| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.5556432517758485, | |
| "eval_steps": 88, | |
| "global_step": 176, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0031570639305445935, | |
| "grad_norm": 3.7962698936462402, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 1.4078, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.006314127861089187, | |
| "grad_norm": 4.4436869621276855, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 1.4492, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.009471191791633781, | |
| "grad_norm": 3.5391576290130615, | |
| "learning_rate": 6e-06, | |
| "loss": 1.4118, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.012628255722178374, | |
| "grad_norm": 4.092630386352539, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 1.4062, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.01578531965272297, | |
| "grad_norm": 2.91861629486084, | |
| "learning_rate": 1e-05, | |
| "loss": 1.4268, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.018942383583267563, | |
| "grad_norm": 2.2911760807037354, | |
| "learning_rate": 9.989395546129376e-06, | |
| "loss": 1.3445, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.022099447513812154, | |
| "grad_norm": 2.0452730655670166, | |
| "learning_rate": 9.97879109225875e-06, | |
| "loss": 1.362, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.025256511444356748, | |
| "grad_norm": 1.7814005613327026, | |
| "learning_rate": 9.968186638388125e-06, | |
| "loss": 1.3063, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.028413575374901343, | |
| "grad_norm": 1.812359094619751, | |
| "learning_rate": 9.957582184517498e-06, | |
| "loss": 1.288, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.03157063930544594, | |
| "grad_norm": 2.5271124839782715, | |
| "learning_rate": 9.946977730646873e-06, | |
| "loss": 1.2618, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03472770323599053, | |
| "grad_norm": 2.614506483078003, | |
| "learning_rate": 9.936373276776247e-06, | |
| "loss": 1.2681, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.037884767166535126, | |
| "grad_norm": 2.6435418128967285, | |
| "learning_rate": 9.925768822905622e-06, | |
| "loss": 1.2784, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.04104183109707971, | |
| "grad_norm": 1.5790609121322632, | |
| "learning_rate": 9.915164369034995e-06, | |
| "loss": 1.1511, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.04419889502762431, | |
| "grad_norm": 2.8828725814819336, | |
| "learning_rate": 9.90455991516437e-06, | |
| "loss": 1.1558, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.0473559589581689, | |
| "grad_norm": 4.099935531616211, | |
| "learning_rate": 9.893955461293744e-06, | |
| "loss": 1.1788, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.050513022888713496, | |
| "grad_norm": 4.893162727355957, | |
| "learning_rate": 9.88335100742312e-06, | |
| "loss": 1.1736, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.05367008681925809, | |
| "grad_norm": 2.095832109451294, | |
| "learning_rate": 9.872746553552493e-06, | |
| "loss": 1.1583, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.056827150749802685, | |
| "grad_norm": 1.7522943019866943, | |
| "learning_rate": 9.862142099681868e-06, | |
| "loss": 1.1384, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.05998421468034728, | |
| "grad_norm": 2.5089237689971924, | |
| "learning_rate": 9.851537645811241e-06, | |
| "loss": 1.164, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.06314127861089187, | |
| "grad_norm": 2.7380177974700928, | |
| "learning_rate": 9.840933191940616e-06, | |
| "loss": 1.0902, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.06629834254143646, | |
| "grad_norm": 1.619511604309082, | |
| "learning_rate": 9.83032873806999e-06, | |
| "loss": 1.1151, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.06945540647198106, | |
| "grad_norm": 2.883584499359131, | |
| "learning_rate": 9.819724284199365e-06, | |
| "loss": 1.0905, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.07261247040252565, | |
| "grad_norm": 3.4861960411071777, | |
| "learning_rate": 9.809119830328739e-06, | |
| "loss": 1.0622, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.07576953433307025, | |
| "grad_norm": 2.4595322608947754, | |
| "learning_rate": 9.798515376458114e-06, | |
| "loss": 1.0453, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.07892659826361484, | |
| "grad_norm": 1.5077670812606812, | |
| "learning_rate": 9.787910922587487e-06, | |
| "loss": 1.0265, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.08208366219415943, | |
| "grad_norm": 3.2110612392425537, | |
| "learning_rate": 9.777306468716862e-06, | |
| "loss": 0.9906, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.08524072612470403, | |
| "grad_norm": 2.8840646743774414, | |
| "learning_rate": 9.766702014846236e-06, | |
| "loss": 0.9937, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.08839779005524862, | |
| "grad_norm": 2.3813140392303467, | |
| "learning_rate": 9.756097560975611e-06, | |
| "loss": 0.9922, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.09155485398579322, | |
| "grad_norm": 3.828608989715576, | |
| "learning_rate": 9.745493107104984e-06, | |
| "loss": 0.9627, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.0947119179163378, | |
| "grad_norm": 3.9936025142669678, | |
| "learning_rate": 9.73488865323436e-06, | |
| "loss": 0.9693, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.0978689818468824, | |
| "grad_norm": 6.44516134262085, | |
| "learning_rate": 9.724284199363733e-06, | |
| "loss": 0.9617, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.10102604577742699, | |
| "grad_norm": 1.657881498336792, | |
| "learning_rate": 9.713679745493108e-06, | |
| "loss": 0.9215, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.10418310970797158, | |
| "grad_norm": 3.1391677856445312, | |
| "learning_rate": 9.703075291622482e-06, | |
| "loss": 0.8923, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.10734017363851618, | |
| "grad_norm": 2.5600392818450928, | |
| "learning_rate": 9.692470837751857e-06, | |
| "loss": 0.8785, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.11049723756906077, | |
| "grad_norm": 3.348170757293701, | |
| "learning_rate": 9.68186638388123e-06, | |
| "loss": 0.8893, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.11365430149960537, | |
| "grad_norm": 2.7278106212615967, | |
| "learning_rate": 9.671261930010606e-06, | |
| "loss": 0.9022, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.11681136543014996, | |
| "grad_norm": 1.878319263458252, | |
| "learning_rate": 9.660657476139979e-06, | |
| "loss": 0.8433, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.11996842936069456, | |
| "grad_norm": 2.206759214401245, | |
| "learning_rate": 9.650053022269354e-06, | |
| "loss": 0.8602, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.12312549329123915, | |
| "grad_norm": 3.3622727394104004, | |
| "learning_rate": 9.639448568398728e-06, | |
| "loss": 0.8443, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.12628255722178375, | |
| "grad_norm": 2.4518015384674072, | |
| "learning_rate": 9.628844114528103e-06, | |
| "loss": 0.8519, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.12943962115232832, | |
| "grad_norm": 1.4743808507919312, | |
| "learning_rate": 9.618239660657476e-06, | |
| "loss": 0.7868, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.13259668508287292, | |
| "grad_norm": 1.5029278993606567, | |
| "learning_rate": 9.607635206786851e-06, | |
| "loss": 0.8163, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.13575374901341752, | |
| "grad_norm": 1.9015096426010132, | |
| "learning_rate": 9.597030752916225e-06, | |
| "loss": 0.8053, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.13891081294396213, | |
| "grad_norm": 1.7063859701156616, | |
| "learning_rate": 9.5864262990456e-06, | |
| "loss": 0.7561, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.1420678768745067, | |
| "grad_norm": 2.0188612937927246, | |
| "learning_rate": 9.575821845174974e-06, | |
| "loss": 0.7417, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.1452249408050513, | |
| "grad_norm": 1.2272390127182007, | |
| "learning_rate": 9.565217391304349e-06, | |
| "loss": 0.7212, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.1483820047355959, | |
| "grad_norm": 1.6369354724884033, | |
| "learning_rate": 9.554612937433722e-06, | |
| "loss": 0.712, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.1515390686661405, | |
| "grad_norm": 1.8548651933670044, | |
| "learning_rate": 9.544008483563097e-06, | |
| "loss": 0.717, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.15469613259668508, | |
| "grad_norm": 1.2353498935699463, | |
| "learning_rate": 9.53340402969247e-06, | |
| "loss": 0.7179, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.15785319652722968, | |
| "grad_norm": 1.025134563446045, | |
| "learning_rate": 9.522799575821846e-06, | |
| "loss": 0.7316, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.16101026045777428, | |
| "grad_norm": 1.2098180055618286, | |
| "learning_rate": 9.51219512195122e-06, | |
| "loss": 0.6877, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.16416732438831885, | |
| "grad_norm": 0.8435678482055664, | |
| "learning_rate": 9.501590668080595e-06, | |
| "loss": 0.6833, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.16732438831886345, | |
| "grad_norm": 2.2340638637542725, | |
| "learning_rate": 9.490986214209968e-06, | |
| "loss": 0.6624, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.17048145224940806, | |
| "grad_norm": 1.3673986196517944, | |
| "learning_rate": 9.480381760339343e-06, | |
| "loss": 0.6664, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.17363851617995266, | |
| "grad_norm": 3.292865037918091, | |
| "learning_rate": 9.469777306468718e-06, | |
| "loss": 0.6353, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.17679558011049723, | |
| "grad_norm": 2.751636505126953, | |
| "learning_rate": 9.459172852598092e-06, | |
| "loss": 0.6413, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.17995264404104183, | |
| "grad_norm": 2.4514176845550537, | |
| "learning_rate": 9.448568398727467e-06, | |
| "loss": 0.678, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.18310970797158643, | |
| "grad_norm": 1.9484950304031372, | |
| "learning_rate": 9.43796394485684e-06, | |
| "loss": 0.629, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.186266771902131, | |
| "grad_norm": 3.3723437786102295, | |
| "learning_rate": 9.427359490986216e-06, | |
| "loss": 0.6095, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.1894238358326756, | |
| "grad_norm": 1.590402364730835, | |
| "learning_rate": 9.416755037115589e-06, | |
| "loss": 0.6239, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.1925808997632202, | |
| "grad_norm": 2.504424810409546, | |
| "learning_rate": 9.406150583244964e-06, | |
| "loss": 0.6, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.1957379636937648, | |
| "grad_norm": 3.6238129138946533, | |
| "learning_rate": 9.395546129374338e-06, | |
| "loss": 0.6264, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.19889502762430938, | |
| "grad_norm": 1.490251064300537, | |
| "learning_rate": 9.384941675503713e-06, | |
| "loss": 0.5917, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.20205209155485399, | |
| "grad_norm": 1.043030858039856, | |
| "learning_rate": 9.374337221633086e-06, | |
| "loss": 0.5993, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.2052091554853986, | |
| "grad_norm": 0.9352568984031677, | |
| "learning_rate": 9.363732767762462e-06, | |
| "loss": 0.6105, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.20836621941594316, | |
| "grad_norm": 1.1061240434646606, | |
| "learning_rate": 9.353128313891835e-06, | |
| "loss": 0.58, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.21152328334648776, | |
| "grad_norm": 1.3794325590133667, | |
| "learning_rate": 9.34252386002121e-06, | |
| "loss": 0.6215, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.21468034727703236, | |
| "grad_norm": 1.6792662143707275, | |
| "learning_rate": 9.331919406150584e-06, | |
| "loss": 0.5623, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.21783741120757696, | |
| "grad_norm": 0.7364373803138733, | |
| "learning_rate": 9.321314952279959e-06, | |
| "loss": 0.6247, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.22099447513812154, | |
| "grad_norm": 2.3978946208953857, | |
| "learning_rate": 9.310710498409332e-06, | |
| "loss": 0.5861, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.22415153906866614, | |
| "grad_norm": 1.1635463237762451, | |
| "learning_rate": 9.300106044538707e-06, | |
| "loss": 0.5699, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.22730860299921074, | |
| "grad_norm": 2.0012998580932617, | |
| "learning_rate": 9.289501590668083e-06, | |
| "loss": 0.562, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.23046566692975531, | |
| "grad_norm": 0.9145333170890808, | |
| "learning_rate": 9.278897136797456e-06, | |
| "loss": 0.5533, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.23362273086029992, | |
| "grad_norm": 1.4702812433242798, | |
| "learning_rate": 9.268292682926831e-06, | |
| "loss": 0.5525, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.23677979479084452, | |
| "grad_norm": 1.0124057531356812, | |
| "learning_rate": 9.257688229056205e-06, | |
| "loss": 0.5679, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.23993685872138912, | |
| "grad_norm": 0.8423709869384766, | |
| "learning_rate": 9.24708377518558e-06, | |
| "loss": 0.5477, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.2430939226519337, | |
| "grad_norm": 0.9099079370498657, | |
| "learning_rate": 9.236479321314953e-06, | |
| "loss": 0.5673, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.2462509865824783, | |
| "grad_norm": 0.9014184474945068, | |
| "learning_rate": 9.225874867444329e-06, | |
| "loss": 0.5445, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.2494080505130229, | |
| "grad_norm": 0.946077287197113, | |
| "learning_rate": 9.215270413573702e-06, | |
| "loss": 0.5402, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.2525651144435675, | |
| "grad_norm": 1.0112385749816895, | |
| "learning_rate": 9.204665959703075e-06, | |
| "loss": 0.5464, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.2557221783741121, | |
| "grad_norm": 0.5525750517845154, | |
| "learning_rate": 9.19406150583245e-06, | |
| "loss": 0.5439, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.25887924230465664, | |
| "grad_norm": 0.6257532238960266, | |
| "learning_rate": 9.183457051961824e-06, | |
| "loss": 0.5659, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.26203630623520124, | |
| "grad_norm": 0.6489551067352295, | |
| "learning_rate": 9.1728525980912e-06, | |
| "loss": 0.5862, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.26519337016574585, | |
| "grad_norm": 0.6500658988952637, | |
| "learning_rate": 9.162248144220573e-06, | |
| "loss": 0.5823, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.26835043409629045, | |
| "grad_norm": 1.1931203603744507, | |
| "learning_rate": 9.151643690349948e-06, | |
| "loss": 0.5576, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.27150749802683505, | |
| "grad_norm": 0.7734494209289551, | |
| "learning_rate": 9.141039236479321e-06, | |
| "loss": 0.561, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.27466456195737965, | |
| "grad_norm": 0.6840056777000427, | |
| "learning_rate": 9.130434782608697e-06, | |
| "loss": 0.5605, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.27782162588792425, | |
| "grad_norm": 0.8087795376777649, | |
| "learning_rate": 9.11983032873807e-06, | |
| "loss": 0.6044, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.27782162588792425, | |
| "eval_runtime": 384.5737, | |
| "eval_samples_per_second": 1.389, | |
| "eval_steps_per_second": 0.694, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.2809786898184688, | |
| "grad_norm": 1.1406381130218506, | |
| "learning_rate": 9.109225874867445e-06, | |
| "loss": 0.536, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.2841357537490134, | |
| "grad_norm": 0.5932796001434326, | |
| "learning_rate": 9.098621420996819e-06, | |
| "loss": 0.544, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.287292817679558, | |
| "grad_norm": 0.6477290987968445, | |
| "learning_rate": 9.088016967126194e-06, | |
| "loss": 0.4996, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.2904498816101026, | |
| "grad_norm": 0.5976299047470093, | |
| "learning_rate": 9.077412513255567e-06, | |
| "loss": 0.5729, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.2936069455406472, | |
| "grad_norm": 0.6203446388244629, | |
| "learning_rate": 9.066808059384942e-06, | |
| "loss": 0.5519, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.2967640094711918, | |
| "grad_norm": 0.5263274908065796, | |
| "learning_rate": 9.056203605514316e-06, | |
| "loss": 0.5721, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.2999210734017364, | |
| "grad_norm": 0.541584849357605, | |
| "learning_rate": 9.045599151643691e-06, | |
| "loss": 0.5184, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.303078137332281, | |
| "grad_norm": 0.9796550869941711, | |
| "learning_rate": 9.034994697773065e-06, | |
| "loss": 0.5204, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.30623520126282555, | |
| "grad_norm": 0.8481103777885437, | |
| "learning_rate": 9.02439024390244e-06, | |
| "loss": 0.4985, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.30939226519337015, | |
| "grad_norm": 0.6354321241378784, | |
| "learning_rate": 9.013785790031813e-06, | |
| "loss": 0.5262, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.31254932912391475, | |
| "grad_norm": 0.6235267519950867, | |
| "learning_rate": 9.003181336161188e-06, | |
| "loss": 0.5487, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.31570639305445936, | |
| "grad_norm": 0.570245087146759, | |
| "learning_rate": 8.992576882290562e-06, | |
| "loss": 0.5596, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.31886345698500396, | |
| "grad_norm": 0.5904330611228943, | |
| "learning_rate": 8.981972428419937e-06, | |
| "loss": 0.5168, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.32202052091554856, | |
| "grad_norm": 0.6091082692146301, | |
| "learning_rate": 8.97136797454931e-06, | |
| "loss": 0.5572, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.32517758484609316, | |
| "grad_norm": 0.622562050819397, | |
| "learning_rate": 8.960763520678686e-06, | |
| "loss": 0.5143, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.3283346487766377, | |
| "grad_norm": 0.554428219795227, | |
| "learning_rate": 8.950159066808059e-06, | |
| "loss": 0.4945, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.3314917127071823, | |
| "grad_norm": 0.6873175501823425, | |
| "learning_rate": 8.939554612937434e-06, | |
| "loss": 0.545, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.3346487766377269, | |
| "grad_norm": 0.8151006698608398, | |
| "learning_rate": 8.92895015906681e-06, | |
| "loss": 0.5397, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.3378058405682715, | |
| "grad_norm": 0.557426929473877, | |
| "learning_rate": 8.918345705196183e-06, | |
| "loss": 0.5296, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.3409629044988161, | |
| "grad_norm": 0.5671313405036926, | |
| "learning_rate": 8.907741251325558e-06, | |
| "loss": 0.5624, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.3441199684293607, | |
| "grad_norm": 0.5837043523788452, | |
| "learning_rate": 8.897136797454932e-06, | |
| "loss": 0.5118, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.3472770323599053, | |
| "grad_norm": 0.5173460245132446, | |
| "learning_rate": 8.886532343584307e-06, | |
| "loss": 0.4991, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.35043409629044986, | |
| "grad_norm": 0.6434454917907715, | |
| "learning_rate": 8.87592788971368e-06, | |
| "loss": 0.5171, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.35359116022099446, | |
| "grad_norm": 0.7504900693893433, | |
| "learning_rate": 8.865323435843055e-06, | |
| "loss": 0.5389, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.35674822415153906, | |
| "grad_norm": 0.5454843640327454, | |
| "learning_rate": 8.854718981972429e-06, | |
| "loss": 0.5199, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.35990528808208366, | |
| "grad_norm": 0.579954981803894, | |
| "learning_rate": 8.844114528101804e-06, | |
| "loss": 0.5465, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.36306235201262826, | |
| "grad_norm": 0.49721452593803406, | |
| "learning_rate": 8.833510074231177e-06, | |
| "loss": 0.5328, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.36621941594317287, | |
| "grad_norm": 0.5003721117973328, | |
| "learning_rate": 8.822905620360553e-06, | |
| "loss": 0.5291, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.36937647987371747, | |
| "grad_norm": 0.5092913508415222, | |
| "learning_rate": 8.812301166489926e-06, | |
| "loss": 0.5366, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.372533543804262, | |
| "grad_norm": 0.6879066824913025, | |
| "learning_rate": 8.801696712619301e-06, | |
| "loss": 0.5395, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.3756906077348066, | |
| "grad_norm": 0.5802463889122009, | |
| "learning_rate": 8.791092258748675e-06, | |
| "loss": 0.5252, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.3788476716653512, | |
| "grad_norm": 0.5767027139663696, | |
| "learning_rate": 8.78048780487805e-06, | |
| "loss": 0.5034, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.3820047355958958, | |
| "grad_norm": 0.7206706404685974, | |
| "learning_rate": 8.769883351007425e-06, | |
| "loss": 0.5342, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.3851617995264404, | |
| "grad_norm": 0.5273686647415161, | |
| "learning_rate": 8.759278897136799e-06, | |
| "loss": 0.5148, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.388318863456985, | |
| "grad_norm": 0.6316848397254944, | |
| "learning_rate": 8.748674443266174e-06, | |
| "loss": 0.5348, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.3914759273875296, | |
| "grad_norm": 0.5844541788101196, | |
| "learning_rate": 8.738069989395547e-06, | |
| "loss": 0.4776, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.39463299131807417, | |
| "grad_norm": 0.575629711151123, | |
| "learning_rate": 8.727465535524922e-06, | |
| "loss": 0.5311, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.39779005524861877, | |
| "grad_norm": 0.5883514285087585, | |
| "learning_rate": 8.716861081654296e-06, | |
| "loss": 0.5179, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.40094711917916337, | |
| "grad_norm": 0.46903765201568604, | |
| "learning_rate": 8.706256627783671e-06, | |
| "loss": 0.5095, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.40410418310970797, | |
| "grad_norm": 0.5590436458587646, | |
| "learning_rate": 8.695652173913044e-06, | |
| "loss": 0.5355, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.40726124704025257, | |
| "grad_norm": 0.752703070640564, | |
| "learning_rate": 8.68504772004242e-06, | |
| "loss": 0.5597, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.4104183109707972, | |
| "grad_norm": 0.4821108281612396, | |
| "learning_rate": 8.674443266171793e-06, | |
| "loss": 0.5383, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.4135753749013418, | |
| "grad_norm": 0.4688114523887634, | |
| "learning_rate": 8.663838812301168e-06, | |
| "loss": 0.5152, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.4167324388318863, | |
| "grad_norm": 0.49247807264328003, | |
| "learning_rate": 8.653234358430542e-06, | |
| "loss": 0.4942, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.4198895027624309, | |
| "grad_norm": 0.47178471088409424, | |
| "learning_rate": 8.642629904559917e-06, | |
| "loss": 0.5551, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.4230465666929755, | |
| "grad_norm": 0.48254281282424927, | |
| "learning_rate": 8.63202545068929e-06, | |
| "loss": 0.5416, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.4262036306235201, | |
| "grad_norm": 0.48389244079589844, | |
| "learning_rate": 8.621420996818664e-06, | |
| "loss": 0.5381, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.4293606945540647, | |
| "grad_norm": 0.6082923412322998, | |
| "learning_rate": 8.610816542948039e-06, | |
| "loss": 0.532, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.4325177584846093, | |
| "grad_norm": 0.7032110691070557, | |
| "learning_rate": 8.600212089077412e-06, | |
| "loss": 0.5164, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.43567482241515393, | |
| "grad_norm": 0.6164286732673645, | |
| "learning_rate": 8.589607635206788e-06, | |
| "loss": 0.5268, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.4388318863456985, | |
| "grad_norm": 0.48358678817749023, | |
| "learning_rate": 8.579003181336161e-06, | |
| "loss": 0.5307, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.4419889502762431, | |
| "grad_norm": 0.4471558928489685, | |
| "learning_rate": 8.568398727465536e-06, | |
| "loss": 0.5715, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.4451460142067877, | |
| "grad_norm": 0.6362801194190979, | |
| "learning_rate": 8.55779427359491e-06, | |
| "loss": 0.4971, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.4483030781373323, | |
| "grad_norm": 0.7430746555328369, | |
| "learning_rate": 8.547189819724285e-06, | |
| "loss": 0.529, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.4514601420678769, | |
| "grad_norm": 0.4701765179634094, | |
| "learning_rate": 8.536585365853658e-06, | |
| "loss": 0.4956, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.4546172059984215, | |
| "grad_norm": 0.5046619176864624, | |
| "learning_rate": 8.525980911983033e-06, | |
| "loss": 0.5224, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.4577742699289661, | |
| "grad_norm": 0.5452792048454285, | |
| "learning_rate": 8.515376458112407e-06, | |
| "loss": 0.5182, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.46093133385951063, | |
| "grad_norm": 0.47501707077026367, | |
| "learning_rate": 8.504772004241782e-06, | |
| "loss": 0.4918, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.46408839779005523, | |
| "grad_norm": 0.4950338304042816, | |
| "learning_rate": 8.494167550371156e-06, | |
| "loss": 0.4824, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.46724546172059983, | |
| "grad_norm": 0.47150734066963196, | |
| "learning_rate": 8.48356309650053e-06, | |
| "loss": 0.5281, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.47040252565114443, | |
| "grad_norm": 0.7325678467750549, | |
| "learning_rate": 8.472958642629904e-06, | |
| "loss": 0.5249, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.47355958958168903, | |
| "grad_norm": 0.49232226610183716, | |
| "learning_rate": 8.46235418875928e-06, | |
| "loss": 0.4958, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.47671665351223363, | |
| "grad_norm": 0.5911719799041748, | |
| "learning_rate": 8.451749734888653e-06, | |
| "loss": 0.5001, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.47987371744277824, | |
| "grad_norm": 0.6609810590744019, | |
| "learning_rate": 8.441145281018028e-06, | |
| "loss": 0.5322, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.48303078137332284, | |
| "grad_norm": 0.5445145964622498, | |
| "learning_rate": 8.430540827147402e-06, | |
| "loss": 0.4798, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.4861878453038674, | |
| "grad_norm": 0.4578617215156555, | |
| "learning_rate": 8.419936373276777e-06, | |
| "loss": 0.5002, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.489344909234412, | |
| "grad_norm": 0.46951398253440857, | |
| "learning_rate": 8.409331919406152e-06, | |
| "loss": 0.4904, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.4925019731649566, | |
| "grad_norm": 0.4709085524082184, | |
| "learning_rate": 8.398727465535525e-06, | |
| "loss": 0.4949, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.4956590370955012, | |
| "grad_norm": 0.45145004987716675, | |
| "learning_rate": 8.3881230116649e-06, | |
| "loss": 0.5194, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.4988161010260458, | |
| "grad_norm": 0.4602022171020508, | |
| "learning_rate": 8.377518557794274e-06, | |
| "loss": 0.492, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.5019731649565904, | |
| "grad_norm": 0.5150194764137268, | |
| "learning_rate": 8.366914103923649e-06, | |
| "loss": 0.4971, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.505130228887135, | |
| "grad_norm": 0.46711647510528564, | |
| "learning_rate": 8.356309650053023e-06, | |
| "loss": 0.4792, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.5082872928176796, | |
| "grad_norm": 0.4433223605155945, | |
| "learning_rate": 8.345705196182398e-06, | |
| "loss": 0.4974, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.5114443567482242, | |
| "grad_norm": 0.4475487768650055, | |
| "learning_rate": 8.335100742311771e-06, | |
| "loss": 0.5083, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.5146014206787688, | |
| "grad_norm": 0.4782634377479553, | |
| "learning_rate": 8.324496288441146e-06, | |
| "loss": 0.5219, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.5177584846093133, | |
| "grad_norm": 0.4378846287727356, | |
| "learning_rate": 8.31389183457052e-06, | |
| "loss": 0.4875, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.5209155485398579, | |
| "grad_norm": 0.5564777851104736, | |
| "learning_rate": 8.303287380699895e-06, | |
| "loss": 0.5076, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.5240726124704025, | |
| "grad_norm": 0.4310687184333801, | |
| "learning_rate": 8.292682926829268e-06, | |
| "loss": 0.4913, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.5272296764009471, | |
| "grad_norm": 0.46949031949043274, | |
| "learning_rate": 8.282078472958644e-06, | |
| "loss": 0.4939, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.5303867403314917, | |
| "grad_norm": 0.45217466354370117, | |
| "learning_rate": 8.271474019088017e-06, | |
| "loss": 0.5162, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.5335438042620363, | |
| "grad_norm": 0.48358985781669617, | |
| "learning_rate": 8.260869565217392e-06, | |
| "loss": 0.498, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.5367008681925809, | |
| "grad_norm": 0.46768447756767273, | |
| "learning_rate": 8.250265111346766e-06, | |
| "loss": 0.4761, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.5398579321231255, | |
| "grad_norm": 0.5213289260864258, | |
| "learning_rate": 8.239660657476141e-06, | |
| "loss": 0.4934, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.5430149960536701, | |
| "grad_norm": 0.5909478664398193, | |
| "learning_rate": 8.229056203605516e-06, | |
| "loss": 0.5271, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.5461720599842147, | |
| "grad_norm": 0.644679605960846, | |
| "learning_rate": 8.21845174973489e-06, | |
| "loss": 0.5419, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.5493291239147593, | |
| "grad_norm": 0.44776424765586853, | |
| "learning_rate": 8.207847295864265e-06, | |
| "loss": 0.5491, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.5524861878453039, | |
| "grad_norm": 0.4359639286994934, | |
| "learning_rate": 8.197242841993638e-06, | |
| "loss": 0.4907, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.5556432517758485, | |
| "grad_norm": 0.4502175748348236, | |
| "learning_rate": 8.186638388123013e-06, | |
| "loss": 0.5004, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.5556432517758485, | |
| "eval_runtime": 384.6601, | |
| "eval_samples_per_second": 1.388, | |
| "eval_steps_per_second": 0.694, | |
| "step": 176 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 948, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 88, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.8460717263625912e+18, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |