| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 7.0, | |
| "eval_steps": 500, | |
| "global_step": 231, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.030303030303030304, | |
| "grad_norm": 7.664393449160768, | |
| "learning_rate": 8.333333333333333e-07, | |
| "loss": 1.1182, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.06060606060606061, | |
| "grad_norm": 6.8397192039989365, | |
| "learning_rate": 1.6666666666666667e-06, | |
| "loss": 1.08, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.09090909090909091, | |
| "grad_norm": 6.9847483234923455, | |
| "learning_rate": 2.5e-06, | |
| "loss": 1.0667, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.12121212121212122, | |
| "grad_norm": 6.369759762749378, | |
| "learning_rate": 3.3333333333333333e-06, | |
| "loss": 1.1114, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.15151515151515152, | |
| "grad_norm": 5.523424082900236, | |
| "learning_rate": 4.166666666666667e-06, | |
| "loss": 1.098, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.18181818181818182, | |
| "grad_norm": 3.0376044926238, | |
| "learning_rate": 5e-06, | |
| "loss": 0.9639, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.21212121212121213, | |
| "grad_norm": 2.586825356379709, | |
| "learning_rate": 5.833333333333334e-06, | |
| "loss": 0.9554, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.24242424242424243, | |
| "grad_norm": 3.6367518268416537, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 0.9294, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.2727272727272727, | |
| "grad_norm": 3.697168705151131, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.9616, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.30303030303030304, | |
| "grad_norm": 3.7564848723722313, | |
| "learning_rate": 8.333333333333334e-06, | |
| "loss": 0.8989, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.3333333333333333, | |
| "grad_norm": 3.459713399168143, | |
| "learning_rate": 9.166666666666666e-06, | |
| "loss": 0.8653, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 3.088988952445496, | |
| "learning_rate": 1e-05, | |
| "loss": 0.8643, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.3939393939393939, | |
| "grad_norm": 2.58433403944635, | |
| "learning_rate": 1.0833333333333334e-05, | |
| "loss": 0.8659, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.42424242424242425, | |
| "grad_norm": 2.336425108200707, | |
| "learning_rate": 1.1666666666666668e-05, | |
| "loss": 0.8524, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.45454545454545453, | |
| "grad_norm": 2.6510425625906238, | |
| "learning_rate": 1.25e-05, | |
| "loss": 0.8494, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.48484848484848486, | |
| "grad_norm": 2.694382062720332, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 0.8016, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.5151515151515151, | |
| "grad_norm": 1.9687607197834662, | |
| "learning_rate": 1.416666666666667e-05, | |
| "loss": 0.759, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.5454545454545454, | |
| "grad_norm": 2.033487518841393, | |
| "learning_rate": 1.5000000000000002e-05, | |
| "loss": 0.7688, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.5757575757575758, | |
| "grad_norm": 1.9021423014168446, | |
| "learning_rate": 1.5833333333333333e-05, | |
| "loss": 0.7663, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.6060606060606061, | |
| "grad_norm": 1.3654012966074662, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 0.7822, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.6363636363636364, | |
| "grad_norm": 1.377840889866603, | |
| "learning_rate": 1.7500000000000002e-05, | |
| "loss": 0.7701, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 1.1868797945575211, | |
| "learning_rate": 1.8333333333333333e-05, | |
| "loss": 0.7546, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.696969696969697, | |
| "grad_norm": 1.1600210957594432, | |
| "learning_rate": 1.916666666666667e-05, | |
| "loss": 0.751, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 1.1113263950821217, | |
| "learning_rate": 2e-05, | |
| "loss": 0.7522, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.7575757575757576, | |
| "grad_norm": 0.9755768173886403, | |
| "learning_rate": 1.999884834944106e-05, | |
| "loss": 0.7728, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.7878787878787878, | |
| "grad_norm": 1.0491773298145084, | |
| "learning_rate": 1.9995393663024054e-05, | |
| "loss": 0.7668, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.8181818181818182, | |
| "grad_norm": 0.8916975646778106, | |
| "learning_rate": 1.9989636736467278e-05, | |
| "loss": 0.7392, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.8484848484848485, | |
| "grad_norm": 0.8481134167269097, | |
| "learning_rate": 1.9981578895764272e-05, | |
| "loss": 0.7944, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.8787878787878788, | |
| "grad_norm": 0.9111917695542912, | |
| "learning_rate": 1.9971221996878395e-05, | |
| "loss": 0.7244, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "grad_norm": 0.8184437024373057, | |
| "learning_rate": 1.9958568425315316e-05, | |
| "loss": 0.7645, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.9393939393939394, | |
| "grad_norm": 0.802838678644513, | |
| "learning_rate": 1.9943621095573588e-05, | |
| "loss": 0.7673, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.9696969696969697, | |
| "grad_norm": 0.8006470216904298, | |
| "learning_rate": 1.9926383450473344e-05, | |
| "loss": 0.7154, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.837577801370185, | |
| "learning_rate": 1.9906859460363307e-05, | |
| "loss": 0.744, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 1.0303030303030303, | |
| "grad_norm": 0.8788290619470983, | |
| "learning_rate": 1.9885053622206305e-05, | |
| "loss": 0.6342, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 1.0606060606060606, | |
| "grad_norm": 0.7680508761273809, | |
| "learning_rate": 1.986097095854347e-05, | |
| "loss": 0.589, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.0909090909090908, | |
| "grad_norm": 0.6965668959919553, | |
| "learning_rate": 1.9834617016337424e-05, | |
| "loss": 0.598, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.121212121212121, | |
| "grad_norm": 0.7480432201738144, | |
| "learning_rate": 1.9805997865694616e-05, | |
| "loss": 0.6216, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 1.1515151515151516, | |
| "grad_norm": 0.742483855360199, | |
| "learning_rate": 1.9775120098467212e-05, | |
| "loss": 0.5776, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 1.1818181818181819, | |
| "grad_norm": 0.7986004967815837, | |
| "learning_rate": 1.9741990826734793e-05, | |
| "loss": 0.5696, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 1.2121212121212122, | |
| "grad_norm": 0.7664130855503278, | |
| "learning_rate": 1.970661768116622e-05, | |
| "loss": 0.609, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.2424242424242424, | |
| "grad_norm": 0.843867335966922, | |
| "learning_rate": 1.9669008809262064e-05, | |
| "loss": 0.5789, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 1.2727272727272727, | |
| "grad_norm": 0.7559987969483992, | |
| "learning_rate": 1.9629172873477995e-05, | |
| "loss": 0.5888, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.303030303030303, | |
| "grad_norm": 0.7825069327775015, | |
| "learning_rate": 1.9587119049229558e-05, | |
| "loss": 0.573, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 0.764435652973946, | |
| "learning_rate": 1.954285702277879e-05, | |
| "loss": 0.6057, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 1.3636363636363638, | |
| "grad_norm": 0.8416247176458859, | |
| "learning_rate": 1.9496396989003195e-05, | |
| "loss": 0.5763, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.393939393939394, | |
| "grad_norm": 0.8411450442338159, | |
| "learning_rate": 1.944774964904754e-05, | |
| "loss": 0.5724, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.4242424242424243, | |
| "grad_norm": 58.85300762641492, | |
| "learning_rate": 1.9396926207859085e-05, | |
| "loss": 0.6045, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.4545454545454546, | |
| "grad_norm": 0.989513487513671, | |
| "learning_rate": 1.9343938371606714e-05, | |
| "loss": 0.5425, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.4848484848484849, | |
| "grad_norm": 0.7660144458659521, | |
| "learning_rate": 1.9288798344984673e-05, | |
| "loss": 0.5815, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.5151515151515151, | |
| "grad_norm": 0.7996308078769786, | |
| "learning_rate": 1.9231518828401458e-05, | |
| "loss": 0.5741, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.5454545454545454, | |
| "grad_norm": 0.7421870124694202, | |
| "learning_rate": 1.917211301505453e-05, | |
| "loss": 0.5497, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 1.5757575757575757, | |
| "grad_norm": 0.7174748006849765, | |
| "learning_rate": 1.911059458789152e-05, | |
| "loss": 0.5881, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 1.606060606060606, | |
| "grad_norm": 0.7218164041816885, | |
| "learning_rate": 1.9046977716458627e-05, | |
| "loss": 0.5714, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 1.6363636363636362, | |
| "grad_norm": 0.7219084796232362, | |
| "learning_rate": 1.8981277053636963e-05, | |
| "loss": 0.5762, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 1.6666666666666665, | |
| "grad_norm": 0.6669987521807462, | |
| "learning_rate": 1.891350773226754e-05, | |
| "loss": 0.5246, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.696969696969697, | |
| "grad_norm": 0.7887668785965309, | |
| "learning_rate": 1.8843685361665724e-05, | |
| "loss": 0.585, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.7272727272727273, | |
| "grad_norm": 0.8068988167466868, | |
| "learning_rate": 1.8771826024025944e-05, | |
| "loss": 0.5755, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 1.7575757575757576, | |
| "grad_norm": 0.7977045147813666, | |
| "learning_rate": 1.8697946270717468e-05, | |
| "loss": 0.5949, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 1.7878787878787878, | |
| "grad_norm": 0.8046590769844791, | |
| "learning_rate": 1.8622063118472135e-05, | |
| "loss": 0.5686, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 1.8181818181818183, | |
| "grad_norm": 0.7228155425388395, | |
| "learning_rate": 1.8544194045464888e-05, | |
| "loss": 0.5751, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.8484848484848486, | |
| "grad_norm": 0.7343783239370106, | |
| "learning_rate": 1.8464356987288012e-05, | |
| "loss": 0.5288, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 1.878787878787879, | |
| "grad_norm": 0.7603317442003767, | |
| "learning_rate": 1.8382570332820045e-05, | |
| "loss": 0.5329, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.9090909090909092, | |
| "grad_norm": 0.7402203807135053, | |
| "learning_rate": 1.8298852919990254e-05, | |
| "loss": 0.585, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.9393939393939394, | |
| "grad_norm": 0.7448765178073259, | |
| "learning_rate": 1.821322403143969e-05, | |
| "loss": 0.5313, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.9696969696969697, | |
| "grad_norm": 0.8414943725942962, | |
| "learning_rate": 1.812570339007983e-05, | |
| "loss": 0.5628, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.6613338086533583, | |
| "learning_rate": 1.8036311154549783e-05, | |
| "loss": 0.549, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 2.0303030303030303, | |
| "grad_norm": 0.9910169384083473, | |
| "learning_rate": 1.7945067914573147e-05, | |
| "loss": 0.4305, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 2.0606060606060606, | |
| "grad_norm": 0.8198864215264449, | |
| "learning_rate": 1.7851994686215592e-05, | |
| "loss": 0.4309, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 2.090909090909091, | |
| "grad_norm": 0.8334247414220182, | |
| "learning_rate": 1.77571129070442e-05, | |
| "loss": 0.3827, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 2.121212121212121, | |
| "grad_norm": 0.6647544802335993, | |
| "learning_rate": 1.766044443118978e-05, | |
| "loss": 0.4468, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.1515151515151514, | |
| "grad_norm": 0.7755056316754402, | |
| "learning_rate": 1.7562011524313187e-05, | |
| "loss": 0.418, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 2.1818181818181817, | |
| "grad_norm": 0.7361621818580593, | |
| "learning_rate": 1.7461836858476858e-05, | |
| "loss": 0.4142, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 2.212121212121212, | |
| "grad_norm": 0.7283076300282773, | |
| "learning_rate": 1.7359943506922775e-05, | |
| "loss": 0.4257, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 2.242424242424242, | |
| "grad_norm": 0.6589685200571774, | |
| "learning_rate": 1.725635493875799e-05, | |
| "loss": 0.3462, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 2.2727272727272725, | |
| "grad_norm": 0.6310713769774094, | |
| "learning_rate": 1.7151095013548996e-05, | |
| "loss": 0.3726, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 2.303030303030303, | |
| "grad_norm": 0.6321052067014651, | |
| "learning_rate": 1.7044187975826126e-05, | |
| "loss": 0.4045, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 2.3333333333333335, | |
| "grad_norm": 0.5974858261461334, | |
| "learning_rate": 1.693565844949933e-05, | |
| "loss": 0.3586, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 2.3636363636363638, | |
| "grad_norm": 0.7547514178459003, | |
| "learning_rate": 1.6825531432186545e-05, | |
| "loss": 0.4155, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 2.393939393939394, | |
| "grad_norm": 0.6731605329468637, | |
| "learning_rate": 1.671383228945597e-05, | |
| "loss": 0.397, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 2.4242424242424243, | |
| "grad_norm": 0.6077791209818072, | |
| "learning_rate": 1.6600586748983642e-05, | |
| "loss": 0.3694, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 2.4545454545454546, | |
| "grad_norm": 0.6686590129987983, | |
| "learning_rate": 1.648582089462756e-05, | |
| "loss": 0.4216, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 2.484848484848485, | |
| "grad_norm": 0.642133639309677, | |
| "learning_rate": 1.6369561160419783e-05, | |
| "loss": 0.3635, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 2.515151515151515, | |
| "grad_norm": 0.7043055854005518, | |
| "learning_rate": 1.625183432447789e-05, | |
| "loss": 0.4006, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 2.5454545454545454, | |
| "grad_norm": 0.6929964167577031, | |
| "learning_rate": 1.6132667502837164e-05, | |
| "loss": 0.3794, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 2.5757575757575757, | |
| "grad_norm": 0.7477023331840109, | |
| "learning_rate": 1.6012088143204953e-05, | |
| "loss": 0.4149, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 2.606060606060606, | |
| "grad_norm": 0.6053848977683766, | |
| "learning_rate": 1.589012401863864e-05, | |
| "loss": 0.3685, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 2.6363636363636362, | |
| "grad_norm": 0.6987805424721913, | |
| "learning_rate": 1.5766803221148676e-05, | |
| "loss": 0.3774, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 2.6666666666666665, | |
| "grad_norm": 0.6522340734117946, | |
| "learning_rate": 1.5642154155228124e-05, | |
| "loss": 0.3713, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 2.6969696969696972, | |
| "grad_norm": 0.6243591268202322, | |
| "learning_rate": 1.5516205531310272e-05, | |
| "loss": 0.3843, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 2.7272727272727275, | |
| "grad_norm": 0.6729764888407868, | |
| "learning_rate": 1.538898635915576e-05, | |
| "loss": 0.4195, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 2.757575757575758, | |
| "grad_norm": 0.6367288137397891, | |
| "learning_rate": 1.526052594117071e-05, | |
| "loss": 0.3839, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 2.787878787878788, | |
| "grad_norm": 0.6219557556210868, | |
| "learning_rate": 1.513085386565758e-05, | |
| "loss": 0.4001, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 2.8181818181818183, | |
| "grad_norm": 0.6995782551314211, | |
| "learning_rate": 1.5000000000000002e-05, | |
| "loss": 0.3711, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 2.8484848484848486, | |
| "grad_norm": 0.6206524418283439, | |
| "learning_rate": 1.4867994483783485e-05, | |
| "loss": 0.3885, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 2.878787878787879, | |
| "grad_norm": 0.6034110513686886, | |
| "learning_rate": 1.4734867721853341e-05, | |
| "loss": 0.3662, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 2.909090909090909, | |
| "grad_norm": 0.6117772819381968, | |
| "learning_rate": 1.4600650377311523e-05, | |
| "loss": 0.3784, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 2.9393939393939394, | |
| "grad_norm": 0.6459508908856219, | |
| "learning_rate": 1.4465373364454001e-05, | |
| "loss": 0.3825, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 2.9696969696969697, | |
| "grad_norm": 0.6121870346418952, | |
| "learning_rate": 1.4329067841650274e-05, | |
| "loss": 0.3922, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.5945934088594842, | |
| "learning_rate": 1.4191765204166643e-05, | |
| "loss": 0.3662, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 3.0303030303030303, | |
| "grad_norm": 0.8099290156167358, | |
| "learning_rate": 1.4053497076934948e-05, | |
| "loss": 0.3415, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.0606060606060606, | |
| "grad_norm": 0.5150715052366444, | |
| "learning_rate": 1.3914295307268396e-05, | |
| "loss": 0.2791, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 3.090909090909091, | |
| "grad_norm": 0.7797992886476148, | |
| "learning_rate": 1.3774191957526144e-05, | |
| "loss": 0.2581, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 3.121212121212121, | |
| "grad_norm": 0.9278149908141471, | |
| "learning_rate": 1.3633219297728415e-05, | |
| "loss": 0.3015, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 3.1515151515151514, | |
| "grad_norm": 0.5714507590633585, | |
| "learning_rate": 1.3491409798123687e-05, | |
| "loss": 0.2568, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 3.1818181818181817, | |
| "grad_norm": 0.7549149907792995, | |
| "learning_rate": 1.3348796121709862e-05, | |
| "loss": 0.2676, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 3.212121212121212, | |
| "grad_norm": 0.6472558289271755, | |
| "learning_rate": 1.3205411116710973e-05, | |
| "loss": 0.2752, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 3.242424242424242, | |
| "grad_norm": 0.5633279307573167, | |
| "learning_rate": 1.3061287809011243e-05, | |
| "loss": 0.2739, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 3.2727272727272725, | |
| "grad_norm": 0.5785396190291751, | |
| "learning_rate": 1.291645939454825e-05, | |
| "loss": 0.2738, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 3.303030303030303, | |
| "grad_norm": 0.6230402986042578, | |
| "learning_rate": 1.277095923166689e-05, | |
| "loss": 0.2864, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 3.3333333333333335, | |
| "grad_norm": 0.5800228026086771, | |
| "learning_rate": 1.2624820833435939e-05, | |
| "loss": 0.2708, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 3.3636363636363638, | |
| "grad_norm": 0.5536822867762624, | |
| "learning_rate": 1.2478077859929e-05, | |
| "loss": 0.2559, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 3.393939393939394, | |
| "grad_norm": 0.5760656027333697, | |
| "learning_rate": 1.2330764110471567e-05, | |
| "loss": 0.2823, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 3.4242424242424243, | |
| "grad_norm": 0.5543164962740332, | |
| "learning_rate": 1.2182913515856016e-05, | |
| "loss": 0.2958, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 3.4545454545454546, | |
| "grad_norm": 0.5598002501612586, | |
| "learning_rate": 1.2034560130526341e-05, | |
| "loss": 0.2762, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 3.484848484848485, | |
| "grad_norm": 0.5427601594632117, | |
| "learning_rate": 1.1885738124734359e-05, | |
| "loss": 0.2869, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 3.515151515151515, | |
| "grad_norm": 0.5533179623685557, | |
| "learning_rate": 1.1736481776669307e-05, | |
| "loss": 0.234, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 3.5454545454545454, | |
| "grad_norm": 0.5319723931816306, | |
| "learning_rate": 1.1586825464562515e-05, | |
| "loss": 0.2823, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 3.5757575757575757, | |
| "grad_norm": 0.5245724809469451, | |
| "learning_rate": 1.1436803658769082e-05, | |
| "loss": 0.3014, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 3.606060606060606, | |
| "grad_norm": 0.5411443507792344, | |
| "learning_rate": 1.1286450913828313e-05, | |
| "loss": 0.2687, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 3.6363636363636362, | |
| "grad_norm": 0.5163357572823526, | |
| "learning_rate": 1.113580186050475e-05, | |
| "loss": 0.2728, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 3.6666666666666665, | |
| "grad_norm": 0.5899283053322925, | |
| "learning_rate": 1.0984891197811686e-05, | |
| "loss": 0.2459, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 3.6969696969696972, | |
| "grad_norm": 0.5056828061425238, | |
| "learning_rate": 1.0833753685018935e-05, | |
| "loss": 0.271, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 3.7272727272727275, | |
| "grad_norm": 0.6052763785820251, | |
| "learning_rate": 1.0682424133646712e-05, | |
| "loss": 0.272, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 3.757575757575758, | |
| "grad_norm": 0.5637995961655816, | |
| "learning_rate": 1.0530937399447496e-05, | |
| "loss": 0.2672, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 3.787878787878788, | |
| "grad_norm": 0.5736363544698516, | |
| "learning_rate": 1.0379328374377715e-05, | |
| "loss": 0.2715, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 3.8181818181818183, | |
| "grad_norm": 0.4833069296671861, | |
| "learning_rate": 1.0227631978561057e-05, | |
| "loss": 0.2757, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 3.8484848484848486, | |
| "grad_norm": 0.5465310709540051, | |
| "learning_rate": 1.0075883152245334e-05, | |
| "loss": 0.2851, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 3.878787878787879, | |
| "grad_norm": 0.4792454780979557, | |
| "learning_rate": 9.92411684775467e-06, | |
| "loss": 0.2663, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 3.909090909090909, | |
| "grad_norm": 0.48912202932443005, | |
| "learning_rate": 9.772368021438943e-06, | |
| "loss": 0.2578, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 3.9393939393939394, | |
| "grad_norm": 0.4934070386848243, | |
| "learning_rate": 9.620671625622287e-06, | |
| "loss": 0.2745, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 3.9696969696969697, | |
| "grad_norm": 0.49899682718769883, | |
| "learning_rate": 9.469062600552509e-06, | |
| "loss": 0.2453, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.5112131493673205, | |
| "learning_rate": 9.317575866353293e-06, | |
| "loss": 0.2436, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 4.03030303030303, | |
| "grad_norm": 0.6504212910352821, | |
| "learning_rate": 9.166246314981066e-06, | |
| "loss": 0.187, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 4.0606060606060606, | |
| "grad_norm": 0.4705903819193525, | |
| "learning_rate": 9.015108802188314e-06, | |
| "loss": 0.2116, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 4.090909090909091, | |
| "grad_norm": 0.43306814721746406, | |
| "learning_rate": 8.86419813949525e-06, | |
| "loss": 0.21, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 4.121212121212121, | |
| "grad_norm": 0.5551013058456373, | |
| "learning_rate": 8.71354908617169e-06, | |
| "loss": 0.1871, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 4.151515151515151, | |
| "grad_norm": 0.7186068408401892, | |
| "learning_rate": 8.56319634123092e-06, | |
| "loss": 0.2082, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 4.181818181818182, | |
| "grad_norm": 0.5051811218960339, | |
| "learning_rate": 8.413174535437486e-06, | |
| "loss": 0.2214, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 4.212121212121212, | |
| "grad_norm": 0.5142092027697022, | |
| "learning_rate": 8.263518223330698e-06, | |
| "loss": 0.2175, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 4.242424242424242, | |
| "grad_norm": 0.5981256829736894, | |
| "learning_rate": 8.114261875265643e-06, | |
| "loss": 0.2101, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 4.2727272727272725, | |
| "grad_norm": 0.5019288468754832, | |
| "learning_rate": 7.965439869473664e-06, | |
| "loss": 0.1756, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 4.303030303030303, | |
| "grad_norm": 0.4439156615123574, | |
| "learning_rate": 7.817086484143987e-06, | |
| "loss": 0.1904, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 4.333333333333333, | |
| "grad_norm": 0.45895645378470623, | |
| "learning_rate": 7.669235889528436e-06, | |
| "loss": 0.1698, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 4.363636363636363, | |
| "grad_norm": 0.48011561168439476, | |
| "learning_rate": 7.521922140071003e-06, | |
| "loss": 0.1796, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 4.393939393939394, | |
| "grad_norm": 0.48925867980481696, | |
| "learning_rate": 7.375179166564062e-06, | |
| "loss": 0.1801, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 4.424242424242424, | |
| "grad_norm": 0.44138758840447456, | |
| "learning_rate": 7.2290407683331154e-06, | |
| "loss": 0.1703, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 4.454545454545454, | |
| "grad_norm": 0.4332510625214356, | |
| "learning_rate": 7.0835406054517505e-06, | |
| "loss": 0.2, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 4.484848484848484, | |
| "grad_norm": 0.463874000428395, | |
| "learning_rate": 6.93871219098876e-06, | |
| "loss": 0.2049, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 4.515151515151516, | |
| "grad_norm": 0.43649644239555635, | |
| "learning_rate": 6.79458888328903e-06, | |
| "loss": 0.1961, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 4.545454545454545, | |
| "grad_norm": 0.42871805529474544, | |
| "learning_rate": 6.651203878290139e-06, | |
| "loss": 0.1895, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 4.575757575757576, | |
| "grad_norm": 0.442897745275202, | |
| "learning_rate": 6.508590201876317e-06, | |
| "loss": 0.2183, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 4.606060606060606, | |
| "grad_norm": 0.43863906607775893, | |
| "learning_rate": 6.366780702271589e-06, | |
| "loss": 0.1904, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 4.636363636363637, | |
| "grad_norm": 0.42766689534113494, | |
| "learning_rate": 6.225808042473857e-06, | |
| "loss": 0.2108, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 4.666666666666667, | |
| "grad_norm": 0.43659250162300034, | |
| "learning_rate": 6.085704692731609e-06, | |
| "loss": 0.214, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 4.696969696969697, | |
| "grad_norm": 0.4352431421040145, | |
| "learning_rate": 5.946502923065054e-06, | |
| "loss": 0.1867, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 4.7272727272727275, | |
| "grad_norm": 0.4035382334968924, | |
| "learning_rate": 5.8082347958333625e-06, | |
| "loss": 0.1966, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 4.757575757575758, | |
| "grad_norm": 0.4056938873427896, | |
| "learning_rate": 5.670932158349732e-06, | |
| "loss": 0.2032, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 4.787878787878788, | |
| "grad_norm": 0.4303608924411298, | |
| "learning_rate": 5.534626635546e-06, | |
| "loss": 0.2133, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 4.818181818181818, | |
| "grad_norm": 0.4730389346705751, | |
| "learning_rate": 5.399349622688479e-06, | |
| "loss": 0.2138, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 4.848484848484849, | |
| "grad_norm": 0.3993250434892082, | |
| "learning_rate": 5.2651322781466606e-06, | |
| "loss": 0.19, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 4.878787878787879, | |
| "grad_norm": 0.36731649843808584, | |
| "learning_rate": 5.132005516216512e-06, | |
| "loss": 0.2112, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 4.909090909090909, | |
| "grad_norm": 0.3906896046771335, | |
| "learning_rate": 5.000000000000003e-06, | |
| "loss": 0.1714, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 4.9393939393939394, | |
| "grad_norm": 0.4268824521398622, | |
| "learning_rate": 4.869146134342426e-06, | |
| "loss": 0.1809, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 4.96969696969697, | |
| "grad_norm": 0.3889911141628613, | |
| "learning_rate": 4.739474058829288e-06, | |
| "loss": 0.1663, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.4316054039243355, | |
| "learning_rate": 4.611013640844245e-06, | |
| "loss": 0.1936, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 5.03030303030303, | |
| "grad_norm": 0.3832772405938606, | |
| "learning_rate": 4.483794468689728e-06, | |
| "loss": 0.1673, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 5.0606060606060606, | |
| "grad_norm": 0.36800013616948724, | |
| "learning_rate": 4.357845844771881e-06, | |
| "loss": 0.1245, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 5.090909090909091, | |
| "grad_norm": 0.33751973892440384, | |
| "learning_rate": 4.2331967788513295e-06, | |
| "loss": 0.169, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 5.121212121212121, | |
| "grad_norm": 0.3322939157192326, | |
| "learning_rate": 4.109875981361363e-06, | |
| "loss": 0.1756, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 5.151515151515151, | |
| "grad_norm": 0.3812314112782908, | |
| "learning_rate": 3.987911856795047e-06, | |
| "loss": 0.184, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 5.181818181818182, | |
| "grad_norm": 0.3652321148627788, | |
| "learning_rate": 3.867332497162836e-06, | |
| "loss": 0.1566, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 5.212121212121212, | |
| "grad_norm": 0.3879576882201218, | |
| "learning_rate": 3.748165675522113e-06, | |
| "loss": 0.1753, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 5.242424242424242, | |
| "grad_norm": 0.3728359869994215, | |
| "learning_rate": 3.630438839580217e-06, | |
| "loss": 0.1724, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 5.2727272727272725, | |
| "grad_norm": 0.33306017240961666, | |
| "learning_rate": 3.5141791053724405e-06, | |
| "loss": 0.1605, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 5.303030303030303, | |
| "grad_norm": 0.3374355334137724, | |
| "learning_rate": 3.399413251016359e-06, | |
| "loss": 0.159, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 5.333333333333333, | |
| "grad_norm": 0.3686488288607569, | |
| "learning_rate": 3.2861677105440335e-06, | |
| "loss": 0.1751, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 5.363636363636363, | |
| "grad_norm": 0.30451056598911924, | |
| "learning_rate": 3.174468567813461e-06, | |
| "loss": 0.1645, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 5.393939393939394, | |
| "grad_norm": 0.29609402601383567, | |
| "learning_rate": 3.0643415505006733e-06, | |
| "loss": 0.1588, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 5.424242424242424, | |
| "grad_norm": 0.3177802255154638, | |
| "learning_rate": 2.9558120241738786e-06, | |
| "loss": 0.1397, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 5.454545454545454, | |
| "grad_norm": 0.3274938463450976, | |
| "learning_rate": 2.8489049864510053e-06, | |
| "loss": 0.1394, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 5.484848484848484, | |
| "grad_norm": 0.33858714379928084, | |
| "learning_rate": 2.7436450612420098e-06, | |
| "loss": 0.1592, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 5.515151515151516, | |
| "grad_norm": 0.33031998226857323, | |
| "learning_rate": 2.640056493077231e-06, | |
| "loss": 0.1585, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 5.545454545454545, | |
| "grad_norm": 0.3293127119831307, | |
| "learning_rate": 2.5381631415231455e-06, | |
| "loss": 0.1705, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 5.575757575757576, | |
| "grad_norm": 0.31808083884309923, | |
| "learning_rate": 2.4379884756868167e-06, | |
| "loss": 0.156, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 5.606060606060606, | |
| "grad_norm": 0.3137855630080014, | |
| "learning_rate": 2.339555568810221e-06, | |
| "loss": 0.1623, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 5.636363636363637, | |
| "grad_norm": 0.34339684594276837, | |
| "learning_rate": 2.2428870929558012e-06, | |
| "loss": 0.1323, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 5.666666666666667, | |
| "grad_norm": 0.31577628788346407, | |
| "learning_rate": 2.1480053137844115e-06, | |
| "loss": 0.1356, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 5.696969696969697, | |
| "grad_norm": 0.29489130691296817, | |
| "learning_rate": 2.054932085426856e-06, | |
| "loss": 0.1368, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 5.7272727272727275, | |
| "grad_norm": 0.3097508163890114, | |
| "learning_rate": 1.963688845450218e-06, | |
| "loss": 0.1395, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 5.757575757575758, | |
| "grad_norm": 0.323250511983245, | |
| "learning_rate": 1.8742966099201699e-06, | |
| "loss": 0.1739, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 5.787878787878788, | |
| "grad_norm": 0.30998522403433326, | |
| "learning_rate": 1.7867759685603115e-06, | |
| "loss": 0.124, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 5.818181818181818, | |
| "grad_norm": 0.2978994920727208, | |
| "learning_rate": 1.7011470800097496e-06, | |
| "loss": 0.1823, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 5.848484848484849, | |
| "grad_norm": 0.30595924523973467, | |
| "learning_rate": 1.6174296671799571e-06, | |
| "loss": 0.1589, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 5.878787878787879, | |
| "grad_norm": 0.31895789080001874, | |
| "learning_rate": 1.5356430127119915e-06, | |
| "loss": 0.1599, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 5.909090909090909, | |
| "grad_norm": 0.3121407589131782, | |
| "learning_rate": 1.4558059545351144e-06, | |
| "loss": 0.1342, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 5.9393939393939394, | |
| "grad_norm": 0.30382285870152187, | |
| "learning_rate": 1.3779368815278648e-06, | |
| "loss": 0.1328, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 5.96969696969697, | |
| "grad_norm": 0.3081643736545958, | |
| "learning_rate": 1.302053729282533e-06, | |
| "loss": 0.1656, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 0.3049150198643163, | |
| "learning_rate": 1.2281739759740575e-06, | |
| "loss": 0.1448, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 6.03030303030303, | |
| "grad_norm": 0.3051897441172669, | |
| "learning_rate": 1.156314638334277e-06, | |
| "loss": 0.1431, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 6.0606060606060606, | |
| "grad_norm": 0.313501804289099, | |
| "learning_rate": 1.086492267732462e-06, | |
| "loss": 0.1565, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 6.090909090909091, | |
| "grad_norm": 0.2864845932542848, | |
| "learning_rate": 1.01872294636304e-06, | |
| "loss": 0.1085, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 6.121212121212121, | |
| "grad_norm": 0.27967933376523796, | |
| "learning_rate": 9.530222835413739e-07, | |
| "loss": 0.1434, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 6.151515151515151, | |
| "grad_norm": 0.2829978682294024, | |
| "learning_rate": 8.894054121084839e-07, | |
| "loss": 0.1399, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 6.181818181818182, | |
| "grad_norm": 0.23088841945153465, | |
| "learning_rate": 8.278869849454718e-07, | |
| "loss": 0.1402, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 6.212121212121212, | |
| "grad_norm": 0.29124755048152023, | |
| "learning_rate": 7.684811715985429e-07, | |
| "loss": 0.1518, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 6.242424242424242, | |
| "grad_norm": 0.2611224816916009, | |
| "learning_rate": 7.1120165501533e-07, | |
| "loss": 0.13, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 6.2727272727272725, | |
| "grad_norm": 0.26811544999082754, | |
| "learning_rate": 6.560616283932897e-07, | |
| "loss": 0.1387, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 6.303030303030303, | |
| "grad_norm": 0.31306543438123136, | |
| "learning_rate": 6.030737921409169e-07, | |
| "loss": 0.1549, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 6.333333333333333, | |
| "grad_norm": 0.26674685667962245, | |
| "learning_rate": 5.522503509524591e-07, | |
| "loss": 0.1109, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 6.363636363636363, | |
| "grad_norm": 0.29409637042181064, | |
| "learning_rate": 5.036030109968082e-07, | |
| "loss": 0.15, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 6.393939393939394, | |
| "grad_norm": 0.27988788633571876, | |
| "learning_rate": 4.5714297722121105e-07, | |
| "loss": 0.1368, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 6.424242424242424, | |
| "grad_norm": 0.30019085167602555, | |
| "learning_rate": 4.128809507704445e-07, | |
| "loss": 0.1636, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 6.454545454545454, | |
| "grad_norm": 0.22576538745694485, | |
| "learning_rate": 3.708271265220087e-07, | |
| "loss": 0.1135, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 6.484848484848484, | |
| "grad_norm": 0.28149880342388367, | |
| "learning_rate": 3.309911907379393e-07, | |
| "loss": 0.1391, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 6.515151515151516, | |
| "grad_norm": 0.30862224909209585, | |
| "learning_rate": 2.9338231883378365e-07, | |
| "loss": 0.1633, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 6.545454545454545, | |
| "grad_norm": 0.2291803571610891, | |
| "learning_rate": 2.5800917326521013e-07, | |
| "loss": 0.1044, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 6.575757575757576, | |
| "grad_norm": 0.2638829364154659, | |
| "learning_rate": 2.248799015327907e-07, | |
| "loss": 0.1302, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 6.606060606060606, | |
| "grad_norm": 0.26335615053431827, | |
| "learning_rate": 1.9400213430538773e-07, | |
| "loss": 0.1377, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 6.636363636363637, | |
| "grad_norm": 0.28349814287233055, | |
| "learning_rate": 1.6538298366257975e-07, | |
| "loss": 0.1713, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 6.666666666666667, | |
| "grad_norm": 0.2567166750075201, | |
| "learning_rate": 1.3902904145653094e-07, | |
| "loss": 0.1478, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 6.696969696969697, | |
| "grad_norm": 0.2680290221460674, | |
| "learning_rate": 1.1494637779369766e-07, | |
| "loss": 0.1182, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 6.7272727272727275, | |
| "grad_norm": 0.29013244444291286, | |
| "learning_rate": 9.314053963669245e-08, | |
| "loss": 0.1542, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 6.757575757575758, | |
| "grad_norm": 0.2552816672881068, | |
| "learning_rate": 7.361654952665608e-08, | |
| "loss": 0.145, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 6.787878787878788, | |
| "grad_norm": 0.26800114048103063, | |
| "learning_rate": 5.637890442641403e-08, | |
| "loss": 0.1504, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 6.818181818181818, | |
| "grad_norm": 0.25971750088525103, | |
| "learning_rate": 4.143157468468717e-08, | |
| "loss": 0.147, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 6.848484848484849, | |
| "grad_norm": 0.24581414228868528, | |
| "learning_rate": 2.8778003121607834e-08, | |
| "loss": 0.147, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 6.878787878787879, | |
| "grad_norm": 0.26922469614869, | |
| "learning_rate": 1.8421104235727406e-08, | |
| "loss": 0.1384, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 6.909090909090909, | |
| "grad_norm": 0.26184348115472483, | |
| "learning_rate": 1.0363263532724433e-08, | |
| "loss": 0.1421, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 6.9393939393939394, | |
| "grad_norm": 0.24927813406249957, | |
| "learning_rate": 4.606336975948589e-09, | |
| "loss": 0.1359, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 6.96969696969697, | |
| "grad_norm": 0.2963181121468789, | |
| "learning_rate": 1.1516505589381777e-09, | |
| "loss": 0.145, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 0.25532476849891694, | |
| "learning_rate": 0.0, | |
| "loss": 0.1566, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "step": 231, | |
| "total_flos": 1.9384640635876147e+17, | |
| "train_loss": 0.369713457341576, | |
| "train_runtime": 3572.7124, | |
| "train_samples_per_second": 6.191, | |
| "train_steps_per_second": 0.065 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 231, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 7, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.9384640635876147e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |