6411320145391166,
+      "learning_rate": 3.633939786020884e-05,
+      "loss": 0.1518,
+      "step": 1001
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6501649573892895,
+      "learning_rate": 3.633118858729203e-05,
+      "loss": 0.1375,
+      "step": 1002
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6743897461008136,
+      "learning_rate": 3.632297104906057e-05,
+      "loss": 0.156,
+      "step": 1003
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6672403479914584,
+      "learning_rate": 3.63147452496734e-05,
+      "loss": 0.1124,
+      "step": 1004
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6429425165843409,
+      "learning_rate": 3.6306511193293636e-05,
+      "loss": 0.1337,
+      "step": 1005
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6776541626704364,
+      "learning_rate": 3.629826888408861e-05,
+      "loss": 0.1574,
+      "step": 1006
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7718525068581041,
+      "learning_rate": 3.629001832622979e-05,
+      "loss": 0.2403,
+      "step": 1007
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6612823182052278,
+      "learning_rate": 3.628175952389283e-05,
+      "loss": 0.1602,
+      "step": 1008
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6752689823461115,
+      "learning_rate": 3.627349248125757e-05,
+      "loss": 0.1424,
+      "step": 1009
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5796774048453173,
+      "learning_rate": 3.6265217202508006e-05,
+      "loss": 0.134,
+      "step": 1010
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6309566599944132,
+      "learning_rate": 3.625693369183231e-05,
+      "loss": 0.1365,
+      "step": 1011
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6002505226450737,
+      "learning_rate": 3.624864195342281e-05,
+      "loss": 0.144,
+      "step": 1012
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7674195414564663,
+      "learning_rate": 3.624034199147602e-05,
+      "loss": 0.1914,
+      "step": 1013
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7049269580203319,
+      "learning_rate": 3.623203381019259e-05,
+      "loss": 0.1618,
+      "step": 1014
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6315551794383809,
+      "learning_rate": 3.6223717413777346e-05,
+      "loss": 0.116,
+      "step": 1015
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6859663850789501,
+      "learning_rate": 3.621539280643926e-05,
+      "loss": 0.1655,
+      "step": 1016
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6495537170312613,
+      "learning_rate": 3.620705999239148e-05,
+      "loss": 0.1481,
+      "step": 1017
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6116131249215796,
+      "learning_rate": 3.619871897585129e-05,
+      "loss": 0.1341,
+      "step": 1018
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5413486818915718,
+      "learning_rate": 3.6190369761040116e-05,
+      "loss": 0.1103,
+      "step": 1019
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5985065221200078,
+      "learning_rate": 3.618201235218356e-05,
+      "loss": 0.1247,
+      "step": 1020
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6556592734863363,
+      "learning_rate": 3.617364675351136e-05,
+      "loss": 0.1522,
+      "step": 1021
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5977013764143297,
+      "learning_rate": 3.61652729692574e-05,
+      "loss": 0.1161,
+      "step": 1022
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7310760747668861,
+      "learning_rate": 3.615689100365968e-05,
+      "loss": 0.1906,
+      "step": 1023
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6738576689062206,
+      "learning_rate": 3.6148500860960386e-05,
+      "loss": 0.1552,
+      "step": 1024
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5956457892169247,
+      "learning_rate": 3.614010254540581e-05,
+      "loss": 0.1199,
+      "step": 1025
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6877198301331007,
+      "learning_rate": 3.6131696061246405e-05,
+      "loss": 0.1557,
+      "step": 1026
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6930864253545237,
+      "learning_rate": 3.612328141273673e-05,
+      "loss": 0.1761,
+      "step": 1027
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.711283547158347,
+      "learning_rate": 3.6114858604135496e-05,
+      "loss": 0.1623,
+      "step": 1028
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.729195653248427,
+      "learning_rate": 3.610642763970553e-05,
+      "loss": 0.1848,
+      "step": 1029
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.563524690420537,
+      "learning_rate": 3.6097988523713816e-05,
+      "loss": 0.1184,
+      "step": 1030
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6577990052567404,
+      "learning_rate": 3.608954126043141e-05,
+      "loss": 0.1756,
+      "step": 1031
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6586686025055664,
+      "learning_rate": 3.608108585413356e-05,
+      "loss": 0.15,
+      "step": 1032
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5790512809345955,
+      "learning_rate": 3.6072622309099566e-05,
+      "loss": 0.1078,
+      "step": 1033
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7169676117411847,
+      "learning_rate": 3.60641506296129e-05,
+      "loss": 0.1548,
+      "step": 1034
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6659877021730962,
+      "learning_rate": 3.605567081996113e-05,
+      "loss": 0.1395,
+      "step": 1035
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6933439870926703,
+      "learning_rate": 3.604718288443593e-05,
+      "loss": 0.162,
+      "step": 1036
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6841192351087376,
+      "learning_rate": 3.60386868273331e-05,
+      "loss": 0.1639,
+      "step": 1037
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7478730320930891,
+      "learning_rate": 3.603018265295255e-05,
+      "loss": 0.1938,
+      "step": 1038
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6873603592288243,
+      "learning_rate": 3.60216703655983e-05,
+      "loss": 0.1807,
+      "step": 1039
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6615960711149769,
+      "learning_rate": 3.601314996957845e-05,
+      "loss": 0.1431,
+      "step": 1040
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6790527141864818,
+      "learning_rate": 3.600462146920525e-05,
+      "loss": 0.1507,
+      "step": 1041
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6228395314339072,
+      "learning_rate": 3.5996084868795015e-05,
+      "loss": 0.1275,
+      "step": 1042
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6983925184063805,
+      "learning_rate": 3.5987540172668164e-05,
+      "loss": 0.1587,
+      "step": 1043
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6764855215492669,
+      "learning_rate": 3.597898738514923e-05,
+      "loss": 0.1606,
+      "step": 1044
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6701212772521431,
+      "learning_rate": 3.5970426510566824e-05,
+      "loss": 0.1587,
+      "step": 1045
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6212399628517273,
+      "learning_rate": 3.5961857553253665e-05,
+      "loss": 0.1464,
+      "step": 1046
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.622063222033276,
+      "learning_rate": 3.595328051754654e-05,
+      "loss": 0.1172,
+      "step": 1047
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6384653346584779,
+      "learning_rate": 3.594469540778637e-05,
+      "loss": 0.1121,
+      "step": 1048
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.704067827281616,
+      "learning_rate": 3.593610222831809e-05,
+      "loss": 0.1264,
+      "step": 1049
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.655922558247646,
+      "learning_rate": 3.59275009834908e-05,
+      "loss": 0.1078,
+      "step": 1050
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7779504143522625,
+      "learning_rate": 3.591889167765762e-05,
+      "loss": 0.1519,
+      "step": 1051
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7507066576498731,
+      "learning_rate": 3.591027431517577e-05,
+      "loss": 0.1231,
+      "step": 1052
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6901421494497663,
+      "learning_rate": 3.590164890040657e-05,
+      "loss": 0.1529,
+      "step": 1053
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7730705084134945,
+      "learning_rate": 3.589301543771537e-05,
+      "loss": 0.1819,
+      "step": 1054
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6758688500577673,
+      "learning_rate": 3.588437393147164e-05,
+      "loss": 0.1753,
+      "step": 1055
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6888931204638528,
+      "learning_rate": 3.587572438604889e-05,
+      "loss": 0.1456,
+      "step": 1056
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7073774385951197,
+      "learning_rate": 3.586706680582471e-05,
+      "loss": 0.1722,
+      "step": 1057
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6509485915727481,
+      "learning_rate": 3.585840119518075e-05,
+      "loss": 0.1305,
+      "step": 1058
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6538960381990047,
+      "learning_rate": 3.584972755850273e-05,
+      "loss": 0.1341,
+      "step": 1059
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6963019242777403,
+      "learning_rate": 3.584104590018044e-05,
+      "loss": 0.1436,
+      "step": 1060
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6382712644732104,
+      "learning_rate": 3.58323562246077e-05,
+      "loss": 0.1604,
+      "step": 1061
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6796882673237294,
+      "learning_rate": 3.5823658536182426e-05,
+      "loss": 0.1758,
+      "step": 1062
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6260051512475788,
+      "learning_rate": 3.5814952839306574e-05,
+      "loss": 0.1561,
+      "step": 1063
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6347693340486589,
+      "learning_rate": 3.580623913838613e-05,
+      "loss": 0.1605,
+      "step": 1064
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6308013133443539,
+      "learning_rate": 3.579751743783118e-05,
+      "loss": 0.1308,
+      "step": 1065
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5812309702967815,
+      "learning_rate": 3.578878774205581e-05,
+      "loss": 0.1118,
+      "step": 1066
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6310631390462856,
+      "learning_rate": 3.578005005547817e-05,
+      "loss": 0.1333,
+      "step": 1067
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6877280203948094,
+      "learning_rate": 3.577130438252046e-05,
+      "loss": 0.1531,
+      "step": 1068
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6483847585163999,
+      "learning_rate": 3.576255072760893e-05,
+      "loss": 0.1143,
+      "step": 1069
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6452634640367501,
+      "learning_rate": 3.575378909517385e-05,
+      "loss": 0.1313,
+      "step": 1070
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6900713690875435,
+      "learning_rate": 3.574501948964954e-05,
+      "loss": 0.1479,
+      "step": 1071
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7639374722338221,
+      "learning_rate": 3.5736241915474345e-05,
+      "loss": 0.1593,
+      "step": 1072
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6134283533254882,
+      "learning_rate": 3.572745637709065e-05,
+      "loss": 0.1205,
+      "step": 1073
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7151564450937763,
+      "learning_rate": 3.5718662878944876e-05,
+      "loss": 0.1524,
+      "step": 1074
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5888968945754794,
+      "learning_rate": 3.570986142548746e-05,
+      "loss": 0.1039,
+      "step": 1075
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6295134178319302,
+      "learning_rate": 3.5701052021172874e-05,
+      "loss": 0.1472,
+      "step": 1076
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6032800307183738,
+      "learning_rate": 3.5692234670459615e-05,
+      "loss": 0.165,
+      "step": 1077
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7610619476509329,
+      "learning_rate": 3.5683409377810185e-05,
+      "loss": 0.1877,
+      "step": 1078
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6134598830225814,
+      "learning_rate": 3.567457614769113e-05,
+      "loss": 0.1307,
+      "step": 1079
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6395293146696429,
+      "learning_rate": 3.566573498457301e-05,
+      "loss": 0.1499,
+      "step": 1080
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6567376459395161,
+      "learning_rate": 3.5656885892930376e-05,
+      "loss": 0.1432,
+      "step": 1081
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6329789825719756,
+      "learning_rate": 3.564802887724181e-05,
+      "loss": 0.1358,
+      "step": 1082
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7292218096636005,
+      "learning_rate": 3.563916394198991e-05,
+      "loss": 0.1604,
+      "step": 1083
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.683673200077468,
+      "learning_rate": 3.5630291091661276e-05,
+      "loss": 0.136,
+      "step": 1084
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5416118215035685,
+      "learning_rate": 3.562141033074649e-05,
+      "loss": 0.1176,
+      "step": 1085
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.61694461534053,
+      "learning_rate": 3.5612521663740183e-05,
+      "loss": 0.1251,
+      "step": 1086
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6421387164295951,
+      "learning_rate": 3.560362509514096e-05,
+      "loss": 0.1551,
+      "step": 1087
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6245773793912889,
+      "learning_rate": 3.5594720629451414e-05,
+      "loss": 0.1375,
+      "step": 1088
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5643563897157868,
+      "learning_rate": 3.558580827117817e-05,
+      "loss": 0.1104,
+      "step": 1089
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6756127207978516,
+      "learning_rate": 3.557688802483181e-05,
+      "loss": 0.1397,
+      "step": 1090
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6775988792391023,
+      "learning_rate": 3.556795989492694e-05,
+      "loss": 0.1342,
+      "step": 1091
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7394963916306223,
+      "learning_rate": 3.555902388598213e-05,
+      "loss": 0.1246,
+      "step": 1092
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6586100511807396,
+      "learning_rate": 3.555008000251995e-05,
+      "loss": 0.1442,
+      "step": 1093
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6084882691317554,
+      "learning_rate": 3.554112824906696e-05,
+      "loss": 0.1148,
+      "step": 1094
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6846720447127864,
+      "learning_rate": 3.55321686301537e-05,
+      "loss": 0.1451,
+      "step": 1095
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7111722537625023,
+      "learning_rate": 3.552320115031468e-05,
+      "loss": 0.1472,
+      "step": 1096
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6502074020788126,
+      "learning_rate": 3.55142258140884e-05,
+      "loss": 0.15,
+      "step": 1097
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6718778610168578,
+      "learning_rate": 3.5505242626017326e-05,
+      "loss": 0.1447,
+      "step": 1098
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6165806889739158,
+      "learning_rate": 3.549625159064792e-05,
+      "loss": 0.133,
+      "step": 1099
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6153270348904686,
+      "learning_rate": 3.5487252712530583e-05,
+      "loss": 0.1295,
+      "step": 1100
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6274481272059175,
+      "learning_rate": 3.547824599621971e-05,
+      "loss": 0.1329,
+      "step": 1101
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.661905555524102,
+      "learning_rate": 3.546923144627366e-05,
+      "loss": 0.1649,
+      "step": 1102
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5984050572617661,
+      "learning_rate": 3.546020906725474e-05,
+      "loss": 0.1405,
+      "step": 1103
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5674648783650594,
+      "learning_rate": 3.5451178863729244e-05,
+      "loss": 0.1001,
+      "step": 1104
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5261440618084035,
+      "learning_rate": 3.5442140840267404e-05,
+      "loss": 0.1103,
+      "step": 1105
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6960247977514382,
+      "learning_rate": 3.543309500144343e-05,
+      "loss": 0.159,
+      "step": 1106
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6214548177062327,
+      "learning_rate": 3.542404135183547e-05,
+      "loss": 0.1372,
+      "step": 1107
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6376840494046262,
+      "learning_rate": 3.541497989602562e-05,
+      "loss": 0.1349,
+      "step": 1108
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6090079449196457,
+      "learning_rate": 3.540591063859996e-05,
+      "loss": 0.124,
+      "step": 1109
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6539680223900682,
+      "learning_rate": 3.539683358414848e-05,
+      "loss": 0.1531,
+      "step": 1110
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7891131847457237,
+      "learning_rate": 3.538774873726514e-05,
+      "loss": 0.172,
+      "step": 1111
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6280573215367934,
+      "learning_rate": 3.537865610254784e-05,
+      "loss": 0.1262,
+      "step": 1112
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6193859444660762,
+      "learning_rate": 3.536955568459841e-05,
+      "loss": 0.1145,
+      "step": 1113
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5785477613220341,
+      "learning_rate": 3.536044748802263e-05,
+      "loss": 0.143,
+      "step": 1114
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5395700856893703,
+      "learning_rate": 3.535133151743022e-05,
+      "loss": 0.1128,
+      "step": 1115
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5701944934567039,
+      "learning_rate": 3.534220777743482e-05,
+      "loss": 0.1186,
+      "step": 1116
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6152525157373263,
+      "learning_rate": 3.5333076272654014e-05,
+      "loss": 0.128,
+      "step": 1117
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6256664061206213,
+      "learning_rate": 3.532393700770932e-05,
+      "loss": 0.1221,
+      "step": 1118
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6757214911173773,
+      "learning_rate": 3.5314789987226156e-05,
+      "loss": 0.1129,
+      "step": 1119
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6246653137541823,
+      "learning_rate": 3.5305635215833914e-05,
+      "loss": 0.1228,
+      "step": 1120
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6707085530900672,
+      "learning_rate": 3.5296472698165856e-05,
+      "loss": 0.1376,
+      "step": 1121
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6179075025066263,
+      "learning_rate": 3.5287302438859204e-05,
+      "loss": 0.1042,
+      "step": 1122
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6481782785733887,
+      "learning_rate": 3.5278124442555066e-05,
+      "loss": 0.1291,
+      "step": 1123
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6074076709129159,
+      "learning_rate": 3.526893871389849e-05,
+      "loss": 0.1111,
+      "step": 1124
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6630142240551968,
+      "learning_rate": 3.5259745257538443e-05,
+      "loss": 0.1322,
+      "step": 1125
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7361292481937083,
+      "learning_rate": 3.525054407812777e-05,
+      "loss": 0.1826,
+      "step": 1126
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5535894189934888,
+      "learning_rate": 3.524133518032325e-05,
+      "loss": 0.1087,
+      "step": 1127
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8351112988813633,
+      "learning_rate": 3.5232118568785565e-05,
+      "loss": 0.244,
+      "step": 1128
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6121295635932749,
+      "learning_rate": 3.52228942481793e-05,
+      "loss": 0.1496,
+      "step": 1129
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5354744737740527,
+      "learning_rate": 3.5213662223172935e-05,
+      "loss": 0.1045,
+      "step": 1130
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6452633485710266,
+      "learning_rate": 3.520442249843887e-05,
+      "loss": 0.1192,
+      "step": 1131
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6178008785901008,
+      "learning_rate": 3.5195175078653355e-05,
+      "loss": 0.1412,
+      "step": 1132
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5797864392513596,
+      "learning_rate": 3.51859199684966e-05,
+      "loss": 0.0976,
+      "step": 1133
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6622076271098327,
+      "learning_rate": 3.517665717265265e-05,
+      "loss": 0.1363,
+      "step": 1134
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6045944053491373,
+      "learning_rate": 3.516738669580947e-05,
+      "loss": 0.1131,
+      "step": 1135
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6398177294162027,
+      "learning_rate": 3.5158108542658915e-05,
+      "loss": 0.1404,
+      "step": 1136
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6773543157930411,
+      "learning_rate": 3.5148822717896694e-05,
+      "loss": 0.1461,
+      "step": 1137
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5875629005848713,
+      "learning_rate": 3.513952922622243e-05,
+      "loss": 0.1133,
+      "step": 1138
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6161178059033141,
+      "learning_rate": 3.513022807233964e-05,
+      "loss": 0.1102,
+      "step": 1139
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6611613235073424,
+      "learning_rate": 3.5120919260955655e-05,
+      "loss": 0.1372,
+      "step": 1140
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6502130397718782,
+      "learning_rate": 3.511160279678174e-05,
+      "loss": 0.1408,
+      "step": 1141
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.592408848966039,
+      "learning_rate": 3.510227868453302e-05,
+      "loss": 0.104,
+      "step": 1142
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6089071776647552,
+      "learning_rate": 3.509294692892847e-05,
+      "loss": 0.1135,
+      "step": 1143
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6034376470564629,
+      "learning_rate": 3.508360753469097e-05,
+      "loss": 0.1398,
+      "step": 1144
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6381865824576154,
+      "learning_rate": 3.5074260506547225e-05,
+      "loss": 0.1279,
+      "step": 1145
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6098214250407791,
+      "learning_rate": 3.506490584922784e-05,
+      "loss": 0.1167,
+      "step": 1146
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.699568101782631,
+      "learning_rate": 3.5055543567467244e-05,
+      "loss": 0.1656,
+      "step": 1147
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5408009953766895,
+      "learning_rate": 3.504617366600376e-05,
+      "loss": 0.115,
+      "step": 1148
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5839136648587737,
+      "learning_rate": 3.503679614957955e-05,
+      "loss": 0.1162,
+      "step": 1149
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7622599708603124,
+      "learning_rate": 3.502741102294063e-05,
+      "loss": 0.1708,
+      "step": 1150
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6580762882445695,
+      "learning_rate": 3.501801829083688e-05,
+      "loss": 0.1396,
+      "step": 1151
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5034837652030538,
+      "learning_rate": 3.500861795802201e-05,
+      "loss": 0.1035,
+      "step": 1152
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6301220343485374,
+      "learning_rate": 3.499921002925357e-05,
+      "loss": 0.134,
+      "step": 1153
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6759966027552546,
+      "learning_rate": 3.4989794509293005e-05,
+      "loss": 0.134,
+      "step": 1154
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5163242338440555,
+      "learning_rate": 3.498037140290555e-05,
+      "loss": 0.0933,
+      "step": 1155
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6237718913913396,
+      "learning_rate": 3.497094071486029e-05,
+      "loss": 0.1475,
+      "step": 1156
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5704668175929202,
+      "learning_rate": 3.4961502449930165e-05,
+      "loss": 0.1297,
+      "step": 1157
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6364792917633192,
+      "learning_rate": 3.495205661289193e-05,
+      "loss": 0.1467,
+      "step": 1158
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6446432305362467,
+      "learning_rate": 3.494260320852619e-05,
+      "loss": 0.1305,
+      "step": 1159
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6163755701743595,
+      "learning_rate": 3.493314224161737e-05,
+      "loss": 0.104,
+      "step": 1160
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5823381370149056,
+      "learning_rate": 3.4923673716953717e-05,
+      "loss": 0.1106,
+      "step": 1161
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5806140158264087,
+      "learning_rate": 3.4914197639327306e-05,
+      "loss": 0.1231,
+      "step": 1162
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5533175408118538,
+      "learning_rate": 3.490471401353405e-05,
+      "loss": 0.0922,
+      "step": 1163
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7623329229694302,
+      "learning_rate": 3.489522284437366e-05,
+      "loss": 0.1804,
+      "step": 1164
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6246931753430318,
+      "learning_rate": 3.488572413664969e-05,
+      "loss": 0.164,
+      "step": 1165
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5766952465211387,
+      "learning_rate": 3.4876217895169474e-05,
+      "loss": 0.1108,
+      "step": 1166
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5636533991904469,
+      "learning_rate": 3.4866704124744196e-05,
+      "loss": 0.1229,
+      "step": 1167
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5361245854759259,
+      "learning_rate": 3.4857182830188816e-05,
+      "loss": 0.0994,
+      "step": 1168
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6510711408936877,
+      "learning_rate": 3.484765401632214e-05,
+      "loss": 0.1251,
+      "step": 1169
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6100764028197836,
+      "learning_rate": 3.483811768796674e-05,
+      "loss": 0.1093,
+      "step": 1170
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6169278286668669,
+      "learning_rate": 3.482857384994903e-05,
+      "loss": 0.1295,
+      "step": 1171
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6201544082164446,
+      "learning_rate": 3.4819022507099184e-05,
+      "loss": 0.122,
+      "step": 1172
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5516199530485807,
+      "learning_rate": 3.480946366425121e-05,
+      "loss": 0.1092,
+      "step": 1173
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7000214215815579,
+      "learning_rate": 3.4799897326242895e-05,
+      "loss": 0.1739,
+      "step": 1174
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.632988563816966,
+      "learning_rate": 3.479032349791581e-05,
+      "loss": 0.1376,
+      "step": 1175
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6082218703833914,
+      "learning_rate": 3.478074218411534e-05,
+      "loss": 0.1378,
+      "step": 1176
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5413108598529882,
+      "learning_rate": 3.477115338969065e-05,
+      "loss": 0.0979,
+      "step": 1177
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5812498390033458,
+      "learning_rate": 3.476155711949467e-05,
+      "loss": 0.1012,
+      "step": 1178
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6965512588572845,
+      "learning_rate": 3.475195337838415e-05,
+      "loss": 0.1541,
+      "step": 1179
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5625723951164735,
+      "learning_rate": 3.474234217121959e-05,
+      "loss": 0.1207,
+      "step": 1180
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.596036773798448,
+      "learning_rate": 3.473272350286529e-05,
+      "loss": 0.1274,
+      "step": 1181
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7216373167904762,
+      "learning_rate": 3.4723097378189306e-05,
+      "loss": 0.1523,
+      "step": 1182
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6263755919104075,
+      "learning_rate": 3.471346380206349e-05,
+      "loss": 0.1208,
+      "step": 1183
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5691475139297871,
+      "learning_rate": 3.470382277936345e-05,
+      "loss": 0.113,
+      "step": 1184
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5948601182830424,
+      "learning_rate": 3.4694174314968564e-05,
+      "loss": 0.1156,
+      "step": 1185
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7207090796247909,
+      "learning_rate": 3.468451841376198e-05,
+      "loss": 0.1529,
+      "step": 1186
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.634722053793067,
+      "learning_rate": 3.467485508063061e-05,
+      "loss": 0.1158,
+      "step": 1187
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5649868774578078,
+      "learning_rate": 3.466518432046512e-05,
+      "loss": 0.1024,
+      "step": 1188
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6426538115570382,
+      "learning_rate": 3.4655506138159954e-05,
+      "loss": 0.1087,
+      "step": 1189
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.555267433780788,
+      "learning_rate": 3.464582053861329e-05,
+      "loss": 0.1083,
+      "step": 1190
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5448565153350091,
+      "learning_rate": 3.463612752672707e-05,
+      "loss": 0.133,
+      "step": 1191
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5585144360127174,
+      "learning_rate": 3.462642710740699e-05,
+      "loss": 0.1085,
+      "step": 1192
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6569064580859775,
+      "learning_rate": 3.461671928556248e-05,
+      "loss": 0.136,
+      "step": 1193
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6260130301816523,
+      "learning_rate": 3.4607004066106754e-05,
+      "loss": 0.1196,
+      "step": 1194
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5934602130643271,
+      "learning_rate": 3.459728145395671e-05,
+      "loss": 0.1346,
+      "step": 1195
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7028667611513228,
+      "learning_rate": 3.458755145403306e-05,
+      "loss": 0.1429,
+      "step": 1196
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.648635856899249,
+      "learning_rate": 3.457781407126018e-05,
+      "loss": 0.1357,
+      "step": 1197
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5471637780816928,
+      "learning_rate": 3.456806931056624e-05,
+      "loss": 0.1034,
+      "step": 1198
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.642561405045521,
+      "learning_rate": 3.4558317176883116e-05,
+      "loss": 0.1435,
+      "step": 1199
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6153785899846203,
+      "learning_rate": 3.454855767514643e-05,
+      "loss": 0.1289,
+      "step": 1200
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5295401151516663,
+      "learning_rate": 3.453879081029552e-05,
+      "loss": 0.0893,
+      "step": 1201
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6474491768483753,
+      "learning_rate": 3.452901658727345e-05,
+      "loss": 0.1256,
+      "step": 1202
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6736327841856329,
+      "learning_rate": 3.451923501102703e-05,
+      "loss": 0.1364,
+      "step": 1203
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5755152177033249,
+      "learning_rate": 3.450944608650677e-05,
+      "loss": 0.1296,
+      "step": 1204
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5771117222914954,
+      "learning_rate": 3.449964981866689e-05,
+      "loss": 0.1182,
+      "step": 1205
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6209119854673251,
+      "learning_rate": 3.4489846212465356e-05,
+      "loss": 0.1205,
+      "step": 1206
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7319852214009278,
+      "learning_rate": 3.448003527286383e-05,
+      "loss": 0.1527,
+      "step": 1207
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5456084707930992,
+      "learning_rate": 3.447021700482769e-05,
+      "loss": 0.106,
+      "step": 1208
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6069089189394817,
+      "learning_rate": 3.446039141332602e-05,
+      "loss": 0.0915,
+      "step": 1209
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6142817069593046,
+      "learning_rate": 3.4450558503331606e-05,
+      "loss": 0.1182,
+      "step": 1210
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7099143830243355,
+      "learning_rate": 3.444071827982096e-05,
+      "loss": 0.1767,
+      "step": 1211
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5684238934684104,
+      "learning_rate": 3.4430870747774266e-05,
+      "loss": 0.1236,
+      "step": 1212
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6007230524614405,
+      "learning_rate": 3.442101591217542e-05,
+      "loss": 0.1112,
+      "step": 1213
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6021835292176488,
+      "learning_rate": 3.441115377801202e-05,
+      "loss": 0.1253,
+      "step": 1214
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5635790541604533,
+      "learning_rate": 3.440128435027536e-05,
+      "loss": 0.1188,
+      "step": 1215
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.57701656687504,
+      "learning_rate": 3.43914076339604e-05,
+      "loss": 0.1268,
+      "step": 1216
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6599674519186035,
+      "learning_rate": 3.438152363406582e-05,
+      "loss": 0.1276,
+      "step": 1217
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6034804645653913,
+      "learning_rate": 3.437163235559396e-05,
+      "loss": 0.1168,
+      "step": 1218
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6539808050788924,
+      "learning_rate": 3.4361733803550874e-05,
+      "loss": 0.1581,
+      "step": 1219
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6012704313617491,
+      "learning_rate": 3.4351827982946274e-05,
+      "loss": 0.1041,
+      "step": 1220
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6296357135917684,
+      "learning_rate": 3.434191489879355e-05,
+      "loss": 0.1284,
+      "step": 1221
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6326536049743969,
+      "learning_rate": 3.433199455610978e-05,
+      "loss": 0.1221,
+      "step": 1222
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6212516919732799,
+      "learning_rate": 3.43220669599157e-05,
+      "loss": 0.1399,
+      "step": 1223
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5812341236694485,
+      "learning_rate": 3.431213211523574e-05,
+      "loss": 0.1036,
+      "step": 1224
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6630829711308373,
+      "learning_rate": 3.430219002709799e-05,
+      "loss": 0.123,
+      "step": 1225
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6440885931112753,
+      "learning_rate": 3.429224070053419e-05,
+      "loss": 0.136,
+      "step": 1226
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6477891829287996,
+      "learning_rate": 3.428228414057975e-05,
+      "loss": 0.1256,
+      "step": 1227
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.589110113997143,
+      "learning_rate": 3.427232035227377e-05,
+      "loss": 0.1169,
+      "step": 1228
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6328822438930977,
+      "learning_rate": 3.426234934065896e-05,
+      "loss": 0.1068,
+      "step": 1229
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5505090947435014,
+      "learning_rate": 3.4252371110781716e-05,
+      "loss": 0.142,
+      "step": 1230
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6575949374797614,
+      "learning_rate": 3.424238566769209e-05,
+      "loss": 0.1311,
+      "step": 1231
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5399678961164646,
+      "learning_rate": 3.423239301644377e-05,
+      "loss": 0.108,
+      "step": 1232
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.636209225537825,
+      "learning_rate": 3.42223931620941e-05,
+      "loss": 0.1432,
+      "step": 1233
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5430976659746752,
+      "learning_rate": 3.421238610970406e-05,
+      "loss": 0.0976,
+      "step": 1234
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.654773185785552,
+      "learning_rate": 3.4202371864338295e-05,
+      "loss": 0.1374,
+      "step": 1235
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6577328323362479,
+      "learning_rate": 3.419235043106506e-05,
+      "loss": 0.1299,
+      "step": 1236
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6343967950769432,
+      "learning_rate": 3.4182321814956274e-05,
+      "loss": 0.1374,
+      "step": 1237
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5689775703313454,
+      "learning_rate": 3.4172286021087475e-05,
+      "loss": 0.1115,
+      "step": 1238
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5321393422744918,
+      "learning_rate": 3.416224305453785e-05,
+      "loss": 0.0976,
+      "step": 1239
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5895353104496781,
+      "learning_rate": 3.4152192920390195e-05,
+      "loss": 0.115,
+      "step": 1240
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6693416117685603,
+      "learning_rate": 3.4142135623730954e-05,
+      "loss": 0.1299,
+      "step": 1241
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5223463271451472,
+      "learning_rate": 3.413207116965018e-05,
+      "loss": 0.0913,
+      "step": 1242
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5742671712584116,
+      "learning_rate": 3.412199956324155e-05,
+      "loss": 0.1103,
+      "step": 1243
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6471738567231234,
+      "learning_rate": 3.4111920809602374e-05,
+      "loss": 0.1202,
+      "step": 1244
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5738606536557671,
+      "learning_rate": 3.4101834913833576e-05,
+      "loss": 0.1253,
+      "step": 1245
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5329715099951039,
+      "learning_rate": 3.4091741881039677e-05,
+      "loss": 0.0931,
+      "step": 1246
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6004768334508961,
+      "learning_rate": 3.4081641716328826e-05,
+      "loss": 0.1212,
+      "step": 1247
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6238993728835703,
+      "learning_rate": 3.407153442481278e-05,
+      "loss": 0.1188,
+      "step": 1248
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6391662877126049,
+      "learning_rate": 3.4061420011606906e-05,
+      "loss": 0.1327,
+      "step": 1249
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6850578541931815,
+      "learning_rate": 3.405129848183017e-05,
+      "loss": 0.1346,
+      "step": 1250
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6620741753237898,
+      "learning_rate": 3.404116984060513e-05,
+      "loss": 0.1424,
+      "step": 1251
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5884064094161954,
+      "learning_rate": 3.403103409305796e-05,
+      "loss": 0.1271,
+      "step": 1252
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.592150818385704,
+      "learning_rate": 3.402089124431843e-05,
+      "loss": 0.1224,
+      "step": 1253
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5933580610751684,
+      "learning_rate": 3.4010741299519885e-05,
+      "loss": 0.0974,
+      "step": 1254
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5395084968672039,
+      "learning_rate": 3.400058426379929e-05,
+      "loss": 0.0956,
+      "step": 1255
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5568274468905576,
+      "learning_rate": 3.3990420142297165e-05,
+      "loss": 0.0934,
+      "step": 1256
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6464044573387644,
+      "learning_rate": 3.398024894015764e-05,
+      "loss": 0.1278,
+      "step": 1257
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6397880110723696,
+      "learning_rate": 3.3970070662528436e-05,
+      "loss": 0.1261,
+      "step": 1258
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.580808956213012,
+      "learning_rate": 3.395988531456083e-05,
+      "loss": 0.1214,
+      "step": 1259
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6264461475891724,
+      "learning_rate": 3.394969290140969e-05,
+      "loss": 0.1241,
+      "step": 1260
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6984251622896039,
+      "learning_rate": 3.393949342823346e-05,
+      "loss": 0.1567,
+      "step": 1261
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5048838159783609,
+      "learning_rate": 3.3929286900194154e-05,
+      "loss": 0.0969,
+      "step": 1262
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5949825243427104,
+      "learning_rate": 3.3919073322457364e-05,
+      "loss": 0.1053,
+      "step": 1263
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5917974541286745,
+      "learning_rate": 3.3908852700192236e-05,
+      "loss": 0.1162,
+      "step": 1264
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6482849401229682,
+      "learning_rate": 3.38986250385715e-05,
+      "loss": 0.1305,
+      "step": 1265
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5296304643919447,
+      "learning_rate": 3.388839034277142e-05,
+      "loss": 0.0985,
+      "step": 1266
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6002721447464341,
+      "learning_rate": 3.387814861797186e-05,
+      "loss": 0.1077,
+      "step": 1267
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5918012813995788,
+      "learning_rate": 3.386789986935621e-05,
+      "loss": 0.1221,
+      "step": 1268
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5493203124861112,
+      "learning_rate": 3.385764410211143e-05,
+      "loss": 0.1078,
+      "step": 1269
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5263530608990402,
+      "learning_rate": 3.3847381321428e-05,
+      "loss": 0.1154,
+      "step": 1270
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5414700579233093,
+      "learning_rate": 3.383711153250002e-05,
+      "loss": 0.1058,
+      "step": 1271
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6080252787344901,
+      "learning_rate": 3.382683474052506e-05,
+      "loss": 0.1299,
+      "step": 1272
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5519697744311083,
+      "learning_rate": 3.381655095070428e-05,
+      "loss": 0.0927,
+      "step": 1273
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5399789621528959,
+      "learning_rate": 3.3806260168242365e-05,
+      "loss": 0.0976,
+      "step": 1274
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6127765546933293,
+      "learning_rate": 3.379596239834755e-05,
+      "loss": 0.1375,
+      "step": 1275
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5594481223279179,
+      "learning_rate": 3.3785657646231596e-05,
+      "loss": 0.105,
+      "step": 1276
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5644887629151611,
+      "learning_rate": 3.37753459171098e-05,
+      "loss": 0.1016,
+      "step": 1277
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5622478820791422,
+      "learning_rate": 3.376502721620098e-05,
+      "loss": 0.118,
+      "step": 1278
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6754545844501156,
+      "learning_rate": 3.375470154872751e-05,
+      "loss": 0.13,
+      "step": 1279
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6377381197945493,
+      "learning_rate": 3.3744368919915275e-05,
+      "loss": 0.1602,
+      "step": 1280
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6207918116959827,
+      "learning_rate": 3.3734029334993675e-05,
+      "loss": 0.1371,
+      "step": 1281
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.538156355942291,
+      "learning_rate": 3.372368279919563e-05,
+      "loss": 0.1094,
+      "step": 1282
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7058682228466938,
+      "learning_rate": 3.3713329317757594e-05,
+      "loss": 0.163,
+      "step": 1283
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5479527616275841,
+      "learning_rate": 3.370296889591953e-05,
+      "loss": 0.093,
+      "step": 1284
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5901569015180179,
+      "learning_rate": 3.369260153892491e-05,
+      "loss": 0.1216,
+      "step": 1285
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5840498973615459,
+      "learning_rate": 3.3682227252020716e-05,
+      "loss": 0.1125,
+      "step": 1286
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.522568565049669,
+      "learning_rate": 3.367184604045743e-05,
+      "loss": 0.1061,
+      "step": 1287
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5738316483256367,
+      "learning_rate": 3.3661457909489056e-05,
+      "loss": 0.1105,
+      "step": 1288
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6034257939219156,
+      "learning_rate": 3.365106286437309e-05,
+      "loss": 0.1189,
+      "step": 1289
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5368924066832583,
+      "learning_rate": 3.364066091037052e-05,
+      "loss": 0.0988,
+      "step": 1290
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6711078744452189,
+      "learning_rate": 3.3630252052745844e-05,
+      "loss": 0.1559,
+      "step": 1291
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5399400230496249,
+      "learning_rate": 3.361983629676705e-05,
+      "loss": 0.1077,
+      "step": 1292
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6108261338608086,
+      "learning_rate": 3.360941364770562e-05,
+      "loss": 0.1294,
+      "step": 1293
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5468824113616102,
+      "learning_rate": 3.359898411083652e-05,
+      "loss": 0.1225,
+      "step": 1294
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6249837634838636,
+      "learning_rate": 3.358854769143819e-05,
+      "loss": 0.1312,
+      "step": 1295
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5940600639666558,
+      "learning_rate": 3.357810439479258e-05,
+      "loss": 0.1086,
+      "step": 1296
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5821095356360554,
+      "learning_rate": 3.356765422618509e-05,
+      "loss": 0.108,
+      "step": 1297
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6157257055801524,
+      "learning_rate": 3.355719719090465e-05,
+      "loss": 0.1548,
+      "step": 1298
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5744449922340548,
+      "learning_rate": 3.3546733294243585e-05,
+      "loss": 0.1333,
+      "step": 1299
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6436444510808119,
+      "learning_rate": 3.353626254149776e-05,
+      "loss": 0.1208,
+      "step": 1300
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5962607853391843,
+      "learning_rate": 3.3525784937966474e-05,
+      "loss": 0.1112,
+      "step": 1301
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5523535289195513,
+      "learning_rate": 3.3515300488952534e-05,
+      "loss": 0.0944,
+      "step": 1302
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.501809006719086,
+      "learning_rate": 3.350480919976216e-05,
+      "loss": 0.087,
+      "step": 1303
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5981814750628635,
+      "learning_rate": 3.349431107570506e-05,
+      "loss": 0.1275,
+      "step": 1304
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6180237524967851,
+      "learning_rate": 3.348380612209441e-05,
+      "loss": 0.1173,
+      "step": 1305
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6130160104393875,
+      "learning_rate": 3.347329434424683e-05,
+      "loss": 0.1229,
+      "step": 1306
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.544583943908318,
+      "learning_rate": 3.346277574748238e-05,
+      "loss": 0.1194,
+      "step": 1307
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5317881607186777,
+      "learning_rate": 3.345225033712459e-05,
+      "loss": 0.0983,
+      "step": 1308
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5799913180868896,
+      "learning_rate": 3.344171811850045e-05,
+      "loss": 0.1168,
+      "step": 1309
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5500312731261546,
+      "learning_rate": 3.3431179096940375e-05,
+      "loss": 0.1332,
+      "step": 1310
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6583755222113519,
+      "learning_rate": 3.3420633277778214e-05,
+      "loss": 0.1362,
+      "step": 1311
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5739367824277034,
+      "learning_rate": 3.341008066635129e-05,
+      "loss": 0.114,
+      "step": 1312
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6181973031072495,
+      "learning_rate": 3.339952126800033e-05,
+      "loss": 0.1252,
+      "step": 1313
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5774385397685252,
+      "learning_rate": 3.3388955088069524e-05,
+      "loss": 0.1402,
+      "step": 1314
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6174955926961799,
+      "learning_rate": 3.3378382131906465e-05,
+      "loss": 0.1398,
+      "step": 1315
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5129309703011087,
+      "learning_rate": 3.33678024048622e-05,
+      "loss": 0.0974,
+      "step": 1316
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6284531566427997,
+      "learning_rate": 3.335721591229119e-05,
+      "loss": 0.1331,
+      "step": 1317
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7647961002582618,
+      "learning_rate": 3.334662265955133e-05,
+      "loss": 0.1739,
+      "step": 1318
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6942168615205172,
+      "learning_rate": 3.3336022652003924e-05,
+      "loss": 0.1737,
+      "step": 1319
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5579231077747826,
+      "learning_rate": 3.33254158950137e-05,
+      "loss": 0.0998,
+      "step": 1320
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.500234608445358,
+      "learning_rate": 3.331480239394881e-05,
+      "loss": 0.1234,
+      "step": 1321
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5624056313086268,
+      "learning_rate": 3.330418215418081e-05,
+      "loss": 0.1177,
+      "step": 1322
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.544769976921429,
+      "learning_rate": 3.329355518108466e-05,
+      "loss": 0.086,
+      "step": 1323
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6480806356860086,
+      "learning_rate": 3.328292148003875e-05,
+      "loss": 0.1699,
+      "step": 1324
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5556130008667703,
+      "learning_rate": 3.3272281056424854e-05,
+      "loss": 0.117,
+      "step": 1325
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.546150790661389,
+      "learning_rate": 3.326163391562814e-05,
+      "loss": 0.1028,
+      "step": 1326
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5950563265080882,
+      "learning_rate": 3.325098006303722e-05,
+      "loss": 0.1133,
+      "step": 1327
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.606291157023492,
+      "learning_rate": 3.324031950404406e-05,
+      "loss": 0.0992,
+      "step": 1328
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6167163257536733,
+      "learning_rate": 3.322965224404403e-05,
+      "loss": 0.1478,
+      "step": 1329
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5856656270301086,
+      "learning_rate": 3.3218978288435896e-05,
+      "loss": 0.1117,
+      "step": 1330
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5518055042821899,
+      "learning_rate": 3.3208297642621824e-05,
+      "loss": 0.0941,
+      "step": 1331
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.553741508057109,
+      "learning_rate": 3.319761031200735e-05,
+      "loss": 0.1246,
+      "step": 1332
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5034305485838049,
+      "learning_rate": 3.318691630200138e-05,
+      "loss": 0.0952,
+      "step": 1333
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.49705691216473424,
+      "learning_rate": 3.317621561801624e-05,
+      "loss": 0.0896,
+      "step": 1334
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5258869535349345,
+      "learning_rate": 3.316550826546761e-05,
+      "loss": 0.0988,
+      "step": 1335
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5489673641506772,
+      "learning_rate": 3.315479424977453e-05,
+      "loss": 0.1149,
+      "step": 1336
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5114559925851911,
+      "learning_rate": 3.3144073576359455e-05,
+      "loss": 0.1157,
+      "step": 1337
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5566812333135119,
+      "learning_rate": 3.313334625064816e-05,
+      "loss": 0.1118,
+      "step": 1338
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5873765440338292,
+      "learning_rate": 3.312261227806982e-05,
+      "loss": 0.1208,
+      "step": 1339
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5055540186849343,
+      "learning_rate": 3.311187166405696e-05,
+      "loss": 0.086,
+      "step": 1340
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5983374210667861,
+      "learning_rate": 3.310112441404548e-05,
+      "loss": 0.1095,
+      "step": 1341
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5834239707649026,
+      "learning_rate": 3.309037053347462e-05,
+      "loss": 0.1269,
+      "step": 1342
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5894128349595298,
+      "learning_rate": 3.3079610027786985e-05,
+      "loss": 0.1095,
+      "step": 1343
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5633359895686817,
+      "learning_rate": 3.306884290242854e-05,
+      "loss": 0.1089,
+      "step": 1344
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5123450254265632,
+      "learning_rate": 3.3058069162848586e-05,
+      "loss": 0.1056,
+      "step": 1345
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6086173605103516,
+      "learning_rate": 3.3047288814499786e-05,
+      "loss": 0.0937,
+      "step": 1346
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6991284701068382,
+      "learning_rate": 3.3036501862838125e-05,
+      "loss": 0.1692,
+      "step": 1347
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5441333712760471,
+      "learning_rate": 3.302570831332297e-05,
+      "loss": 0.108,
+      "step": 1348
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5533949069537027,
+      "learning_rate": 3.301490817141698e-05,
+      "loss": 0.1199,
+      "step": 1349
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5405876312130093,
+      "learning_rate": 3.300410144258619e-05,
+      "loss": 0.1341,
+      "step": 1350
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4227404547560645,
+      "learning_rate": 3.2993288132299935e-05,
+      "loss": 0.0827,
+      "step": 1351
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.627268037704672,
+      "learning_rate": 3.298246824603091e-05,
+      "loss": 0.1427,
+      "step": 1352
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5433477821457182,
+      "learning_rate": 3.297164178925512e-05,
+      "loss": 0.1175,
+      "step": 1353
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.525409683959608,
+      "learning_rate": 3.2960808767451905e-05,
+      "loss": 0.1077,
+      "step": 1354
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4823978124632984,
+      "learning_rate": 3.294996918610393e-05,
+      "loss": 0.0896,
+      "step": 1355
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5202600057428901,
+      "learning_rate": 3.293912305069715e-05,
+      "loss": 0.0906,
+      "step": 1356
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6476774932359233,
+      "learning_rate": 3.292827036672089e-05,
+      "loss": 0.1117,
+      "step": 1357
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5763221674018213,
+      "learning_rate": 3.291741113966773e-05,
+      "loss": 0.1023,
+      "step": 1358
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6494560056062909,
+      "learning_rate": 3.290654537503362e-05,
+      "loss": 0.1113,
+      "step": 1359
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7034436033874668,
+      "learning_rate": 3.2895673078317775e-05,
+      "loss": 0.1229,
+      "step": 1360
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6143205425695772,
+      "learning_rate": 3.288479425502273e-05,
+      "loss": 0.1194,
+      "step": 1361
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6209591173806414,
+      "learning_rate": 3.287390891065433e-05,
+      "loss": 0.1485,
+      "step": 1362
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5999097786066504,
+      "learning_rate": 3.2863017050721715e-05,
+      "loss": 0.1246,
+      "step": 1363
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.48418642803392065,
+      "learning_rate": 3.2852118680737306e-05,
+      "loss": 0.0786,
+      "step": 1364
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.540208482983803,
+      "learning_rate": 3.2841213806216864e-05,
+      "loss": 0.1044,
+      "step": 1365
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5094061859507117,
+      "learning_rate": 3.283030243267939e-05,
+      "loss": 0.0803,
+      "step": 1366
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5646400010314275,
+      "learning_rate": 3.281938456564721e-05,
+      "loss": 0.1106,
+      "step": 1367
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5686819203471248,
+      "learning_rate": 3.2808460210645906e-05,
+      "loss": 0.1155,
+      "step": 1368
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6112084804857806,
+      "learning_rate": 3.2797529373204375e-05,
+      "loss": 0.1225,
+      "step": 1369
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5162515878999319,
+      "learning_rate": 3.278659205885479e-05,
+      "loss": 0.0996,
+      "step": 1370
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6340510175256542,
+      "learning_rate": 3.2775648273132574e-05,
+      "loss": 0.1383,
+      "step": 1371
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5671948435074174,
+      "learning_rate": 3.2764698021576446e-05,
+      "loss": 0.1048,
+      "step": 1372
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6487673644312671,
+      "learning_rate": 3.27537413097284e-05,
+      "loss": 0.1625,
+      "step": 1373
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5601474787877964,
+      "learning_rate": 3.27427781431337e-05,
+      "loss": 0.1142,
+      "step": 1374
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5813905402472606,
+      "learning_rate": 3.273180852734087e-05,
+      "loss": 0.1012,
+      "step": 1375
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4734692792644643,
+      "learning_rate": 3.27208324679017e-05,
+      "loss": 0.0724,
+      "step": 1376
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5780253453324077,
+      "learning_rate": 3.270984997037123e-05,
+      "loss": 0.1006,
+      "step": 1377
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5225369977400518,
+      "learning_rate": 3.269886104030778e-05,
+      "loss": 0.1089,
+      "step": 1378
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5194797060071085,
+      "learning_rate": 3.268786568327291e-05,
+      "loss": 0.1096,
+      "step": 1379
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5409759066217806,
+      "learning_rate": 3.2676863904831444e-05,
+      "loss": 0.1027,
+      "step": 1380
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5970778315366738,
+      "learning_rate": 3.266585571055145e-05,
+      "loss": 0.1028,
+      "step": 1381
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6425400928500881,
+      "learning_rate": 3.2654841106004225e-05,
+      "loss": 0.1467,
+      "step": 1382
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5410733519236234,
+      "learning_rate": 3.264382009676435e-05,
+      "loss": 0.1025,
+      "step": 1383
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5964767435581186,
+      "learning_rate": 3.263279268840961e-05,
+      "loss": 0.1334,
+      "step": 1384
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5198936345823255,
+      "learning_rate": 3.262175888652106e-05,
+      "loss": 0.0953,
+      "step": 1385
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.656659815755759,
+      "learning_rate": 3.261071869668296e-05,
+      "loss": 0.1355,
+      "step": 1386
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.588144393341704,
+      "learning_rate": 3.259967212448282e-05,
+      "loss": 0.1398,
+      "step": 1387
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5605675563918083,
+      "learning_rate": 3.2588619175511387e-05,
+      "loss": 0.1048,
+      "step": 1388
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.452692499602161,
+      "learning_rate": 3.2577559855362614e-05,
+      "loss": 0.0811,
+      "step": 1389
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5681878057077676,
+      "learning_rate": 3.2566494169633693e-05,
+      "loss": 0.1066,
+      "step": 1390
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5795194716502409,
+      "learning_rate": 3.255542212392505e-05,
+      "loss": 0.1226,
+      "step": 1391
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.565216129538394,
+      "learning_rate": 3.2544343723840296e-05,
+      "loss": 0.0917,
+      "step": 1392
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4864753285882939,
+      "learning_rate": 3.253325897498629e-05,
+      "loss": 0.0918,
+      "step": 1393
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5830497790492641,
+      "learning_rate": 3.2522167882973085e-05,
+      "loss": 0.0995,
+      "step": 1394
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5171751360841577,
+      "learning_rate": 3.251107045341395e-05,
+      "loss": 0.0975,
+      "step": 1395
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6417617249571287,
+      "learning_rate": 3.249996669192537e-05,
+      "loss": 0.1543,
+      "step": 1396
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5739693132332222,
+      "learning_rate": 3.248885660412701e-05,
+      "loss": 0.115,
+      "step": 1397
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5815615711071502,
+      "learning_rate": 3.247774019564178e-05,
+      "loss": 0.0829,
+      "step": 1398
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6249244644296427,
+      "learning_rate": 3.2466617472095736e-05,
+      "loss": 0.0991,
+      "step": 1399
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5152293334436637,
+      "learning_rate": 3.245548843911817e-05,
+      "loss": 0.0931,
+      "step": 1400
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5970091713898942,
+      "learning_rate": 3.244435310234156e-05,
+      "loss": 0.1193,
+      "step": 1401
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5920520899415288,
+      "learning_rate": 3.243321146740155e-05,
+      "loss": 0.0919,
+      "step": 1402
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5128584538616451,
+      "learning_rate": 3.2422063539937006e-05,
+      "loss": 0.0919,
+      "step": 1403
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5298398319280635,
+      "learning_rate": 3.2410909325589954e-05,
+      "loss": 0.085,
+      "step": 1404
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5807596435304535,
+      "learning_rate": 3.239974883000561e-05,
+      "loss": 0.1002,
+      "step": 1405
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5915929109199566,
+      "learning_rate": 3.2388582058832375e-05,
+      "loss": 0.117,
+      "step": 1406
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5410412944116243,
+      "learning_rate": 3.237740901772181e-05,
+      "loss": 0.1044,
+      "step": 1407
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5061377860140891,
+      "learning_rate": 3.2366229712328675e-05,
+      "loss": 0.0859,
+      "step": 1408
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5453665501606407,
+      "learning_rate": 3.235504414831087e-05,
+      "loss": 0.1005,
+      "step": 1409
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6237674002810024,
+      "learning_rate": 3.234385233132949e-05,
+      "loss": 0.1188,
+      "step": 1410
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5550178833391438,
+      "learning_rate": 3.233265426704877e-05,
+      "loss": 0.0948,
+      "step": 1411
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5620625702416668,
+      "learning_rate": 3.232144996113613e-05,
+      "loss": 0.0876,
+      "step": 1412
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6278011493027039,
+      "learning_rate": 3.231023941926213e-05,
+      "loss": 0.1235,
+      "step": 1413
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6123232800643602,
+      "learning_rate": 3.22990226471005e-05,
+      "loss": 0.1085,
+      "step": 1414
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5878673054578677,
+      "learning_rate": 3.2287799650328116e-05,
+      "loss": 0.1088,
+      "step": 1415
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5936979973757809,
+      "learning_rate": 3.2276570434625e-05,
+      "loss": 0.108,
+      "step": 1416
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5982325150532656,
+      "learning_rate": 3.226533500567433e-05,
+      "loss": 0.121,
+      "step": 1417
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5867558677097133,
+      "learning_rate": 3.2254093369162425e-05,
+      "loss": 0.115,
+      "step": 1418
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5545845473477063,
+      "learning_rate": 3.2242845530778755e-05,
+      "loss": 0.1124,
+      "step": 1419
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4871739458666332,
+      "learning_rate": 3.22315914962159e-05,
+      "loss": 0.0867,
+      "step": 1420
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4646669259868776,
+      "learning_rate": 3.2220331271169614e-05,
+      "loss": 0.1008,
+      "step": 1421
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4921344698236326,
+      "learning_rate": 3.220906486133876e-05,
+      "loss": 0.0877,
+      "step": 1422
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.506246499649945,
+      "learning_rate": 3.219779227242534e-05,
+      "loss": 0.1104,
+      "step": 1423
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5522790387999251,
+      "learning_rate": 3.218651351013447e-05,
+      "loss": 0.0856,
+      "step": 1424
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.571981261756722,
+      "learning_rate": 3.217522858017442e-05,
+      "loss": 0.0969,
+      "step": 1425
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6090202766293858,
+      "learning_rate": 3.216393748825654e-05,
+      "loss": 0.1176,
+      "step": 1426
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.44838768752167724,
+      "learning_rate": 3.2152640240095335e-05,
+      "loss": 0.0705,
+      "step": 1427
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.46354196491303296,
+      "learning_rate": 3.2141336841408406e-05,
+      "loss": 0.0671,
+      "step": 1428
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5880174468365021,
+      "learning_rate": 3.2130027297916476e-05,
+      "loss": 0.1175,
+      "step": 1429
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5015029372740631,
+      "learning_rate": 3.2118711615343366e-05,
+      "loss": 0.0792,
+      "step": 1430
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6066306502092649,
+      "learning_rate": 3.210738979941603e-05,
+      "loss": 0.1205,
+      "step": 1431
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5556678704800044,
+      "learning_rate": 3.2096061855864485e-05,
+      "loss": 0.0861,
+      "step": 1432
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5559779455558638,
+      "learning_rate": 3.2084727790421895e-05,
+      "loss": 0.1002,
+      "step": 1433
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5594356834272175,
+      "learning_rate": 3.207338760882448e-05,
+      "loss": 0.0945,
+      "step": 1434
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.49659919521560014,
+      "learning_rate": 3.20620413168116e-05,
+      "loss": 0.0853,
+      "step": 1435
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5243465444052824,
+      "learning_rate": 3.205068892012565e-05,
+      "loss": 0.0922,
+      "step": 1436
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5947183943564954,
+      "learning_rate": 3.203933042451218e-05,
+      "loss": 0.1038,
+      "step": 1437
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5286617057434441,
+      "learning_rate": 3.202796583571977e-05,
+      "loss": 0.0944,
+      "step": 1438
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6296246613226489,
+      "learning_rate": 3.2016595159500127e-05,
+      "loss": 0.1377,
+      "step": 1439
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6802923646189153,
+      "learning_rate": 3.2005218401608006e-05,
+      "loss": 0.1245,
+      "step": 1440
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4712275674018901,
+      "learning_rate": 3.1993835567801266e-05,
+      "loss": 0.0807,
+      "step": 1441
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6099906647859015,
+      "learning_rate": 3.19824466638408e-05,
+      "loss": 0.1082,
+      "step": 1442
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5531651196773758,
+      "learning_rate": 3.1971051695490644e-05,
+      "loss": 0.1041,
+      "step": 1443
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5385036889218184,
+      "learning_rate": 3.195965066851784e-05,
+      "loss": 0.1112,
+      "step": 1444
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5480856709740979,
+      "learning_rate": 3.194824358869252e-05,
+      "loss": 0.0909,
+      "step": 1445
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5267635403138891,
+      "learning_rate": 3.1936830461787866e-05,
+      "loss": 0.1064,
+      "step": 1446
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5324786787734438,
+      "learning_rate": 3.192541129358014e-05,
+      "loss": 0.0897,
+      "step": 1447
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5094036410184004,
+      "learning_rate": 3.191398608984867e-05,
+      "loss": 0.093,
+      "step": 1448
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5502984927559366,
+      "learning_rate": 3.19025548563758e-05,
+      "loss": 0.1035,
+      "step": 1449
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.579324277408475,
+      "learning_rate": 3.189111759894695e-05,
+      "loss": 0.1183,
+      "step": 1450
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5289617793855,
+      "learning_rate": 3.1879674323350594e-05,
+      "loss": 0.0944,
+      "step": 1451
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.560668692826493,
+      "learning_rate": 3.186822503537823e-05,
+      "loss": 0.1117,
+      "step": 1452
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5176155796534162,
+      "learning_rate": 3.1856769740824426e-05,
+      "loss": 0.0911,
+      "step": 1453
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.49774506343824937,
+      "learning_rate": 3.184530844548678e-05,
+      "loss": 0.0838,
+      "step": 1454
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.512447042752987,
+      "learning_rate": 3.183384115516591e-05,
+      "loss": 0.09,
+      "step": 1455
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6195166856706205,
+      "learning_rate": 3.182236787566549e-05,
+      "loss": 0.1248,
+      "step": 1456
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.49141038368318024,
+      "learning_rate": 3.181088861279222e-05,
+      "loss": 0.0838,
+      "step": 1457
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5814312908518064,
+      "learning_rate": 3.179940337235582e-05,
+      "loss": 0.1247,
+      "step": 1458
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5920081186495002,
+      "learning_rate": 3.178791216016904e-05,
+      "loss": 0.1222,
+      "step": 1459
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5004816119982474,
+      "learning_rate": 3.177641498204765e-05,
+      "loss": 0.0956,
+      "step": 1460
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.682944398095459,
+      "learning_rate": 3.1764911843810456e-05,
+      "loss": 0.1432,
+      "step": 1461
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.547596346725736,
+      "learning_rate": 3.175340275127925e-05,
+      "loss": 0.0909,
+      "step": 1462
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4748960029795314,
+      "learning_rate": 3.1741887710278874e-05,
+      "loss": 0.0804,
+      "step": 1463
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5348561413037483,
+      "learning_rate": 3.173036672663714e-05,
+      "loss": 0.0884,
+      "step": 1464
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5017013215085425,
+      "learning_rate": 3.17188398061849e-05,
+      "loss": 0.1001,
+      "step": 1465
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5777986223814436,
+      "learning_rate": 3.170730695475599e-05,
+      "loss": 0.126,
+      "step": 1466
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.533350410659888,
+      "learning_rate": 3.1695768178187267e-05,
+      "loss": 0.1064,
+      "step": 1467
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5274072962020079,
+      "learning_rate": 3.168422348231857e-05,
+      "loss": 0.095,
+      "step": 1468
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5806139901619206,
+      "learning_rate": 3.1672672872992755e-05,
+      "loss": 0.1023,
+      "step": 1469
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5420230616118438,
+      "learning_rate": 3.166111635605564e-05,
+      "loss": 0.1033,
+      "step": 1470
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5674809224959466,
+      "learning_rate": 3.164955393735605e-05,
+      "loss": 0.0948,
+      "step": 1471
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5598604472015583,
+      "learning_rate": 3.1637985622745795e-05,
+      "loss": 0.0861,
+      "step": 1472
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.48932312245861176,
+      "learning_rate": 3.1626411418079684e-05,
+      "loss": 0.0933,
+      "step": 1473
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.49542347239489665,
+      "learning_rate": 3.1614831329215475e-05,
+      "loss": 0.0905,
+      "step": 1474
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5307744365947702,
+      "learning_rate": 3.160324536201393e-05,
+      "loss": 0.0922,
+      "step": 1475
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6242590804075524,
+      "learning_rate": 3.159165352233879e-05,
+      "loss": 0.1098,
+      "step": 1476
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5361180260014112,
+      "learning_rate": 3.158005581605673e-05,
+      "loss": 0.0815,
+      "step": 1477
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5969933467395213,
+      "learning_rate": 3.156845224903745e-05,
+      "loss": 0.1015,
+      "step": 1478
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5662072456926389,
+      "learning_rate": 3.1556842827153556e-05,
+      "loss": 0.1035,
+      "step": 1479
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5801233344651101,
+      "learning_rate": 3.154522755628067e-05,
+      "loss": 0.0885,
+      "step": 1480
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5516999070831768,
+      "learning_rate": 3.153360644229735e-05,
+      "loss": 0.0943,
+      "step": 1481
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5700633928952555,
+      "learning_rate": 3.1521979491085095e-05,
+      "loss": 0.1038,
+      "step": 1482
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5798399465153998,
+      "learning_rate": 3.15103467085284e-05,
+      "loss": 0.1173,
+      "step": 1483
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5076361716853368,
+      "learning_rate": 3.149870810051467e-05,
+      "loss": 0.0959,
+      "step": 1484
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5473599327454586,
+      "learning_rate": 3.148706367293428e-05,
+      "loss": 0.0934,
+      "step": 1485
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5861413728164516,
+      "learning_rate": 3.147541343168055e-05,
+      "loss": 0.1274,
+      "step": 1486
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5542745530605213,
+      "learning_rate": 3.146375738264975e-05,
+      "loss": 0.0949,
+      "step": 1487
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5921784986551651,
+      "learning_rate": 3.145209553174105e-05,
+      "loss": 0.1391,
+      "step": 1488
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5181984692655839,
+      "learning_rate": 3.14404278848566e-05,
+      "loss": 0.0863,
+      "step": 1489
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5962280462130014,
+      "learning_rate": 3.142875444790147e-05,
+      "loss": 0.0991,
+      "step": 1490
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.48386436125900834,
+      "learning_rate": 3.141707522678365e-05,
+      "loss": 0.0813,
+      "step": 1491
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.48607849700451455,
+      "learning_rate": 3.140539022741408e-05,
+      "loss": 0.0923,
+      "step": 1492
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4443590095314898,
+      "learning_rate": 3.139369945570659e-05,
+      "loss": 0.0806,
+      "step": 1493
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.605275350877237,
+      "learning_rate": 3.138200291757797e-05,
+      "loss": 0.1126,
+      "step": 1494
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5292985641621455,
+      "learning_rate": 3.137030061894789e-05,
+      "loss": 0.0933,
+      "step": 1495
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5313127705010521,
+      "learning_rate": 3.135859256573898e-05,
+      "loss": 0.0958,
+      "step": 1496
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5485840930268028,
+      "learning_rate": 3.134687876387673e-05,
+      "loss": 0.096,
+      "step": 1497
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5495443320540766,
+      "learning_rate": 3.1335159219289585e-05,
+      "loss": 0.0903,
+      "step": 1498
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.49349570123777126,
+      "learning_rate": 3.132343393790887e-05,
+      "loss": 0.0879,
+      "step": 1499
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.48489367494468105,
+      "learning_rate": 3.131170292566883e-05,
+      "loss": 0.092,
+      "step": 1500
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5636963836988381,
+      "learning_rate": 3.12999661885066e-05,
+      "loss": 0.1014,
+      "step": 1501
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5906622809294992,
+      "learning_rate": 3.12882237323622e-05,
+      "loss": 0.107,
+      "step": 1502
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.49269801109528827,
+      "learning_rate": 3.127647556317858e-05,
+      "loss": 0.0805,
+      "step": 1503
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5129773917096273,
+      "learning_rate": 3.126472168690156e-05,
+      "loss": 0.0737,
+      "step": 1504
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4865049017917022,
+      "learning_rate": 3.125296210947983e-05,
+      "loss": 0.0882,
+      "step": 1505
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5390699634864613,
+      "learning_rate": 3.1241196836865e-05,
+      "loss": 0.0939,
+      "step": 1506
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4798041608770732,
+      "learning_rate": 3.1229425875011534e-05,
+      "loss": 0.1018,
+      "step": 1507
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.47996977350808995,
+      "learning_rate": 3.12176492298768e-05,
+      "loss": 0.0691,
+      "step": 1508
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.44069173253425836,
+      "learning_rate": 3.120586690742102e-05,
+      "loss": 0.0631,
+      "step": 1509
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5452199141375765,
+      "learning_rate": 3.119407891360732e-05,
+      "loss": 0.1006,
+      "step": 1510
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5161302576367768,
+      "learning_rate": 3.118228525440165e-05,
+      "loss": 0.0915,
+      "step": 1511
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.49472325169915915,
+      "learning_rate": 3.1170485935772864e-05,
+      "loss": 0.0884,
+      "step": 1512
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5517885722810728,
+      "learning_rate": 3.1158680963692676e-05,
+      "loss": 0.0884,
+      "step": 1513
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.46487889838167595,
+      "learning_rate": 3.114687034413564e-05,
+      "loss": 0.0756,
+      "step": 1514
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5892782721730733,
+      "learning_rate": 3.1135054083079194e-05,
+      "loss": 0.1095,
+      "step": 1515
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.513953012326265,
+      "learning_rate": 3.112323218650362e-05,
+      "loss": 0.0992,
+      "step": 1516
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4966780008099141,
+      "learning_rate": 3.111140466039205e-05,
+      "loss": 0.08,
+      "step": 1517
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.494700088726365,
+      "learning_rate": 3.1099571510730466e-05,
+      "loss": 0.0836,
+      "step": 1518
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.46990742839572375,
+      "learning_rate": 3.1087732743507704e-05,
+      "loss": 0.0791,
+      "step": 1519
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.45231846028789485,
+      "learning_rate": 3.107588836471542e-05,
+      "loss": 0.0751,
+      "step": 1520
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5732737179976433,
+      "learning_rate": 3.106403838034815e-05,
+      "loss": 0.0994,
+      "step": 1521
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.555422497418479,
+      "learning_rate": 3.1052182796403225e-05,
+      "loss": 0.1109,
+      "step": 1522
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.44664638246672456,
+      "learning_rate": 3.104032161888084e-05,
+      "loss": 0.0782,
+      "step": 1523
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.49068254145545215,
+      "learning_rate": 3.1028454853784e-05,
+      "loss": 0.0817,
+      "step": 1524
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5007696784195629,
+      "learning_rate": 3.101658250711856e-05,
+      "loss": 0.0835,
+      "step": 1525
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.47998469089890194,
+      "learning_rate": 3.100470458489318e-05,
+      "loss": 0.0867,
+      "step": 1526
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5557227346756548,
+      "learning_rate": 3.099282109311934e-05,
+      "loss": 0.1048,
+      "step": 1527
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5085157659837836,
+      "learning_rate": 3.098093203781137e-05,
+      "loss": 0.1068,
+      "step": 1528
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4919272294724993,
+      "learning_rate": 3.0969037424986376e-05,
+      "loss": 0.0793,
+      "step": 1529
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4970668800100589,
+      "learning_rate": 3.09571372606643e-05,
+      "loss": 0.0718,
+      "step": 1530
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6050020449580161,
+      "learning_rate": 3.09452315508679e-05,
+      "loss": 0.1104,
+      "step": 1531
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5542740153786689,
+      "learning_rate": 3.09333203016227e-05,
+      "loss": 0.0835,
+      "step": 1532
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5153961251662009,
+      "learning_rate": 3.0921403518957076e-05,
+      "loss": 0.0965,
+      "step": 1533
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6651957547171047,
+      "learning_rate": 3.0909481208902185e-05,
+      "loss": 0.1351,
+      "step": 1534
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5472869956005133,
+      "learning_rate": 3.089755337749198e-05,
+      "loss": 0.0888,
+      "step": 1535
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6644839351985476,
+      "learning_rate": 3.08856200307632e-05,
+      "loss": 0.114,
+      "step": 1536
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5567185196834562,
+      "learning_rate": 3.08736811747554e-05,
+      "loss": 0.0958,
+      "step": 1537
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5060256625780866,
+      "learning_rate": 3.08617368155109e-05,
+      "loss": 0.0637,
+      "step": 1538
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5294189030931142,
+      "learning_rate": 3.084978695907482e-05,
+      "loss": 0.0915,
+      "step": 1539
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5011807744937476,
+      "learning_rate": 3.0837831611495036e-05,
+      "loss": 0.0851,
+      "step": 1540
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5303344691440417,
+      "learning_rate": 3.082587077882225e-05,
+      "loss": 0.0788,
+      "step": 1541
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5673118455006193,
+      "learning_rate": 3.081390446710989e-05,
+      "loss": 0.0943,
+      "step": 1542
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5125749137834809,
+      "learning_rate": 3.080193268241419e-05,
+      "loss": 0.0832,
+      "step": 1543
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5498719250505671,
+      "learning_rate": 3.0789955430794145e-05,
+      "loss": 0.1065,
+      "step": 1544
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5784083522581387,
+      "learning_rate": 3.077797271831152e-05,
+      "loss": 0.104,
+      "step": 1545
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5247577414829844,
+      "learning_rate": 3.076598455103081e-05,
+      "loss": 0.0856,
+      "step": 1546
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5081610437163209,
+      "learning_rate": 3.0753990935019345e-05,
+      "loss": 0.0854,
+      "step": 1547
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.47914656479452084,
+      "learning_rate": 3.074199187634713e-05,
+      "loss": 0.0884,
+      "step": 1548
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4793126533479205,
+      "learning_rate": 3.072998738108699e-05,
+      "loss": 0.084,
+      "step": 1549
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5118582440755662,
+      "learning_rate": 3.071797745531445e-05,
+      "loss": 0.0909,
+      "step": 1550
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5628049871355234,
+      "learning_rate": 3.070596210510783e-05,
+      "loss": 0.1084,
+      "step": 1551
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6168438884461107,
+      "learning_rate": 3.069394133654815e-05,
+      "loss": 0.1213,
+      "step": 1552
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5079984177861392,
+      "learning_rate": 3.068191515571921e-05,
+      "loss": 0.0857,
+      "step": 1553
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.46789837405160123,
+      "learning_rate": 3.066988356870752e-05,
+      "loss": 0.0895,
+      "step": 1554
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5426420148085699,
+      "learning_rate": 3.0657846581602355e-05,
+      "loss": 0.1254,
+      "step": 1555
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.558099434620568,
+      "learning_rate": 3.06458042004957e-05,
+      "loss": 0.0789,
+      "step": 1556
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4035657639202378,
+      "learning_rate": 3.063375643148228e-05,
+      "loss": 0.062,
+      "step": 1557
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5096574573562189,
+      "learning_rate": 3.062170328065954e-05,
+      "loss": 0.1382,
+      "step": 1558
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4840329254405113,
+      "learning_rate": 3.060964475412766e-05,
+      "loss": 0.0796,
+      "step": 1559
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.49819952925350103,
+      "learning_rate": 3.059758085798954e-05,
+      "loss": 0.0909,
+      "step": 1560
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5484340243759227,
+      "learning_rate": 3.058551159835078e-05,
+      "loss": 0.0892,
+      "step": 1561
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.553949052585398,
+      "learning_rate": 3.057343698131971e-05,
+      "loss": 0.0914,
+      "step": 1562
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5722146906660561,
+      "learning_rate": 3.056135701300736e-05,
+      "loss": 0.0987,
+      "step": 1563
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5633963225303233,
+      "learning_rate": 3.054927169952749e-05,
+      "loss": 0.103,
+      "step": 1564
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6369758742073995,
+      "learning_rate": 3.053718104699654e-05,
+      "loss": 0.1306,
+      "step": 1565
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5158833810020567,
+      "learning_rate": 3.052508506153368e-05,
+      "loss": 0.0986,
+      "step": 1566
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.46353267453790725,
+      "learning_rate": 3.051298374926074e-05,
+      "loss": 0.0915,
+      "step": 1567
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5066324461289279,
+      "learning_rate": 3.0500877116302284e-05,
+      "loss": 0.0788,
+      "step": 1568
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5822538423849566,
+      "learning_rate": 3.0488765168785544e-05,
+      "loss": 0.111,
+      "step": 1569
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5673723334758027,
+      "learning_rate": 3.047664791284046e-05,
+      "loss": 0.1206,
+      "step": 1570
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5087601549133268,
+      "learning_rate": 3.046452535459963e-05,
+      "loss": 0.0846,
+      "step": 1571
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5376127346602497,
+      "learning_rate": 3.045239750019839e-05,
+      "loss": 0.0835,
+      "step": 1572
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5174346149067873,
+      "learning_rate": 3.044026435577469e-05,
+      "loss": 0.0971,
+      "step": 1573
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5149142539760165,
+      "learning_rate": 3.0428125927469198e-05,
+      "loss": 0.0895,
+      "step": 1574
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5280456957441996,
+      "learning_rate": 3.0415982221425257e-05,
+      "loss": 0.1009,
+      "step": 1575
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4897789378386961,
+      "learning_rate": 3.040383324378885e-05,
+      "loss": 0.0722,
+      "step": 1576
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5358871703246333,
+      "learning_rate": 3.0391679000708673e-05,
+      "loss": 0.1011,
+      "step": 1577
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5310148953074585,
+      "learning_rate": 3.0379519498336054e-05,
+      "loss": 0.0898,
+      "step": 1578
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5301812866397277,
+      "learning_rate": 3.036735474282498e-05,
+      "loss": 0.0927,
+      "step": 1579
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6411726169481955,
+      "learning_rate": 3.035518474033212e-05,
+      "loss": 0.1478,
+      "step": 1580
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5214365676511248,
+      "learning_rate": 3.0343009497016787e-05,
+      "loss": 0.0989,
+      "step": 1581
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5757916505202934,
+      "learning_rate": 3.0330829019040945e-05,
+      "loss": 0.1243,
+      "step": 1582
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5416318503218875,
+      "learning_rate": 3.0318643312569204e-05,
+      "loss": 0.1017,
+      "step": 1583
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5471932166567245,
+      "learning_rate": 3.0306452383768833e-05,
+      "loss": 0.0784,
+      "step": 1584
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5177630400072355,
+      "learning_rate": 3.0294256238809727e-05,
+      "loss": 0.1019,
+      "step": 1585
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.45850777738104637,
+      "learning_rate": 3.0282054883864434e-05,
+      "loss": 0.0764,
+      "step": 1586
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4849117443070667,
+      "learning_rate": 3.026984832510814e-05,
+      "loss": 0.0741,
+      "step": 1587
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5921277169521008,
+      "learning_rate": 3.025763656871865e-05,
+      "loss": 0.1184,
+      "step": 1588
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6143843819104651,
+      "learning_rate": 3.024541962087641e-05,
+      "loss": 0.1227,
+      "step": 1589
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4186554019034123,
+      "learning_rate": 3.0233197487764494e-05,
+      "loss": 0.0758,
+      "step": 1590
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4419283184812778,
+      "learning_rate": 3.0220970175568604e-05,
+      "loss": 0.0936,
+      "step": 1591
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.49122642356168894,
+      "learning_rate": 3.020873769047705e-05,
+      "loss": 0.1048,
+      "step": 1592
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5101086861605791,
+      "learning_rate": 3.019650003868077e-05,
+      "loss": 0.0856,
+      "step": 1593
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5561429928068005,
+      "learning_rate": 3.0184257226373317e-05,
+      "loss": 0.0963,
+      "step": 1594
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4809035682977758,
+      "learning_rate": 3.0172009259750852e-05,
+      "loss": 0.0893,
+      "step": 1595
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5498611936108522,
+      "learning_rate": 3.015975614501214e-05,
+      "loss": 0.1033,
+      "step": 1596
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5622349220448989,
+      "learning_rate": 3.0147497888358564e-05,
+      "loss": 0.0949,
+      "step": 1597
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.479641613334266,
+      "learning_rate": 3.0135234495994107e-05,
+      "loss": 0.0908,
+      "step": 1598
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5252805686745193,
+      "learning_rate": 3.0122965974125335e-05,
+      "loss": 0.0927,
+      "step": 1599
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.598178610317444,
+      "learning_rate": 3.0110692328961435e-05,
+      "loss": 0.118,
+      "step": 1600
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5620250816314273,
+      "learning_rate": 3.0098413566714165e-05,
+      "loss": 0.1106,
+      "step": 1601
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5314593183248383,
+      "learning_rate": 3.008612969359788e-05,
+      "loss": 0.0918,
+      "step": 1602
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4614640800284727,
+      "learning_rate": 3.0073840715829532e-05,
+      "loss": 0.0706,
+      "step": 1603
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4640629244972625,
+      "learning_rate": 3.006154663962865e-05,
+      "loss": 0.0763,
+      "step": 1604
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4099952563441726,
+      "learning_rate": 3.0049247471217326e-05,
+      "loss": 0.0592,
+      "step": 1605
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5684856523583793,
+      "learning_rate": 3.0036943216820256e-05,
+      "loss": 0.1109,
+      "step": 1606
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.47359148612090135,
+      "learning_rate": 3.00246338826647e-05,
+      "loss": 0.0738,
+      "step": 1607
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.47715175657175857,
+      "learning_rate": 3.001231947498048e-05,
+      "loss": 0.0643,
+      "step": 1608
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5851278178459499,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.1059,
+      "step": 1609
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6081181794555448,
+      "learning_rate": 2.998767546395822e-05,
+      "loss": 0.1035,
+      "step": 1610
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.47406968880158457,
+      "learning_rate": 2.9975345873092662e-05,
+      "loss": 0.0911,
+      "step": 1611
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.511908372296521,
+      "learning_rate": 2.996301123364341e-05,
+      "loss": 0.0761,
+      "step": 1612
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5312066902009361,
+      "learning_rate": 2.9950671551853094e-05,
+      "loss": 0.0902,
+      "step": 1613
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5479439778163137,
+      "learning_rate": 2.9938326833966914e-05,
+      "loss": 0.1056,
+      "step": 1614
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5139200169640844,
+      "learning_rate": 2.992597708623259e-05,
+      "loss": 0.1016,
+      "step": 1615
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4303417426917004,
+      "learning_rate": 2.991362231490042e-05,
+      "loss": 0.0837,
+      "step": 1616
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5217671859754873,
+      "learning_rate": 2.990126252622323e-05,
+      "loss": 0.091,
+      "step": 1617
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5249282333731543,
+      "learning_rate": 2.9888897726456374e-05,
+      "loss": 0.0917,
+      "step": 1618
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5065982386168,
+      "learning_rate": 2.9876527921857756e-05,
+      "loss": 0.1009,
+      "step": 1619
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5157414796063478,
+      "learning_rate": 2.986415311868782e-05,
+      "loss": 0.091,
+      "step": 1620
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4295084493558709,
+      "learning_rate": 2.985177332320952e-05,
+      "loss": 0.0606,
+      "step": 1621
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5135798788910292,
+      "learning_rate": 2.9839388541688352e-05,
+      "loss": 0.0907,
+      "step": 1622
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5477632937661516,
+      "learning_rate": 2.9826998780392324e-05,
+      "loss": 0.1087,
+      "step": 1623
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5377968013052618,
+      "learning_rate": 2.9814604045591974e-05,
+      "loss": 0.082,
+      "step": 1624
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4219926210962844,
+      "learning_rate": 2.980220434356035e-05,
+      "loss": 0.0731,
+      "step": 1625
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.45800952057021754,
+      "learning_rate": 2.9789799680573014e-05,
+      "loss": 0.0815,
+      "step": 1626
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4190818901895578,
+      "learning_rate": 2.9777390062908056e-05,
+      "loss": 0.0612,
+      "step": 1627
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.654880934842561,
+      "learning_rate": 2.976497549684605e-05,
+      "loss": 0.0936,
+      "step": 1628
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.44223417433418194,
+      "learning_rate": 2.9752555988670084e-05,
+      "loss": 0.064,
+      "step": 1629
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5239911069788755,
+      "learning_rate": 2.9740131544665748e-05,
+      "loss": 0.0842,
+      "step": 1630
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4960591161457972,
+      "learning_rate": 2.9727702171121125e-05,
+      "loss": 0.082,
+      "step": 1631
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4905212900166773,
+      "learning_rate": 2.9715267874326805e-05,
+      "loss": 0.0901,
+      "step": 1632
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.46224708730131725,
+      "learning_rate": 2.970282866057586e-05,
+      "loss": 0.0767,
+      "step": 1633
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5464459507312848,
+      "learning_rate": 2.969038453616385e-05,
+      "loss": 0.0953,
+      "step": 1634
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4984873899167725,
+      "learning_rate": 2.9677935507388817e-05,
+      "loss": 0.0708,
+      "step": 1635
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.534677360835388,
+      "learning_rate": 2.96654815805513e-05,
+      "loss": 0.0912,
+      "step": 1636
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4562672481150706,
+      "learning_rate": 2.965302276195431e-05,
+      "loss": 0.0751,
+      "step": 1637
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.37727624751445854,
+      "learning_rate": 2.9640559057903325e-05,
+      "loss": 0.0692,
+      "step": 1638
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.47634405851813993,
+      "learning_rate": 2.9628090474706304e-05,
+      "loss": 0.088,
+      "step": 1639
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.45687423356353307,
+      "learning_rate": 2.9615617018673663e-05,
+      "loss": 0.0689,
+      "step": 1640
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4847616221641512,
+      "learning_rate": 2.9603138696118315e-05,
+      "loss": 0.0768,
+      "step": 1641
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4417883982633577,
+      "learning_rate": 2.9590655513355598e-05,
+      "loss": 0.0727,
+      "step": 1642
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.44089015395594955,
+      "learning_rate": 2.957816747670334e-05,
+      "loss": 0.06,
+      "step": 1643
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.42681674855841306,
+      "learning_rate": 2.956567459248181e-05,
+      "loss": 0.0702,
+      "step": 1644
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.566858413689165,
+      "learning_rate": 2.9553176867013714e-05,
+      "loss": 0.102,
+      "step": 1645
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5197223799349769,
+      "learning_rate": 2.9540674306624262e-05,
+      "loss": 0.0942,
+      "step": 1646
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5280340692011758,
+      "learning_rate": 2.9528166917641048e-05,
+      "loss": 0.0882,
+      "step": 1647
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5924065599923164,
+      "learning_rate": 2.951565470639415e-05,
+      "loss": 0.1015,
+      "step": 1648
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5174930718552494,
+      "learning_rate": 2.9503137679216073e-05,
+      "loss": 0.0926,
+      "step": 1649
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5359374699717699,
+      "learning_rate": 2.9490615842441764e-05,
+      "loss": 0.0907,
+      "step": 1650
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4148972931125151,
+      "learning_rate": 2.94780892024086e-05,
+      "loss": 0.0677,
+      "step": 1651
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5461000946942457,
+      "learning_rate": 2.9465557765456387e-05,
+      "loss": 0.0973,
+      "step": 1652
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4813988529715445,
+      "learning_rate": 2.9453021537927363e-05,
+      "loss": 0.077,
+      "step": 1653
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.46848046182768466,
+      "learning_rate": 2.9440480526166193e-05,
+      "loss": 0.0686,
+      "step": 1654
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5593676507989175,
+      "learning_rate": 2.9427934736519962e-05,
+      "loss": 0.0954,
+      "step": 1655
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5000285200091341,
+      "learning_rate": 2.9415384175338154e-05,
+      "loss": 0.0729,
+      "step": 1656
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4873686387291929,
+      "learning_rate": 2.9402828848972706e-05,
+      "loss": 0.0838,
+      "step": 1657
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4687000724905962,
+      "learning_rate": 2.9390268763777938e-05,
+      "loss": 0.0793,
+      "step": 1658
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.47911341170825583,
+      "learning_rate": 2.937770392611058e-05,
+      "loss": 0.089,
+      "step": 1659
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5289099148013381,
+      "learning_rate": 2.9365134342329783e-05,
+      "loss": 0.0908,
+      "step": 1660
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5187240295343717,
+      "learning_rate": 2.935256001879709e-05,
+      "loss": 0.0853,
+      "step": 1661
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5361045176781944,
+      "learning_rate": 2.9339980961876434e-05,
+      "loss": 0.1128,
+      "step": 1662
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5844729876855916,
+      "learning_rate": 2.932739717793416e-05,
+      "loss": 0.0931,
+      "step": 1663
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.48204295918542,
+      "learning_rate": 2.9314808673338997e-05,
+      "loss": 0.0693,
+      "step": 1664
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5227459566825613,
+      "learning_rate": 2.9302215454462063e-05,
+      "loss": 0.0818,
+      "step": 1665
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.49666751512139234,
+      "learning_rate": 2.928961752767686e-05,
+      "loss": 0.0805,
+      "step": 1666
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5397645000579105,
+      "learning_rate": 2.9277014899359284e-05,
+      "loss": 0.0955,
+      "step": 1667
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.49132051250609765,
+      "learning_rate": 2.92644075758876e-05,
+      "loss": 0.0943,
+      "step": 1668
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.49830369498658783,
+      "learning_rate": 2.9251795563642445e-05,
+      "loss": 0.0933,
+      "step": 1669
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4875517622048043,
+      "learning_rate": 2.923917886900685e-05,
+      "loss": 0.0717,
+      "step": 1670
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.45240061506285206,
+      "learning_rate": 2.922655749836618e-05,
+      "loss": 0.0757,
+      "step": 1671
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4454040265973597,
+      "learning_rate": 2.921393145810821e-05,
+      "loss": 0.0709,
+      "step": 1672
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.511675650841848,
+      "learning_rate": 2.9201300754623046e-05,
+      "loss": 0.0962,
+      "step": 1673
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5622649231419086,
+      "learning_rate": 2.9188665394303163e-05,
+      "loss": 0.1,
+      "step": 1674
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5357272311190564,
+      "learning_rate": 2.9176025383543395e-05,
+      "loss": 0.0753,
+      "step": 1675
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4992327167306387,
+      "learning_rate": 2.916338072874093e-05,
+      "loss": 0.0919,
+      "step": 1676
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5875316043225938,
+      "learning_rate": 2.915073143629531e-05,
+      "loss": 0.111,
+      "step": 1677
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5132310956027424,
+      "learning_rate": 2.9138077512608417e-05,
+      "loss": 0.0993,
+      "step": 1678
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4525387029271844,
+      "learning_rate": 2.9125418964084474e-05,
+      "loss": 0.0912,
+      "step": 1679
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5493623027701254,
+      "learning_rate": 2.9112755797130052e-05,
+      "loss": 0.1015,
+      "step": 1680
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.47496973493428624,
+      "learning_rate": 2.910008801815406e-05,
+      "loss": 0.0692,
+      "step": 1681
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.487745387722754,
+      "learning_rate": 2.908741563356774e-05,
+      "loss": 0.0816,
+      "step": 1682
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5300960630503472,
+      "learning_rate": 2.9074738649784665e-05,
+      "loss": 0.0796,
+      "step": 1683
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4775093397006017,
+      "learning_rate": 2.9062057073220723e-05,
+      "loss": 0.0837,
+      "step": 1684
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4841079591075937,
+      "learning_rate": 2.9049370910294143e-05,
+      "loss": 0.0792,
+      "step": 1685
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.44417570809304496,
+      "learning_rate": 2.9036680167425476e-05,
+      "loss": 0.0658,
+      "step": 1686
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5095910836808137,
+      "learning_rate": 2.902398485103758e-05,
+      "loss": 0.1016,
+      "step": 1687
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4611040557981612,
+      "learning_rate": 2.901128496755564e-05,
+      "loss": 0.0745,
+      "step": 1688
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5023244234107879,
+      "learning_rate": 2.899858052340713e-05,
+      "loss": 0.077,
+      "step": 1689
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4784058581203675,
+      "learning_rate": 2.8985871525021857e-05,
+      "loss": 0.0832,
+      "step": 1690
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5384177623104506,
+      "learning_rate": 2.897315797883192e-05,
+      "loss": 0.1033,
+      "step": 1691
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4415949696551052,
+      "learning_rate": 2.896043989127172e-05,
+      "loss": 0.0798,
+      "step": 1692
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.46916304510129275,
+      "learning_rate": 2.8947717268777968e-05,
+      "loss": 0.0854,
+      "step": 1693
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5489348718405892,
+      "learning_rate": 2.8934990117789658e-05,
+      "loss": 0.1041,
+      "step": 1694
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4737551187122979,
+      "learning_rate": 2.8922258444748074e-05,
+      "loss": 0.0661,
+      "step": 1695
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.529193290906736,
+      "learning_rate": 2.8909522256096795e-05,
+      "loss": 0.0812,
+      "step": 1696
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.47969049521141716,
+      "learning_rate": 2.8896781558281688e-05,
+      "loss": 0.0744,
+      "step": 1697
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4855578500331324,
+      "learning_rate": 2.88840363577509e-05,
+      "loss": 0.0651,
+      "step": 1698
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5289464261954826,
+      "learning_rate": 2.8871286660954846e-05,
+      "loss": 0.0764,
+      "step": 1699
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.45181143211991037,
+      "learning_rate": 2.8858532474346232e-05,
+      "loss": 0.0728,
+      "step": 1700
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.48186171593767424,
+      "learning_rate": 2.8845773804380028e-05,
+      "loss": 0.0715,
+      "step": 1701
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5211316176442896,
+      "learning_rate": 2.883301065751348e-05,
+      "loss": 0.1015,
+      "step": 1702
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4493606509118241,
+      "learning_rate": 2.882024304020609e-05,
+      "loss": 0.0631,
+      "step": 1703
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.45422305871096313,
+      "learning_rate": 2.8807470958919626e-05,
+      "loss": 0.0674,
+      "step": 1704
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4887822899310735,
+      "learning_rate": 2.8794694420118117e-05,
+      "loss": 0.0814,
+      "step": 1705
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5160722238022565,
+      "learning_rate": 2.8781913430267857e-05,
+      "loss": 0.0777,
+      "step": 1706
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5285768573530069,
+      "learning_rate": 2.876912799583737e-05,
+      "loss": 0.0714,
+      "step": 1707
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6155821286682815,
+      "learning_rate": 2.8756338123297455e-05,
+      "loss": 0.1196,
+      "step": 1708
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.48713356943010355,
+      "learning_rate": 2.8743543819121132e-05,
+      "loss": 0.0688,
+      "step": 1709
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5582549394781079,
+      "learning_rate": 2.8730745089783686e-05,
+      "loss": 0.0951,
+      "step": 1710
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5635245846493944,
+      "learning_rate": 2.871794194176263e-05,
+      "loss": 0.1159,
+      "step": 1711
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4709760578978905,
+      "learning_rate": 2.8705134381537718e-05,
+      "loss": 0.0635,
+      "step": 1712
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5130103896057514,
+      "learning_rate": 2.869232241559093e-05,
+      "loss": 0.0764,
+      "step": 1713
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4696199133462722,
+      "learning_rate": 2.8679506050406475e-05,
+      "loss": 0.0656,
+      "step": 1714
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5931310187899642,
+      "learning_rate": 2.8666685292470814e-05,
+      "loss": 0.1018,
+      "step": 1715
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5262454428214967,
+      "learning_rate": 2.8653860148272596e-05,
+      "loss": 0.0946,
+      "step": 1716
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5290788996994328,
+      "learning_rate": 2.8641030624302704e-05,
+      "loss": 0.0929,
+      "step": 1717
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4861553758570568,
+      "learning_rate": 2.8628196727054244e-05,
+      "loss": 0.0931,
+      "step": 1718
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.44621232790038645,
+      "learning_rate": 2.8615358463022533e-05,
+      "loss": 0.0729,
+      "step": 1719
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4687295909253299,
+      "learning_rate": 2.860251583870509e-05,
+      "loss": 0.095,
+      "step": 1720
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.41306681848024107,
+      "learning_rate": 2.8589668860601643e-05,
+      "loss": 0.0783,
+      "step": 1721
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4714748814598099,
+      "learning_rate": 2.857681753521413e-05,
+      "loss": 0.0862,
+      "step": 1722
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.47383239316472453,
+      "learning_rate": 2.856396186904669e-05,
+      "loss": 0.0586,
+      "step": 1723
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.49578643651846827,
+      "learning_rate": 2.8551101868605644e-05,
+      "loss": 0.081,
+      "step": 1724
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5084300467725035,
+      "learning_rate": 2.8538237540399528e-05,
+      "loss": 0.0899,
+      "step": 1725
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5145259762148203,
+      "learning_rate": 2.8525368890939055e-05,
+      "loss": 0.1061,
+      "step": 1726
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4566929861417621,
+      "learning_rate": 2.851249592673712e-05,
+      "loss": 0.0661,
+      "step": 1727
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5720207810244956,
+      "learning_rate": 2.8499618654308815e-05,
+      "loss": 0.1057,
+      "step": 1728
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5089245284548529,
+      "learning_rate": 2.8486737080171405e-05,
+      "loss": 0.0875,
+      "step": 1729
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4464601818882488,
+      "learning_rate": 2.847385121084434e-05,
+      "loss": 0.059,
+      "step": 1730
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.47582491093389556,
+      "learning_rate": 2.8460961052849222e-05,
+      "loss": 0.0879,
+      "step": 1731
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.426657448845374,
+      "learning_rate": 2.8448066612709854e-05,
+      "loss": 0.0684,
+      "step": 1732
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.44690866676969737,
+      "learning_rate": 2.843516789695219e-05,
+      "loss": 0.0584,
+      "step": 1733
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5096894422876526,
+      "learning_rate": 2.842226491210434e-05,
+      "loss": 0.0747,
+      "step": 1734
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4899342442689138,
+      "learning_rate": 2.8409357664696585e-05,
+      "loss": 0.0949,
+      "step": 1735
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5672997103580923,
+      "learning_rate": 2.8396446161261372e-05,
+      "loss": 0.0891,
+      "step": 1736
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5439526542017444,
+      "learning_rate": 2.8383530408333285e-05,
+      "loss": 0.0949,
+      "step": 1737
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.48880725756917676,
+      "learning_rate": 2.8370610412449066e-05,
+      "loss": 0.0777,
+      "step": 1738
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.46023548137096015,
+      "learning_rate": 2.8357686180147604e-05,
+      "loss": 0.0708,
+      "step": 1739
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.44760904912110805,
+      "learning_rate": 2.834475771796993e-05,
+      "loss": 0.0571,
+      "step": 1740
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5765089956129296,
+      "learning_rate": 2.8331825032459228e-05,
+      "loss": 0.1086,
+      "step": 1741
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.44519501525006766,
+      "learning_rate": 2.8318888130160796e-05,
+      "loss": 0.076,
+      "step": 1742
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4456507586007182,
+      "learning_rate": 2.830594701762209e-05,
+      "loss": 0.0683,
+      "step": 1743
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4807542557401542,
+      "learning_rate": 2.8293001701392677e-05,
+      "loss": 0.0814,
+      "step": 1744
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.46088174709060936,
+      "learning_rate": 2.828005218802427e-05,
+      "loss": 0.0681,
+      "step": 1745
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.48053604523088983,
+      "learning_rate": 2.8267098484070693e-05,
+      "loss": 0.0894,
+      "step": 1746
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6099883440733267,
+      "learning_rate": 2.8254140596087897e-05,
+      "loss": 0.1019,
+      "step": 1747
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4979950821817546,
+      "learning_rate": 2.8241178530633947e-05,
+      "loss": 0.099,
+      "step": 1748
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5786325700981236,
+      "learning_rate": 2.822821229426902e-05,
+      "loss": 0.1113,
+      "step": 1749
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5114477182388685,
+      "learning_rate": 2.8215241893555415e-05,
+      "loss": 0.0924,
+      "step": 1750
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6553502044963802,
+      "learning_rate": 2.8202267335057522e-05,
+      "loss": 0.1022,
+      "step": 1751
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.44788021856062976,
+      "learning_rate": 2.818928862534185e-05,
+      "loss": 0.0545,
+      "step": 1752
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5089788979669672,
+      "learning_rate": 2.817630577097701e-05,
+      "loss": 0.0936,
+      "step": 1753
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5125735183646032,
+      "learning_rate": 2.8163318778533692e-05,
+      "loss": 0.0874,
+      "step": 1754
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.47712190033503193,
+      "learning_rate": 2.81503276545847e-05,
+      "loss": 0.0893,
+      "step": 1755
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4760484424362048,
+      "learning_rate": 2.8137332405704922e-05,
+      "loss": 0.0736,
+      "step": 1756
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4860070942294226,
+      "learning_rate": 2.812433303847133e-05,
+      "loss": 0.083,
+      "step": 1757
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4364222125430108,
+      "learning_rate": 2.811132955946298e-05,
+      "loss": 0.0692,
+      "step": 1758
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.476793749000669,
+      "learning_rate": 2.8098321975261026e-05,
+      "loss": 0.071,
+      "step": 1759
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4861955576919296,
+      "learning_rate": 2.8085310292448666e-05,
+      "loss": 0.0799,
+      "step": 1760
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.541403703559733,
+      "learning_rate": 2.8072294517611208e-05,
+      "loss": 0.094,
+      "step": 1761
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4389854315999451,
+      "learning_rate": 2.805927465733601e-05,
+      "loss": 0.0738,
+      "step": 1762
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4030064060918374,
+      "learning_rate": 2.8046250718212507e-05,
+      "loss": 0.0583,
+      "step": 1763
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4475027397007963,
+      "learning_rate": 2.8033222706832187e-05,
+      "loss": 0.0786,
+      "step": 1764
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5462551418734595,
+      "learning_rate": 2.802019062978861e-05,
+      "loss": 0.0936,
+      "step": 1765
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.45795786896413304,
+      "learning_rate": 2.80071544936774e-05,
+      "loss": 0.0723,
+      "step": 1766
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4787774284563339,
+      "learning_rate": 2.7994114305096208e-05,
+      "loss": 0.0738,
+      "step": 1767
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.47482042368927674,
+      "learning_rate": 2.7981070070644764e-05,
+      "loss": 0.0885,
+      "step": 1768
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.512290693201746,
+      "learning_rate": 2.7968021796924834e-05,
+      "loss": 0.1083,
+      "step": 1769
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5731261621710955,
+      "learning_rate": 2.7954969490540223e-05,
+      "loss": 0.0942,
+      "step": 1770
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4821992429933812,
+      "learning_rate": 2.7941913158096792e-05,
+      "loss": 0.0758,
+      "step": 1771
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.44004913602999246,
+      "learning_rate": 2.7928852806202424e-05,
+      "loss": 0.0578,
+      "step": 1772
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4511290690935369,
+      "learning_rate": 2.7915788441467052e-05,
+      "loss": 0.0722,
+      "step": 1773
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5211129169284835,
+      "learning_rate": 2.790272007050262e-05,
+      "loss": 0.0794,
+      "step": 1774
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4790374868523092,
+      "learning_rate": 2.7889647699923114e-05,
+      "loss": 0.0806,
+      "step": 1775
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4536289832412529,
+      "learning_rate": 2.7876571336344546e-05,
+      "loss": 0.0626,
+      "step": 1776
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4069042439652636,
+      "learning_rate": 2.7863490986384945e-05,
+      "loss": 0.0739,
+      "step": 1777
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4723581682503797,
+      "learning_rate": 2.7850406656664346e-05,
+      "loss": 0.0899,
+      "step": 1778
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5206130388403967,
+      "learning_rate": 2.783731835380482e-05,
+      "loss": 0.0877,
+      "step": 1779
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4326390418050192,
+      "learning_rate": 2.782422608443043e-05,
+      "loss": 0.0569,
+      "step": 1780
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4592044870680303,
+      "learning_rate": 2.781112985516725e-05,
+      "loss": 0.0694,
+      "step": 1781
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5556496080285102,
+      "learning_rate": 2.7798029672643375e-05,
+      "loss": 0.0923,
+      "step": 1782
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5149313856452441,
+      "learning_rate": 2.778492554348887e-05,
+      "loss": 0.0843,
+      "step": 1783
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5557231100718248,
+      "learning_rate": 2.7771817474335835e-05,
+      "loss": 0.09,
+      "step": 1784
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4594406943803711,
+      "learning_rate": 2.7758705471818327e-05,
+      "loss": 0.0775,
+      "step": 1785
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4263839655470079,
+      "learning_rate": 2.7745589542572424e-05,
+      "loss": 0.0609,
+      "step": 1786
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.43457029535529285,
+      "learning_rate": 2.7732469693236166e-05,
+      "loss": 0.0683,
+      "step": 1787
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.549615261585891,
+      "learning_rate": 2.77193459304496e-05,
+      "loss": 0.0943,
+      "step": 1788
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4156646680903677,
+      "learning_rate": 2.7706218260854738e-05,
+      "loss": 0.0583,
+      "step": 1789
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4454491221674145,
+      "learning_rate": 2.7693086691095573e-05,
+      "loss": 0.0791,
+      "step": 1790
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4934414527256401,
+      "learning_rate": 2.767995122781807e-05,
+      "loss": 0.0768,
+      "step": 1791
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.47652166222870884,
+      "learning_rate": 2.7666811877670177e-05,
+      "loss": 0.0723,
+      "step": 1792
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4011121621863688,
+      "learning_rate": 2.7653668647301797e-05,
+      "loss": 0.0605,
+      "step": 1793
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4426381298575097,
+      "learning_rate": 2.7640521543364797e-05,
+      "loss": 0.0705,
+      "step": 1794
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5711811713692367,
+      "learning_rate": 2.7627370572513005e-05,
+      "loss": 0.0787,
+      "step": 1795
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5513816158664855,
+      "learning_rate": 2.7614215741402204e-05,
+      "loss": 0.0921,
+      "step": 1796
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4770801109582723,
+      "learning_rate": 2.7601057056690148e-05,
+      "loss": 0.0572,
+      "step": 1797
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4465907973766518,
+      "learning_rate": 2.7587894525036517e-05,
+      "loss": 0.0676,
+      "step": 1798
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.43528794602088433,
+      "learning_rate": 2.7574728153102956e-05,
+      "loss": 0.064,
+      "step": 1799
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.45708921668069113,
+      "learning_rate": 2.7561557947553037e-05,
+      "loss": 0.0721,
+      "step": 1800
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.49661595351522414,
+      "learning_rate": 2.7548383915052287e-05,
+      "loss": 0.0833,
+      "step": 1801
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4647576709194282,
+      "learning_rate": 2.7535206062268174e-05,
+      "loss": 0.0874,
+      "step": 1802
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4758847529637113,
+      "learning_rate": 2.7522024395870075e-05,
+      "loss": 0.073,
+      "step": 1803
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5512991289922461,
+      "learning_rate": 2.7508838922529316e-05,
+      "loss": 0.0728,
+      "step": 1804
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.45794397490796895,
+      "learning_rate": 2.7495649648919153e-05,
+      "loss": 0.082,
+      "step": 1805
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5781542280902109,
+      "learning_rate": 2.7482456581714757e-05,
+      "loss": 0.1036,
+      "step": 1806
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5296405647661102,
+      "learning_rate": 2.7469259727593213e-05,
+      "loss": 0.0862,
+      "step": 1807
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.43297235105373694,
+      "learning_rate": 2.7456059093233537e-05,
+      "loss": 0.0723,
+      "step": 1808
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.47428964905001236,
+      "learning_rate": 2.7442854685316643e-05,
+      "loss": 0.082,
+      "step": 1809
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.45791286631225797,
+      "learning_rate": 2.7429646510525373e-05,
+      "loss": 0.0568,
+      "step": 1810
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.442750554782858,
+      "learning_rate": 2.7416434575544455e-05,
+      "loss": 0.0675,
+      "step": 1811
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.47779904642868215,
+      "learning_rate": 2.7403218887060538e-05,
+      "loss": 0.0828,
+      "step": 1812
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4937883108703817,
+      "learning_rate": 2.738999945176215e-05,
+      "loss": 0.0599,
+      "step": 1813
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.46984544546305385,
+      "learning_rate": 2.7376776276339745e-05,
+      "loss": 0.0701,
+      "step": 1814
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4381758034347087,
+      "learning_rate": 2.7363549367485648e-05,
+      "loss": 0.0646,
+      "step": 1815
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.46041729230823847,
+      "learning_rate": 2.7350318731894075e-05,
+      "loss": 0.0812,
+      "step": 1816
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4205301713824055,
+      "learning_rate": 2.7337084376261135e-05,
+      "loss": 0.0705,
+      "step": 1817
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4711574720111425,
+      "learning_rate": 2.7323846307284814e-05,
+      "loss": 0.0731,
+      "step": 1818
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.46499916491894955,
+      "learning_rate": 2.7310604531664983e-05,
+      "loss": 0.0679,
+      "step": 1819
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4927366614179062,
+      "learning_rate": 2.7297359056103378e-05,
+      "loss": 0.085,
+      "step": 1820
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.45650688477575435,
+      "learning_rate": 2.7284109887303628e-05,
+      "loss": 0.071,
+      "step": 1821
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5742329446504342,
+      "learning_rate": 2.7270857031971203e-05,
+      "loss": 0.1062,
+      "step": 1822
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.531793288511195,
+      "learning_rate": 2.7257600496813475e-05,
+      "loss": 0.0912,
+      "step": 1823
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.502106461315447,
+      "learning_rate": 2.7244340288539638e-05,
+      "loss": 0.0726,
+      "step": 1824
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5183084194087579,
+      "learning_rate": 2.7231076413860774e-05,
+      "loss": 0.0635,
+      "step": 1825
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.45798770559459295,
+      "learning_rate": 2.721780887948981e-05,
+      "loss": 0.0539,
+      "step": 1826
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4696202623789847,
+      "learning_rate": 2.7204537692141526e-05,
+      "loss": 0.0771,
+      "step": 1827
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.43115573073857216,
+      "learning_rate": 2.7191262858532552e-05,
+      "loss": 0.0682,
+      "step": 1828
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.49340640619357506,
+      "learning_rate": 2.7177984385381366e-05,
+      "loss": 0.0853,
+      "step": 1829
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.47025112752865444,
+      "learning_rate": 2.7164702279408275e-05,
+      "loss": 0.0615,
+      "step": 1830
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4015113404633209,
+      "learning_rate": 2.715141654733544e-05,
+      "loss": 0.0517,
+      "step": 1831
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5527488248304472,
+      "learning_rate": 2.7138127195886856e-05,
+      "loss": 0.0959,
+      "step": 1832
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4134978456925406,
+      "learning_rate": 2.712483423178834e-05,
+      "loss": 0.0639,
+      "step": 1833
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.38267705428554666,
+      "learning_rate": 2.7111537661767537e-05,
+      "loss": 0.0503,
+      "step": 1834
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.502958604296117,
+      "learning_rate": 2.7098237492553937e-05,
+      "loss": 0.0683,
+      "step": 1835
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4738071710618732,
+      "learning_rate": 2.7084933730878824e-05,
+      "loss": 0.0812,
+      "step": 1836
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5045485432184268,
+      "learning_rate": 2.7071626383475327e-05,
+      "loss": 0.0827,
+      "step": 1837
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4528071340923516,
+      "learning_rate": 2.7058315457078358e-05,
+      "loss": 0.0705,
+      "step": 1838
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4197753118195681,
+      "learning_rate": 2.7045000958424674e-05,
+      "loss": 0.0679,
+      "step": 1839
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5344421188410078,
+      "learning_rate": 2.7031682894252816e-05,
+      "loss": 0.0901,
+      "step": 1840
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.48017108039036727,
+      "learning_rate": 2.701836127130314e-05,
+      "loss": 0.0642,
+      "step": 1841
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4774242801446497,
+      "learning_rate": 2.7005036096317802e-05,
+      "loss": 0.0709,
+      "step": 1842
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4355363374128159,
+      "learning_rate": 2.6991707376040755e-05,
+      "loss": 0.0642,
+      "step": 1843
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5430117802755907,
+      "learning_rate": 2.6978375117217743e-05,
+      "loss": 0.0812,
+      "step": 1844
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.44822825823647566,
+      "learning_rate": 2.696503932659631e-05,
+      "loss": 0.0683,
+      "step": 1845
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.44747005604877643,
+      "learning_rate": 2.6951700010925774e-05,
+      "loss": 0.0627,
+      "step": 1846
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6233900076002594,
+      "learning_rate": 2.6938357176957243e-05,
+      "loss": 0.0998,
+      "step": 1847
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5089348933656639,
+      "learning_rate": 2.692501083144362e-05,
+      "loss": 0.0581,
+      "step": 1848
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4220702284742675,
+      "learning_rate": 2.6911660981139563e-05,
+      "loss": 0.0589,
+      "step": 1849
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4690353796410688,
+      "learning_rate": 2.6898307632801515e-05,
+      "loss": 0.0624,
+      "step": 1850
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4031684164575929,
+      "learning_rate": 2.6884950793187684e-05,
+      "loss": 0.0523,
+      "step": 1851
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4621623301535883,
+      "learning_rate": 2.6871590469058052e-05,
+      "loss": 0.0745,
+      "step": 1852
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.46212234801320057,
+      "learning_rate": 2.6858226667174362e-05,
+      "loss": 0.0803,
+      "step": 1853
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4780853049312022,
+      "learning_rate": 2.684485939430011e-05,
+      "loss": 0.0621,
+      "step": 1854
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5359598516579986,
+      "learning_rate": 2.683148865720056e-05,
+      "loss": 0.1006,
+      "step": 1855
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3983820708371192,
+      "learning_rate": 2.6818114462642726e-05,
+      "loss": 0.0472,
+      "step": 1856
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4445445650382663,
+      "learning_rate": 2.6804736817395362e-05,
+      "loss": 0.0584,
+      "step": 1857
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5937913830038436,
+      "learning_rate": 2.6791355728228986e-05,
+      "loss": 0.1188,
+      "step": 1858
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.49584486117172055,
+      "learning_rate": 2.6777971201915843e-05,
+      "loss": 0.0693,
+      "step": 1859
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5385538271988911,
+      "learning_rate": 2.676458324522992e-05,
+      "loss": 0.0734,
+      "step": 1860
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5635618148652323,
+      "learning_rate": 2.675119186494696e-05,
+      "loss": 0.0983,
+      "step": 1861
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.48687003097642834,
+      "learning_rate": 2.6737797067844403e-05,
+      "loss": 0.068,
+      "step": 1862
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5590274434322308,
+      "learning_rate": 2.6724398860701453e-05,
+      "loss": 0.0995,
+      "step": 1863
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4434149114972751,
+      "learning_rate": 2.6710997250299012e-05,
+      "loss": 0.0634,
+      "step": 1864
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.41477458829925434,
+      "learning_rate": 2.6697592243419723e-05,
+      "loss": 0.0588,
+      "step": 1865
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.41729900141627835,
+      "learning_rate": 2.668418384684795e-05,
+      "loss": 0.0632,
+      "step": 1866
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4770965866541249,
+      "learning_rate": 2.6670772067369754e-05,
+      "loss": 0.0783,
+      "step": 1867
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4641929366087564,
+      "learning_rate": 2.6657356911772922e-05,
+      "loss": 0.0657,
+      "step": 1868
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.38559123115511384,
+      "learning_rate": 2.6643938386846945e-05,
+      "loss": 0.0578,
+      "step": 1869
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.47675891643241647,
+      "learning_rate": 2.663051649938303e-05,
+      "loss": 0.0861,
+      "step": 1870
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4886254585109774,
+      "learning_rate": 2.6617091256174058e-05,
+      "loss": 0.0805,
+      "step": 1871
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.45498374647935974,
+      "learning_rate": 2.6603662664014644e-05,
+      "loss": 0.08,
+      "step": 1872
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4548434360441476,
+      "learning_rate": 2.659023072970107e-05,
+      "loss": 0.0832,
+      "step": 1873
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4640783853091864,
+      "learning_rate": 2.657679546003133e-05,
+      "loss": 0.0824,
+      "step": 1874
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.44275094182535174,
+      "learning_rate": 2.656335686180509e-05,
+      "loss": 0.0513,
+      "step": 1875
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4445752851132524,
+      "learning_rate": 2.6549914941823713e-05,
+      "loss": 0.0763,
+      "step": 1876
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4123212080066735,
+      "learning_rate": 2.6536469706890226e-05,
+      "loss": 0.084,
+      "step": 1877
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5086670654238233,
+      "learning_rate": 2.652302116380935e-05,
+      "loss": 0.087,
+      "step": 1878
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.48448247640076997,
+      "learning_rate": 2.6509569319387477e-05,
+      "loss": 0.0784,
+      "step": 1879
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4387107172929702,
+      "learning_rate": 2.6496114180432672e-05,
+      "loss": 0.062,
+      "step": 1880
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4821219496508101,
+      "learning_rate": 2.6482655753754657e-05,
+      "loss": 0.0841,
+      "step": 1881
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5209558215749556,
+      "learning_rate": 2.6469194046164818e-05,
+      "loss": 0.0942,
+      "step": 1882
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.40424438725563716,
+      "learning_rate": 2.6455729064476227e-05,
+      "loss": 0.0701,
+      "step": 1883
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4744976367419273,
+      "learning_rate": 2.6442260815503575e-05,
+      "loss": 0.0695,
+      "step": 1884
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.45511068792452036,
+      "learning_rate": 2.6428789306063233e-05,
+      "loss": 0.0689,
+      "step": 1885
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.39648076699405155,
+      "learning_rate": 2.6415314542973214e-05,
+      "loss": 0.0569,
+      "step": 1886
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.39929021953304733,
+      "learning_rate": 2.6401836533053186e-05,
+      "loss": 0.0633,
+      "step": 1887
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5730543722482679,
+      "learning_rate": 2.6388355283124435e-05,
+      "loss": 0.0875,
+      "step": 1888
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.47096256371662787,
+      "learning_rate": 2.637487080000992e-05,
+      "loss": 0.0705,
+      "step": 1889
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4661904390045287,
+      "learning_rate": 2.636138309053421e-05,
+      "loss": 0.0533,
+      "step": 1890
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.503316933883678,
+      "learning_rate": 2.634789216152353e-05,
+      "loss": 0.07,
+      "step": 1891
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5061427615048223,
+      "learning_rate": 2.63343980198057e-05,
+      "loss": 0.0905,
+      "step": 1892
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4891768824481601,
+      "learning_rate": 2.6320900672210216e-05,
+      "loss": 0.0663,
+      "step": 1893
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.46323637785237165,
+      "learning_rate": 2.6307400125568147e-05,
+      "loss": 0.0685,
+      "step": 1894
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5383597506348967,
+      "learning_rate": 2.629389638671221e-05,
+      "loss": 0.0838,
+      "step": 1895
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.47678387299932706,
+      "learning_rate": 2.6280389462476733e-05,
+      "loss": 0.0821,
+      "step": 1896
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.46968283023403673,
+      "learning_rate": 2.6266879359697647e-05,
+      "loss": 0.0678,
+      "step": 1897
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4568098298074498,
+      "learning_rate": 2.6253366085212503e-05,
+      "loss": 0.075,
+      "step": 1898
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4514430333537384,
+      "learning_rate": 2.6239849645860447e-05,
+      "loss": 0.0744,
+      "step": 1899
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.44452307815951564,
+      "learning_rate": 2.6226330048482233e-05,
+      "loss": 0.0718,
+      "step": 1900
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4951100245452337,
+      "learning_rate": 2.6212807299920218e-05,
+      "loss": 0.0713,
+      "step": 1901
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4672619725199249,
+      "learning_rate": 2.6199281407018338e-05,
+      "loss": 0.0758,
+      "step": 1902
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.35243519444327964,
+      "learning_rate": 2.618575237662214e-05,
+      "loss": 0.0512,
+      "step": 1903
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.40736542307882356,
+      "learning_rate": 2.6172220215578743e-05,
+      "loss": 0.0544,
+      "step": 1904
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4704904194392552,
+      "learning_rate": 2.615868493073686e-05,
+      "loss": 0.0806,
+      "step": 1905
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.36351862709614324,
+      "learning_rate": 2.614514652894678e-05,
+      "loss": 0.0582,
+      "step": 1906
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4428274181493983,
+      "learning_rate": 2.613160501706037e-05,
+      "loss": 0.062,
+      "step": 1907
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.37773973289563717,
+      "learning_rate": 2.6118060401931073e-05,
+      "loss": 0.057,
+      "step": 1908
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4885177039781334,
+      "learning_rate": 2.6104512690413906e-05,
+      "loss": 0.0975,
+      "step": 1909
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.49160308049558293,
+      "learning_rate": 2.609096188936544e-05,
+      "loss": 0.0669,
+      "step": 1910
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4757599091402448,
+      "learning_rate": 2.607740800564383e-05,
+      "loss": 0.0663,
+      "step": 1911
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4879069304967567,
+      "learning_rate": 2.6063851046108766e-05,
+      "loss": 0.0749,
+      "step": 1912
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6329272601773189,
+      "learning_rate": 2.605029101762152e-05,
+      "loss": 0.1222,
+      "step": 1913
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4395667713293501,
+      "learning_rate": 2.6036727927044897e-05,
+      "loss": 0.0693,
+      "step": 1914
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.42974012226050445,
+      "learning_rate": 2.602316178124327e-05,
+      "loss": 0.0624,
+      "step": 1915
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5026687685967235,
+      "learning_rate": 2.6009592587082538e-05,
+      "loss": 0.0715,
+      "step": 1916
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4787859406190129,
+      "learning_rate": 2.5996020351430163e-05,
+      "loss": 0.0772,
+      "step": 1917
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.49640720707144015,
+      "learning_rate": 2.598244508115513e-05,
+      "loss": 0.0869,
+      "step": 1918
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.44704691774444677,
+      "learning_rate": 2.596886678312797e-05,
+      "loss": 0.0717,
+      "step": 1919
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.566518259483125,
+      "learning_rate": 2.5955285464220738e-05,
+      "loss": 0.1036,
+      "step": 1920
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.49886906395355324,
+      "learning_rate": 2.594170113130703e-05,
+      "loss": 0.0717,
+      "step": 1921
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3631885369239595,
+      "learning_rate": 2.5928113791261952e-05,
+      "loss": 0.0446,
+      "step": 1922
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.42634171162088436,
+      "learning_rate": 2.5914523450962147e-05,
+      "loss": 0.0643,
+      "step": 1923
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.49661806889154125,
+      "learning_rate": 2.590093011728577e-05,
+      "loss": 0.0865,
+      "step": 1924
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.37300980468627076,
+      "learning_rate": 2.588733379711248e-05,
+      "loss": 0.0465,
+      "step": 1925
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.41554669022488777,
+      "learning_rate": 2.587373449732347e-05,
+      "loss": 0.051,
+      "step": 1926
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.45649580278378604,
+      "learning_rate": 2.5860132224801424e-05,
+      "loss": 0.0712,
+      "step": 1927
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.40900949293536104,
+      "learning_rate": 2.584652698643054e-05,
+      "loss": 0.0591,
+      "step": 1928
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3939289367631103,
+      "learning_rate": 2.58329187890965e-05,
+      "loss": 0.0472,
+      "step": 1929
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.46501062106749136,
+      "learning_rate": 2.581930763968651e-05,
+      "loss": 0.0741,
+      "step": 1930
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4760442792819372,
+      "learning_rate": 2.580569354508925e-05,
+      "loss": 0.0687,
+      "step": 1931
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4572477561474321,
+      "learning_rate": 2.5792076512194895e-05,
+      "loss": 0.075,
+      "step": 1932
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5661992714329125,
+      "learning_rate": 2.5778456547895117e-05,
+      "loss": 0.0905,
+      "step": 1933
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4468749226389998,
+      "learning_rate": 2.5764833659083053e-05,
+      "loss": 0.0656,
+      "step": 1934
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5459863800019874,
+      "learning_rate": 2.5751207852653334e-05,
+      "loss": 0.1169,
+      "step": 1935
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4620661092127289,
+      "learning_rate": 2.5737579135502068e-05,
+      "loss": 0.0691,
+      "step": 1936
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.47962413766985756,
+      "learning_rate": 2.572394751452683e-05,
+      "loss": 0.0738,
+      "step": 1937
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4390047116787088,
+      "learning_rate": 2.5710312996626667e-05,
+      "loss": 0.0751,
+      "step": 1938
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4369903218236427,
+      "learning_rate": 2.569667558870209e-05,
+      "loss": 0.0652,
+      "step": 1939
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.45743852384713024,
+      "learning_rate": 2.5683035297655076e-05,
+      "loss": 0.075,
+      "step": 1940
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4280689098342205,
+      "learning_rate": 2.566939213038906e-05,
+      "loss": 0.0711,
+      "step": 1941
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.46608190968404795,
+      "learning_rate": 2.5655746093808934e-05,
+      "loss": 0.0796,
+      "step": 1942
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.440689855902259,
+      "learning_rate": 2.564209719482104e-05,
+      "loss": 0.0509,
+      "step": 1943
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.47352029309397264,
+      "learning_rate": 2.5628445440333164e-05,
+      "loss": 0.0917,
+      "step": 1944
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.44388221799202127,
+      "learning_rate": 2.5614790837254555e-05,
+      "loss": 0.0565,
+      "step": 1945
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4410939757738066,
+      "learning_rate": 2.5601133392495886e-05,
+      "loss": 0.0636,
+      "step": 1946
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4855035279152683,
+      "learning_rate": 2.558747311296926e-05,
+      "loss": 0.0593,
+      "step": 1947
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4578283320218462,
+      "learning_rate": 2.5573810005588245e-05,
+      "loss": 0.0786,
+      "step": 1948
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.49591954995008763,
+      "learning_rate": 2.5560144077267826e-05,
+      "loss": 0.0731,
+      "step": 1949
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4308347471916173,
+      "learning_rate": 2.5546475334924398e-05,
+      "loss": 0.0584,
+      "step": 1950
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4204925562690745,
+      "learning_rate": 2.5532803785475802e-05,
+      "loss": 0.0656,
+      "step": 1951
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.49925856453889894,
+      "learning_rate": 2.5519129435841298e-05,
+      "loss": 0.077,
+      "step": 1952
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4948463978529427,
+      "learning_rate": 2.550545229294155e-05,
+      "loss": 0.0663,
+      "step": 1953
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4999508982628857,
+      "learning_rate": 2.549177236369865e-05,
+      "loss": 0.0838,
+      "step": 1954
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.45857363239203563,
+      "learning_rate": 2.5478089655036086e-05,
+      "loss": 0.061,
+      "step": 1955
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.35649923674406764,
+      "learning_rate": 2.5464404173878775e-05,
+      "loss": 0.0626,
+      "step": 1956
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4048285926917149,
+      "learning_rate": 2.5450715927153012e-05,
+      "loss": 0.0536,
+      "step": 1957
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.38869547173262164,
+      "learning_rate": 2.54370249217865e-05,
+      "loss": 0.0501,
+      "step": 1958
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5734012519543342,
+      "learning_rate": 2.542333116470835e-05,
+      "loss": 0.0768,
+      "step": 1959
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4480250545333302,
+      "learning_rate": 2.5409634662849053e-05,
+      "loss": 0.0644,
+      "step": 1960
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.47188119947390483,
+      "learning_rate": 2.5395935423140487e-05,
+      "loss": 0.0771,
+      "step": 1961
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4498040408396079,
+      "learning_rate": 2.5382233452515927e-05,
+      "loss": 0.0795,
+      "step": 1962
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5292149441641056,
+      "learning_rate": 2.5368528757910027e-05,
+      "loss": 0.0749,
+      "step": 1963
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.44870011618221406,
+      "learning_rate": 2.5354821346258813e-05,
+      "loss": 0.0622,
+      "step": 1964
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4956262894189136,
+      "learning_rate": 2.534111122449969e-05,
+      "loss": 0.0914,
+      "step": 1965
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5129120286948241,
+      "learning_rate": 2.532739839957143e-05,
+      "loss": 0.0824,
+      "step": 1966
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4525574549928641,
+      "learning_rate": 2.5313682878414185e-05,
+      "loss": 0.0588,
+      "step": 1967
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.49355155904583753,
+      "learning_rate": 2.529996466796946e-05,
+      "loss": 0.0858,
+      "step": 1968
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.47739849869196777,
+      "learning_rate": 2.5286243775180128e-05,
+      "loss": 0.0738,
+      "step": 1969
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3867738087309617,
+      "learning_rate": 2.5272520206990418e-05,
+      "loss": 0.0636,
+      "step": 1970
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.45189343176114144,
+      "learning_rate": 2.5258793970345908e-05,
+      "loss": 0.0667,
+      "step": 1971
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.45433940606363016,
+      "learning_rate": 2.5245065072193534e-05,
+      "loss": 0.069,
+      "step": 1972
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.519966287528791,
+      "learning_rate": 2.5231333519481577e-05,
+      "loss": 0.0744,
+      "step": 1973
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4226544385437567,
+      "learning_rate": 2.5217599319159654e-05,
+      "loss": 0.0551,
+      "step": 1974
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3707658825119907,
+      "learning_rate": 2.5203862478178732e-05,
+      "loss": 0.0505,
+      "step": 1975
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5165477933826751,
+      "learning_rate": 2.519012300349111e-05,
+      "loss": 0.0848,
+      "step": 1976
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4305647132097901,
+      "learning_rate": 2.5176380902050418e-05,
+      "loss": 0.0624,
+      "step": 1977
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4972505673558224,
+      "learning_rate": 2.516263618081162e-05,
+      "loss": 0.0629,
+      "step": 1978
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.46207126902088547,
+      "learning_rate": 2.5148888846731007e-05,
+      "loss": 0.0777,
+      "step": 1979
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.44448269945982694,
+      "learning_rate": 2.5135138906766185e-05,
+      "loss": 0.0675,
+      "step": 1980
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4387582668384212,
+      "learning_rate": 2.512138636787608e-05,
+      "loss": 0.0724,
+      "step": 1981
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.48427690004796514,
+      "learning_rate": 2.510763123702094e-05,
+      "loss": 0.0745,
+      "step": 1982
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.401680074912499,
+      "learning_rate": 2.5093873521162323e-05,
+      "loss": 0.05,
+      "step": 1983
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.48928307565120804,
+      "learning_rate": 2.5080113227263093e-05,
+      "loss": 0.0863,
+      "step": 1984
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3701655575188742,
+      "learning_rate": 2.5066350362287407e-05,
+      "loss": 0.0556,
+      "step": 1985
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.37219148199736224,
+      "learning_rate": 2.5052584933200756e-05,
+      "loss": 0.0562,
+      "step": 1986
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.522504487406247,
+      "learning_rate": 2.5038816946969894e-05,
+      "loss": 0.0867,
+      "step": 1987
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4148666203116047,
+      "learning_rate": 2.5025046410562888e-05,
+      "loss": 0.076,
+      "step": 1988
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4241673390970852,
+      "learning_rate": 2.501127333094909e-05,
+      "loss": 0.058,
+      "step": 1989
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.524471741528751,
+      "learning_rate": 2.4997497715099134e-05,
+      "loss": 0.0695,
+      "step": 1990
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.48102112938397373,
+      "learning_rate": 2.4983719569984955e-05,
+      "loss": 0.0668,
+      "step": 1991
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3793570795807171,
+      "learning_rate": 2.496993890257975e-05,
+      "loss": 0.0638,
+      "step": 1992
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3804833732473676,
+      "learning_rate": 2.4956155719858e-05,
+      "loss": 0.0473,
+      "step": 1993
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4429668419744718,
+      "learning_rate": 2.4942370028795456e-05,
+      "loss": 0.0663,
+      "step": 1994
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5238030139613532,
+      "learning_rate": 2.4928581836369147e-05,
+      "loss": 0.089,
+      "step": 1995
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4618954483378638,
+      "learning_rate": 2.4914791149557358e-05,
+      "loss": 0.0645,
+      "step": 1996
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.40761575008590367,
+      "learning_rate": 2.490099797533964e-05,
+      "loss": 0.0684,
+      "step": 1997
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.39386441748205986,
+      "learning_rate": 2.48872023206968e-05,
+      "loss": 0.0582,
+      "step": 1998
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.580610602439528,
+      "learning_rate": 2.487340419261091e-05,
+      "loss": 0.1076,
+      "step": 1999
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.41888062660197195,
+      "learning_rate": 2.485960359806528e-05,
+      "loss": 0.0584,
+      "step": 2000
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4820066510667588,
+      "learning_rate": 2.4845800544044483e-05,
+      "loss": 0.0792,
+      "step": 2001
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4113433733628816,
+      "learning_rate": 2.4831995037534325e-05,
+      "loss": 0.0639,
+      "step": 2002
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4688180556167094,
+      "learning_rate": 2.481818708552185e-05,
+      "loss": 0.0657,
+      "step": 2003
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.557453565759295,
+      "learning_rate": 2.480437669499537e-05,
+      "loss": 0.0744,
+      "step": 2004
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.45971494928556966,
+      "learning_rate": 2.479056387294438e-05,
+      "loss": 0.0745,
+      "step": 2005
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3851151984198529,
+      "learning_rate": 2.4776748626359656e-05,
+      "loss": 0.0543,
+      "step": 2006
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5287876565298385,
+      "learning_rate": 2.4762930962233164e-05,
+      "loss": 0.0922,
+      "step": 2007
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5038474884074328,
+      "learning_rate": 2.4749110887558114e-05,
+      "loss": 0.0628,
+      "step": 2008
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.43435565919150204,
+      "learning_rate": 2.4735288409328937e-05,
+      "loss": 0.0619,
+      "step": 2009
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4305378735493745,
+      "learning_rate": 2.472146353454127e-05,
+      "loss": 0.0666,
+      "step": 2010
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.42052916150574776,
+      "learning_rate": 2.4707636270191956e-05,
+      "loss": 0.0669,
+      "step": 2011
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4320426553114052,
+      "learning_rate": 2.4693806623279074e-05,
+      "loss": 0.0643,
+      "step": 2012
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.45603134128190864,
+      "learning_rate": 2.4679974600801882e-05,
+      "loss": 0.0756,
+      "step": 2013
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5432398825807966,
+      "learning_rate": 2.4666140209760862e-05,
+      "loss": 0.1047,
+      "step": 2014
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3501122018088189,
+      "learning_rate": 2.4652303457157677e-05,
+      "loss": 0.0458,
+      "step": 2015
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4849934014382745,
+      "learning_rate": 2.4638464349995186e-05,
+      "loss": 0.0739,
+      "step": 2016
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4127435189367342,
+      "learning_rate": 2.4624622895277462e-05,
+      "loss": 0.0532,
+      "step": 2017
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3852790853300477,
+      "learning_rate": 2.461077910000974e-05,
+      "loss": 0.0526,
+      "step": 2018
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.39799662114630846,
+      "learning_rate": 2.4596932971198446e-05,
+      "loss": 0.0678,
+      "step": 2019
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.42710675392442127,
+      "learning_rate": 2.4583084515851194e-05,
+      "loss": 0.0618,
+      "step": 2020
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4730950517637139,
+      "learning_rate": 2.456923374097678e-05,
+      "loss": 0.0679,
+      "step": 2021
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3588081118074976,
+      "learning_rate": 2.4555380653585158e-05,
+      "loss": 0.0578,
+      "step": 2022
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4513238923236032,
+      "learning_rate": 2.4541525260687468e-05,
+      "loss": 0.0626,
+      "step": 2023
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5074836840144351,
+      "learning_rate": 2.4527667569295996e-05,
+      "loss": 0.0722,
+      "step": 2024
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.45947534801574375,
+      "learning_rate": 2.4513807586424214e-05,
+      "loss": 0.0532,
+      "step": 2025
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.41736844885552515,
+      "learning_rate": 2.449994531908675e-05,
+      "loss": 0.0625,
+      "step": 2026
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4226561308347636,
+      "learning_rate": 2.4486080774299364e-05,
+      "loss": 0.0602,
+      "step": 2027
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.44196260886816885,
+      "learning_rate": 2.4472213959079002e-05,
+      "loss": 0.0608,
+      "step": 2028
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.48246632112555954,
+      "learning_rate": 2.4458344880443735e-05,
+      "loss": 0.0785,
+      "step": 2029
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.43183535502932363,
+      "learning_rate": 2.4444473545412804e-05,
+      "loss": 0.0519,
+      "step": 2030
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.47867407219085056,
+      "learning_rate": 2.4430599961006563e-05,
+      "loss": 0.0679,
+      "step": 2031
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.385196541309476,
+      "learning_rate": 2.441672413424652e-05,
+      "loss": 0.0545,
+      "step": 2032
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.38971940926330956,
+      "learning_rate": 2.4402846072155313e-05,
+      "loss": 0.0598,
+      "step": 2033
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4747944694621555,
+      "learning_rate": 2.4388965781756727e-05,
+      "loss": 0.0663,
+      "step": 2034
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5031381711953609,
+      "learning_rate": 2.437508327007565e-05,
+      "loss": 0.0781,
+      "step": 2035
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3809346246962901,
+      "learning_rate": 2.4361198544138117e-05,
+      "loss": 0.0534,
+      "step": 2036
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.34526724195138403,
+      "learning_rate": 2.4347311610971255e-05,
+      "loss": 0.0496,
+      "step": 2037
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4932194744743591,
+      "learning_rate": 2.4333422477603342e-05,
+      "loss": 0.0835,
+      "step": 2038
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.48806754967168864,
+      "learning_rate": 2.4319531151063753e-05,
+      "loss": 0.0631,
+      "step": 2039
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3881418457614026,
+      "learning_rate": 2.4305637638382967e-05,
+      "loss": 0.0453,
+      "step": 2040
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3958838664809399,
+      "learning_rate": 2.4291741946592575e-05,
+      "loss": 0.0618,
+      "step": 2041
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4232946985585955,
+      "learning_rate": 2.427784408272528e-05,
+      "loss": 0.0558,
+      "step": 2042
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.45405005706414564,
+      "learning_rate": 2.4263944053814866e-05,
+      "loss": 0.072,
+      "step": 2043
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.41057790245168874,
+      "learning_rate": 2.4250041866896234e-05,
+      "loss": 0.0633,
+      "step": 2044
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.49894964639299666,
+      "learning_rate": 2.4236137529005355e-05,
+      "loss": 0.0733,
+      "step": 2045
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4828462008717174,
+      "learning_rate": 2.4222231047179303e-05,
+      "loss": 0.0645,
+      "step": 2046
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4851688985358276,
+      "learning_rate": 2.420832242845624e-05,
+      "loss": 0.0518,
+      "step": 2047
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.47643881896966234,
+      "learning_rate": 2.41944116798754e-05,
+      "loss": 0.0719,
+      "step": 2048
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.46149104227028975,
+      "learning_rate": 2.4180498808477096e-05,
+      "loss": 0.0696,
+      "step": 2049
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4211329716125816,
+      "learning_rate": 2.4166583821302712e-05,
+      "loss": 0.0714,
+      "step": 2050
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3851136120143572,
+      "learning_rate": 2.4152666725394717e-05,
+      "loss": 0.0594,
+      "step": 2051
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4271611243632751,
+      "learning_rate": 2.413874752779664e-05,
+      "loss": 0.0637,
+      "step": 2052
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4608236592549507,
+      "learning_rate": 2.412482623555307e-05,
+      "loss": 0.0786,
+      "step": 2053
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4462314293039966,
+      "learning_rate": 2.411090285570965e-05,
+      "loss": 0.0636,
+      "step": 2054
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.41341095170918657,
+      "learning_rate": 2.4096977395313096e-05,
+      "loss": 0.0691,
+      "step": 2055
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3892252639047788,
+      "learning_rate": 2.4083049861411173e-05,
+      "loss": 0.0557,
+      "step": 2056
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.43831609495639573,
+      "learning_rate": 2.4069120261052682e-05,
+      "loss": 0.0713,
+      "step": 2057
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4160579785202587,
+      "learning_rate": 2.4055188601287483e-05,
+      "loss": 0.0649,
+      "step": 2058
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.35123628740593144,
+      "learning_rate": 2.404125488916647e-05,
+      "loss": 0.0519,
+      "step": 2059
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3956505788576386,
+      "learning_rate": 2.402731913174159e-05,
+      "loss": 0.0516,
+      "step": 2060
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5417443335922905,
+      "learning_rate": 2.4013381336065805e-05,
+      "loss": 0.0875,
+      "step": 2061
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4087644851229479,
+      "learning_rate": 2.399944150919313e-05,
+      "loss": 0.0595,
+      "step": 2062
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4386785504995785,
+      "learning_rate": 2.398549965817858e-05,
+      "loss": 0.0634,
+      "step": 2063
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3772173533599839,
+      "learning_rate": 2.3971555790078228e-05,
+      "loss": 0.0626,
+      "step": 2064
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.49008808764987444,
+      "learning_rate": 2.3957609911949146e-05,
+      "loss": 0.0765,
+      "step": 2065
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3784129283938517,
+      "learning_rate": 2.3943662030849426e-05,
+      "loss": 0.0541,
+      "step": 2066
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.42390100230516714,
+      "learning_rate": 2.3929712153838173e-05,
+      "loss": 0.0645,
+      "step": 2067
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5508192333353973,
+      "learning_rate": 2.3915760287975515e-05,
+      "loss": 0.0865,
+      "step": 2068
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.48057344117933043,
+      "learning_rate": 2.390180644032257e-05,
+      "loss": 0.0718,
+      "step": 2069
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.44469013559146137,
+      "learning_rate": 2.3887850617941464e-05,
+      "loss": 0.0731,
+      "step": 2070
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.37449150577991236,
+      "learning_rate": 2.3873892827895332e-05,
+      "loss": 0.0592,
+      "step": 2071
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3453121668611188,
+      "learning_rate": 2.3859933077248285e-05,
+      "loss": 0.0572,
+      "step": 2072
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3831193628254262,
+      "learning_rate": 2.3845971373065452e-05,
+      "loss": 0.0448,
+      "step": 2073
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4014324311553023,
+      "learning_rate": 2.3832007722412934e-05,
+      "loss": 0.066,
+      "step": 2074
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4756276701803358,
+      "learning_rate": 2.3818042132357812e-05,
+      "loss": 0.0615,
+      "step": 2075
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.46100292711300944,
+      "learning_rate": 2.3804074609968158e-05,
+      "loss": 0.0621,
+      "step": 2076
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.40052272199260675,
+      "learning_rate": 2.3790105162313032e-05,
+      "loss": 0.0569,
+      "step": 2077
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.47979434684151695,
+      "learning_rate": 2.3776133796462446e-05,
+      "loss": 0.0566,
+      "step": 2078
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.503394081000372,
+      "learning_rate": 2.3762160519487402e-05,
+      "loss": 0.089,
+      "step": 2079
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.47883710377584393,
+      "learning_rate": 2.3748185338459847e-05,
+      "loss": 0.0656,
+      "step": 2080
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4445039536286361,
+      "learning_rate": 2.3734208260452727e-05,
+      "loss": 0.0679,
+      "step": 2081
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6269723526807839,
+      "learning_rate": 2.372022929253991e-05,
+      "loss": 0.0928,
+      "step": 2082
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.45459800573575326,
+      "learning_rate": 2.3706248441796246e-05,
+      "loss": 0.0678,
+      "step": 2083
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.48578360533979176,
+      "learning_rate": 2.369226571529752e-05,
+      "loss": 0.0953,
+      "step": 2084
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3936335860836555,
+      "learning_rate": 2.3678281120120485e-05,
+      "loss": 0.0661,
+      "step": 2085
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3621791611864308,
+      "learning_rate": 2.366429466334283e-05,
+      "loss": 0.0416,
+      "step": 2086
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4244046662682168,
+      "learning_rate": 2.3650306352043182e-05,
+      "loss": 0.0617,
+      "step": 2087
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3979055930722182,
+      "learning_rate": 2.3636316193301107e-05,
+      "loss": 0.047,
+      "step": 2088
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5196930482789566,
+      "learning_rate": 2.3622324194197118e-05,
+      "loss": 0.0734,
+      "step": 2089
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3873937091870021,
+      "learning_rate": 2.3608330361812652e-05,
+      "loss": 0.0708,
+      "step": 2090
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3548939847473483,
+      "learning_rate": 2.3594334703230065e-05,
+      "loss": 0.0524,
+      "step": 2091
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3971874411438089,
+      "learning_rate": 2.3580337225532663e-05,
+      "loss": 0.0568,
+      "step": 2092
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4571903636475039,
+      "learning_rate": 2.356633793580463e-05,
+      "loss": 0.0694,
+      "step": 2093
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.38851891196183885,
+      "learning_rate": 2.355233684113111e-05,
+      "loss": 0.0735,
+      "step": 2094
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4311069970136712,
+      "learning_rate": 2.3538333948598142e-05,
+      "loss": 0.0633,
+      "step": 2095
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5541316457548243,
+      "learning_rate": 2.3524329265292668e-05,
+      "loss": 0.0989,
+      "step": 2096
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5119881248531218,
+      "learning_rate": 2.3510322798302553e-05,
+      "loss": 0.0733,
+      "step": 2097
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.35231347937292995,
+      "learning_rate": 2.3496314554716543e-05,
+      "loss": 0.0499,
+      "step": 2098
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4671012811618288,
+      "learning_rate": 2.348230454162431e-05,
+      "loss": 0.0646,
+      "step": 2099
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.41502455742116545,
+      "learning_rate": 2.34682927661164e-05,
+      "loss": 0.0609,
+      "step": 2100
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.36580421750435826,
+      "learning_rate": 2.3454279235284264e-05,
+      "loss": 0.0516,
+      "step": 2101
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3288414490352913,
+      "learning_rate": 2.344026395622023e-05,
+      "loss": 0.0411,
+      "step": 2102
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5474729266505735,
+      "learning_rate": 2.3426246936017514e-05,
+      "loss": 0.0827,
+      "step": 2103
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3939901361416779,
+      "learning_rate": 2.3412228181770224e-05,
+      "loss": 0.0615,
+      "step": 2104
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4041682605449802,
+      "learning_rate": 2.3398207700573336e-05,
+      "loss": 0.066,
+      "step": 2105
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.45185146919493696,
+      "learning_rate": 2.3384185499522696e-05,
+      "loss": 0.0771,
+      "step": 2106
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.39996471547427803,
+      "learning_rate": 2.337016158571503e-05,
+      "loss": 0.0537,
+      "step": 2107
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.441245767232677,
+      "learning_rate": 2.335613596624793e-05,
+      "loss": 0.0729,
+      "step": 2108
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4054840643605466,
+      "learning_rate": 2.334210864821984e-05,
+      "loss": 0.0524,
+      "step": 2109
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.48563810981326716,
+      "learning_rate": 2.3328079638730073e-05,
+      "loss": 0.0801,
+      "step": 2110
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4039880983221286,
+      "learning_rate": 2.3314048944878804e-05,
+      "loss": 0.0548,
+      "step": 2111
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.48363093170054183,
+      "learning_rate": 2.330001657376705e-05,
+      "loss": 0.0823,
+      "step": 2112
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3865247490288611,
+      "learning_rate": 2.3285982532496676e-05,
+      "loss": 0.0684,
+      "step": 2113
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.37701545951017845,
+      "learning_rate": 2.32719468281704e-05,
+      "loss": 0.0563,
+      "step": 2114
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4805666350852702,
+      "learning_rate": 2.325790946789178e-05,
+      "loss": 0.0873,
+      "step": 2115
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.394559689242731,
+      "learning_rate": 2.32438704587652e-05,
+      "loss": 0.0542,
+      "step": 2116
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3775259419236919,
+      "learning_rate": 2.3229829807895904e-05,
+      "loss": 0.0574,
+      "step": 2117
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.39554686176948745,
+      "learning_rate": 2.3215787522389935e-05,
+      "loss": 0.0521,
+      "step": 2118
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.41370662654906654,
+      "learning_rate": 2.3201743609354187e-05,
+      "loss": 0.0639,
+      "step": 2119
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.34958462121375417,
+      "learning_rate": 2.3187698075896378e-05,
+      "loss": 0.0453,
+      "step": 2120
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4804601591008989,
+      "learning_rate": 2.317365092912503e-05,
+      "loss": 0.0712,
+      "step": 2121
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4395080871120076,
+      "learning_rate": 2.3159602176149493e-05,
+      "loss": 0.078,
+      "step": 2122
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.49353773105297966,
+      "learning_rate": 2.314555182407992e-05,
+      "loss": 0.0694,
+      "step": 2123
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.41822454165715695,
+      "learning_rate": 2.3131499880027294e-05,
+      "loss": 0.0651,
+      "step": 2124
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4385989215954988,
+      "learning_rate": 2.311744635110338e-05,
+      "loss": 0.0643,
+      "step": 2125
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.45052014793681966,
+      "learning_rate": 2.3103391244420754e-05,
+      "loss": 0.0683,
+      "step": 2126
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.454793816685309,
+      "learning_rate": 2.30893345670928e-05,
+      "loss": 0.0699,
+      "step": 2127
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.44657509817367813,
+      "learning_rate": 2.3075276326233676e-05,
+      "loss": 0.0803,
+      "step": 2128
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.40947483315993416,
+      "learning_rate": 2.306121652895836e-05,
+      "loss": 0.062,
+      "step": 2129
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.38736266655873886,
+      "learning_rate": 2.3047155182382584e-05,
+      "loss": 0.0543,
+      "step": 2130
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3548728643355111,
+      "learning_rate": 2.3033092293622903e-05,
+      "loss": 0.0476,
+      "step": 2131
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4771604694937133,
+      "learning_rate": 2.3019027869796607e-05,
+      "loss": 0.0645,
+      "step": 2132
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4048655468083591,
+      "learning_rate": 2.3004961918021804e-05,
+      "loss": 0.0533,
+      "step": 2133
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3945217886819489,
+      "learning_rate": 2.2990894445417355e-05,
+      "loss": 0.0698,
+      "step": 2134
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.44724783383095956,
+      "learning_rate": 2.2976825459102898e-05,
+      "loss": 0.0602,
+      "step": 2135
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4704754068846899,
+      "learning_rate": 2.2962754966198815e-05,
+      "loss": 0.0769,
+      "step": 2136
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4445963002428909,
+      "learning_rate": 2.2948682973826292e-05,
+      "loss": 0.072,
+      "step": 2137
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.38627250047225226,
+      "learning_rate": 2.2934609489107236e-05,
+      "loss": 0.0499,
+      "step": 2138
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5143966493187458,
+      "learning_rate": 2.292053451916433e-05,
+      "loss": 0.0722,
+      "step": 2139
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.42852994882712525,
+      "learning_rate": 2.2906458071121e-05,
+      "loss": 0.0503,
+      "step": 2140
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3497800144203384,
+      "learning_rate": 2.289238015210142e-05,
+      "loss": 0.0421,
+      "step": 2141
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4088490317765311,
+      "learning_rate": 2.2878300769230522e-05,
+      "loss": 0.0426,
+      "step": 2142
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.45912554578699355,
+      "learning_rate": 2.2864219929633956e-05,
+      "loss": 0.0635,
+      "step": 2143
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4630192606083607,
+      "learning_rate": 2.2850137640438126e-05,
+      "loss": 0.0781,
+      "step": 2144
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.45750754553403933,
+      "learning_rate": 2.2836053908770165e-05,
+      "loss": 0.0743,
+      "step": 2145
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.46505834931196594,
+      "learning_rate": 2.2821968741757935e-05,
+      "loss": 0.0754,
+      "step": 2146
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.38697667623807835,
+      "learning_rate": 2.280788214653003e-05,
+      "loss": 0.0662,
+      "step": 2147
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.376535846827464,
+      "learning_rate": 2.2793794130215753e-05,
+      "loss": 0.0484,
+      "step": 2148
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.41352409770681026,
+      "learning_rate": 2.2779704699945136e-05,
+      "loss": 0.0645,
+      "step": 2149
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3455735087539365,
+      "learning_rate": 2.2765613862848936e-05,
+      "loss": 0.0504,
+      "step": 2150
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.42004527563202043,
+      "learning_rate": 2.2751521626058607e-05,
+      "loss": 0.0643,
+      "step": 2151
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.40863652926214156,
+      "learning_rate": 2.2737427996706316e-05,
+      "loss": 0.0544,
+      "step": 2152
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3377745875628734,
+      "learning_rate": 2.2723332981924937e-05,
+      "loss": 0.0459,
+      "step": 2153
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.39150200622356357,
+      "learning_rate": 2.2709236588848036e-05,
+      "loss": 0.0535,
+      "step": 2154
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5176623006064641,
+      "learning_rate": 2.269513882460989e-05,
+      "loss": 0.0888,
+      "step": 2155
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.40359070514234835,
+      "learning_rate": 2.268103969634547e-05,
+      "loss": 0.0482,
+      "step": 2156
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.45307968176111113,
+      "learning_rate": 2.266693921119042e-05,
+      "loss": 0.0661,
+      "step": 2157
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6422601162212864,
+      "learning_rate": 2.2652837376281087e-05,
+      "loss": 0.0816,
+      "step": 2158
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3880241018409379,
+      "learning_rate": 2.2638734198754496e-05,
+      "loss": 0.0642,
+      "step": 2159
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3337084539500318,
+      "learning_rate": 2.2624629685748353e-05,
+      "loss": 0.0465,
+      "step": 2160
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.37604125816273715,
+      "learning_rate": 2.261052384440104e-05,
+      "loss": 0.0629,
+      "step": 2161
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.41380834876811373,
+      "learning_rate": 2.2596416681851595e-05,
+      "loss": 0.0567,
+      "step": 2162
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4765554959142064,
+      "learning_rate": 2.2582308205239757e-05,
+      "loss": 0.0697,
+      "step": 2163
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.34652462567446174,
+      "learning_rate": 2.256819842170591e-05,
+      "loss": 0.0428,
+      "step": 2164
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3953894167167263,
+      "learning_rate": 2.2554087338391098e-05,
+      "loss": 0.0622,
+      "step": 2165
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.47163784244279544,
+      "learning_rate": 2.2539974962437022e-05,
+      "loss": 0.0849,
+      "step": 2166
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4278849506541713,
+      "learning_rate": 2.252586130098605e-05,
+      "loss": 0.0663,
+      "step": 2167
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.49671398648944554,
+      "learning_rate": 2.251174636118119e-05,
+      "loss": 0.0574,
+      "step": 2168
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.40812914237217485,
+      "learning_rate": 2.2497630150166102e-05,
+      "loss": 0.0607,
+      "step": 2169
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4092688451970953,
+      "learning_rate": 2.2483512675085085e-05,
+      "loss": 0.0657,
+      "step": 2170
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.41064415023368733,
+      "learning_rate": 2.2469393943083068e-05,
+      "loss": 0.0525,
+      "step": 2171
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4334403445391836,
+      "learning_rate": 2.245527396130565e-05,
+      "loss": 0.062,
+      "step": 2172
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4424358614004642,
+      "learning_rate": 2.2441152736899026e-05,
+      "loss": 0.0805,
+      "step": 2173
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4298174228183831,
+      "learning_rate": 2.242703027701004e-05,
+      "loss": 0.0587,
+      "step": 2174
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3311299151354249,
+      "learning_rate": 2.2412906588786147e-05,
+      "loss": 0.0548,
+      "step": 2175
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5231035220386947,
+      "learning_rate": 2.2398781679375445e-05,
+      "loss": 0.0883,
+      "step": 2176
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.432308870762196,
+      "learning_rate": 2.2384655555926625e-05,
+      "loss": 0.0757,
+      "step": 2177
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4268838796627569,
+      "learning_rate": 2.237052822558901e-05,
+      "loss": 0.0714,
+      "step": 2178
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4251954835217631,
+      "learning_rate": 2.235639969551253e-05,
+      "loss": 0.0518,
+      "step": 2179
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.38094984104985813,
+      "learning_rate": 2.2342269972847718e-05,
+      "loss": 0.0544,
+      "step": 2180
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4359203499322573,
+      "learning_rate": 2.232813906474572e-05,
+      "loss": 0.0518,
+      "step": 2181
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.42064737147089587,
+      "learning_rate": 2.2314006978358263e-05,
+      "loss": 0.0462,
+      "step": 2182
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.426380243595148,
+      "learning_rate": 2.2299873720837692e-05,
+      "loss": 0.0452,
+      "step": 2183
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3966205907889756,
+      "learning_rate": 2.2285739299336933e-05,
+      "loss": 0.0465,
+      "step": 2184
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3828731994773382,
+      "learning_rate": 2.22716037210095e-05,
+      "loss": 0.0639,
+      "step": 2185
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.39471942832875806,
+      "learning_rate": 2.2257466993009503e-05,
+      "loss": 0.0642,
+      "step": 2186
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4215792042821033,
+      "learning_rate": 2.2243329122491617e-05,
+      "loss": 0.0574,
+      "step": 2187
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.49431560268048147,
+      "learning_rate": 2.222919011661111e-05,
+      "loss": 0.077,
+      "step": 2188
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4424619456539286,
+      "learning_rate": 2.2215049982523827e-05,
+      "loss": 0.0567,
+      "step": 2189
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4930441228976846,
+      "learning_rate": 2.2200908727386167e-05,
+      "loss": 0.0829,
+      "step": 2190
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.43646680221292344,
+      "learning_rate": 2.2186766358355106e-05,
+      "loss": 0.0687,
+      "step": 2191
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3875802010940222,
+      "learning_rate": 2.217262288258818e-05,
+      "loss": 0.0563,
+      "step": 2192
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4571296553349933,
+      "learning_rate": 2.2158478307243507e-05,
+      "loss": 0.0694,
+      "step": 2193
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.40632144593168007,
+      "learning_rate": 2.2144332639479722e-05,
+      "loss": 0.0535,
+      "step": 2194
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4048639641812654,
+      "learning_rate": 2.213018588645605e-05,
+      "loss": 0.0625,
+      "step": 2195
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4720688945731608,
+      "learning_rate": 2.2116038055332238e-05,
+      "loss": 0.0681,
+      "step": 2196
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4141419892215765,
+      "learning_rate": 2.2101889153268595e-05,
+      "loss": 0.0575,
+      "step": 2197
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.38017011040985954,
+      "learning_rate": 2.2087739187425967e-05,
+      "loss": 0.0561,
+      "step": 2198
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.42091886466391487,
+      "learning_rate": 2.2073588164965737e-05,
+      "loss": 0.0536,
+      "step": 2199
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.395803308184746,
+      "learning_rate": 2.205943609304983e-05,
+      "loss": 0.0507,
+      "step": 2200
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.41410210258103936,
+      "learning_rate": 2.2045282978840684e-05,
+      "loss": 0.0505,
+      "step": 2201
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.31556133204089304,
+      "learning_rate": 2.2031128829501293e-05,
+      "loss": 0.0407,
+      "step": 2202
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4462056154997179,
+      "learning_rate": 2.2016973652195145e-05,
+      "loss": 0.0724,
+      "step": 2203
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4439268303853345,
+      "learning_rate": 2.200281745408627e-05,
+      "loss": 0.0596,
+      "step": 2204
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4004929016486724,
+      "learning_rate": 2.1988660242339205e-05,
+      "loss": 0.0568,
+      "step": 2205
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3522127284277402,
+      "learning_rate": 2.1974502024119002e-05,
+      "loss": 0.0456,
+      "step": 2206
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5424500327237654,
+      "learning_rate": 2.196034280659122e-05,
+      "loss": 0.0967,
+      "step": 2207
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3807729378451628,
+      "learning_rate": 2.1946182596921917e-05,
+      "loss": 0.0496,
+      "step": 2208
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.45126173437400224,
+      "learning_rate": 2.1932021402277682e-05,
+      "loss": 0.0748,
+      "step": 2209
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4133062834630545,
+      "learning_rate": 2.1917859229825565e-05,
+      "loss": 0.0555,
+      "step": 2210
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.41212271718488375,
+      "learning_rate": 2.1903696086733142e-05,
+      "loss": 0.0635,
+      "step": 2211
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.36619925925389807,
+      "learning_rate": 2.188953198016846e-05,
+      "loss": 0.0457,
+      "step": 2212
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3598676704542792,
+      "learning_rate": 2.1875366917300057e-05,
+      "loss": 0.0431,
+      "step": 2213
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5540880806015249,
+      "learning_rate": 2.1861200905296952e-05,
+      "loss": 0.0816,
+      "step": 2214
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.43048077962859,
+      "learning_rate": 2.1847033951328673e-05,
+      "loss": 0.0618,
+      "step": 2215
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3485028107336857,
+      "learning_rate": 2.1832866062565183e-05,
+      "loss": 0.0458,
+      "step": 2216
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.45006451806515346,
+      "learning_rate": 2.1818697246176943e-05,
+      "loss": 0.0561,
+      "step": 2217
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.414669886935835,
+      "learning_rate": 2.1804527509334875e-05,
+      "loss": 0.0639,
+      "step": 2218
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.38722038486539806,
+      "learning_rate": 2.1790356859210378e-05,
+      "loss": 0.0507,
+      "step": 2219
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3876689642497525,
+      "learning_rate": 2.17761853029753e-05,
+      "loss": 0.0625,
+      "step": 2220
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3570540990550638,
+      "learning_rate": 2.176201284780195e-05,
+      "loss": 0.0452,
+      "step": 2221
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4030564116747038,
+      "learning_rate": 2.1747839500863096e-05,
+      "loss": 0.0664,
+      "step": 2222
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4450630526663112,
+      "learning_rate": 2.1733665269331953e-05,
+      "loss": 0.0569,
+      "step": 2223
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.34117537533148296,
+      "learning_rate": 2.1719490160382196e-05,
+      "loss": 0.043,
+      "step": 2224
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.43426080890571217,
+      "learning_rate": 2.1705314181187922e-05,
+      "loss": 0.057,
+      "step": 2225
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3532074486008631,
+      "learning_rate": 2.169113733892369e-05,
+      "loss": 0.0622,
+      "step": 2226
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5107080218410214,
+      "learning_rate": 2.1676959640764484e-05,
+      "loss": 0.0657,
+      "step": 2227
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.37515124608793443,
+      "learning_rate": 2.166278109388572e-05,
+      "loss": 0.0508,
+      "step": 2228
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3836904306804695,
+      "learning_rate": 2.1648601705463263e-05,
+      "loss": 0.0482,
+      "step": 2229
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3251608294742402,
+      "learning_rate": 2.1634421482673368e-05,
+      "loss": 0.0361,
+      "step": 2230
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3571923357885689,
+      "learning_rate": 2.1620240432692737e-05,
+      "loss": 0.062,
+      "step": 2231
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.44233513042496453,
+      "learning_rate": 2.1606058562698496e-05,
+      "loss": 0.0776,
+      "step": 2232
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.44027822389523236,
+      "learning_rate": 2.1591875879868177e-05,
+      "loss": 0.0623,
+      "step": 2233
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4128785029628535,
+      "learning_rate": 2.157769239137971e-05,
+      "loss": 0.0465,
+      "step": 2234
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3351935421486037,
+      "learning_rate": 2.1563508104411457e-05,
+      "loss": 0.0516,
+      "step": 2235
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3493431292244364,
+      "learning_rate": 2.1549323026142168e-05,
+      "loss": 0.0478,
+      "step": 2236
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4243760676082615,
+      "learning_rate": 2.153513716375099e-05,
+      "loss": 0.0732,
+      "step": 2237
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3524577186976111,
+      "learning_rate": 2.1520950524417484e-05,
+      "loss": 0.0507,
+      "step": 2238
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4432719980591418,
+      "learning_rate": 2.1506763115321602e-05,
+      "loss": 0.0595,
+      "step": 2239
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4883926569683455,
+      "learning_rate": 2.1492574943643666e-05,
+      "loss": 0.0712,
+      "step": 2240
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.41592459958861955,
+      "learning_rate": 2.1478386016564406e-05,
+      "loss": 0.0588,
+      "step": 2241
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3325416680808275,
+      "learning_rate": 2.1464196341264915e-05,
+      "loss": 0.0545,
+      "step": 2242
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3711408484837303,
+      "learning_rate": 2.145000592492668e-05,
+      "loss": 0.0559,
+      "step": 2243
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.38396880621795976,
+      "learning_rate": 2.1435814774731557e-05,
+      "loss": 0.0454,
+      "step": 2244
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4628181078800962,
+      "learning_rate": 2.1421622897861777e-05,
+      "loss": 0.0704,
+      "step": 2245
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3477668479133176,
+      "learning_rate": 2.1407430301499934e-05,
+      "loss": 0.051,
+      "step": 2246
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4742206818542849,
+      "learning_rate": 2.139323699282899e-05,
+      "loss": 0.0551,
+      "step": 2247
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.44695137678370234,
+      "learning_rate": 2.1379042979032256e-05,
+      "loss": 0.0703,
+      "step": 2248
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4000663002023182,
+      "learning_rate": 2.1364848267293424e-05,
+      "loss": 0.0516,
+      "step": 2249
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.43719202509645305,
+      "learning_rate": 2.1350652864796513e-05,
+      "loss": 0.0638,
+      "step": 2250
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.39235092344908234,
+      "learning_rate": 2.133645677872591e-05,
+      "loss": 0.0548,
+      "step": 2251
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3668704388739481,
+      "learning_rate": 2.1322260016266337e-05,
+      "loss": 0.0408,
+      "step": 2252
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3150238402367162,
+      "learning_rate": 2.1308062584602865e-05,
+      "loss": 0.041,
+      "step": 2253
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5245566289748852,
+      "learning_rate": 2.1293864490920897e-05,
+      "loss": 0.0728,
+      "step": 2254
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3644941243520977,
+      "learning_rate": 2.1279665742406187e-05,
+      "loss": 0.0399,
+      "step": 2255
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.40779250084958746,
+      "learning_rate": 2.126546634624479e-05,
+      "loss": 0.0548,
+      "step": 2256
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3788580323047344,
+      "learning_rate": 2.125126630962312e-05,
+      "loss": 0.0603,
+      "step": 2257
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.39045975050796067,
+      "learning_rate": 2.1237065639727906e-05,
+      "loss": 0.0525,
+      "step": 2258
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4137295517953831,
+      "learning_rate": 2.1222864343746185e-05,
+      "loss": 0.052,
+      "step": 2259
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4671759084073663,
+      "learning_rate": 2.1208662428865326e-05,
+      "loss": 0.0558,
+      "step": 2260
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3834009707770895,
+      "learning_rate": 2.1194459902272997e-05,
+      "loss": 0.0619,
+      "step": 2261
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.33960149615028973,
+      "learning_rate": 2.1180256771157194e-05,
+      "loss": 0.0547,
+      "step": 2262
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3576002665943287,
+      "learning_rate": 2.1166053042706204e-05,
+      "loss": 0.0446,
+      "step": 2263
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.35063333642538436,
+      "learning_rate": 2.115184872410862e-05,
+      "loss": 0.0528,
+      "step": 2264
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.39879788200945676,
+      "learning_rate": 2.113764382255334e-05,
+      "loss": 0.0545,
+      "step": 2265
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5747178691662079,
+      "learning_rate": 2.1123438345229537e-05,
+      "loss": 0.0866,
+      "step": 2266
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.572943588549069,
+      "learning_rate": 2.110923229932671e-05,
+      "loss": 0.0792,
+      "step": 2267
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3623014794385008,
+      "learning_rate": 2.1095025692034614e-05,
+      "loss": 0.0482,
+      "step": 2268
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5828839910191299,
+      "learning_rate": 2.1080818530543304e-05,
+      "loss": 0.0825,
+      "step": 2269
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.478978707586908,
+      "learning_rate": 2.106661082204311e-05,
+      "loss": 0.0578,
+      "step": 2270
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.36041172424779067,
+      "learning_rate": 2.105240257372464e-05,
+      "loss": 0.0578,
+      "step": 2271
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.38219240140228306,
+      "learning_rate": 2.1038193792778775e-05,
+      "loss": 0.057,
+      "step": 2272
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.36689056402462356,
+      "learning_rate": 2.102398448639667e-05,
+      "loss": 0.0555,
+      "step": 2273
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.41942277567971076,
+      "learning_rate": 2.100977466176973e-05,
+      "loss": 0.0474,
+      "step": 2274
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4291063713877314,
+      "learning_rate": 2.099556432608965e-05,
+      "loss": 0.069,
+      "step": 2275
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.32794329743943423,
+      "learning_rate": 2.0981353486548363e-05,
+      "loss": 0.0444,
+      "step": 2276
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4401056489035421,
+      "learning_rate": 2.096714215033806e-05,
+      "loss": 0.0674,
+      "step": 2277
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3810406766864367,
+      "learning_rate": 2.095293032465119e-05,
+      "loss": 0.0634,
+      "step": 2278
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.42038816246072847,
+      "learning_rate": 2.0938718016680433e-05,
+      "loss": 0.069,
+      "step": 2279
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3479869212907461,
+      "learning_rate": 2.0924505233618734e-05,
+      "loss": 0.05,
+      "step": 2280
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3993746637811574,
+      "learning_rate": 2.0910291982659277e-05,
+      "loss": 0.06,
+      "step": 2281
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4005270748702094,
+      "learning_rate": 2.0896078270995463e-05,
+      "loss": 0.0489,
+      "step": 2282
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.396981721331143,
+      "learning_rate": 2.0881864105820936e-05,
+      "loss": 0.0685,
+      "step": 2283
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.36236271560600947,
+      "learning_rate": 2.0867649494329587e-05,
+      "loss": 0.0481,
+      "step": 2284
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3550763741797328,
+      "learning_rate": 2.085343444371551e-05,
+      "loss": 0.051,
+      "step": 2285
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3787923425353629,
+      "learning_rate": 2.083921896117303e-05,
+      "loss": 0.0529,
+      "step": 2286
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.39855603717447186,
+      "learning_rate": 2.0825003053896686e-05,
+      "loss": 0.0469,
+      "step": 2287
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.36050130735945973,
+      "learning_rate": 2.0810786729081237e-05,
+      "loss": 0.0589,
+      "step": 2288
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.373236063972357,
+      "learning_rate": 2.079656999392166e-05,
+      "loss": 0.0405,
+      "step": 2289
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4681168094320828,
+      "learning_rate": 2.0782352855613128e-05,
+      "loss": 0.075,
+      "step": 2290
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4689835125713308,
+      "learning_rate": 2.0768135321351016e-05,
+      "loss": 0.0728,
+      "step": 2291
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.331002402888966,
+      "learning_rate": 2.0753917398330902e-05,
+      "loss": 0.0493,
+      "step": 2292
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.36350895298160313,
+      "learning_rate": 2.073969909374858e-05,
+      "loss": 0.0413,
+      "step": 2293
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.43343697540317466,
+      "learning_rate": 2.0725480414800012e-05,
+      "loss": 0.0616,
+      "step": 2294
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3888972027018617,
+      "learning_rate": 2.0711261368681356e-05,
+      "loss": 0.0509,
+      "step": 2295
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3704147551794404,
+      "learning_rate": 2.069704196258896e-05,
+      "loss": 0.0412,
+      "step": 2296
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.390403971128073,
+      "learning_rate": 2.068282220371936e-05,
+      "loss": 0.0567,
+      "step": 2297
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.42857761960439805,
+      "learning_rate": 2.066860209926925e-05,
+      "loss": 0.0492,
+      "step": 2298
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4495829059886345,
+      "learning_rate": 2.0654381656435526e-05,
+      "loss": 0.0656,
+      "step": 2299
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.314731332667737,
+      "learning_rate": 2.064016088241523e-05,
+      "loss": 0.0397,
+      "step": 2300
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4139443714850748,
+      "learning_rate": 2.0625939784405586e-05,
+      "loss": 0.0552,
+      "step": 2301
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4813590291607367,
+      "learning_rate": 2.0611718369603982e-05,
+      "loss": 0.0897,
+      "step": 2302
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4462453207203422,
+      "learning_rate": 2.0597496645207964e-05,
+      "loss": 0.0727,
+      "step": 2303
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4211409505176882,
+      "learning_rate": 2.0583274618415227e-05,
+      "loss": 0.0639,
+      "step": 2304
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.35422693001649885,
+      "learning_rate": 2.056905229642363e-05,
+      "loss": 0.0498,
+      "step": 2305
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3386480188822363,
+      "learning_rate": 2.055482968643118e-05,
+      "loss": 0.0424,
+      "step": 2306
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4217757355504114,
+      "learning_rate": 2.0540606795636022e-05,
+      "loss": 0.0708,
+      "step": 2307
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.38629292635355156,
+      "learning_rate": 2.0526383631236454e-05,
+      "loss": 0.0576,
+      "step": 2308
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3800503672611192,
+      "learning_rate": 2.0512160200430896e-05,
+      "loss": 0.0543,
+      "step": 2309
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3909405768720875,
+      "learning_rate": 2.0497936510417928e-05,
+      "loss": 0.065,
+      "step": 2310
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.30695811315431687,
+      "learning_rate": 2.048371256839624e-05,
+      "loss": 0.0409,
+      "step": 2311
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.45774239859423366,
+      "learning_rate": 2.046948838156465e-05,
+      "loss": 0.0712,
+      "step": 2312
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.40685103045526805,
+      "learning_rate": 2.0455263957122113e-05,
+      "loss": 0.0633,
+      "step": 2313
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3732654071254558,
+      "learning_rate": 2.04410393022677e-05,
+      "loss": 0.0518,
+      "step": 2314
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4410804964372272,
+      "learning_rate": 2.0426814424200592e-05,
+      "loss": 0.0498,
+      "step": 2315
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.48450435941465114,
+      "learning_rate": 2.041258933012009e-05,
+      "loss": 0.0722,
+      "step": 2316
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.33728466935387436,
+      "learning_rate": 2.0398364027225593e-05,
+      "loss": 0.0481,
+      "step": 2317
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.42652390832959736,
+      "learning_rate": 2.0384138522716626e-05,
+      "loss": 0.0631,
+      "step": 2318
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4007601451708101,
+      "learning_rate": 2.036991282379279e-05,
+      "loss": 0.0543,
+      "step": 2319
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3920582037638183,
+      "learning_rate": 2.0355686937653818e-05,
+      "loss": 0.048,
+      "step": 2320
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.400337680944842,
+      "learning_rate": 2.03414608714995e-05,
+      "loss": 0.0511,
+      "step": 2321
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3276742041075169,
+      "learning_rate": 2.0327234632529738e-05,
+      "loss": 0.042,
+      "step": 2322
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.38644815861691056,
+      "learning_rate": 2.0313008227944527e-05,
+      "loss": 0.0457,
+      "step": 2323
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.42529913528824476,
+      "learning_rate": 2.029878166494393e-05,
+      "loss": 0.0631,
+      "step": 2324
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4331816871780329,
+      "learning_rate": 2.0284554950728106e-05,
+      "loss": 0.0625,
+      "step": 2325
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3881718279896845,
+      "learning_rate": 2.0270328092497266e-05,
+      "loss": 0.0331,
+      "step": 2326
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3789899674574855,
+      "learning_rate": 2.025610109745173e-05,
+      "loss": 0.064,
+      "step": 2327
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3858736417812594,
+      "learning_rate": 2.024187397279186e-05,
+      "loss": 0.0475,
+      "step": 2328
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.36311207185275024,
+      "learning_rate": 2.0227646725718085e-05,
+      "loss": 0.0522,
+      "step": 2329
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.42080811418689446,
+      "learning_rate": 2.021341936343091e-05,
+      "loss": 0.0669,
+      "step": 2330
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4367811223292118,
+      "learning_rate": 2.0199191893130893e-05,
+      "loss": 0.0642,
+      "step": 2331
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.44304571946339333,
+      "learning_rate": 2.018496432201863e-05,
+      "loss": 0.0667,
+      "step": 2332
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.39111846272706985,
+      "learning_rate": 2.017073665729479e-05,
+      "loss": 0.0576,
+      "step": 2333
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3756661458893297,
+      "learning_rate": 2.0156508906160083e-05,
+      "loss": 0.0558,
+      "step": 2334
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.36196762405032007,
+      "learning_rate": 2.0142281075815253e-05,
+      "loss": 0.0538,
+      "step": 2335
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.41753657449199966,
+      "learning_rate": 2.0128053173461105e-05,
+      "loss": 0.0533,
+      "step": 2336
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.35695799445414533,
+      "learning_rate": 2.0113825206298458e-05,
+      "loss": 0.049,
+      "step": 2337
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.43629295869942647,
+      "learning_rate": 2.009959718152818e-05,
+      "loss": 0.0629,
+      "step": 2338
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.36733210536684613,
+      "learning_rate": 2.008536910635115e-05,
+      "loss": 0.0533,
+      "step": 2339
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.34239426858627797,
+      "learning_rate": 2.00711409879683e-05,
+      "loss": 0.0422,
+      "step": 2340
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.370170287497713,
+      "learning_rate": 2.0056912833580557e-05,
+      "loss": 0.0583,
+      "step": 2341
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.37309516474357096,
+      "learning_rate": 2.0042684650388882e-05,
+      "loss": 0.0557,
+      "step": 2342
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3704571936199428,
+      "learning_rate": 2.0028456445594234e-05,
+      "loss": 0.0485,
+      "step": 2343
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5111271590786414,
+      "learning_rate": 2.0014228226397618e-05,
+      "loss": 0.0692,
+      "step": 2344
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.408704732406459,
+      "learning_rate": 2e-05,
+      "loss": 0.0522,
+      "step": 2345
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4179396931182973,
+      "learning_rate": 1.998577177360239e-05,
+      "loss": 0.0465,
+      "step": 2346
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4131803652646927,
+      "learning_rate": 1.997154355440577e-05,
+      "loss": 0.0621,
+      "step": 2347
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5090762625579097,
+      "learning_rate": 1.995731534961113e-05,
+      "loss": 0.0812,
+      "step": 2348
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3399041878463311,
+      "learning_rate": 1.9943087166419453e-05,
+      "loss": 0.0424,
+      "step": 2349
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4759776482082502,
+      "learning_rate": 1.9928859012031703e-05,
+      "loss": 0.0664,
+      "step": 2350
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.40435171446462115,
+      "learning_rate": 1.991463089364885e-05,
+      "loss": 0.0494,
+      "step": 2351
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.34789546494594176,
+      "learning_rate": 1.9900402818471825e-05,
+      "loss": 0.0511,
+      "step": 2352
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.33811444934171764,
+      "learning_rate": 1.9886174793701546e-05,
+      "loss": 0.0375,
+      "step": 2353
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.44538547520252336,
+      "learning_rate": 1.98719468265389e-05,
+      "loss": 0.0686,
+      "step": 2354
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.35413298259696435,
+      "learning_rate": 1.985771892418475e-05,
+      "loss": 0.0517,
+      "step": 2355
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3550469127739452,
+      "learning_rate": 1.9843491093839927e-05,
+      "loss": 0.0567,
+      "step": 2356
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.39031416923953993,
+      "learning_rate": 1.982926334270522e-05,
+      "loss": 0.0485,
+      "step": 2357
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.35131035345630074,
+      "learning_rate": 1.9815035677981378e-05,
+      "loss": 0.0486,
+      "step": 2358
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4054292677985494,
+      "learning_rate": 1.9800808106869117e-05,
+      "loss": 0.0615,
+      "step": 2359
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.29001962872784276,
+      "learning_rate": 1.9786580636569092e-05,
+      "loss": 0.0312,
+      "step": 2360
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3415296109097626,
+      "learning_rate": 1.9772353274281918e-05,
+      "loss": 0.035,
+      "step": 2361
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.30966855244264235,
+      "learning_rate": 1.9758126027208146e-05,
+      "loss": 0.031,
+      "step": 2362
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3744315965633129,
+      "learning_rate": 1.9743898902548273e-05,
+      "loss": 0.0589,
+      "step": 2363
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4780805205555588,
+      "learning_rate": 1.972967190750274e-05,
+      "loss": 0.0575,
+      "step": 2364
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4121418980786573,
+      "learning_rate": 1.9715445049271907e-05,
+      "loss": 0.0533,
+      "step": 2365
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.46827083574793354,
+      "learning_rate": 1.9701218335056076e-05,
+      "loss": 0.0667,
+      "step": 2366
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3978848832437258,
+      "learning_rate": 1.9686991772055476e-05,
+      "loss": 0.0542,
+      "step": 2367
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.39031993397702963,
+      "learning_rate": 1.9672765367470265e-05,
+      "loss": 0.0574,
+      "step": 2368
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3806681226158581,
+      "learning_rate": 1.9658539128500507e-05,
+      "loss": 0.0535,
+      "step": 2369
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5512545560791239,
+      "learning_rate": 1.964431306234619e-05,
+      "loss": 0.0692,
+      "step": 2370
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.39036734666751816,
+      "learning_rate": 1.9630087176207212e-05,
+      "loss": 0.0645,
+      "step": 2371
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3950232657691014,
+      "learning_rate": 1.9615861477283384e-05,
+      "loss": 0.0426,
+      "step": 2372
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3738941337640487,
+      "learning_rate": 1.9601635972774414e-05,
+      "loss": 0.0541,
+      "step": 2373
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.36719159874759005,
+      "learning_rate": 1.958741066987992e-05,
+      "loss": 0.061,
+      "step": 2374
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3758156370104676,
+      "learning_rate": 1.9573185575799414e-05,
+      "loss": 0.0546,
+      "step": 2375
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3452507119112734,
+      "learning_rate": 1.95589606977323e-05,
+      "loss": 0.0481,
+      "step": 2376
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.36708523969916607,
+      "learning_rate": 1.9544736042877886e-05,
+      "loss": 0.0505,
+      "step": 2377
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.35413298259696435,
+      "learning_rate": 1.9530511618435352e-05,
+      "loss": 0.0432,
+      "step": 2378
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.35248344384197294,
+      "learning_rate": 1.9516287431603767e-05,
+      "loss": 0.0411,
+      "step": 2379
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.41347060093390803,
+      "learning_rate": 1.950206348958208e-05,
+      "loss": 0.06,
+      "step": 2380
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3132393316635138,
+      "learning_rate": 1.948783979956911e-05,
+      "loss": 0.0455,
+      "step": 2381
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.48297148065402884,
+      "learning_rate": 1.9473616368763556e-05,
+      "loss": 0.0713,
+      "step": 2382
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3520791601840363,
+      "learning_rate": 1.9459393204363988e-05,
+      "loss": 0.0344,
+      "step": 2383
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.44459028405128037,
+      "learning_rate": 1.944517031356882e-05,
+      "loss": 0.0627,
+      "step": 2384
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3420051594032704,
+      "learning_rate": 1.9430947703576373e-05,
+      "loss": 0.0582,
+      "step": 2385
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3404259134894278,
+      "learning_rate": 1.9416725381584777e-05,
+      "loss": 0.0393,
+      "step": 2386
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.289782799432669,
+      "learning_rate": 1.9402503354792043e-05,
+      "loss": 0.048,
+      "step": 2387
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3397732584839841,
+      "learning_rate": 1.938828163039602e-05,
+      "loss": 0.0401,
+      "step": 2388
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.32293331451981655,
+      "learning_rate": 1.9374060215594417e-05,
+      "loss": 0.0371,
+      "step": 2389
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4402223424612593,
+      "learning_rate": 1.9359839117584775e-05,
+      "loss": 0.0523,
+      "step": 2390
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3906156538798913,
+      "learning_rate": 1.934561834356448e-05,
+      "loss": 0.0559,
+      "step": 2391
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4286245550140536,
+      "learning_rate": 1.9331397900730754e-05,
+      "loss": 0.0484,
+      "step": 2392
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3395809275399548,
+      "learning_rate": 1.9317177796280643e-05,
+      "loss": 0.0467,
+      "step": 2393
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3323436613670127,
+      "learning_rate": 1.930295803741104e-05,
+      "loss": 0.0462,
+      "step": 2394
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.40455667008457696,
+      "learning_rate": 1.9288738631318648e-05,
+      "loss": 0.0459,
+      "step": 2395
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.47881007574553464,
+      "learning_rate": 1.9274519585199995e-05,
+      "loss": 0.0668,
+      "step": 2396
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4284875239822807,
+      "learning_rate": 1.9260300906251422e-05,
+      "loss": 0.0618,
+      "step": 2397
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.41261309532758883,
+      "learning_rate": 1.92460826016691e-05,
+      "loss": 0.0531,
+      "step": 2398
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3862499130890881,
+      "learning_rate": 1.9231864678648994e-05,
+      "loss": 0.0604,
+      "step": 2399
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4244319113266689,
+      "learning_rate": 1.9217647144386885e-05,
+      "loss": 0.0706,
+      "step": 2400
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.38182235654596364,
+      "learning_rate": 1.9203430006078348e-05,
+      "loss": 0.0493,
+      "step": 2401
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3904403440806351,
+      "learning_rate": 1.918921327091876e-05,
+      "loss": 0.055,
+      "step": 2402
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3772309223455192,
+      "learning_rate": 1.9174996946103318e-05,
+      "loss": 0.0557,
+      "step": 2403
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3871718555457978,
+      "learning_rate": 1.9160781038826973e-05,
+      "loss": 0.044,
+      "step": 2404
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3547079727381328,
+      "learning_rate": 1.9146565556284492e-05,
+      "loss": 0.0448,
+      "step": 2405
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.37406062645948895,
+      "learning_rate": 1.9132350505670416e-05,
+      "loss": 0.0505,
+      "step": 2406
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.35909978030619516,
+      "learning_rate": 1.9118135894179067e-05,
+      "loss": 0.0537,
+      "step": 2407
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3432187613770574,
+      "learning_rate": 1.910392172900455e-05,
+      "loss": 0.0558,
+      "step": 2408
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.36632735372906744,
+      "learning_rate": 1.9089708017340733e-05,
+      "loss": 0.0451,
+      "step": 2409
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.37284915737457436,
+      "learning_rate": 1.9075494766381263e-05,
+      "loss": 0.0422,
+      "step": 2410
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3631603901305983,
+      "learning_rate": 1.906128198331957e-05,
+      "loss": 0.0448,
+      "step": 2411
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4002529372751417,
+      "learning_rate": 1.9047069675348816e-05,
+      "loss": 0.0488,
+      "step": 2412
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.34655052250355045,
+      "learning_rate": 1.9032857849661942e-05,
+      "loss": 0.0377,
+      "step": 2413
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.392345625319532,
+      "learning_rate": 1.901864651345164e-05,
+      "loss": 0.0498,
+      "step": 2414
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.37201830482951675,
+      "learning_rate": 1.9004435673910356e-05,
+      "loss": 0.0472,
+      "step": 2415
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.34400427690331264,
+      "learning_rate": 1.8990225338230276e-05,
+      "loss": 0.0372,
+      "step": 2416
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.34003686350944584,
+      "learning_rate": 1.8976015513603344e-05,
+      "loss": 0.0426,
+      "step": 2417
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4402325478679173,
+      "learning_rate": 1.8961806207221235e-05,
+      "loss": 0.0523,
+      "step": 2418
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.377458855742547,
+      "learning_rate": 1.8947597426275368e-05,
+      "loss": 0.0394,
+      "step": 2419
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3755874006904432,
+      "learning_rate": 1.8933389177956896e-05,
+      "loss": 0.0408,
+      "step": 2420
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4364706771337179,
+      "learning_rate": 1.8919181469456703e-05,
+      "loss": 0.0583,
+      "step": 2421
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3461203132234935,
+      "learning_rate": 1.8904974307965393e-05,
+      "loss": 0.0426,
+      "step": 2422
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.34278681584956155,
+      "learning_rate": 1.8890767700673296e-05,
+      "loss": 0.0375,
+      "step": 2423
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.38381606533579277,
+      "learning_rate": 1.8876561654770466e-05,
+      "loss": 0.0504,
+      "step": 2424
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.31069624805201673,
+      "learning_rate": 1.8862356177446667e-05,
+      "loss": 0.0406,
+      "step": 2425
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.37474832434021943,
+      "learning_rate": 1.8848151275891383e-05,
+      "loss": 0.0469,
+      "step": 2426
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3399929508362034,
+      "learning_rate": 1.8833946957293796e-05,
+      "loss": 0.035,
+      "step": 2427
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.29861208018557206,
+      "learning_rate": 1.8819743228842806e-05,
+      "loss": 0.0445,
+      "step": 2428
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3823212371359132,
+      "learning_rate": 1.8805540097727003e-05,
+      "loss": 0.0523,
+      "step": 2429
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3661808052382058,
+      "learning_rate": 1.8791337571134677e-05,
+      "loss": 0.0437,
+      "step": 2430
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.43966059555670706,
+      "learning_rate": 1.877713565625382e-05,
+      "loss": 0.0717,
+      "step": 2431
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.42020291527145803,
+      "learning_rate": 1.8762934360272097e-05,
+      "loss": 0.0399,
+      "step": 2432
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.33925193469677256,
+      "learning_rate": 1.8748733690376883e-05,
+      "loss": 0.0396,
+      "step": 2433
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3561620344345107,
+      "learning_rate": 1.8734533653755216e-05,
+      "loss": 0.0548,
+      "step": 2434
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3905977048874348,
+      "learning_rate": 1.8720334257593826e-05,
+      "loss": 0.0555,
+      "step": 2435
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3188506005050462,
+      "learning_rate": 1.8706135509079103e-05,
+      "loss": 0.0429,
+      "step": 2436
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.29873289198951014,
+      "learning_rate": 1.869193741539714e-05,
+      "loss": 0.0397,
+      "step": 2437
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.33672384891056945,
+      "learning_rate": 1.8677739983733666e-05,
+      "loss": 0.0382,
+      "step": 2438
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.38256970315878586,
+      "learning_rate": 1.8663543221274096e-05,
+      "loss": 0.0551,
+      "step": 2439
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.357453475138913,
+      "learning_rate": 1.8649347135203494e-05,
+      "loss": 0.0482,
+      "step": 2440
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.41836751650342313,
+      "learning_rate": 1.8635151732706586e-05,
+      "loss": 0.0528,
+      "step": 2441
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.45123432602435154,
+      "learning_rate": 1.862095702096775e-05,
+      "loss": 0.0814,
+      "step": 2442
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3612207407800378,
+      "learning_rate": 1.860676300717102e-05,
+      "loss": 0.0443,
+      "step": 2443
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.38921682215656894,
+      "learning_rate": 1.8592569698500076e-05,
+      "loss": 0.0546,
+      "step": 2444
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4375983876905605,
+      "learning_rate": 1.8578377102138223e-05,
+      "loss": 0.0591,
+      "step": 2445
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3719401094194151,
+      "learning_rate": 1.8564185225268446e-05,
+      "loss": 0.0482,
+      "step": 2446
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3118690319675185,
+      "learning_rate": 1.8549994075073327e-05,
+      "loss": 0.0492,
+      "step": 2447
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.30169929010979746,
+      "learning_rate": 1.853580365873509e-05,
+      "loss": 0.035,
+      "step": 2448
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.31670811464494275,
+      "learning_rate": 1.8521613983435604e-05,
+      "loss": 0.0383,
+      "step": 2449
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.39067542704295666,
+      "learning_rate": 1.8507425056356338e-05,
+      "loss": 0.0519,
+      "step": 2450
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4427692334043251,
+      "learning_rate": 1.8493236884678405e-05,
+      "loss": 0.0681,
+      "step": 2451
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.33943585877574,
+      "learning_rate": 1.847904947558252e-05,
+      "loss": 0.0401,
+      "step": 2452
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3351146467659251,
+      "learning_rate": 1.8464862836249014e-05,
+      "loss": 0.0492,
+      "step": 2453
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2713231355761402,
+      "learning_rate": 1.8450676973857842e-05,
+      "loss": 0.0354,
+      "step": 2454
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.35652747056555917,
+      "learning_rate": 1.843649189558855e-05,
+      "loss": 0.0473,
+      "step": 2455
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.37071698788929364,
+      "learning_rate": 1.8422307608620292e-05,
+      "loss": 0.047,
+      "step": 2456
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3180103963623487,
+      "learning_rate": 1.840812412013183e-05,
+      "loss": 0.0448,
+      "step": 2457
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.371256983142258,
+      "learning_rate": 1.8393941437301507e-05,
+      "loss": 0.0463,
+      "step": 2458
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3359779843620849,
+      "learning_rate": 1.8379759567307266e-05,
+      "loss": 0.049,
+      "step": 2459
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3967052464630342,
+      "learning_rate": 1.8365578517326642e-05,
+      "loss": 0.0514,
+      "step": 2460
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.35972747931859106,
+      "learning_rate": 1.8351398294536747e-05,
+      "loss": 0.0389,
+      "step": 2461
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4577699218323339,
+      "learning_rate": 1.833721890611428e-05,
+      "loss": 0.0621,
+      "step": 2462
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4291514086865876,
+      "learning_rate": 1.832304035923552e-05,
+      "loss": 0.0661,
+      "step": 2463
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.43516989503352316,
+      "learning_rate": 1.8308862661076313e-05,
+      "loss": 0.0589,
+      "step": 2464
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.38741815302530797,
+      "learning_rate": 1.829468581881208e-05,
+      "loss": 0.0564,
+      "step": 2465
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.37852651700468604,
+      "learning_rate": 1.8280509839617814e-05,
+      "loss": 0.0466,
+      "step": 2466
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.32719829805658424,
+      "learning_rate": 1.8266334730668054e-05,
+      "loss": 0.0385,
+      "step": 2467
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.40737122086407546,
+      "learning_rate": 1.8252160499136914e-05,
+      "loss": 0.051,
+      "step": 2468
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.36836061366207723,
+      "learning_rate": 1.8237987152198063e-05,
+      "loss": 0.0511,
+      "step": 2469
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4453552292518177,
+      "learning_rate": 1.822381469702471e-05,
+      "loss": 0.0661,
+      "step": 2470
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.32987103935269346,
+      "learning_rate": 1.8209643140789622e-05,
+      "loss": 0.0463,
+      "step": 2471
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4150262987754753,
+      "learning_rate": 1.8195472490665125e-05,
+      "loss": 0.0531,
+      "step": 2472
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4306926761039924,
+      "learning_rate": 1.8181302753823064e-05,
+      "loss": 0.064,
+      "step": 2473
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.39084008970839546,
+      "learning_rate": 1.8167133937434823e-05,
+      "loss": 0.0502,
+      "step": 2474
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5909688967670853,
+      "learning_rate": 1.8152966048671334e-05,
+      "loss": 0.0722,
+      "step": 2475
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.38884734980182967,
+      "learning_rate": 1.813879909470305e-05,
+      "loss": 0.0555,
+      "step": 2476
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.37595676518621624,
+      "learning_rate": 1.8124633082699956e-05,
+      "loss": 0.0502,
+      "step": 2477
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.474914844935941,
+      "learning_rate": 1.8110468019831553e-05,
+      "loss": 0.0721,
+      "step": 2478
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3843573907369612,
+      "learning_rate": 1.8096303913266864e-05,
+      "loss": 0.0756,
+      "step": 2479
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.40612778292424834,
+      "learning_rate": 1.808214077017444e-05,
+      "loss": 0.0581,
+      "step": 2480
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3720001394825335,
+      "learning_rate": 1.8067978597722325e-05,
+      "loss": 0.0549,
+      "step": 2481
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3975497809613756,
+      "learning_rate": 1.8053817403078087e-05,
+      "loss": 0.0665,
+      "step": 2482
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3122717977815116,
+      "learning_rate": 1.8039657193408788e-05,
+      "loss": 0.043,
+      "step": 2483
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.40600406098096514,
+      "learning_rate": 1.8025497975881004e-05,
+      "loss": 0.0543,
+      "step": 2484
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.300010101327112,
+      "learning_rate": 1.8011339757660798e-05,
+      "loss": 0.0365,
+      "step": 2485
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3967849268788347,
+      "learning_rate": 1.7997182545913732e-05,
+      "loss": 0.0409,
+      "step": 2486
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.29865259737449684,
+      "learning_rate": 1.798302634780486e-05,
+      "loss": 0.0439,
+      "step": 2487
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3211444888899818,
+      "learning_rate": 1.796887117049871e-05,
+      "loss": 0.0387,
+      "step": 2488
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3812861472331901,
+      "learning_rate": 1.7954717021159316e-05,
+      "loss": 0.055,
+      "step": 2489
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3748673760814513,
+      "learning_rate": 1.7940563906950175e-05,
+      "loss": 0.0447,
+      "step": 2490
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3460907783122389,
+      "learning_rate": 1.7926411835034267e-05,
+      "loss": 0.0545,
+      "step": 2491
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.41721838147453766,
+      "learning_rate": 1.791226081257404e-05,
+      "loss": 0.0578,
+      "step": 2492
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3697827924639646,
+      "learning_rate": 1.7898110846731415e-05,
+      "loss": 0.0587,
+      "step": 2493
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4420947385348038,
+      "learning_rate": 1.7883961944667772e-05,
+      "loss": 0.0588,
+      "step": 2494
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.33188920067903177,
+      "learning_rate": 1.786981411354396e-05,
+      "loss": 0.0486,
+      "step": 2495
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.40870752154700785,
+      "learning_rate": 1.7855667360520277e-05,
+      "loss": 0.0512,
+      "step": 2496
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.35365349625987297,
+      "learning_rate": 1.7841521692756497e-05,
+      "loss": 0.05,
+      "step": 2497
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3862683148655958,
+      "learning_rate": 1.782737711741182e-05,
+      "loss": 0.0457,
+      "step": 2498
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.34301279953855635,
+      "learning_rate": 1.7813233641644904e-05,
+      "loss": 0.0419,
+      "step": 2499
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.326006270529698,
+      "learning_rate": 1.7799091272613843e-05,
+      "loss": 0.0404,
+      "step": 2500
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.38251507162182874,
+      "learning_rate": 1.778495001747618e-05,
+      "loss": 0.0534,
+      "step": 2501
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3813334717597466,
+      "learning_rate": 1.7770809883388896e-05,
+      "loss": 0.044,
+      "step": 2502
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.39002479217662644,
+      "learning_rate": 1.775667087750839e-05,
+      "loss": 0.0435,
+      "step": 2503
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3337172281948741,
+      "learning_rate": 1.774253300699051e-05,
+      "loss": 0.0428,
+      "step": 2504
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.36579148748737944,
+      "learning_rate": 1.77283962789905e-05,
+      "loss": 0.0432,
+      "step": 2505
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3800404866196099,
+      "learning_rate": 1.771426070066307e-05,
+      "loss": 0.0534,
+      "step": 2506
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.35188721389983824,
+      "learning_rate": 1.770012627916231e-05,
+      "loss": 0.0443,
+      "step": 2507
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.35454603054259937,
+      "learning_rate": 1.768599302164174e-05,
+      "loss": 0.0421,
+      "step": 2508
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.37317982646238296,
+      "learning_rate": 1.7671860935254285e-05,
+      "loss": 0.0587,
+      "step": 2509
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2835275173347284,
+      "learning_rate": 1.7657730027152286e-05,
+      "loss": 0.036,
+      "step": 2510
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.35067588481115647,
+      "learning_rate": 1.7643600304487475e-05,
+      "loss": 0.042,
+      "step": 2511
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3442711239562008,
+      "learning_rate": 1.7629471774410997e-05,
+      "loss": 0.0367,
+      "step": 2512
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3680934484137127,
+      "learning_rate": 1.7615344444073385e-05,
+      "loss": 0.0415,
+      "step": 2513
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.35494281301469954,
+      "learning_rate": 1.7601218320624562e-05,
+      "loss": 0.047,
+      "step": 2514
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.41312054788957775,
+      "learning_rate": 1.7587093411213856e-05,
+      "loss": 0.0511,
+      "step": 2515
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.32683657684756035,
+      "learning_rate": 1.7572969722989967e-05,
+      "loss": 0.0441,
+      "step": 2516
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4399027579856521,
+      "learning_rate": 1.755884726310098e-05,
+      "loss": 0.062,
+      "step": 2517
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4171422826714955,
+      "learning_rate": 1.754472603869436e-05,
+      "loss": 0.0502,
+      "step": 2518
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.38591179318083907,
+      "learning_rate": 1.7530606056916935e-05,
+      "loss": 0.0452,
+      "step": 2519
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.32631634465151743,
+      "learning_rate": 1.751648732491493e-05,
+      "loss": 0.046,
+      "step": 2520
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3409260255227103,
+      "learning_rate": 1.7502369849833908e-05,
+      "loss": 0.0461,
+      "step": 2521
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3786171075383073,
+      "learning_rate": 1.748825363881881e-05,
+      "loss": 0.046,
+      "step": 2522
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.35722992125319664,
+      "learning_rate": 1.7474138699013953e-05,
+      "loss": 0.0563,
+      "step": 2523
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.298713450151254,
+      "learning_rate": 1.746002503756298e-05,
+      "loss": 0.0374,
+      "step": 2524
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.30910242384037684,
+      "learning_rate": 1.7445912661608912e-05,
+      "loss": 0.0407,
+      "step": 2525
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.46861530593996237,
+      "learning_rate": 1.7431801578294097e-05,
+      "loss": 0.0517,
+      "step": 2526
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.394621111767584,
+      "learning_rate": 1.7417691794760247e-05,
+      "loss": 0.0524,
+      "step": 2527
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3686938129873323,
+      "learning_rate": 1.740358331814841e-05,
+      "loss": 0.0452,
+      "step": 2528
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4074832829749992,
+      "learning_rate": 1.7389476155598974e-05,
+      "loss": 0.0514,
+      "step": 2529
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.342856394410026,
+      "learning_rate": 1.7375370314251657e-05,
+      "loss": 0.0429,
+      "step": 2530
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3690525333308805,
+      "learning_rate": 1.7361265801245504e-05,
+      "loss": 0.0548,
+      "step": 2531
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.37886233911028505,
+      "learning_rate": 1.7347162623718913e-05,
+      "loss": 0.0615,
+      "step": 2532
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.35522498691184834,
+      "learning_rate": 1.7333060788809582e-05,
+      "loss": 0.0384,
+      "step": 2533
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.35634166307916276,
+      "learning_rate": 1.7318960303654534e-05,
+      "loss": 0.0404,
+      "step": 2534
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4526984574928489,
+      "learning_rate": 1.7304861175390112e-05,
+      "loss": 0.0644,
+      "step": 2535
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3635237509931962,
+      "learning_rate": 1.729076341115197e-05,
+      "loss": 0.0689,
+      "step": 2536
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3394390414952252,
+      "learning_rate": 1.7276667018075073e-05,
+      "loss": 0.0521,
+      "step": 2537
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.358893528049773,
+      "learning_rate": 1.726257200329369e-05,
+      "loss": 0.0439,
+      "step": 2538
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3475280696866417,
+      "learning_rate": 1.72484783739414e-05,
+      "loss": 0.0516,
+      "step": 2539
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.31888040371514104,
+      "learning_rate": 1.7234386137151067e-05,
+      "loss": 0.0389,
+      "step": 2540
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3433524563974395,
+      "learning_rate": 1.7220295300054867e-05,
+      "loss": 0.0416,
+      "step": 2541
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3505328039852608,
+      "learning_rate": 1.7206205869784254e-05,
+      "loss": 0.0434,
+      "step": 2542
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.31793390406469835,
+      "learning_rate": 1.719211785346998e-05,
+      "loss": 0.0378,
+      "step": 2543
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3443520648747606,
+      "learning_rate": 1.717803125824207e-05,
+      "loss": 0.0513,
+      "step": 2544
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.328554734885512,
+      "learning_rate": 1.716394609122984e-05,
+      "loss": 0.0448,
+      "step": 2545
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2885518589813259,
+      "learning_rate": 1.714986235956188e-05,
+      "loss": 0.0401,
+      "step": 2546
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4041937728850449,
+      "learning_rate": 1.713578007036605e-05,
+      "loss": 0.0582,
+      "step": 2547
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.39700011361334303,
+      "learning_rate": 1.712169923076948e-05,
+      "loss": 0.058,
+      "step": 2548
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.29387935265903975,
+      "learning_rate": 1.710761984789858e-05,
+      "loss": 0.0382,
+      "step": 2549
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.503105355110246,
+      "learning_rate": 1.7093541928879004e-05,
+      "loss": 0.0946,
+      "step": 2550
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.33110208087717735,
+      "learning_rate": 1.7079465480835677e-05,
+      "loss": 0.0399,
+      "step": 2551
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.26865055107203595,
+      "learning_rate": 1.7065390510892767e-05,
+      "loss": 0.0423,
+      "step": 2552
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4193389039629841,
+      "learning_rate": 1.7051317026173715e-05,
+      "loss": 0.0548,
+      "step": 2553
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3646464384051765,
+      "learning_rate": 1.703724503380119e-05,
+      "loss": 0.0496,
+      "step": 2554
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.328030254672491,
+      "learning_rate": 1.7023174540897112e-05,
+      "loss": 0.0367,
+      "step": 2555
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.37608506259623486,
+      "learning_rate": 1.7009105554582652e-05,
+      "loss": 0.0529,
+      "step": 2556
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3777206197480758,
+      "learning_rate": 1.6995038081978193e-05,
+      "loss": 0.0584,
+      "step": 2557
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.30499260860226335,
+      "learning_rate": 1.6980972130203396e-05,
+      "loss": 0.0367,
+      "step": 2558
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3273897106889949,
+      "learning_rate": 1.6966907706377103e-05,
+      "loss": 0.0396,
+      "step": 2559
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3536332498520532,
+      "learning_rate": 1.695284481761742e-05,
+      "loss": 0.0504,
+      "step": 2560
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.388388218718895,
+      "learning_rate": 1.6938783471041647e-05,
+      "loss": 0.0537,
+      "step": 2561
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2954232205110963,
+      "learning_rate": 1.692472367376633e-05,
+      "loss": 0.039,
+      "step": 2562
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.464982372739429,
+      "learning_rate": 1.691066543290721e-05,
+      "loss": 0.0805,
+      "step": 2563
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.27277565284516625,
+      "learning_rate": 1.6896608755579256e-05,
+      "loss": 0.0413,
+      "step": 2564
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.35453117303355675,
+      "learning_rate": 1.6882553648896625e-05,
+      "loss": 0.0425,
+      "step": 2565
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.27941979857681537,
+      "learning_rate": 1.686850011997271e-05,
+      "loss": 0.0342,
+      "step": 2566
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3715773793955695,
+      "learning_rate": 1.685444817592008e-05,
+      "loss": 0.0618,
+      "step": 2567
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.373749675112682,
+      "learning_rate": 1.6840397823850513e-05,
+      "loss": 0.0532,
+      "step": 2568
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3291121008569529,
+      "learning_rate": 1.6826349070874973e-05,
+      "loss": 0.0406,
+      "step": 2569
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.35252667772513235,
+      "learning_rate": 1.6812301924103626e-05,
+      "loss": 0.0529,
+      "step": 2570
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3799685501586278,
+      "learning_rate": 1.6798256390645816e-05,
+      "loss": 0.0451,
+      "step": 2571
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.35622731688714665,
+      "learning_rate": 1.6784212477610075e-05,
+      "loss": 0.0442,
+      "step": 2572
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3117283831568278,
+      "learning_rate": 1.6770170192104107e-05,
+      "loss": 0.0331,
+      "step": 2573
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.32447007790578225,
+      "learning_rate": 1.67561295412348e-05,
+      "loss": 0.0452,
+      "step": 2574
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3309813336600599,
+      "learning_rate": 1.6742090532108228e-05,
+      "loss": 0.0392,
+      "step": 2575
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.40694373758744323,
+      "learning_rate": 1.6728053171829603e-05,
+      "loss": 0.0606,
+      "step": 2576
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5453939819669776,
+      "learning_rate": 1.6714017467503328e-05,
+      "loss": 0.1091,
+      "step": 2577
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4217280732488526,
+      "learning_rate": 1.6699983426232955e-05,
+      "loss": 0.048,
+      "step": 2578
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3787519396364005,
+      "learning_rate": 1.6685951055121203e-05,
+      "loss": 0.0446,
+      "step": 2579
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.335666791384736,
+      "learning_rate": 1.667192036126993e-05,
+      "loss": 0.047,
+      "step": 2580
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.26844197075243886,
+      "learning_rate": 1.665789135178017e-05,
+      "loss": 0.0343,
+      "step": 2581
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.36392243519255474,
+      "learning_rate": 1.664386403375208e-05,
+      "loss": 0.0453,
+      "step": 2582
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4301809598543906,
+      "learning_rate": 1.6629838414284972e-05,
+      "loss": 0.0472,
+      "step": 2583
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3268916702396021,
+      "learning_rate": 1.6615814500477307e-05,
+      "loss": 0.0423,
+      "step": 2584
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3298473680154127,
+      "learning_rate": 1.6601792299426668e-05,
+      "loss": 0.0391,
+      "step": 2585
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3484045718761079,
+      "learning_rate": 1.658777181822978e-05,
+      "loss": 0.0436,
+      "step": 2586
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3389652780822943,
+      "learning_rate": 1.6573753063982492e-05,
+      "loss": 0.0448,
+      "step": 2587
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.37909805712879996,
+      "learning_rate": 1.655973604377978e-05,
+      "loss": 0.06,
+      "step": 2588
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3678902536237636,
+      "learning_rate": 1.6545720764715746e-05,
+      "loss": 0.039,
+      "step": 2589
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3260284611076184,
+      "learning_rate": 1.6531707233883607e-05,
+      "loss": 0.0459,
+      "step": 2590
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.40296197818244606,
+      "learning_rate": 1.651769545837569e-05,
+      "loss": 0.0643,
+      "step": 2591
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2994351050540486,
+      "learning_rate": 1.650368544528346e-05,
+      "loss": 0.03,
+      "step": 2592
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.352678202806525,
+      "learning_rate": 1.6489677201697453e-05,
+      "loss": 0.0436,
+      "step": 2593
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.35567778425230934,
+      "learning_rate": 1.6475670734707336e-05,
+      "loss": 0.0538,
+      "step": 2594
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.31933516756017416,
+      "learning_rate": 1.6461666051401865e-05,
+      "loss": 0.0378,
+      "step": 2595
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3431270003998748,
+      "learning_rate": 1.6447663158868897e-05,
+      "loss": 0.0368,
+      "step": 2596
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3797904247919991,
+      "learning_rate": 1.6433662064195378e-05,
+      "loss": 0.0427,
+      "step": 2597
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.40201406952923524,
+      "learning_rate": 1.641966277446735e-05,
+      "loss": 0.0773,
+      "step": 2598
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3433373423549848,
+      "learning_rate": 1.6405665296769942e-05,
+      "loss": 0.0334,
+      "step": 2599
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.41390034361412825,
+      "learning_rate": 1.6391669638187355e-05,
+      "loss": 0.0634,
+      "step": 2600
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3183565198150102,
+      "learning_rate": 1.6377675805802882e-05,
+      "loss": 0.0384,
+      "step": 2601
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.34967596602657924,
+      "learning_rate": 1.6363683806698896e-05,
+      "loss": 0.0466,
+      "step": 2602
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3991341204438977,
+      "learning_rate": 1.6349693647956824e-05,
+      "loss": 0.0667,
+      "step": 2603
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3746840417493953,
+      "learning_rate": 1.6335705336657176e-05,
+      "loss": 0.0417,
+      "step": 2604
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.30447485423593174,
+      "learning_rate": 1.632171887987952e-05,
+      "loss": 0.0398,
+      "step": 2605
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.27851170434104927,
+      "learning_rate": 1.6307734284702484e-05,
+      "loss": 0.0446,
+      "step": 2606
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3622261229518271,
+      "learning_rate": 1.6293751558203764e-05,
+      "loss": 0.0417,
+      "step": 2607
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3671122128835008,
+      "learning_rate": 1.6279770707460096e-05,
+      "loss": 0.0481,
+      "step": 2608
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.29716355956407814,
+      "learning_rate": 1.6265791739547276e-05,
+      "loss": 0.0413,
+      "step": 2609
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3207904459828551,
+      "learning_rate": 1.625181466154015e-05,
+      "loss": 0.0335,
+      "step": 2610
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3842392433599277,
+      "learning_rate": 1.62378394805126e-05,
+      "loss": 0.0432,
+      "step": 2611
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3508987306493617,
+      "learning_rate": 1.6223866203537558e-05,
+      "loss": 0.0439,
+      "step": 2612
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3887107101284892,
+      "learning_rate": 1.6209894837686974e-05,
+      "loss": 0.0493,
+      "step": 2613
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.34506361480253667,
+      "learning_rate": 1.6195925390031845e-05,
+      "loss": 0.0385,
+      "step": 2614
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3061896731845957,
+      "learning_rate": 1.61819578676422e-05,
+      "loss": 0.0326,
+      "step": 2615
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.32517755591683667,
+      "learning_rate": 1.616799227758708e-05,
+      "loss": 0.0453,
+      "step": 2616
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3009063658647218,
+      "learning_rate": 1.6154028626934548e-05,
+      "loss": 0.0353,
+      "step": 2617
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.34151820130246274,
+      "learning_rate": 1.6140066922751715e-05,
+      "loss": 0.0423,
+      "step": 2618
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3868738884362936,
+      "learning_rate": 1.612610717210467e-05,
+      "loss": 0.0526,
+      "step": 2619
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.29858877532184536,
+      "learning_rate": 1.611214938205854e-05,
+      "loss": 0.0415,
+      "step": 2620
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.30091310063498156,
+      "learning_rate": 1.609819355967744e-05,
+      "loss": 0.0368,
+      "step": 2621
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.43227457296134053,
+      "learning_rate": 1.6084239712024492e-05,
+      "loss": 0.0533,
+      "step": 2622
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.36195687927499137,
+      "learning_rate": 1.6070287846161834e-05,
+      "loss": 0.041,
+      "step": 2623
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2939982699218436,
+      "learning_rate": 1.6056337969150584e-05,
+      "loss": 0.0298,
+      "step": 2624
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.33352795025660326,
+      "learning_rate": 1.6042390088050864e-05,
+      "loss": 0.0448,
+      "step": 2625
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.49970352800821244,
+      "learning_rate": 1.6028444209921775e-05,
+      "loss": 0.0493,
+      "step": 2626
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3028238717966314,
+      "learning_rate": 1.601450034182142e-05,
+      "loss": 0.0352,
+      "step": 2627
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.33227471793707164,
+      "learning_rate": 1.6000558490806877e-05,
+      "loss": 0.0548,
+      "step": 2628
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2640987190823617,
+      "learning_rate": 1.59866186639342e-05,
+      "loss": 0.0317,
+      "step": 2629
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3609816875079501,
+      "learning_rate": 1.597268086825842e-05,
+      "loss": 0.0461,
+      "step": 2630
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3688559266875854,
+      "learning_rate": 1.5958745110833536e-05,
+      "loss": 0.0521,
+      "step": 2631
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.366729450630457,
+      "learning_rate": 1.5944811398712527e-05,
+      "loss": 0.052,
+      "step": 2632
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.35198386689963646,
+      "learning_rate": 1.5930879738947328e-05,
+      "loss": 0.0561,
+      "step": 2633
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.32985824394095403,
+      "learning_rate": 1.5916950138588834e-05,
+      "loss": 0.0519,
+      "step": 2634
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3273032320388355,
+      "learning_rate": 1.5903022604686908e-05,
+      "loss": 0.047,
+      "step": 2635
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2991084361099232,
+      "learning_rate": 1.5889097144290357e-05,
+      "loss": 0.0357,
+      "step": 2636
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3294866894962564,
+      "learning_rate": 1.587517376444694e-05,
+      "loss": 0.0484,
+      "step": 2637
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.34033152773148057,
+      "learning_rate": 1.5861252472203367e-05,
+      "loss": 0.0433,
+      "step": 2638
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.32795379350134185,
+      "learning_rate": 1.5847333274605286e-05,
+      "loss": 0.0445,
+      "step": 2639
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.44996915420142114,
+      "learning_rate": 1.5833416178697298e-05,
+      "loss": 0.082,
+      "step": 2640
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.310305219470742,
+      "learning_rate": 1.5819501191522917e-05,
+      "loss": 0.0355,
+      "step": 2641
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2931443006346977,
+      "learning_rate": 1.5805588320124607e-05,
+      "loss": 0.0338,
+      "step": 2642
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3440901958568532,
+      "learning_rate": 1.5791677571543762e-05,
+      "loss": 0.0477,
+      "step": 2643
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.43826924543888984,
+      "learning_rate": 1.5777768952820697e-05,
+      "loss": 0.0781,
+      "step": 2644
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.38238867828954753,
+      "learning_rate": 1.576386247099465e-05,
+      "loss": 0.0561,
+      "step": 2645
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3758116521453048,
+      "learning_rate": 1.5749958133103772e-05,
+      "loss": 0.0616,
+      "step": 2646
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2914145634669556,
+      "learning_rate": 1.5736055946185137e-05,
+      "loss": 0.0326,
+      "step": 2647
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3146572398322556,
+      "learning_rate": 1.572215591727473e-05,
+      "loss": 0.0393,
+      "step": 2648
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3275491907399743,
+      "learning_rate": 1.570825805340743e-05,
+      "loss": 0.0462,
+      "step": 2649
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3821830245198099,
+      "learning_rate": 1.5694362361617043e-05,
+      "loss": 0.0479,
+      "step": 2650
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.38612555273676535,
+      "learning_rate": 1.568046884893626e-05,
+      "loss": 0.0585,
+      "step": 2651
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.35829701664699615,
+      "learning_rate": 1.5666577522396658e-05,
+      "loss": 0.0501,
+      "step": 2652
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3198429364267204,
+      "learning_rate": 1.565268838902875e-05,
+      "loss": 0.039,
+      "step": 2653
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.33457842637181534,
+      "learning_rate": 1.5638801455861893e-05,
+      "loss": 0.0462,
+      "step": 2654
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.34628602363815075,
+      "learning_rate": 1.5624916729924354e-05,
+      "loss": 0.0505,
+      "step": 2655
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.28053922378784835,
+      "learning_rate": 1.561103421824328e-05,
+      "loss": 0.0305,
+      "step": 2656
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3882918488804002,
+      "learning_rate": 1.5597153927844693e-05,
+      "loss": 0.0448,
+      "step": 2657
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4070758311921904,
+      "learning_rate": 1.5583275865753492e-05,
+      "loss": 0.0483,
+      "step": 2658
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3473070941618986,
+      "learning_rate": 1.556940003899345e-05,
+      "loss": 0.0444,
+      "step": 2659
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4384647699142405,
+      "learning_rate": 1.55555264545872e-05,
+      "loss": 0.0759,
+      "step": 2660
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.321066527129186,
+      "learning_rate": 1.5541655119556262e-05,
+      "loss": 0.0393,
+      "step": 2661
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.32812148046876627,
+      "learning_rate": 1.5527786040921e-05,
+      "loss": 0.0379,
+      "step": 2662
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2679898790309503,
+      "learning_rate": 1.551391922570064e-05,
+      "loss": 0.0298,
+      "step": 2663
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4259319213815324,
+      "learning_rate": 1.550005468091326e-05,
+      "loss": 0.0726,
+      "step": 2664
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2978625936497866,
+      "learning_rate": 1.548619241357579e-05,
+      "loss": 0.0356,
+      "step": 2665
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4194003392848951,
+      "learning_rate": 1.5472332430704007e-05,
+      "loss": 0.0549,
+      "step": 2666
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.30088993684977516,
+      "learning_rate": 1.545847473931254e-05,
+      "loss": 0.0426,
+      "step": 2667
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.44909686217909617,
+      "learning_rate": 1.5444619346414845e-05,
+      "loss": 0.0627,
+      "step": 2668
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3775639896284503,
+      "learning_rate": 1.543076625902322e-05,
+      "loss": 0.0522,
+      "step": 2669
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3054246299862562,
+      "learning_rate": 1.5416915484148805e-05,
+      "loss": 0.0386,
+      "step": 2670
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.39041315056117604,
+      "learning_rate": 1.5403067028801558e-05,
+      "loss": 0.0507,
+      "step": 2671
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.31897717776320117,
+      "learning_rate": 1.5389220899990267e-05,
+      "loss": 0.0503,
+      "step": 2672
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.307557708802915,
+      "learning_rate": 1.5375377104722545e-05,
+      "loss": 0.043,
+      "step": 2673
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3631592617515664,
+      "learning_rate": 1.5361535650004818e-05,
+      "loss": 0.0489,
+      "step": 2674
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.33431035022896627,
+      "learning_rate": 1.5347696542842333e-05,
+      "loss": 0.0384,
+      "step": 2675
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3262609257458062,
+      "learning_rate": 1.5333859790239148e-05,
+      "loss": 0.042,
+      "step": 2676
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3015285843470328,
+      "learning_rate": 1.5320025399198125e-05,
+      "loss": 0.0466,
+      "step": 2677
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4392304937366175,
+      "learning_rate": 1.530619337672093e-05,
+      "loss": 0.0576,
+      "step": 2678
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3149699353448315,
+      "learning_rate": 1.5292363729808048e-05,
+      "loss": 0.0371,
+      "step": 2679
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3775983831905447,
+      "learning_rate": 1.5278536465458738e-05,
+      "loss": 0.0641,
+      "step": 2680
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3950621366217049,
+      "learning_rate": 1.5264711590671067e-05,
+      "loss": 0.0524,
+      "step": 2681
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3429047857260409,
+      "learning_rate": 1.5250889112441889e-05,
+      "loss": 0.0512,
+      "step": 2682
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2749659869443981,
+      "learning_rate": 1.5237069037766843e-05,
+      "loss": 0.0361,
+      "step": 2683
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3164070859356422,
+      "learning_rate": 1.5223251373640354e-05,
+      "loss": 0.0386,
+      "step": 2684
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.28880691828118865,
+      "learning_rate": 1.5209436127055627e-05,
+      "loss": 0.0301,
+      "step": 2685
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3426884602907715,
+      "learning_rate": 1.5195623305004637e-05,
+      "loss": 0.037,
+      "step": 2686
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.32675668975035227,
+      "learning_rate": 1.5181812914478146e-05,
+      "loss": 0.039,
+      "step": 2687
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.36544928118454606,
+      "learning_rate": 1.5168004962465681e-05,
+      "loss": 0.0533,
+      "step": 2688
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.27877802104980726,
+      "learning_rate": 1.5154199455955523e-05,
+      "loss": 0.0308,
+      "step": 2689
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3260941327124316,
+      "learning_rate": 1.5140396401934725e-05,
+      "loss": 0.0497,
+      "step": 2690
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.29156792479249044,
+      "learning_rate": 1.5126595807389098e-05,
+      "loss": 0.0418,
+      "step": 2691
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3199316994412665,
+      "learning_rate": 1.5112797679303206e-05,
+      "loss": 0.0384,
+      "step": 2692
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.35845826192788094,
+      "learning_rate": 1.5099002024660368e-05,
+      "loss": 0.0579,
+      "step": 2693
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3281439367007041,
+      "learning_rate": 1.5085208850442649e-05,
+      "loss": 0.0431,
+      "step": 2694
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.38845412705942295,
+      "learning_rate": 1.5071418163630855e-05,
+      "loss": 0.0475,
+      "step": 2695
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.38170965111248956,
+      "learning_rate": 1.5057629971204546e-05,
+      "loss": 0.0535,
+      "step": 2696
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3093982061679945,
+      "learning_rate": 1.5043844280142005e-05,
+      "loss": 0.0325,
+      "step": 2697
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.28981425502855296,
+      "learning_rate": 1.5030061097420255e-05,
+      "loss": 0.0433,
+      "step": 2698
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.31525880655108296,
+      "learning_rate": 1.5016280430015052e-05,
+      "loss": 0.0359,
+      "step": 2699
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3033155093331658,
+      "learning_rate": 1.5002502284900871e-05,
+      "loss": 0.0319,
+      "step": 2700
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3376865101202634,
+      "learning_rate": 1.4988726669050917e-05,
+      "loss": 0.0498,
+      "step": 2701
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.41858342686618616,
+      "learning_rate": 1.4974953589437117e-05,
+      "loss": 0.0576,
+      "step": 2702
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3182429118187405,
+      "learning_rate": 1.4961183053030106e-05,
+      "loss": 0.0403,
+      "step": 2703
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3506493152510276,
+      "learning_rate": 1.4947415066799247e-05,
+      "loss": 0.0492,
+      "step": 2704
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2502868020514377,
+      "learning_rate": 1.4933649637712593e-05,
+      "loss": 0.03,
+      "step": 2705
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.33765236497062373,
+      "learning_rate": 1.4919886772736915e-05,
+      "loss": 0.0325,
+      "step": 2706
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.29693614179525,
+      "learning_rate": 1.4906126478837683e-05,
+      "loss": 0.0356,
+      "step": 2707
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3101886626634362,
+      "learning_rate": 1.4892368762979067e-05,
+      "loss": 0.0302,
+      "step": 2708
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3130714198962901,
+      "learning_rate": 1.4878613632123928e-05,
+      "loss": 0.0318,
+      "step": 2709
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3793925281604337,
+      "learning_rate": 1.4864861093233827e-05,
+      "loss": 0.0504,
+      "step": 2710
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3023329996568408,
+      "learning_rate": 1.4851111153269005e-05,
+      "loss": 0.0421,
+      "step": 2711
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3567602360211933,
+      "learning_rate": 1.4837363819188379e-05,
+      "loss": 0.0402,
+      "step": 2712
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.33821174488992095,
+      "learning_rate": 1.4823619097949584e-05,
+      "loss": 0.0482,
+      "step": 2713
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3185716751967196,
+      "learning_rate": 1.4809876996508897e-05,
+      "loss": 0.0481,
+      "step": 2714
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3710762421342836,
+      "learning_rate": 1.4796137521821274e-05,
+      "loss": 0.0469,
+      "step": 2715
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2994532560113849,
+      "learning_rate": 1.4782400680840352e-05,
+      "loss": 0.0335,
+      "step": 2716
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.32841658125444484,
+      "learning_rate": 1.4768666480518432e-05,
+      "loss": 0.0407,
+      "step": 2717
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3524634367329856,
+      "learning_rate": 1.4754934927806473e-05,
+      "loss": 0.0466,
+      "step": 2718
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3246530817861148,
+      "learning_rate": 1.4741206029654098e-05,
+      "loss": 0.0337,
+      "step": 2719
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3791839722292734,
+      "learning_rate": 1.472747979300959e-05,
+      "loss": 0.0434,
+      "step": 2720
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.336279440002278,
+      "learning_rate": 1.4713756224819872e-05,
+      "loss": 0.0417,
+      "step": 2721
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2882385625480761,
+      "learning_rate": 1.4700035332030545e-05,
+      "loss": 0.0374,
+      "step": 2722
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.28726647419164664,
+      "learning_rate": 1.468631712158582e-05,
+      "loss": 0.044,
+      "step": 2723
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.49525735590255704,
+      "learning_rate": 1.4672601600428578e-05,
+      "loss": 0.0589,
+      "step": 2724
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3239146725506776,
+      "learning_rate": 1.465888877550032e-05,
+      "loss": 0.0361,
+      "step": 2725
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.31194783543268184,
+      "learning_rate": 1.4645178653741194e-05,
+      "loss": 0.0417,
+      "step": 2726
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.39087053211734085,
+      "learning_rate": 1.4631471242089978e-05,
+      "loss": 0.0435,
+      "step": 2727
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.34705907989168727,
+      "learning_rate": 1.4617766547484075e-05,
+      "loss": 0.0473,
+      "step": 2728
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.32899760973128733,
+      "learning_rate": 1.4604064576859513e-05,
+      "loss": 0.0309,
+      "step": 2729
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4117188795693934,
+      "learning_rate": 1.459036533715095e-05,
+      "loss": 0.0578,
+      "step": 2730
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.37993141000758085,
+      "learning_rate": 1.4576668835291654e-05,
+      "loss": 0.0381,
+      "step": 2731
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.35835120282232763,
+      "learning_rate": 1.4562975078213504e-05,
+      "loss": 0.0458,
+      "step": 2732
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4180931726097148,
+      "learning_rate": 1.4549284072846996e-05,
+      "loss": 0.0523,
+      "step": 2733
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.368766958848891,
+      "learning_rate": 1.4535595826121233e-05,
+      "loss": 0.0461,
+      "step": 2734
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.37347475448631995,
+      "learning_rate": 1.4521910344963918e-05,
+      "loss": 0.0523,
+      "step": 2735
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2811875141454974,
+      "learning_rate": 1.450822763630136e-05,
+      "loss": 0.0296,
+      "step": 2736
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3818052430755057,
+      "learning_rate": 1.4494547707058459e-05,
+      "loss": 0.0525,
+      "step": 2737
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2714308816365113,
+      "learning_rate": 1.4480870564158704e-05,
+      "loss": 0.0256,
+      "step": 2738
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3279851547325715,
+      "learning_rate": 1.44671962145242e-05,
+      "loss": 0.0396,
+      "step": 2739
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3168864080154981,
+      "learning_rate": 1.4453524665075607e-05,
+      "loss": 0.0375,
+      "step": 2740
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2895756110984442,
+      "learning_rate": 1.4439855922732182e-05,
+      "loss": 0.0352,
+      "step": 2741
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3121535049197099,
+      "learning_rate": 1.4426189994411756e-05,
+      "loss": 0.0447,
+      "step": 2742
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3011568251237467,
+      "learning_rate": 1.4412526887030745e-05,
+      "loss": 0.0447,
+      "step": 2743
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.34098788844481187,
+      "learning_rate": 1.4398866607504128e-05,
+      "loss": 0.0515,
+      "step": 2744
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3039500285882022,
+      "learning_rate": 1.4385209162745453e-05,
+      "loss": 0.035,
+      "step": 2745
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.36113540046057563,
+      "learning_rate": 1.4371554559666843e-05,
+      "loss": 0.0576,
+      "step": 2746
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4289938442926769,
+      "learning_rate": 1.4357902805178965e-05,
+      "loss": 0.0448,
+      "step": 2747
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.39897293689915525,
+      "learning_rate": 1.434425390619107e-05,
+      "loss": 0.0571,
+      "step": 2748
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3216986321138276,
+      "learning_rate": 1.4330607869610945e-05,
+      "loss": 0.0371,
+      "step": 2749
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4036541127822171,
+      "learning_rate": 1.431696470234493e-05,
+      "loss": 0.051,
+      "step": 2750
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.31114819926250076,
+      "learning_rate": 1.4303324411297918e-05,
+      "loss": 0.0492,
+      "step": 2751
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3387100998933343,
+      "learning_rate": 1.4289687003373342e-05,
+      "loss": 0.0468,
+      "step": 2752
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2328871549688841,
+      "learning_rate": 1.4276052485473177e-05,
+      "loss": 0.0239,
+      "step": 2753
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3202228653331459,
+      "learning_rate": 1.4262420864497939e-05,
+      "loss": 0.0432,
+      "step": 2754
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2696762524200947,
+      "learning_rate": 1.4248792147346668e-05,
+      "loss": 0.0317,
+      "step": 2755
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.31967655879116696,
+      "learning_rate": 1.4235166340916955e-05,
+      "loss": 0.045,
+      "step": 2756
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2970760568117321,
+      "learning_rate": 1.4221543452104891e-05,
+      "loss": 0.0308,
+      "step": 2757
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.42458892253619795,
+      "learning_rate": 1.4207923487805108e-05,
+      "loss": 0.0672,
+      "step": 2758
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.33230513328162714,
+      "learning_rate": 1.4194306454910757e-05,
+      "loss": 0.0397,
+      "step": 2759
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.42762986739777514,
+      "learning_rate": 1.4180692360313494e-05,
+      "loss": 0.0554,
+      "step": 2760
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.27931077295684037,
+      "learning_rate": 1.4167081210903501e-05,
+      "loss": 0.04,
+      "step": 2761
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.41547331587630704,
+      "learning_rate": 1.4153473013569468e-05,
+      "loss": 0.0458,
+      "step": 2762
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3026604592932078,
+      "learning_rate": 1.413986777519858e-05,
+      "loss": 0.0389,
+      "step": 2763
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.40036944821808274,
+      "learning_rate": 1.412626550267653e-05,
+      "loss": 0.0751,
+      "step": 2764
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.47896027439194794,
+      "learning_rate": 1.4112666202887522e-05,
+      "loss": 0.0722,
+      "step": 2765
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.27886565471312885,
+      "learning_rate": 1.4099069882714236e-05,
+      "loss": 0.038,
+      "step": 2766
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.30368425797912313,
+      "learning_rate": 1.4085476549037856e-05,
+      "loss": 0.0446,
+      "step": 2767
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2981647908149259,
+      "learning_rate": 1.4071886208738053e-05,
+      "loss": 0.0388,
+      "step": 2768
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.33278599396225034,
+      "learning_rate": 1.4058298868692979e-05,
+      "loss": 0.0512,
+      "step": 2769
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.33682190007711665,
+      "learning_rate": 1.4044714535779269e-05,
+      "loss": 0.0424,
+      "step": 2770
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.30438516973860114,
+      "learning_rate": 1.403113321687204e-05,
+      "loss": 0.0483,
+      "step": 2771
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.33630166165650016,
+      "learning_rate": 1.4017554918844872e-05,
+      "loss": 0.0401,
+      "step": 2772
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.340924135150723,
+      "learning_rate": 1.4003979648569839e-05,
+      "loss": 0.0489,
+      "step": 2773
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.28300769245153057,
+      "learning_rate": 1.3990407412917462e-05,
+      "loss": 0.0283,
+      "step": 2774
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3466167661491617,
+      "learning_rate": 1.3976838218756733e-05,
+      "loss": 0.0487,
+      "step": 2775
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3346830945121258,
+      "learning_rate": 1.3963272072955106e-05,
+      "loss": 0.0462,
+      "step": 2776
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.29863913799445185,
+      "learning_rate": 1.3949708982378487e-05,
+      "loss": 0.0415,
+      "step": 2777
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.253393359450197,
+      "learning_rate": 1.3936148953891242e-05,
+      "loss": 0.0321,
+      "step": 2778
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3637070570173247,
+      "learning_rate": 1.392259199435618e-05,
+      "loss": 0.0524,
+      "step": 2779
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3759873425808452,
+      "learning_rate": 1.3909038110634567e-05,
+      "loss": 0.0573,
+      "step": 2780
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.28541006262766017,
+      "learning_rate": 1.3895487309586097e-05,
+      "loss": 0.0416,
+      "step": 2781
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.40106802601906233,
+      "learning_rate": 1.388193959806893e-05,
+      "loss": 0.0618,
+      "step": 2782
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3662019652221195,
+      "learning_rate": 1.3868394982939636e-05,
+      "loss": 0.0489,
+      "step": 2783
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.34370404716597247,
+      "learning_rate": 1.3854853471053225e-05,
+      "loss": 0.0477,
+      "step": 2784
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3543753733624572,
+      "learning_rate": 1.3841315069263146e-05,
+      "loss": 0.0482,
+      "step": 2785
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.35525774719252445,
+      "learning_rate": 1.3827779784421262e-05,
+      "loss": 0.0406,
+      "step": 2786
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3808691561494414,
+      "learning_rate": 1.3814247623377868e-05,
+      "loss": 0.0578,
+      "step": 2787
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.23788466046753176,
+      "learning_rate": 1.3800718592981668e-05,
+      "loss": 0.0244,
+      "step": 2788
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3964649815819659,
+      "learning_rate": 1.3787192700079792e-05,
+      "loss": 0.0508,
+      "step": 2789
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.26043611771738767,
+      "learning_rate": 1.377366995151777e-05,
+      "loss": 0.0312,
+      "step": 2790
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.37760732145197373,
+      "learning_rate": 1.3760150354139558e-05,
+      "loss": 0.0401,
+      "step": 2791
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2952347920625057,
+      "learning_rate": 1.3746633914787504e-05,
+      "loss": 0.036,
+      "step": 2792
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.27096026759420505,
+      "learning_rate": 1.3733120640302358e-05,
+      "loss": 0.0276,
+      "step": 2793
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.28976262851872026,
+      "learning_rate": 1.3719610537523274e-05,
+      "loss": 0.0368,
+      "step": 2794
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2947502773848208,
+      "learning_rate": 1.3706103613287796e-05,
+      "loss": 0.0366,
+      "step": 2795
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.36182000985970075,
+      "learning_rate": 1.369259987443186e-05,
+      "loss": 0.0538,
+      "step": 2796
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2963917463149579,
+      "learning_rate": 1.3679099327789794e-05,
+      "loss": 0.0267,
+      "step": 2797
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3148398893701771,
+      "learning_rate": 1.3665601980194297e-05,
+      "loss": 0.0496,
+      "step": 2798
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3313389348355963,
+      "learning_rate": 1.3652107838476476e-05,
+      "loss": 0.0375,
+      "step": 2799
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2519413849217148,
+      "learning_rate": 1.3638616909465791e-05,
+      "loss": 0.0359,
+      "step": 2800
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.32790310486681473,
+      "learning_rate": 1.3625129199990083e-05,
+      "loss": 0.0364,
+      "step": 2801
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.37081440902313717,
+      "learning_rate": 1.3611644716875568e-05,
+      "loss": 0.061,
+      "step": 2802
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3476633542652982,
+      "learning_rate": 1.3598163466946823e-05,
+      "loss": 0.0408,
+      "step": 2803
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3780332911227297,
+      "learning_rate": 1.3584685457026789e-05,
+      "loss": 0.0504,
+      "step": 2804
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.31770609896556845,
+      "learning_rate": 1.3571210693936774e-05,
+      "loss": 0.0388,
+      "step": 2805
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.31575286182698364,
+      "learning_rate": 1.3557739184496435e-05,
+      "loss": 0.0437,
+      "step": 2806
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3594767177759614,
+      "learning_rate": 1.3544270935523778e-05,
+      "loss": 0.0447,
+      "step": 2807
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.36754325109400565,
+      "learning_rate": 1.3530805953835182e-05,
+      "loss": 0.0515,
+      "step": 2808
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4071580934482072,
+      "learning_rate": 1.351734424624535e-05,
+      "loss": 0.0578,
+      "step": 2809
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3646757576366893,
+      "learning_rate": 1.3503885819567335e-05,
+      "loss": 0.0547,
+      "step": 2810
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2684636603321318,
+      "learning_rate": 1.3490430680612528e-05,
+      "loss": 0.0354,
+      "step": 2811
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.28134491431539715,
+      "learning_rate": 1.3476978836190658e-05,
+      "loss": 0.0264,
+      "step": 2812
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.33385682598409755,
+      "learning_rate": 1.3463530293109783e-05,
+      "loss": 0.0379,
+      "step": 2813
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.32932744636579747,
+      "learning_rate": 1.34500850581763e-05,
+      "loss": 0.0367,
+      "step": 2814
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.37019697552233266,
+      "learning_rate": 1.3436643138194918e-05,
+      "loss": 0.0495,
+      "step": 2815
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.358547314897313,
+      "learning_rate": 1.3423204539968677e-05,
+      "loss": 0.0321,
+      "step": 2816
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2801113922919793,
+      "learning_rate": 1.3409769270298934e-05,
+      "loss": 0.029,
+      "step": 2817
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3508957261880292,
+      "learning_rate": 1.3396337335985361e-05,
+      "loss": 0.0336,
+      "step": 2818
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3909390712800989,
+      "learning_rate": 1.3382908743825947e-05,
+      "loss": 0.0522,
+      "step": 2819
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2928226106459401,
+      "learning_rate": 1.336948350061698e-05,
+      "loss": 0.0346,
+      "step": 2820
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2954126026903477,
+      "learning_rate": 1.335606161315306e-05,
+      "loss": 0.0261,
+      "step": 2821
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2717804015217529,
+      "learning_rate": 1.3342643088227085e-05,
+      "loss": 0.0398,
+      "step": 2822
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.33073288121978467,
+      "learning_rate": 1.3329227932630255e-05,
+      "loss": 0.0388,
+      "step": 2823
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.33370518308471864,
+      "learning_rate": 1.3315816153152055e-05,
+      "loss": 0.0439,
+      "step": 2824
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.30608677504159165,
+      "learning_rate": 1.3302407756580278e-05,
+      "loss": 0.0387,
+      "step": 2825
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.33089457778381837,
+      "learning_rate": 1.3289002749700992e-05,
+      "loss": 0.0325,
+      "step": 2826
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4467097997419248,
+      "learning_rate": 1.3275601139298556e-05,
+      "loss": 0.0457,
+      "step": 2827
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3706083636868158,
+      "learning_rate": 1.3262202932155602e-05,
+      "loss": 0.0484,
+      "step": 2828
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3404835345982751,
+      "learning_rate": 1.3248808135053048e-05,
+      "loss": 0.0498,
+      "step": 2829
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2949020555922916,
+      "learning_rate": 1.3235416754770082e-05,
+      "loss": 0.0475,
+      "step": 2830
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.377446064790745,
+      "learning_rate": 1.3222028798084165e-05,
+      "loss": 0.0534,
+      "step": 2831
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.32032310654361856,
+      "learning_rate": 1.3208644271771026e-05,
+      "loss": 0.0443,
+      "step": 2832
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.37997190318260876,
+      "learning_rate": 1.3195263182604638e-05,
+      "loss": 0.0436,
+      "step": 2833
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3834620822777257,
+      "learning_rate": 1.3181885537357277e-05,
+      "loss": 0.0493,
+      "step": 2834
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2942071173338933,
+      "learning_rate": 1.3168511342799444e-05,
+      "loss": 0.0337,
+      "step": 2835
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.30761294891828145,
+      "learning_rate": 1.3155140605699894e-05,
+      "loss": 0.0422,
+      "step": 2836
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3319158914665372,
+      "learning_rate": 1.3141773332825647e-05,
+      "loss": 0.0331,
+      "step": 2837
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.44578035773700425,
+      "learning_rate": 1.3128409530941957e-05,
+      "loss": 0.0581,
+      "step": 2838
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.32801149312387384,
+      "learning_rate": 1.3115049206812325e-05,
+      "loss": 0.0507,
+      "step": 2839
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.382789377565198,
+      "learning_rate": 1.3101692367198498e-05,
+      "loss": 0.0921,
+      "step": 2840
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.30446166447700507,
+      "learning_rate": 1.3088339018860439e-05,
+      "loss": 0.0435,
+      "step": 2841
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.39309102804653384,
+      "learning_rate": 1.307498916855638e-05,
+      "loss": 0.0453,
+      "step": 2842
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2857678139199307,
+      "learning_rate": 1.3061642823042757e-05,
+      "loss": 0.039,
+      "step": 2843
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3226794735584289,
+      "learning_rate": 1.3048299989074234e-05,
+      "loss": 0.0423,
+      "step": 2844
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.22917456866136676,
+      "learning_rate": 1.3034960673403699e-05,
+      "loss": 0.0292,
+      "step": 2845
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2897688766383443,
+      "learning_rate": 1.3021624882782262e-05,
+      "loss": 0.0356,
+      "step": 2846
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3024013533413087,
+      "learning_rate": 1.3008292623959253e-05,
+      "loss": 0.0429,
+      "step": 2847
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3064032443695261,
+      "learning_rate": 1.2994963903682205e-05,
+      "loss": 0.0392,
+      "step": 2848
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.30547864610100867,
+      "learning_rate": 1.2981638728696868e-05,
+      "loss": 0.0475,
+      "step": 2849
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3186415609393602,
+      "learning_rate": 1.2968317105747189e-05,
+      "loss": 0.043,
+      "step": 2850
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.32282427137613623,
+      "learning_rate": 1.2954999041575331e-05,
+      "loss": 0.0412,
+      "step": 2851
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3826593365122257,
+      "learning_rate": 1.2941684542921646e-05,
+      "loss": 0.0493,
+      "step": 2852
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2961884893065571,
+      "learning_rate": 1.2928373616524682e-05,
+      "loss": 0.0288,
+      "step": 2853
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3475935590133547,
+      "learning_rate": 1.291506626912118e-05,
+      "loss": 0.0475,
+      "step": 2854
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.35536169179632393,
+      "learning_rate": 1.290176250744607e-05,
+      "loss": 0.0425,
+      "step": 2855
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3053531955003352,
+      "learning_rate": 1.2888462338232466e-05,
+      "loss": 0.0371,
+      "step": 2856
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2528767475118705,
+      "learning_rate": 1.287516576821167e-05,
+      "loss": 0.0245,
+      "step": 2857
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2888324182721138,
+      "learning_rate": 1.2861872804113154e-05,
+      "loss": 0.0376,
+      "step": 2858
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3111029151397963,
+      "learning_rate": 1.284858345266456e-05,
+      "loss": 0.0434,
+      "step": 2859
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.42479953105033164,
+      "learning_rate": 1.2835297720591729e-05,
+      "loss": 0.0472,
+      "step": 2860
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2913783456363997,
+      "learning_rate": 1.282201561461864e-05,
+      "loss": 0.0347,
+      "step": 2861
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.37374209985708284,
+      "learning_rate": 1.2808737141467451e-05,
+      "loss": 0.0481,
+      "step": 2862
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3775900958717164,
+      "learning_rate": 1.2795462307858478e-05,
+      "loss": 0.0506,
+      "step": 2863
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.31289014304287066,
+      "learning_rate": 1.2782191120510196e-05,
+      "loss": 0.0361,
+      "step": 2864
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2958720733681559,
+      "learning_rate": 1.2768923586139232e-05,
+      "loss": 0.0393,
+      "step": 2865
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3559906869055891,
+      "learning_rate": 1.275565971146037e-05,
+      "loss": 0.048,
+      "step": 2866
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2940382066111115,
+      "learning_rate": 1.2742399503186528e-05,
+      "loss": 0.0433,
+      "step": 2867
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3047524162975901,
+      "learning_rate": 1.2729142968028793e-05,
+      "loss": 0.0368,
+      "step": 2868
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.28501375112905175,
+      "learning_rate": 1.2715890112696379e-05,
+      "loss": 0.0292,
+      "step": 2869
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2598049162784958,
+      "learning_rate": 1.2702640943896625e-05,
+      "loss": 0.0312,
+      "step": 2870
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3311882312332492,
+      "learning_rate": 1.2689395468335027e-05,
+      "loss": 0.0424,
+      "step": 2871
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3100638922291191,
+      "learning_rate": 1.2676153692715195e-05,
+      "loss": 0.0322,
+      "step": 2872
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3158769536699534,
+      "learning_rate": 1.2662915623738874e-05,
+      "loss": 0.0401,
+      "step": 2873
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3905736317237826,
+      "learning_rate": 1.2649681268105933e-05,
+      "loss": 0.0494,
+      "step": 2874
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3623962492290295,
+      "learning_rate": 1.263645063251436e-05,
+      "loss": 0.0494,
+      "step": 2875
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3047706172742518,
+      "learning_rate": 1.2623223723660258e-05,
+      "loss": 0.0402,
+      "step": 2876
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.29039474153923495,
+      "learning_rate": 1.2610000548237851e-05,
+      "loss": 0.0345,
+      "step": 2877
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.42741917183463485,
+      "learning_rate": 1.259678111293947e-05,
+      "loss": 0.0456,
+      "step": 2878
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.36327702509566523,
+      "learning_rate": 1.2583565424455552e-05,
+      "loss": 0.0591,
+      "step": 2879
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.25609254768094003,
+      "learning_rate": 1.2570353489474637e-05,
+      "loss": 0.0385,
+      "step": 2880
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3580431515431179,
+      "learning_rate": 1.2557145314683364e-05,
+      "loss": 0.0451,
+      "step": 2881
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.24723126894418496,
+      "learning_rate": 1.254394090676647e-05,
+      "loss": 0.0337,
+      "step": 2882
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.24767297062551186,
+      "learning_rate": 1.2530740272406792e-05,
+      "loss": 0.0307,
+      "step": 2883
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2776789184659262,
+      "learning_rate": 1.2517543418285247e-05,
+      "loss": 0.0334,
+      "step": 2884
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.32881618544241237,
+      "learning_rate": 1.2504350351080845e-05,
+      "loss": 0.0403,
+      "step": 2885
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3183871649098128,
+      "learning_rate": 1.2491161077470682e-05,
+      "loss": 0.0448,
+      "step": 2886
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2888848042097438,
+      "learning_rate": 1.2477975604129929e-05,
+      "loss": 0.0389,
+      "step": 2887
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3045754593476319,
+      "learning_rate": 1.2464793937731831e-05,
+      "loss": 0.0383,
+      "step": 2888
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.42008218588417245,
+      "learning_rate": 1.2451616084947714e-05,
+      "loss": 0.0645,
+      "step": 2889
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2813345199320012,
+      "learning_rate": 1.243844205244697e-05,
+      "loss": 0.0334,
+      "step": 2890
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.37685954459211185,
+      "learning_rate": 1.2425271846897053e-05,
+      "loss": 0.0558,
+      "step": 2891
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4008627736124671,
+      "learning_rate": 1.2412105474963491e-05,
+      "loss": 0.0533,
+      "step": 2892
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4143105789384026,
+      "learning_rate": 1.2398942943309855e-05,
+      "loss": 0.066,
+      "step": 2893
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.34944816154299846,
+      "learning_rate": 1.2385784258597796e-05,
+      "loss": 0.0427,
+      "step": 2894
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3303130854129566,
+      "learning_rate": 1.2372629427487e-05,
+      "loss": 0.05,
+      "step": 2895
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2697725459850731,
+      "learning_rate": 1.235947845663521e-05,
+      "loss": 0.027,
+      "step": 2896
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3182834932663021,
+      "learning_rate": 1.2346331352698206e-05,
+      "loss": 0.0363,
+      "step": 2897
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.24598310647552704,
+      "learning_rate": 1.2333188122329824e-05,
+      "loss": 0.0408,
+      "step": 2898
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.29446130119261393,
+      "learning_rate": 1.2320048772181932e-05,
+      "loss": 0.0401,
+      "step": 2899
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.27915071784669715,
+      "learning_rate": 1.2306913308904435e-05,
+      "loss": 0.0265,
+      "step": 2900
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.31153922205430945,
+      "learning_rate": 1.2293781739145274e-05,
+      "loss": 0.0347,
+      "step": 2901
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2740717752120284,
+      "learning_rate": 1.2280654069550404e-05,
+      "loss": 0.0375,
+      "step": 2902
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.32661056712366027,
+      "learning_rate": 1.2267530306763837e-05,
+      "loss": 0.0329,
+      "step": 2903
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.25680366284318407,
+      "learning_rate": 1.2254410457427581e-05,
+      "loss": 0.0272,
+      "step": 2904
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.40490566249909893,
+      "learning_rate": 1.2241294528181678e-05,
+      "loss": 0.0406,
+      "step": 2905
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.24799753434259125,
+      "learning_rate": 1.2228182525664175e-05,
+      "loss": 0.0265,
+      "step": 2906
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.32278382224724494,
+      "learning_rate": 1.2215074456511136e-05,
+      "loss": 0.0349,
+      "step": 2907
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2880573193834743,
+      "learning_rate": 1.2201970327356639e-05,
+      "loss": 0.0378,
+      "step": 2908
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3387098689255451,
+      "learning_rate": 1.2188870144832758e-05,
+      "loss": 0.0433,
+      "step": 2909
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3689953957324699,
+      "learning_rate": 1.217577391556958e-05,
+      "loss": 0.0433,
+      "step": 2910
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.28615291381637514,
+      "learning_rate": 1.2162681646195187e-05,
+      "loss": 0.0269,
+      "step": 2911
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.37840625375641673,
+      "learning_rate": 1.2149593343335658e-05,
+      "loss": 0.0417,
+      "step": 2912
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.33880744429424287,
+      "learning_rate": 1.2136509013615063e-05,
+      "loss": 0.0418,
+      "step": 2913
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2916809260197858,
+      "learning_rate": 1.2123428663655457e-05,
+      "loss": 0.0345,
+      "step": 2914
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.349445347161046,
+      "learning_rate": 1.211035230007689e-05,
+      "loss": 0.0398,
+      "step": 2915
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.30144971408574184,
+      "learning_rate": 1.209727992949739e-05,
+      "loss": 0.0345,
+      "step": 2916
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.28327437195210436,
+      "learning_rate": 1.2084211558532958e-05,
+      "loss": 0.0269,
+      "step": 2917
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.35992758188097324,
+      "learning_rate": 1.2071147193797578e-05,
+      "loss": 0.0456,
+      "step": 2918
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3139806835397025,
+      "learning_rate": 1.2058086841903211e-05,
+      "loss": 0.0362,
+      "step": 2919
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.33326614964467255,
+      "learning_rate": 1.204503050945978e-05,
+      "loss": 0.0426,
+      "step": 2920
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.26053383893253446,
+      "learning_rate": 1.2031978203075172e-05,
+      "loss": 0.032,
+      "step": 2921
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2218677210621221,
+      "learning_rate": 1.2018929929355241e-05,
+      "loss": 0.0185,
+      "step": 2922
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4643167666122263,
+      "learning_rate": 1.2005885694903796e-05,
+      "loss": 0.0613,
+      "step": 2923
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3133421636518249,
+      "learning_rate": 1.1992845506322607e-05,
+      "loss": 0.0306,
+      "step": 2924
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3368456784936427,
+      "learning_rate": 1.1979809370211392e-05,
+      "loss": 0.0441,
+      "step": 2925
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3515826537395515,
+      "learning_rate": 1.196677729316782e-05,
+      "loss": 0.0424,
+      "step": 2926
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.32891554060474937,
+      "learning_rate": 1.1953749281787502e-05,
+      "loss": 0.0388,
+      "step": 2927
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.383314698567998,
+      "learning_rate": 1.194072534266399e-05,
+      "loss": 0.0481,
+      "step": 2928
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2887688253301843,
+      "learning_rate": 1.1927705482388794e-05,
+      "loss": 0.0403,
+      "step": 2929
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3175047889483473,
+      "learning_rate": 1.1914689707551337e-05,
+      "loss": 0.0392,
+      "step": 2930
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3006459960029491,
+      "learning_rate": 1.1901678024738983e-05,
+      "loss": 0.0337,
+      "step": 2931
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3369660599983346,
+      "learning_rate": 1.1888670440537025e-05,
+      "loss": 0.0335,
+      "step": 2932
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.32840089325379707,
+      "learning_rate": 1.1875666961528679e-05,
+      "loss": 0.0336,
+      "step": 2933
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.31666973116713637,
+      "learning_rate": 1.1862667594295086e-05,
+      "loss": 0.043,
+      "step": 2934
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.28063185632481785,
+      "learning_rate": 1.1849672345415306e-05,
+      "loss": 0.0425,
+      "step": 2935
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.34532325693352134,
+      "learning_rate": 1.1836681221466308e-05,
+      "loss": 0.0393,
+      "step": 2936
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.30076662250967534,
+      "learning_rate": 1.1823694229022995e-05,
+      "loss": 0.0391,
+      "step": 2937
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3062019003833736,
+      "learning_rate": 1.181071137465815e-05,
+      "loss": 0.035,
+      "step": 2938
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.37222397001806645,
+      "learning_rate": 1.1797732664942481e-05,
+      "loss": 0.0744,
+      "step": 2939
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2864970600075523,
+      "learning_rate": 1.1784758106444594e-05,
+      "loss": 0.039,
+      "step": 2940
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2879502575008263,
+      "learning_rate": 1.1771787705730983e-05,
+      "loss": 0.0304,
+      "step": 2941
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.31231118697034904,
+      "learning_rate": 1.175882146936606e-05,
+      "loss": 0.0406,
+      "step": 2942
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3410455347601863,
+      "learning_rate": 1.1745859403912108e-05,
+      "loss": 0.0366,
+      "step": 2943
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.30274079838983087,
+      "learning_rate": 1.1732901515929312e-05,
+      "loss": 0.0391,
+      "step": 2944
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3174696348718507,
+      "learning_rate": 1.1719947811975732e-05,
+      "loss": 0.0403,
+      "step": 2945
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.31613914376879376,
+      "learning_rate": 1.1706998298607325e-05,
+      "loss": 0.0428,
+      "step": 2946
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.43372456407309057,
+      "learning_rate": 1.1694052982377915e-05,
+      "loss": 0.0508,
+      "step": 2947
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.27547772700698475,
+      "learning_rate": 1.1681111869839209e-05,
+      "loss": 0.0354,
+      "step": 2948
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.23440111332696512,
+      "learning_rate": 1.166817496754078e-05,
+      "loss": 0.0248,
+      "step": 2949
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6419558309085079,
+      "learning_rate": 1.1655242282030068e-05,
+      "loss": 0.0934,
+      "step": 2950
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4207470437930287,
+      "learning_rate": 1.1642313819852405e-05,
+      "loss": 0.0574,
+      "step": 2951
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2677592757227379,
+      "learning_rate": 1.1629389587550939e-05,
+      "loss": 0.0438,
+      "step": 2952
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.30039862038000953,
+      "learning_rate": 1.1616469591666725e-05,
+      "loss": 0.0329,
+      "step": 2953
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2554436429253239,
+      "learning_rate": 1.1603553838738635e-05,
+      "loss": 0.0307,
+      "step": 2954
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3277463772091227,
+      "learning_rate": 1.1590642335303417e-05,
+      "loss": 0.0348,
+      "step": 2955
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.28809162716177045,
+      "learning_rate": 1.1577735087895664e-05,
+      "loss": 0.0267,
+      "step": 2956
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2916931356045432,
+      "learning_rate": 1.1564832103047818e-05,
+      "loss": 0.0361,
+      "step": 2957
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2786283695535751,
+      "learning_rate": 1.1551933387290149e-05,
+      "loss": 0.0338,
+      "step": 2958
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3756814169764887,
+      "learning_rate": 1.1539038947150783e-05,
+      "loss": 0.0556,
+      "step": 2959
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3044743158896093,
+      "learning_rate": 1.152614878915567e-05,
+      "loss": 0.0357,
+      "step": 2960
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2520853479294366,
+      "learning_rate": 1.1513262919828603e-05,
+      "loss": 0.0318,
+      "step": 2961
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.30014479788558823,
+      "learning_rate": 1.1500381345691192e-05,
+      "loss": 0.0374,
+      "step": 2962
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.37943443373256613,
+      "learning_rate": 1.1487504073262886e-05,
+      "loss": 0.0519,
+      "step": 2963
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.26828896988529727,
+      "learning_rate": 1.1474631109060957e-05,
+      "loss": 0.0327,
+      "step": 2964
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3523671584378296,
+      "learning_rate": 1.1461762459600476e-05,
+      "loss": 0.0442,
+      "step": 2965
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.26789693565200207,
+      "learning_rate": 1.1448898131394364e-05,
+      "loss": 0.0293,
+      "step": 2966
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2616347989618018,
+      "learning_rate": 1.1436038130953317e-05,
+      "loss": 0.0356,
+      "step": 2967
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.23439360385810168,
+      "learning_rate": 1.142318246478588e-05,
+      "loss": 0.0297,
+      "step": 2968
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.27644543837366153,
+      "learning_rate": 1.1410331139398365e-05,
+      "loss": 0.0425,
+      "step": 2969
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.31879390713338973,
+      "learning_rate": 1.1397484161294924e-05,
+      "loss": 0.0387,
+      "step": 2970
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2862545054093661,
+      "learning_rate": 1.138464153697747e-05,
+      "loss": 0.0386,
+      "step": 2971
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3208737689501007,
+      "learning_rate": 1.1371803272945759e-05,
+      "loss": 0.0434,
+      "step": 2972
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3828787259796405,
+      "learning_rate": 1.1358969375697297e-05,
+      "loss": 0.0419,
+      "step": 2973
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.34019136737319144,
+      "learning_rate": 1.1346139851727412e-05,
+      "loss": 0.0394,
+      "step": 2974
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.28714105859439243,
+      "learning_rate": 1.1333314707529188e-05,
+      "loss": 0.0325,
+      "step": 2975
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.24338669906714824,
+      "learning_rate": 1.1320493949593528e-05,
+      "loss": 0.0251,
+      "step": 2976
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.23301690332500383,
+      "learning_rate": 1.1307677584409076e-05,
+      "loss": 0.0299,
+      "step": 2977
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3014458336750867,
+      "learning_rate": 1.1294865618462294e-05,
+      "loss": 0.0355,
+      "step": 2978
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2848679116820537,
+      "learning_rate": 1.128205805823737e-05,
+      "loss": 0.0279,
+      "step": 2979
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3149477107873396,
+      "learning_rate": 1.1269254910216316e-05,
+      "loss": 0.0406,
+      "step": 2980
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3981554491624712,
+      "learning_rate": 1.1256456180878867e-05,
+      "loss": 0.0541,
+      "step": 2981
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3531659752357583,
+      "learning_rate": 1.1243661876702552e-05,
+      "loss": 0.0536,
+      "step": 2982
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3250805534889989,
+      "learning_rate": 1.1230872004162631e-05,
+      "loss": 0.0537,
+      "step": 2983
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2604980611756454,
+      "learning_rate": 1.1218086569732152e-05,
+      "loss": 0.0323,
+      "step": 2984
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.26375732450122863,
+      "learning_rate": 1.1205305579881883e-05,
+      "loss": 0.0361,
+      "step": 2985
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2597553996891595,
+      "learning_rate": 1.1192529041080382e-05,
+      "loss": 0.0243,
+      "step": 2986
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.30060744527767147,
+      "learning_rate": 1.1179756959793918e-05,
+      "loss": 0.0306,
+      "step": 2987
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.35249120117934213,
+      "learning_rate": 1.1166989342486524e-05,
+      "loss": 0.0486,
+      "step": 2988
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.26966911052098275,
+      "learning_rate": 1.1154226195619979e-05,
+      "loss": 0.0249,
+      "step": 2989
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3111691508172709,
+      "learning_rate": 1.1141467525653773e-05,
+      "loss": 0.0332,
+      "step": 2990
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2694851241517991,
+      "learning_rate": 1.1128713339045162e-05,
+      "loss": 0.0269,
+      "step": 2991
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4213832001964649,
+      "learning_rate": 1.1115963642249107e-05,
+      "loss": 0.0444,
+      "step": 2992
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2772025299513124,
+      "learning_rate": 1.110321844171832e-05,
+      "loss": 0.0401,
+      "step": 2993
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3647805520649611,
+      "learning_rate": 1.1090477743903212e-05,
+      "loss": 0.0505,
+      "step": 2994
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.26937476051921777,
+      "learning_rate": 1.1077741555251938e-05,
+      "loss": 0.0311,
+      "step": 2995
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2771986729672004,
+      "learning_rate": 1.1065009882210352e-05,
+      "loss": 0.0404,
+      "step": 2996
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2800261839050899,
+      "learning_rate": 1.1052282731222035e-05,
+      "loss": 0.0341,
+      "step": 2997
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.24857849312844407,
+      "learning_rate": 1.1039560108728277e-05,
+      "loss": 0.0315,
+      "step": 2998
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3018654689903294,
+      "learning_rate": 1.1026842021168088e-05,
+      "loss": 0.032,
+      "step": 2999
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3464088564074809,
+      "learning_rate": 1.101412847497815e-05,
+      "loss": 0.0424,
+      "step": 3000
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3691593750727993,
+      "learning_rate": 1.100141947659288e-05,
+      "loss": 0.0518,
+      "step": 3001
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3224610978512565,
+      "learning_rate": 1.0988715032444369e-05,
+      "loss": 0.0417,
+      "step": 3002
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2211300173153049,
+      "learning_rate": 1.0976015148962427e-05,
+      "loss": 0.0245,
+      "step": 3003
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3657974757425489,
+      "learning_rate": 1.0963319832574528e-05,
+      "loss": 0.0489,
+      "step": 3004
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2633445827705902,
+      "learning_rate": 1.0950629089705857e-05,
+      "loss": 0.028,
+      "step": 3005
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.24596810539431577,
+      "learning_rate": 1.0937942926779279e-05,
+      "loss": 0.0301,
+      "step": 3006
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.266475549896758,
+      "learning_rate": 1.0925261350215344e-05,
+      "loss": 0.0342,
+      "step": 3007
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.31072261324263634,
+      "learning_rate": 1.091258436643226e-05,
+      "loss": 0.0427,
+      "step": 3008
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2857607873897929,
+      "learning_rate": 1.0899911981845946e-05,
+      "loss": 0.038,
+      "step": 3009
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.24947990020506072,
+      "learning_rate": 1.0887244202869951e-05,
+      "loss": 0.0364,
+      "step": 3010
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.27201095119687674,
+      "learning_rate": 1.0874581035915534e-05,
+      "loss": 0.0297,
+      "step": 3011
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2753170131482302,
+      "learning_rate": 1.0861922487391588e-05,
+      "loss": 0.0312,
+      "step": 3012
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.29611178219925743,
+      "learning_rate": 1.0849268563704696e-05,
+      "loss": 0.034,
+      "step": 3013
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3435229722174912,
+      "learning_rate": 1.0836619271259072e-05,
+      "loss": 0.0441,
+      "step": 3014
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2658495654982007,
+      "learning_rate": 1.0823974616456607e-05,
+      "loss": 0.0338,
+      "step": 3015
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.29094652981374264,
+      "learning_rate": 1.0811334605696837e-05,
+      "loss": 0.0371,
+      "step": 3016
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.252277608889838,
+      "learning_rate": 1.0798699245376959e-05,
+      "loss": 0.0273,
+      "step": 3017
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2975109341359133,
+      "learning_rate": 1.078606854189179e-05,
+      "loss": 0.0383,
+      "step": 3018
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.36072595267535207,
+      "learning_rate": 1.0773442501633822e-05,
+      "loss": 0.0433,
+      "step": 3019
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3628036271553774,
+      "learning_rate": 1.0760821130993157e-05,
+      "loss": 0.0625,
+      "step": 3020
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3527759379658487,
+      "learning_rate": 1.0748204436357562e-05,
+      "loss": 0.0415,
+      "step": 3021
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.37273677699622976,
+      "learning_rate": 1.0735592424112404e-05,
+      "loss": 0.0464,
+      "step": 3022
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3031672797261527,
+      "learning_rate": 1.0722985100640717e-05,
+      "loss": 0.0437,
+      "step": 3023
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4999697348256345,
+      "learning_rate": 1.0710382472323145e-05,
+      "loss": 0.099,
+      "step": 3024
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.29739183060223257,
+      "learning_rate": 1.0697784545537943e-05,
+      "loss": 0.0391,
+      "step": 3025
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3408796045707868,
+      "learning_rate": 1.0685191326661015e-05,
+      "loss": 0.0495,
+      "step": 3026
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3168906283615832,
+      "learning_rate": 1.0672602822065845e-05,
+      "loss": 0.0412,
+      "step": 3027
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.22778330125223972,
+      "learning_rate": 1.0660019038123577e-05,
+      "loss": 0.0265,
+      "step": 3028
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.31655710650685553,
+      "learning_rate": 1.0647439981202918e-05,
+      "loss": 0.0323,
+      "step": 3029
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2903265250999385,
+      "learning_rate": 1.0634865657670227e-05,
+      "loss": 0.0455,
+      "step": 3030
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2576086654583902,
+      "learning_rate": 1.0622296073889417e-05,
+      "loss": 0.0364,
+      "step": 3031
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2611284573747479,
+      "learning_rate": 1.0609731236222069e-05,
+      "loss": 0.0346,
+      "step": 3032
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2360144101401728,
+      "learning_rate": 1.0597171151027297e-05,
+      "loss": 0.0349,
+      "step": 3033
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2767069046268152,
+      "learning_rate": 1.058461582466185e-05,
+      "loss": 0.036,
+      "step": 3034
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.29520680387895365,
+      "learning_rate": 1.0572065263480046e-05,
+      "loss": 0.0328,
+      "step": 3035
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2730848217792552,
+      "learning_rate": 1.0559519473833815e-05,
+      "loss": 0.0235,
+      "step": 3036
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3385652741735248,
+      "learning_rate": 1.0546978462072642e-05,
+      "loss": 0.0386,
+      "step": 3037
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2996355629707367,
+      "learning_rate": 1.0534442234543623e-05,
+      "loss": 0.0436,
+      "step": 3038
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3374086702363696,
+      "learning_rate": 1.0521910797591408e-05,
+      "loss": 0.0393,
+      "step": 3039
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.34725073405930423,
+      "learning_rate": 1.0509384157558236e-05,
+      "loss": 0.0497,
+      "step": 3040
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3113323809669054,
+      "learning_rate": 1.0496862320783926e-05,
+      "loss": 0.043,
+      "step": 3041
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.32109315459110344,
+      "learning_rate": 1.0484345293605853e-05,
+      "loss": 0.0342,
+      "step": 3042
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2767139456583379,
+      "learning_rate": 1.0471833082358954e-05,
+      "loss": 0.0296,
+      "step": 3043
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3111162424314804,
+      "learning_rate": 1.0459325693375746e-05,
+      "loss": 0.042,
+      "step": 3044
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3189332388597514,
+      "learning_rate": 1.0446823132986283e-05,
+      "loss": 0.0361,
+      "step": 3045
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.31497175676644285,
+      "learning_rate": 1.0434325407518204e-05,
+      "loss": 0.0347,
+      "step": 3046
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2779880749191047,
+      "learning_rate": 1.0421832523296665e-05,
+      "loss": 0.0261,
+      "step": 3047
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.31046564011343325,
+      "learning_rate": 1.04093444866444e-05,
+      "loss": 0.0403,
+      "step": 3048
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2831393860747274,
+      "learning_rate": 1.0396861303881691e-05,
+      "loss": 0.0286,
+      "step": 3049
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2727575294371441,
+      "learning_rate": 1.0384382981326336e-05,
+      "loss": 0.0359,
+      "step": 3050
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3663008312520523,
+      "learning_rate": 1.0371909525293709e-05,
+      "loss": 0.0523,
+      "step": 3051
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.30116656010510984,
+      "learning_rate": 1.0359440942096682e-05,
+      "loss": 0.0309,
+      "step": 3052
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3353678619920413,
+      "learning_rate": 1.0346977238045699e-05,
+      "loss": 0.0397,
+      "step": 3053
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3399476298151822,
+      "learning_rate": 1.0334518419448703e-05,
+      "loss": 0.0396,
+      "step": 3054
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.23916380456187727,
+      "learning_rate": 1.0322064492611195e-05,
+      "loss": 0.0299,
+      "step": 3055
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.25943841618555563,
+      "learning_rate": 1.0309615463836162e-05,
+      "loss": 0.0241,
+      "step": 3056
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3557983798400247,
+      "learning_rate": 1.0297171339424148e-05,
+      "loss": 0.049,
+      "step": 3057
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.324009885498815,
+      "learning_rate": 1.0284732125673198e-05,
+      "loss": 0.0409,
+      "step": 3058
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.20160362060773568,
+      "learning_rate": 1.0272297828878881e-05,
+      "loss": 0.0209,
+      "step": 3059
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.26653484585635645,
+      "learning_rate": 1.0259868455334259e-05,
+      "loss": 0.0284,
+      "step": 3060
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2883275328288837,
+      "learning_rate": 1.0247444011329928e-05,
+      "loss": 0.0343,
+      "step": 3061
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.30784072020467396,
+      "learning_rate": 1.0235024503153956e-05,
+      "loss": 0.0337,
+      "step": 3062
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.39213557944528954,
+      "learning_rate": 1.0222609937091952e-05,
+      "loss": 0.0571,
+      "step": 3063
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.35248830540852255,
+      "learning_rate": 1.0210200319426988e-05,
+      "loss": 0.0289,
+      "step": 3064
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.24872786988966386,
+      "learning_rate": 1.0197795656439662e-05,
+      "loss": 0.0281,
+      "step": 3065
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3033114439975192,
+      "learning_rate": 1.0185395954408031e-05,
+      "loss": 0.0372,
+      "step": 3066
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.28971036273483797,
+      "learning_rate": 1.0173001219607683e-05,
+      "loss": 0.0481,
+      "step": 3067
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3235645357995982,
+      "learning_rate": 1.0160611458311651e-05,
+      "loss": 0.0369,
+      "step": 3068
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.405914241250726,
+      "learning_rate": 1.0148226676790482e-05,
+      "loss": 0.047,
+      "step": 3069
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.33128543565256136,
+      "learning_rate": 1.013584688131218e-05,
+      "loss": 0.0559,
+      "step": 3070
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2835106988154603,
+      "learning_rate": 1.0123472078142248e-05,
+      "loss": 0.0301,
+      "step": 3071
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3415746784045918,
+      "learning_rate": 1.011110227354363e-05,
+      "loss": 0.047,
+      "step": 3072
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2912476657806638,
+      "learning_rate": 1.0098737473776781e-05,
+      "loss": 0.0398,
+      "step": 3073
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.24532364133573092,
+      "learning_rate": 1.0086377685099578e-05,
+      "loss": 0.0244,
+      "step": 3074
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.32993961565654917,
+      "learning_rate": 1.0074022913767411e-05,
+      "loss": 0.0449,
+      "step": 3075
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3160682568379425,
+      "learning_rate": 1.006167316603309e-05,
+      "loss": 0.042,
+      "step": 3076
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.36620874022778316,
+      "learning_rate": 1.0049328448146908e-05,
+      "loss": 0.0427,
+      "step": 3077
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2819488901785845,
+      "learning_rate": 1.0036988766356592e-05,
+      "loss": 0.0289,
+      "step": 3078
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3035661039266114,
+      "learning_rate": 1.0024654126907343e-05,
+      "loss": 0.0415,
+      "step": 3079
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2880209639389726,
+      "learning_rate": 1.0012324536041781e-05,
+      "loss": 0.0335,
+      "step": 3080
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2384137582520422,
+      "learning_rate": 1.0000000000000006e-05,
+      "loss": 0.0306,
+      "step": 3081
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3243717499020729,
+      "learning_rate": 9.987680525019521e-06,
+      "loss": 0.0373,
+      "step": 3082
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.37508876067967445,
+      "learning_rate": 9.975366117335301e-06,
+      "loss": 0.0556,
+      "step": 3083
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.25705657300879875,
+      "learning_rate": 9.96305678317975e-06,
+      "loss": 0.0302,
+      "step": 3084
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2486956514718051,
+      "learning_rate": 9.950752528782679e-06,
+      "loss": 0.0277,
+      "step": 3085
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.31257510474818884,
+      "learning_rate": 9.938453360371363e-06,
+      "loss": 0.0405,
+      "step": 3086
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2958070972835022,
+      "learning_rate": 9.926159284170471e-06,
+      "loss": 0.0378,
+      "step": 3087
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2978037311107582,
+      "learning_rate": 9.913870306402129e-06,
+      "loss": 0.0359,
+      "step": 3088
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.24685634047575178,
+      "learning_rate": 9.901586433285845e-06,
+      "loss": 0.0246,
+      "step": 3089
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.29914513770600243,
+      "learning_rate": 9.889307671038579e-06,
+      "loss": 0.0297,
+      "step": 3090
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.25373522060085896,
+      "learning_rate": 9.877034025874675e-06,
+      "loss": 0.0279,
+      "step": 3091
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.28816754753091417,
+      "learning_rate": 9.864765504005901e-06,
+      "loss": 0.0401,
+      "step": 3092
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2882007434774494,
+      "learning_rate": 9.852502111641438e-06,
+      "loss": 0.0426,
+      "step": 3093
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.25232177201688216,
+      "learning_rate": 9.840243854987868e-06,
+      "loss": 0.0345,
+      "step": 3094
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.24840152348033523,
+      "learning_rate": 9.827990740249156e-06,
+      "loss": 0.03,
+      "step": 3095
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2506674410256598,
+      "learning_rate": 9.815742773626693e-06,
+      "loss": 0.0342,
+      "step": 3096
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.23513415096530796,
+      "learning_rate": 9.803499961319234e-06,
+      "loss": 0.0199,
+      "step": 3097
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3594340193010657,
+      "learning_rate": 9.791262309522959e-06,
+      "loss": 0.0534,
+      "step": 3098
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.26262614250660854,
+      "learning_rate": 9.779029824431403e-06,
+      "loss": 0.0328,
+      "step": 3099
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3819389106850532,
+      "learning_rate": 9.766802512235507e-06,
+      "loss": 0.0734,
+      "step": 3100
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3023854983385063,
+      "learning_rate": 9.75458037912359e-06,
+      "loss": 0.0339,
+      "step": 3101
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3121884819870665,
+      "learning_rate": 9.742363431281356e-06,
+      "loss": 0.0437,
+      "step": 3102
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.25550331208570076,
+      "learning_rate": 9.73015167489186e-06,
+      "loss": 0.0253,
+      "step": 3103
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3430055880911585,
+      "learning_rate": 9.717945116135568e-06,
+      "loss": 0.0373,
+      "step": 3104
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.31624401313030304,
+      "learning_rate": 9.705743761190273e-06,
+      "loss": 0.0406,
+      "step": 3105
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2834614200655747,
+      "learning_rate": 9.693547616231173e-06,
+      "loss": 0.0361,
+      "step": 3106
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3513387179755282,
+      "learning_rate": 9.681356687430798e-06,
+      "loss": 0.0434,
+      "step": 3107
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.28196666062468867,
+      "learning_rate": 9.669170980959063e-06,
+      "loss": 0.0335,
+      "step": 3108
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.33908513780252647,
+      "learning_rate": 9.656990502983216e-06,
+      "loss": 0.0317,
+      "step": 3109
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.313221265991117,
+      "learning_rate": 9.644815259667881e-06,
+      "loss": 0.035,
+      "step": 3110
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.26410877623606116,
+      "learning_rate": 9.632645257175027e-06,
+      "loss": 0.0354,
+      "step": 3111
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.28774523227496873,
+      "learning_rate": 9.620480501663954e-06,
+      "loss": 0.0274,
+      "step": 3112
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3178335188120151,
+      "learning_rate": 9.608320999291333e-06,
+      "loss": 0.0353,
+      "step": 3113
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3229114765344905,
+      "learning_rate": 9.59616675621115e-06,
+      "loss": 0.0352,
+      "step": 3114
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.27446791096771567,
+      "learning_rate": 9.58401777857475e-06,
+      "loss": 0.026,
+      "step": 3115
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.28892722701999873,
+      "learning_rate": 9.571874072530809e-06,
+      "loss": 0.0355,
+      "step": 3116
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.31541426299089276,
+      "learning_rate": 9.559735644225316e-06,
+      "loss": 0.0263,
+      "step": 3117
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.25088864578936493,
+      "learning_rate": 9.547602499801616e-06,
+      "loss": 0.026,
+      "step": 3118
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.30759405624372854,
+      "learning_rate": 9.53547464540037e-06,
+      "loss": 0.0459,
+      "step": 3119
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2569584862348946,
+      "learning_rate": 9.523352087159548e-06,
+      "loss": 0.0328,
+      "step": 3120
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2631748601423334,
+      "learning_rate": 9.511234831214464e-06,
+      "loss": 0.0301,
+      "step": 3121
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.22610547568516606,
+      "learning_rate": 9.499122883697724e-06,
+      "loss": 0.031,
+      "step": 3122
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.30749758862891285,
+      "learning_rate": 9.487016250739269e-06,
+      "loss": 0.0391,
+      "step": 3123
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2865122859893621,
+      "learning_rate": 9.474914938466328e-06,
+      "loss": 0.0266,
+      "step": 3124
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3161344066840338,
+      "learning_rate": 9.462818953003465e-06,
+      "loss": 0.0372,
+      "step": 3125
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3230785751011246,
+      "learning_rate": 9.45072830047251e-06,
+      "loss": 0.0471,
+      "step": 3126
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3371732737427641,
+      "learning_rate": 9.438642986992641e-06,
+      "loss": 0.0389,
+      "step": 3127
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.25125590412357773,
+      "learning_rate": 9.426563018680293e-06,
+      "loss": 0.0219,
+      "step": 3128
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3338988345174679,
+      "learning_rate": 9.414488401649227e-06,
+      "loss": 0.0361,
+      "step": 3129
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3217851467083712,
+      "learning_rate": 9.40241914201046e-06,
+      "loss": 0.0336,
+      "step": 3130
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.35707314994576855,
+      "learning_rate": 9.390355245872337e-06,
+      "loss": 0.0482,
+      "step": 3131
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2898732748699959,
+      "learning_rate": 9.378296719340459e-06,
+      "loss": 0.0295,
+      "step": 3132
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2828287214822192,
+      "learning_rate": 9.366243568517726e-06,
+      "loss": 0.0373,
+      "step": 3133
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2516824465950768,
+      "learning_rate": 9.354195799504305e-06,
+      "loss": 0.0264,
+      "step": 3134
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2505504389748185,
+      "learning_rate": 9.342153418397647e-06,
+      "loss": 0.0372,
+      "step": 3135
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.29056852673906025,
+      "learning_rate": 9.330116431292478e-06,
+      "loss": 0.0407,
+      "step": 3136
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.26675860352481984,
+      "learning_rate": 9.318084844280798e-06,
+      "loss": 0.0393,
+      "step": 3137
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.26130883298179075,
+      "learning_rate": 9.306058663451852e-06,
+      "loss": 0.0258,
+      "step": 3138
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2881059025457402,
+      "learning_rate": 9.294037894892178e-06,
+      "loss": 0.0365,
+      "step": 3139
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3193162684551438,
+      "learning_rate": 9.28202254468555e-06,
+      "loss": 0.0387,
+      "step": 3140
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.31524669429337604,
+      "learning_rate": 9.270012618913018e-06,
+      "loss": 0.0431,
+      "step": 3141
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.29463283952180785,
+      "learning_rate": 9.258008123652868e-06,
+      "loss": 0.0413,
+      "step": 3142
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3625294895344755,
+      "learning_rate": 9.246009064980657e-06,
+      "loss": 0.0479,
+      "step": 3143
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.29733082010017053,
+      "learning_rate": 9.23401544896919e-06,
+      "loss": 0.0338,
+      "step": 3144
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2953035650970452,
+      "learning_rate": 9.22202728168849e-06,
+      "loss": 0.0384,
+      "step": 3145
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.26778874148341164,
+      "learning_rate": 9.210044569205863e-06,
+      "loss": 0.0262,
+      "step": 3146
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2789062088611049,
+      "learning_rate": 9.198067317585816e-06,
+      "loss": 0.0368,
+      "step": 3147
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.27855798043463303,
+      "learning_rate": 9.186095532890121e-06,
+      "loss": 0.0423,
+      "step": 3148
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.24564879928196806,
+      "learning_rate": 9.174129221177762e-06,
+      "loss": 0.034,
+      "step": 3149
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2798131159847416,
+      "learning_rate": 9.162168388504972e-06,
+      "loss": 0.0291,
+      "step": 3150
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.29632821676191,
+      "learning_rate": 9.150213040925193e-06,
+      "loss": 0.0328,
+      "step": 3151
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.42087156559602495,
+      "learning_rate": 9.138263184489104e-06,
+      "loss": 0.0482,
+      "step": 3152
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.30010754823131197,
+      "learning_rate": 9.1263188252446e-06,
+      "loss": 0.0341,
+      "step": 3153
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3008349234511671,
+      "learning_rate": 9.114379969236802e-06,
+      "loss": 0.0331,
+      "step": 3154
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2628119822938154,
+      "learning_rate": 9.102446622508025e-06,
+      "loss": 0.0341,
+      "step": 3155
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2439599627576946,
+      "learning_rate": 9.090518791097822e-06,
+      "loss": 0.0304,
+      "step": 3156
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3738203169466667,
+      "learning_rate": 9.078596481042927e-06,
+      "loss": 0.0526,
+      "step": 3157
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.28116721683341717,
+      "learning_rate": 9.066679698377311e-06,
+      "loss": 0.0219,
+      "step": 3158
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.26721846640871477,
+      "learning_rate": 9.054768449132115e-06,
+      "loss": 0.0354,
+      "step": 3159
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.30702480620011646,
+      "learning_rate": 9.042862739335707e-06,
+      "loss": 0.0344,
+      "step": 3160
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.26577193740057614,
+      "learning_rate": 9.030962575013622e-06,
+      "loss": 0.0343,
+      "step": 3161
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.27656957541860233,
+      "learning_rate": 9.019067962188634e-06,
+      "loss": 0.0274,
+      "step": 3162
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.24374774601701155,
+      "learning_rate": 9.007178906880655e-06,
+      "loss": 0.0234,
+      "step": 3163
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2787367800070968,
+      "learning_rate": 8.995295415106829e-06,
+      "loss": 0.0299,
+      "step": 3164
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2559452349619959,
+      "learning_rate": 8.983417492881443e-06,
+      "loss": 0.0324,
+      "step": 3165
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2008699540297915,
+      "learning_rate": 8.971545146216005e-06,
+      "loss": 0.0235,
+      "step": 3166
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.23118591258238033,
+      "learning_rate": 8.959678381119166e-06,
+      "loss": 0.0305,
+      "step": 3167
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.32931410946889816,
+      "learning_rate": 8.947817203596785e-06,
+      "loss": 0.0272,
+      "step": 3168
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.27844357354404126,
+      "learning_rate": 8.935961619651859e-06,
+      "loss": 0.0347,
+      "step": 3169
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3423909501339809,
+      "learning_rate": 8.924111635284582e-06,
+      "loss": 0.0372,
+      "step": 3170
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.36290249512540756,
+      "learning_rate": 8.91226725649231e-06,
+      "loss": 0.0532,
+      "step": 3171
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2738942418836486,
+      "learning_rate": 8.900428489269541e-06,
+      "loss": 0.0275,
+      "step": 3172
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3044333253627893,
+      "learning_rate": 8.888595339607961e-06,
+      "loss": 0.0354,
+      "step": 3173
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.22113366457544695,
+      "learning_rate": 8.876767813496388e-06,
+      "loss": 0.0244,
+      "step": 3174
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.33908047958026905,
+      "learning_rate": 8.86494591692081e-06,
+      "loss": 0.0385,
+      "step": 3175
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2877423063537248,
+      "learning_rate": 8.85312965586437e-06,
+      "loss": 0.0375,
+      "step": 3176
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.29744477556872667,
+      "learning_rate": 8.841319036307334e-06,
+      "loss": 0.0412,
+      "step": 3177
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.39751737601611026,
+      "learning_rate": 8.829514064227138e-06,
+      "loss": 0.0539,
+      "step": 3178
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.28259364348374433,
+      "learning_rate": 8.817714745598358e-06,
+      "loss": 0.0314,
+      "step": 3179
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.25615170210053695,
+      "learning_rate": 8.805921086392686e-06,
+      "loss": 0.0305,
+      "step": 3180
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.27951022982259094,
+      "learning_rate": 8.79413309257898e-06,
+      "loss": 0.0299,
+      "step": 3181
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.27529645892553,
+      "learning_rate": 8.782350770123202e-06,
+      "loss": 0.033,
+      "step": 3182
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.27557445369722716,
+      "learning_rate": 8.770574124988474e-06,
+      "loss": 0.0348,
+      "step": 3183
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2881043250471307,
+      "learning_rate": 8.758803163135008e-06,
+      "loss": 0.0392,
+      "step": 3184
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.25932623335834537,
+      "learning_rate": 8.74703789052018e-06,
+      "loss": 0.0322,
+      "step": 3185
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.28605773253215566,
+      "learning_rate": 8.73527831309844e-06,
+      "loss": 0.0305,
+      "step": 3186
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3632188353707027,
+      "learning_rate": 8.723524436821418e-06,
+      "loss": 0.0359,
+      "step": 3187
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.30155488632157246,
+      "learning_rate": 8.711776267637794e-06,
+      "loss": 0.0304,
+      "step": 3188
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.29998057173364495,
+      "learning_rate": 8.700033811493407e-06,
+      "loss": 0.0343,
+      "step": 3189
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3343495163164025,
+      "learning_rate": 8.688297074331171e-06,
+      "loss": 0.0513,
+      "step": 3190
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2502815627983121,
+      "learning_rate": 8.676566062091135e-06,
+      "loss": 0.0284,
+      "step": 3191
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.27547491420485215,
+      "learning_rate": 8.66484078071042e-06,
+      "loss": 0.0343,
+      "step": 3192
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3151790461619625,
+      "learning_rate": 8.653121236123278e-06,
+      "loss": 0.0319,
+      "step": 3193
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2732800984344735,
+      "learning_rate": 8.641407434261031e-06,
+      "loss": 0.0316,
+      "step": 3194
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3315842255890883,
+      "learning_rate": 8.62969938105211e-06,
+      "loss": 0.0378,
+      "step": 3195
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.26288651665797536,
+      "learning_rate": 8.617997082422031e-06,
+      "loss": 0.0311,
+      "step": 3196
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2804867268488633,
+      "learning_rate": 8.606300544293412e-06,
+      "loss": 0.0364,
+      "step": 3197
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.34573039793205407,
+      "learning_rate": 8.594609772585922e-06,
+      "loss": 0.0449,
+      "step": 3198
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.24380722932954038,
+      "learning_rate": 8.582924773216353e-06,
+      "loss": 0.0244,
+      "step": 3199
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.27146298166414884,
+      "learning_rate": 8.571245552098533e-06,
+      "loss": 0.0351,
+      "step": 3200
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.291641356300972,
+      "learning_rate": 8.559572115143406e-06,
+      "loss": 0.0251,
+      "step": 3201
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2927027066573506,
+      "learning_rate": 8.547904468258957e-06,
+      "loss": 0.0287,
+      "step": 3202
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.26510552491525224,
+      "learning_rate": 8.536242617350265e-06,
+      "loss": 0.0352,
+      "step": 3203
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.286560637094647,
+      "learning_rate": 8.524586568319451e-06,
+      "loss": 0.041,
+      "step": 3204
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2858820001489877,
+      "learning_rate": 8.51293632706572e-06,
+      "loss": 0.0278,
+      "step": 3205
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.38231843088964895,
+      "learning_rate": 8.501291899485337e-06,
+      "loss": 0.0433,
+      "step": 3206
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4060582662170565,
+      "learning_rate": 8.489653291471607e-06,
+      "loss": 0.0549,
+      "step": 3207
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.28511768237017565,
+      "learning_rate": 8.47802050891491e-06,
+      "loss": 0.035,
+      "step": 3208
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2831822880939062,
+      "learning_rate": 8.466393557702659e-06,
+      "loss": 0.0385,
+      "step": 3209
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3482690938759456,
+      "learning_rate": 8.454772443719339e-06,
+      "loss": 0.0435,
+      "step": 3210
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2907088281881407,
+      "learning_rate": 8.443157172846448e-06,
+      "loss": 0.0307,
+      "step": 3211
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.26492070343199503,
+      "learning_rate": 8.43154775096256e-06,
+      "loss": 0.0347,
+      "step": 3212
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3079662703787326,
+      "learning_rate": 8.419944183943266e-06,
+      "loss": 0.0357,
+      "step": 3213
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2970429372118634,
+      "learning_rate": 8.408346477661218e-06,
+      "loss": 0.029,
+      "step": 3214
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.295109165165925,
+      "learning_rate": 8.39675463798607e-06,
+      "loss": 0.0315,
+      "step": 3215
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.26527102109089074,
+      "learning_rate": 8.385168670784532e-06,
+      "loss": 0.0337,
+      "step": 3216
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3220551699472881,
+      "learning_rate": 8.373588581920325e-06,
+      "loss": 0.0315,
+      "step": 3217
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2518849539387254,
+      "learning_rate": 8.362014377254213e-06,
+      "loss": 0.0275,
+      "step": 3218
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2867164128387573,
+      "learning_rate": 8.35044606264396e-06,
+      "loss": 0.035,
+      "step": 3219
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3426814485669959,
+      "learning_rate": 8.338883643944375e-06,
+      "loss": 0.0445,
+      "step": 3220
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.22337205335160346,
+      "learning_rate": 8.327327127007247e-06,
+      "loss": 0.0261,
+      "step": 3221
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.31098813071247333,
+      "learning_rate": 8.315776517681428e-06,
+      "loss": 0.0372,
+      "step": 3222
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.27028504987487467,
+      "learning_rate": 8.304231821812733e-06,
+      "loss": 0.0302,
+      "step": 3223
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3178525060794622,
+      "learning_rate": 8.292693045244016e-06,
+      "loss": 0.0364,
+      "step": 3224
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3628950219355067,
+      "learning_rate": 8.281160193815108e-06,
+      "loss": 0.045,
+      "step": 3225
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.44063406048921605,
+      "learning_rate": 8.269633273362872e-06,
+      "loss": 0.0511,
+      "step": 3226
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2285235476954415,
+      "learning_rate": 8.258112289721134e-06,
+      "loss": 0.021,
+      "step": 3227
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3376012235303488,
+      "learning_rate": 8.246597248720756e-06,
+      "loss": 0.0408,
+      "step": 3228
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2039304225668567,
+      "learning_rate": 8.23508815618955e-06,
+      "loss": 0.0212,
+      "step": 3229
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.357302119303284,
+      "learning_rate": 8.22358501795235e-06,
+      "loss": 0.0481,
+      "step": 3230
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.31132384935073115,
+      "learning_rate": 8.212087839830968e-06,
+      "loss": 0.0267,
+      "step": 3231
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2606308660144364,
+      "learning_rate": 8.200596627644187e-06,
+      "loss": 0.0324,
+      "step": 3232
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.30183494841210046,
+      "learning_rate": 8.189111387207782e-06,
+      "loss": 0.0378,
+      "step": 3233
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2965052083452161,
+      "learning_rate": 8.177632124334513e-06,
+      "loss": 0.05,
+      "step": 3234
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3487436348153068,
+      "learning_rate": 8.16615884483409e-06,
+      "loss": 0.0393,
+      "step": 3235
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.24122902188482195,
+      "learning_rate": 8.154691554513228e-06,
+      "loss": 0.0295,
+      "step": 3236
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3403639155904013,
+      "learning_rate": 8.143230259175574e-06,
+      "loss": 0.0389,
+      "step": 3237
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2303562859744233,
+      "learning_rate": 8.13177496462177e-06,
+      "loss": 0.0251,
+      "step": 3238
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.28828770958504507,
+      "learning_rate": 8.120325676649416e-06,
+      "loss": 0.0324,
+      "step": 3239
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3063426422594093,
+      "learning_rate": 8.108882401053055e-06,
+      "loss": 0.0243,
+      "step": 3240
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.31969377026789886,
+      "learning_rate": 8.09744514362421e-06,
+      "loss": 0.0453,
+      "step": 3241
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.28191090126797036,
+      "learning_rate": 8.086013910151334e-06,
+      "loss": 0.0288,
+      "step": 3242
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2359073376480803,
+      "learning_rate": 8.07458870641986e-06,
+      "loss": 0.0261,
+      "step": 3243
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.31954006938375107,
+      "learning_rate": 8.063169538212139e-06,
+      "loss": 0.0395,
+      "step": 3244
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2573348880995303,
+      "learning_rate": 8.051756411307494e-06,
+      "loss": 0.0297,
+      "step": 3245
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.22732547837337838,
+      "learning_rate": 8.040349331482167e-06,
+      "loss": 0.0315,
+      "step": 3246
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2483432531029029,
+      "learning_rate": 8.028948304509356e-06,
+      "loss": 0.0351,
+      "step": 3247
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.36573764989291757,
+      "learning_rate": 8.017553336159192e-06,
+      "loss": 0.0394,
+      "step": 3248
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2565588816574629,
+      "learning_rate": 8.006164432198747e-06,
+      "loss": 0.0366,
+      "step": 3249
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.28825887887882923,
+      "learning_rate": 7.994781598391995e-06,
+      "loss": 0.0302,
+      "step": 3250
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3507524907614378,
+      "learning_rate": 7.983404840499882e-06,
+      "loss": 0.0483,
+      "step": 3251
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2725297711015889,
+      "learning_rate": 7.972034164280231e-06,
+      "loss": 0.0336,
+      "step": 3252
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2762138156797559,
+      "learning_rate": 7.96066957548783e-06,
+      "loss": 0.0378,
+      "step": 3253
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2705830355830464,
+      "learning_rate": 7.949311079874352e-06,
+      "loss": 0.0311,
+      "step": 3254
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3045217601223277,
+      "learning_rate": 7.937958683188407e-06,
+      "loss": 0.0322,
+      "step": 3255
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.26781638181581713,
+      "learning_rate": 7.926612391175516e-06,
+      "loss": 0.0326,
+      "step": 3256
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3282592589589826,
+      "learning_rate": 7.915272209578112e-06,
+      "loss": 0.0357,
+      "step": 3257
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.25635829358510154,
+      "learning_rate": 7.903938144135515e-06,
+      "loss": 0.0265,
+      "step": 3258
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2292502081203452,
+      "learning_rate": 7.892610200583979e-06,
+      "loss": 0.0252,
+      "step": 3259
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2563374980548708,
+      "learning_rate": 7.881288384656634e-06,
+      "loss": 0.0226,
+      "step": 3260
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3228283910172808,
+      "learning_rate": 7.869972702083532e-06,
+      "loss": 0.0306,
+      "step": 3261
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2937913662617634,
+      "learning_rate": 7.8586631585916e-06,
+      "loss": 0.0297,
+      "step": 3262
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.27880336936959554,
+      "learning_rate": 7.847359759904675e-06,
+      "loss": 0.0303,
+      "step": 3263
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2654205404899196,
+      "learning_rate": 7.836062511743468e-06,
+      "loss": 0.0291,
+      "step": 3264
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.35197320895624984,
+      "learning_rate": 7.824771419825588e-06,
+      "loss": 0.0528,
+      "step": 3265
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3555739110978662,
+      "learning_rate": 7.813486489865534e-06,
+      "loss": 0.0394,
+      "step": 3266
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.27169094871366306,
+      "learning_rate": 7.802207727574665e-06,
+      "loss": 0.029,
+      "step": 3267
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.24856026153573255,
+      "learning_rate": 7.790935138661246e-06,
+      "loss": 0.0354,
+      "step": 3268
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2589370304520512,
+      "learning_rate": 7.779668728830389e-06,
+      "loss": 0.0287,
+      "step": 3269
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2636273240492979,
+      "learning_rate": 7.768408503784108e-06,
+      "loss": 0.0259,
+      "step": 3270
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.27505692087975914,
+      "learning_rate": 7.757154469221257e-06,
+      "loss": 0.035,
+      "step": 3271
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.26944460374515017,
+      "learning_rate": 7.745906630837586e-06,
+      "loss": 0.0297,
+      "step": 3272
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2780305927222107,
+      "learning_rate": 7.734664994325672e-06,
+      "loss": 0.0289,
+      "step": 3273
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.33649667510576525,
+      "learning_rate": 7.723429565375006e-06,
+      "loss": 0.0413,
+      "step": 3274
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3260934244253876,
+      "learning_rate": 7.71220034967189e-06,
+      "loss": 0.0342,
+      "step": 3275
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.34872161845458016,
+      "learning_rate": 7.700977352899506e-06,
+      "loss": 0.0428,
+      "step": 3276
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.29277538276010856,
+      "learning_rate": 7.68976058073787e-06,
+      "loss": 0.0297,
+      "step": 3277
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.27543608648700496,
+      "learning_rate": 7.678550038863877e-06,
+      "loss": 0.0332,
+      "step": 3278
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2683624967035358,
+      "learning_rate": 7.667345732951233e-06,
+      "loss": 0.0341,
+      "step": 3279
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.29395466532988246,
+      "learning_rate": 7.656147668670519e-06,
+      "loss": 0.0345,
+      "step": 3280
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2438548131709034,
+      "learning_rate": 7.644955851689129e-06,
+      "loss": 0.0278,
+      "step": 3281
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.23469584280035205,
+      "learning_rate": 7.63377028767133e-06,
+      "loss": 0.0252,
+      "step": 3282
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.26294336342626873,
+      "learning_rate": 7.622590982278189e-06,
+      "loss": 0.0297,
+      "step": 3283
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2637056117197377,
+      "learning_rate": 7.611417941167634e-06,
+      "loss": 0.0414,
+      "step": 3284
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.24425409350483882,
+      "learning_rate": 7.600251169994392e-06,
+      "loss": 0.0275,
+      "step": 3285
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3440145969738492,
+      "learning_rate": 7.589090674410056e-06,
+      "loss": 0.0407,
+      "step": 3286
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.32265773384916846,
+      "learning_rate": 7.577936460063e-06,
+      "loss": 0.0457,
+      "step": 3287
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.21486533663414784,
+      "learning_rate": 7.566788532598457e-06,
+      "loss": 0.0242,
+      "step": 3288
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3426672398743925,
+      "learning_rate": 7.555646897658448e-06,
+      "loss": 0.0418,
+      "step": 3289
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2978583663413471,
+      "learning_rate": 7.544511560881829e-06,
+      "loss": 0.0507,
+      "step": 3290
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.29657699033028645,
+      "learning_rate": 7.533382527904263e-06,
+      "loss": 0.0372,
+      "step": 3291
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.30881177467626963,
+      "learning_rate": 7.5222598043582274e-06,
+      "loss": 0.0374,
+      "step": 3292
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.26934463841005973,
+      "learning_rate": 7.511143395872986e-06,
+      "loss": 0.0333,
+      "step": 3293
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.28768593124767466,
+      "learning_rate": 7.500033308074639e-06,
+      "loss": 0.0252,
+      "step": 3294
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3364082852313981,
+      "learning_rate": 7.488929546586053e-06,
+      "loss": 0.037,
+      "step": 3295
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.22541449997196128,
+      "learning_rate": 7.477832117026924e-06,
+      "loss": 0.0245,
+      "step": 3296
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.24872789984440977,
+      "learning_rate": 7.466741025013715e-06,
+      "loss": 0.0285,
+      "step": 3297
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.22554073458836682,
+      "learning_rate": 7.455656276159713e-06,
+      "loss": 0.025,
+      "step": 3298
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.29784291989654366,
+      "learning_rate": 7.444577876074956e-06,
+      "loss": 0.0274,
+      "step": 3299
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2899906362057722,
+      "learning_rate": 7.4335058303663056e-06,
+      "loss": 0.0322,
+      "step": 3300
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.29269440839549676,
+      "learning_rate": 7.422440144637395e-06,
+      "loss": 0.0427,
+      "step": 3301
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3068639585835522,
+      "learning_rate": 7.411380824488621e-06,
+      "loss": 0.0359,
+      "step": 3302
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.25979565324964654,
+      "learning_rate": 7.400327875517188e-06,
+      "loss": 0.03,
+      "step": 3303
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.31275196646339726,
+      "learning_rate": 7.389281303317046e-06,
+      "loss": 0.0399,
+      "step": 3304
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.247167598282481,
+      "learning_rate": 7.37824111347895e-06,
+      "loss": 0.0331,
+      "step": 3305
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.25559863425551843,
+      "learning_rate": 7.367207311590392e-06,
+      "loss": 0.0323,
+      "step": 3306
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2560772877837243,
+      "learning_rate": 7.356179903235654e-06,
+      "loss": 0.0357,
+      "step": 3307
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.25963678248029315,
+      "learning_rate": 7.345158893995774e-06,
+      "loss": 0.0271,
+      "step": 3308
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.30030508869803,
+      "learning_rate": 7.33414428944856e-06,
+      "loss": 0.0376,
+      "step": 3309
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3277322712263529,
+      "learning_rate": 7.3231360951685574e-06,
+      "loss": 0.0448,
+      "step": 3310
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.23406026688899606,
+      "learning_rate": 7.312134316727093e-06,
+      "loss": 0.0227,
+      "step": 3311
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.27825149220668854,
+      "learning_rate": 7.301138959692225e-06,
+      "loss": 0.0339,
+      "step": 3312
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2623883975162797,
+      "learning_rate": 7.290150029628777e-06,
+      "loss": 0.0306,
+      "step": 3313
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.31569686298159066,
+      "learning_rate": 7.2791675320983076e-06,
+      "loss": 0.0406,
+      "step": 3314
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2641515959588397,
+      "learning_rate": 7.268191472659136e-06,
+      "loss": 0.0256,
+      "step": 3315
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.26511555792736424,
+      "learning_rate": 7.257221856866295e-06,
+      "loss": 0.0291,
+      "step": 3316
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.25792506679317007,
+      "learning_rate": 7.246258690271599e-06,
+      "loss": 0.0268,
+      "step": 3317
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5339755656834957,
+      "learning_rate": 7.235301978423555e-06,
+      "loss": 0.0309,
+      "step": 3318
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2740130090122586,
+      "learning_rate": 7.224351726867433e-06,
+      "loss": 0.032,
+      "step": 3319
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.22585862262855183,
+      "learning_rate": 7.213407941145214e-06,
+      "loss": 0.0195,
+      "step": 3320
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2225228462116293,
+      "learning_rate": 7.202470626795626e-06,
+      "loss": 0.0243,
+      "step": 3321
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2747098746967936,
+      "learning_rate": 7.191539789354096e-06,
+      "loss": 0.0319,
+      "step": 3322
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.24156826594726816,
+      "learning_rate": 7.180615434352802e-06,
+      "loss": 0.0217,
+      "step": 3323
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2107368857465454,
+      "learning_rate": 7.1696975673206125e-06,
+      "loss": 0.0284,
+      "step": 3324
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2335866384795216,
+      "learning_rate": 7.158786193783138e-06,
+      "loss": 0.0234,
+      "step": 3325
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2748073054489228,
+      "learning_rate": 7.147881319262695e-06,
+      "loss": 0.0379,
+      "step": 3326
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2523174461142641,
+      "learning_rate": 7.136982949278293e-06,
+      "loss": 0.0279,
+      "step": 3327
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.25476723458425904,
+      "learning_rate": 7.126091089345679e-06,
+      "loss": 0.0313,
+      "step": 3328
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.26992223549141114,
+      "learning_rate": 7.115205744977276e-06,
+      "loss": 0.0309,
+      "step": 3329
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.28147717414007484,
+      "learning_rate": 7.104326921682236e-06,
+      "loss": 0.0256,
+      "step": 3330
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2750101911217013,
+      "learning_rate": 7.093454624966387e-06,
+      "loss": 0.0323,
+      "step": 3331
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2562045469910545,
+      "learning_rate": 7.082588860332271e-06,
+      "loss": 0.0309,
+      "step": 3332
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.27352463832409785,
+      "learning_rate": 7.071729633279118e-06,
+      "loss": 0.0288,
+      "step": 3333
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2512556075898497,
+      "learning_rate": 7.060876949302855e-06,
+      "loss": 0.0236,
+      "step": 3334
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2677385726256808,
+      "learning_rate": 7.050030813896078e-06,
+      "loss": 0.0278,
+      "step": 3335
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.28504892184507447,
+      "learning_rate": 7.0391912325481e-06,
+      "loss": 0.0325,
+      "step": 3336
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.36006169432920915,
+      "learning_rate": 7.028358210744881e-06,
+      "loss": 0.0477,
+      "step": 3337
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.21469547183354776,
+      "learning_rate": 7.017531753969098e-06,
+      "loss": 0.024,
+      "step": 3338
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.259636954657241,
+      "learning_rate": 7.006711867700069e-06,
+      "loss": 0.0307,
+      "step": 3339
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.24688085446478022,
+      "learning_rate": 6.995898557413823e-06,
+      "loss": 0.0296,
+      "step": 3340
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2831645150009331,
+      "learning_rate": 6.985091828583024e-06,
+      "loss": 0.0363,
+      "step": 3341
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.278823985728811,
+      "learning_rate": 6.974291686677035e-06,
+      "loss": 0.0413,
+      "step": 3342
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2746523572511557,
+      "learning_rate": 6.96349813716187e-06,
+      "loss": 0.0298,
+      "step": 3343
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2719467808247288,
+      "learning_rate": 6.952711185500223e-06,
+      "loss": 0.0312,
+      "step": 3344
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.23624465193321353,
+      "learning_rate": 6.941930837151416e-06,
+      "loss": 0.0232,
+      "step": 3345
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.21903564332030248,
+      "learning_rate": 6.931157097571468e-06,
+      "loss": 0.0294,
+      "step": 3346
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2691619388462333,
+      "learning_rate": 6.920389972213017e-06,
+      "loss": 0.0357,
+      "step": 3347
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.28274667722063346,
+      "learning_rate": 6.909629466525389e-06,
+      "loss": 0.0351,
+      "step": 3348
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.30060128612135306,
+      "learning_rate": 6.898875585954527e-06,
+      "loss": 0.0315,
+      "step": 3349
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3506542341179284,
+      "learning_rate": 6.888128335943041e-06,
+      "loss": 0.046,
+      "step": 3350
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3748067914417686,
+      "learning_rate": 6.877387721930182e-06,
+      "loss": 0.0425,
+      "step": 3351
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3375190279153201,
+      "learning_rate": 6.866653749351846e-06,
+      "loss": 0.0399,
+      "step": 3352
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.29486608925001806,
+      "learning_rate": 6.855926423640549e-06,
+      "loss": 0.0347,
+      "step": 3353
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.280396105674734,
+      "learning_rate": 6.84520575022547e-06,
+      "loss": 0.0352,
+      "step": 3354
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.28398588460864693,
+      "learning_rate": 6.8344917345323935e-06,
+      "loss": 0.0439,
+      "step": 3355
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.31521055564088574,
+      "learning_rate": 6.823784381983764e-06,
+      "loss": 0.0381,
+      "step": 3356
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2925692186413323,
+      "learning_rate": 6.8130836979986236e-06,
+      "loss": 0.0437,
+      "step": 3357
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2670446214626016,
+      "learning_rate": 6.802389687992666e-06,
+      "loss": 0.0302,
+      "step": 3358
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3065052462551992,
+      "learning_rate": 6.791702357378185e-06,
+      "loss": 0.0364,
+      "step": 3359
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.30816710228955385,
+      "learning_rate": 6.781021711564107e-06,
+      "loss": 0.0401,
+      "step": 3360
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.26630850813961854,
+      "learning_rate": 6.770347755955982e-06,
+      "loss": 0.0307,
+      "step": 3361
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.26040227214448936,
+      "learning_rate": 6.7596804959559494e-06,
+      "loss": 0.0314,
+      "step": 3362
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.23877831503346528,
+      "learning_rate": 6.749019936962791e-06,
+      "loss": 0.0354,
+      "step": 3363
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2416854392082955,
+      "learning_rate": 6.7383660843718635e-06,
+      "loss": 0.0322,
+      "step": 3364
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3053072470737272,
+      "learning_rate": 6.727718943575161e-06,
+      "loss": 0.0345,
+      "step": 3365
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2022255644424565,
+      "learning_rate": 6.717078519961257e-06,
+      "loss": 0.0174,
+      "step": 3366
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2389068286781463,
+      "learning_rate": 6.706444818915345e-06,
+      "loss": 0.024,
+      "step": 3367
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.30114718879034874,
+      "learning_rate": 6.695817845819188e-06,
+      "loss": 0.0386,
+      "step": 3368
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2640765581395411,
+      "learning_rate": 6.68519760605119e-06,
+      "loss": 0.0272,
+      "step": 3369
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.254462659503167,
+      "learning_rate": 6.674584104986295e-06,
+      "loss": 0.0303,
+      "step": 3370
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.353073391476766,
+      "learning_rate": 6.66397734799608e-06,
+      "loss": 0.0402,
+      "step": 3371
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2137880832225216,
+      "learning_rate": 6.653377340448673e-06,
+      "loss": 0.0215,
+      "step": 3372
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2648998909920871,
+      "learning_rate": 6.642784087708814e-06,
+      "loss": 0.0256,
+      "step": 3373
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2506035433595592,
+      "learning_rate": 6.6321975951378034e-06,
+      "loss": 0.0279,
+      "step": 3374
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.22480728717961235,
+      "learning_rate": 6.621617868093544e-06,
+      "loss": 0.0218,
+      "step": 3375
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2735224591828485,
+      "learning_rate": 6.611044911930477e-06,
+      "loss": 0.0343,
+      "step": 3376
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.23876396906016184,
+      "learning_rate": 6.6004787319996714e-06,
+      "loss": 0.0291,
+      "step": 3377
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3272250071635811,
+      "learning_rate": 6.589919333648711e-06,
+      "loss": 0.0406,
+      "step": 3378
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2897342145880577,
+      "learning_rate": 6.579366722221789e-06,
+      "loss": 0.0368,
+      "step": 3379
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3224209613385021,
+      "learning_rate": 6.568820903059632e-06,
+      "loss": 0.0395,
+      "step": 3380
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.26667334739442633,
+      "learning_rate": 6.558281881499556e-06,
+      "loss": 0.0408,
+      "step": 3381
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2810547733903214,
+      "learning_rate": 6.547749662875411e-06,
+      "loss": 0.0356,
+      "step": 3382
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.27188158136930185,
+      "learning_rate": 6.537224252517633e-06,
+      "loss": 0.0232,
+      "step": 3383
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.24156934543400135,
+      "learning_rate": 6.526705655753183e-06,
+      "loss": 0.034,
+      "step": 3384
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.23455296753242358,
+      "learning_rate": 6.516193877905592e-06,
+      "loss": 0.029,
+      "step": 3385
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3370798117890245,
+      "learning_rate": 6.505688924294944e-06,
+      "loss": 0.0356,
+      "step": 3386
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2723097159370534,
+      "learning_rate": 6.495190800237845e-06,
+      "loss": 0.0322,
+      "step": 3387
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.23418612817301912,
+      "learning_rate": 6.484699511047474e-06,
+      "loss": 0.0225,
+      "step": 3388
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2419803256251913,
+      "learning_rate": 6.474215062033527e-06,
+      "loss": 0.0231,
+      "step": 3389
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.30069783527957566,
+      "learning_rate": 6.463737458502255e-06,
+      "loss": 0.0358,
+      "step": 3390
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.26570232051378745,
+      "learning_rate": 6.453266705756427e-06,
+      "loss": 0.03,
+      "step": 3391
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2755429272846124,
+      "learning_rate": 6.442802809095363e-06,
+      "loss": 0.0344,
+      "step": 3392
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2552849093532269,
+      "learning_rate": 6.4323457738149034e-06,
+      "loss": 0.0264,
+      "step": 3393
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3070690782068707,
+      "learning_rate": 6.421895605207427e-06,
+      "loss": 0.0228,
+      "step": 3394
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.21608959849101006,
+      "learning_rate": 6.41145230856181e-06,
+      "loss": 0.0219,
+      "step": 3395
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3256684718703453,
+      "learning_rate": 6.401015889163489e-06,
+      "loss": 0.0457,
+      "step": 3396
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.27969183039376333,
+      "learning_rate": 6.3905863522943786e-06,
+      "loss": 0.0294,
+      "step": 3397
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.32161444562324737,
+      "learning_rate": 6.380163703232953e-06,
+      "loss": 0.0523,
+      "step": 3398
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.30481554666388455,
+      "learning_rate": 6.369747947254159e-06,
+      "loss": 0.038,
+      "step": 3399
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2658260510208518,
+      "learning_rate": 6.35933908962949e-06,
+      "loss": 0.0303,
+      "step": 3400
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2595659506770729,
+      "learning_rate": 6.348937135626922e-06,
+      "loss": 0.0311,
+      "step": 3401
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.20705685367249238,
+      "learning_rate": 6.338542090510951e-06,
+      "loss": 0.022,
+      "step": 3402
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.28922769620260047,
+      "learning_rate": 6.328153959542573e-06,
+      "loss": 0.0294,
+      "step": 3403
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.32985597391757204,
+      "learning_rate": 6.3177727479792914e-06,
+      "loss": 0.0394,
+      "step": 3404
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.23242230775937095,
+      "learning_rate": 6.307398461075091e-06,
+      "loss": 0.0285,
+      "step": 3405
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.27369074674458943,
+      "learning_rate": 6.297031104080471e-06,
+      "loss": 0.0238,
+      "step": 3406
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.262031994012767,
+      "learning_rate": 6.286670682242404e-06,
+      "loss": 0.0308,
+      "step": 3407
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.26526526326030525,
+      "learning_rate": 6.276317200804376e-06,
+      "loss": 0.0274,
+      "step": 3408
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3540060319507838,
+      "learning_rate": 6.265970665006334e-06,
+      "loss": 0.0469,
+      "step": 3409
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2258842114700864,
+      "learning_rate": 6.255631080084735e-06,
+      "loss": 0.027,
+      "step": 3410
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.31748298824722165,
+      "learning_rate": 6.245298451272486e-06,
+      "loss": 0.0358,
+      "step": 3411
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2888240733028864,
+      "learning_rate": 6.234972783799023e-06,
+      "loss": 0.0362,
+      "step": 3412
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2723556916884701,
+      "learning_rate": 6.224654082890207e-06,
+      "loss": 0.03,
+      "step": 3413
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2978367536021447,
+      "learning_rate": 6.214342353768412e-06,
+      "loss": 0.0324,
+      "step": 3414
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.29954612094059346,
+      "learning_rate": 6.2040376016524506e-06,
+      "loss": 0.0492,
+      "step": 3415
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.30504933904023807,
+      "learning_rate": 6.193739831757637e-06,
+      "loss": 0.0362,
+      "step": 3416
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3184118403165737,
+      "learning_rate": 6.183449049295722e-06,
+      "loss": 0.0334,
+      "step": 3417
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2485101608567878,
+      "learning_rate": 6.1731652594749465e-06,
+      "loss": 0.0288,
+      "step": 3418
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.23786661938343767,
+      "learning_rate": 6.162888467499988e-06,
+      "loss": 0.0282,
+      "step": 3419
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.29388923997343264,
+      "learning_rate": 6.152618678571996e-06,
+      "loss": 0.0257,
+      "step": 3420
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.414291264659481,
+      "learning_rate": 6.1423558978885836e-06,
+      "loss": 0.0706,
+      "step": 3421
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.24517493308692564,
+      "learning_rate": 6.1321001306437946e-06,
+      "loss": 0.0268,
+      "step": 3422
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.27522385058717846,
+      "learning_rate": 6.121851382028146e-06,
+      "loss": 0.0316,
+      "step": 3423
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2731825730117969,
+      "learning_rate": 6.111609657228581e-06,
+      "loss": 0.04,
+      "step": 3424
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3421778054801919,
+      "learning_rate": 6.101374961428512e-06,
+      "loss": 0.0295,
+      "step": 3425
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2716664314464027,
+      "learning_rate": 6.091147299807769e-06,
+      "loss": 0.0325,
+      "step": 3426
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.37625543729839095,
+      "learning_rate": 6.080926677542646e-06,
+      "loss": 0.0461,
+      "step": 3427
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2710786161128589,
+      "learning_rate": 6.070713099805845e-06,
+      "loss": 0.0327,
+      "step": 3428
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2750153250108817,
+      "learning_rate": 6.0605065717665445e-06,
+      "loss": 0.0272,
+      "step": 3429
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.27125907715758363,
+      "learning_rate": 6.050307098590311e-06,
+      "loss": 0.0327,
+      "step": 3430
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.24568234978981157,
+      "learning_rate": 6.040114685439175e-06,
+      "loss": 0.0278,
+      "step": 3431
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.26473579513348117,
+      "learning_rate": 6.029929337471565e-06,
+      "loss": 0.0251,
+      "step": 3432
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2821258180316564,
+      "learning_rate": 6.019751059842362e-06,
+      "loss": 0.0373,
+      "step": 3433
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2629471744972731,
+      "learning_rate": 6.009579857702843e-06,
+      "loss": 0.0273,
+      "step": 3434
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2585106845912787,
+      "learning_rate": 5.999415736200724e-06,
+      "loss": 0.0311,
+      "step": 3435
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3633553009034456,
+      "learning_rate": 5.98925870048012e-06,
+      "loss": 0.0366,
+      "step": 3436
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.35135252298030384,
+      "learning_rate": 5.979108755681575e-06,
+      "loss": 0.0429,
+      "step": 3437
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2768956715291986,
+      "learning_rate": 5.968965906942039e-06,
+      "loss": 0.0386,
+      "step": 3438
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2373179757750094,
+      "learning_rate": 5.958830159394875e-06,
+      "loss": 0.0292,
+      "step": 3439
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.27913695872408384,
+      "learning_rate": 5.948701518169835e-06,
+      "loss": 0.0353,
+      "step": 3440
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.26605618765585465,
+      "learning_rate": 5.938579988393099e-06,
+      "loss": 0.0366,
+      "step": 3441
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.23421451319195605,
+      "learning_rate": 5.928465575187221e-06,
+      "loss": 0.0304,
+      "step": 3442
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2615844894609264,
+      "learning_rate": 5.918358283671182e-06,
+      "loss": 0.0386,
+      "step": 3443
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.27336979436518566,
+      "learning_rate": 5.90825811896033e-06,
+      "loss": 0.0379,
+      "step": 3444
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.365486384468251,
+      "learning_rate": 5.89816508616643e-06,
+      "loss": 0.0427,
+      "step": 3445
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.29642779145819365,
+      "learning_rate": 5.888079190397628e-06,
+      "loss": 0.0477,
+      "step": 3446
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.22215723703476092,
+      "learning_rate": 5.878000436758453e-06,
+      "loss": 0.0203,
+      "step": 3447
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.21405507026085965,
+      "learning_rate": 5.86792883034983e-06,
+      "loss": 0.0232,
+      "step": 3448
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2628546164396591,
+      "learning_rate": 5.857864376269051e-06,
+      "loss": 0.0271,
+      "step": 3449
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2734873453757123,
+      "learning_rate": 5.847807079609804e-06,
+      "loss": 0.0305,
+      "step": 3450
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.22837991841975297,
+      "learning_rate": 5.837756945462154e-06,
+      "loss": 0.0294,
+      "step": 3451
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2840929849505443,
+      "learning_rate": 5.827713978912524e-06,
+      "loss": 0.0252,
+      "step": 3452
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.21994348626473945,
+      "learning_rate": 5.817678185043733e-06,
+      "loss": 0.0356,
+      "step": 3453
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2650177690558583,
+      "learning_rate": 5.807649568934945e-06,
+      "loss": 0.0319,
+      "step": 3454
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2915132735135509,
+      "learning_rate": 5.79762813566171e-06,
+      "loss": 0.0342,
+      "step": 3455
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2420591891239867,
+      "learning_rate": 5.7876138902959445e-06,
+      "loss": 0.0258,
+      "step": 3456
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.272814886513111,
+      "learning_rate": 5.777606837905905e-06,
+      "loss": 0.0363,
+      "step": 3457
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.22757156696554626,
+      "learning_rate": 5.767606983556237e-06,
+      "loss": 0.0327,
+      "step": 3458
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.25811792835034714,
+      "learning_rate": 5.757614332307912e-06,
+      "loss": 0.025,
+      "step": 3459
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2748675417510592,
+      "learning_rate": 5.7476288892182905e-06,
+      "loss": 0.0406,
+      "step": 3460
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.29713312020558075,
+      "learning_rate": 5.737650659341048e-06,
+      "loss": 0.0299,
+      "step": 3461
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.27745443472639314,
+      "learning_rate": 5.7276796477262365e-06,
+      "loss": 0.0344,
+      "step": 3462
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.17985481267953204,
+      "learning_rate": 5.717715859420246e-06,
+      "loss": 0.018,
+      "step": 3463
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.19989098468629776,
+      "learning_rate": 5.707759299465816e-06,
+      "loss": 0.0179,
+      "step": 3464
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2669343513181597,
+      "learning_rate": 5.6978099729020105e-06,
+      "loss": 0.0304,
+      "step": 3465
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.21945043379538026,
+      "learning_rate": 5.68786788476426e-06,
+      "loss": 0.0249,
+      "step": 3466
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.27184858522676725,
+      "learning_rate": 5.6779330400843e-06,
+      "loss": 0.0297,
+      "step": 3467
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.26262306439441285,
+      "learning_rate": 5.66800544389023e-06,
+      "loss": 0.023,
+      "step": 3468
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.32370174504924254,
+      "learning_rate": 5.658085101206456e-06,
+      "loss": 0.0523,
+      "step": 3469
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.27778446749075747,
+      "learning_rate": 5.648172017053737e-06,
+      "loss": 0.0287,
+      "step": 3470
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4630715543716684,
+      "learning_rate": 5.638266196449123e-06,
+      "loss": 0.0579,
+      "step": 3471
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2009417597347354,
+      "learning_rate": 5.628367644406039e-06,
+      "loss": 0.0205,
+      "step": 3472
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3200980375658781,
+      "learning_rate": 5.618476365934184e-06,
+      "loss": 0.0349,
+      "step": 3473
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.28005963986738336,
+      "learning_rate": 5.608592366039607e-06,
+      "loss": 0.0403,
+      "step": 3474
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.25529178241170714,
+      "learning_rate": 5.598715649724647e-06,
+      "loss": 0.0283,
+      "step": 3475
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2284862303779472,
+      "learning_rate": 5.588846221987982e-06,
+      "loss": 0.0334,
+      "step": 3476
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.23788867724142732,
+      "learning_rate": 5.578984087824581e-06,
+      "loss": 0.0219,
+      "step": 3477
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2614670582312282,
+      "learning_rate": 5.569129252225745e-06,
+      "loss": 0.0258,
+      "step": 3478
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.24618739662400665,
+      "learning_rate": 5.559281720179046e-06,
+      "loss": 0.0221,
+      "step": 3479
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2834080185115215,
+      "learning_rate": 5.549441496668393e-06,
+      "loss": 0.0385,
+      "step": 3480
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.20664769757287502,
+      "learning_rate": 5.539608586673988e-06,
+      "loss": 0.0302,
+      "step": 3481
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.25517500283450656,
+      "learning_rate": 5.529782995172315e-06,
+      "loss": 0.0257,
+      "step": 3482
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2605643648285357,
+      "learning_rate": 5.519964727136178e-06,
+      "loss": 0.0308,
+      "step": 3483
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2539689646809514,
+      "learning_rate": 5.510153787534651e-06,
+      "loss": 0.0201,
+      "step": 3484
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2587773590320521,
+      "learning_rate": 5.500350181333121e-06,
+      "loss": 0.034,
+      "step": 3485
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.24054653541860618,
+      "learning_rate": 5.490553913493242e-06,
+      "loss": 0.0216,
+      "step": 3486
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.33018462590373493,
+      "learning_rate": 5.48076498897298e-06,
+      "loss": 0.0315,
+      "step": 3487
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.27005919291089847,
+      "learning_rate": 5.470983412726547e-06,
+      "loss": 0.0232,
+      "step": 3488
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.20541222015878996,
+      "learning_rate": 5.461209189704486e-06,
+      "loss": 0.0172,
+      "step": 3489
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.23395304844311773,
+      "learning_rate": 5.451442324853571e-06,
+      "loss": 0.0257,
+      "step": 3490
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.24403252306408385,
+      "learning_rate": 5.441682823116887e-06,
+      "loss": 0.0237,
+      "step": 3491
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2592484625124243,
+      "learning_rate": 5.431930689433762e-06,
+      "loss": 0.0246,
+      "step": 3492
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.26405085419475327,
+      "learning_rate": 5.422185928739827e-06,
+      "loss": 0.0247,
+      "step": 3493
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.23156080301978307,
+      "learning_rate": 5.4124485459669485e-06,
+      "loss": 0.0264,
+      "step": 3494
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2927319905143484,
+      "learning_rate": 5.402718546043293e-06,
+      "loss": 0.0275,
+      "step": 3495
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3207400538826871,
+      "learning_rate": 5.392995933893255e-06,
+      "loss": 0.0259,
+      "step": 3496
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.28107603308849083,
+      "learning_rate": 5.383280714437518e-06,
+      "loss": 0.0252,
+      "step": 3497
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.28216600923457863,
+      "learning_rate": 5.373572892593013e-06,
+      "loss": 0.0379,
+      "step": 3498
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.22892613503596634,
+      "learning_rate": 5.363872473272935e-06,
+      "loss": 0.0222,
+      "step": 3499
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.24876891199660664,
+      "learning_rate": 5.354179461386712e-06,
+      "loss": 0.039,
+      "step": 3500
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.38444579333155693,
+      "learning_rate": 5.3444938618400524e-06,
+      "loss": 0.0455,
+      "step": 3501
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.26324160755669274,
+      "learning_rate": 5.334815679534882e-06,
+      "loss": 0.0241,
+      "step": 3502
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.24986183550745963,
+      "learning_rate": 5.325144919369398e-06,
+      "loss": 0.0287,
+      "step": 3503
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2715735803501177,
+      "learning_rate": 5.315481586238025e-06,
+      "loss": 0.0253,
+      "step": 3504
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2895472817426594,
+      "learning_rate": 5.305825685031445e-06,
+      "loss": 0.0296,
+      "step": 3505
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.25753565576129356,
+      "learning_rate": 5.296177220636556e-06,
+      "loss": 0.0263,
+      "step": 3506
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.25278871061247116,
+      "learning_rate": 5.286536197936512e-06,
+      "loss": 0.0349,
+      "step": 3507
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3275757120245091,
+      "learning_rate": 5.276902621810691e-06,
+      "loss": 0.0298,
+      "step": 3508
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2513118834223532,
+      "learning_rate": 5.267276497134715e-06,
+      "loss": 0.0292,
+      "step": 3509
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2804987463676603,
+      "learning_rate": 5.257657828780409e-06,
+      "loss": 0.0289,
+      "step": 3510
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2999060270032961,
+      "learning_rate": 5.2480466216158565e-06,
+      "loss": 0.0376,
+      "step": 3511
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2903810148621985,
+      "learning_rate": 5.2384428805053325e-06,
+      "loss": 0.0302,
+      "step": 3512
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3013708108225558,
+      "learning_rate": 5.228846610309359e-06,
+      "loss": 0.0365,
+      "step": 3513
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3061993089879895,
+      "learning_rate": 5.219257815884662e-06,
+      "loss": 0.0314,
+      "step": 3514
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3321456487709721,
+      "learning_rate": 5.209676502084191e-06,
+      "loss": 0.0341,
+      "step": 3515
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2771528150926492,
+      "learning_rate": 5.200102673757115e-06,
+      "loss": 0.033,
+      "step": 3516
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2362254842017043,
+      "learning_rate": 5.190536335748792e-06,
+      "loss": 0.0235,
+      "step": 3517
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.32312161596950045,
+      "learning_rate": 5.180977492900823e-06,
+      "loss": 0.0386,
+      "step": 3518
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2725945558677096,
+      "learning_rate": 5.171426150050977e-06,
+      "loss": 0.0316,
+      "step": 3519
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.35931874954513726,
+      "learning_rate": 5.161882312033264e-06,
+      "loss": 0.0579,
+      "step": 3520
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.23317933272713304,
+      "learning_rate": 5.152345983677866e-06,
+      "loss": 0.0182,
+      "step": 3521
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.24431754758865426,
+      "learning_rate": 5.142817169811189e-06,
+      "loss": 0.0259,
+      "step": 3522
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.23309584279315684,
+      "learning_rate": 5.133295875255808e-06,
+      "loss": 0.0228,
+      "step": 3523
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.24801257038945074,
+      "learning_rate": 5.1237821048305305e-06,
+      "loss": 0.0295,
+      "step": 3524
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.22829640298379072,
+      "learning_rate": 5.114275863350313e-06,
+      "loss": 0.0222,
+      "step": 3525
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.29244651024135737,
+      "learning_rate": 5.104777155626341e-06,
+      "loss": 0.0262,
+      "step": 3526
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2948691845239951,
+      "learning_rate": 5.095285986465952e-06,
+      "loss": 0.0307,
+      "step": 3527
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.25266239963235654,
+      "learning_rate": 5.085802360672701e-06,
+      "loss": 0.0258,
+      "step": 3528
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.19868677769673954,
+      "learning_rate": 5.076326283046291e-06,
+      "loss": 0.02,
+      "step": 3529
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.361861026704966,
+      "learning_rate": 5.066857758382642e-06,
+      "loss": 0.0369,
+      "step": 3530
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2777206385309938,
+      "learning_rate": 5.057396791473807e-06,
+      "loss": 0.027,
+      "step": 3531
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.24217204844498547,
+      "learning_rate": 5.047943387108072e-06,
+      "loss": 0.0343,
+      "step": 3532
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.27128613046774036,
+      "learning_rate": 5.038497550069836e-06,
+      "loss": 0.0323,
+      "step": 3533
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.25555773418961775,
+      "learning_rate": 5.029059285139715e-06,
+      "loss": 0.0259,
+      "step": 3534
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.26456868282566975,
+      "learning_rate": 5.019628597094455e-06,
+      "loss": 0.0216,
+      "step": 3535
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1992101760964003,
+      "learning_rate": 5.010205490706998e-06,
+      "loss": 0.0241,
+      "step": 3536
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2213759420941348,
+      "learning_rate": 5.000789970746427e-06,
+      "loss": 0.0179,
+      "step": 3537
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3218004163420789,
+      "learning_rate": 4.9913820419780035e-06,
+      "loss": 0.0291,
+      "step": 3538
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.24229933094903675,
+      "learning_rate": 4.981981709163126e-06,
+      "loss": 0.025,
+      "step": 3539
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.26500260138134524,
+      "learning_rate": 4.972588977059369e-06,
+      "loss": 0.0312,
+      "step": 3540
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2412525558089232,
+      "learning_rate": 4.963203850420455e-06,
+      "loss": 0.0268,
+      "step": 3541
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2521682973733539,
+      "learning_rate": 4.953826333996243e-06,
+      "loss": 0.0292,
+      "step": 3542
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2608333813928269,
+      "learning_rate": 4.944456432532765e-06,
+      "loss": 0.0319,
+      "step": 3543
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2768863210021235,
+      "learning_rate": 4.93509415077217e-06,
+      "loss": 0.0426,
+      "step": 3544
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.28003648051964447,
+      "learning_rate": 4.925739493452783e-06,
+      "loss": 0.0377,
+      "step": 3545
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2550543284744202,
+      "learning_rate": 4.916392465309037e-06,
+      "loss": 0.0293,
+      "step": 3546
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.35349700415028396,
+      "learning_rate": 4.907053071071535e-06,
+      "loss": 0.0384,
+      "step": 3547
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2566989489840341,
+      "learning_rate": 4.89772131546699e-06,
+      "loss": 0.0259,
+      "step": 3548
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.21212567254501036,
+      "learning_rate": 4.888397203218265e-06,
+      "loss": 0.0248,
+      "step": 3549
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.28878337683718536,
+      "learning_rate": 4.879080739044351e-06,
+      "loss": 0.0309,
+      "step": 3550
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.24755283025469316,
+      "learning_rate": 4.869771927660371e-06,
+      "loss": 0.0331,
+      "step": 3551
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.21122812630478413,
+      "learning_rate": 4.860470773777566e-06,
+      "loss": 0.0178,
+      "step": 3552
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2693034328794336,
+      "learning_rate": 4.851177282103312e-06,
+      "loss": 0.0347,
+      "step": 3553
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.23671622762815486,
+      "learning_rate": 4.841891457341095e-06,
+      "loss": 0.0317,
+      "step": 3554
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.26564499835858235,
+      "learning_rate": 4.832613304190537e-06,
+      "loss": 0.0306,
+      "step": 3555
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.276596284408431,
+      "learning_rate": 4.823342827347357e-06,
+      "loss": 0.0291,
+      "step": 3556
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2284591882249527,
+      "learning_rate": 4.814080031503407e-06,
+      "loss": 0.0222,
+      "step": 3557
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.288734481898206,
+      "learning_rate": 4.804824921346645e-06,
+      "loss": 0.0365,
+      "step": 3558
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3201243964751489,
+      "learning_rate": 4.795577501561144e-06,
+      "loss": 0.0397,
+      "step": 3559
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.24722216014706502,
+      "learning_rate": 4.786337776827066e-06,
+      "loss": 0.0272,
+      "step": 3560
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.217336635706269,
+      "learning_rate": 4.777105751820708e-06,
+      "loss": 0.0205,
+      "step": 3561
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.22388667061618933,
+      "learning_rate": 4.767881431214441e-06,
+      "loss": 0.0247,
+      "step": 3562
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2911109512169312,
+      "learning_rate": 4.758664819676759e-06,
+      "loss": 0.0424,
+      "step": 3563
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2558790738042841,
+      "learning_rate": 4.7494559218722395e-06,
+      "loss": 0.0295,
+      "step": 3564
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.271038856225442,
+      "learning_rate": 4.74025474246157e-06,
+      "loss": 0.0316,
+      "step": 3565
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.21785685733395657,
+      "learning_rate": 4.7310612861015125e-06,
+      "loss": 0.0276,
+      "step": 3566
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3025124505530959,
+      "learning_rate": 4.7218755574449394e-06,
+      "loss": 0.0348,
+      "step": 3567
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2646960535885467,
+      "learning_rate": 4.712697561140802e-06,
+      "loss": 0.0415,
+      "step": 3568
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2712372677535224,
+      "learning_rate": 4.703527301834148e-06,
+      "loss": 0.0301,
+      "step": 3569
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3025306754676668,
+      "learning_rate": 4.69436478416609e-06,
+      "loss": 0.0361,
+      "step": 3570
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2173128089024891,
+      "learning_rate": 4.685210012773844e-06,
+      "loss": 0.0255,
+      "step": 3571
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.279395026186865,
+      "learning_rate": 4.676062992290686e-06,
+      "loss": 0.0243,
+      "step": 3572
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3091795467682432,
+      "learning_rate": 4.666923727345991e-06,
+      "loss": 0.0309,
+      "step": 3573
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2500044405065996,
+      "learning_rate": 4.657792222565185e-06,
+      "loss": 0.0299,
+      "step": 3574
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.27663942010892373,
+      "learning_rate": 4.6486684825697845e-06,
+      "loss": 0.0426,
+      "step": 3575
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.32554082300480536,
+      "learning_rate": 4.639552511977374e-06,
+      "loss": 0.0415,
+      "step": 3576
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.23337846135799326,
+      "learning_rate": 4.630444315401594e-06,
+      "loss": 0.0297,
+      "step": 3577
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.21699514000752693,
+      "learning_rate": 4.621343897452169e-06,
+      "loss": 0.0246,
+      "step": 3578
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2296995565273345,
+      "learning_rate": 4.612251262734864e-06,
+      "loss": 0.0282,
+      "step": 3579
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2650624798273476,
+      "learning_rate": 4.603166415851527e-06,
+      "loss": 0.0381,
+      "step": 3580
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.24872530874554657,
+      "learning_rate": 4.594089361400047e-06,
+      "loss": 0.0313,
+      "step": 3581
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.22568069053377085,
+      "learning_rate": 4.585020103974387e-06,
+      "loss": 0.0187,
+      "step": 3582
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2579961469440179,
+      "learning_rate": 4.575958648164536e-06,
+      "loss": 0.0306,
+      "step": 3583
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.25226387554943125,
+      "learning_rate": 4.5669049985565735e-06,
+      "loss": 0.0305,
+      "step": 3584
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2301970755656676,
+      "learning_rate": 4.5578591597325935e-06,
+      "loss": 0.0235,
+      "step": 3585
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.273026429858453,
+      "learning_rate": 4.54882113627076e-06,
+      "loss": 0.0295,
+      "step": 3586
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.25671518772600055,
+      "learning_rate": 4.53979093274526e-06,
+      "loss": 0.0295,
+      "step": 3587
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.24103823359464405,
+      "learning_rate": 4.530768553726348e-06,
+      "loss": 0.0281,
+      "step": 3588
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2895857997258674,
+      "learning_rate": 4.521754003780294e-06,
+      "loss": 0.0335,
+      "step": 3589
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.23744475387192743,
+      "learning_rate": 4.512747287469426e-06,
+      "loss": 0.0282,
+      "step": 3590
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.22834568551656584,
+      "learning_rate": 4.503748409352089e-06,
+      "loss": 0.0291,
+      "step": 3591
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.24846882112866903,
+      "learning_rate": 4.494757373982674e-06,
+      "loss": 0.0255,
+      "step": 3592
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.24942828909284545,
+      "learning_rate": 4.4857741859116024e-06,
+      "loss": 0.028,
+      "step": 3593
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2613375150307862,
+      "learning_rate": 4.476798849685322e-06,
+      "loss": 0.0219,
+      "step": 3594
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.26663221804814036,
+      "learning_rate": 4.467831369846301e-06,
+      "loss": 0.0359,
+      "step": 3595
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.21777006753956984,
+      "learning_rate": 4.458871750933038e-06,
+      "loss": 0.0219,
+      "step": 3596
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.25977810134181817,
+      "learning_rate": 4.449919997480047e-06,
+      "loss": 0.0283,
+      "step": 3597
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.276873794837206,
+      "learning_rate": 4.4409761140178765e-06,
+      "loss": 0.026,
+      "step": 3598
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.31331411650670776,
+      "learning_rate": 4.432040105073065e-06,
+      "loss": 0.0369,
+      "step": 3599
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.25877502691246485,
+      "learning_rate": 4.4231119751681885e-06,
+      "loss": 0.0414,
+      "step": 3600
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2487674519392905,
+      "learning_rate": 4.414191728821838e-06,
+      "loss": 0.0343,
+      "step": 3601
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.234888951728711,
+      "learning_rate": 4.405279370548587e-06,
+      "loss": 0.0313,
+      "step": 3602
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2263207543502551,
+      "learning_rate": 4.396374904859051e-06,
+      "loss": 0.0324,
+      "step": 3603
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.34164101452982987,
+      "learning_rate": 4.387478336259821e-06,
+      "loss": 0.0471,
+      "step": 3604
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2783808859779696,
+      "learning_rate": 4.3785896692535165e-06,
+      "loss": 0.0417,
+      "step": 3605
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3006577671775209,
+      "learning_rate": 4.369708908338735e-06,
+      "loss": 0.0285,
+      "step": 3606
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.20801565665914815,
+      "learning_rate": 4.360836058010096e-06,
+      "loss": 0.0216,
+      "step": 3607
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2478282971656624,
+      "learning_rate": 4.351971122758194e-06,
+      "loss": 0.0242,
+      "step": 3608
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.30743639038019077,
+      "learning_rate": 4.343114107069628e-06,
+      "loss": 0.0414,
+      "step": 3609
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.28202020301212055,
+      "learning_rate": 4.334265015426993e-06,
+      "loss": 0.0363,
+      "step": 3610
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3470685577777562,
+      "learning_rate": 4.3254238523088695e-06,
+      "loss": 0.0313,
+      "step": 3611
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.28720296255665057,
+      "learning_rate": 4.316590622189815e-06,
+      "loss": 0.0329,
+      "step": 3612
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2573910794628964,
+      "learning_rate": 4.307765329540394e-06,
+      "loss": 0.0252,
+      "step": 3613
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.23434111032246852,
+      "learning_rate": 4.298947978827128e-06,
+      "loss": 0.033,
+      "step": 3614
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.28824083722117905,
+      "learning_rate": 4.290138574512546e-06,
+      "loss": 0.0279,
+      "step": 3615
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.24409707253133334,
+      "learning_rate": 4.2813371210551294e-06,
+      "loss": 0.0283,
+      "step": 3616
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.21835811457671117,
+      "learning_rate": 4.272543622909355e-06,
+      "loss": 0.0252,
+      "step": 3617
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1936975565814078,
+      "learning_rate": 4.263758084525656e-06,
+      "loss": 0.0187,
+      "step": 3618
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.27033475993678024,
+      "learning_rate": 4.254980510350464e-06,
+      "loss": 0.0257,
+      "step": 3619
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2880795106332577,
+      "learning_rate": 4.246210904826149e-06,
+      "loss": 0.0263,
+      "step": 3620
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.21054254750349924,
+      "learning_rate": 4.237449272391072e-06,
+      "loss": 0.0221,
+      "step": 3621
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.23125726971927127,
+      "learning_rate": 4.228695617479541e-06,
+      "loss": 0.0217,
+      "step": 3622
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.22920103881009266,
+      "learning_rate": 4.219949944521842e-06,
+      "loss": 0.0282,
+      "step": 3623
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2115455622915119,
+      "learning_rate": 4.2112122579442015e-06,
+      "loss": 0.0296,
+      "step": 3624
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2582877143193991,
+      "learning_rate": 4.202482562168832e-06,
+      "loss": 0.0284,
+      "step": 3625
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.257448546398354,
+      "learning_rate": 4.193760861613865e-06,
+      "loss": 0.0287,
+      "step": 3626
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2342688001988582,
+      "learning_rate": 4.185047160693432e-06,
+      "loss": 0.0344,
+      "step": 3627
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.19579876449193206,
+      "learning_rate": 4.176341463817573e-06,
+      "loss": 0.0227,
+      "step": 3628
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.20072648826742953,
+      "learning_rate": 4.167643775392305e-06,
+      "loss": 0.0227,
+      "step": 3629
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.27461042878774905,
+      "learning_rate": 4.1589540998195695e-06,
+      "loss": 0.0346,
+      "step": 3630
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.23646947085972797,
+      "learning_rate": 4.150272441497276e-06,
+      "loss": 0.0307,
+      "step": 3631
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2571087392858843,
+      "learning_rate": 4.141598804819256e-06,
+      "loss": 0.0355,
+      "step": 3632
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.20020142522644224,
+      "learning_rate": 4.132933194175299e-06,
+      "loss": 0.0203,
+      "step": 3633
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.40853096605429784,
+      "learning_rate": 4.124275613951114e-06,
+      "loss": 0.0491,
+      "step": 3634
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.36319887601451634,
+      "learning_rate": 4.115626068528362e-06,
+      "loss": 0.0279,
+      "step": 3635
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.27277745555992244,
+      "learning_rate": 4.106984562284633e-06,
+      "loss": 0.0317,
+      "step": 3636
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.21533601026253893,
+      "learning_rate": 4.0983510995934365e-06,
+      "loss": 0.0253,
+      "step": 3637
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.28834596956729397,
+      "learning_rate": 4.089725684824235e-06,
+      "loss": 0.029,
+      "step": 3638
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.23269264092864095,
+      "learning_rate": 4.081108322342389e-06,
+      "loss": 0.0249,
+      "step": 3639
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2555612909769337,
+      "learning_rate": 4.07249901650921e-06,
+      "loss": 0.0271,
+      "step": 3640
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.32616875698403835,
+      "learning_rate": 4.0638977716819105e-06,
+      "loss": 0.0393,
+      "step": 3641
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.22646813236034966,
+      "learning_rate": 4.055304592213645e-06,
+      "loss": 0.0183,
+      "step": 3642
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2554617697085855,
+      "learning_rate": 4.046719482453461e-06,
+      "loss": 0.0242,
+      "step": 3643
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.32966515704916,
+      "learning_rate": 4.038142446746342e-06,
+      "loss": 0.0483,
+      "step": 3644
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.22952046321610411,
+      "learning_rate": 4.029573489433179e-06,
+      "loss": 0.0205,
+      "step": 3645
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.17493135242590682,
+      "learning_rate": 4.021012614850779e-06,
+      "loss": 0.0172,
+      "step": 3646
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1928107541498316,
+      "learning_rate": 4.012459827331841e-06,
+      "loss": 0.0206,
+      "step": 3647
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2356499327867729,
+      "learning_rate": 4.003915131204996e-06,
+      "loss": 0.0277,
+      "step": 3648
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.39366916856177986,
+      "learning_rate": 3.995378530794754e-06,
+      "loss": 0.05,
+      "step": 3649
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.23756317691259934,
+      "learning_rate": 3.986850030421554e-06,
+      "loss": 0.0241,
+      "step": 3650
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.29357554855893153,
+      "learning_rate": 3.97832963440171e-06,
+      "loss": 0.0287,
+      "step": 3651
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2522710228865532,
+      "learning_rate": 3.969817347047451e-06,
+      "loss": 0.0264,
+      "step": 3652
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2546196525397989,
+      "learning_rate": 3.961313172666898e-06,
+      "loss": 0.0284,
+      "step": 3653
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.19070338607451412,
+      "learning_rate": 3.952817115564076e-06,
+      "loss": 0.0169,
+      "step": 3654
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2269978205123538,
+      "learning_rate": 3.944329180038875e-06,
+      "loss": 0.0285,
+      "step": 3655
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2821964657057303,
+      "learning_rate": 3.935849370387104e-06,
+      "loss": 0.0272,
+      "step": 3656
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.22076824375348655,
+      "learning_rate": 3.927377690900436e-06,
+      "loss": 0.0247,
+      "step": 3657
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.21780141300077158,
+      "learning_rate": 3.91891414586645e-06,
+      "loss": 0.0282,
+      "step": 3658
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3060406446126063,
+      "learning_rate": 3.91045873956859e-06,
+      "loss": 0.0352,
+      "step": 3659
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2893880608826457,
+      "learning_rate": 3.902011476286196e-06,
+      "loss": 0.0318,
+      "step": 3660
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.25120418334859473,
+      "learning_rate": 3.893572360294471e-06,
+      "loss": 0.0281,
+      "step": 3661
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.27315269351638816,
+      "learning_rate": 3.885141395864509e-06,
+      "loss": 0.0232,
+      "step": 3662
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.27229528279275966,
+      "learning_rate": 3.876718587263278e-06,
+      "loss": 0.0318,
+      "step": 3663
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2782702484612129,
+      "learning_rate": 3.868303938753599e-06,
+      "loss": 0.0335,
+      "step": 3664
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.18907673797905694,
+      "learning_rate": 3.859897454594192e-06,
+      "loss": 0.0235,
+      "step": 3665
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.23696425237289911,
+      "learning_rate": 3.851499139039618e-06,
+      "loss": 0.0255,
+      "step": 3666
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2530445441962811,
+      "learning_rate": 3.843108996340323e-06,
+      "loss": 0.0231,
+      "step": 3667
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1827468819515304,
+      "learning_rate": 3.834727030742613e-06,
+      "loss": 0.0173,
+      "step": 3668
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.25870133838473247,
+      "learning_rate": 3.826353246488641e-06,
+      "loss": 0.0331,
+      "step": 3669
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.26622786548554345,
+      "learning_rate": 3.817987647816437e-06,
+      "loss": 0.0368,
+      "step": 3670
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.32275130924443896,
+      "learning_rate": 3.809630238959887e-06,
+      "loss": 0.0367,
+      "step": 3671
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.25701007813794574,
+      "learning_rate": 3.8012810241487175e-06,
+      "loss": 0.0246,
+      "step": 3672
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.22454690479132108,
+      "learning_rate": 3.7929400076085255e-06,
+      "loss": 0.0302,
+      "step": 3673
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.26587275563033913,
+      "learning_rate": 3.7846071935607408e-06,
+      "loss": 0.0326,
+      "step": 3674
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2474751604624967,
+      "learning_rate": 3.7762825862226637e-06,
+      "loss": 0.0312,
+      "step": 3675
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.30273048644491024,
+      "learning_rate": 3.767966189807415e-06,
+      "loss": 0.0369,
+      "step": 3676
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2579598294846661,
+      "learning_rate": 3.7596580085239897e-06,
+      "loss": 0.0294,
+      "step": 3677
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2648196493565275,
+      "learning_rate": 3.7513580465771893e-06,
+      "loss": 0.0258,
+      "step": 3678
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3436867699123688,
+      "learning_rate": 3.7430663081676977e-06,
+      "loss": 0.0559,
+      "step": 3679
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3231971801587294,
+      "learning_rate": 3.734782797491998e-06,
+      "loss": 0.0412,
+      "step": 3680
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.21947458017702118,
+      "learning_rate": 3.7265075187424373e-06,
+      "loss": 0.0209,
+      "step": 3681
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.23357392740506586,
+      "learning_rate": 3.7182404761071735e-06,
+      "loss": 0.0304,
+      "step": 3682
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.23760398389086507,
+      "learning_rate": 3.7099816737702197e-06,
+      "loss": 0.0267,
+      "step": 3683
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.29003768817820047,
+      "learning_rate": 3.7017311159113956e-06,
+      "loss": 0.0271,
+      "step": 3684
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.24743537921386893,
+      "learning_rate": 3.6934888067063667e-06,
+      "loss": 0.0256,
+      "step": 3685
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2274675625296988,
+      "learning_rate": 3.68525475032661e-06,
+      "loss": 0.0284,
+      "step": 3686
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2360116794638995,
+      "learning_rate": 3.677028950939434e-06,
+      "loss": 0.023,
+      "step": 3687
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.24521057659949916,
+      "learning_rate": 3.6688114127079665e-06,
+      "loss": 0.0228,
+      "step": 3688
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2635946796584002,
+      "learning_rate": 3.6606021397911605e-06,
+      "loss": 0.0273,
+      "step": 3689
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.24556618863648857,
+      "learning_rate": 3.652401136343768e-06,
+      "loss": 0.03,
+      "step": 3690
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2609608908122948,
+      "learning_rate": 3.6442084065163784e-06,
+      "loss": 0.0344,
+      "step": 3691
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.28327570018181836,
+      "learning_rate": 3.636023954455372e-06,
+      "loss": 0.0301,
+      "step": 3692
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.23041111823593596,
+      "learning_rate": 3.6278477843029603e-06,
+      "loss": 0.0287,
+      "step": 3693
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2625289730703892,
+      "learning_rate": 3.6196799001971416e-06,
+      "loss": 0.0256,
+      "step": 3694
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1630802210645409,
+      "learning_rate": 3.6115203062717386e-06,
+      "loss": 0.0228,
+      "step": 3695
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3190133451470103,
+      "learning_rate": 3.6033690066563765e-06,
+      "loss": 0.0414,
+      "step": 3696
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2435520707321006,
+      "learning_rate": 3.5952260054764663e-06,
+      "loss": 0.0283,
+      "step": 3697
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.22822638897501876,
+      "learning_rate": 3.5870913068532455e-06,
+      "loss": 0.0257,
+      "step": 3698
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2763722961816296,
+      "learning_rate": 3.5789649149037197e-06,
+      "loss": 0.0305,
+      "step": 3699
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2695684269130349,
+      "learning_rate": 3.5708468337407177e-06,
+      "loss": 0.0291,
+      "step": 3700
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.20398215857725774,
+      "learning_rate": 3.562737067472841e-06,
+      "loss": 0.0195,
+      "step": 3701
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2502541740080814,
+      "learning_rate": 3.554635620204503e-06,
+      "loss": 0.0315,
+      "step": 3702
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.24771777420698718,
+      "learning_rate": 3.546542496035883e-06,
+      "loss": 0.0328,
+      "step": 3703
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.25344037083741355,
+      "learning_rate": 3.5384576990629672e-06,
+      "loss": 0.0224,
+      "step": 3704
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.26966825403295247,
+      "learning_rate": 3.53038123337752e-06,
+      "loss": 0.0457,
+      "step": 3705
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4245201824319466,
+      "learning_rate": 3.5223131030670942e-06,
+      "loss": 0.0387,
+      "step": 3706
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2605416601937401,
+      "learning_rate": 3.5142533122150147e-06,
+      "loss": 0.0341,
+      "step": 3707
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2730210811890883,
+      "learning_rate": 3.506201864900396e-06,
+      "loss": 0.0277,
+      "step": 3708
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2157021971740995,
+      "learning_rate": 3.4981587651981185e-06,
+      "loss": 0.0288,
+      "step": 3709
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2760890194326243,
+      "learning_rate": 3.490124017178851e-06,
+      "loss": 0.0322,
+      "step": 3710
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.27453212981064085,
+      "learning_rate": 3.482097624909022e-06,
+      "loss": 0.0333,
+      "step": 3711
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.253931058992105,
+      "learning_rate": 3.474079592450845e-06,
+      "loss": 0.0282,
+      "step": 3712
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.24944489661127137,
+      "learning_rate": 3.466069923862283e-06,
+      "loss": 0.0248,
+      "step": 3713
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2315743324149911,
+      "learning_rate": 3.458068623197097e-06,
+      "loss": 0.0299,
+      "step": 3714
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.23188678469876328,
+      "learning_rate": 3.4500756945047774e-06,
+      "loss": 0.0175,
+      "step": 3715
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.23487875366002967,
+      "learning_rate": 3.442091141830608e-06,
+      "loss": 0.026,
+      "step": 3716
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2520699045582025,
+      "learning_rate": 3.4341149692156074e-06,
+      "loss": 0.0262,
+      "step": 3717
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.22399663690028476,
+      "learning_rate": 3.426147180696577e-06,
+      "loss": 0.0334,
+      "step": 3718
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.25049011051335446,
+      "learning_rate": 3.4181877803060528e-06,
+      "loss": 0.0299,
+      "step": 3719
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2646873698859137,
+      "learning_rate": 3.4102367720723438e-06,
+      "loss": 0.0315,
+      "step": 3720
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2442497314795245,
+      "learning_rate": 3.402294160019499e-06,
+      "loss": 0.0297,
+      "step": 3721
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3376258407979168,
+      "learning_rate": 3.394359948167325e-06,
+      "loss": 0.0407,
+      "step": 3722
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2851709205302547,
+      "learning_rate": 3.386434140531378e-06,
+      "loss": 0.0291,
+      "step": 3723
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3110857792115305,
+      "learning_rate": 3.3785167411229523e-06,
+      "loss": 0.0428,
+      "step": 3724
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.24018609717101114,
+      "learning_rate": 3.3706077539490933e-06,
+      "loss": 0.0259,
+      "step": 3725
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.24668818539972917,
+      "learning_rate": 3.362707183012597e-06,
+      "loss": 0.0232,
+      "step": 3726
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.23018877351239234,
+      "learning_rate": 3.354815032311978e-06,
+      "loss": 0.0303,
+      "step": 3727
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.22833097773982017,
+      "learning_rate": 3.34693130584151e-06,
+      "loss": 0.0215,
+      "step": 3728
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.24650305059085142,
+      "learning_rate": 3.3390560075911906e-06,
+      "loss": 0.0277,
+      "step": 3729
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.17666722183901298,
+      "learning_rate": 3.331189141546758e-06,
+      "loss": 0.0202,
+      "step": 3730
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.26592299631376487,
+      "learning_rate": 3.3233307116896874e-06,
+      "loss": 0.0443,
+      "step": 3731
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.22041344832436627,
+      "learning_rate": 3.3154807219971684e-06,
+      "loss": 0.024,
+      "step": 3732
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.34140833978428003,
+      "learning_rate": 3.307639176442137e-06,
+      "loss": 0.0325,
+      "step": 3733
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.23034179552941636,
+      "learning_rate": 3.299806078993242e-06,
+      "loss": 0.0202,
+      "step": 3734
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1835631081191456,
+      "learning_rate": 3.2919814336148657e-06,
+      "loss": 0.0205,
+      "step": 3735
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.20295680858908874,
+      "learning_rate": 3.2841652442671033e-06,
+      "loss": 0.0222,
+      "step": 3736
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2775041089741753,
+      "learning_rate": 3.276357514905788e-06,
+      "loss": 0.0322,
+      "step": 3737
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.22677679450652197,
+      "learning_rate": 3.2685582494824386e-06,
+      "loss": 0.0285,
+      "step": 3738
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.28460758151071236,
+      "learning_rate": 3.260767451944338e-06,
+      "loss": 0.0383,
+      "step": 3739
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.23390609406113763,
+      "learning_rate": 3.252985126234434e-06,
+      "loss": 0.0309,
+      "step": 3740
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.20212903136937763,
+      "learning_rate": 3.245211276291427e-06,
+      "loss": 0.0258,
+      "step": 3741
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.26395564880049716,
+      "learning_rate": 3.237445906049694e-06,
+      "loss": 0.0457,
+      "step": 3742
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.29903038510175467,
+      "learning_rate": 3.229689019439348e-06,
+      "loss": 0.0315,
+      "step": 3743
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.25464142224590275,
+      "learning_rate": 3.2219406203861903e-06,
+      "loss": 0.0321,
+      "step": 3744
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.28081431020683456,
+      "learning_rate": 3.2142007128117393e-06,
+      "loss": 0.0265,
+      "step": 3745
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.281312485217518,
+      "learning_rate": 3.2064693006332013e-06,
+      "loss": 0.028,
+      "step": 3746
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.22573948873242403,
+      "learning_rate": 3.1987463877634962e-06,
+      "loss": 0.0204,
+      "step": 3747
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.22625983509268627,
+      "learning_rate": 3.1910319781112364e-06,
+      "loss": 0.0235,
+      "step": 3748
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2566258691874267,
+      "learning_rate": 3.1833260755807392e-06,
+      "loss": 0.0261,
+      "step": 3749
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.216310359009026,
+      "learning_rate": 3.1756286840719987e-06,
+      "loss": 0.0293,
+      "step": 3750
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2803729209720778,
+      "learning_rate": 3.16793980748072e-06,
+      "loss": 0.0373,
+      "step": 3751
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2628738760749563,
+      "learning_rate": 3.160259449698282e-06,
+      "loss": 0.0312,
+      "step": 3752
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.23420699774006093,
+      "learning_rate": 3.1525876146117707e-06,
+      "loss": 0.0315,
+      "step": 3753
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2373883684596302,
+      "learning_rate": 3.144924306103938e-06,
+      "loss": 0.0253,
+      "step": 3754
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.21807232689411402,
+      "learning_rate": 3.1372695280532415e-06,
+      "loss": 0.0314,
+      "step": 3755
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.26653987743804225,
+      "learning_rate": 3.129623284333805e-06,
+      "loss": 0.0357,
+      "step": 3756
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.25989470465701053,
+      "learning_rate": 3.1219855788154385e-06,
+      "loss": 0.0293,
+      "step": 3757
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.23634236663278327,
+      "learning_rate": 3.1143564153636395e-06,
+      "loss": 0.0179,
+      "step": 3758
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.23741298135441422,
+      "learning_rate": 3.1067357978395663e-06,
+      "loss": 0.0263,
+      "step": 3759
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.20682910731405757,
+      "learning_rate": 3.0991237301000664e-06,
+      "loss": 0.0218,
+      "step": 3760
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.26802454553833704,
+      "learning_rate": 3.0915202159976453e-06,
+      "loss": 0.0309,
+      "step": 3761
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.21710761028010478,
+      "learning_rate": 3.083925259380498e-06,
+      "loss": 0.0217,
+      "step": 3762
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2641650215307287,
+      "learning_rate": 3.0763388640924698e-06,
+      "loss": 0.032,
+      "step": 3763
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.24072832019634527,
+      "learning_rate": 3.068761033973087e-06,
+      "loss": 0.0305,
+      "step": 3764
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.27969197690566716,
+      "learning_rate": 3.0611917728575347e-06,
+      "loss": 0.0324,
+      "step": 3765
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.20708361449276919,
+      "learning_rate": 3.053631084576667e-06,
+      "loss": 0.0265,
+      "step": 3766
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2899234297188242,
+      "learning_rate": 3.046078972956985e-06,
+      "loss": 0.0331,
+      "step": 3767
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.24798203168945265,
+      "learning_rate": 3.038535441820669e-06,
+      "loss": 0.0205,
+      "step": 3768
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.20767294205972725,
+      "learning_rate": 3.0310004949855366e-06,
+      "loss": 0.025,
+      "step": 3769
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.21678776764856741,
+      "learning_rate": 3.0234741362650787e-06,
+      "loss": 0.0276,
+      "step": 3770
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3020937016231474,
+      "learning_rate": 3.0159563694684245e-06,
+      "loss": 0.0505,
+      "step": 3771
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.23336752682863882,
+      "learning_rate": 3.008447198400368e-06,
+      "loss": 0.0223,
+      "step": 3772
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.21756051435717852,
+      "learning_rate": 3.0009466268613384e-06,
+      "loss": 0.027,
+      "step": 3773
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2660732974190698,
+      "learning_rate": 2.9934546586474346e-06,
+      "loss": 0.0283,
+      "step": 3774
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.18405623723241546,
+      "learning_rate": 2.985971297550374e-06,
+      "loss": 0.0187,
+      "step": 3775
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.21215769391278613,
+      "learning_rate": 2.9784965473575434e-06,
+      "loss": 0.0231,
+      "step": 3776
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.19297202644192088,
+      "learning_rate": 2.9710304118519473e-06,
+      "loss": 0.0198,
+      "step": 3777
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2579171806283081,
+      "learning_rate": 2.9635728948122542e-06,
+      "loss": 0.023,
+      "step": 3778
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.28781569138454693,
+      "learning_rate": 2.95612400001275e-06,
+      "loss": 0.0362,
+      "step": 3779
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.24115771119516768,
+      "learning_rate": 2.9486837312233742e-06,
+      "loss": 0.0222,
+      "step": 3780
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.26588425886678607,
+      "learning_rate": 2.9412520922096834e-06,
+      "loss": 0.0305,
+      "step": 3781
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.25585837036401493,
+      "learning_rate": 2.93382908673288e-06,
+      "loss": 0.0304,
+      "step": 3782
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3162272853760593,
+      "learning_rate": 2.926414718549797e-06,
+      "loss": 0.0363,
+      "step": 3783
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2875859075196345,
+      "learning_rate": 2.9190089914128837e-06,
+      "loss": 0.0312,
+      "step": 3784
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2679100483916866,
+      "learning_rate": 2.911611909070229e-06,
+      "loss": 0.0248,
+      "step": 3785
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.23112737974506606,
+      "learning_rate": 2.9042234752655417e-06,
+      "loss": 0.0205,
+      "step": 3786
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.20902051345813139,
+      "learning_rate": 2.8968436937381515e-06,
+      "loss": 0.024,
+      "step": 3787
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.28679984138719145,
+      "learning_rate": 2.889472568223015e-06,
+      "loss": 0.0317,
+      "step": 3788
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.38947517921805536,
+      "learning_rate": 2.8821101024506947e-06,
+      "loss": 0.0397,
+      "step": 3789
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2625566137662427,
+      "learning_rate": 2.874756300147388e-06,
+      "loss": 0.0259,
+      "step": 3790
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3466975462705927,
+      "learning_rate": 2.867411165034901e-06,
+      "loss": 0.0433,
+      "step": 3791
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.30872740446645297,
+      "learning_rate": 2.8600747008306417e-06,
+      "loss": 0.0287,
+      "step": 3792
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.24879147066498217,
+      "learning_rate": 2.8527469112476524e-06,
+      "loss": 0.0271,
+      "step": 3793
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.22946090485031406,
+      "learning_rate": 2.8454277999945603e-06,
+      "loss": 0.0186,
+      "step": 3794
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3252357251599501,
+      "learning_rate": 2.8381173707756214e-06,
+      "loss": 0.0337,
+      "step": 3795
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.22712515976584763,
+      "learning_rate": 2.8308156272906794e-06,
+      "loss": 0.0229,
+      "step": 3796
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.20611607509466556,
+      "learning_rate": 2.8235225732352043e-06,
+      "loss": 0.031,
+      "step": 3797
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2323427626660013,
+      "learning_rate": 2.8162382123002418e-06,
+      "loss": 0.0179,
+      "step": 3798
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2634233928672666,
+      "learning_rate": 2.8089625481724604e-06,
+      "loss": 0.027,
+      "step": 3799
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.23105383007828828,
+      "learning_rate": 2.8016955845341143e-06,
+      "loss": 0.0235,
+      "step": 3800
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2677735359976356,
+      "learning_rate": 2.794437325063064e-06,
+      "loss": 0.0227,
+      "step": 3801
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.24656583525090428,
+      "learning_rate": 2.7871877734327514e-06,
+      "loss": 0.0297,
+      "step": 3802
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.23939083838791006,
+      "learning_rate": 2.7799469333122275e-06,
+      "loss": 0.0255,
+      "step": 3803
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.24008755733334192,
+      "learning_rate": 2.772714808366115e-06,
+      "loss": 0.0316,
+      "step": 3804
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2231639043966675,
+      "learning_rate": 2.76549140225465e-06,
+      "loss": 0.0303,
+      "step": 3805
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.25595756277460696,
+      "learning_rate": 2.758276718633628e-06,
+      "loss": 0.027,
+      "step": 3806
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.211849651128429,
+      "learning_rate": 2.751070761154453e-06,
+      "loss": 0.0201,
+      "step": 3807
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.26129634418740444,
+      "learning_rate": 2.743873533464105e-06,
+      "loss": 0.0265,
+      "step": 3808
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.24285469137835775,
+      "learning_rate": 2.7366850392051468e-06,
+      "loss": 0.0235,
+      "step": 3809
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.28917501167290083,
+      "learning_rate": 2.7295052820157097e-06,
+      "loss": 0.029,
+      "step": 3810
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.277645363473503,
+      "learning_rate": 2.722334265529527e-06,
+      "loss": 0.0318,
+      "step": 3811
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2101197108159598,
+      "learning_rate": 2.715171993375878e-06,
+      "loss": 0.0219,
+      "step": 3812
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3028963455553397,
+      "learning_rate": 2.7080184691796474e-06,
+      "loss": 0.0292,
+      "step": 3813
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.22889463668631502,
+      "learning_rate": 2.7008736965612658e-06,
+      "loss": 0.018,
+      "step": 3814
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.24035723576142085,
+      "learning_rate": 2.6937376791367566e-06,
+      "loss": 0.0256,
+      "step": 3815
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2424393006312234,
+      "learning_rate": 2.6866104205176925e-06,
+      "loss": 0.0333,
+      "step": 3816
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.24892566481864783,
+      "learning_rate": 2.679491924311226e-06,
+      "loss": 0.0237,
+      "step": 3817
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2524314302875021,
+      "learning_rate": 2.67238219412008e-06,
+      "loss": 0.0269,
+      "step": 3818
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3008899987543034,
+      "learning_rate": 2.6652812335425184e-06,
+      "loss": 0.0384,
+      "step": 3819
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2663023530727557,
+      "learning_rate": 2.6581890461723925e-06,
+      "loss": 0.0364,
+      "step": 3820
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.23916030765402485,
+      "learning_rate": 2.651105635599094e-06,
+      "loss": 0.0288,
+      "step": 3821
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.314366417534471,
+      "learning_rate": 2.6440310054075877e-06,
+      "loss": 0.0396,
+      "step": 3822
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.19393635409315052,
+      "learning_rate": 2.6369651591783774e-06,
+      "loss": 0.0222,
+      "step": 3823
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2970231715412625,
+      "learning_rate": 2.629908100487544e-06,
+      "loss": 0.0367,
+      "step": 3824
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2730624897084373,
+      "learning_rate": 2.6228598329066902e-06,
+      "loss": 0.0329,
+      "step": 3825
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2221304474051573,
+      "learning_rate": 2.6158203600030076e-06,
+      "loss": 0.0208,
+      "step": 3826
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.21295674198523606,
+      "learning_rate": 2.6087896853392037e-06,
+      "loss": 0.0213,
+      "step": 3827
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2634736057395541,
+      "learning_rate": 2.6017678124735545e-06,
+      "loss": 0.0249,
+      "step": 3828
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2801856325424695,
+      "learning_rate": 2.594754744959862e-06,
+      "loss": 0.0331,
+      "step": 3829
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.38273736644602885,
+      "learning_rate": 2.5877504863474933e-06,
+      "loss": 0.0441,
+      "step": 3830
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.23987318038548458,
+      "learning_rate": 2.58075504018134e-06,
+      "loss": 0.0247,
+      "step": 3831
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3165671210408318,
+      "learning_rate": 2.5737684100018446e-06,
+      "loss": 0.0504,
+      "step": 3832
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.26761475194029594,
+      "learning_rate": 2.566790599344973e-06,
+      "loss": 0.0264,
+      "step": 3833
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.28997572133362387,
+      "learning_rate": 2.5598216117422547e-06,
+      "loss": 0.0292,
+      "step": 3834
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.23359317715679853,
+      "learning_rate": 2.552861450720725e-06,
+      "loss": 0.0239,
+      "step": 3835
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.225136801073581,
+      "learning_rate": 2.5459101198029724e-06,
+      "loss": 0.0188,
+      "step": 3836
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2199612868449696,
+      "learning_rate": 2.538967622507098e-06,
+      "loss": 0.0228,
+      "step": 3837
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.263409052655707,
+      "learning_rate": 2.532033962346754e-06,
+      "loss": 0.0244,
+      "step": 3838
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.23126387425634926,
+      "learning_rate": 2.525109142831095e-06,
+      "loss": 0.0264,
+      "step": 3839
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.22309392442868756,
+      "learning_rate": 2.5181931674648265e-06,
+      "loss": 0.022,
+      "step": 3840
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.24698059805352435,
+      "learning_rate": 2.5112860397481553e-06,
+      "loss": 0.0276,
+      "step": 3841
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4268208681918181,
+      "learning_rate": 2.50438776317683e-06,
+      "loss": 0.0494,
+      "step": 3842
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2269383387911252,
+      "learning_rate": 2.497498341242104e-06,
+      "loss": 0.0291,
+      "step": 3843
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.25532006069273766,
+      "learning_rate": 2.490617777430766e-06,
+      "loss": 0.0289,
+      "step": 3844
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.19921207416793138,
+      "learning_rate": 2.4837460752251e-06,
+      "loss": 0.0219,
+      "step": 3845
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.23851431720373856,
+      "learning_rate": 2.476883238102925e-06,
+      "loss": 0.0269,
+      "step": 3846
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.28687080540218757,
+      "learning_rate": 2.4700292695375596e-06,
+      "loss": 0.0336,
+      "step": 3847
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2877835383842551,
+      "learning_rate": 2.4631841729978435e-06,
+      "loss": 0.0386,
+      "step": 3848
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.21595209508503585,
+      "learning_rate": 2.456347951948115e-06,
+      "loss": 0.027,
+      "step": 3849
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.25590705427331534,
+      "learning_rate": 2.449520609848237e-06,
+      "loss": 0.0285,
+      "step": 3850
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2914738983095523,
+      "learning_rate": 2.442702150153562e-06,
+      "loss": 0.0302,
+      "step": 3851
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.22516376244575387,
+      "learning_rate": 2.4358925763149557e-06,
+      "loss": 0.0266,
+      "step": 3852
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2483287171253741,
+      "learning_rate": 2.4290918917787876e-06,
+      "loss": 0.0265,
+      "step": 3853
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.32799978364742516,
+      "learning_rate": 2.4223000999869227e-06,
+      "loss": 0.0379,
+      "step": 3854
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.21206941491648676,
+      "learning_rate": 2.4155172043767337e-06,
+      "loss": 0.0253,
+      "step": 3855
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2549331586084214,
+      "learning_rate": 2.4087432083810792e-06,
+      "loss": 0.0241,
+      "step": 3856
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.26321166106479543,
+      "learning_rate": 2.401978115428325e-06,
+      "loss": 0.0232,
+      "step": 3857
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2478637469868672,
+      "learning_rate": 2.395221928942322e-06,
+      "loss": 0.029,
+      "step": 3858
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.25795024024458996,
+      "learning_rate": 2.388474652342416e-06,
+      "loss": 0.0296,
+      "step": 3859
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2210709536496926,
+      "learning_rate": 2.3817362890434526e-06,
+      "loss": 0.0384,
+      "step": 3860
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.27839401340087205,
+      "learning_rate": 2.375006842455756e-06,
+      "loss": 0.0324,
+      "step": 3861
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.29559086088363035,
+      "learning_rate": 2.3682863159851377e-06,
+      "loss": 0.0322,
+      "step": 3862
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.258958163884098,
+      "learning_rate": 2.3615747130329013e-06,
+      "loss": 0.045,
+      "step": 3863
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.23485219378085304,
+      "learning_rate": 2.3548720369958256e-06,
+      "loss": 0.0288,
+      "step": 3864
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2157679964756096,
+      "learning_rate": 2.3481782912661788e-06,
+      "loss": 0.0228,
+      "step": 3865
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3899911506272396,
+      "learning_rate": 2.3414934792317047e-06,
+      "loss": 0.0413,
+      "step": 3866
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2272902507691261,
+      "learning_rate": 2.334817604275632e-06,
+      "loss": 0.0268,
+      "step": 3867
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.22339759350247212,
+      "learning_rate": 2.3281506697766522e-06,
+      "loss": 0.0252,
+      "step": 3868
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2745319398362205,
+      "learning_rate": 2.3214926791089563e-06,
+      "loss": 0.0263,
+      "step": 3869
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3004511222885687,
+      "learning_rate": 2.3148436356421813e-06,
+      "loss": 0.0405,
+      "step": 3870
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.18937862543847042,
+      "learning_rate": 2.3082035427414585e-06,
+      "loss": 0.0212,
+      "step": 3871
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2781247439008241,
+      "learning_rate": 2.301572403767369e-06,
+      "loss": 0.028,
+      "step": 3872
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.25762808590872704,
+      "learning_rate": 2.2949502220759866e-06,
+      "loss": 0.0247,
+      "step": 3873
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2648198885000602,
+      "learning_rate": 2.2883370010188232e-06,
+      "loss": 0.0364,
+      "step": 3874
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.22559464012633032,
+      "learning_rate": 2.2817327439428836e-06,
+      "loss": 0.0267,
+      "step": 3875
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2344780456991246,
+      "learning_rate": 2.2751374541906122e-06,
+      "loss": 0.0327,
+      "step": 3876
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.22897060453096788,
+      "learning_rate": 2.26855113509993e-06,
+      "loss": 0.0278,
+      "step": 3877
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.23947667601923092,
+      "learning_rate": 2.261973790004217e-06,
+      "loss": 0.0281,
+      "step": 3878
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2541858893822508,
+      "learning_rate": 2.2554054222323018e-06,
+      "loss": 0.0265,
+      "step": 3879
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2625970621268977,
+      "learning_rate": 2.2488460351084827e-06,
+      "loss": 0.0302,
+      "step": 3880
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.23400592354466196,
+      "learning_rate": 2.242295631952496e-06,
+      "loss": 0.0231,
+      "step": 3881
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.24455926478727372,
+      "learning_rate": 2.235754216079551e-06,
+      "loss": 0.0295,
+      "step": 3882
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.29506134373317705,
+      "learning_rate": 2.229221790800291e-06,
+      "loss": 0.0319,
+      "step": 3883
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.25740314988611324,
+      "learning_rate": 2.2226983594208187e-06,
+      "loss": 0.0342,
+      "step": 3884
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.32531390161423046,
+      "learning_rate": 2.216183925242681e-06,
+      "loss": 0.0268,
+      "step": 3885
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.25434132493678374,
+      "learning_rate": 2.209678491562881e-06,
+      "loss": 0.0227,
+      "step": 3886
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2397500094154205,
+      "learning_rate": 2.2031820616738477e-06,
+      "loss": 0.0202,
+      "step": 3887
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2139124016890787,
+      "learning_rate": 2.1966946388634746e-06,
+      "loss": 0.0291,
+      "step": 3888
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2131140088158562,
+      "learning_rate": 2.190216226415074e-06,
+      "loss": 0.0207,
+      "step": 3889
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2961527295114186,
+      "learning_rate": 2.1837468276074227e-06,
+      "loss": 0.0418,
+      "step": 3890
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.23153809405938847,
+      "learning_rate": 2.1772864457147126e-06,
+      "loss": 0.0268,
+      "step": 3891
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.28574910654176955,
+      "learning_rate": 2.1708350840065927e-06,
+      "loss": 0.0344,
+      "step": 3892
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.23484012227863682,
+      "learning_rate": 2.164392745748125e-06,
+      "loss": 0.0257,
+      "step": 3893
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.23499878683943798,
+      "learning_rate": 2.1579594341998235e-06,
+      "loss": 0.0337,
+      "step": 3894
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2245127594514266,
+      "learning_rate": 2.151535152617625e-06,
+      "loss": 0.0246,
+      "step": 3895
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.22720426068665847,
+      "learning_rate": 2.1451199042529035e-06,
+      "loss": 0.0211,
+      "step": 3896
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.17705131105658264,
+      "learning_rate": 2.1387136923524475e-06,
+      "loss": 0.017,
+      "step": 3897
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.25607005756259693,
+      "learning_rate": 2.1323165201584863e-06,
+      "loss": 0.0269,
+      "step": 3898
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.19983322148628807,
+      "learning_rate": 2.125928390908658e-06,
+      "loss": 0.023,
+      "step": 3899
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2594664579030888,
+      "learning_rate": 2.1195493078360486e-06,
+      "loss": 0.0205,
+      "step": 3900
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2109651459131922,
+      "learning_rate": 2.113179274169137e-06,
+      "loss": 0.0227,
+      "step": 3901
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.24432628438712112,
+      "learning_rate": 2.1068182931318424e-06,
+      "loss": 0.0229,
+      "step": 3902
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.19328943586785619,
+      "learning_rate": 2.1004663679434987e-06,
+      "loss": 0.0148,
+      "step": 3903
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.22900019726057594,
+      "learning_rate": 2.0941235018188543e-06,
+      "loss": 0.0318,
+      "step": 3904
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.27477682985090945,
+      "learning_rate": 2.0877896979680654e-06,
+      "loss": 0.0248,
+      "step": 3905
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.26366659088491357,
+      "learning_rate": 2.0814649595967194e-06,
+      "loss": 0.0333,
+      "step": 3906
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.21312036279908078,
+      "learning_rate": 2.0751492899057957e-06,
+      "loss": 0.0204,
+      "step": 3907
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.20255649311395135,
+      "learning_rate": 2.0688426920916992e-06,
+      "loss": 0.0229,
+      "step": 3908
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.29645066302079465,
+      "learning_rate": 2.062545169346235e-06,
+      "loss": 0.0228,
+      "step": 3909
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.29417591617827665,
+      "learning_rate": 2.05625672485662e-06,
+      "loss": 0.032,
+      "step": 3910
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2254233331599424,
+      "learning_rate": 2.049977361805471e-06,
+      "loss": 0.0238,
+      "step": 3911
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.28845767590186433,
+      "learning_rate": 2.043707083370814e-06,
+      "loss": 0.0301,
+      "step": 3912
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.21304851766358,
+      "learning_rate": 2.03744589272608e-06,
+      "loss": 0.0307,
+      "step": 3913
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3292309877269268,
+      "learning_rate": 2.031193793040087e-06,
+      "loss": 0.0344,
+      "step": 3914
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.20086938838265608,
+      "learning_rate": 2.0249507874770714e-06,
+      "loss": 0.0248,
+      "step": 3915
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2349814278411253,
+      "learning_rate": 2.018716879196645e-06,
+      "loss": 0.0286,
+      "step": 3916
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3030606388311023,
+      "learning_rate": 2.0124920713538378e-06,
+      "loss": 0.0334,
+      "step": 3917
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2677680128398089,
+      "learning_rate": 2.006276367099054e-06,
+      "loss": 0.0301,
+      "step": 3918
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.18866716146011084,
+      "learning_rate": 2.000069769578108e-06,
+      "loss": 0.02,
+      "step": 3919
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2781445132064626,
+      "learning_rate": 1.9938722819321854e-06,
+      "loss": 0.0259,
+      "step": 3920
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.21489925534501872,
+      "learning_rate": 1.987683907297888e-06,
+      "loss": 0.0226,
+      "step": 3921
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.22646346887563482,
+      "learning_rate": 1.9815046488071774e-06,
+      "loss": 0.0239,
+      "step": 3922
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2401493666437388,
+      "learning_rate": 1.9753345095874234e-06,
+      "loss": 0.0238,
+      "step": 3923
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.29931022052689477,
+      "learning_rate": 1.9691734927613625e-06,
+      "loss": 0.0297,
+      "step": 3924
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.22429214133725534,
+      "learning_rate": 1.9630216014471326e-06,
+      "loss": 0.023,
+      "step": 3925
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2424895187698589,
+      "learning_rate": 1.9568788387582338e-06,
+      "loss": 0.0251,
+      "step": 3926
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2084886071605321,
+      "learning_rate": 1.950745207803566e-06,
+      "loss": 0.0137,
+      "step": 3927
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.26072221315148936,
+      "learning_rate": 1.9446207116873815e-06,
+      "loss": 0.0333,
+      "step": 3928
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.178197163888115,
+      "learning_rate": 1.9385053535093455e-06,
+      "loss": 0.0212,
+      "step": 3929
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2847289973644325,
+      "learning_rate": 1.9323991363644645e-06,
+      "loss": 0.0288,
+      "step": 3930
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.27226567535181007,
+      "learning_rate": 1.9263020633431416e-06,
+      "loss": 0.0214,
+      "step": 3931
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.31075256061803264,
+      "learning_rate": 1.9202141375311335e-06,
+      "loss": 0.0388,
+      "step": 3932
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2823310579403776,
+      "learning_rate": 1.9141353620095835e-06,
+      "loss": 0.0372,
+      "step": 3933
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.18758217679707598,
+      "learning_rate": 1.9080657398549916e-06,
+      "loss": 0.0209,
+      "step": 3934
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.19805705065045512,
+      "learning_rate": 1.902005274139238e-06,
+      "loss": 0.02,
+      "step": 3935
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.23035887345797168,
+      "learning_rate": 1.8959539679295536e-06,
+      "loss": 0.0225,
+      "step": 3936
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.22014119339755225,
+      "learning_rate": 1.8899118242885462e-06,
+      "loss": 0.0318,
+      "step": 3937
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.230950296961087,
+      "learning_rate": 1.8838788462741852e-06,
+      "loss": 0.024,
+      "step": 3938
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.24027613145127033,
+      "learning_rate": 1.8778550369397886e-06,
+      "loss": 0.0231,
+      "step": 3939
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.27697689347173,
+      "learning_rate": 1.8718403993340528e-06,
+      "loss": 0.024,
+      "step": 3940
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2174718847409269,
+      "learning_rate": 1.865834936501012e-06,
+      "loss": 0.0245,
+      "step": 3941
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.19060331401042857,
+      "learning_rate": 1.8598386514800793e-06,
+      "loss": 0.0176,
+      "step": 3942
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3065493380980718,
+      "learning_rate": 1.8538515473060026e-06,
+      "loss": 0.0351,
+      "step": 3943
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2501655715787742,
+      "learning_rate": 1.847873627008896e-06,
+      "loss": 0.0307,
+      "step": 3944
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.22951989513899249,
+      "learning_rate": 1.8419048936142191e-06,
+      "loss": 0.0315,
+      "step": 3945
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.21929841608063866,
+      "learning_rate": 1.8359453501427916e-06,
+      "loss": 0.0228,
+      "step": 3946
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2332128800695291,
+      "learning_rate": 1.8299949996107646e-06,
+      "loss": 0.0281,
+      "step": 3947
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.19833979529150592,
+      "learning_rate": 1.8240538450296563e-06,
+      "loss": 0.026,
+      "step": 3948
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2472978983664071,
+      "learning_rate": 1.8181218894063146e-06,
+      "loss": 0.0397,
+      "step": 3949
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2029199757606438,
+      "learning_rate": 1.8121991357429425e-06,
+      "loss": 0.0159,
+      "step": 3950
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.24460092258866165,
+      "learning_rate": 1.8062855870370798e-06,
+      "loss": 0.0306,
+      "step": 3951
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.19939599847379869,
+      "learning_rate": 1.8003812462816127e-06,
+      "loss": 0.0234,
+      "step": 3952
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.27247723493064596,
+      "learning_rate": 1.7944861164647576e-06,
+      "loss": 0.0372,
+      "step": 3953
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.22841116971338493,
+      "learning_rate": 1.788600200570083e-06,
+      "loss": 0.0255,
+      "step": 3954
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2392800145649786,
+      "learning_rate": 1.782723501576482e-06,
+      "loss": 0.0317,
+      "step": 3955
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.25418792652631794,
+      "learning_rate": 1.7768560224581955e-06,
+      "loss": 0.0335,
+      "step": 3956
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.21703919623420148,
+      "learning_rate": 1.770997766184781e-06,
+      "loss": 0.0245,
+      "step": 3957
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.16896704715874653,
+      "learning_rate": 1.7651487357211472e-06,
+      "loss": 0.0108,
+      "step": 3958
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2984374241054898,
+      "learning_rate": 1.7593089340275149e-06,
+      "loss": 0.0298,
+      "step": 3959
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.22443900073523565,
+      "learning_rate": 1.7534783640594533e-06,
+      "loss": 0.0202,
+      "step": 3960
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.20475501992519377,
+      "learning_rate": 1.7476570287678396e-06,
+      "loss": 0.0191,
+      "step": 3961
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.20974471025253075,
+      "learning_rate": 1.741844931098895e-06,
+      "loss": 0.0205,
+      "step": 3962
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2384955974233491,
+      "learning_rate": 1.7360420739941486e-06,
+      "loss": 0.0263,
+      "step": 3963
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.22069584994858502,
+      "learning_rate": 1.7302484603904756e-06,
+      "loss": 0.0314,
+      "step": 3964
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.19494000199605882,
+      "learning_rate": 1.7244640932200484e-06,
+      "loss": 0.0234,
+      "step": 3965
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2347603887776966,
+      "learning_rate": 1.7186889754103763e-06,
+      "loss": 0.0297,
+      "step": 3966
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.23574376172896755,
+      "learning_rate": 1.7129231098842791e-06,
+      "loss": 0.0297,
+      "step": 3967
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.26678069527800613,
+      "learning_rate": 1.707166499559898e-06,
+      "loss": 0.0412,
+      "step": 3968
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.25157095142644803,
+      "learning_rate": 1.701419147350687e-06,
+      "loss": 0.026,
+      "step": 3969
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.24909355821890453,
+      "learning_rate": 1.6956810561654213e-06,
+      "loss": 0.0236,
+      "step": 3970
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2158988964917277,
+      "learning_rate": 1.6899522289081737e-06,
+      "loss": 0.0178,
+      "step": 3971
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2413588622411202,
+      "learning_rate": 1.6842326684783473e-06,
+      "loss": 0.0203,
+      "step": 3972
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2026649179715184,
+      "learning_rate": 1.6785223777706482e-06,
+      "loss": 0.0254,
+      "step": 3973
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.33542930621663797,
+      "learning_rate": 1.6728213596750831e-06,
+      "loss": 0.0343,
+      "step": 3974
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.26128103176874784,
+      "learning_rate": 1.667129617076977e-06,
+      "loss": 0.0356,
+      "step": 3975
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.25916595365103207,
+      "learning_rate": 1.661447152856952e-06,
+      "loss": 0.029,
+      "step": 3976
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.21586879341047518,
+      "learning_rate": 1.6557739698909436e-06,
+      "loss": 0.0251,
+      "step": 3977
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.24176543898825847,
+      "learning_rate": 1.650110071050175e-06,
+      "loss": 0.025,
+      "step": 3978
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.25397719345805286,
+      "learning_rate": 1.6444554592011909e-06,
+      "loss": 0.0189,
+      "step": 3979
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.19706409849080536,
+      "learning_rate": 1.638810137205813e-06,
+      "loss": 0.0212,
+      "step": 3980
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.24386950888774606,
+      "learning_rate": 1.6331741079211872e-06,
+      "loss": 0.0263,
+      "step": 3981
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.22030736666982426,
+      "learning_rate": 1.627547374199734e-06,
+      "loss": 0.0244,
+      "step": 3982
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2842806738262328,
+      "learning_rate": 1.6219299388891797e-06,
+      "loss": 0.0219,
+      "step": 3983
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.20590756121999668,
+      "learning_rate": 1.6163218048325413e-06,
+      "loss": 0.0238,
+      "step": 3984
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.38306856349578233,
+      "learning_rate": 1.610722974868133e-06,
+      "loss": 0.0324,
+      "step": 3985
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.26639646808642814,
+      "learning_rate": 1.6051334518295546e-06,
+      "loss": 0.0246,
+      "step": 3986
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2887242116187527,
+      "learning_rate": 1.5995532385456992e-06,
+      "loss": 0.0276,
+      "step": 3987
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.33405953392589066,
+      "learning_rate": 1.5939823378407426e-06,
+      "loss": 0.0417,
+      "step": 3988
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.21000298367378883,
+      "learning_rate": 1.5884207525341566e-06,
+      "loss": 0.0158,
+      "step": 3989
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.26179161765461234,
+      "learning_rate": 1.5828684854406918e-06,
+      "loss": 0.0276,
+      "step": 3990
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.28524516300225977,
+      "learning_rate": 1.577325539370389e-06,
+      "loss": 0.0297,
+      "step": 3991
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.25805574559513955,
+      "learning_rate": 1.57179191712856e-06,
+      "loss": 0.0322,
+      "step": 3992
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2801519256785691,
+      "learning_rate": 1.5662676215158112e-06,
+      "loss": 0.034,
+      "step": 3993
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2271380103121437,
+      "learning_rate": 1.5607526553280172e-06,
+      "loss": 0.0216,
+      "step": 3994
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.23207426122064762,
+      "learning_rate": 1.5552470213563408e-06,
+      "loss": 0.0264,
+      "step": 3995
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2378034099816618,
+      "learning_rate": 1.549750722387211e-06,
+      "loss": 0.0285,
+      "step": 3996
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.35806127587094355,
+      "learning_rate": 1.5442637612023425e-06,
+      "loss": 0.0507,
+      "step": 3997
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2183148961496767,
+      "learning_rate": 1.5387861405787252e-06,
+      "loss": 0.0228,
+      "step": 3998
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2775771945043418,
+      "learning_rate": 1.5333178632886058e-06,
+      "loss": 0.0353,
+      "step": 3999
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1723180176519468,
+      "learning_rate": 1.5278589320995218e-06,
+      "loss": 0.0178,
+      "step": 4000
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2291028880465536,
+      "learning_rate": 1.5224093497742654e-06,
+      "loss": 0.0254,
+      "step": 4001
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.21540815603745844,
+      "learning_rate": 1.5169691190709057e-06,
+      "loss": 0.0336,
+      "step": 4002
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2567826857678153,
+      "learning_rate": 1.5115382427427827e-06,
+      "loss": 0.0345,
+      "step": 4003
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2195455017079472,
+      "learning_rate": 1.5061167235384867e-06,
+      "loss": 0.0308,
+      "step": 4004
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.25006646524482823,
+      "learning_rate": 1.5007045642018868e-06,
+      "loss": 0.0289,
+      "step": 4005
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.21073223652648765,
+      "learning_rate": 1.4953017674721083e-06,
+      "loss": 0.0253,
+      "step": 4006
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.256005805825172,
+      "learning_rate": 1.4899083360835408e-06,
+      "loss": 0.0326,
+      "step": 4007
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.27258169581314906,
+      "learning_rate": 1.484524272765837e-06,
+      "loss": 0.0268,
+      "step": 4008
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2409342982553185,
+      "learning_rate": 1.479149580243895e-06,
+      "loss": 0.0274,
+      "step": 4009
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.20566295612613625,
+      "learning_rate": 1.4737842612378894e-06,
+      "loss": 0.02,
+      "step": 4010
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.18890379873346602,
+      "learning_rate": 1.468428318463233e-06,
+      "loss": 0.0205,
+      "step": 4011
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.20006802862239356,
+      "learning_rate": 1.4630817546306087e-06,
+      "loss": 0.0236,
+      "step": 4012
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.23546096032008776,
+      "learning_rate": 1.4577445724459382e-06,
+      "loss": 0.0379,
+      "step": 4013
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1940759232465805,
+      "learning_rate": 1.4524167746104034e-06,
+      "loss": 0.0301,
+      "step": 4014
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2289597930390186,
+      "learning_rate": 1.4470983638204384e-06,
+      "loss": 0.0319,
+      "step": 4015
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.22939968273863018,
+      "learning_rate": 1.4417893427677276e-06,
+      "loss": 0.0298,
+      "step": 4016
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.25927656787820885,
+      "learning_rate": 1.4364897141391888e-06,
+      "loss": 0.0262,
+      "step": 4017
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.24248916542782697,
+      "learning_rate": 1.4311994806170048e-06,
+      "loss": 0.0365,
+      "step": 4018
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.24264834765102183,
+      "learning_rate": 1.42591864487859e-06,
+      "loss": 0.0251,
+      "step": 4019
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.23210383541824384,
+      "learning_rate": 1.4206472095966107e-06,
+      "loss": 0.029,
+      "step": 4020
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2245166089428411,
+      "learning_rate": 1.4153851774389703e-06,
+      "loss": 0.0233,
+      "step": 4021
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.22277273510764165,
+      "learning_rate": 1.4101325510688192e-06,
+      "loss": 0.0209,
+      "step": 4022
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.24405514555826774,
+      "learning_rate": 1.4048893331445367e-06,
+      "loss": 0.0162,
+      "step": 4023
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.21945974468325163,
+      "learning_rate": 1.3996555263197587e-06,
+      "loss": 0.0211,
+      "step": 4024
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.21372582279354385,
+      "learning_rate": 1.3944311332433368e-06,
+      "loss": 0.0276,
+      "step": 4025
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.32695586956517514,
+      "learning_rate": 1.3892161565593743e-06,
+      "loss": 0.0343,
+      "step": 4026
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.22012335659477245,
+      "learning_rate": 1.3840105989071995e-06,
+      "loss": 0.0294,
+      "step": 4027
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3275218369942323,
+      "learning_rate": 1.3788144629213785e-06,
+      "loss": 0.0367,
+      "step": 4028
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.29462431745081913,
+      "learning_rate": 1.3736277512317076e-06,
+      "loss": 0.0359,
+      "step": 4029
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2192795508608225,
+      "learning_rate": 1.3684504664632137e-06,
+      "loss": 0.0198,
+      "step": 4030
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.280980113131807,
+      "learning_rate": 1.3632826112361497e-06,
+      "loss": 0.0376,
+      "step": 4031
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2036203982346656,
+      "learning_rate": 1.3581241881660011e-06,
+      "loss": 0.0202,
+      "step": 4032
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2446505142993538,
+      "learning_rate": 1.352975199863482e-06,
+      "loss": 0.0273,
+      "step": 4033
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3287036992419058,
+      "learning_rate": 1.3478356489345168e-06,
+      "loss": 0.0291,
+      "step": 4034
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.19690709533655248,
+      "learning_rate": 1.3427055379802733e-06,
+      "loss": 0.0292,
+      "step": 4035
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.20444550805776926,
+      "learning_rate": 1.3375848695971239e-06,
+      "loss": 0.0188,
+      "step": 4036
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.20293176152451714,
+      "learning_rate": 1.3324736463766775e-06,
+      "loss": 0.0229,
+      "step": 4037
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.22411925703638305,
+      "learning_rate": 1.3273718709057493e-06,
+      "loss": 0.0255,
+      "step": 4038
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.20742693462055836,
+      "learning_rate": 1.322279545766385e-06,
+      "loss": 0.0184,
+      "step": 4039
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.22124226303366012,
+      "learning_rate": 1.3171966735358343e-06,
+      "loss": 0.0238,
+      "step": 4040
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.21374453332669846,
+      "learning_rate": 1.3121232567865793e-06,
+      "loss": 0.0222,
+      "step": 4041
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2995226151281877,
+      "learning_rate": 1.3070592980862994e-06,
+      "loss": 0.0282,
+      "step": 4042
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.20906475339624503,
+      "learning_rate": 1.3020047999979002e-06,
+      "loss": 0.0268,
+      "step": 4043
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.22014655769406827,
+      "learning_rate": 1.2969597650794907e-06,
+      "loss": 0.0238,
+      "step": 4044
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.22428625333450491,
+      "learning_rate": 1.2919241958843975e-06,
+      "loss": 0.0227,
+      "step": 4045
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.24069317372127222,
+      "learning_rate": 1.2868980949611486e-06,
+      "loss": 0.0284,
+      "step": 4046
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2820574772017969,
+      "learning_rate": 1.2818814648534895e-06,
+      "loss": 0.0351,
+      "step": 4047
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.24369227630638465,
+      "learning_rate": 1.2768743081003598e-06,
+      "loss": 0.0232,
+      "step": 4048
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.25864173006102703,
+      "learning_rate": 1.2718766272359195e-06,
+      "loss": 0.0289,
+      "step": 4049
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2382527631397271,
+      "learning_rate": 1.266888424789523e-06,
+      "loss": 0.0228,
+      "step": 4050
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3481749832938333,
+      "learning_rate": 1.261909703285733e-06,
+      "loss": 0.0343,
+      "step": 4051
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.22609741882848491,
+      "learning_rate": 1.2569404652443073e-06,
+      "loss": 0.0244,
+      "step": 4052
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2095581360449896,
+      "learning_rate": 1.2519807131802097e-06,
+      "loss": 0.0287,
+      "step": 4053
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2500651244931791,
+      "learning_rate": 1.2470304496035968e-06,
+      "loss": 0.0362,
+      "step": 4054
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.22651535398660863,
+      "learning_rate": 1.2420896770198355e-06,
+      "loss": 0.0238,
+      "step": 4055
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1854365472547319,
+      "learning_rate": 1.237158397929472e-06,
+      "loss": 0.014,
+      "step": 4056
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2610574880424356,
+      "learning_rate": 1.2322366148282638e-06,
+      "loss": 0.0227,
+      "step": 4057
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.19375172698881735,
+      "learning_rate": 1.2273243302071513e-06,
+      "loss": 0.0163,
+      "step": 4058
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.21801280222549693,
+      "learning_rate": 1.2224215465522726e-06,
+      "loss": 0.0292,
+      "step": 4059
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.21858894004737908,
+      "learning_rate": 1.2175282663449584e-06,
+      "loss": 0.0311,
+      "step": 4060
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.23674585923679517,
+      "learning_rate": 1.2126444920617297e-06,
+      "loss": 0.0244,
+      "step": 4061
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.24901589774890992,
+      "learning_rate": 1.2077702261742875e-06,
+      "loss": 0.0328,
+      "step": 4062
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.21182642939812502,
+      "learning_rate": 1.2029054711495358e-06,
+      "loss": 0.0235,
+      "step": 4063
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.24339609680372667,
+      "learning_rate": 1.19805022944955e-06,
+      "loss": 0.0215,
+      "step": 4064
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.23009335122623728,
+      "learning_rate": 1.193204503531602e-06,
+      "loss": 0.0266,
+      "step": 4065
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.21858024822689778,
+      "learning_rate": 1.1883682958481413e-06,
+      "loss": 0.0203,
+      "step": 4066
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1924488225475941,
+      "learning_rate": 1.1835416088468033e-06,
+      "loss": 0.0201,
+      "step": 4067
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2545215926378505,
+      "learning_rate": 1.178724444970405e-06,
+      "loss": 0.0242,
+      "step": 4068
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.22614522029108927,
+      "learning_rate": 1.1739168066569406e-06,
+      "loss": 0.0203,
+      "step": 4069
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.23238722759341868,
+      "learning_rate": 1.1691186963395861e-06,
+      "loss": 0.0219,
+      "step": 4070
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.18852350905397555,
+      "learning_rate": 1.1643301164466926e-06,
+      "loss": 0.0153,
+      "step": 4071
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2369111490974642,
+      "learning_rate": 1.1595510694017943e-06,
+      "loss": 0.0218,
+      "step": 4072
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.21487258370856713,
+      "learning_rate": 1.154781557623592e-06,
+      "loss": 0.027,
+      "step": 4073
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.17336881153748232,
+      "learning_rate": 1.1500215835259664e-06,
+      "loss": 0.0178,
+      "step": 4074
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.23694717100406254,
+      "learning_rate": 1.1452711495179659e-06,
+      "loss": 0.0344,
+      "step": 4075
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.25646757681452675,
+      "learning_rate": 1.1405302580038224e-06,
+      "loss": 0.0252,
+      "step": 4076
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.22545380458878736,
+      "learning_rate": 1.1357989113829237e-06,
+      "loss": 0.0318,
+      "step": 4077
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.26999264077912627,
+      "learning_rate": 1.1310771120498386e-06,
+      "loss": 0.0314,
+      "step": 4078
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2638899995894788,
+      "learning_rate": 1.1263648623942912e-06,
+      "loss": 0.0306,
+      "step": 4079
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.23883802215613728,
+      "learning_rate": 1.1216621648011873e-06,
+      "loss": 0.0292,
+      "step": 4080
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.24956217813602427,
+      "learning_rate": 1.1169690216505846e-06,
+      "loss": 0.0202,
+      "step": 4081
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.30855110936179464,
+      "learning_rate": 1.1122854353177171e-06,
+      "loss": 0.0309,
+      "step": 4082
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.25117188211918606,
+      "learning_rate": 1.1076114081729682e-06,
+      "loss": 0.0228,
+      "step": 4083
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.23851976026841812,
+      "learning_rate": 1.1029469425819039e-06,
+      "loss": 0.0268,
+      "step": 4084
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.20732132365666606,
+      "learning_rate": 1.0982920409052312e-06,
+      "loss": 0.0222,
+      "step": 4085
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1951781001201858,
+      "learning_rate": 1.0936467054988276e-06,
+      "loss": 0.0291,
+      "step": 4086
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.20133334225995106,
+      "learning_rate": 1.0890109387137216e-06,
+      "loss": 0.0231,
+      "step": 4087
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.23568867642602953,
+      "learning_rate": 1.0843847428961074e-06,
+      "loss": 0.0279,
+      "step": 4088
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.22112487061724806,
+      "learning_rate": 1.0797681203873255e-06,
+      "loss": 0.0309,
+      "step": 4089
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2557224897586447,
+      "learning_rate": 1.0751610735238848e-06,
+      "loss": 0.0389,
+      "step": 4090
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.25351980585021766,
+      "learning_rate": 1.0705636046374334e-06,
+      "loss": 0.0269,
+      "step": 4091
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2546228420351522,
+      "learning_rate": 1.0659757160547813e-06,
+      "loss": 0.0255,
+      "step": 4092
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.27955351555875024,
+      "learning_rate": 1.0613974100978885e-06,
+      "loss": 0.0308,
+      "step": 4093
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.22525865998805114,
+      "learning_rate": 1.0568286890838575e-06,
+      "loss": 0.0188,
+      "step": 4094
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2952819419874532,
+      "learning_rate": 1.0522695553249562e-06,
+      "loss": 0.0275,
+      "step": 4095
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2355088938746616,
+      "learning_rate": 1.047720011128579e-06,
+      "loss": 0.032,
+      "step": 4096
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.23432050707250046,
+      "learning_rate": 1.0431800587972862e-06,
+      "loss": 0.0189,
+      "step": 4097
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.22924152239045292,
+      "learning_rate": 1.038649700628771e-06,
+      "loss": 0.0309,
+      "step": 4098
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3333575741576425,
+      "learning_rate": 1.0341289389158793e-06,
+      "loss": 0.0321,
+      "step": 4099
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.21430262129224475,
+      "learning_rate": 1.029617775946592e-06,
+      "loss": 0.0222,
+      "step": 4100
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.21626876395465008,
+      "learning_rate": 1.0251162140040383e-06,
+      "loss": 0.0233,
+      "step": 4101
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.22993769318753188,
+      "learning_rate": 1.0206242553664868e-06,
+      "loss": 0.0222,
+      "step": 4102
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2886638598948931,
+      "learning_rate": 1.016141902307346e-06,
+      "loss": 0.035,
+      "step": 4103
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.23592357847054904,
+      "learning_rate": 1.011669157095161e-06,
+      "loss": 0.032,
+      "step": 4104
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2396318431478083,
+      "learning_rate": 1.0072060219936164e-06,
+      "loss": 0.0183,
+      "step": 4105
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.26774349810437653,
+      "learning_rate": 1.002752499261528e-06,
+      "loss": 0.0318,
+      "step": 4106
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.23850047079925532,
+      "learning_rate": 9.98308591152859e-07,
+      "loss": 0.0339,
+      "step": 4107
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2522215337607944,
+      "learning_rate": 9.93874299916693e-07,
+      "loss": 0.0243,
+      "step": 4108
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.22719153686620427,
+      "learning_rate": 9.894496277972498e-07,
+      "loss": 0.03,
+      "step": 4109
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1965662389439065,
+      "learning_rate": 9.850345770338875e-07,
+      "loss": 0.0282,
+      "step": 4110
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.24668700750247183,
+      "learning_rate": 9.80629149861092e-07,
+      "loss": 0.032,
+      "step": 4111
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.23108659783490207,
+      "learning_rate": 9.76233348508473e-07,
+      "loss": 0.0265,
+      "step": 4112
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2106170339238177,
+      "learning_rate": 9.718471752007753e-07,
+      "loss": 0.0189,
+      "step": 4113
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.21015933222442607,
+      "learning_rate": 9.67470632157863e-07,
+      "loss": 0.0221,
+      "step": 4114
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.20480083611108585,
+      "learning_rate": 9.63103721594738e-07,
+      "loss": 0.0241,
+      "step": 4115
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2944418429963534,
+      "learning_rate": 9.587464457215146e-07,
+      "loss": 0.0288,
+      "step": 4116
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.23625826788366722,
+      "learning_rate": 9.54398806743444e-07,
+      "loss": 0.0224,
+      "step": 4117
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.263884818661961,
+      "learning_rate": 9.500608068608841e-07,
+      "loss": 0.0247,
+      "step": 4118
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.23296904069458949,
+      "learning_rate": 9.457324482693275e-07,
+      "loss": 0.0228,
+      "step": 4119
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.28249035291319474,
+      "learning_rate": 9.414137331593842e-07,
+      "loss": 0.0379,
+      "step": 4120
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.23230560985053014,
+      "learning_rate": 9.371046637167835e-07,
+      "loss": 0.0167,
+      "step": 4121
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.25321784965640365,
+      "learning_rate": 9.328052421223676e-07,
+      "loss": 0.0247,
+      "step": 4122
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.22432202741914464,
+      "learning_rate": 9.285154705521048e-07,
+      "loss": 0.029,
+      "step": 4123
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.21346787402733183,
+      "learning_rate": 9.242353511770697e-07,
+      "loss": 0.0219,
+      "step": 4124
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.23649280894740557,
+      "learning_rate": 9.199648861634625e-07,
+      "loss": 0.0261,
+      "step": 4125
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1963177180843091,
+      "learning_rate": 9.157040776725856e-07,
+      "loss": 0.0204,
+      "step": 4126
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.23517643281117567,
+      "learning_rate": 9.11452927860863e-07,
+      "loss": 0.0234,
+      "step": 4127
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2606570501311804,
+      "learning_rate": 9.072114388798314e-07,
+      "loss": 0.0229,
+      "step": 4128
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2713708297069939,
+      "learning_rate": 9.029796128761292e-07,
+      "loss": 0.0274,
+      "step": 4129
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.21933885064572553,
+      "learning_rate": 8.987574519915121e-07,
+      "loss": 0.0144,
+      "step": 4130
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.24057086391334367,
+      "learning_rate": 8.945449583628396e-07,
+      "loss": 0.0228,
+      "step": 4131
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.20083751483840492,
+      "learning_rate": 8.903421341220842e-07,
+      "loss": 0.0253,
+      "step": 4132
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.23787963353213165,
+      "learning_rate": 8.861489813963154e-07,
+      "loss": 0.0222,
+      "step": 4133
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2820054610447314,
+      "learning_rate": 8.819655023077201e-07,
+      "loss": 0.0304,
+      "step": 4134
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2659333346855394,
+      "learning_rate": 8.777916989735736e-07,
+      "loss": 0.0293,
+      "step": 4135
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2236776098834561,
+      "learning_rate": 8.736275735062749e-07,
+      "loss": 0.0203,
+      "step": 4136
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3545968818018609,
+      "learning_rate": 8.694731280133051e-07,
+      "loss": 0.04,
+      "step": 4137
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.19651733713495328,
+      "learning_rate": 8.653283645972598e-07,
+      "loss": 0.0195,
+      "step": 4138
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.24283739537981863,
+      "learning_rate": 8.611932853558236e-07,
+      "loss": 0.0356,
+      "step": 4139
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.18475203675836635,
+      "learning_rate": 8.570678923817888e-07,
+      "loss": 0.0156,
+      "step": 4140
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2951715057089647,
+      "learning_rate": 8.529521877630409e-07,
+      "loss": 0.0328,
+      "step": 4141
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.21674076419565533,
+      "learning_rate": 8.48846173582567e-07,
+      "loss": 0.0212,
+      "step": 4142
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.24360912451878924,
+      "learning_rate": 8.447498519184405e-07,
+      "loss": 0.0233,
+      "step": 4143
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1933249240586514,
+      "learning_rate": 8.406632248438362e-07,
+      "loss": 0.0147,
+      "step": 4144
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2590958414376381,
+      "learning_rate": 8.365862944270243e-07,
+      "loss": 0.0361,
+      "step": 4145
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.23042954903530702,
+      "learning_rate": 8.325190627313628e-07,
+      "loss": 0.0283,
+      "step": 4146
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2168782742937472,
+      "learning_rate": 8.284615318152988e-07,
+      "loss": 0.0256,
+      "step": 4147
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.23256130203793526,
+      "learning_rate": 8.244137037323807e-07,
+      "loss": 0.0243,
+      "step": 4148
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2065068928120067,
+      "learning_rate": 8.203755805312319e-07,
+      "loss": 0.027,
+      "step": 4149
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2840058099562093,
+      "learning_rate": 8.163471642555798e-07,
+      "loss": 0.0318,
+      "step": 4150
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.24845108378959502,
+      "learning_rate": 8.123284569442203e-07,
+      "loss": 0.0214,
+      "step": 4151
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.26372225244833675,
+      "learning_rate": 8.083194606310507e-07,
+      "loss": 0.0255,
+      "step": 4152
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.222776832053034,
+      "learning_rate": 8.043201773450526e-07,
+      "loss": 0.0328,
+      "step": 4153
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2358572422346872,
+      "learning_rate": 8.003306091102803e-07,
+      "loss": 0.0363,
+      "step": 4154
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.20142653929065527,
+      "learning_rate": 7.963507579458851e-07,
+      "loss": 0.0197,
+      "step": 4155
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.23690345186906567,
+      "learning_rate": 7.923806258660893e-07,
+      "loss": 0.0185,
+      "step": 4156
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.27213312753447716,
+      "learning_rate": 7.884202148802056e-07,
+      "loss": 0.0234,
+      "step": 4157
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.24306988711717298,
+      "learning_rate": 7.844695269926194e-07,
+      "loss": 0.0212,
+      "step": 4158
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2129738146430894,
+      "learning_rate": 7.805285642027983e-07,
+      "loss": 0.0246,
+      "step": 4159
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.25549705710181264,
+      "learning_rate": 7.765973285052863e-07,
+      "loss": 0.0319,
+      "step": 4160
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.22458338379785636,
+      "learning_rate": 7.726758218897079e-07,
+      "loss": 0.0234,
+      "step": 4161
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.23688323660850585,
+      "learning_rate": 7.687640463407597e-07,
+      "loss": 0.0226,
+      "step": 4162
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.20468908374843428,
+      "learning_rate": 7.648620038382204e-07,
+      "loss": 0.0262,
+      "step": 4163
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.23873796624918533,
+      "learning_rate": 7.609696963569325e-07,
+      "loss": 0.0238,
+      "step": 4164
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.25653442844889807,
+      "learning_rate": 7.5708712586682e-07,
+      "loss": 0.027,
+      "step": 4165
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.22679305677521136,
+      "learning_rate": 7.532142943328713e-07,
+      "loss": 0.0226,
+      "step": 4166
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2986472835490037,
+      "learning_rate": 7.493512037151563e-07,
+      "loss": 0.0373,
+      "step": 4167
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.19956714422279945,
+      "learning_rate": 7.454978559688019e-07,
+      "loss": 0.0188,
+      "step": 4168
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2401934798253411,
+      "learning_rate": 7.416542530440174e-07,
+      "loss": 0.0289,
+      "step": 4169
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.25598255133583186,
+      "learning_rate": 7.378203968860643e-07,
+      "loss": 0.031,
+      "step": 4170
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2686609508781585,
+      "learning_rate": 7.339962894352925e-07,
+      "loss": 0.0337,
+      "step": 4171
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2355377600378674,
+      "learning_rate": 7.30181932627101e-07,
+      "loss": 0.0296,
+      "step": 4172
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.23873494683739044,
+      "learning_rate": 7.263773283919584e-07,
+      "loss": 0.029,
+      "step": 4173
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2928635599121976,
+      "learning_rate": 7.225824786553981e-07,
+      "loss": 0.0269,
+      "step": 4174
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2290730321214082,
+      "learning_rate": 7.187973853380215e-07,
+      "loss": 0.021,
+      "step": 4175
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2189374018796047,
+      "learning_rate": 7.150220503554783e-07,
+      "loss": 0.0225,
+      "step": 4176
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.23741704533576782,
+      "learning_rate": 7.112564756184981e-07,
+      "loss": 0.0263,
+      "step": 4177
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.24110244905868689,
+      "learning_rate": 7.075006630328518e-07,
+      "loss": 0.0175,
+      "step": 4178
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2777066073372374,
+      "learning_rate": 7.037546144993901e-07,
+      "loss": 0.0327,
+      "step": 4179
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.28038260696353856,
+      "learning_rate": 7.000183319140053e-07,
+      "loss": 0.0229,
+      "step": 4180
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.28097797855363493,
+      "learning_rate": 6.962918171676536e-07,
+      "loss": 0.033,
+      "step": 4181
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2777447286787186,
+      "learning_rate": 6.925750721463443e-07,
+      "loss": 0.03,
+      "step": 4182
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3036071747282006,
+      "learning_rate": 6.88868098731148e-07,
+      "loss": 0.0404,
+      "step": 4183
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2548057468877089,
+      "learning_rate": 6.851708987981865e-07,
+      "loss": 0.0246,
+      "step": 4184
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.20866771129708744,
+      "learning_rate": 6.814834742186361e-07,
+      "loss": 0.0165,
+      "step": 4185
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.24086860735556023,
+      "learning_rate": 6.778058268587217e-07,
+      "loss": 0.0374,
+      "step": 4186
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.23213997743465795,
+      "learning_rate": 6.741379585797236e-07,
+      "loss": 0.0236,
+      "step": 4187
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2619117480694197,
+      "learning_rate": 6.704798712379768e-07,
+      "loss": 0.0334,
+      "step": 4188
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2228225789458934,
+      "learning_rate": 6.66831566684858e-07,
+      "loss": 0.026,
+      "step": 4189
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.23261587877650092,
+      "learning_rate": 6.631930467667991e-07,
+      "loss": 0.0254,
+      "step": 4190
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.23538675480397928,
+      "learning_rate": 6.595643133252716e-07,
+      "loss": 0.0234,
+      "step": 4191
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.20758722435386293,
+      "learning_rate": 6.559453681968064e-07,
+      "loss": 0.0261,
+      "step": 4192
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.23417337009794903,
+      "learning_rate": 6.523362132129718e-07,
+      "loss": 0.0214,
+      "step": 4193
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2359065086018697,
+      "learning_rate": 6.487368502003821e-07,
+      "loss": 0.0197,
+      "step": 4194
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2548183929766835,
+      "learning_rate": 6.451472809806958e-07,
+      "loss": 0.0254,
+      "step": 4195
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.24662451047698913,
+      "learning_rate": 6.415675073706174e-07,
+      "loss": 0.0226,
+      "step": 4196
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2464448150470119,
+      "learning_rate": 6.379975311818931e-07,
+      "loss": 0.0378,
+      "step": 4197
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2398910472719313,
+      "learning_rate": 6.344373542213112e-07,
+      "loss": 0.0254,
+      "step": 4198
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.195640993439512,
+      "learning_rate": 6.308869782906946e-07,
+      "loss": 0.0181,
+      "step": 4199
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.23935219596327983,
+      "learning_rate": 6.27346405186915e-07,
+      "loss": 0.0238,
+      "step": 4200
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2357560713882237,
+      "learning_rate": 6.238156367018744e-07,
+      "loss": 0.0276,
+      "step": 4201
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2454419207707864,
+      "learning_rate": 6.202946746225191e-07,
+      "loss": 0.0321,
+      "step": 4202
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2008752487806573,
+      "learning_rate": 6.16783520730826e-07,
+      "loss": 0.0179,
+      "step": 4203
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.23999128215324347,
+      "learning_rate": 6.132821768038133e-07,
+      "loss": 0.0412,
+      "step": 4204
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.18943407021881573,
+      "learning_rate": 6.097906446135349e-07,
+      "loss": 0.0196,
+      "step": 4205
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.24820520949456476,
+      "learning_rate": 6.063089259270749e-07,
+      "loss": 0.0241,
+      "step": 4206
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.22108302713818226,
+      "learning_rate": 6.028370225065527e-07,
+      "loss": 0.0256,
+      "step": 4207
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.20942082686440533,
+      "learning_rate": 5.993749361091206e-07,
+      "loss": 0.0233,
+      "step": 4208
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.21624351021780297,
+      "learning_rate": 5.95922668486959e-07,
+      "loss": 0.0221,
+      "step": 4209
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2030324908655009,
+      "learning_rate": 5.92480221387286e-07,
+      "loss": 0.0266,
+      "step": 4210
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2234644610137812,
+      "learning_rate": 5.890475965523412e-07,
+      "loss": 0.0195,
+      "step": 4211
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.20601259497684363,
+      "learning_rate": 5.856247957193995e-07,
+      "loss": 0.021,
+      "step": 4212
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2268001855233441,
+      "learning_rate": 5.822118206207594e-07,
+      "loss": 0.0288,
+      "step": 4213
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.231620561267959,
+      "learning_rate": 5.788086729837505e-07,
+      "loss": 0.014,
+      "step": 4214
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.27735890091896426,
+      "learning_rate": 5.754153545307262e-07,
+      "loss": 0.033,
+      "step": 4215
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.22598401570866886,
+      "learning_rate": 5.720318669790636e-07,
+      "loss": 0.0209,
+      "step": 4216
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2529445187259624,
+      "learning_rate": 5.68658212041171e-07,
+      "loss": 0.0325,
+      "step": 4217
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.28292747763192433,
+      "learning_rate": 5.652943914244713e-07,
+      "loss": 0.0184,
+      "step": 4218
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.23992488277864113,
+      "learning_rate": 5.61940406831416e-07,
+      "loss": 0.0251,
+      "step": 4219
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2524621095413231,
+      "learning_rate": 5.585962599594807e-07,
+      "loss": 0.0261,
+      "step": 4220
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2279851821378975,
+      "learning_rate": 5.552619525011538e-07,
+      "loss": 0.0177,
+      "step": 4221
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.23148640939756462,
+      "learning_rate": 5.519374861439497e-07,
+      "loss": 0.0286,
+      "step": 4222
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.26922703589088404,
+      "learning_rate": 5.486228625704049e-07,
+      "loss": 0.0315,
+      "step": 4223
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.25237760346266275,
+      "learning_rate": 5.453180834580663e-07,
+      "loss": 0.0243,
+      "step": 4224
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.28562854167820906,
+      "learning_rate": 5.42023150479507e-07,
+      "loss": 0.0342,
+      "step": 4225
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2511582366420523,
+      "learning_rate": 5.387380653023066e-07,
+      "loss": 0.029,
+      "step": 4226
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1821205709968757,
+      "learning_rate": 5.354628295890729e-07,
+      "loss": 0.0203,
+      "step": 4227
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.210666880264267,
+      "learning_rate": 5.321974449974198e-07,
+      "loss": 0.0224,
+      "step": 4228
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.19076632553010706,
+      "learning_rate": 5.289419131799811e-07,
+      "loss": 0.0215,
+      "step": 4229
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.21570515041374458,
+      "learning_rate": 5.256962357843942e-07,
+      "loss": 0.0194,
+      "step": 4230
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.23654925822055214,
+      "learning_rate": 5.224604144533274e-07,
+      "loss": 0.0227,
+      "step": 4231
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2083108333517541,
+      "learning_rate": 5.192344508244418e-07,
+      "loss": 0.0241,
+      "step": 4232
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2713138676090215,
+      "learning_rate": 5.160183465304203e-07,
+      "loss": 0.0346,
+      "step": 4233
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.20912935451025808,
+      "learning_rate": 5.128121031989497e-07,
+      "loss": 0.0157,
+      "step": 4234
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.19525838102139567,
+      "learning_rate": 5.096157224527343e-07,
+      "loss": 0.02,
+      "step": 4235
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3220334690689413,
+      "learning_rate": 5.0642920590948e-07,
+      "loss": 0.0321,
+      "step": 4236
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2030117379374764,
+      "learning_rate": 5.032525551819012e-07,
+      "loss": 0.0231,
+      "step": 4237
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.26590538653566287,
+      "learning_rate": 5.000857718777186e-07,
+      "loss": 0.0315,
+      "step": 4238
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3333967292621206,
+      "learning_rate": 4.969288575996656e-07,
+      "loss": 0.0372,
+      "step": 4239
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.282552722045364,
+      "learning_rate": 4.937818139454709e-07,
+      "loss": 0.0245,
+      "step": 4240
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.28079548508313534,
+      "learning_rate": 4.906446425078782e-07,
+      "loss": 0.0313,
+      "step": 4241
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.23541741617190348,
+      "learning_rate": 4.87517344874624e-07,
+      "loss": 0.0271,
+      "step": 4242
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.20766293227191449,
+      "learning_rate": 4.843999226284579e-07,
+      "loss": 0.0203,
+      "step": 4243
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.24228220287830854,
+      "learning_rate": 4.812923773471201e-07,
+      "loss": 0.0349,
+      "step": 4244
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.20977995413316466,
+      "learning_rate": 4.781947106033635e-07,
+      "loss": 0.0138,
+      "step": 4245
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.22777955603260774,
+      "learning_rate": 4.7510692396493197e-07,
+      "loss": 0.0256,
+      "step": 4246
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2576554426894255,
+      "learning_rate": 4.720290189945775e-07,
+      "loss": 0.0282,
+      "step": 4247
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2607966734003366,
+      "learning_rate": 4.689609972500453e-07,
+      "loss": 0.0224,
+      "step": 4248
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.26018610602543085,
+      "learning_rate": 4.659028602840776e-07,
+      "loss": 0.0223,
+      "step": 4249
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.24925940310591385,
+      "learning_rate": 4.628546096444186e-07,
+      "loss": 0.0354,
+      "step": 4250
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.23503375459028097,
+      "learning_rate": 4.5981624687380764e-07,
+      "loss": 0.0189,
+      "step": 4251
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2559029927821459,
+      "learning_rate": 4.567877735099768e-07,
+      "loss": 0.0327,
+      "step": 4252
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.22518056306733947,
+      "learning_rate": 4.5376919108565345e-07,
+      "loss": 0.0331,
+      "step": 4253
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2204743355679495,
+      "learning_rate": 4.507605011285643e-07,
+      "loss": 0.0225,
+      "step": 4254
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.24288990851315317,
+      "learning_rate": 4.477617051614225e-07,
+      "loss": 0.0244,
+      "step": 4255
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.24696868193397514,
+      "learning_rate": 4.4477280470194064e-07,
+      "loss": 0.0357,
+      "step": 4256
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.19652242690213034,
+      "learning_rate": 4.4179380126281533e-07,
+      "loss": 0.0196,
+      "step": 4257
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.26521811456515404,
+      "learning_rate": 4.3882469635174287e-07,
+      "loss": 0.036,
+      "step": 4258
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.22014588081843334,
+      "learning_rate": 4.358654914714033e-07,
+      "loss": 0.0313,
+      "step": 4259
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2394356203137378,
+      "learning_rate": 4.329161881194677e-07,
+      "loss": 0.0297,
+      "step": 4260
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.23013257708396465,
+      "learning_rate": 4.299767877885974e-07,
+      "loss": 0.0272,
+      "step": 4261
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.19900803548497645,
+      "learning_rate": 4.270472919664426e-07,
+      "loss": 0.0279,
+      "step": 4262
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.24045887203637645,
+      "learning_rate": 4.241277021356327e-07,
+      "loss": 0.025,
+      "step": 4263
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2620758637838076,
+      "learning_rate": 4.212180197737992e-07,
+      "loss": 0.0243,
+      "step": 4264
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.23535719742103473,
+      "learning_rate": 4.183182463535418e-07,
+      "loss": 0.0227,
+      "step": 4265
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.20651391007562137,
+      "learning_rate": 4.1542838334245994e-07,
+      "loss": 0.0206,
+      "step": 4266
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.28741586635869343,
+      "learning_rate": 4.1254843220312814e-07,
+      "loss": 0.0243,
+      "step": 4267
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.22432157903277913,
+      "learning_rate": 4.096783943931093e-07,
+      "loss": 0.0302,
+      "step": 4268
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2236200437273739,
+      "learning_rate": 4.0681827136494157e-07,
+      "loss": 0.0234,
+      "step": 4269
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.22921955065597152,
+      "learning_rate": 4.039680645661581e-07,
+      "loss": 0.022,
+      "step": 4270
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2059903699825295,
+      "learning_rate": 4.011277754392606e-07,
+      "loss": 0.0307,
+      "step": 4271
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.24625171405418378,
+      "learning_rate": 3.9829740542174143e-07,
+      "loss": 0.0331,
+      "step": 4272
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.280307993602784,
+      "learning_rate": 3.954769559460614e-07,
+      "loss": 0.0223,
+      "step": 4273
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2639314292337824,
+      "learning_rate": 3.9266642843967415e-07,
+      "loss": 0.0255,
+      "step": 4274
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.15444482880435126,
+      "learning_rate": 3.8986582432500196e-07,
+      "loss": 0.0118,
+      "step": 4275
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2156307803636951,
+      "learning_rate": 3.8707514501944657e-07,
+      "loss": 0.0222,
+      "step": 4276
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.24637660921685095,
+      "learning_rate": 3.842943919353914e-07,
+      "loss": 0.0297,
+      "step": 4277
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.19506373775960947,
+      "learning_rate": 3.815235664801908e-07,
+      "loss": 0.0225,
+      "step": 4278
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.21452376204645704,
+      "learning_rate": 3.787626700561742e-07,
+      "loss": 0.0253,
+      "step": 4279
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2523996551046947,
+      "learning_rate": 3.7601170406065034e-07,
+      "loss": 0.0214,
+      "step": 4280
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.20145856927965056,
+      "learning_rate": 3.732706698859012e-07,
+      "loss": 0.0212,
+      "step": 4281
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2139929832573777,
+      "learning_rate": 3.705395689191771e-07,
+      "loss": 0.0232,
+      "step": 4282
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.23201054149913145,
+      "learning_rate": 3.6781840254271227e-07,
+      "loss": 0.0255,
+      "step": 4283
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2101806378306389,
+      "learning_rate": 3.651071721336963e-07,
+      "loss": 0.0256,
+      "step": 4284
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.258124278577854,
+      "learning_rate": 3.62405879064307e-07,
+      "loss": 0.0235,
+      "step": 4285
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.20574487654342996,
+      "learning_rate": 3.59714524701682e-07,
+      "loss": 0.0196,
+      "step": 4286
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2911754015930831,
+      "learning_rate": 3.5703311040793167e-07,
+      "loss": 0.0259,
+      "step": 4287
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.24479935089522756,
+      "learning_rate": 3.543616375401393e-07,
+      "loss": 0.0251,
+      "step": 4288
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.22234509292754578,
+      "learning_rate": 3.517001074503501e-07,
+      "loss": 0.0296,
+      "step": 4289
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2600904312728723,
+      "learning_rate": 3.490485214855799e-07,
+      "loss": 0.032,
+      "step": 4290
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.22179503947671178,
+      "learning_rate": 3.464068809878196e-07,
+      "loss": 0.023,
+      "step": 4291
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3118749447098188,
+      "learning_rate": 3.4377518729401317e-07,
+      "loss": 0.0336,
+      "step": 4292
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.20872304745239587,
+      "learning_rate": 3.4115344173607957e-07,
+      "loss": 0.0196,
+      "step": 4293
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.23229821706240353,
+      "learning_rate": 3.3854164564089964e-07,
+      "loss": 0.028,
+      "step": 4294
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2519289049439303,
+      "learning_rate": 3.359398003303183e-07,
+      "loss": 0.0213,
+      "step": 4295
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.25250354053828866,
+      "learning_rate": 3.333479071211465e-07,
+      "loss": 0.0296,
+      "step": 4296
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3185874262580183,
+      "learning_rate": 3.307659673251595e-07,
+      "loss": 0.036,
+      "step": 4297
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.267967650630909,
+      "learning_rate": 3.281939822490876e-07,
+      "loss": 0.0359,
+      "step": 4298
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2060807469237559,
+      "learning_rate": 3.256319531946317e-07,
+      "loss": 0.0157,
+      "step": 4299
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.21211101678876754,
+      "learning_rate": 3.230798814584502e-07,
+      "loss": 0.0258,
+      "step": 4300
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.23040547552978455,
+      "learning_rate": 3.2053776833216533e-07,
+      "loss": 0.0255,
+      "step": 4301
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.25107194149449047,
+      "learning_rate": 3.1800561510234805e-07,
+      "loss": 0.0193,
+      "step": 4302
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2353953641280843,
+      "learning_rate": 3.1548342305054435e-07,
+      "loss": 0.0261,
+      "step": 4303
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.21677579863228158,
+      "learning_rate": 3.1297119345324645e-07,
+      "loss": 0.0252,
+      "step": 4304
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2627323790992519,
+      "learning_rate": 3.104689275819128e-07,
+      "loss": 0.0226,
+      "step": 4305
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.18438741916920828,
+      "learning_rate": 3.079766267029527e-07,
+      "loss": 0.0179,
+      "step": 4306
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.18393177001251199,
+      "learning_rate": 3.0549429207773483e-07,
+      "loss": 0.0199,
+      "step": 4307
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.19469604482620562,
+      "learning_rate": 3.030219249625854e-07,
+      "loss": 0.0169,
+      "step": 4308
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.24246372344830378,
+      "learning_rate": 3.005595266087835e-07,
+      "loss": 0.0278,
+      "step": 4309
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.18704342884511982,
+      "learning_rate": 2.9810709826256557e-07,
+      "loss": 0.0179,
+      "step": 4310
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.27612142790059196,
+      "learning_rate": 2.956646411651165e-07,
+      "loss": 0.0339,
+      "step": 4311
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.252322613566112,
+      "learning_rate": 2.932321565525853e-07,
+      "loss": 0.028,
+      "step": 4312
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.23652565007706597,
+      "learning_rate": 2.9080964565606273e-07,
+      "loss": 0.0327,
+      "step": 4313
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.22163762100975448,
+      "learning_rate": 2.883971097015992e-07,
+      "loss": 0.0256,
+      "step": 4314
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.23073831246607718,
+      "learning_rate": 2.859945499101913e-07,
+      "loss": 0.0242,
+      "step": 4315
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.23499828748890664,
+      "learning_rate": 2.8360196749778857e-07,
+      "loss": 0.0294,
+      "step": 4316
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2615324752419736,
+      "learning_rate": 2.8121936367529357e-07,
+      "loss": 0.0289,
+      "step": 4317
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.21855633541573388,
+      "learning_rate": 2.788467396485595e-07,
+      "loss": 0.0237,
+      "step": 4318
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2838499246799449,
+      "learning_rate": 2.7648409661837903e-07,
+      "loss": 0.0325,
+      "step": 4319
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2482569248197878,
+      "learning_rate": 2.7413143578050915e-07,
+      "loss": 0.0243,
+      "step": 4320
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.24498061064465365,
+      "learning_rate": 2.7178875832563734e-07,
+      "loss": 0.0209,
+      "step": 4321
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2504161559849217,
+      "learning_rate": 2.6945606543941073e-07,
+      "loss": 0.0322,
+      "step": 4322
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1695524452953065,
+      "learning_rate": 2.671333583024205e-07,
+      "loss": 0.019,
+      "step": 4323
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.24340795824962608,
+      "learning_rate": 2.6482063809020186e-07,
+      "loss": 0.0296,
+      "step": 4324
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.19713535350830463,
+      "learning_rate": 2.625179059732341e-07,
+      "loss": 0.019,
+      "step": 4325
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.24370940463376908,
+      "learning_rate": 2.6022516311695166e-07,
+      "loss": 0.0241,
+      "step": 4326
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3164895148273076,
+      "learning_rate": 2.579424106817174e-07,
+      "loss": 0.0373,
+      "step": 4327
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.23927869121772133,
+      "learning_rate": 2.556696498228495e-07,
+      "loss": 0.0268,
+      "step": 4328
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.23079495055507038,
+      "learning_rate": 2.5340688169060767e-07,
+      "loss": 0.0241,
+      "step": 4329
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2125157196177094,
+      "learning_rate": 2.511541074301915e-07,
+      "loss": 0.0244,
+      "step": 4330
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4308091868465902,
+      "learning_rate": 2.489113281817424e-07,
+      "loss": 0.0395,
+      "step": 4331
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.21628613495322382,
+      "learning_rate": 2.4667854508034774e-07,
+      "loss": 0.0283,
+      "step": 4332
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2564300695277641,
+      "learning_rate": 2.444557592560304e-07,
+      "loss": 0.0321,
+      "step": 4333
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.22927813997690163,
+      "learning_rate": 2.4224297183375487e-07,
+      "loss": 0.0219,
+      "step": 4334
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2885607540390936,
+      "learning_rate": 2.400401839334299e-07,
+      "loss": 0.0306,
+      "step": 4335
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.18273473210359006,
+      "learning_rate": 2.378473966698991e-07,
+      "loss": 0.0251,
+      "step": 4336
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2038541873515383,
+      "learning_rate": 2.356646111529415e-07,
+      "loss": 0.0163,
+      "step": 4337
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.22756448693732054,
+      "learning_rate": 2.3349182848728447e-07,
+      "loss": 0.0209,
+      "step": 4338
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.25275938272936144,
+      "learning_rate": 2.3132904977258175e-07,
+      "loss": 0.0275,
+      "step": 4339
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.25280394797837497,
+      "learning_rate": 2.291762761034333e-07,
+      "loss": 0.034,
+      "step": 4340
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2656973011126717,
+      "learning_rate": 2.2703350856936534e-07,
+      "loss": 0.0339,
+      "step": 4341
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2092671130621699,
+      "learning_rate": 2.249007482548482e-07,
+      "loss": 0.0234,
+      "step": 4342
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.23992350864382142,
+      "learning_rate": 2.2277799623928953e-07,
+      "loss": 0.0296,
+      "step": 4343
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2881623764785107,
+      "learning_rate": 2.2066525359701885e-07,
+      "loss": 0.04,
+      "step": 4344
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.20662916472588463,
+      "learning_rate": 2.1856252139731637e-07,
+      "loss": 0.0307,
+      "step": 4345
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2559865096959685,
+      "learning_rate": 2.1646980070437973e-07,
+      "loss": 0.0398,
+      "step": 4346
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2727011438698973,
+      "learning_rate": 2.14387092577355e-07,
+      "loss": 0.0281,
+      "step": 4347
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2275752092064004,
+      "learning_rate": 2.1231439807031019e-07,
+      "loss": 0.0227,
+      "step": 4348
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.25661856732154437,
+      "learning_rate": 2.102517182322483e-07,
+      "loss": 0.0212,
+      "step": 4349
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2125043426799214,
+      "learning_rate": 2.0819905410710327e-07,
+      "loss": 0.0185,
+      "step": 4350
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1752174395363728,
+      "learning_rate": 2.0615640673374181e-07,
+      "loss": 0.0182,
+      "step": 4351
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.17804550354493404,
+      "learning_rate": 2.0412377714596365e-07,
+      "loss": 0.0159,
+      "step": 4352
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.24171920077251674,
+      "learning_rate": 2.0210116637249032e-07,
+      "loss": 0.0254,
+      "step": 4353
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.23536818987728242,
+      "learning_rate": 2.0008857543698078e-07,
+      "loss": 0.0203,
+      "step": 4354
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.19003395630424846,
+      "learning_rate": 1.9808600535802024e-07,
+      "loss": 0.0193,
+      "step": 4355
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.21540936662109153,
+      "learning_rate": 1.9609345714911575e-07,
+      "loss": 0.0286,
+      "step": 4356
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.25794112724783735,
+      "learning_rate": 1.941109318187162e-07,
+      "loss": 0.0263,
+      "step": 4357
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.20516243064617495,
+      "learning_rate": 1.9213843037018344e-07,
+      "loss": 0.0215,
+      "step": 4358
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2646495776838115,
+      "learning_rate": 1.9017595380181442e-07,
+      "loss": 0.0318,
+      "step": 4359
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.26166491168237466,
+      "learning_rate": 1.8822350310683246e-07,
+      "loss": 0.0251,
+      "step": 4360
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2060624071455964,
+      "learning_rate": 1.862810792733849e-07,
+      "loss": 0.0249,
+      "step": 4361
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2474567798364588,
+      "learning_rate": 1.843486832845409e-07,
+      "loss": 0.0254,
+      "step": 4362
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2384592704533461,
+      "learning_rate": 1.8242631611830263e-07,
+      "loss": 0.0279,
+      "step": 4363
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.21230885870940985,
+      "learning_rate": 1.8051397874758736e-07,
+      "loss": 0.0316,
+      "step": 4364
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.26554681524874807,
+      "learning_rate": 1.786116721402431e-07,
+      "loss": 0.0214,
+      "step": 4365
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.20386617493254652,
+      "learning_rate": 1.7671939725903752e-07,
+      "loss": 0.0249,
+      "step": 4366
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.23737255742813607,
+      "learning_rate": 1.7483715506166455e-07,
+      "loss": 0.0291,
+      "step": 4367
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.19966449611016848,
+      "learning_rate": 1.729649465007377e-07,
+      "loss": 0.0269,
+      "step": 4368
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.24499828752824235,
+      "learning_rate": 1.7110277252379238e-07,
+      "loss": 0.0312,
+      "step": 4369
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.23564291367368878,
+      "learning_rate": 1.692506340732858e-07,
+      "loss": 0.0234,
+      "step": 4370
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2292955570735456,
+      "learning_rate": 1.6740853208659923e-07,
+      "loss": 0.0176,
+      "step": 4371
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3094994858065757,
+      "learning_rate": 1.655764674960292e-07,
+      "loss": 0.0357,
+      "step": 4372
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.22058824064100482,
+      "learning_rate": 1.6375444122879613e-07,
+      "loss": 0.0266,
+      "step": 4373
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.24158084937870627,
+      "learning_rate": 1.6194245420704025e-07,
+      "loss": 0.0243,
+      "step": 4374
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2617999847509731,
+      "learning_rate": 1.6014050734781461e-07,
+      "loss": 0.0193,
+      "step": 4375
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.21384703310406136,
+      "learning_rate": 1.583486015630986e-07,
+      "loss": 0.0278,
+      "step": 4376
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2059478121015002,
+      "learning_rate": 1.565667377597868e-07,
+      "loss": 0.0222,
+      "step": 4377
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.25593105797118465,
+      "learning_rate": 1.5479491683969117e-07,
+      "loss": 0.0291,
+      "step": 4378
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2607773603511524,
+      "learning_rate": 1.5303313969954103e-07,
+      "loss": 0.0239,
+      "step": 4379
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.18820977971594646,
+      "learning_rate": 1.5128140723098317e-07,
+      "loss": 0.0193,
+      "step": 4380
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.22737894441256667,
+      "learning_rate": 1.4953972032057952e-07,
+      "loss": 0.02,
+      "step": 4381
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.22069038084251644,
+      "learning_rate": 1.4780807984980716e-07,
+      "loss": 0.0256,
+      "step": 4382
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2673301940370567,
+      "learning_rate": 1.4608648669506287e-07,
+      "loss": 0.0311,
+      "step": 4383
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.21062860123586485,
+      "learning_rate": 1.443749417276541e-07,
+      "loss": 0.0309,
+      "step": 4384
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.25730030158581574,
+      "learning_rate": 1.4267344581380127e-07,
+      "loss": 0.03,
+      "step": 4385
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.23933088782605375,
+      "learning_rate": 1.4098199981464887e-07,
+      "loss": 0.0316,
+      "step": 4386
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.27638496638288323,
+      "learning_rate": 1.3930060458624106e-07,
+      "loss": 0.0289,
+      "step": 4387
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.231611256723867,
+      "learning_rate": 1.37629260979546e-07,
+      "loss": 0.0222,
+      "step": 4388
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.21467912609032905,
+      "learning_rate": 1.3596796984044037e-07,
+      "loss": 0.0209,
+      "step": 4389
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.21185405602366125,
+      "learning_rate": 1.3431673200971386e-07,
+      "loss": 0.0229,
+      "step": 4390
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.31429962297338676,
+      "learning_rate": 1.3267554832306463e-07,
+      "loss": 0.0379,
+      "step": 4391
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.25666941476925725,
+      "learning_rate": 1.310444196111127e-07,
+      "loss": 0.0329,
+      "step": 4392
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.24319215788408693,
+      "learning_rate": 1.2942334669937773e-07,
+      "loss": 0.0285,
+      "step": 4393
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.22563324469404789,
+      "learning_rate": 1.2781233040829234e-07,
+      "loss": 0.0186,
+      "step": 4394
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.24097933441907754,
+      "learning_rate": 1.2621137155320872e-07,
+      "loss": 0.0315,
+      "step": 4395
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.22624021660480337,
+      "learning_rate": 1.2462047094437657e-07,
+      "loss": 0.0185,
+      "step": 4396
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2228838276219386,
+      "learning_rate": 1.2303962938696068e-07,
+      "loss": 0.0244,
+      "step": 4397
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.24666175686507735,
+      "learning_rate": 1.2146884768103883e-07,
+      "loss": 0.0245,
+      "step": 4398
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.23184753427529117,
+      "learning_rate": 1.1990812662158846e-07,
+      "loss": 0.0288,
+      "step": 4399
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.31526595551894643,
+      "learning_rate": 1.1835746699850215e-07,
+      "loss": 0.0262,
+      "step": 4400
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2519075219335554,
+      "learning_rate": 1.1681686959657879e-07,
+      "loss": 0.0254,
+      "step": 4401
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2899146535678704,
+      "learning_rate": 1.1528633519552357e-07,
+      "loss": 0.0271,
+      "step": 4402
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1967784891024214,
+      "learning_rate": 1.1376586456994798e-07,
+      "loss": 0.0216,
+      "step": 4403
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.20469461639787942,
+      "learning_rate": 1.1225545848937203e-07,
+      "loss": 0.0271,
+      "step": 4404
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2547110347756392,
+      "learning_rate": 1.1075511771822423e-07,
+      "loss": 0.0289,
+      "step": 4405
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.24842515021044465,
+      "learning_rate": 1.0926484301583273e-07,
+      "loss": 0.0342,
+      "step": 4406
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.27344049724571035,
+      "learning_rate": 1.0778463513643645e-07,
+      "loss": 0.041,
+      "step": 4407
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2866099288819449,
+      "learning_rate": 1.0631449482917833e-07,
+      "loss": 0.0293,
+      "step": 4408
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2351114860967714,
+      "learning_rate": 1.0485442283810321e-07,
+      "loss": 0.0276,
+      "step": 4409
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.21697494136147136,
+      "learning_rate": 1.0340441990216443e-07,
+      "loss": 0.022,
+      "step": 4410
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.21252921688956675,
+      "learning_rate": 1.019644867552172e-07,
+      "loss": 0.024,
+      "step": 4411
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.23640651821365824,
+      "learning_rate": 1.0053462412601855e-07,
+      "loss": 0.0249,
+      "step": 4412
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.20857305245457847,
+      "learning_rate": 9.911483273823408e-08,
+      "loss": 0.021,
+      "step": 4413
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.23242865480423908,
+      "learning_rate": 9.770511331042454e-08,
+      "loss": 0.0347,
+      "step": 4414
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2642728949258274,
+      "learning_rate": 9.630546655606365e-08,
+      "loss": 0.0216,
+      "step": 4415
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.28584540713271855,
+      "learning_rate": 9.491589318351368e-08,
+      "loss": 0.0381,
+      "step": 4416
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.19898143358756218,
+      "learning_rate": 9.353639389605207e-08,
+      "loss": 0.0191,
+      "step": 4417
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.23981266690114933,
+      "learning_rate": 9.216696939184922e-08,
+      "loss": 0.0248,
+      "step": 4418
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2955660826219277,
+      "learning_rate": 9.080762036398184e-08,
+      "loss": 0.0298,
+      "step": 4419
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.27698898459903903,
+      "learning_rate": 8.94583475004196e-08,
+      "loss": 0.0337,
+      "step": 4420
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2443484680475507,
+      "learning_rate": 8.811915148404294e-08,
+      "loss": 0.03,
+      "step": 4421
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.20734399888944313,
+      "learning_rate": 8.679003299262523e-08,
+      "loss": 0.0265,
+      "step": 4422
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2711587228885377,
+      "learning_rate": 8.547099269884396e-08,
+      "loss": 0.029,
+      "step": 4423
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.22715330372892692,
+      "learning_rate": 8.416203127026734e-08,
+      "loss": 0.0272,
+      "step": 4424
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.18137627510868737,
+      "learning_rate": 8.286314936937434e-08,
+      "loss": 0.0207,
+      "step": 4425
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.25726488504265727,
+      "learning_rate": 8.157434765353466e-08,
+      "loss": 0.0222,
+      "step": 4426
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.22919501685576507,
+      "learning_rate": 8.029562677502212e-08,
+      "loss": 0.0216,
+      "step": 4427
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.22722395988706445,
+      "learning_rate": 7.902698738099901e-08,
+      "loss": 0.0251,
+      "step": 4428
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.25666707801063404,
+      "learning_rate": 7.776843011353619e-08,
+      "loss": 0.0306,
+      "step": 4429
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.26100535452765306,
+      "learning_rate": 7.651995560959525e-08,
+      "loss": 0.0265,
+      "step": 4430
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.21609393419766798,
+      "learning_rate": 7.528156450103963e-08,
+      "loss": 0.0225,
+      "step": 4431
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.21134781666393448,
+      "learning_rate": 7.405325741462354e-08,
+      "loss": 0.0201,
+      "step": 4432
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.17877245193474503,
+      "learning_rate": 7.283503497200083e-08,
+      "loss": 0.0137,
+      "step": 4433
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.22853090773120874,
+      "learning_rate": 7.162689778972276e-08,
+      "loss": 0.0273,
+      "step": 4434
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2219781169228068,
+      "learning_rate": 7.042884647923353e-08,
+      "loss": 0.0245,
+      "step": 4435
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.21269905150098456,
+      "learning_rate": 6.924088164687703e-08,
+      "loss": 0.0239,
+      "step": 4436
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2551449125986645,
+      "learning_rate": 6.806300389388565e-08,
+      "loss": 0.0367,
+      "step": 4437
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.24152648610438776,
+      "learning_rate": 6.689521381639363e-08,
+      "loss": 0.0222,
+      "step": 4438
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2695999196245218,
+      "learning_rate": 6.573751200542599e-08,
+      "loss": 0.0194,
+      "step": 4439
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.18773310793839185,
+      "learning_rate": 6.458989904690072e-08,
+      "loss": 0.0177,
+      "step": 4440
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3021901684419358,
+      "learning_rate": 6.345237552163541e-08,
+      "loss": 0.026,
+      "step": 4441
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.16928288465784194,
+      "learning_rate": 6.232494200533623e-08,
+      "loss": 0.0136,
+      "step": 4442
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3627623677343884,
+      "learning_rate": 6.120759906860008e-08,
+      "loss": 0.0468,
+      "step": 4443
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2693787295266967,
+      "learning_rate": 6.010034727692792e-08,
+      "loss": 0.0301,
+      "step": 4444
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2592147780396339,
+      "learning_rate": 5.900318719070264e-08,
+      "loss": 0.0259,
+      "step": 4445
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3031344323804535,
+      "learning_rate": 5.791611936520447e-08,
+      "loss": 0.0305,
+      "step": 4446
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.227992224592611,
+      "learning_rate": 5.683914435060445e-08,
+      "loss": 0.0258,
+      "step": 4447
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.19873245575682397,
+      "learning_rate": 5.577226269196656e-08,
+      "loss": 0.0151,
+      "step": 4448
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2479018741222908,
+      "learning_rate": 5.471547492924778e-08,
+      "loss": 0.0242,
+      "step": 4449
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2322767429661271,
+      "learning_rate": 5.3668781597291386e-08,
+      "loss": 0.0269,
+      "step": 4450
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.21112779581354238,
+      "learning_rate": 5.263218322584029e-08,
+      "loss": 0.0245,
+      "step": 4451
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.23287870886567938,
+      "learning_rate": 5.160568033951929e-08,
+      "loss": 0.0309,
+      "step": 4452
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2579778516701443,
+      "learning_rate": 5.058927345784836e-08,
+      "loss": 0.0317,
+      "step": 4453
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2629862737455073,
+      "learning_rate": 4.9582963095238247e-08,
+      "loss": 0.0267,
+      "step": 4454
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2353377120505107,
+      "learning_rate": 4.8586749760985987e-08,
+      "loss": 0.0295,
+      "step": 4455
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.26850899047468857,
+      "learning_rate": 4.7600633959286044e-08,
+      "loss": 0.0267,
+      "step": 4456
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.25478691547878657,
+      "learning_rate": 4.6624616189214765e-08,
+      "loss": 0.025,
+      "step": 4457
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.25562566901681366,
+      "learning_rate": 4.565869694474367e-08,
+      "loss": 0.0298,
+      "step": 4458
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2124237838973614,
+      "learning_rate": 4.470287671472395e-08,
+      "loss": 0.0176,
+      "step": 4459
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.24257266273430095,
+      "learning_rate": 4.375715598290864e-08,
+      "loss": 0.0304,
+      "step": 4460
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2193001317922933,
+      "learning_rate": 4.2821535227930424e-08,
+      "loss": 0.0226,
+      "step": 4461
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.23471578618040925,
+      "learning_rate": 4.1896014923310525e-08,
+      "loss": 0.0235,
+      "step": 4462
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.17518535913199723,
+      "learning_rate": 4.098059553746536e-08,
+      "loss": 0.0186,
+      "step": 4463
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.22191256420057429,
+      "learning_rate": 4.0075277533688784e-08,
+      "loss": 0.0261,
+      "step": 4464
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.26132643888696494,
+      "learning_rate": 3.918006137017205e-08,
+      "loss": 0.0243,
+      "step": 4465
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.18770673401504426,
+      "learning_rate": 3.829494749998608e-08,
+      "loss": 0.0165,
+      "step": 4466
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.20540545543875058,
+      "learning_rate": 3.7419936371094756e-08,
+      "loss": 0.0281,
+      "step": 4467
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.22627322869373032,
+      "learning_rate": 3.655502842634606e-08,
+      "loss": 0.039,
+      "step": 4468
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.23842989084493904,
+      "learning_rate": 3.570022410347651e-08,
+      "loss": 0.0272,
+      "step": 4469
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2099454474834303,
+      "learning_rate": 3.485552383510671e-08,
+      "loss": 0.0195,
+      "step": 4470
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.21160685337741267,
+      "learning_rate": 3.402092804874357e-08,
+      "loss": 0.0223,
+      "step": 4471
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.22340505571162192,
+      "learning_rate": 3.3196437166780336e-08,
+      "loss": 0.0287,
+      "step": 4472
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.19631909382666662,
+      "learning_rate": 3.2382051606500986e-08,
+      "loss": 0.0196,
+      "step": 4473
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2376322898599393,
+      "learning_rate": 3.1577771780066936e-08,
+      "loss": 0.0277,
+      "step": 4474
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2635491827444607,
+      "learning_rate": 3.078359809453257e-08,
+      "loss": 0.0348,
+      "step": 4475
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.18508521419488638,
+      "learning_rate": 2.999953095182972e-08,
+      "loss": 0.0123,
+      "step": 4476
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2867455805144428,
+      "learning_rate": 2.9225570748785402e-08,
+      "loss": 0.0334,
+      "step": 4477
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.21363303896832175,
+      "learning_rate": 2.8461717877099615e-08,
+      "loss": 0.0207,
+      "step": 4478
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.21129233926357469,
+      "learning_rate": 2.770797272336756e-08,
+      "loss": 0.0248,
+      "step": 4479
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2801433088352883,
+      "learning_rate": 2.696433566905965e-08,
+      "loss": 0.0351,
+      "step": 4480
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2532927363554509,
+      "learning_rate": 2.623080709054149e-08,
+      "loss": 0.0282,
+      "step": 4481
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2864289817800324,
+      "learning_rate": 2.550738735905167e-08,
+      "loss": 0.0324,
+      "step": 4482
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2373376988185147,
+      "learning_rate": 2.479407684071733e-08,
+      "loss": 0.0258,
+      "step": 4483
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2127057068581972,
+      "learning_rate": 2.4090875896551903e-08,
+      "loss": 0.0177,
+      "step": 4484
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.23692122823597622,
+      "learning_rate": 2.3397784882448483e-08,
+      "loss": 0.0299,
+      "step": 4485
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.20988079581968236,
+      "learning_rate": 2.2714804149184256e-08,
+      "loss": 0.0262,
+      "step": 4486
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.23558350368400716,
+      "learning_rate": 2.2041934042420497e-08,
+      "loss": 0.0209,
+      "step": 4487
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2822744071559274,
+      "learning_rate": 2.137917490269814e-08,
+      "loss": 0.031,
+      "step": 4488
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.32218783046509364,
+      "learning_rate": 2.0726527065448865e-08,
+      "loss": 0.0352,
+      "step": 4489
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2496145301608871,
+      "learning_rate": 2.0083990860977343e-08,
+      "loss": 0.033,
+      "step": 4490
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2534664306167664,
+      "learning_rate": 1.9451566614479e-08,
+      "loss": 0.0314,
+      "step": 4491
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.21548093461183634,
+      "learning_rate": 1.8829254646022256e-08,
+      "loss": 0.0247,
+      "step": 4492
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2361792103104299,
+      "learning_rate": 1.8217055270568497e-08,
+      "loss": 0.0211,
+      "step": 4493
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2320042231582333,
+      "learning_rate": 1.7614968797952102e-08,
+      "loss": 0.0239,
+      "step": 4494
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2202331890576311,
+      "learning_rate": 1.702299553289377e-08,
+      "loss": 0.0336,
+      "step": 4495
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.23760916560119846,
+      "learning_rate": 1.6441135774996066e-08,
+      "loss": 0.0317,
+      "step": 4496
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.29437743706311953,
+      "learning_rate": 1.586938981873898e-08,
+      "loss": 0.0276,
+      "step": 4497
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.18441166189966846,
+      "learning_rate": 1.530775795348882e-08,
+      "loss": 0.0143,
+      "step": 4498
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.23832741274571254,
+      "learning_rate": 1.4756240463491555e-08,
+      "loss": 0.0232,
+      "step": 4499
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2247094470461646,
+      "learning_rate": 1.421483762787057e-08,
+      "loss": 0.0274,
+      "step": 4500
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3082288685382838,
+      "learning_rate": 1.368354972063557e-08,
+      "loss": 0.0349,
+      "step": 4501
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.19854662215208568,
+      "learning_rate": 1.3162377010673689e-08,
+      "loss": 0.027,
+      "step": 4502
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.31196752721804216,
+      "learning_rate": 1.2651319761753933e-08,
+      "loss": 0.0464,
+      "step": 4503
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.30917682369252936,
+      "learning_rate": 1.2150378232527183e-08,
+      "loss": 0.0323,
+      "step": 4504
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.22311947137196755,
+      "learning_rate": 1.1659552676519525e-08,
+      "loss": 0.0258,
+      "step": 4505
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2824374931944922,
+      "learning_rate": 1.1178843342143363e-08,
+      "loss": 0.03,
+      "step": 4506
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2217890767832452,
+      "learning_rate": 1.070825047268631e-08,
+      "loss": 0.0223,
+      "step": 4507
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2394927603447689,
+      "learning_rate": 1.024777430632229e-08,
+      "loss": 0.0262,
+      "step": 4508
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.30117619581870964,
+      "learning_rate": 9.797415076095996e-09,
+      "loss": 0.0328,
+      "step": 4509
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.28627625082660674,
+      "learning_rate": 9.357173009942878e-09,
+      "loss": 0.0275,
+      "step": 4510
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.20651706686452403,
+      "learning_rate": 8.927048330666932e-09,
+      "loss": 0.0214,
+      "step": 4511
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2013347207359353,
+      "learning_rate": 8.50704125595847e-09,
+      "loss": 0.0292,
+      "step": 4512
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.24912173257265088,
+      "learning_rate": 8.097151998387453e-09,
+      "loss": 0.0236,
+      "step": 4513
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2045073513613524,
+      "learning_rate": 7.697380765399053e-09,
+      "loss": 0.0181,
+      "step": 4514
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2715851713271294,
+      "learning_rate": 7.3077277593203155e-09,
+      "loss": 0.0358,
+      "step": 4515
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.24233401389028633,
+      "learning_rate": 6.928193177360154e-09,
+      "loss": 0.0288,
+      "step": 4516
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2407937244795581,
+      "learning_rate": 6.558777211598255e-09,
+      "loss": 0.0264,
+      "step": 4517
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.27028214168412523,
+      "learning_rate": 6.199480049000617e-09,
+      "loss": 0.0313,
+      "step": 4518
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.22730838559569844,
+      "learning_rate": 5.850301871410668e-09,
+      "loss": 0.0315,
+      "step": 4519
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.22956373032952396,
+      "learning_rate": 5.51124285554927e-09,
+      "loss": 0.0258,
+      "step": 4520
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.22859658331428653,
+      "learning_rate": 5.182303173016934e-09,
+      "loss": 0.0302,
+      "step": 4521
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.20019196298169717,
+      "learning_rate": 4.8634829902893806e-09,
+      "loss": 0.02,
+      "step": 4522
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.256797976281126,
+      "learning_rate": 4.554782468726426e-09,
+      "loss": 0.0336,
+      "step": 4523
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.24382401345259824,
+      "learning_rate": 4.256201764560874e-09,
+      "loss": 0.0276,
+      "step": 4524
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.26355353631491457,
+      "learning_rate": 3.967741028907401e-09,
+      "loss": 0.0321,
+      "step": 4525
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.236170992355893,
+      "learning_rate": 3.6894004077558942e-09,
+      "loss": 0.0261,
+      "step": 4526
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.21118369570046402,
+      "learning_rate": 3.421180041980332e-09,
+      "loss": 0.026,
+      "step": 4527
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.215920213733391,
+      "learning_rate": 3.1630800673254636e-09,
+      "loss": 0.0233,
+      "step": 4528
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.341207246372847,
+      "learning_rate": 2.9151006144201298e-09,
+      "loss": 0.0292,
+      "step": 4529
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.27762589408390465,
+      "learning_rate": 2.6772418087639417e-09,
+      "loss": 0.0251,
+      "step": 4530
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.19320572422128685,
+      "learning_rate": 2.4495037707428226e-09,
+      "loss": 0.0163,
+      "step": 4531
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.26428603242243764,
+      "learning_rate": 2.2318866156134654e-09,
+      "loss": 0.0248,
+      "step": 4532
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.33253742231504263,
+      "learning_rate": 2.0243904535144353e-09,
+      "loss": 0.0453,
+      "step": 4533
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.22122698196730783,
+      "learning_rate": 1.8270153894617282e-09,
+      "loss": 0.0256,
+      "step": 4534
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.23089916641515254,
+      "learning_rate": 1.6397615233465503e-09,
+      "loss": 0.0264,
+      "step": 4535
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.22087355590784324,
+      "learning_rate": 1.4626289499397596e-09,
+      "loss": 0.0225,
+      "step": 4536
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.24501147023180128,
+      "learning_rate": 1.2956177588896445e-09,
+      "loss": 0.0271,
+      "step": 4537
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.23313726392912676,
+      "learning_rate": 1.138728034719705e-09,
+      "loss": 0.0281,
+      "step": 4538
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.24477027569640544,
+      "learning_rate": 9.919598568353118e-10,
+      "loss": 0.0282,
+      "step": 4539
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.29386165607098685,
+      "learning_rate": 8.553132995170466e-10,
+      "loss": 0.0264,
+      "step": 4540
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.36618149702624436,
+      "learning_rate": 7.287884319184813e-10,
+      "loss": 0.0326,
+      "step": 4541
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.28540606856102807,
+      "learning_rate": 6.123853180795003e-10,
+      "loss": 0.0362,
+      "step": 4542
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.20529983887905334,
+      "learning_rate": 5.061040169107578e-10,
+      "loss": 0.0167,
+      "step": 4543
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2795280753393063,
+      "learning_rate": 4.0994458220033896e-10,
+      "loss": 0.0294,
+      "step": 4544
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2661258659536841,
+      "learning_rate": 3.2390706261598015e-10,
+      "loss": 0.0309,
+      "step": 4545
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2145268096519866,
+      "learning_rate": 2.479915017028489e-10,
+      "loss": 0.0283,
+      "step": 4546
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.18357432039405988,
+      "learning_rate": 1.8219793788132322e-10,
+      "loss": 0.0201,
+      "step": 4547
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2166040353315691,
+      "learning_rate": 1.265264044514325e-10,
+      "loss": 0.0244,
+      "step": 4548
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.23382594647015617,
+      "learning_rate": 8.097692958619619e-11,
+      "loss": 0.0268,
+      "step": 4549
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.24069606796661666,
+      "learning_rate": 4.5549536340505627e-11,
+      "loss": 0.0201,
+      "step": 4550
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.22354764024397278,
+      "learning_rate": 2.024424264224223e-11,
+      "loss": 0.0247,
+      "step": 4551
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.22004441054170246,
+      "learning_rate": 5.061061301159242e-12,
+      "loss": 0.0244,
+      "step": 4552
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.27832210607118085,
+      "learning_rate": 0.0,
+      "loss": 0.0444,
+      "step": 4553
+    },
+    {
+      "epoch": 1.0,
+      "step": 4553,
+      "total_flos": 1.7101442093126403e+22,
+      "train_loss": 0.1402213812993695,
+      "train_runtime": 25492.8985,
+      "train_samples_per_second": 11.43,
+      "train_steps_per_second": 0.179
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 4553,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 2000,
+  "total_flos": 1.7101442093126403e+22,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}