| { | |
| "best_metric": 3884.240234375, | |
| "best_model_checkpoint": "./ckpts/tinyllama_v1.1/ternary-g128/checkpoint-400", | |
| "epoch": 4.0, | |
| "eval_steps": 40, | |
| "global_step": 400, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 0.0, | |
| "loss": 11333.3047, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 2e-05, | |
| "loss": 12740.3945, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 2e-05, | |
| "loss": 12751.8154, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 2e-05, | |
| "loss": 13332.6777, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 2e-05, | |
| "loss": 12041.1602, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 2e-05, | |
| "loss": 11759.3701, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 2e-05, | |
| "loss": 7579.3691, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 2e-05, | |
| "loss": 7834.7402, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 2e-05, | |
| "loss": 10647.7695, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 2e-05, | |
| "loss": 7894.1113, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 2e-05, | |
| "loss": 9336.8496, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 2e-05, | |
| "loss": 9224.625, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 2e-05, | |
| "loss": 8379.665, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 2e-05, | |
| "loss": 9123.9082, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 2e-05, | |
| "loss": 10331.1914, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 2e-05, | |
| "loss": 8695.0479, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 2e-05, | |
| "loss": 7838.9502, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 2e-05, | |
| "loss": 7516.7783, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 2e-05, | |
| "loss": 8220.8701, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 2e-05, | |
| "loss": 6244.4844, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 2e-05, | |
| "loss": 7569.1367, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 2e-05, | |
| "loss": 6951.0103, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 2e-05, | |
| "loss": 7107.9727, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 2e-05, | |
| "loss": 7161.4819, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 2e-05, | |
| "loss": 7868.584, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 2e-05, | |
| "loss": 6625.4795, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 2e-05, | |
| "loss": 6222.7422, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 2e-05, | |
| "loss": 6319.2148, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 2e-05, | |
| "loss": 5998.3438, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 2e-05, | |
| "loss": 6406.7197, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 2e-05, | |
| "loss": 6865.812, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 2e-05, | |
| "loss": 5550.9609, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 2e-05, | |
| "loss": 6530.6548, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 2e-05, | |
| "loss": 5644.4668, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 2e-05, | |
| "loss": 6787.8086, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 2e-05, | |
| "loss": 6230.1162, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 2e-05, | |
| "loss": 5976.9404, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 2e-05, | |
| "loss": 5261.2661, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 2e-05, | |
| "loss": 5871.9727, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 2e-05, | |
| "loss": 5244.6387, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "eval_loss": 5789.15625, | |
| "eval_runtime": 153.2221, | |
| "eval_samples_per_second": 10.416, | |
| "eval_steps_per_second": 0.653, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 2e-05, | |
| "loss": 5647.2305, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 2e-05, | |
| "loss": 5769.7246, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 2e-05, | |
| "loss": 5806.084, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 2e-05, | |
| "loss": 6338.6992, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 2e-05, | |
| "loss": 6663.2988, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 2e-05, | |
| "loss": 5164.0898, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 2e-05, | |
| "loss": 5250.1279, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 2e-05, | |
| "loss": 4920.438, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 2e-05, | |
| "loss": 5083.2578, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 2e-05, | |
| "loss": 5441.7505, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 2e-05, | |
| "loss": 5271.9443, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 2e-05, | |
| "loss": 6324.0801, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 2e-05, | |
| "loss": 5859.5479, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 2e-05, | |
| "loss": 5835.2354, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 2e-05, | |
| "loss": 5948.1357, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 2e-05, | |
| "loss": 4974.9961, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 2e-05, | |
| "loss": 5055.021, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 2e-05, | |
| "loss": 5852.9814, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 2e-05, | |
| "loss": 4996.5635, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 2e-05, | |
| "loss": 4621.3154, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 2e-05, | |
| "loss": 4878.3037, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 2e-05, | |
| "loss": 4970.6401, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 2e-05, | |
| "loss": 4931.6152, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 2e-05, | |
| "loss": 5476.6318, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 2e-05, | |
| "loss": 5228.8818, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 2e-05, | |
| "loss": 4760.2158, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 2e-05, | |
| "loss": 4873.3926, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 2e-05, | |
| "loss": 5457.2378, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 2e-05, | |
| "loss": 5250.7954, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 2e-05, | |
| "loss": 4680.854, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 2e-05, | |
| "loss": 4848.3828, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 2e-05, | |
| "loss": 5283.5181, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 2e-05, | |
| "loss": 4919.6475, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 2e-05, | |
| "loss": 5075.8682, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 2e-05, | |
| "loss": 4698.2148, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 2e-05, | |
| "loss": 4929.0264, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 2e-05, | |
| "loss": 5158.9775, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 2e-05, | |
| "loss": 5029.708, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 2e-05, | |
| "loss": 5271.4033, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 2e-05, | |
| "loss": 4820.6875, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "eval_loss": 4961.72412109375, | |
| "eval_runtime": 153.3149, | |
| "eval_samples_per_second": 10.41, | |
| "eval_steps_per_second": 0.652, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 2e-05, | |
| "loss": 4819.8237, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 2e-05, | |
| "loss": 4838.397, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 2e-05, | |
| "loss": 5228.1221, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 2e-05, | |
| "loss": 5043.3447, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 2e-05, | |
| "loss": 5215.2246, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 2e-05, | |
| "loss": 4427.6909, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 2e-05, | |
| "loss": 5120.7344, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 2e-05, | |
| "loss": 5044.043, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "learning_rate": 2e-05, | |
| "loss": 4083.1196, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 2e-05, | |
| "loss": 4404.4829, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 2e-05, | |
| "loss": 5224.4292, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 2e-05, | |
| "loss": 5386.4004, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 2e-05, | |
| "loss": 5193.6807, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 2e-05, | |
| "loss": 5194.8838, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 2e-05, | |
| "loss": 5112.0229, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 2e-05, | |
| "loss": 4970.9795, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "learning_rate": 2e-05, | |
| "loss": 4510.7227, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 2e-05, | |
| "loss": 5206.0742, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 2e-05, | |
| "loss": 5560.2061, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 2e-05, | |
| "loss": 4309.6255, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "learning_rate": 2e-05, | |
| "loss": 4270.1172, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "learning_rate": 2e-05, | |
| "loss": 4307.147, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "learning_rate": 2e-05, | |
| "loss": 5392.7363, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "learning_rate": 2e-05, | |
| "loss": 4613.896, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "learning_rate": 2e-05, | |
| "loss": 4642.5605, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "learning_rate": 2e-05, | |
| "loss": 4990.2676, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "learning_rate": 2e-05, | |
| "loss": 4562.7651, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "learning_rate": 2e-05, | |
| "loss": 4420.7632, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "learning_rate": 2e-05, | |
| "loss": 4600.2861, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "learning_rate": 2e-05, | |
| "loss": 4772.0371, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "learning_rate": 2e-05, | |
| "loss": 5246.8555, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "learning_rate": 2e-05, | |
| "loss": 4700.6631, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "learning_rate": 2e-05, | |
| "loss": 4570.7441, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "learning_rate": 2e-05, | |
| "loss": 4589.0493, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "learning_rate": 2e-05, | |
| "loss": 4846.958, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "learning_rate": 2e-05, | |
| "loss": 4793.291, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "learning_rate": 2e-05, | |
| "loss": 5359.541, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "learning_rate": 2e-05, | |
| "loss": 4714.4346, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "learning_rate": 2e-05, | |
| "loss": 4767.8203, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "learning_rate": 2e-05, | |
| "loss": 5251.5449, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "eval_loss": 4679.81640625, | |
| "eval_runtime": 153.1502, | |
| "eval_samples_per_second": 10.421, | |
| "eval_steps_per_second": 0.653, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "learning_rate": 2e-05, | |
| "loss": 4597.8896, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "learning_rate": 2e-05, | |
| "loss": 4525.1045, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "learning_rate": 2e-05, | |
| "loss": 4740.7261, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "learning_rate": 2e-05, | |
| "loss": 4635.7217, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "learning_rate": 2e-05, | |
| "loss": 4253.7402, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "learning_rate": 2e-05, | |
| "loss": 4565.4546, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "learning_rate": 2e-05, | |
| "loss": 4249.1875, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "learning_rate": 2e-05, | |
| "loss": 5051.9473, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "learning_rate": 2e-05, | |
| "loss": 4128.4834, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "learning_rate": 2e-05, | |
| "loss": 4775.8789, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "learning_rate": 2e-05, | |
| "loss": 4661.0552, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "learning_rate": 2e-05, | |
| "loss": 4616.229, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "learning_rate": 2e-05, | |
| "loss": 5068.2056, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "learning_rate": 2e-05, | |
| "loss": 5291.3818, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "learning_rate": 2e-05, | |
| "loss": 4666.6348, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "learning_rate": 2e-05, | |
| "loss": 5022.856, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "learning_rate": 2e-05, | |
| "loss": 4555.9136, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "learning_rate": 2e-05, | |
| "loss": 4543.4004, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "learning_rate": 2e-05, | |
| "loss": 5208.0181, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "learning_rate": 2e-05, | |
| "loss": 4202.6709, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "learning_rate": 2e-05, | |
| "loss": 4570.6206, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "learning_rate": 2e-05, | |
| "loss": 5114.9287, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "learning_rate": 2e-05, | |
| "loss": 4141.6299, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "learning_rate": 2e-05, | |
| "loss": 4849.125, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "learning_rate": 2e-05, | |
| "loss": 4281.4214, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "learning_rate": 2e-05, | |
| "loss": 4833.1548, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "learning_rate": 2e-05, | |
| "loss": 4661.8291, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "learning_rate": 2e-05, | |
| "loss": 4485.2344, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "learning_rate": 2e-05, | |
| "loss": 4128.439, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "learning_rate": 2e-05, | |
| "loss": 4300.2852, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "learning_rate": 2e-05, | |
| "loss": 4689.8813, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "learning_rate": 2e-05, | |
| "loss": 4261.6787, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "learning_rate": 2e-05, | |
| "loss": 4410.7871, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "learning_rate": 2e-05, | |
| "loss": 4859.6416, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "learning_rate": 2e-05, | |
| "loss": 4945.5195, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "learning_rate": 2e-05, | |
| "loss": 4227.752, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "learning_rate": 2e-05, | |
| "loss": 4758.1689, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "learning_rate": 2e-05, | |
| "loss": 4772.7354, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "learning_rate": 2e-05, | |
| "loss": 4114.6689, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "learning_rate": 2e-05, | |
| "loss": 4792.8496, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "eval_loss": 4522.2705078125, | |
| "eval_runtime": 153.0119, | |
| "eval_samples_per_second": 10.431, | |
| "eval_steps_per_second": 0.654, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "learning_rate": 2e-05, | |
| "loss": 4813.7637, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "learning_rate": 2e-05, | |
| "loss": 3838.3706, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "learning_rate": 2e-05, | |
| "loss": 4562.4961, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "learning_rate": 2e-05, | |
| "loss": 4465.833, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "learning_rate": 2e-05, | |
| "loss": 4186.3252, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "learning_rate": 2e-05, | |
| "loss": 4580.1899, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "learning_rate": 2e-05, | |
| "loss": 4878.4688, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "learning_rate": 2e-05, | |
| "loss": 4298.249, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.69, | |
| "learning_rate": 2e-05, | |
| "loss": 4254.103, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "learning_rate": 2e-05, | |
| "loss": 3883.5581, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "learning_rate": 2e-05, | |
| "loss": 4455.5347, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "learning_rate": 2e-05, | |
| "loss": 4322.6348, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "learning_rate": 2e-05, | |
| "loss": 4249.686, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "learning_rate": 2e-05, | |
| "loss": 4670.4375, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "learning_rate": 2e-05, | |
| "loss": 3984.6772, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "learning_rate": 2e-05, | |
| "loss": 4424.6416, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "learning_rate": 2e-05, | |
| "loss": 4487.6416, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "learning_rate": 2e-05, | |
| "loss": 4786.0723, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "learning_rate": 2e-05, | |
| "loss": 4320.7349, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "learning_rate": 2e-05, | |
| "loss": 4392.415, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "learning_rate": 2e-05, | |
| "loss": 4128.5469, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "learning_rate": 2e-05, | |
| "loss": 3805.3416, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "learning_rate": 2e-05, | |
| "loss": 5295.7549, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "learning_rate": 2e-05, | |
| "loss": 4073.2388, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "learning_rate": 2e-05, | |
| "loss": 4689.1528, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "learning_rate": 2e-05, | |
| "loss": 4331.627, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "learning_rate": 2e-05, | |
| "loss": 4956.3774, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "learning_rate": 2e-05, | |
| "loss": 3998.3621, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "learning_rate": 2e-05, | |
| "loss": 4095.9175, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "learning_rate": 2e-05, | |
| "loss": 3817.9526, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "learning_rate": 2e-05, | |
| "loss": 4632.4111, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "learning_rate": 2e-05, | |
| "loss": 3853.741, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "learning_rate": 2e-05, | |
| "loss": 3862.7646, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "learning_rate": 2e-05, | |
| "loss": 4408.2993, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "learning_rate": 2e-05, | |
| "loss": 3478.6504, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "learning_rate": 2e-05, | |
| "loss": 4667.4932, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "learning_rate": 2e-05, | |
| "loss": 3785.4988, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "learning_rate": 2e-05, | |
| "loss": 4366.5952, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.99, | |
| "learning_rate": 2e-05, | |
| "loss": 4249.0308, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 2e-05, | |
| "loss": 4039.4888, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 4252.13720703125, | |
| "eval_runtime": 152.5389, | |
| "eval_samples_per_second": 10.463, | |
| "eval_steps_per_second": 0.656, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "learning_rate": 2e-05, | |
| "loss": 4239.4941, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "learning_rate": 2e-05, | |
| "loss": 3973.5347, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "learning_rate": 2e-05, | |
| "loss": 4343.3428, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "learning_rate": 2e-05, | |
| "loss": 3764.2117, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "learning_rate": 2e-05, | |
| "loss": 4449.8027, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 2.06, | |
| "learning_rate": 2e-05, | |
| "loss": 4788.9707, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 2.07, | |
| "learning_rate": 2e-05, | |
| "loss": 4126.4805, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "learning_rate": 2e-05, | |
| "loss": 4510.4546, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.09, | |
| "learning_rate": 2e-05, | |
| "loss": 3748.04, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "learning_rate": 2e-05, | |
| "loss": 4639.4795, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.11, | |
| "learning_rate": 2e-05, | |
| "loss": 4275.8872, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.12, | |
| "learning_rate": 2e-05, | |
| "loss": 4072.8267, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "learning_rate": 2e-05, | |
| "loss": 4090.3594, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.14, | |
| "learning_rate": 2e-05, | |
| "loss": 4549.5156, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.15, | |
| "learning_rate": 2e-05, | |
| "loss": 4239.3809, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "learning_rate": 2e-05, | |
| "loss": 4693.9189, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "learning_rate": 2e-05, | |
| "loss": 3877.6008, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.18, | |
| "learning_rate": 2e-05, | |
| "loss": 4515.2139, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "learning_rate": 2e-05, | |
| "loss": 4458.5381, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "learning_rate": 2e-05, | |
| "loss": 4279.3545, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.21, | |
| "learning_rate": 2e-05, | |
| "loss": 4421.5928, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "learning_rate": 2e-05, | |
| "loss": 4403.8462, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.23, | |
| "learning_rate": 2e-05, | |
| "loss": 3923.6372, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "learning_rate": 2e-05, | |
| "loss": 3892.7231, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.25, | |
| "learning_rate": 2e-05, | |
| "loss": 4327.0947, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.26, | |
| "learning_rate": 2e-05, | |
| "loss": 4771.2183, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.27, | |
| "learning_rate": 2e-05, | |
| "loss": 4328.8911, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.28, | |
| "learning_rate": 2e-05, | |
| "loss": 4228.5918, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.29, | |
| "learning_rate": 2e-05, | |
| "loss": 4549.1064, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "learning_rate": 2e-05, | |
| "loss": 4156.6675, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.31, | |
| "learning_rate": 2e-05, | |
| "loss": 4196.0254, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "learning_rate": 2e-05, | |
| "loss": 4296.939, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.33, | |
| "learning_rate": 2e-05, | |
| "loss": 4569.1528, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.34, | |
| "learning_rate": 2e-05, | |
| "loss": 3441.3088, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "learning_rate": 2e-05, | |
| "loss": 3583.4678, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 2.36, | |
| "learning_rate": 2e-05, | |
| "loss": 3813.4893, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 2.37, | |
| "learning_rate": 2e-05, | |
| "loss": 4058.2871, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 2.38, | |
| "learning_rate": 2e-05, | |
| "loss": 3981.6729, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 2.39, | |
| "learning_rate": 2e-05, | |
| "loss": 3813.9546, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "learning_rate": 2e-05, | |
| "loss": 4029.8252, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "eval_loss": 4160.52294921875, | |
| "eval_runtime": 152.8787, | |
| "eval_samples_per_second": 10.44, | |
| "eval_steps_per_second": 0.654, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.41, | |
| "learning_rate": 2e-05, | |
| "loss": 4221.4492, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 2.42, | |
| "learning_rate": 2e-05, | |
| "loss": 4045.0322, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 2.43, | |
| "learning_rate": 2e-05, | |
| "loss": 4226.5371, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 2.44, | |
| "learning_rate": 2e-05, | |
| "loss": 3511.958, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "learning_rate": 2e-05, | |
| "loss": 4395.8027, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 2.46, | |
| "learning_rate": 2e-05, | |
| "loss": 4735.3145, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 2.47, | |
| "learning_rate": 2e-05, | |
| "loss": 3843.616, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "learning_rate": 2e-05, | |
| "loss": 4264.6924, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 2.49, | |
| "learning_rate": 2e-05, | |
| "loss": 4039.4785, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "learning_rate": 2e-05, | |
| "loss": 4019.9854, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.51, | |
| "learning_rate": 2e-05, | |
| "loss": 4662.4287, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 2.52, | |
| "learning_rate": 2e-05, | |
| "loss": 4328.1245, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "learning_rate": 2e-05, | |
| "loss": 3456.5835, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 2.54, | |
| "learning_rate": 2e-05, | |
| "loss": 4076.3733, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 2.55, | |
| "learning_rate": 2e-05, | |
| "loss": 3871.9949, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "learning_rate": 2e-05, | |
| "loss": 3940.6655, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 2.57, | |
| "learning_rate": 2e-05, | |
| "loss": 3684.9651, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 2.58, | |
| "learning_rate": 2e-05, | |
| "loss": 3649.6597, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 2.59, | |
| "learning_rate": 2e-05, | |
| "loss": 4184.002, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "learning_rate": 2e-05, | |
| "loss": 3991.2231, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.61, | |
| "learning_rate": 2e-05, | |
| "loss": 3957.1196, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 2.62, | |
| "learning_rate": 2e-05, | |
| "loss": 4392.5942, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 2.63, | |
| "learning_rate": 2e-05, | |
| "loss": 3652.1855, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "learning_rate": 2e-05, | |
| "loss": 4256.6865, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 2.65, | |
| "learning_rate": 2e-05, | |
| "loss": 3684.3928, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 2.66, | |
| "learning_rate": 2e-05, | |
| "loss": 3780.9175, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "learning_rate": 2e-05, | |
| "loss": 4062.6343, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 2.68, | |
| "learning_rate": 2e-05, | |
| "loss": 4281.2744, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 2.69, | |
| "learning_rate": 2e-05, | |
| "loss": 4767.3281, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "learning_rate": 2e-05, | |
| "loss": 4602.7568, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.71, | |
| "learning_rate": 2e-05, | |
| "loss": 4051.676, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 2.72, | |
| "learning_rate": 2e-05, | |
| "loss": 4140.875, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 2.73, | |
| "learning_rate": 2e-05, | |
| "loss": 3707.1763, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 2.74, | |
| "learning_rate": 2e-05, | |
| "loss": 3892.2764, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 2.75, | |
| "learning_rate": 2e-05, | |
| "loss": 3792.2725, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 2.76, | |
| "learning_rate": 2e-05, | |
| "loss": 4242.418, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 2.77, | |
| "learning_rate": 2e-05, | |
| "loss": 4035.6191, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 2.78, | |
| "learning_rate": 2e-05, | |
| "loss": 4154.3467, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 2.79, | |
| "learning_rate": 2e-05, | |
| "loss": 3960.1741, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "learning_rate": 2e-05, | |
| "loss": 3626.5103, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "eval_loss": 4051.461669921875, | |
| "eval_runtime": 152.381, | |
| "eval_samples_per_second": 10.474, | |
| "eval_steps_per_second": 0.656, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.81, | |
| "learning_rate": 2e-05, | |
| "loss": 4040.4492, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 2.82, | |
| "learning_rate": 2e-05, | |
| "loss": 3889.311, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 2.83, | |
| "learning_rate": 2e-05, | |
| "loss": 3271.3108, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 2.84, | |
| "learning_rate": 2e-05, | |
| "loss": 4014.6853, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 2.85, | |
| "learning_rate": 2e-05, | |
| "loss": 4074.5417, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 2.86, | |
| "learning_rate": 2e-05, | |
| "loss": 4337.8213, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 2.87, | |
| "learning_rate": 2e-05, | |
| "loss": 3792.6384, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "learning_rate": 2e-05, | |
| "loss": 3692.104, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 2.89, | |
| "learning_rate": 2e-05, | |
| "loss": 4260.9463, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "learning_rate": 2e-05, | |
| "loss": 3741.4792, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.91, | |
| "learning_rate": 2e-05, | |
| "loss": 3809.4341, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "learning_rate": 2e-05, | |
| "loss": 4053.488, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "learning_rate": 2e-05, | |
| "loss": 3743.4941, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 2.94, | |
| "learning_rate": 2e-05, | |
| "loss": 4357.4707, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 2.95, | |
| "learning_rate": 2e-05, | |
| "loss": 4021.8262, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "learning_rate": 2e-05, | |
| "loss": 3761.1348, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 2.97, | |
| "learning_rate": 2e-05, | |
| "loss": 4424.8853, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 2.98, | |
| "learning_rate": 2e-05, | |
| "loss": 3830.8997, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 2.99, | |
| "learning_rate": 2e-05, | |
| "loss": 4062.6968, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 2e-05, | |
| "loss": 4436.0918, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.01, | |
| "learning_rate": 2e-05, | |
| "loss": 4233.5381, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 3.02, | |
| "learning_rate": 2e-05, | |
| "loss": 4343.8237, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 3.03, | |
| "learning_rate": 2e-05, | |
| "loss": 4052.5039, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 3.04, | |
| "learning_rate": 2e-05, | |
| "loss": 4048.3113, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 3.05, | |
| "learning_rate": 2e-05, | |
| "loss": 3583.0908, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 3.06, | |
| "learning_rate": 2e-05, | |
| "loss": 3624.552, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 3.07, | |
| "learning_rate": 2e-05, | |
| "loss": 4182.4458, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 3.08, | |
| "learning_rate": 2e-05, | |
| "loss": 3741.467, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 3.09, | |
| "learning_rate": 2e-05, | |
| "loss": 4421.3584, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 3.1, | |
| "learning_rate": 2e-05, | |
| "loss": 3731.9844, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 3.11, | |
| "learning_rate": 2e-05, | |
| "loss": 3778.5786, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 3.12, | |
| "learning_rate": 2e-05, | |
| "loss": 3995.1738, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 3.13, | |
| "learning_rate": 2e-05, | |
| "loss": 3414.417, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 3.14, | |
| "learning_rate": 2e-05, | |
| "loss": 3799.9009, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 3.15, | |
| "learning_rate": 2e-05, | |
| "loss": 3363.4624, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 3.16, | |
| "learning_rate": 2e-05, | |
| "loss": 3599.3823, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 3.17, | |
| "learning_rate": 2e-05, | |
| "loss": 4243.3867, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 3.18, | |
| "learning_rate": 2e-05, | |
| "loss": 4223.4482, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 3.19, | |
| "learning_rate": 2e-05, | |
| "loss": 3808.5315, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "learning_rate": 2e-05, | |
| "loss": 4241.2925, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "eval_loss": 3989.93701171875, | |
| "eval_runtime": 152.2822, | |
| "eval_samples_per_second": 10.481, | |
| "eval_steps_per_second": 0.657, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 3.21, | |
| "learning_rate": 2e-05, | |
| "loss": 4326.0718, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 3.22, | |
| "learning_rate": 2e-05, | |
| "loss": 4092.2671, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 3.23, | |
| "learning_rate": 2e-05, | |
| "loss": 3987.1335, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 3.24, | |
| "learning_rate": 2e-05, | |
| "loss": 3692.0537, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 3.25, | |
| "learning_rate": 2e-05, | |
| "loss": 4115.0986, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 3.26, | |
| "learning_rate": 2e-05, | |
| "loss": 3721.7102, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 3.27, | |
| "learning_rate": 2e-05, | |
| "loss": 4032.0439, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 3.28, | |
| "learning_rate": 2e-05, | |
| "loss": 3985.3152, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 3.29, | |
| "learning_rate": 2e-05, | |
| "loss": 4192.1748, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 3.3, | |
| "learning_rate": 2e-05, | |
| "loss": 4013.9749, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 3.31, | |
| "learning_rate": 2e-05, | |
| "loss": 3703.3384, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 3.32, | |
| "learning_rate": 2e-05, | |
| "loss": 3750.4739, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 3.33, | |
| "learning_rate": 2e-05, | |
| "loss": 3136.6565, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 3.34, | |
| "learning_rate": 2e-05, | |
| "loss": 3656.863, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 3.35, | |
| "learning_rate": 2e-05, | |
| "loss": 3817.365, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 3.36, | |
| "learning_rate": 2e-05, | |
| "loss": 3881.1313, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 3.37, | |
| "learning_rate": 2e-05, | |
| "loss": 4028.1912, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 3.38, | |
| "learning_rate": 2e-05, | |
| "loss": 3855.8794, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 3.39, | |
| "learning_rate": 2e-05, | |
| "loss": 4090.366, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 3.4, | |
| "learning_rate": 2e-05, | |
| "loss": 4035.5522, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 3.41, | |
| "learning_rate": 2e-05, | |
| "loss": 3345.8081, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 3.42, | |
| "learning_rate": 2e-05, | |
| "loss": 4388.6914, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 3.43, | |
| "learning_rate": 2e-05, | |
| "loss": 3672.5535, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 3.44, | |
| "learning_rate": 2e-05, | |
| "loss": 4601.6753, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 3.45, | |
| "learning_rate": 2e-05, | |
| "loss": 3514.8469, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 3.46, | |
| "learning_rate": 2e-05, | |
| "loss": 3831.9482, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 3.47, | |
| "learning_rate": 2e-05, | |
| "loss": 3722.3574, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 3.48, | |
| "learning_rate": 2e-05, | |
| "loss": 3708.3647, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 3.49, | |
| "learning_rate": 2e-05, | |
| "loss": 3739.094, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 3.5, | |
| "learning_rate": 2e-05, | |
| "loss": 3696.3413, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 3.51, | |
| "learning_rate": 2e-05, | |
| "loss": 4573.146, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 3.52, | |
| "learning_rate": 2e-05, | |
| "loss": 4276.0264, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 3.53, | |
| "learning_rate": 2e-05, | |
| "loss": 3745.4736, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 3.54, | |
| "learning_rate": 2e-05, | |
| "loss": 3880.3018, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 3.55, | |
| "learning_rate": 2e-05, | |
| "loss": 3983.3145, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 3.56, | |
| "learning_rate": 2e-05, | |
| "loss": 3436.9373, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 3.57, | |
| "learning_rate": 2e-05, | |
| "loss": 4162.9082, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 3.58, | |
| "learning_rate": 2e-05, | |
| "loss": 3945.7412, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 3.59, | |
| "learning_rate": 2e-05, | |
| "loss": 3507.0898, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "learning_rate": 2e-05, | |
| "loss": 3669.127, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "eval_loss": 4000.498046875, | |
| "eval_runtime": 152.4801, | |
| "eval_samples_per_second": 10.467, | |
| "eval_steps_per_second": 0.656, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 3.61, | |
| "learning_rate": 2e-05, | |
| "loss": 4219.0439, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 3.62, | |
| "learning_rate": 2e-05, | |
| "loss": 3937.8618, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 3.63, | |
| "learning_rate": 2e-05, | |
| "loss": 3954.0127, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 3.64, | |
| "learning_rate": 2e-05, | |
| "loss": 3861.4602, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 3.65, | |
| "learning_rate": 2e-05, | |
| "loss": 3659.9412, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 3.66, | |
| "learning_rate": 2e-05, | |
| "loss": 3923.0183, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 3.67, | |
| "learning_rate": 2e-05, | |
| "loss": 3869.6533, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 3.68, | |
| "learning_rate": 2e-05, | |
| "loss": 4066.0229, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 3.69, | |
| "learning_rate": 2e-05, | |
| "loss": 4978.6533, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 3.7, | |
| "learning_rate": 2e-05, | |
| "loss": 3747.0063, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 3.71, | |
| "learning_rate": 2e-05, | |
| "loss": 3962.3235, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 3.72, | |
| "learning_rate": 2e-05, | |
| "loss": 4163.8613, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 3.73, | |
| "learning_rate": 2e-05, | |
| "loss": 4133.5288, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 3.74, | |
| "learning_rate": 2e-05, | |
| "loss": 4478.0767, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 3.75, | |
| "learning_rate": 2e-05, | |
| "loss": 3869.3635, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 3.76, | |
| "learning_rate": 2e-05, | |
| "loss": 4138.4487, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 3.77, | |
| "learning_rate": 2e-05, | |
| "loss": 4073.0718, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 3.78, | |
| "learning_rate": 2e-05, | |
| "loss": 4123.4678, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 3.79, | |
| "learning_rate": 2e-05, | |
| "loss": 4103.5117, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 3.8, | |
| "learning_rate": 2e-05, | |
| "loss": 3554.8037, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 3.81, | |
| "learning_rate": 2e-05, | |
| "loss": 4128.582, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 3.82, | |
| "learning_rate": 2e-05, | |
| "loss": 3922.896, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 3.83, | |
| "learning_rate": 2e-05, | |
| "loss": 3866.6106, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "learning_rate": 2e-05, | |
| "loss": 3856.5063, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 3.85, | |
| "learning_rate": 2e-05, | |
| "loss": 3534.6477, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 3.86, | |
| "learning_rate": 2e-05, | |
| "loss": 3878.1855, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 3.87, | |
| "learning_rate": 2e-05, | |
| "loss": 3660.9155, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 3.88, | |
| "learning_rate": 2e-05, | |
| "loss": 3628.4751, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 3.89, | |
| "learning_rate": 2e-05, | |
| "loss": 3467.9219, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 3.9, | |
| "learning_rate": 2e-05, | |
| "loss": 3975.7178, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 3.91, | |
| "learning_rate": 2e-05, | |
| "loss": 3414.8813, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 3.92, | |
| "learning_rate": 2e-05, | |
| "loss": 4234.2061, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 3.93, | |
| "learning_rate": 2e-05, | |
| "loss": 4341.897, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 3.94, | |
| "learning_rate": 2e-05, | |
| "loss": 4035.8455, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 3.95, | |
| "learning_rate": 2e-05, | |
| "loss": 3405.0151, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 3.96, | |
| "learning_rate": 2e-05, | |
| "loss": 3455.5244, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 3.97, | |
| "learning_rate": 2e-05, | |
| "loss": 3791.8367, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 3.98, | |
| "learning_rate": 2e-05, | |
| "loss": 3939.1094, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 3.99, | |
| "learning_rate": 2e-05, | |
| "loss": 4370.543, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "learning_rate": 2e-05, | |
| "loss": 4121.397, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 3884.240234375, | |
| "eval_runtime": 152.4512, | |
| "eval_samples_per_second": 10.469, | |
| "eval_steps_per_second": 0.656, | |
| "step": 400 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 400, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 40, | |
| "total_flos": 1.624099600371548e+17, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |