| { | |
| "best_global_step": 648, | |
| "best_metric": 0.00023728572705294937, | |
| "best_model_checkpoint": "Qwen2.5_3B_VL_flip_detection/checkpoint-648", | |
| "epoch": 1.0, | |
| "eval_steps": 72, | |
| "global_step": 725, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.001380738695201933, | |
| "grad_norm": 318.0, | |
| "learning_rate": 0.0, | |
| "loss": 6.0874, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.002761477390403866, | |
| "grad_norm": 326.0, | |
| "learning_rate": 9.090909090909091e-07, | |
| "loss": 6.1564, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.004142216085605799, | |
| "grad_norm": 312.0, | |
| "learning_rate": 1.8181818181818183e-06, | |
| "loss": 6.0274, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.005522954780807732, | |
| "grad_norm": 338.0, | |
| "learning_rate": 2.7272727272727272e-06, | |
| "loss": 5.9388, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0069036934760096655, | |
| "grad_norm": 282.0, | |
| "learning_rate": 3.6363636363636366e-06, | |
| "loss": 5.6881, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.008284432171211598, | |
| "grad_norm": 290.0, | |
| "learning_rate": 4.5454545454545455e-06, | |
| "loss": 5.5135, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.009665170866413532, | |
| "grad_norm": 219.0, | |
| "learning_rate": 5.4545454545454545e-06, | |
| "loss": 4.9453, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.011045909561615464, | |
| "grad_norm": 140.0, | |
| "learning_rate": 6.363636363636364e-06, | |
| "loss": 4.438, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.012426648256817397, | |
| "grad_norm": 125.0, | |
| "learning_rate": 7.272727272727273e-06, | |
| "loss": 4.0221, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.013807386952019331, | |
| "grad_norm": 152.0, | |
| "learning_rate": 8.181818181818183e-06, | |
| "loss": 3.5748, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.015188125647221263, | |
| "grad_norm": 84.0, | |
| "learning_rate": 9.090909090909091e-06, | |
| "loss": 3.0537, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.016568864342423197, | |
| "grad_norm": 72.0, | |
| "learning_rate": 1e-05, | |
| "loss": 2.5683, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.01794960303762513, | |
| "grad_norm": 62.0, | |
| "learning_rate": 1.0909090909090909e-05, | |
| "loss": 2.1136, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.019330341732827064, | |
| "grad_norm": 63.75, | |
| "learning_rate": 1.181818181818182e-05, | |
| "loss": 1.8072, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.020711080428028994, | |
| "grad_norm": 66.0, | |
| "learning_rate": 1.2727272727272728e-05, | |
| "loss": 1.4696, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.022091819123230928, | |
| "grad_norm": 49.25, | |
| "learning_rate": 1.3636363636363637e-05, | |
| "loss": 1.1919, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.02347255781843286, | |
| "grad_norm": 53.75, | |
| "learning_rate": 1.4545454545454546e-05, | |
| "loss": 0.9495, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.024853296513634795, | |
| "grad_norm": 46.5, | |
| "learning_rate": 1.5454545454545454e-05, | |
| "loss": 0.8079, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.02623403520883673, | |
| "grad_norm": 58.75, | |
| "learning_rate": 1.6363636363636366e-05, | |
| "loss": 0.6889, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.027614773904038662, | |
| "grad_norm": 31.625, | |
| "learning_rate": 1.7272727272727274e-05, | |
| "loss": 0.5888, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.028995512599240592, | |
| "grad_norm": 61.75, | |
| "learning_rate": 1.8181818181818182e-05, | |
| "loss": 0.5087, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.030376251294442526, | |
| "grad_norm": 21.625, | |
| "learning_rate": 1.9090909090909094e-05, | |
| "loss": 0.4641, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.03175698998964446, | |
| "grad_norm": 37.25, | |
| "learning_rate": 2e-05, | |
| "loss": 0.4248, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.03313772868484639, | |
| "grad_norm": 19.25, | |
| "learning_rate": 1.997155049786629e-05, | |
| "loss": 0.3976, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.03451846738004832, | |
| "grad_norm": 54.25, | |
| "learning_rate": 1.9943100995732575e-05, | |
| "loss": 0.3755, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.03589920607525026, | |
| "grad_norm": 37.75, | |
| "learning_rate": 1.9914651493598865e-05, | |
| "loss": 0.3674, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.03727994477045219, | |
| "grad_norm": 12.3125, | |
| "learning_rate": 1.9886201991465152e-05, | |
| "loss": 0.3343, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.03866068346565413, | |
| "grad_norm": 23.875, | |
| "learning_rate": 1.985775248933144e-05, | |
| "loss": 0.3108, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.04004142216085606, | |
| "grad_norm": 16.0, | |
| "learning_rate": 1.9829302987197725e-05, | |
| "loss": 0.3134, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.04142216085605799, | |
| "grad_norm": 140.0, | |
| "learning_rate": 1.9800853485064012e-05, | |
| "loss": 0.2831, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.042802899551259925, | |
| "grad_norm": 12.875, | |
| "learning_rate": 1.97724039829303e-05, | |
| "loss": 0.2804, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.044183638246461855, | |
| "grad_norm": 70.0, | |
| "learning_rate": 1.974395448079659e-05, | |
| "loss": 0.2625, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.04556437694166379, | |
| "grad_norm": 20.625, | |
| "learning_rate": 1.9715504978662876e-05, | |
| "loss": 0.2517, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.04694511563686572, | |
| "grad_norm": 33.25, | |
| "learning_rate": 1.9687055476529162e-05, | |
| "loss": 0.2638, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.04832585433206766, | |
| "grad_norm": 7.1875, | |
| "learning_rate": 1.965860597439545e-05, | |
| "loss": 0.2471, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.04970659302726959, | |
| "grad_norm": 4.53125, | |
| "learning_rate": 1.9630156472261736e-05, | |
| "loss": 0.228, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.05108733172247152, | |
| "grad_norm": 3.828125, | |
| "learning_rate": 1.9601706970128026e-05, | |
| "loss": 0.2205, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.05246807041767346, | |
| "grad_norm": 8.8125, | |
| "learning_rate": 1.957325746799431e-05, | |
| "loss": 0.221, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.05384880911287539, | |
| "grad_norm": 22.25, | |
| "learning_rate": 1.95448079658606e-05, | |
| "loss": 0.2166, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.055229547808077324, | |
| "grad_norm": 3.203125, | |
| "learning_rate": 1.9516358463726886e-05, | |
| "loss": 0.206, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.056610286503279254, | |
| "grad_norm": 10.5, | |
| "learning_rate": 1.9487908961593173e-05, | |
| "loss": 0.2014, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.057991025198481184, | |
| "grad_norm": 3.265625, | |
| "learning_rate": 1.9459459459459463e-05, | |
| "loss": 0.1947, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.05937176389368312, | |
| "grad_norm": 8.875, | |
| "learning_rate": 1.943100995732575e-05, | |
| "loss": 0.1937, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.06075250258888505, | |
| "grad_norm": 4.375, | |
| "learning_rate": 1.9402560455192037e-05, | |
| "loss": 0.1874, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.06213324128408699, | |
| "grad_norm": 3.625, | |
| "learning_rate": 1.9374110953058323e-05, | |
| "loss": 0.1797, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.06351397997928893, | |
| "grad_norm": 3.359375, | |
| "learning_rate": 1.934566145092461e-05, | |
| "loss": 0.177, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.06489471867449086, | |
| "grad_norm": 4.0, | |
| "learning_rate": 1.9317211948790897e-05, | |
| "loss": 0.1697, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.06627545736969279, | |
| "grad_norm": 3.359375, | |
| "learning_rate": 1.9288762446657187e-05, | |
| "loss": 0.163, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.06765619606489472, | |
| "grad_norm": 3.46875, | |
| "learning_rate": 1.926031294452347e-05, | |
| "loss": 0.1563, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.06903693476009665, | |
| "grad_norm": 3.8125, | |
| "learning_rate": 1.923186344238976e-05, | |
| "loss": 0.1531, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.07041767345529859, | |
| "grad_norm": 5.1875, | |
| "learning_rate": 1.9203413940256047e-05, | |
| "loss": 0.1461, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.07179841215050052, | |
| "grad_norm": 3.265625, | |
| "learning_rate": 1.9174964438122334e-05, | |
| "loss": 0.1408, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.07317915084570245, | |
| "grad_norm": 6.21875, | |
| "learning_rate": 1.9146514935988624e-05, | |
| "loss": 0.1403, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.07455988954090438, | |
| "grad_norm": 3.15625, | |
| "learning_rate": 1.9118065433854907e-05, | |
| "loss": 0.1285, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.07594062823610631, | |
| "grad_norm": 3.125, | |
| "learning_rate": 1.9089615931721197e-05, | |
| "loss": 0.1205, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.07732136693130826, | |
| "grad_norm": 3.3125, | |
| "learning_rate": 1.9061166429587484e-05, | |
| "loss": 0.1147, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.07870210562651019, | |
| "grad_norm": 3.0625, | |
| "learning_rate": 1.903271692745377e-05, | |
| "loss": 0.1054, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.08008284432171212, | |
| "grad_norm": 5.15625, | |
| "learning_rate": 1.9004267425320058e-05, | |
| "loss": 0.1014, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.08146358301691405, | |
| "grad_norm": 3.3125, | |
| "learning_rate": 1.8975817923186348e-05, | |
| "loss": 0.0919, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.08284432171211598, | |
| "grad_norm": 3.078125, | |
| "learning_rate": 1.894736842105263e-05, | |
| "loss": 0.0857, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.08422506040731792, | |
| "grad_norm": 2.921875, | |
| "learning_rate": 1.891891891891892e-05, | |
| "loss": 0.0781, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.08560579910251985, | |
| "grad_norm": 2.890625, | |
| "learning_rate": 1.8890469416785208e-05, | |
| "loss": 0.0702, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.08698653779772178, | |
| "grad_norm": 2.734375, | |
| "learning_rate": 1.8862019914651495e-05, | |
| "loss": 0.0619, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.08836727649292371, | |
| "grad_norm": 2.53125, | |
| "learning_rate": 1.8833570412517785e-05, | |
| "loss": 0.055, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.08974801518812564, | |
| "grad_norm": 2.46875, | |
| "learning_rate": 1.8805120910384068e-05, | |
| "loss": 0.049, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.09112875388332758, | |
| "grad_norm": 2.53125, | |
| "learning_rate": 1.8776671408250358e-05, | |
| "loss": 0.0431, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.09250949257852951, | |
| "grad_norm": 2.734375, | |
| "learning_rate": 1.8748221906116645e-05, | |
| "loss": 0.0366, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.09389023127373144, | |
| "grad_norm": 4.0625, | |
| "learning_rate": 1.871977240398293e-05, | |
| "loss": 0.0433, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.09527096996893337, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.869132290184922e-05, | |
| "loss": 0.0275, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.09665170866413532, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 1.8662873399715505e-05, | |
| "loss": 0.0239, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.09803244735933725, | |
| "grad_norm": 1.53125, | |
| "learning_rate": 1.8634423897581795e-05, | |
| "loss": 0.0238, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.09941318605453918, | |
| "grad_norm": 2.984375, | |
| "learning_rate": 1.8605974395448082e-05, | |
| "loss": 0.019, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.09941318605453918, | |
| "eval_loss": 0.016371596604585648, | |
| "eval_runtime": 592.2167, | |
| "eval_samples_per_second": 2.175, | |
| "eval_steps_per_second": 2.175, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.10079392474974111, | |
| "grad_norm": 1.515625, | |
| "learning_rate": 1.857752489331437e-05, | |
| "loss": 0.0154, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.10217466344494304, | |
| "grad_norm": 0.9296875, | |
| "learning_rate": 1.8549075391180655e-05, | |
| "loss": 0.013, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.10355540214014498, | |
| "grad_norm": 0.91796875, | |
| "learning_rate": 1.8520625889046942e-05, | |
| "loss": 0.0154, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.10493614083534691, | |
| "grad_norm": 1.2734375, | |
| "learning_rate": 1.849217638691323e-05, | |
| "loss": 0.0101, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.10631687953054884, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 1.846372688477952e-05, | |
| "loss": 0.0098, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.10769761822575077, | |
| "grad_norm": 6.3125, | |
| "learning_rate": 1.8435277382645806e-05, | |
| "loss": 0.0114, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.1090783569209527, | |
| "grad_norm": 0.51171875, | |
| "learning_rate": 1.8406827880512092e-05, | |
| "loss": 0.0065, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.11045909561615465, | |
| "grad_norm": 0.6796875, | |
| "learning_rate": 1.8378378378378383e-05, | |
| "loss": 0.0067, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.11183983431135658, | |
| "grad_norm": 1.0859375, | |
| "learning_rate": 1.8349928876244666e-05, | |
| "loss": 0.0109, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.11322057300655851, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 1.8321479374110956e-05, | |
| "loss": 0.0045, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.11460131170176044, | |
| "grad_norm": 0.6875, | |
| "learning_rate": 1.8293029871977243e-05, | |
| "loss": 0.0065, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.11598205039696237, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 1.826458036984353e-05, | |
| "loss": 0.0036, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.11736278909216431, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 1.8236130867709816e-05, | |
| "loss": 0.0029, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.11874352778736624, | |
| "grad_norm": 0.22265625, | |
| "learning_rate": 1.8207681365576103e-05, | |
| "loss": 0.0025, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.12012426648256817, | |
| "grad_norm": 0.208984375, | |
| "learning_rate": 1.817923186344239e-05, | |
| "loss": 0.0023, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.1215050051777701, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 1.815078236130868e-05, | |
| "loss": 0.0021, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.12288574387297203, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 1.8122332859174966e-05, | |
| "loss": 0.0019, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.12426648256817398, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 1.8093883357041253e-05, | |
| "loss": 0.0016, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.1256472212633759, | |
| "grad_norm": 0.134765625, | |
| "learning_rate": 1.806543385490754e-05, | |
| "loss": 0.0014, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.12702795995857785, | |
| "grad_norm": 0.96484375, | |
| "learning_rate": 1.8036984352773827e-05, | |
| "loss": 0.0027, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.12840869865377977, | |
| "grad_norm": 0.1162109375, | |
| "learning_rate": 1.8008534850640117e-05, | |
| "loss": 0.0012, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.1297894373489817, | |
| "grad_norm": 0.107421875, | |
| "learning_rate": 1.7980085348506404e-05, | |
| "loss": 0.0011, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.13117017604418363, | |
| "grad_norm": 0.6015625, | |
| "learning_rate": 1.795163584637269e-05, | |
| "loss": 0.0025, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.13255091473938557, | |
| "grad_norm": 0.0869140625, | |
| "learning_rate": 1.7923186344238977e-05, | |
| "loss": 0.0009, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.13393165343458752, | |
| "grad_norm": 0.1396484375, | |
| "learning_rate": 1.7894736842105264e-05, | |
| "loss": 0.001, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.13531239212978943, | |
| "grad_norm": 0.078125, | |
| "learning_rate": 1.7866287339971554e-05, | |
| "loss": 0.0008, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.13669313082499138, | |
| "grad_norm": 0.1689453125, | |
| "learning_rate": 1.783783783783784e-05, | |
| "loss": 0.0015, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.1380738695201933, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 1.7809388335704127e-05, | |
| "loss": 0.001, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.13945460821539524, | |
| "grad_norm": 0.06640625, | |
| "learning_rate": 1.7780938833570414e-05, | |
| "loss": 0.0007, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.14083534691059718, | |
| "grad_norm": 0.0634765625, | |
| "learning_rate": 1.77524893314367e-05, | |
| "loss": 0.0006, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.1422160856057991, | |
| "grad_norm": 0.058837890625, | |
| "learning_rate": 1.7724039829302988e-05, | |
| "loss": 0.0006, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.14359682430100104, | |
| "grad_norm": 0.053955078125, | |
| "learning_rate": 1.7695590327169278e-05, | |
| "loss": 0.0005, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.14497756299620296, | |
| "grad_norm": 0.054931640625, | |
| "learning_rate": 1.766714082503556e-05, | |
| "loss": 0.0005, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.1463583016914049, | |
| "grad_norm": 0.054443359375, | |
| "learning_rate": 1.763869132290185e-05, | |
| "loss": 0.0005, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.14773904038660685, | |
| "grad_norm": 0.053466796875, | |
| "learning_rate": 1.7610241820768138e-05, | |
| "loss": 0.0005, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.14911977908180876, | |
| "grad_norm": 0.81640625, | |
| "learning_rate": 1.7581792318634425e-05, | |
| "loss": 0.001, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.1505005177770107, | |
| "grad_norm": 0.048583984375, | |
| "learning_rate": 1.7553342816500715e-05, | |
| "loss": 0.0005, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.15188125647221262, | |
| "grad_norm": 0.05322265625, | |
| "learning_rate": 1.7524893314367e-05, | |
| "loss": 0.0005, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.15326199516741457, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 1.7496443812233288e-05, | |
| "loss": 0.0005, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.1546427338626165, | |
| "grad_norm": 2.5625, | |
| "learning_rate": 1.7467994310099575e-05, | |
| "loss": 0.0025, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.15602347255781843, | |
| "grad_norm": 0.4609375, | |
| "learning_rate": 1.743954480796586e-05, | |
| "loss": 0.0029, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.15740421125302037, | |
| "grad_norm": 0.56640625, | |
| "learning_rate": 1.741109530583215e-05, | |
| "loss": 0.0007, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.1587849499482223, | |
| "grad_norm": 0.044189453125, | |
| "learning_rate": 1.738264580369844e-05, | |
| "loss": 0.0004, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.16016568864342423, | |
| "grad_norm": 0.04443359375, | |
| "learning_rate": 1.7354196301564722e-05, | |
| "loss": 0.0004, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.16154642733862618, | |
| "grad_norm": 0.045654296875, | |
| "learning_rate": 1.7325746799431012e-05, | |
| "loss": 0.0004, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.1629271660338281, | |
| "grad_norm": 0.09033203125, | |
| "learning_rate": 1.72972972972973e-05, | |
| "loss": 0.0005, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.16430790472903004, | |
| "grad_norm": 0.046875, | |
| "learning_rate": 1.7268847795163585e-05, | |
| "loss": 0.0004, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.16568864342423195, | |
| "grad_norm": 0.0439453125, | |
| "learning_rate": 1.7240398293029875e-05, | |
| "loss": 0.0004, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.1670693821194339, | |
| "grad_norm": 0.41015625, | |
| "learning_rate": 1.721194879089616e-05, | |
| "loss": 0.0056, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.16845012081463584, | |
| "grad_norm": 0.07275390625, | |
| "learning_rate": 1.718349928876245e-05, | |
| "loss": 0.0005, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.16983085950983776, | |
| "grad_norm": 0.042236328125, | |
| "learning_rate": 1.7155049786628736e-05, | |
| "loss": 0.0004, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.1712115982050397, | |
| "grad_norm": 0.10107421875, | |
| "learning_rate": 1.7126600284495022e-05, | |
| "loss": 0.0005, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.17259233690024162, | |
| "grad_norm": 0.040283203125, | |
| "learning_rate": 1.709815078236131e-05, | |
| "loss": 0.0004, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.17397307559544356, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 1.7069701280227596e-05, | |
| "loss": 0.0034, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.1753538142906455, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 1.7041251778093886e-05, | |
| "loss": 0.0008, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.17673455298584742, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 1.7012802275960173e-05, | |
| "loss": 0.0019, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.17811529168104936, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 1.698435277382646e-05, | |
| "loss": 0.0018, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.17949603037625128, | |
| "grad_norm": 0.10498046875, | |
| "learning_rate": 1.6955903271692746e-05, | |
| "loss": 0.0006, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.18087676907145323, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 1.6927453769559036e-05, | |
| "loss": 0.001, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.18225750776665517, | |
| "grad_norm": 0.12060546875, | |
| "learning_rate": 1.689900426742532e-05, | |
| "loss": 0.0011, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.18363824646185709, | |
| "grad_norm": 0.034912109375, | |
| "learning_rate": 1.687055476529161e-05, | |
| "loss": 0.0003, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.18501898515705903, | |
| "grad_norm": 0.03515625, | |
| "learning_rate": 1.6842105263157896e-05, | |
| "loss": 0.0003, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.18639972385226097, | |
| "grad_norm": 0.08544921875, | |
| "learning_rate": 1.6813655761024183e-05, | |
| "loss": 0.0007, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.1877804625474629, | |
| "grad_norm": 0.03466796875, | |
| "learning_rate": 1.6785206258890473e-05, | |
| "loss": 0.0003, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.18916120124266483, | |
| "grad_norm": 0.033203125, | |
| "learning_rate": 1.6756756756756757e-05, | |
| "loss": 0.0003, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.19054193993786675, | |
| "grad_norm": 0.1650390625, | |
| "learning_rate": 1.6728307254623047e-05, | |
| "loss": 0.0004, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.1919226786330687, | |
| "grad_norm": 0.0703125, | |
| "learning_rate": 1.6699857752489334e-05, | |
| "loss": 0.0004, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.19330341732827064, | |
| "grad_norm": 0.03955078125, | |
| "learning_rate": 1.667140825035562e-05, | |
| "loss": 0.0004, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.19468415602347255, | |
| "grad_norm": 0.138671875, | |
| "learning_rate": 1.6642958748221907e-05, | |
| "loss": 0.0004, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.1960648947186745, | |
| "grad_norm": 1.015625, | |
| "learning_rate": 1.6614509246088194e-05, | |
| "loss": 0.0013, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.19744563341387641, | |
| "grad_norm": 0.03271484375, | |
| "learning_rate": 1.658605974395448e-05, | |
| "loss": 0.0003, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.19882637210907836, | |
| "grad_norm": 0.03369140625, | |
| "learning_rate": 1.655761024182077e-05, | |
| "loss": 0.0003, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.19882637210907836, | |
| "eval_loss": 0.0005994443781673908, | |
| "eval_runtime": 594.9971, | |
| "eval_samples_per_second": 2.165, | |
| "eval_steps_per_second": 2.165, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.2002071108042803, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 1.6529160739687057e-05, | |
| "loss": 0.0005, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.20158784949948222, | |
| "grad_norm": 0.054443359375, | |
| "learning_rate": 1.6500711237553344e-05, | |
| "loss": 0.0005, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.20296858819468416, | |
| "grad_norm": 0.494140625, | |
| "learning_rate": 1.647226173541963e-05, | |
| "loss": 0.0038, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.20434932688988608, | |
| "grad_norm": 0.09228515625, | |
| "learning_rate": 1.6443812233285917e-05, | |
| "loss": 0.0005, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.20573006558508802, | |
| "grad_norm": 0.0380859375, | |
| "learning_rate": 1.6415362731152208e-05, | |
| "loss": 0.0003, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.20711080428028997, | |
| "grad_norm": 0.044189453125, | |
| "learning_rate": 1.6386913229018494e-05, | |
| "loss": 0.0005, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.20849154297549188, | |
| "grad_norm": 0.50390625, | |
| "learning_rate": 1.635846372688478e-05, | |
| "loss": 0.0018, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.20987228167069383, | |
| "grad_norm": 0.2275390625, | |
| "learning_rate": 1.6330014224751068e-05, | |
| "loss": 0.0006, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.21125302036589574, | |
| "grad_norm": 0.08349609375, | |
| "learning_rate": 1.6301564722617355e-05, | |
| "loss": 0.0005, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.2126337590610977, | |
| "grad_norm": 0.1689453125, | |
| "learning_rate": 1.627311522048364e-05, | |
| "loss": 0.0007, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.21401449775629963, | |
| "grad_norm": 0.89453125, | |
| "learning_rate": 1.624466571834993e-05, | |
| "loss": 0.0041, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.21539523645150155, | |
| "grad_norm": 1.0625, | |
| "learning_rate": 1.6216216216216218e-05, | |
| "loss": 0.0024, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.2167759751467035, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 1.6187766714082505e-05, | |
| "loss": 0.0014, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.2181567138419054, | |
| "grad_norm": 0.09716796875, | |
| "learning_rate": 1.615931721194879e-05, | |
| "loss": 0.0005, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.21953745253710735, | |
| "grad_norm": 0.2431640625, | |
| "learning_rate": 1.6130867709815078e-05, | |
| "loss": 0.001, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.2209181912323093, | |
| "grad_norm": 0.047119140625, | |
| "learning_rate": 1.610241820768137e-05, | |
| "loss": 0.0004, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.2222989299275112, | |
| "grad_norm": 0.0303955078125, | |
| "learning_rate": 1.6073968705547652e-05, | |
| "loss": 0.0003, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.22367966862271316, | |
| "grad_norm": 0.07470703125, | |
| "learning_rate": 1.6045519203413942e-05, | |
| "loss": 0.0004, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.22506040731791507, | |
| "grad_norm": 0.042724609375, | |
| "learning_rate": 1.601706970128023e-05, | |
| "loss": 0.0003, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.22644114601311702, | |
| "grad_norm": 0.02880859375, | |
| "learning_rate": 1.5988620199146515e-05, | |
| "loss": 0.0003, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.22782188470831896, | |
| "grad_norm": 0.029541015625, | |
| "learning_rate": 1.5960170697012805e-05, | |
| "loss": 0.0003, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.22920262340352088, | |
| "grad_norm": 0.028076171875, | |
| "learning_rate": 1.5931721194879092e-05, | |
| "loss": 0.0003, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.23058336209872282, | |
| "grad_norm": 0.0291748046875, | |
| "learning_rate": 1.590327169274538e-05, | |
| "loss": 0.0003, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.23196410079392474, | |
| "grad_norm": 0.0284423828125, | |
| "learning_rate": 1.5874822190611666e-05, | |
| "loss": 0.0003, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.23334483948912668, | |
| "grad_norm": 0.0279541015625, | |
| "learning_rate": 1.5846372688477952e-05, | |
| "loss": 0.0003, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.23472557818432863, | |
| "grad_norm": 0.0277099609375, | |
| "learning_rate": 1.581792318634424e-05, | |
| "loss": 0.0003, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.23610631687953054, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 1.578947368421053e-05, | |
| "loss": 0.0036, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.2374870555747325, | |
| "grad_norm": 0.0274658203125, | |
| "learning_rate": 1.5761024182076813e-05, | |
| "loss": 0.0003, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.2388677942699344, | |
| "grad_norm": 1.0625, | |
| "learning_rate": 1.5732574679943103e-05, | |
| "loss": 0.0027, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.24024853296513635, | |
| "grad_norm": 0.0274658203125, | |
| "learning_rate": 1.570412517780939e-05, | |
| "loss": 0.0003, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.2416292716603383, | |
| "grad_norm": 0.0277099609375, | |
| "learning_rate": 1.5675675675675676e-05, | |
| "loss": 0.0003, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.2430100103555402, | |
| "grad_norm": 0.0262451171875, | |
| "learning_rate": 1.5647226173541966e-05, | |
| "loss": 0.0003, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.24439074905074215, | |
| "grad_norm": 0.0264892578125, | |
| "learning_rate": 1.561877667140825e-05, | |
| "loss": 0.0003, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.24577148774594407, | |
| "grad_norm": 0.0274658203125, | |
| "learning_rate": 1.559032716927454e-05, | |
| "loss": 0.0003, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.247152226441146, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 1.5561877667140826e-05, | |
| "loss": 0.001, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.24853296513634796, | |
| "grad_norm": 0.034423828125, | |
| "learning_rate": 1.5533428165007113e-05, | |
| "loss": 0.0003, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.24991370383154987, | |
| "grad_norm": 0.038330078125, | |
| "learning_rate": 1.55049786628734e-05, | |
| "loss": 0.0003, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.2512944425267518, | |
| "grad_norm": 0.0262451171875, | |
| "learning_rate": 1.547652916073969e-05, | |
| "loss": 0.0003, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.25267518122195376, | |
| "grad_norm": 1.484375, | |
| "learning_rate": 1.5448079658605977e-05, | |
| "loss": 0.0017, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.2540559199171557, | |
| "grad_norm": 0.025634765625, | |
| "learning_rate": 1.5419630156472263e-05, | |
| "loss": 0.0002, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.2554366586123576, | |
| "grad_norm": 0.02685546875, | |
| "learning_rate": 1.539118065433855e-05, | |
| "loss": 0.0003, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.25681739730755954, | |
| "grad_norm": 3.921875, | |
| "learning_rate": 1.5362731152204837e-05, | |
| "loss": 0.0042, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.2581981360027615, | |
| "grad_norm": 0.81640625, | |
| "learning_rate": 1.5334281650071127e-05, | |
| "loss": 0.0012, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.2595788746979634, | |
| "grad_norm": 0.0390625, | |
| "learning_rate": 1.530583214793741e-05, | |
| "loss": 0.0004, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.26095961339316537, | |
| "grad_norm": 0.0262451171875, | |
| "learning_rate": 1.52773826458037e-05, | |
| "loss": 0.0003, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.26234035208836726, | |
| "grad_norm": 0.0250244140625, | |
| "learning_rate": 1.5248933143669986e-05, | |
| "loss": 0.0002, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.2637210907835692, | |
| "grad_norm": 0.1689453125, | |
| "learning_rate": 1.5220483641536274e-05, | |
| "loss": 0.0004, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.26510182947877114, | |
| "grad_norm": 0.0294189453125, | |
| "learning_rate": 1.5192034139402562e-05, | |
| "loss": 0.0003, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.2664825681739731, | |
| "grad_norm": 0.419921875, | |
| "learning_rate": 1.5163584637268849e-05, | |
| "loss": 0.0013, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.26786330686917503, | |
| "grad_norm": 0.0260009765625, | |
| "learning_rate": 1.5135135135135138e-05, | |
| "loss": 0.0002, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.2692440455643769, | |
| "grad_norm": 0.02685546875, | |
| "learning_rate": 1.5106685633001423e-05, | |
| "loss": 0.0003, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.27062478425957887, | |
| "grad_norm": 0.025634765625, | |
| "learning_rate": 1.5078236130867711e-05, | |
| "loss": 0.0002, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.2720055229547808, | |
| "grad_norm": 0.0244140625, | |
| "learning_rate": 1.5049786628733998e-05, | |
| "loss": 0.0002, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.27338626164998275, | |
| "grad_norm": 0.029296875, | |
| "learning_rate": 1.5021337126600286e-05, | |
| "loss": 0.0003, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.2747670003451847, | |
| "grad_norm": 0.02880859375, | |
| "learning_rate": 1.4992887624466573e-05, | |
| "loss": 0.0003, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.2761477390403866, | |
| "grad_norm": 0.024658203125, | |
| "learning_rate": 1.4964438122332861e-05, | |
| "loss": 0.0002, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.27752847773558853, | |
| "grad_norm": 0.0732421875, | |
| "learning_rate": 1.4935988620199146e-05, | |
| "loss": 0.0004, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.2789092164307905, | |
| "grad_norm": 0.03466796875, | |
| "learning_rate": 1.4907539118065435e-05, | |
| "loss": 0.0003, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.2802899551259924, | |
| "grad_norm": 0.0235595703125, | |
| "learning_rate": 1.4879089615931723e-05, | |
| "loss": 0.0002, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.28167069382119436, | |
| "grad_norm": 0.0257568359375, | |
| "learning_rate": 1.485064011379801e-05, | |
| "loss": 0.0002, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.28305143251639625, | |
| "grad_norm": 0.0263671875, | |
| "learning_rate": 1.4822190611664298e-05, | |
| "loss": 0.0002, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.2844321712115982, | |
| "grad_norm": 0.045166015625, | |
| "learning_rate": 1.4793741109530583e-05, | |
| "loss": 0.0004, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.28581290990680014, | |
| "grad_norm": 0.0252685546875, | |
| "learning_rate": 1.4765291607396872e-05, | |
| "loss": 0.0002, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.2871936486020021, | |
| "grad_norm": 0.703125, | |
| "learning_rate": 1.4736842105263159e-05, | |
| "loss": 0.0015, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.288574387297204, | |
| "grad_norm": 0.0233154296875, | |
| "learning_rate": 1.4708392603129447e-05, | |
| "loss": 0.0002, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.2899551259924059, | |
| "grad_norm": 0.0255126953125, | |
| "learning_rate": 1.4679943100995732e-05, | |
| "loss": 0.0003, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.29133586468760786, | |
| "grad_norm": 0.0233154296875, | |
| "learning_rate": 1.465149359886202e-05, | |
| "loss": 0.0002, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.2927166033828098, | |
| "grad_norm": 0.023193359375, | |
| "learning_rate": 1.4623044096728309e-05, | |
| "loss": 0.0002, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.29409734207801175, | |
| "grad_norm": 0.51953125, | |
| "learning_rate": 1.4594594594594596e-05, | |
| "loss": 0.0004, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.2954780807732137, | |
| "grad_norm": 0.52734375, | |
| "learning_rate": 1.4566145092460884e-05, | |
| "loss": 0.0003, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.2968588194684156, | |
| "grad_norm": 0.02392578125, | |
| "learning_rate": 1.453769559032717e-05, | |
| "loss": 0.0002, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.2982395581636175, | |
| "grad_norm": 0.0234375, | |
| "learning_rate": 1.4509246088193457e-05, | |
| "loss": 0.0002, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.2982395581636175, | |
| "eval_loss": 0.000599406601395458, | |
| "eval_runtime": 586.817, | |
| "eval_samples_per_second": 2.195, | |
| "eval_steps_per_second": 2.195, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.29962029685881947, | |
| "grad_norm": 0.024658203125, | |
| "learning_rate": 1.4480796586059744e-05, | |
| "loss": 0.0002, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.3010010355540214, | |
| "grad_norm": 0.16796875, | |
| "learning_rate": 1.4452347083926033e-05, | |
| "loss": 0.0028, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.30238177424922336, | |
| "grad_norm": 0.0230712890625, | |
| "learning_rate": 1.442389758179232e-05, | |
| "loss": 0.0002, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.30376251294442524, | |
| "grad_norm": 0.024658203125, | |
| "learning_rate": 1.4395448079658608e-05, | |
| "loss": 0.0002, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.3051432516396272, | |
| "grad_norm": 0.0238037109375, | |
| "learning_rate": 1.4366998577524896e-05, | |
| "loss": 0.0002, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.30652399033482913, | |
| "grad_norm": 0.4765625, | |
| "learning_rate": 1.4338549075391181e-05, | |
| "loss": 0.0009, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.3079047290300311, | |
| "grad_norm": 0.023193359375, | |
| "learning_rate": 1.431009957325747e-05, | |
| "loss": 0.0002, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.309285467725233, | |
| "grad_norm": 0.023193359375, | |
| "learning_rate": 1.4281650071123756e-05, | |
| "loss": 0.0002, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.3106662064204349, | |
| "grad_norm": 0.0223388671875, | |
| "learning_rate": 1.4253200568990045e-05, | |
| "loss": 0.0002, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.31204694511563685, | |
| "grad_norm": 0.022216796875, | |
| "learning_rate": 1.422475106685633e-05, | |
| "loss": 0.0002, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.3134276838108388, | |
| "grad_norm": 0.02294921875, | |
| "learning_rate": 1.4196301564722618e-05, | |
| "loss": 0.0002, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.31480842250604074, | |
| "grad_norm": 0.0233154296875, | |
| "learning_rate": 1.4167852062588905e-05, | |
| "loss": 0.0002, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.3161891612012427, | |
| "grad_norm": 0.0238037109375, | |
| "learning_rate": 1.4139402560455193e-05, | |
| "loss": 0.0002, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.3175698998964446, | |
| "grad_norm": 1.25, | |
| "learning_rate": 1.4110953058321482e-05, | |
| "loss": 0.012, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.3189506385916465, | |
| "grad_norm": 0.0224609375, | |
| "learning_rate": 1.4082503556187767e-05, | |
| "loss": 0.0002, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.32033137728684846, | |
| "grad_norm": 0.0218505859375, | |
| "learning_rate": 1.4054054054054055e-05, | |
| "loss": 0.0002, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.3217121159820504, | |
| "grad_norm": 0.06005859375, | |
| "learning_rate": 1.4025604551920342e-05, | |
| "loss": 0.0004, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.32309285467725235, | |
| "grad_norm": 0.042236328125, | |
| "learning_rate": 1.399715504978663e-05, | |
| "loss": 0.0004, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.32447359337245424, | |
| "grad_norm": 0.031494140625, | |
| "learning_rate": 1.3968705547652917e-05, | |
| "loss": 0.0003, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.3258543320676562, | |
| "grad_norm": 0.0264892578125, | |
| "learning_rate": 1.3940256045519206e-05, | |
| "loss": 0.0003, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.3272350707628581, | |
| "grad_norm": 0.1318359375, | |
| "learning_rate": 1.391180654338549e-05, | |
| "loss": 0.0007, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.32861580945806007, | |
| "grad_norm": 0.034423828125, | |
| "learning_rate": 1.3883357041251779e-05, | |
| "loss": 0.0003, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.329996548153262, | |
| "grad_norm": 0.036865234375, | |
| "learning_rate": 1.3854907539118068e-05, | |
| "loss": 0.0003, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.3313772868484639, | |
| "grad_norm": 0.039794921875, | |
| "learning_rate": 1.3826458036984354e-05, | |
| "loss": 0.0003, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.33275802554366585, | |
| "grad_norm": 0.0390625, | |
| "learning_rate": 1.3798008534850643e-05, | |
| "loss": 0.0003, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.3341387642388678, | |
| "grad_norm": 0.03759765625, | |
| "learning_rate": 1.3769559032716928e-05, | |
| "loss": 0.0003, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.33551950293406974, | |
| "grad_norm": 0.045654296875, | |
| "learning_rate": 1.3741109530583216e-05, | |
| "loss": 0.0004, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.3369002416292717, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 1.3712660028449503e-05, | |
| "loss": 0.0004, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.33828098032447357, | |
| "grad_norm": 0.0260009765625, | |
| "learning_rate": 1.3684210526315791e-05, | |
| "loss": 0.0003, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.3396617190196755, | |
| "grad_norm": 0.051513671875, | |
| "learning_rate": 1.3655761024182076e-05, | |
| "loss": 0.0004, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.34104245771487746, | |
| "grad_norm": 0.023193359375, | |
| "learning_rate": 1.3627311522048365e-05, | |
| "loss": 0.0002, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.3424231964100794, | |
| "grad_norm": 0.062255859375, | |
| "learning_rate": 1.3598862019914653e-05, | |
| "loss": 0.0003, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.34380393510528134, | |
| "grad_norm": 0.0250244140625, | |
| "learning_rate": 1.357041251778094e-05, | |
| "loss": 0.0003, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.34518467380048323, | |
| "grad_norm": 0.0235595703125, | |
| "learning_rate": 1.3541963015647228e-05, | |
| "loss": 0.0002, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.3465654124956852, | |
| "grad_norm": 0.021728515625, | |
| "learning_rate": 1.3513513513513515e-05, | |
| "loss": 0.0002, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.3479461511908871, | |
| "grad_norm": 1.6015625, | |
| "learning_rate": 1.3485064011379802e-05, | |
| "loss": 0.0012, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.34932688988608906, | |
| "grad_norm": 0.0224609375, | |
| "learning_rate": 1.3456614509246089e-05, | |
| "loss": 0.0002, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.350707628581291, | |
| "grad_norm": 0.034912109375, | |
| "learning_rate": 1.3428165007112377e-05, | |
| "loss": 0.0002, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.3520883672764929, | |
| "grad_norm": 0.034423828125, | |
| "learning_rate": 1.3399715504978664e-05, | |
| "loss": 0.0003, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.35346910597169484, | |
| "grad_norm": 0.0206298828125, | |
| "learning_rate": 1.3371266002844952e-05, | |
| "loss": 0.0002, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.3548498446668968, | |
| "grad_norm": 0.0247802734375, | |
| "learning_rate": 1.3342816500711237e-05, | |
| "loss": 0.0002, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.35623058336209873, | |
| "grad_norm": 0.023193359375, | |
| "learning_rate": 1.3314366998577526e-05, | |
| "loss": 0.0002, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.3576113220573007, | |
| "grad_norm": 0.0208740234375, | |
| "learning_rate": 1.3285917496443814e-05, | |
| "loss": 0.0002, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.35899206075250256, | |
| "grad_norm": 0.02099609375, | |
| "learning_rate": 1.32574679943101e-05, | |
| "loss": 0.0002, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.3603727994477045, | |
| "grad_norm": 0.70703125, | |
| "learning_rate": 1.322901849217639e-05, | |
| "loss": 0.0005, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.36175353814290645, | |
| "grad_norm": 0.02001953125, | |
| "learning_rate": 1.3200568990042674e-05, | |
| "loss": 0.0002, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.3631342768381084, | |
| "grad_norm": 0.021728515625, | |
| "learning_rate": 1.3172119487908963e-05, | |
| "loss": 0.0002, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.36451501553331034, | |
| "grad_norm": 0.0203857421875, | |
| "learning_rate": 1.314366998577525e-05, | |
| "loss": 0.0002, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.3658957542285122, | |
| "grad_norm": 0.03564453125, | |
| "learning_rate": 1.3115220483641538e-05, | |
| "loss": 0.0003, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.36727649292371417, | |
| "grad_norm": 0.020263671875, | |
| "learning_rate": 1.3086770981507825e-05, | |
| "loss": 0.0002, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.3686572316189161, | |
| "grad_norm": 0.0201416015625, | |
| "learning_rate": 1.3058321479374111e-05, | |
| "loss": 0.0002, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.37003797031411806, | |
| "grad_norm": 0.1396484375, | |
| "learning_rate": 1.30298719772404e-05, | |
| "loss": 0.0002, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.37141870900932, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 1.3001422475106686e-05, | |
| "loss": 0.0006, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.37279944770452195, | |
| "grad_norm": 0.02001953125, | |
| "learning_rate": 1.2972972972972975e-05, | |
| "loss": 0.0002, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.37418018639972384, | |
| "grad_norm": 0.0264892578125, | |
| "learning_rate": 1.2944523470839262e-05, | |
| "loss": 0.0002, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.3755609250949258, | |
| "grad_norm": 0.019775390625, | |
| "learning_rate": 1.291607396870555e-05, | |
| "loss": 0.0002, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.3769416637901277, | |
| "grad_norm": 0.019775390625, | |
| "learning_rate": 1.2887624466571835e-05, | |
| "loss": 0.0002, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.37832240248532967, | |
| "grad_norm": 0.019287109375, | |
| "learning_rate": 1.2859174964438123e-05, | |
| "loss": 0.0002, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.3797031411805316, | |
| "grad_norm": 0.02978515625, | |
| "learning_rate": 1.283072546230441e-05, | |
| "loss": 0.0002, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.3810838798757335, | |
| "grad_norm": 0.0189208984375, | |
| "learning_rate": 1.2802275960170699e-05, | |
| "loss": 0.0002, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.38246461857093544, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 1.2773826458036987e-05, | |
| "loss": 0.0008, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.3838453572661374, | |
| "grad_norm": 0.018798828125, | |
| "learning_rate": 1.2745376955903272e-05, | |
| "loss": 0.0002, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.38522609596133933, | |
| "grad_norm": 0.019775390625, | |
| "learning_rate": 1.271692745376956e-05, | |
| "loss": 0.0002, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.3866068346565413, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 1.2688477951635847e-05, | |
| "loss": 0.0031, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.38798757335174316, | |
| "grad_norm": 0.0224609375, | |
| "learning_rate": 1.2660028449502136e-05, | |
| "loss": 0.0002, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.3893683120469451, | |
| "grad_norm": 0.0194091796875, | |
| "learning_rate": 1.263157894736842e-05, | |
| "loss": 0.0002, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.39074905074214705, | |
| "grad_norm": 0.041259765625, | |
| "learning_rate": 1.2603129445234709e-05, | |
| "loss": 0.0002, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.392129789437349, | |
| "grad_norm": 0.0245361328125, | |
| "learning_rate": 1.2574679943100996e-05, | |
| "loss": 0.0003, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.39351052813255094, | |
| "grad_norm": 0.0186767578125, | |
| "learning_rate": 1.2546230440967284e-05, | |
| "loss": 0.0002, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.39489126682775283, | |
| "grad_norm": 0.0194091796875, | |
| "learning_rate": 1.2517780938833573e-05, | |
| "loss": 0.0002, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.3962720055229548, | |
| "grad_norm": 0.018798828125, | |
| "learning_rate": 1.248933143669986e-05, | |
| "loss": 0.0002, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.3976527442181567, | |
| "grad_norm": 0.02001953125, | |
| "learning_rate": 1.2460881934566146e-05, | |
| "loss": 0.0002, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.3976527442181567, | |
| "eval_loss": 0.00039300136268138885, | |
| "eval_runtime": 581.6065, | |
| "eval_samples_per_second": 2.215, | |
| "eval_steps_per_second": 2.215, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.39903348291335866, | |
| "grad_norm": 0.2041015625, | |
| "learning_rate": 1.2432432432432433e-05, | |
| "loss": 0.0005, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.4004142216085606, | |
| "grad_norm": 0.0208740234375, | |
| "learning_rate": 1.2403982930298721e-05, | |
| "loss": 0.0002, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.4017949603037625, | |
| "grad_norm": 0.09619140625, | |
| "learning_rate": 1.2375533428165008e-05, | |
| "loss": 0.0002, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.40317569899896444, | |
| "grad_norm": 0.15625, | |
| "learning_rate": 1.2347083926031296e-05, | |
| "loss": 0.0021, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.4045564376941664, | |
| "grad_norm": 0.0196533203125, | |
| "learning_rate": 1.2318634423897581e-05, | |
| "loss": 0.0002, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.4059371763893683, | |
| "grad_norm": 0.0189208984375, | |
| "learning_rate": 1.229018492176387e-05, | |
| "loss": 0.0002, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.40731791508457027, | |
| "grad_norm": 0.018310546875, | |
| "learning_rate": 1.2261735419630158e-05, | |
| "loss": 0.0002, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.40869865377977216, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 1.2233285917496445e-05, | |
| "loss": 0.0033, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.4100793924749741, | |
| "grad_norm": 0.068359375, | |
| "learning_rate": 1.2204836415362733e-05, | |
| "loss": 0.0003, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.41146013117017605, | |
| "grad_norm": 0.0196533203125, | |
| "learning_rate": 1.2176386913229019e-05, | |
| "loss": 0.0002, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.412840869865378, | |
| "grad_norm": 0.019287109375, | |
| "learning_rate": 1.2147937411095307e-05, | |
| "loss": 0.0002, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.41422160856057993, | |
| "grad_norm": 0.0223388671875, | |
| "learning_rate": 1.2119487908961594e-05, | |
| "loss": 0.0002, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.4156023472557818, | |
| "grad_norm": 0.0186767578125, | |
| "learning_rate": 1.2091038406827882e-05, | |
| "loss": 0.0002, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.41698308595098377, | |
| "grad_norm": 0.66796875, | |
| "learning_rate": 1.2062588904694169e-05, | |
| "loss": 0.0017, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.4183638246461857, | |
| "grad_norm": 0.018798828125, | |
| "learning_rate": 1.2034139402560456e-05, | |
| "loss": 0.0002, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.41974456334138766, | |
| "grad_norm": 0.0196533203125, | |
| "learning_rate": 1.2005689900426742e-05, | |
| "loss": 0.0002, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.4211253020365896, | |
| "grad_norm": 0.020751953125, | |
| "learning_rate": 1.197724039829303e-05, | |
| "loss": 0.0002, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.4225060407317915, | |
| "grad_norm": 0.03662109375, | |
| "learning_rate": 1.1948790896159319e-05, | |
| "loss": 0.0003, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.42388677942699343, | |
| "grad_norm": 0.0196533203125, | |
| "learning_rate": 1.1920341394025606e-05, | |
| "loss": 0.0002, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.4252675181221954, | |
| "grad_norm": 0.036376953125, | |
| "learning_rate": 1.1891891891891894e-05, | |
| "loss": 0.0002, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.4266482568173973, | |
| "grad_norm": 0.197265625, | |
| "learning_rate": 1.186344238975818e-05, | |
| "loss": 0.0002, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.42802899551259926, | |
| "grad_norm": 0.02197265625, | |
| "learning_rate": 1.1834992887624468e-05, | |
| "loss": 0.0002, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.42940973420780115, | |
| "grad_norm": 0.0198974609375, | |
| "learning_rate": 1.1806543385490754e-05, | |
| "loss": 0.0002, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.4307904729030031, | |
| "grad_norm": 0.02001953125, | |
| "learning_rate": 1.1778093883357043e-05, | |
| "loss": 0.0002, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.43217121159820504, | |
| "grad_norm": 0.02197265625, | |
| "learning_rate": 1.1749644381223328e-05, | |
| "loss": 0.0002, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.433551950293407, | |
| "grad_norm": 0.0208740234375, | |
| "learning_rate": 1.1721194879089616e-05, | |
| "loss": 0.0002, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.43493268898860893, | |
| "grad_norm": 0.05322265625, | |
| "learning_rate": 1.1692745376955905e-05, | |
| "loss": 0.0005, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.4363134276838108, | |
| "grad_norm": 0.0179443359375, | |
| "learning_rate": 1.1664295874822192e-05, | |
| "loss": 0.0002, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.43769416637901276, | |
| "grad_norm": 0.01904296875, | |
| "learning_rate": 1.163584637268848e-05, | |
| "loss": 0.0002, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.4390749050742147, | |
| "grad_norm": 0.0184326171875, | |
| "learning_rate": 1.1607396870554765e-05, | |
| "loss": 0.0002, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.44045564376941665, | |
| "grad_norm": 0.03125, | |
| "learning_rate": 1.1578947368421053e-05, | |
| "loss": 0.0003, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.4418363824646186, | |
| "grad_norm": 0.031005859375, | |
| "learning_rate": 1.155049786628734e-05, | |
| "loss": 0.0002, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.4432171211598205, | |
| "grad_norm": 0.0712890625, | |
| "learning_rate": 1.1522048364153629e-05, | |
| "loss": 0.0003, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.4445978598550224, | |
| "grad_norm": 0.024169921875, | |
| "learning_rate": 1.1493598862019915e-05, | |
| "loss": 0.0002, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.44597859855022437, | |
| "grad_norm": 0.019287109375, | |
| "learning_rate": 1.1465149359886204e-05, | |
| "loss": 0.0002, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.4473593372454263, | |
| "grad_norm": 0.040283203125, | |
| "learning_rate": 1.143669985775249e-05, | |
| "loss": 0.0003, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.44874007594062826, | |
| "grad_norm": 0.0179443359375, | |
| "learning_rate": 1.1408250355618777e-05, | |
| "loss": 0.0002, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.45012081463583015, | |
| "grad_norm": 0.016845703125, | |
| "learning_rate": 1.1379800853485066e-05, | |
| "loss": 0.0002, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.4515015533310321, | |
| "grad_norm": 0.0179443359375, | |
| "learning_rate": 1.1351351351351352e-05, | |
| "loss": 0.0002, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.45288229202623403, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 1.132290184921764e-05, | |
| "loss": 0.0007, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.454263030721436, | |
| "grad_norm": 0.0206298828125, | |
| "learning_rate": 1.1294452347083926e-05, | |
| "loss": 0.0002, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.4556437694166379, | |
| "grad_norm": 0.07421875, | |
| "learning_rate": 1.1266002844950214e-05, | |
| "loss": 0.0003, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.4570245081118398, | |
| "grad_norm": 0.0177001953125, | |
| "learning_rate": 1.1237553342816501e-05, | |
| "loss": 0.0002, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.45840524680704176, | |
| "grad_norm": 0.016845703125, | |
| "learning_rate": 1.120910384068279e-05, | |
| "loss": 0.0002, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.4597859855022437, | |
| "grad_norm": 0.032958984375, | |
| "learning_rate": 1.1180654338549078e-05, | |
| "loss": 0.0002, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.46116672419744564, | |
| "grad_norm": 0.2109375, | |
| "learning_rate": 1.1152204836415363e-05, | |
| "loss": 0.0002, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.4625474628926476, | |
| "grad_norm": 0.0181884765625, | |
| "learning_rate": 1.1123755334281651e-05, | |
| "loss": 0.0002, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.4639282015878495, | |
| "grad_norm": 0.01708984375, | |
| "learning_rate": 1.1095305832147938e-05, | |
| "loss": 0.0002, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.4653089402830514, | |
| "grad_norm": 0.0167236328125, | |
| "learning_rate": 1.1066856330014226e-05, | |
| "loss": 0.0002, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.46668967897825336, | |
| "grad_norm": 0.0255126953125, | |
| "learning_rate": 1.1038406827880513e-05, | |
| "loss": 0.0002, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.4680704176734553, | |
| "grad_norm": 0.019287109375, | |
| "learning_rate": 1.10099573257468e-05, | |
| "loss": 0.0002, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.46945115636865725, | |
| "grad_norm": 0.01708984375, | |
| "learning_rate": 1.0981507823613087e-05, | |
| "loss": 0.0002, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.47083189506385914, | |
| "grad_norm": 0.0172119140625, | |
| "learning_rate": 1.0953058321479375e-05, | |
| "loss": 0.0002, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.4722126337590611, | |
| "grad_norm": 0.04541015625, | |
| "learning_rate": 1.0924608819345663e-05, | |
| "loss": 0.0004, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.47359337245426303, | |
| "grad_norm": 0.038818359375, | |
| "learning_rate": 1.089615931721195e-05, | |
| "loss": 0.0003, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.474974111149465, | |
| "grad_norm": 0.059326171875, | |
| "learning_rate": 1.0867709815078239e-05, | |
| "loss": 0.0003, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.4763548498446669, | |
| "grad_norm": 0.017333984375, | |
| "learning_rate": 1.0839260312944524e-05, | |
| "loss": 0.0002, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.4777355885398688, | |
| "grad_norm": 0.02001953125, | |
| "learning_rate": 1.0810810810810812e-05, | |
| "loss": 0.0002, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.47911632723507075, | |
| "grad_norm": 0.0419921875, | |
| "learning_rate": 1.0782361308677099e-05, | |
| "loss": 0.0002, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.4804970659302727, | |
| "grad_norm": 0.021240234375, | |
| "learning_rate": 1.0753911806543387e-05, | |
| "loss": 0.0002, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.48187780462547464, | |
| "grad_norm": 0.044677734375, | |
| "learning_rate": 1.0725462304409672e-05, | |
| "loss": 0.0004, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.4832585433206766, | |
| "grad_norm": 0.0174560546875, | |
| "learning_rate": 1.069701280227596e-05, | |
| "loss": 0.0002, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.48463928201587847, | |
| "grad_norm": 0.01708984375, | |
| "learning_rate": 1.0668563300142247e-05, | |
| "loss": 0.0002, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.4860200207110804, | |
| "grad_norm": 0.016357421875, | |
| "learning_rate": 1.0640113798008536e-05, | |
| "loss": 0.0002, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.48740075940628236, | |
| "grad_norm": 0.02001953125, | |
| "learning_rate": 1.0611664295874824e-05, | |
| "loss": 0.0002, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.4887814981014843, | |
| "grad_norm": 0.0172119140625, | |
| "learning_rate": 1.058321479374111e-05, | |
| "loss": 0.0002, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.49016223679668625, | |
| "grad_norm": 0.0159912109375, | |
| "learning_rate": 1.0554765291607398e-05, | |
| "loss": 0.0002, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.49154297549188813, | |
| "grad_norm": 0.5, | |
| "learning_rate": 1.0526315789473684e-05, | |
| "loss": 0.0021, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.4929237141870901, | |
| "grad_norm": 0.01611328125, | |
| "learning_rate": 1.0497866287339973e-05, | |
| "loss": 0.0002, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.494304452882292, | |
| "grad_norm": 0.016357421875, | |
| "learning_rate": 1.046941678520626e-05, | |
| "loss": 0.0002, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.49568519157749397, | |
| "grad_norm": 0.016845703125, | |
| "learning_rate": 1.0440967283072548e-05, | |
| "loss": 0.0002, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.4970659302726959, | |
| "grad_norm": 0.0194091796875, | |
| "learning_rate": 1.0412517780938833e-05, | |
| "loss": 0.0002, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.4970659302726959, | |
| "eval_loss": 0.00029756984440609813, | |
| "eval_runtime": 582.415, | |
| "eval_samples_per_second": 2.211, | |
| "eval_steps_per_second": 2.211, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.4984466689678978, | |
| "grad_norm": 0.0169677734375, | |
| "learning_rate": 1.0384068278805121e-05, | |
| "loss": 0.0002, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.49982740766309974, | |
| "grad_norm": 0.020263671875, | |
| "learning_rate": 1.035561877667141e-05, | |
| "loss": 0.0002, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.5012081463583017, | |
| "grad_norm": 0.016357421875, | |
| "learning_rate": 1.0327169274537697e-05, | |
| "loss": 0.0002, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.5025888850535036, | |
| "grad_norm": 0.01953125, | |
| "learning_rate": 1.0298719772403985e-05, | |
| "loss": 0.0002, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.5039696237487056, | |
| "grad_norm": 0.0159912109375, | |
| "learning_rate": 1.027027027027027e-05, | |
| "loss": 0.0002, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.5053503624439075, | |
| "grad_norm": 0.0166015625, | |
| "learning_rate": 1.0241820768136559e-05, | |
| "loss": 0.0002, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.5067311011391095, | |
| "grad_norm": 0.04296875, | |
| "learning_rate": 1.0213371266002845e-05, | |
| "loss": 0.0002, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.5081118398343114, | |
| "grad_norm": 0.0673828125, | |
| "learning_rate": 1.0184921763869134e-05, | |
| "loss": 0.0003, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.5094925785295132, | |
| "grad_norm": 0.0164794921875, | |
| "learning_rate": 1.0156472261735419e-05, | |
| "loss": 0.0002, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.5108733172247152, | |
| "grad_norm": 0.0174560546875, | |
| "learning_rate": 1.0128022759601707e-05, | |
| "loss": 0.0002, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.5122540559199171, | |
| "grad_norm": 0.0224609375, | |
| "learning_rate": 1.0099573257467996e-05, | |
| "loss": 0.0002, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.5136347946151191, | |
| "grad_norm": 0.42578125, | |
| "learning_rate": 1.0071123755334282e-05, | |
| "loss": 0.0002, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.515015533310321, | |
| "grad_norm": 0.126953125, | |
| "learning_rate": 1.004267425320057e-05, | |
| "loss": 0.0014, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.516396272005523, | |
| "grad_norm": 0.015869140625, | |
| "learning_rate": 1.0014224751066857e-05, | |
| "loss": 0.0002, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.5177770107007249, | |
| "grad_norm": 0.1337890625, | |
| "learning_rate": 9.985775248933144e-06, | |
| "loss": 0.0014, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.5191577493959268, | |
| "grad_norm": 0.040771484375, | |
| "learning_rate": 9.957325746799433e-06, | |
| "loss": 0.0004, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.5205384880911288, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 9.92887624466572e-06, | |
| "loss": 0.0003, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.5219192267863307, | |
| "grad_norm": 0.01708984375, | |
| "learning_rate": 9.900426742532006e-06, | |
| "loss": 0.0002, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.5232999654815326, | |
| "grad_norm": 0.1884765625, | |
| "learning_rate": 9.871977240398294e-06, | |
| "loss": 0.0008, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.5246807041767345, | |
| "grad_norm": 0.016357421875, | |
| "learning_rate": 9.843527738264581e-06, | |
| "loss": 0.0002, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.5260614428719365, | |
| "grad_norm": 0.0162353515625, | |
| "learning_rate": 9.815078236130868e-06, | |
| "loss": 0.0002, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.5274421815671384, | |
| "grad_norm": 0.06591796875, | |
| "learning_rate": 9.786628733997155e-06, | |
| "loss": 0.0004, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.5288229202623403, | |
| "grad_norm": 0.015625, | |
| "learning_rate": 9.758179231863443e-06, | |
| "loss": 0.0001, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.5302036589575423, | |
| "grad_norm": 0.0230712890625, | |
| "learning_rate": 9.729729729729732e-06, | |
| "loss": 0.0002, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.5315843976527442, | |
| "grad_norm": 0.016357421875, | |
| "learning_rate": 9.701280227596018e-06, | |
| "loss": 0.0002, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.5329651363479462, | |
| "grad_norm": 0.01611328125, | |
| "learning_rate": 9.672830725462305e-06, | |
| "loss": 0.0002, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.5343458750431481, | |
| "grad_norm": 0.0223388671875, | |
| "learning_rate": 9.644381223328593e-06, | |
| "loss": 0.0002, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.5357266137383501, | |
| "grad_norm": 0.0299072265625, | |
| "learning_rate": 9.61593172119488e-06, | |
| "loss": 0.0002, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.5371073524335519, | |
| "grad_norm": 0.0174560546875, | |
| "learning_rate": 9.587482219061167e-06, | |
| "loss": 0.0002, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.5384880911287538, | |
| "grad_norm": 0.016845703125, | |
| "learning_rate": 9.559032716927454e-06, | |
| "loss": 0.0002, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.5398688298239558, | |
| "grad_norm": 0.63671875, | |
| "learning_rate": 9.530583214793742e-06, | |
| "loss": 0.002, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.5412495685191577, | |
| "grad_norm": 0.0225830078125, | |
| "learning_rate": 9.502133712660029e-06, | |
| "loss": 0.0002, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.5426303072143597, | |
| "grad_norm": 0.0296630859375, | |
| "learning_rate": 9.473684210526315e-06, | |
| "loss": 0.0002, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.5440110459095616, | |
| "grad_norm": 0.0771484375, | |
| "learning_rate": 9.445234708392604e-06, | |
| "loss": 0.0011, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.5453917846047636, | |
| "grad_norm": 0.0673828125, | |
| "learning_rate": 9.416785206258892e-06, | |
| "loss": 0.0005, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.5467725232999655, | |
| "grad_norm": 0.0213623046875, | |
| "learning_rate": 9.388335704125179e-06, | |
| "loss": 0.0002, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.5481532619951675, | |
| "grad_norm": 0.039306640625, | |
| "learning_rate": 9.359886201991466e-06, | |
| "loss": 0.0002, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.5495340006903694, | |
| "grad_norm": 0.0162353515625, | |
| "learning_rate": 9.331436699857753e-06, | |
| "loss": 0.0002, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.5509147393855712, | |
| "grad_norm": 0.01519775390625, | |
| "learning_rate": 9.302987197724041e-06, | |
| "loss": 0.0001, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.5522954780807732, | |
| "grad_norm": 0.0224609375, | |
| "learning_rate": 9.274537695590328e-06, | |
| "loss": 0.0002, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.5536762167759751, | |
| "grad_norm": 0.033935546875, | |
| "learning_rate": 9.246088193456614e-06, | |
| "loss": 0.0002, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.5550569554711771, | |
| "grad_norm": 0.019287109375, | |
| "learning_rate": 9.217638691322903e-06, | |
| "loss": 0.0002, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.556437694166379, | |
| "grad_norm": 0.11962890625, | |
| "learning_rate": 9.189189189189191e-06, | |
| "loss": 0.0007, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.557818432861581, | |
| "grad_norm": 0.0179443359375, | |
| "learning_rate": 9.160739687055478e-06, | |
| "loss": 0.0001, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.5591991715567829, | |
| "grad_norm": 0.0400390625, | |
| "learning_rate": 9.132290184921765e-06, | |
| "loss": 0.0002, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.5605799102519848, | |
| "grad_norm": 0.61328125, | |
| "learning_rate": 9.103840682788051e-06, | |
| "loss": 0.0011, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.5619606489471868, | |
| "grad_norm": 0.0174560546875, | |
| "learning_rate": 9.07539118065434e-06, | |
| "loss": 0.0002, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.5633413876423887, | |
| "grad_norm": 0.061767578125, | |
| "learning_rate": 9.046941678520627e-06, | |
| "loss": 0.0003, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.5647221263375906, | |
| "grad_norm": 0.01708984375, | |
| "learning_rate": 9.018492176386913e-06, | |
| "loss": 0.0002, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.5661028650327925, | |
| "grad_norm": 0.1650390625, | |
| "learning_rate": 8.990042674253202e-06, | |
| "loss": 0.0011, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.5674836037279944, | |
| "grad_norm": 0.01519775390625, | |
| "learning_rate": 8.961593172119488e-06, | |
| "loss": 0.0001, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.5688643424231964, | |
| "grad_norm": 0.0194091796875, | |
| "learning_rate": 8.933143669985777e-06, | |
| "loss": 0.0002, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.5702450811183983, | |
| "grad_norm": 0.0206298828125, | |
| "learning_rate": 8.904694167852064e-06, | |
| "loss": 0.0001, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.5716258198136003, | |
| "grad_norm": 0.01483154296875, | |
| "learning_rate": 8.87624466571835e-06, | |
| "loss": 0.0001, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.5730065585088022, | |
| "grad_norm": 0.018798828125, | |
| "learning_rate": 8.847795163584639e-06, | |
| "loss": 0.0002, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.5743872972040042, | |
| "grad_norm": 0.040771484375, | |
| "learning_rate": 8.819345661450926e-06, | |
| "loss": 0.0002, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.5757680358992061, | |
| "grad_norm": 0.015380859375, | |
| "learning_rate": 8.790896159317212e-06, | |
| "loss": 0.0001, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.577148774594408, | |
| "grad_norm": 0.01495361328125, | |
| "learning_rate": 8.7624466571835e-06, | |
| "loss": 0.0001, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.5785295132896099, | |
| "grad_norm": 0.0152587890625, | |
| "learning_rate": 8.733997155049787e-06, | |
| "loss": 0.0001, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.5799102519848118, | |
| "grad_norm": 0.03515625, | |
| "learning_rate": 8.705547652916074e-06, | |
| "loss": 0.0002, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.5812909906800138, | |
| "grad_norm": 0.0167236328125, | |
| "learning_rate": 8.677098150782361e-06, | |
| "loss": 0.0002, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.5826717293752157, | |
| "grad_norm": 0.0263671875, | |
| "learning_rate": 8.64864864864865e-06, | |
| "loss": 0.0002, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.5840524680704177, | |
| "grad_norm": 0.015625, | |
| "learning_rate": 8.620199146514938e-06, | |
| "loss": 0.0001, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.5854332067656196, | |
| "grad_norm": 0.0198974609375, | |
| "learning_rate": 8.591749644381224e-06, | |
| "loss": 0.0002, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.5868139454608216, | |
| "grad_norm": 0.0225830078125, | |
| "learning_rate": 8.563300142247511e-06, | |
| "loss": 0.0002, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.5881946841560235, | |
| "grad_norm": 0.03125, | |
| "learning_rate": 8.534850640113798e-06, | |
| "loss": 0.0002, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.5895754228512254, | |
| "grad_norm": 0.01470947265625, | |
| "learning_rate": 8.506401137980086e-06, | |
| "loss": 0.0001, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.5909561615464274, | |
| "grad_norm": 0.02392578125, | |
| "learning_rate": 8.477951635846373e-06, | |
| "loss": 0.0002, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.5923369002416292, | |
| "grad_norm": 0.01531982421875, | |
| "learning_rate": 8.44950213371266e-06, | |
| "loss": 0.0001, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.5937176389368312, | |
| "grad_norm": 0.01446533203125, | |
| "learning_rate": 8.421052631578948e-06, | |
| "loss": 0.0001, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.5950983776320331, | |
| "grad_norm": 0.053466796875, | |
| "learning_rate": 8.392603129445237e-06, | |
| "loss": 0.0002, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.596479116327235, | |
| "grad_norm": 0.017578125, | |
| "learning_rate": 8.364153627311523e-06, | |
| "loss": 0.0001, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.596479116327235, | |
| "eval_loss": 0.0002615667472127825, | |
| "eval_runtime": 581.6024, | |
| "eval_samples_per_second": 2.215, | |
| "eval_steps_per_second": 2.215, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.597859855022437, | |
| "grad_norm": 0.01483154296875, | |
| "learning_rate": 8.33570412517781e-06, | |
| "loss": 0.0001, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.5992405937176389, | |
| "grad_norm": 0.0211181640625, | |
| "learning_rate": 8.307254623044097e-06, | |
| "loss": 0.0002, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.6006213324128409, | |
| "grad_norm": 0.01953125, | |
| "learning_rate": 8.278805120910385e-06, | |
| "loss": 0.0002, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.6020020711080428, | |
| "grad_norm": 0.0157470703125, | |
| "learning_rate": 8.250355618776672e-06, | |
| "loss": 0.0001, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.6033828098032448, | |
| "grad_norm": 0.08447265625, | |
| "learning_rate": 8.221906116642959e-06, | |
| "loss": 0.0005, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.6047635484984467, | |
| "grad_norm": 0.01470947265625, | |
| "learning_rate": 8.193456614509247e-06, | |
| "loss": 0.0001, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.6061442871936487, | |
| "grad_norm": 0.015869140625, | |
| "learning_rate": 8.165007112375534e-06, | |
| "loss": 0.0002, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.6075250258888505, | |
| "grad_norm": 0.0162353515625, | |
| "learning_rate": 8.13655761024182e-06, | |
| "loss": 0.0001, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.6089057645840524, | |
| "grad_norm": 0.0223388671875, | |
| "learning_rate": 8.108108108108109e-06, | |
| "loss": 0.0002, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.6102865032792544, | |
| "grad_norm": 0.062255859375, | |
| "learning_rate": 8.079658605974396e-06, | |
| "loss": 0.0005, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.6116672419744563, | |
| "grad_norm": 0.020263671875, | |
| "learning_rate": 8.051209103840684e-06, | |
| "loss": 0.0001, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.6130479806696583, | |
| "grad_norm": 0.01544189453125, | |
| "learning_rate": 8.022759601706971e-06, | |
| "loss": 0.0001, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.6144287193648602, | |
| "grad_norm": 0.6953125, | |
| "learning_rate": 7.994310099573258e-06, | |
| "loss": 0.0031, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.6158094580600622, | |
| "grad_norm": 0.0262451171875, | |
| "learning_rate": 7.965860597439546e-06, | |
| "loss": 0.0002, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.6171901967552641, | |
| "grad_norm": 0.060791015625, | |
| "learning_rate": 7.937411095305833e-06, | |
| "loss": 0.0003, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.618570935450466, | |
| "grad_norm": 0.0172119140625, | |
| "learning_rate": 7.90896159317212e-06, | |
| "loss": 0.0002, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.619951674145668, | |
| "grad_norm": 0.0146484375, | |
| "learning_rate": 7.880512091038406e-06, | |
| "loss": 0.0001, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 0.6213324128408698, | |
| "grad_norm": 0.014892578125, | |
| "learning_rate": 7.852062588904695e-06, | |
| "loss": 0.0001, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.6227131515360718, | |
| "grad_norm": 0.0145263671875, | |
| "learning_rate": 7.823613086770983e-06, | |
| "loss": 0.0001, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 0.6240938902312737, | |
| "grad_norm": 0.014892578125, | |
| "learning_rate": 7.79516358463727e-06, | |
| "loss": 0.0001, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 0.6254746289264757, | |
| "grad_norm": 0.017333984375, | |
| "learning_rate": 7.766714082503557e-06, | |
| "loss": 0.0001, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 0.6268553676216776, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 7.738264580369845e-06, | |
| "loss": 0.002, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 0.6282361063168795, | |
| "grad_norm": 0.73828125, | |
| "learning_rate": 7.709815078236132e-06, | |
| "loss": 0.0019, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.6296168450120815, | |
| "grad_norm": 0.01531982421875, | |
| "learning_rate": 7.681365576102418e-06, | |
| "loss": 0.0001, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.6309975837072834, | |
| "grad_norm": 0.015625, | |
| "learning_rate": 7.652916073968705e-06, | |
| "loss": 0.0001, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 0.6323783224024854, | |
| "grad_norm": 0.01531982421875, | |
| "learning_rate": 7.624466571834993e-06, | |
| "loss": 0.0001, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 0.6337590610976873, | |
| "grad_norm": 0.014404296875, | |
| "learning_rate": 7.596017069701281e-06, | |
| "loss": 0.0001, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 0.6351397997928891, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 7.567567567567569e-06, | |
| "loss": 0.0005, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.6365205384880911, | |
| "grad_norm": 0.0191650390625, | |
| "learning_rate": 7.5391180654338555e-06, | |
| "loss": 0.0002, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 0.637901277183293, | |
| "grad_norm": 0.018798828125, | |
| "learning_rate": 7.510668563300143e-06, | |
| "loss": 0.0002, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.639282015878495, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 7.482219061166431e-06, | |
| "loss": 0.0017, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 0.6406627545736969, | |
| "grad_norm": 0.013916015625, | |
| "learning_rate": 7.453769559032717e-06, | |
| "loss": 0.0001, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 0.6420434932688989, | |
| "grad_norm": 0.01806640625, | |
| "learning_rate": 7.425320056899005e-06, | |
| "loss": 0.0001, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.6434242319641008, | |
| "grad_norm": 0.013916015625, | |
| "learning_rate": 7.396870554765292e-06, | |
| "loss": 0.0001, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 0.6448049706593028, | |
| "grad_norm": 0.06201171875, | |
| "learning_rate": 7.368421052631579e-06, | |
| "loss": 0.0004, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 0.6461857093545047, | |
| "grad_norm": 0.0179443359375, | |
| "learning_rate": 7.339971550497866e-06, | |
| "loss": 0.0001, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 0.6475664480497066, | |
| "grad_norm": 0.0225830078125, | |
| "learning_rate": 7.3115220483641544e-06, | |
| "loss": 0.0002, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 0.6489471867449085, | |
| "grad_norm": 0.0242919921875, | |
| "learning_rate": 7.283072546230442e-06, | |
| "loss": 0.0002, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.6503279254401104, | |
| "grad_norm": 0.0242919921875, | |
| "learning_rate": 7.254623044096729e-06, | |
| "loss": 0.0002, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 0.6517086641353124, | |
| "grad_norm": 0.01446533203125, | |
| "learning_rate": 7.226173541963016e-06, | |
| "loss": 0.0001, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 0.6530894028305143, | |
| "grad_norm": 0.039306640625, | |
| "learning_rate": 7.197724039829304e-06, | |
| "loss": 0.0003, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 0.6544701415257163, | |
| "grad_norm": 0.1201171875, | |
| "learning_rate": 7.169274537695591e-06, | |
| "loss": 0.0002, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 0.6558508802209182, | |
| "grad_norm": 0.01483154296875, | |
| "learning_rate": 7.140825035561878e-06, | |
| "loss": 0.0001, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.6572316189161201, | |
| "grad_norm": 0.035888671875, | |
| "learning_rate": 7.112375533428165e-06, | |
| "loss": 0.0002, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 0.6586123576113221, | |
| "grad_norm": 0.08447265625, | |
| "learning_rate": 7.0839260312944525e-06, | |
| "loss": 0.0004, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 0.659993096306524, | |
| "grad_norm": 0.01556396484375, | |
| "learning_rate": 7.055476529160741e-06, | |
| "loss": 0.0001, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 0.661373835001726, | |
| "grad_norm": 0.0240478515625, | |
| "learning_rate": 7.027027027027028e-06, | |
| "loss": 0.0002, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 0.6627545736969278, | |
| "grad_norm": 0.0181884765625, | |
| "learning_rate": 6.998577524893315e-06, | |
| "loss": 0.0001, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.6641353123921298, | |
| "grad_norm": 0.0145263671875, | |
| "learning_rate": 6.970128022759603e-06, | |
| "loss": 0.0001, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 0.6655160510873317, | |
| "grad_norm": 0.029541015625, | |
| "learning_rate": 6.9416785206258896e-06, | |
| "loss": 0.0002, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 0.6668967897825336, | |
| "grad_norm": 0.04736328125, | |
| "learning_rate": 6.913229018492177e-06, | |
| "loss": 0.0003, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 0.6682775284777356, | |
| "grad_norm": 0.0242919921875, | |
| "learning_rate": 6.884779516358464e-06, | |
| "loss": 0.0002, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 0.6696582671729375, | |
| "grad_norm": 0.01611328125, | |
| "learning_rate": 6.8563300142247514e-06, | |
| "loss": 0.0002, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.6710390058681395, | |
| "grad_norm": 0.11474609375, | |
| "learning_rate": 6.827880512091038e-06, | |
| "loss": 0.001, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 0.6724197445633414, | |
| "grad_norm": 0.0230712890625, | |
| "learning_rate": 6.799431009957327e-06, | |
| "loss": 0.0002, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 0.6738004832585434, | |
| "grad_norm": 0.11083984375, | |
| "learning_rate": 6.770981507823614e-06, | |
| "loss": 0.0005, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 0.6751812219537453, | |
| "grad_norm": 0.0184326171875, | |
| "learning_rate": 6.742532005689901e-06, | |
| "loss": 0.0001, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 0.6765619606489471, | |
| "grad_norm": 0.01409912109375, | |
| "learning_rate": 6.7140825035561885e-06, | |
| "loss": 0.0001, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.6779426993441491, | |
| "grad_norm": 0.01531982421875, | |
| "learning_rate": 6.685633001422476e-06, | |
| "loss": 0.0001, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 0.679323438039351, | |
| "grad_norm": 0.055908203125, | |
| "learning_rate": 6.657183499288763e-06, | |
| "loss": 0.0004, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 0.680704176734553, | |
| "grad_norm": 0.0152587890625, | |
| "learning_rate": 6.62873399715505e-06, | |
| "loss": 0.0001, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 0.6820849154297549, | |
| "grad_norm": 0.01458740234375, | |
| "learning_rate": 6.600284495021337e-06, | |
| "loss": 0.0001, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 0.6834656541249569, | |
| "grad_norm": 0.0244140625, | |
| "learning_rate": 6.571834992887625e-06, | |
| "loss": 0.0002, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.6848463928201588, | |
| "grad_norm": 0.1591796875, | |
| "learning_rate": 6.543385490753912e-06, | |
| "loss": 0.0003, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 0.6862271315153607, | |
| "grad_norm": 0.01397705078125, | |
| "learning_rate": 6.5149359886202e-06, | |
| "loss": 0.0001, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 0.6876078702105627, | |
| "grad_norm": 0.0142822265625, | |
| "learning_rate": 6.486486486486487e-06, | |
| "loss": 0.0001, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 0.6889886089057646, | |
| "grad_norm": 0.013916015625, | |
| "learning_rate": 6.458036984352775e-06, | |
| "loss": 0.0001, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 0.6903693476009665, | |
| "grad_norm": 0.07470703125, | |
| "learning_rate": 6.429587482219062e-06, | |
| "loss": 0.0002, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.6917500862961684, | |
| "grad_norm": 0.018798828125, | |
| "learning_rate": 6.401137980085349e-06, | |
| "loss": 0.0002, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 0.6931308249913704, | |
| "grad_norm": 0.0206298828125, | |
| "learning_rate": 6.372688477951636e-06, | |
| "loss": 0.0002, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 0.6945115636865723, | |
| "grad_norm": 0.01544189453125, | |
| "learning_rate": 6.344238975817924e-06, | |
| "loss": 0.0001, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 0.6958923023817742, | |
| "grad_norm": 0.0137939453125, | |
| "learning_rate": 6.31578947368421e-06, | |
| "loss": 0.0001, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 0.6958923023817742, | |
| "eval_loss": 0.00025013022241182625, | |
| "eval_runtime": 580.5157, | |
| "eval_samples_per_second": 2.219, | |
| "eval_steps_per_second": 2.219, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 0.6972730410769762, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 6.287339971550498e-06, | |
| "loss": 0.0005, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.6986537797721781, | |
| "grad_norm": 0.01806640625, | |
| "learning_rate": 6.258890469416786e-06, | |
| "loss": 0.0001, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 0.7000345184673801, | |
| "grad_norm": 0.0169677734375, | |
| "learning_rate": 6.230440967283073e-06, | |
| "loss": 0.0001, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 0.701415257162582, | |
| "grad_norm": 0.014892578125, | |
| "learning_rate": 6.201991465149361e-06, | |
| "loss": 0.0001, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 0.702795995857784, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 6.173541963015648e-06, | |
| "loss": 0.0009, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 0.7041767345529858, | |
| "grad_norm": 0.01373291015625, | |
| "learning_rate": 6.145092460881935e-06, | |
| "loss": 0.0001, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.7055574732481877, | |
| "grad_norm": 0.0164794921875, | |
| "learning_rate": 6.1166429587482225e-06, | |
| "loss": 0.0001, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 0.7069382119433897, | |
| "grad_norm": 0.016357421875, | |
| "learning_rate": 6.088193456614509e-06, | |
| "loss": 0.0001, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 0.7083189506385916, | |
| "grad_norm": 0.0181884765625, | |
| "learning_rate": 6.059743954480797e-06, | |
| "loss": 0.0001, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 0.7096996893337936, | |
| "grad_norm": 0.0230712890625, | |
| "learning_rate": 6.031294452347084e-06, | |
| "loss": 0.0002, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 0.7110804280289955, | |
| "grad_norm": 0.0146484375, | |
| "learning_rate": 6.002844950213371e-06, | |
| "loss": 0.0001, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.7124611667241975, | |
| "grad_norm": 0.11328125, | |
| "learning_rate": 5.9743954480796596e-06, | |
| "loss": 0.0007, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 0.7138419054193994, | |
| "grad_norm": 0.02099609375, | |
| "learning_rate": 5.945945945945947e-06, | |
| "loss": 0.0001, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 0.7152226441146013, | |
| "grad_norm": 0.015625, | |
| "learning_rate": 5.917496443812234e-06, | |
| "loss": 0.0001, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 0.7166033828098033, | |
| "grad_norm": 0.01422119140625, | |
| "learning_rate": 5.8890469416785214e-06, | |
| "loss": 0.0001, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 0.7179841215050051, | |
| "grad_norm": 0.01361083984375, | |
| "learning_rate": 5.860597439544808e-06, | |
| "loss": 0.0001, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.7193648602002071, | |
| "grad_norm": 0.013916015625, | |
| "learning_rate": 5.832147937411096e-06, | |
| "loss": 0.0001, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 0.720745598895409, | |
| "grad_norm": 0.044677734375, | |
| "learning_rate": 5.8036984352773825e-06, | |
| "loss": 0.0002, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 0.722126337590611, | |
| "grad_norm": 0.01611328125, | |
| "learning_rate": 5.77524893314367e-06, | |
| "loss": 0.0001, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 0.7235070762858129, | |
| "grad_norm": 0.01806640625, | |
| "learning_rate": 5.746799431009958e-06, | |
| "loss": 0.0001, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 0.7248878149810148, | |
| "grad_norm": 0.0137939453125, | |
| "learning_rate": 5.718349928876245e-06, | |
| "loss": 0.0001, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.7262685536762168, | |
| "grad_norm": 0.07861328125, | |
| "learning_rate": 5.689900426742533e-06, | |
| "loss": 0.0002, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 0.7276492923714187, | |
| "grad_norm": 0.12060546875, | |
| "learning_rate": 5.66145092460882e-06, | |
| "loss": 0.0005, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 0.7290300310666207, | |
| "grad_norm": 0.0137939453125, | |
| "learning_rate": 5.633001422475107e-06, | |
| "loss": 0.0001, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 0.7304107697618226, | |
| "grad_norm": 0.0162353515625, | |
| "learning_rate": 5.604551920341395e-06, | |
| "loss": 0.0001, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 0.7317915084570245, | |
| "grad_norm": 0.017333984375, | |
| "learning_rate": 5.576102418207681e-06, | |
| "loss": 0.0002, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.7331722471522264, | |
| "grad_norm": 0.0458984375, | |
| "learning_rate": 5.547652916073969e-06, | |
| "loss": 0.0003, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 0.7345529858474283, | |
| "grad_norm": 0.01373291015625, | |
| "learning_rate": 5.5192034139402566e-06, | |
| "loss": 0.0001, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 0.7359337245426303, | |
| "grad_norm": 0.01397705078125, | |
| "learning_rate": 5.490753911806543e-06, | |
| "loss": 0.0001, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 0.7373144632378322, | |
| "grad_norm": 0.01361083984375, | |
| "learning_rate": 5.462304409672832e-06, | |
| "loss": 0.0001, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 0.7386952019330342, | |
| "grad_norm": 0.0634765625, | |
| "learning_rate": 5.433854907539119e-06, | |
| "loss": 0.0003, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.7400759406282361, | |
| "grad_norm": 0.12353515625, | |
| "learning_rate": 5.405405405405406e-06, | |
| "loss": 0.0012, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 0.7414566793234381, | |
| "grad_norm": 0.0157470703125, | |
| "learning_rate": 5.376955903271694e-06, | |
| "loss": 0.0001, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 0.74283741801864, | |
| "grad_norm": 0.01416015625, | |
| "learning_rate": 5.34850640113798e-06, | |
| "loss": 0.0001, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 0.744218156713842, | |
| "grad_norm": 0.11376953125, | |
| "learning_rate": 5.320056899004268e-06, | |
| "loss": 0.0003, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 0.7455988954090439, | |
| "grad_norm": 0.01373291015625, | |
| "learning_rate": 5.291607396870555e-06, | |
| "loss": 0.0001, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.7469796341042457, | |
| "grad_norm": 0.08984375, | |
| "learning_rate": 5.263157894736842e-06, | |
| "loss": 0.0003, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 0.7483603727994477, | |
| "grad_norm": 0.0137939453125, | |
| "learning_rate": 5.23470839260313e-06, | |
| "loss": 0.0001, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 0.7497411114946496, | |
| "grad_norm": 0.048583984375, | |
| "learning_rate": 5.2062588904694165e-06, | |
| "loss": 0.0004, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 0.7511218501898516, | |
| "grad_norm": 0.0135498046875, | |
| "learning_rate": 5.177809388335705e-06, | |
| "loss": 0.0001, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 0.7525025888850535, | |
| "grad_norm": 0.01397705078125, | |
| "learning_rate": 5.1493598862019925e-06, | |
| "loss": 0.0001, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.7538833275802554, | |
| "grad_norm": 0.01446533203125, | |
| "learning_rate": 5.120910384068279e-06, | |
| "loss": 0.0001, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 0.7552640662754574, | |
| "grad_norm": 0.040283203125, | |
| "learning_rate": 5.092460881934567e-06, | |
| "loss": 0.0002, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 0.7566448049706593, | |
| "grad_norm": 0.01422119140625, | |
| "learning_rate": 5.0640113798008536e-06, | |
| "loss": 0.0001, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 0.7580255436658613, | |
| "grad_norm": 0.01361083984375, | |
| "learning_rate": 5.035561877667141e-06, | |
| "loss": 0.0001, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 0.7594062823610632, | |
| "grad_norm": 0.013916015625, | |
| "learning_rate": 5.007112375533429e-06, | |
| "loss": 0.0001, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.7607870210562651, | |
| "grad_norm": 0.03857421875, | |
| "learning_rate": 4.978662873399716e-06, | |
| "loss": 0.0002, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 0.762167759751467, | |
| "grad_norm": 0.013671875, | |
| "learning_rate": 4.950213371266003e-06, | |
| "loss": 0.0001, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 0.7635484984466689, | |
| "grad_norm": 0.0205078125, | |
| "learning_rate": 4.921763869132291e-06, | |
| "loss": 0.0001, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 0.7649292371418709, | |
| "grad_norm": 0.0145263671875, | |
| "learning_rate": 4.893314366998577e-06, | |
| "loss": 0.0001, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 0.7663099758370728, | |
| "grad_norm": 0.01336669921875, | |
| "learning_rate": 4.864864864864866e-06, | |
| "loss": 0.0001, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.7676907145322748, | |
| "grad_norm": 0.01336669921875, | |
| "learning_rate": 4.8364153627311525e-06, | |
| "loss": 0.0001, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 0.7690714532274767, | |
| "grad_norm": 0.01336669921875, | |
| "learning_rate": 4.80796586059744e-06, | |
| "loss": 0.0001, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 0.7704521919226787, | |
| "grad_norm": 0.01324462890625, | |
| "learning_rate": 4.779516358463727e-06, | |
| "loss": 0.0001, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 0.7718329306178806, | |
| "grad_norm": 0.01361083984375, | |
| "learning_rate": 4.751066856330014e-06, | |
| "loss": 0.0001, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 0.7732136693130826, | |
| "grad_norm": 0.01348876953125, | |
| "learning_rate": 4.722617354196302e-06, | |
| "loss": 0.0001, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.7745944080082844, | |
| "grad_norm": 0.0142822265625, | |
| "learning_rate": 4.6941678520625895e-06, | |
| "loss": 0.0001, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 0.7759751467034863, | |
| "grad_norm": 0.119140625, | |
| "learning_rate": 4.665718349928876e-06, | |
| "loss": 0.0007, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 0.7773558853986883, | |
| "grad_norm": 0.0223388671875, | |
| "learning_rate": 4.637268847795164e-06, | |
| "loss": 0.0002, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 0.7787366240938902, | |
| "grad_norm": 0.01318359375, | |
| "learning_rate": 4.608819345661451e-06, | |
| "loss": 0.0001, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 0.7801173627890922, | |
| "grad_norm": 0.013671875, | |
| "learning_rate": 4.580369843527739e-06, | |
| "loss": 0.0001, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.7814981014842941, | |
| "grad_norm": 0.013427734375, | |
| "learning_rate": 4.551920341394026e-06, | |
| "loss": 0.0001, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 0.782878840179496, | |
| "grad_norm": 0.0174560546875, | |
| "learning_rate": 4.523470839260313e-06, | |
| "loss": 0.0001, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 0.784259578874698, | |
| "grad_norm": 0.0203857421875, | |
| "learning_rate": 4.495021337126601e-06, | |
| "loss": 0.0001, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 0.7856403175698999, | |
| "grad_norm": 0.013427734375, | |
| "learning_rate": 4.4665718349928885e-06, | |
| "loss": 0.0001, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 0.7870210562651019, | |
| "grad_norm": 0.01318359375, | |
| "learning_rate": 4.438122332859175e-06, | |
| "loss": 0.0001, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.7884017949603037, | |
| "grad_norm": 0.162109375, | |
| "learning_rate": 4.409672830725463e-06, | |
| "loss": 0.0005, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 0.7897825336555057, | |
| "grad_norm": 0.0133056640625, | |
| "learning_rate": 4.38122332859175e-06, | |
| "loss": 0.0001, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 0.7911632723507076, | |
| "grad_norm": 0.013427734375, | |
| "learning_rate": 4.352773826458037e-06, | |
| "loss": 0.0001, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 0.7925440110459095, | |
| "grad_norm": 0.053466796875, | |
| "learning_rate": 4.324324324324325e-06, | |
| "loss": 0.0003, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 0.7939247497411115, | |
| "grad_norm": 0.01373291015625, | |
| "learning_rate": 4.295874822190612e-06, | |
| "loss": 0.0001, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.7953054884363134, | |
| "grad_norm": 0.0224609375, | |
| "learning_rate": 4.267425320056899e-06, | |
| "loss": 0.0002, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 0.7953054884363134, | |
| "eval_loss": 0.0002442169061396271, | |
| "eval_runtime": 582.9379, | |
| "eval_samples_per_second": 2.209, | |
| "eval_steps_per_second": 2.209, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 0.7966862271315154, | |
| "grad_norm": 0.01361083984375, | |
| "learning_rate": 4.2389758179231865e-06, | |
| "loss": 0.0001, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 0.7980669658267173, | |
| "grad_norm": 0.01953125, | |
| "learning_rate": 4.210526315789474e-06, | |
| "loss": 0.0002, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 0.7994477045219193, | |
| "grad_norm": 0.0201416015625, | |
| "learning_rate": 4.182076813655762e-06, | |
| "loss": 0.0002, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 0.8008284432171212, | |
| "grad_norm": 0.01336669921875, | |
| "learning_rate": 4.1536273115220484e-06, | |
| "loss": 0.0001, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.802209181912323, | |
| "grad_norm": 0.0130615234375, | |
| "learning_rate": 4.125177809388336e-06, | |
| "loss": 0.0001, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 0.803589920607525, | |
| "grad_norm": 0.0133056640625, | |
| "learning_rate": 4.096728307254624e-06, | |
| "loss": 0.0001, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 0.8049706593027269, | |
| "grad_norm": 0.01373291015625, | |
| "learning_rate": 4.06827880512091e-06, | |
| "loss": 0.0001, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 0.8063513979979289, | |
| "grad_norm": 0.03173828125, | |
| "learning_rate": 4.039829302987198e-06, | |
| "loss": 0.0002, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 0.8077321366931308, | |
| "grad_norm": 0.212890625, | |
| "learning_rate": 4.0113798008534855e-06, | |
| "loss": 0.001, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.8091128753883328, | |
| "grad_norm": 0.0439453125, | |
| "learning_rate": 3.982930298719773e-06, | |
| "loss": 0.0004, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 0.8104936140835347, | |
| "grad_norm": 0.04443359375, | |
| "learning_rate": 3.95448079658606e-06, | |
| "loss": 0.0003, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 0.8118743527787367, | |
| "grad_norm": 0.078125, | |
| "learning_rate": 3.926031294452347e-06, | |
| "loss": 0.0004, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 0.8132550914739386, | |
| "grad_norm": 0.0233154296875, | |
| "learning_rate": 3.897581792318635e-06, | |
| "loss": 0.0002, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 0.8146358301691405, | |
| "grad_norm": 0.01416015625, | |
| "learning_rate": 3.8691322901849225e-06, | |
| "loss": 0.0001, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.8160165688643424, | |
| "grad_norm": 0.044677734375, | |
| "learning_rate": 3.840682788051209e-06, | |
| "loss": 0.0002, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 0.8173973075595443, | |
| "grad_norm": 0.17578125, | |
| "learning_rate": 3.8122332859174964e-06, | |
| "loss": 0.0004, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 0.8187780462547463, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 3.7837837837837844e-06, | |
| "loss": 0.0003, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 0.8201587849499482, | |
| "grad_norm": 0.01416015625, | |
| "learning_rate": 3.7553342816500715e-06, | |
| "loss": 0.0001, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 0.8215395236451501, | |
| "grad_norm": 0.0137939453125, | |
| "learning_rate": 3.7268847795163587e-06, | |
| "loss": 0.0001, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.8229202623403521, | |
| "grad_norm": 0.01373291015625, | |
| "learning_rate": 3.698435277382646e-06, | |
| "loss": 0.0001, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 0.824301001035554, | |
| "grad_norm": 0.01385498046875, | |
| "learning_rate": 3.669985775248933e-06, | |
| "loss": 0.0001, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 0.825681739730756, | |
| "grad_norm": 0.0133056640625, | |
| "learning_rate": 3.641536273115221e-06, | |
| "loss": 0.0001, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 0.8270624784259579, | |
| "grad_norm": 0.04248046875, | |
| "learning_rate": 3.613086770981508e-06, | |
| "loss": 0.0003, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 0.8284432171211599, | |
| "grad_norm": 0.126953125, | |
| "learning_rate": 3.5846372688477953e-06, | |
| "loss": 0.0013, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.8298239558163617, | |
| "grad_norm": 0.0145263671875, | |
| "learning_rate": 3.5561877667140825e-06, | |
| "loss": 0.0001, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 0.8312046945115636, | |
| "grad_norm": 0.043212890625, | |
| "learning_rate": 3.5277382645803705e-06, | |
| "loss": 0.0004, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 0.8325854332067656, | |
| "grad_norm": 0.02099609375, | |
| "learning_rate": 3.4992887624466576e-06, | |
| "loss": 0.0001, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 0.8339661719019675, | |
| "grad_norm": 0.55859375, | |
| "learning_rate": 3.4708392603129448e-06, | |
| "loss": 0.0004, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 0.8353469105971695, | |
| "grad_norm": 0.045166015625, | |
| "learning_rate": 3.442389758179232e-06, | |
| "loss": 0.0004, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.8367276492923714, | |
| "grad_norm": 0.080078125, | |
| "learning_rate": 3.413940256045519e-06, | |
| "loss": 0.0002, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 0.8381083879875734, | |
| "grad_norm": 0.01416015625, | |
| "learning_rate": 3.385490753911807e-06, | |
| "loss": 0.0001, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 0.8394891266827753, | |
| "grad_norm": 0.01336669921875, | |
| "learning_rate": 3.3570412517780942e-06, | |
| "loss": 0.0001, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 0.8408698653779773, | |
| "grad_norm": 0.048828125, | |
| "learning_rate": 3.3285917496443814e-06, | |
| "loss": 0.0002, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 0.8422506040731792, | |
| "grad_norm": 0.016357421875, | |
| "learning_rate": 3.3001422475106685e-06, | |
| "loss": 0.0001, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.843631342768381, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 3.271692745376956e-06, | |
| "loss": 0.0008, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 0.845012081463583, | |
| "grad_norm": 0.01336669921875, | |
| "learning_rate": 3.2432432432432437e-06, | |
| "loss": 0.0001, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 0.8463928201587849, | |
| "grad_norm": 0.01336669921875, | |
| "learning_rate": 3.214793741109531e-06, | |
| "loss": 0.0001, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 0.8477735588539869, | |
| "grad_norm": 0.072265625, | |
| "learning_rate": 3.186344238975818e-06, | |
| "loss": 0.0002, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 0.8491542975491888, | |
| "grad_norm": 0.0133056640625, | |
| "learning_rate": 3.157894736842105e-06, | |
| "loss": 0.0001, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.8505350362443908, | |
| "grad_norm": 0.01397705078125, | |
| "learning_rate": 3.129445234708393e-06, | |
| "loss": 0.0001, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 0.8519157749395927, | |
| "grad_norm": 0.05615234375, | |
| "learning_rate": 3.1009957325746803e-06, | |
| "loss": 0.0002, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 0.8532965136347946, | |
| "grad_norm": 0.0257568359375, | |
| "learning_rate": 3.0725462304409675e-06, | |
| "loss": 0.0001, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 0.8546772523299966, | |
| "grad_norm": 0.21875, | |
| "learning_rate": 3.0440967283072546e-06, | |
| "loss": 0.0014, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 0.8560579910251985, | |
| "grad_norm": 0.013671875, | |
| "learning_rate": 3.015647226173542e-06, | |
| "loss": 0.0001, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.8574387297204004, | |
| "grad_norm": 0.08203125, | |
| "learning_rate": 2.9871977240398298e-06, | |
| "loss": 0.0003, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 0.8588194684156023, | |
| "grad_norm": 0.0135498046875, | |
| "learning_rate": 2.958748221906117e-06, | |
| "loss": 0.0001, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 0.8602002071108042, | |
| "grad_norm": 0.053955078125, | |
| "learning_rate": 2.930298719772404e-06, | |
| "loss": 0.0003, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 0.8615809458060062, | |
| "grad_norm": 0.01348876953125, | |
| "learning_rate": 2.9018492176386912e-06, | |
| "loss": 0.0001, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 0.8629616845012081, | |
| "grad_norm": 0.01385498046875, | |
| "learning_rate": 2.873399715504979e-06, | |
| "loss": 0.0001, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.8643424231964101, | |
| "grad_norm": 0.01348876953125, | |
| "learning_rate": 2.8449502133712664e-06, | |
| "loss": 0.0001, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 0.865723161891612, | |
| "grad_norm": 0.01409912109375, | |
| "learning_rate": 2.8165007112375536e-06, | |
| "loss": 0.0001, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 0.867103900586814, | |
| "grad_norm": 0.018798828125, | |
| "learning_rate": 2.7880512091038407e-06, | |
| "loss": 0.0001, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 0.8684846392820159, | |
| "grad_norm": 0.01397705078125, | |
| "learning_rate": 2.7596017069701283e-06, | |
| "loss": 0.0001, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 0.8698653779772179, | |
| "grad_norm": 0.01348876953125, | |
| "learning_rate": 2.731152204836416e-06, | |
| "loss": 0.0001, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.8712461166724197, | |
| "grad_norm": 0.0133056640625, | |
| "learning_rate": 2.702702702702703e-06, | |
| "loss": 0.0001, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 0.8726268553676216, | |
| "grad_norm": 0.012939453125, | |
| "learning_rate": 2.67425320056899e-06, | |
| "loss": 0.0001, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 0.8740075940628236, | |
| "grad_norm": 0.0135498046875, | |
| "learning_rate": 2.6458036984352773e-06, | |
| "loss": 0.0001, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 0.8753883327580255, | |
| "grad_norm": 0.01300048828125, | |
| "learning_rate": 2.617354196301565e-06, | |
| "loss": 0.0001, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 0.8767690714532275, | |
| "grad_norm": 0.16015625, | |
| "learning_rate": 2.5889046941678525e-06, | |
| "loss": 0.0004, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.8781498101484294, | |
| "grad_norm": 0.01336669921875, | |
| "learning_rate": 2.5604551920341396e-06, | |
| "loss": 0.0001, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 0.8795305488436314, | |
| "grad_norm": 0.091796875, | |
| "learning_rate": 2.5320056899004268e-06, | |
| "loss": 0.0006, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 0.8809112875388333, | |
| "grad_norm": 0.0189208984375, | |
| "learning_rate": 2.5035561877667144e-06, | |
| "loss": 0.0002, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 0.8822920262340352, | |
| "grad_norm": 0.06494140625, | |
| "learning_rate": 2.4751066856330015e-06, | |
| "loss": 0.0003, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 0.8836727649292372, | |
| "grad_norm": 0.01324462890625, | |
| "learning_rate": 2.4466571834992887e-06, | |
| "loss": 0.0001, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.885053503624439, | |
| "grad_norm": 0.19921875, | |
| "learning_rate": 2.4182076813655762e-06, | |
| "loss": 0.0013, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 0.886434242319641, | |
| "grad_norm": 0.027099609375, | |
| "learning_rate": 2.3897581792318634e-06, | |
| "loss": 0.0002, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 0.8878149810148429, | |
| "grad_norm": 0.1884765625, | |
| "learning_rate": 2.361308677098151e-06, | |
| "loss": 0.0003, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 0.8891957197100449, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 2.332859174964438e-06, | |
| "loss": 0.0014, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 0.8905764584052468, | |
| "grad_norm": 0.0732421875, | |
| "learning_rate": 2.3044096728307257e-06, | |
| "loss": 0.0002, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.8919571971004487, | |
| "grad_norm": 0.01361083984375, | |
| "learning_rate": 2.275960170697013e-06, | |
| "loss": 0.0001, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 0.8933379357956507, | |
| "grad_norm": 0.01318359375, | |
| "learning_rate": 2.2475106685633004e-06, | |
| "loss": 0.0001, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 0.8947186744908526, | |
| "grad_norm": 0.10302734375, | |
| "learning_rate": 2.2190611664295876e-06, | |
| "loss": 0.0005, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 0.8947186744908526, | |
| "eval_loss": 0.00023728572705294937, | |
| "eval_runtime": 582.6042, | |
| "eval_samples_per_second": 2.211, | |
| "eval_steps_per_second": 2.211, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 0.8960994131860546, | |
| "grad_norm": 0.0244140625, | |
| "learning_rate": 2.190611664295875e-06, | |
| "loss": 0.0002, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 0.8974801518812565, | |
| "grad_norm": 0.0218505859375, | |
| "learning_rate": 2.1621621621621623e-06, | |
| "loss": 0.0002, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.8988608905764585, | |
| "grad_norm": 0.04736328125, | |
| "learning_rate": 2.1337126600284495e-06, | |
| "loss": 0.0004, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 0.9002416292716603, | |
| "grad_norm": 0.0137939453125, | |
| "learning_rate": 2.105263157894737e-06, | |
| "loss": 0.0001, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 0.9016223679668622, | |
| "grad_norm": 0.0196533203125, | |
| "learning_rate": 2.0768136557610242e-06, | |
| "loss": 0.0002, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 0.9030031066620642, | |
| "grad_norm": 0.0133056640625, | |
| "learning_rate": 2.048364153627312e-06, | |
| "loss": 0.0001, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 0.9043838453572661, | |
| "grad_norm": 0.01422119140625, | |
| "learning_rate": 2.019914651493599e-06, | |
| "loss": 0.0001, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.9057645840524681, | |
| "grad_norm": 0.1435546875, | |
| "learning_rate": 1.9914651493598865e-06, | |
| "loss": 0.0007, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 0.90714532274767, | |
| "grad_norm": 0.01531982421875, | |
| "learning_rate": 1.9630156472261737e-06, | |
| "loss": 0.0001, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 0.908526061442872, | |
| "grad_norm": 0.07470703125, | |
| "learning_rate": 1.9345661450924613e-06, | |
| "loss": 0.0003, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 0.9099068001380739, | |
| "grad_norm": 0.023681640625, | |
| "learning_rate": 1.9061166429587482e-06, | |
| "loss": 0.0002, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 0.9112875388332758, | |
| "grad_norm": 0.052001953125, | |
| "learning_rate": 1.8776671408250358e-06, | |
| "loss": 0.0002, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.9126682775284778, | |
| "grad_norm": 0.045654296875, | |
| "learning_rate": 1.849217638691323e-06, | |
| "loss": 0.0003, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 0.9140490162236796, | |
| "grad_norm": 0.1142578125, | |
| "learning_rate": 1.8207681365576105e-06, | |
| "loss": 0.0004, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 0.9154297549188816, | |
| "grad_norm": 0.01324462890625, | |
| "learning_rate": 1.7923186344238977e-06, | |
| "loss": 0.0001, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 0.9168104936140835, | |
| "grad_norm": 0.01348876953125, | |
| "learning_rate": 1.7638691322901852e-06, | |
| "loss": 0.0001, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 0.9181912323092855, | |
| "grad_norm": 0.0133056640625, | |
| "learning_rate": 1.7354196301564724e-06, | |
| "loss": 0.0001, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.9195719710044874, | |
| "grad_norm": 0.01544189453125, | |
| "learning_rate": 1.7069701280227595e-06, | |
| "loss": 0.0001, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 0.9209527096996893, | |
| "grad_norm": 0.01507568359375, | |
| "learning_rate": 1.6785206258890471e-06, | |
| "loss": 0.0001, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 0.9223334483948913, | |
| "grad_norm": 0.0135498046875, | |
| "learning_rate": 1.6500711237553343e-06, | |
| "loss": 0.0001, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 0.9237141870900932, | |
| "grad_norm": 0.0142822265625, | |
| "learning_rate": 1.6216216216216219e-06, | |
| "loss": 0.0001, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 0.9250949257852952, | |
| "grad_norm": 0.126953125, | |
| "learning_rate": 1.593172119487909e-06, | |
| "loss": 0.0012, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.9264756644804971, | |
| "grad_norm": 0.126953125, | |
| "learning_rate": 1.5647226173541966e-06, | |
| "loss": 0.0003, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 0.927856403175699, | |
| "grad_norm": 0.01312255859375, | |
| "learning_rate": 1.5362731152204837e-06, | |
| "loss": 0.0001, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 0.9292371418709009, | |
| "grad_norm": 0.037109375, | |
| "learning_rate": 1.507823613086771e-06, | |
| "loss": 0.0002, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 0.9306178805661028, | |
| "grad_norm": 0.0225830078125, | |
| "learning_rate": 1.4793741109530585e-06, | |
| "loss": 0.0001, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 0.9319986192613048, | |
| "grad_norm": 0.01324462890625, | |
| "learning_rate": 1.4509246088193456e-06, | |
| "loss": 0.0001, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.9333793579565067, | |
| "grad_norm": 0.01416015625, | |
| "learning_rate": 1.4224751066856332e-06, | |
| "loss": 0.0001, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 0.9347600966517087, | |
| "grad_norm": 0.01611328125, | |
| "learning_rate": 1.3940256045519204e-06, | |
| "loss": 0.0001, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 0.9361408353469106, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 1.365576102418208e-06, | |
| "loss": 0.0012, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 0.9375215740421126, | |
| "grad_norm": 0.0157470703125, | |
| "learning_rate": 1.337126600284495e-06, | |
| "loss": 0.0001, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 0.9389023127373145, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 1.3086770981507825e-06, | |
| "loss": 0.0017, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.9402830514325164, | |
| "grad_norm": 0.130859375, | |
| "learning_rate": 1.2802275960170698e-06, | |
| "loss": 0.0004, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 0.9416637901277183, | |
| "grad_norm": 0.0166015625, | |
| "learning_rate": 1.2517780938833572e-06, | |
| "loss": 0.0001, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 0.9430445288229202, | |
| "grad_norm": 0.01458740234375, | |
| "learning_rate": 1.2233285917496443e-06, | |
| "loss": 0.0001, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 0.9444252675181222, | |
| "grad_norm": 0.01373291015625, | |
| "learning_rate": 1.1948790896159317e-06, | |
| "loss": 0.0001, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 0.9458060062133241, | |
| "grad_norm": 0.01336669921875, | |
| "learning_rate": 1.166429587482219e-06, | |
| "loss": 0.0001, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.9471867449085261, | |
| "grad_norm": 0.0213623046875, | |
| "learning_rate": 1.1379800853485064e-06, | |
| "loss": 0.0001, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 0.948567483603728, | |
| "grad_norm": 0.038818359375, | |
| "learning_rate": 1.1095305832147938e-06, | |
| "loss": 0.0002, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 0.94994822229893, | |
| "grad_norm": 0.0157470703125, | |
| "learning_rate": 1.0810810810810812e-06, | |
| "loss": 0.0001, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 0.9513289609941319, | |
| "grad_norm": 0.01373291015625, | |
| "learning_rate": 1.0526315789473685e-06, | |
| "loss": 0.0001, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 0.9527096996893338, | |
| "grad_norm": 0.0137939453125, | |
| "learning_rate": 1.024182076813656e-06, | |
| "loss": 0.0001, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.9540904383845358, | |
| "grad_norm": 0.05078125, | |
| "learning_rate": 9.957325746799433e-07, | |
| "loss": 0.0002, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 0.9554711770797376, | |
| "grad_norm": 0.015869140625, | |
| "learning_rate": 9.672830725462306e-07, | |
| "loss": 0.0002, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 0.9568519157749396, | |
| "grad_norm": 0.042236328125, | |
| "learning_rate": 9.388335704125179e-07, | |
| "loss": 0.0002, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 0.9582326544701415, | |
| "grad_norm": 0.01287841796875, | |
| "learning_rate": 9.103840682788053e-07, | |
| "loss": 0.0001, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 0.9596133931653434, | |
| "grad_norm": 0.01422119140625, | |
| "learning_rate": 8.819345661450926e-07, | |
| "loss": 0.0001, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.9609941318605454, | |
| "grad_norm": 0.01513671875, | |
| "learning_rate": 8.534850640113798e-07, | |
| "loss": 0.0001, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 0.9623748705557473, | |
| "grad_norm": 0.0771484375, | |
| "learning_rate": 8.250355618776671e-07, | |
| "loss": 0.0003, | |
| "step": 697 | |
| }, | |
| { | |
| "epoch": 0.9637556092509493, | |
| "grad_norm": 0.01300048828125, | |
| "learning_rate": 7.965860597439545e-07, | |
| "loss": 0.0001, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 0.9651363479461512, | |
| "grad_norm": 0.01324462890625, | |
| "learning_rate": 7.681365576102419e-07, | |
| "loss": 0.0001, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 0.9665170866413532, | |
| "grad_norm": 0.01336669921875, | |
| "learning_rate": 7.396870554765292e-07, | |
| "loss": 0.0001, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.9678978253365551, | |
| "grad_norm": 0.04736328125, | |
| "learning_rate": 7.112375533428166e-07, | |
| "loss": 0.0002, | |
| "step": 701 | |
| }, | |
| { | |
| "epoch": 0.9692785640317569, | |
| "grad_norm": 0.020751953125, | |
| "learning_rate": 6.82788051209104e-07, | |
| "loss": 0.0002, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 0.9706593027269589, | |
| "grad_norm": 0.037841796875, | |
| "learning_rate": 6.543385490753912e-07, | |
| "loss": 0.0001, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 0.9720400414221608, | |
| "grad_norm": 0.01483154296875, | |
| "learning_rate": 6.258890469416786e-07, | |
| "loss": 0.0001, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 0.9734207801173628, | |
| "grad_norm": 0.0208740234375, | |
| "learning_rate": 5.974395448079659e-07, | |
| "loss": 0.0001, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.9748015188125647, | |
| "grad_norm": 0.01300048828125, | |
| "learning_rate": 5.689900426742532e-07, | |
| "loss": 0.0001, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 0.9761822575077667, | |
| "grad_norm": 0.01409912109375, | |
| "learning_rate": 5.405405405405406e-07, | |
| "loss": 0.0001, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 0.9775629962029686, | |
| "grad_norm": 0.0137939453125, | |
| "learning_rate": 5.12091038406828e-07, | |
| "loss": 0.0001, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 0.9789437348981705, | |
| "grad_norm": 0.044921875, | |
| "learning_rate": 4.836415362731153e-07, | |
| "loss": 0.0002, | |
| "step": 709 | |
| }, | |
| { | |
| "epoch": 0.9803244735933725, | |
| "grad_norm": 0.1357421875, | |
| "learning_rate": 4.551920341394026e-07, | |
| "loss": 0.0003, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.9817052122885744, | |
| "grad_norm": 0.01422119140625, | |
| "learning_rate": 4.267425320056899e-07, | |
| "loss": 0.0001, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 0.9830859509837763, | |
| "grad_norm": 0.040283203125, | |
| "learning_rate": 3.9829302987197725e-07, | |
| "loss": 0.0003, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 0.9844666896789782, | |
| "grad_norm": 0.02734375, | |
| "learning_rate": 3.698435277382646e-07, | |
| "loss": 0.0002, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 0.9858474283741802, | |
| "grad_norm": 0.01422119140625, | |
| "learning_rate": 3.41394025604552e-07, | |
| "loss": 0.0001, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 0.9872281670693821, | |
| "grad_norm": 0.091796875, | |
| "learning_rate": 3.129445234708393e-07, | |
| "loss": 0.0006, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.988608905764584, | |
| "grad_norm": 0.01611328125, | |
| "learning_rate": 2.844950213371266e-07, | |
| "loss": 0.0001, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 0.989989644459786, | |
| "grad_norm": 0.054443359375, | |
| "learning_rate": 2.56045519203414e-07, | |
| "loss": 0.0003, | |
| "step": 717 | |
| }, | |
| { | |
| "epoch": 0.9913703831549879, | |
| "grad_norm": 0.014892578125, | |
| "learning_rate": 2.275960170697013e-07, | |
| "loss": 0.0001, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 0.9927511218501899, | |
| "grad_norm": 0.0137939453125, | |
| "learning_rate": 1.9914651493598863e-07, | |
| "loss": 0.0001, | |
| "step": 719 | |
| }, | |
| { | |
| "epoch": 0.9941318605453918, | |
| "grad_norm": 0.01361083984375, | |
| "learning_rate": 1.70697012802276e-07, | |
| "loss": 0.0001, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.9941318605453918, | |
| "eval_loss": 0.00024354300694540143, | |
| "eval_runtime": 582.0573, | |
| "eval_samples_per_second": 2.213, | |
| "eval_steps_per_second": 2.213, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.9955125992405938, | |
| "grad_norm": 0.013671875, | |
| "learning_rate": 1.422475106685633e-07, | |
| "loss": 0.0001, | |
| "step": 721 | |
| }, | |
| { | |
| "epoch": 0.9968933379357956, | |
| "grad_norm": 0.0133056640625, | |
| "learning_rate": 1.1379800853485066e-07, | |
| "loss": 0.0001, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 0.9982740766309975, | |
| "grad_norm": 0.01318359375, | |
| "learning_rate": 8.5348506401138e-08, | |
| "loss": 0.0001, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 0.9996548153261995, | |
| "grad_norm": 0.0211181640625, | |
| "learning_rate": 5.689900426742533e-08, | |
| "loss": 0.0001, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.003204345703125, | |
| "learning_rate": 2.8449502133712664e-08, | |
| "loss": 0.0, | |
| "step": 725 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 725, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 72, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3.298593420780503e+17, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |