| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.3333613374779467, | |
| "eval_steps": 248, | |
| "global_step": 248, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0013441989414433337, | |
| "grad_norm": 9.1875, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 1.0378, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0013441989414433337, | |
| "eval_loss": 3.0436718463897705, | |
| "eval_runtime": 5734.1832, | |
| "eval_samples_per_second": 3.861, | |
| "eval_steps_per_second": 0.483, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0026883978828866673, | |
| "grad_norm": 10.3125, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 1.1143, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.004032596824330001, | |
| "grad_norm": 8.75, | |
| "learning_rate": 3e-06, | |
| "loss": 1.0933, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.005376795765773335, | |
| "grad_norm": 6.40625, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 1.1014, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.006720994707216668, | |
| "grad_norm": 5.34375, | |
| "learning_rate": 5e-06, | |
| "loss": 1.0504, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.008065193648660002, | |
| "grad_norm": 3.453125, | |
| "learning_rate": 6e-06, | |
| "loss": 1.0031, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.009409392590103335, | |
| "grad_norm": 3.0, | |
| "learning_rate": 7e-06, | |
| "loss": 1.0001, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.01075359153154667, | |
| "grad_norm": 2.640625, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 1.0196, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.012097790472990002, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 9e-06, | |
| "loss": 0.9207, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.013441989414433336, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 1e-05, | |
| "loss": 0.9502, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.01478618835587667, | |
| "grad_norm": 2.34375, | |
| "learning_rate": 9.999954076906038e-06, | |
| "loss": 0.9674, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.016130387297320005, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 9.99981630846772e-06, | |
| "loss": 0.8617, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.017474586238763336, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 9.999586697215748e-06, | |
| "loss": 0.9222, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.01881878518020667, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 9.999265247367909e-06, | |
| "loss": 0.884, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.020162984121650004, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 9.998851964828987e-06, | |
| "loss": 0.9086, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.02150718306309334, | |
| "grad_norm": 1.6015625, | |
| "learning_rate": 9.99834685719067e-06, | |
| "loss": 0.8723, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.022851382004536673, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 9.997749933731397e-06, | |
| "loss": 0.8678, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.024195580945980004, | |
| "grad_norm": 1.5625, | |
| "learning_rate": 9.997061205416203e-06, | |
| "loss": 0.8497, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.025539779887423338, | |
| "grad_norm": 1.515625, | |
| "learning_rate": 9.996280684896496e-06, | |
| "loss": 0.8475, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.026883978828866673, | |
| "grad_norm": 1.625, | |
| "learning_rate": 9.995408386509846e-06, | |
| "loss": 0.8673, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.028228177770310007, | |
| "grad_norm": 1.5625, | |
| "learning_rate": 9.99444432627971e-06, | |
| "loss": 0.8704, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.02957237671175334, | |
| "grad_norm": 1.53125, | |
| "learning_rate": 9.993388521915134e-06, | |
| "loss": 0.813, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.030916575653196672, | |
| "grad_norm": 1.453125, | |
| "learning_rate": 9.992240992810445e-06, | |
| "loss": 0.8497, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.03226077459464001, | |
| "grad_norm": 1.453125, | |
| "learning_rate": 9.991001760044877e-06, | |
| "loss": 0.8027, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.03360497353608334, | |
| "grad_norm": 1.375, | |
| "learning_rate": 9.989670846382189e-06, | |
| "loss": 0.8099, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.03494917247752667, | |
| "grad_norm": 1.328125, | |
| "learning_rate": 9.98824827627025e-06, | |
| "loss": 0.7818, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.03629337141897001, | |
| "grad_norm": 1.5, | |
| "learning_rate": 9.986734075840591e-06, | |
| "loss": 0.8078, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.03763757036041334, | |
| "grad_norm": 1.5625, | |
| "learning_rate": 9.985128272907917e-06, | |
| "loss": 0.819, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.03898176930185668, | |
| "grad_norm": 1.34375, | |
| "learning_rate": 9.983430896969606e-06, | |
| "loss": 0.7949, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.04032596824330001, | |
| "grad_norm": 1.46875, | |
| "learning_rate": 9.981641979205158e-06, | |
| "loss": 0.8044, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.04167016718474334, | |
| "grad_norm": 1.4921875, | |
| "learning_rate": 9.97976155247563e-06, | |
| "loss": 0.8025, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.04301436612618668, | |
| "grad_norm": 1.3828125, | |
| "learning_rate": 9.977789651323025e-06, | |
| "loss": 0.7461, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.04435856506763001, | |
| "grad_norm": 1.40625, | |
| "learning_rate": 9.975726311969664e-06, | |
| "loss": 0.787, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.045702764009073346, | |
| "grad_norm": 1.4609375, | |
| "learning_rate": 9.973571572317519e-06, | |
| "loss": 0.7837, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.04704696295051668, | |
| "grad_norm": 1.4453125, | |
| "learning_rate": 9.971325471947518e-06, | |
| "loss": 0.8101, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.04839116189196001, | |
| "grad_norm": 1.4921875, | |
| "learning_rate": 9.968988052118804e-06, | |
| "loss": 0.783, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.049735360833403346, | |
| "grad_norm": 1.4609375, | |
| "learning_rate": 9.966559355768005e-06, | |
| "loss": 0.8141, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.051079559774846676, | |
| "grad_norm": 1.359375, | |
| "learning_rate": 9.964039427508418e-06, | |
| "loss": 0.7753, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.052423758716290014, | |
| "grad_norm": 1.375, | |
| "learning_rate": 9.961428313629203e-06, | |
| "loss": 0.773, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.053767957657733345, | |
| "grad_norm": 1.3515625, | |
| "learning_rate": 9.958726062094533e-06, | |
| "loss": 0.7557, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.055112156599176676, | |
| "grad_norm": 1.3203125, | |
| "learning_rate": 9.955932722542709e-06, | |
| "loss": 0.7241, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.056456355540620014, | |
| "grad_norm": 1.4296875, | |
| "learning_rate": 9.953048346285245e-06, | |
| "loss": 0.781, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.057800554482063345, | |
| "grad_norm": 1.3984375, | |
| "learning_rate": 9.950072986305938e-06, | |
| "loss": 0.7644, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.05914475342350668, | |
| "grad_norm": 1.3515625, | |
| "learning_rate": 9.947006697259881e-06, | |
| "loss": 0.7406, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.06048895236495001, | |
| "grad_norm": 1.3828125, | |
| "learning_rate": 9.943849535472468e-06, | |
| "loss": 0.762, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.061833151306393344, | |
| "grad_norm": 1.3046875, | |
| "learning_rate": 9.940601558938348e-06, | |
| "loss": 0.7448, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.06317735024783667, | |
| "grad_norm": 1.390625, | |
| "learning_rate": 9.93726282732038e-06, | |
| "loss": 0.7658, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.06452154918928002, | |
| "grad_norm": 1.4765625, | |
| "learning_rate": 9.933833401948514e-06, | |
| "loss": 0.7876, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.06586574813072335, | |
| "grad_norm": 1.3984375, | |
| "learning_rate": 9.930313345818683e-06, | |
| "loss": 0.8166, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.06720994707216668, | |
| "grad_norm": 1.421875, | |
| "learning_rate": 9.92670272359163e-06, | |
| "loss": 0.7927, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.06855414601361001, | |
| "grad_norm": 1.359375, | |
| "learning_rate": 9.923001601591738e-06, | |
| "loss": 0.7906, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.06989834495505334, | |
| "grad_norm": 1.296875, | |
| "learning_rate": 9.919210047805791e-06, | |
| "loss": 0.7576, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.07124254389649669, | |
| "grad_norm": 1.4453125, | |
| "learning_rate": 9.915328131881745e-06, | |
| "loss": 0.7847, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.07258674283794002, | |
| "grad_norm": 1.4375, | |
| "learning_rate": 9.911355925127433e-06, | |
| "loss": 0.7728, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.07393094177938335, | |
| "grad_norm": 1.4375, | |
| "learning_rate": 9.907293500509268e-06, | |
| "loss": 0.7947, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.07527514072082668, | |
| "grad_norm": 1.3515625, | |
| "learning_rate": 9.903140932650891e-06, | |
| "loss": 0.7503, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.07661933966227001, | |
| "grad_norm": 1.3671875, | |
| "learning_rate": 9.898898297831808e-06, | |
| "loss": 0.7519, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.07796353860371336, | |
| "grad_norm": 1.4296875, | |
| "learning_rate": 9.894565673985986e-06, | |
| "loss": 0.7438, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.07930773754515669, | |
| "grad_norm": 1.4375, | |
| "learning_rate": 9.890143140700419e-06, | |
| "loss": 0.7635, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.08065193648660002, | |
| "grad_norm": 1.375, | |
| "learning_rate": 9.885630779213678e-06, | |
| "loss": 0.7176, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.08199613542804335, | |
| "grad_norm": 1.4296875, | |
| "learning_rate": 9.881028672414397e-06, | |
| "loss": 0.7536, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.08334033436948668, | |
| "grad_norm": 1.4375, | |
| "learning_rate": 9.876336904839772e-06, | |
| "loss": 0.7572, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.08468453331093002, | |
| "grad_norm": 1.4609375, | |
| "learning_rate": 9.871555562673996e-06, | |
| "loss": 0.7767, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.08602873225237335, | |
| "grad_norm": 1.4921875, | |
| "learning_rate": 9.86668473374668e-06, | |
| "loss": 0.763, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.08737293119381669, | |
| "grad_norm": 1.3984375, | |
| "learning_rate": 9.861724507531234e-06, | |
| "loss": 0.7576, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.08871713013526002, | |
| "grad_norm": 1.421875, | |
| "learning_rate": 9.856674975143237e-06, | |
| "loss": 0.7342, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.09006132907670335, | |
| "grad_norm": 1.4453125, | |
| "learning_rate": 9.851536229338747e-06, | |
| "loss": 0.7487, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.09140552801814669, | |
| "grad_norm": 1.390625, | |
| "learning_rate": 9.846308364512607e-06, | |
| "loss": 0.7604, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.09274972695959002, | |
| "grad_norm": 1.5078125, | |
| "learning_rate": 9.840991476696707e-06, | |
| "loss": 0.7537, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.09409392590103335, | |
| "grad_norm": 1.4296875, | |
| "learning_rate": 9.835585663558221e-06, | |
| "loss": 0.7608, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.09543812484247668, | |
| "grad_norm": 1.4140625, | |
| "learning_rate": 9.830091024397818e-06, | |
| "loss": 0.7298, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.09678232378392002, | |
| "grad_norm": 1.4140625, | |
| "learning_rate": 9.824507660147831e-06, | |
| "loss": 0.7646, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.09812652272536336, | |
| "grad_norm": 1.3984375, | |
| "learning_rate": 9.8188356733704e-06, | |
| "loss": 0.7679, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.09947072166680669, | |
| "grad_norm": 1.4765625, | |
| "learning_rate": 9.813075168255601e-06, | |
| "loss": 0.7263, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.10081492060825002, | |
| "grad_norm": 1.484375, | |
| "learning_rate": 9.807226250619522e-06, | |
| "loss": 0.7589, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.10215911954969335, | |
| "grad_norm": 1.515625, | |
| "learning_rate": 9.801289027902316e-06, | |
| "loss": 0.7216, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.10350331849113668, | |
| "grad_norm": 1.4765625, | |
| "learning_rate": 9.795263609166243e-06, | |
| "loss": 0.7779, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.10484751743258003, | |
| "grad_norm": 1.4609375, | |
| "learning_rate": 9.789150105093647e-06, | |
| "loss": 0.6941, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.10619171637402336, | |
| "grad_norm": 1.46875, | |
| "learning_rate": 9.78294862798494e-06, | |
| "loss": 0.7362, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.10753591531546669, | |
| "grad_norm": 1.4296875, | |
| "learning_rate": 9.776659291756528e-06, | |
| "loss": 0.7103, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.10888011425691002, | |
| "grad_norm": 1.359375, | |
| "learning_rate": 9.77028221193872e-06, | |
| "loss": 0.6866, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.11022431319835335, | |
| "grad_norm": 1.53125, | |
| "learning_rate": 9.763817505673614e-06, | |
| "loss": 0.7517, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.1115685121397967, | |
| "grad_norm": 1.453125, | |
| "learning_rate": 9.75726529171293e-06, | |
| "loss": 0.7508, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.11291271108124003, | |
| "grad_norm": 1.5, | |
| "learning_rate": 9.750625690415848e-06, | |
| "loss": 0.7275, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.11425691002268336, | |
| "grad_norm": 1.484375, | |
| "learning_rate": 9.74389882374678e-06, | |
| "loss": 0.6994, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.11560110896412669, | |
| "grad_norm": 1.4453125, | |
| "learning_rate": 9.737084815273137e-06, | |
| "loss": 0.7365, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.11694530790557002, | |
| "grad_norm": 1.5078125, | |
| "learning_rate": 9.730183790163061e-06, | |
| "loss": 0.7294, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.11828950684701336, | |
| "grad_norm": 1.375, | |
| "learning_rate": 9.72319587518312e-06, | |
| "loss": 0.7226, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.1196337057884567, | |
| "grad_norm": 1.390625, | |
| "learning_rate": 9.716121198695987e-06, | |
| "loss": 0.7126, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.12097790472990003, | |
| "grad_norm": 1.484375, | |
| "learning_rate": 9.708959890658074e-06, | |
| "loss": 0.7118, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.12232210367134336, | |
| "grad_norm": 1.3515625, | |
| "learning_rate": 9.70171208261715e-06, | |
| "loss": 0.7164, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.12366630261278669, | |
| "grad_norm": 1.4765625, | |
| "learning_rate": 9.69437790770992e-06, | |
| "loss": 0.7015, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.12501050155423002, | |
| "grad_norm": 1.3828125, | |
| "learning_rate": 9.68695750065959e-06, | |
| "loss": 0.7212, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.12635470049567335, | |
| "grad_norm": 1.4140625, | |
| "learning_rate": 9.679450997773378e-06, | |
| "loss": 0.7301, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.12769889943711668, | |
| "grad_norm": 1.53125, | |
| "learning_rate": 9.67185853694002e-06, | |
| "loss": 0.7434, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.12904309837856004, | |
| "grad_norm": 1.3828125, | |
| "learning_rate": 9.664180257627231e-06, | |
| "loss": 0.7503, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.13038729732000337, | |
| "grad_norm": 1.3984375, | |
| "learning_rate": 9.656416300879147e-06, | |
| "loss": 0.704, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.1317314962614467, | |
| "grad_norm": 1.3515625, | |
| "learning_rate": 9.648566809313738e-06, | |
| "loss": 0.7091, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.13307569520289003, | |
| "grad_norm": 1.25, | |
| "learning_rate": 9.640631927120177e-06, | |
| "loss": 0.6939, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.13441989414433336, | |
| "grad_norm": 1.25, | |
| "learning_rate": 9.632611800056202e-06, | |
| "loss": 0.6645, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.1357640930857767, | |
| "grad_norm": 1.3671875, | |
| "learning_rate": 9.62450657544543e-06, | |
| "loss": 0.7314, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.13710829202722002, | |
| "grad_norm": 1.359375, | |
| "learning_rate": 9.616316402174657e-06, | |
| "loss": 0.7022, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.13845249096866336, | |
| "grad_norm": 1.390625, | |
| "learning_rate": 9.608041430691126e-06, | |
| "loss": 0.7015, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.1397966899101067, | |
| "grad_norm": 1.3046875, | |
| "learning_rate": 9.59968181299975e-06, | |
| "loss": 0.6831, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.14114088885155002, | |
| "grad_norm": 1.28125, | |
| "learning_rate": 9.591237702660335e-06, | |
| "loss": 0.6903, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.14248508779299338, | |
| "grad_norm": 1.2421875, | |
| "learning_rate": 9.58270925478475e-06, | |
| "loss": 0.7236, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.1438292867344367, | |
| "grad_norm": 1.3125, | |
| "learning_rate": 9.574096626034077e-06, | |
| "loss": 0.7375, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.14517348567588004, | |
| "grad_norm": 1.2421875, | |
| "learning_rate": 9.565399974615744e-06, | |
| "loss": 0.7051, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.14651768461732337, | |
| "grad_norm": 1.3515625, | |
| "learning_rate": 9.556619460280605e-06, | |
| "loss": 0.6961, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.1478618835587667, | |
| "grad_norm": 1.2890625, | |
| "learning_rate": 9.547755244320013e-06, | |
| "loss": 0.731, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.14920608250021003, | |
| "grad_norm": 1.3125, | |
| "learning_rate": 9.53880748956286e-06, | |
| "loss": 0.6992, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.15055028144165336, | |
| "grad_norm": 1.2578125, | |
| "learning_rate": 9.529776360372576e-06, | |
| "loss": 0.6954, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.1518944803830967, | |
| "grad_norm": 1.140625, | |
| "learning_rate": 9.52066202264412e-06, | |
| "loss": 0.6932, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.15323867932454002, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 9.511464643800926e-06, | |
| "loss": 0.7117, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.15458287826598335, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 9.502184392791834e-06, | |
| "loss": 0.6992, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.1559270772074267, | |
| "grad_norm": 1.09375, | |
| "learning_rate": 9.492821440087978e-06, | |
| "loss": 0.6744, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.15727127614887004, | |
| "grad_norm": 1.1796875, | |
| "learning_rate": 9.48337595767966e-06, | |
| "loss": 0.6553, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.15861547509031337, | |
| "grad_norm": 1.1484375, | |
| "learning_rate": 9.473848119073188e-06, | |
| "loss": 0.7206, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.1599596740317567, | |
| "grad_norm": 1.125, | |
| "learning_rate": 9.4642380992877e-06, | |
| "loss": 0.7014, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.16130387297320004, | |
| "grad_norm": 1.1171875, | |
| "learning_rate": 9.454546074851927e-06, | |
| "loss": 0.7249, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.16264807191464337, | |
| "grad_norm": 1.140625, | |
| "learning_rate": 9.444772223800972e-06, | |
| "loss": 0.7142, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.1639922708560867, | |
| "grad_norm": 1.0390625, | |
| "learning_rate": 9.434916725673023e-06, | |
| "loss": 0.6778, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.16533646979753003, | |
| "grad_norm": 1.0390625, | |
| "learning_rate": 9.42497976150607e-06, | |
| "loss": 0.6831, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.16668066873897336, | |
| "grad_norm": 1.0546875, | |
| "learning_rate": 9.414961513834569e-06, | |
| "loss": 0.6744, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.1680248676804167, | |
| "grad_norm": 1.1015625, | |
| "learning_rate": 9.404862166686089e-06, | |
| "loss": 0.7059, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.16936906662186005, | |
| "grad_norm": 1.0390625, | |
| "learning_rate": 9.394681905577938e-06, | |
| "loss": 0.7245, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.17071326556330338, | |
| "grad_norm": 1.0390625, | |
| "learning_rate": 9.384420917513752e-06, | |
| "loss": 0.7064, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.1720574645047467, | |
| "grad_norm": 1.0, | |
| "learning_rate": 9.374079390980058e-06, | |
| "loss": 0.7177, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.17340166344619004, | |
| "grad_norm": 1.0078125, | |
| "learning_rate": 9.363657515942814e-06, | |
| "loss": 0.7031, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.17474586238763337, | |
| "grad_norm": 0.96875, | |
| "learning_rate": 9.35315548384392e-06, | |
| "loss": 0.6752, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.1760900613290767, | |
| "grad_norm": 1.0, | |
| "learning_rate": 9.342573487597696e-06, | |
| "loss": 0.7189, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.17743426027052003, | |
| "grad_norm": 1.0234375, | |
| "learning_rate": 9.331911721587345e-06, | |
| "loss": 0.6873, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.17877845921196336, | |
| "grad_norm": 1.0, | |
| "learning_rate": 9.321170381661383e-06, | |
| "loss": 0.7056, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.1801226581534067, | |
| "grad_norm": 0.984375, | |
| "learning_rate": 9.310349665130035e-06, | |
| "loss": 0.7024, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.18146685709485003, | |
| "grad_norm": 1.015625, | |
| "learning_rate": 9.299449770761612e-06, | |
| "loss": 0.708, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.18281105603629338, | |
| "grad_norm": 0.8984375, | |
| "learning_rate": 9.288470898778863e-06, | |
| "loss": 0.6624, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.18415525497773672, | |
| "grad_norm": 0.984375, | |
| "learning_rate": 9.277413250855296e-06, | |
| "loss": 0.6622, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.18549945391918005, | |
| "grad_norm": 0.9609375, | |
| "learning_rate": 9.266277030111474e-06, | |
| "loss": 0.7074, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.18684365286062338, | |
| "grad_norm": 0.97265625, | |
| "learning_rate": 9.25506244111128e-06, | |
| "loss": 0.6745, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.1881878518020667, | |
| "grad_norm": 0.93359375, | |
| "learning_rate": 9.243769689858167e-06, | |
| "loss": 0.6627, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.18953205074351004, | |
| "grad_norm": 0.9609375, | |
| "learning_rate": 9.232398983791363e-06, | |
| "loss": 0.6563, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.19087624968495337, | |
| "grad_norm": 1.0234375, | |
| "learning_rate": 9.220950531782069e-06, | |
| "loss": 0.688, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.1922204486263967, | |
| "grad_norm": 1.046875, | |
| "learning_rate": 9.209424544129621e-06, | |
| "loss": 0.7021, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.19356464756784003, | |
| "grad_norm": 0.87890625, | |
| "learning_rate": 9.197821232557625e-06, | |
| "loss": 0.6826, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.19490884650928336, | |
| "grad_norm": 0.9140625, | |
| "learning_rate": 9.186140810210066e-06, | |
| "loss": 0.6904, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.19625304545072672, | |
| "grad_norm": 1.0, | |
| "learning_rate": 9.1743834916474e-06, | |
| "loss": 0.6747, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.19759724439217005, | |
| "grad_norm": 0.9609375, | |
| "learning_rate": 9.162549492842603e-06, | |
| "loss": 0.7073, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.19894144333361338, | |
| "grad_norm": 0.9765625, | |
| "learning_rate": 9.150639031177211e-06, | |
| "loss": 0.6773, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.2002856422750567, | |
| "grad_norm": 0.94921875, | |
| "learning_rate": 9.138652325437326e-06, | |
| "loss": 0.689, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.20162984121650004, | |
| "grad_norm": 1.015625, | |
| "learning_rate": 9.12658959580959e-06, | |
| "loss": 0.6874, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.20297404015794338, | |
| "grad_norm": 0.90234375, | |
| "learning_rate": 9.114451063877152e-06, | |
| "loss": 0.6617, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.2043182390993867, | |
| "grad_norm": 1.0, | |
| "learning_rate": 9.102236952615588e-06, | |
| "loss": 0.6882, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.20566243804083004, | |
| "grad_norm": 1.03125, | |
| "learning_rate": 9.08994748638881e-06, | |
| "loss": 0.6919, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.20700663698227337, | |
| "grad_norm": 0.97265625, | |
| "learning_rate": 9.077582890944945e-06, | |
| "loss": 0.6713, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.2083508359237167, | |
| "grad_norm": 0.921875, | |
| "learning_rate": 9.065143393412179e-06, | |
| "loss": 0.6504, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.20969503486516006, | |
| "grad_norm": 0.98046875, | |
| "learning_rate": 9.052629222294605e-06, | |
| "loss": 0.6674, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.2110392338066034, | |
| "grad_norm": 0.9375, | |
| "learning_rate": 9.040040607467999e-06, | |
| "loss": 0.7111, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.21238343274804672, | |
| "grad_norm": 1.03125, | |
| "learning_rate": 9.02737778017562e-06, | |
| "loss": 0.6596, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.21372763168949005, | |
| "grad_norm": 0.921875, | |
| "learning_rate": 9.014640973023951e-06, | |
| "loss": 0.6846, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.21507183063093338, | |
| "grad_norm": 0.94921875, | |
| "learning_rate": 9.00183041997843e-06, | |
| "loss": 0.6788, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.2164160295723767, | |
| "grad_norm": 0.890625, | |
| "learning_rate": 8.988946356359147e-06, | |
| "loss": 0.6888, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.21776022851382004, | |
| "grad_norm": 0.92578125, | |
| "learning_rate": 8.97598901883653e-06, | |
| "loss": 0.6854, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.21910442745526337, | |
| "grad_norm": 0.94140625, | |
| "learning_rate": 8.962958645426989e-06, | |
| "loss": 0.7137, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.2204486263967067, | |
| "grad_norm": 0.8984375, | |
| "learning_rate": 8.949855475488549e-06, | |
| "loss": 0.7041, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.22179282533815003, | |
| "grad_norm": 0.98046875, | |
| "learning_rate": 8.936679749716452e-06, | |
| "loss": 0.6573, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.2231370242795934, | |
| "grad_norm": 0.8984375, | |
| "learning_rate": 8.923431710138735e-06, | |
| "loss": 0.6694, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.22448122322103672, | |
| "grad_norm": 0.94140625, | |
| "learning_rate": 8.910111600111786e-06, | |
| "loss": 0.6904, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.22582542216248006, | |
| "grad_norm": 0.953125, | |
| "learning_rate": 8.896719664315866e-06, | |
| "loss": 0.679, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.22716962110392339, | |
| "grad_norm": 0.9140625, | |
| "learning_rate": 8.883256148750634e-06, | |
| "loss": 0.6631, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.22851382004536672, | |
| "grad_norm": 0.9765625, | |
| "learning_rate": 8.869721300730596e-06, | |
| "loss": 0.6576, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.22985801898681005, | |
| "grad_norm": 0.9296875, | |
| "learning_rate": 8.856115368880598e-06, | |
| "loss": 0.6714, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.23120221792825338, | |
| "grad_norm": 0.97265625, | |
| "learning_rate": 8.842438603131232e-06, | |
| "loss": 0.6572, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.2325464168696967, | |
| "grad_norm": 0.91796875, | |
| "learning_rate": 8.828691254714259e-06, | |
| "loss": 0.6798, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.23389061581114004, | |
| "grad_norm": 0.90234375, | |
| "learning_rate": 8.814873576157988e-06, | |
| "loss": 0.6934, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.23523481475258337, | |
| "grad_norm": 1.0390625, | |
| "learning_rate": 8.800985821282637e-06, | |
| "loss": 0.6648, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.23657901369402673, | |
| "grad_norm": 0.8203125, | |
| "learning_rate": 8.787028245195676e-06, | |
| "loss": 0.6705, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.23792321263547006, | |
| "grad_norm": 0.8984375, | |
| "learning_rate": 8.773001104287137e-06, | |
| "loss": 0.6661, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.2392674115769134, | |
| "grad_norm": 0.94140625, | |
| "learning_rate": 8.758904656224904e-06, | |
| "loss": 0.6968, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.24061161051835672, | |
| "grad_norm": 0.94140625, | |
| "learning_rate": 8.744739159949981e-06, | |
| "loss": 0.6817, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.24195580945980005, | |
| "grad_norm": 0.9296875, | |
| "learning_rate": 8.730504875671732e-06, | |
| "loss": 0.6704, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.24330000840124338, | |
| "grad_norm": 0.90625, | |
| "learning_rate": 8.71620206486311e-06, | |
| "loss": 0.6484, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.24464420734268671, | |
| "grad_norm": 0.8515625, | |
| "learning_rate": 8.701830990255843e-06, | |
| "loss": 0.6711, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.24598840628413005, | |
| "grad_norm": 1.0234375, | |
| "learning_rate": 8.687391915835617e-06, | |
| "loss": 0.6526, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.24733260522557338, | |
| "grad_norm": 0.87890625, | |
| "learning_rate": 8.672885106837216e-06, | |
| "loss": 0.6907, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.2486768041670167, | |
| "grad_norm": 0.9609375, | |
| "learning_rate": 8.658310829739666e-06, | |
| "loss": 0.7003, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.25002100310846004, | |
| "grad_norm": 0.85546875, | |
| "learning_rate": 8.643669352261321e-06, | |
| "loss": 0.6353, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.25136520204990337, | |
| "grad_norm": 0.921875, | |
| "learning_rate": 8.628960943354965e-06, | |
| "loss": 0.6566, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.2527094009913467, | |
| "grad_norm": 0.92578125, | |
| "learning_rate": 8.614185873202852e-06, | |
| "loss": 0.6676, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.25405359993279003, | |
| "grad_norm": 0.953125, | |
| "learning_rate": 8.599344413211755e-06, | |
| "loss": 0.6647, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.25539779887423336, | |
| "grad_norm": 0.9140625, | |
| "learning_rate": 8.58443683600798e-06, | |
| "loss": 0.6771, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.2567419978156767, | |
| "grad_norm": 0.90234375, | |
| "learning_rate": 8.569463415432356e-06, | |
| "loss": 0.6629, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.2580861967571201, | |
| "grad_norm": 0.91015625, | |
| "learning_rate": 8.554424426535202e-06, | |
| "loss": 0.7327, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.2594303956985634, | |
| "grad_norm": 0.94921875, | |
| "learning_rate": 8.539320145571277e-06, | |
| "loss": 0.6809, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.26077459464000674, | |
| "grad_norm": 0.875, | |
| "learning_rate": 8.524150849994708e-06, | |
| "loss": 0.6895, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.26211879358145007, | |
| "grad_norm": 0.9140625, | |
| "learning_rate": 8.50891681845389e-06, | |
| "loss": 0.6739, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.2634629925228934, | |
| "grad_norm": 0.96875, | |
| "learning_rate": 8.493618330786365e-06, | |
| "loss": 0.7063, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.26480719146433673, | |
| "grad_norm": 0.85546875, | |
| "learning_rate": 8.47825566801369e-06, | |
| "loss": 0.6676, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.26615139040578006, | |
| "grad_norm": 0.81640625, | |
| "learning_rate": 8.462829112336266e-06, | |
| "loss": 0.6842, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.2674955893472234, | |
| "grad_norm": 0.90625, | |
| "learning_rate": 8.44733894712816e-06, | |
| "loss": 0.7045, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.2688397882886667, | |
| "grad_norm": 1.015625, | |
| "learning_rate": 8.431785456931898e-06, | |
| "loss": 0.6569, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.27018398723011006, | |
| "grad_norm": 0.91015625, | |
| "learning_rate": 8.416168927453237e-06, | |
| "loss": 0.679, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.2715281861715534, | |
| "grad_norm": 0.8671875, | |
| "learning_rate": 8.400489645555914e-06, | |
| "loss": 0.6452, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.2728723851129967, | |
| "grad_norm": 0.9296875, | |
| "learning_rate": 8.384747899256386e-06, | |
| "loss": 0.6744, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.27421658405444005, | |
| "grad_norm": 0.91015625, | |
| "learning_rate": 8.368943977718528e-06, | |
| "loss": 0.6501, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.2755607829958834, | |
| "grad_norm": 0.890625, | |
| "learning_rate": 8.353078171248335e-06, | |
| "loss": 0.666, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.2769049819373267, | |
| "grad_norm": 0.87109375, | |
| "learning_rate": 8.337150771288571e-06, | |
| "loss": 0.6663, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.27824918087877004, | |
| "grad_norm": 0.9296875, | |
| "learning_rate": 8.32116207041343e-06, | |
| "loss": 0.6481, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.2795933798202134, | |
| "grad_norm": 0.921875, | |
| "learning_rate": 8.30511236232316e-06, | |
| "loss": 0.6884, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.2809375787616567, | |
| "grad_norm": 0.890625, | |
| "learning_rate": 8.289001941838659e-06, | |
| "loss": 0.6562, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.28228177770310003, | |
| "grad_norm": 0.8984375, | |
| "learning_rate": 8.27283110489607e-06, | |
| "loss": 0.7106, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.28362597664454336, | |
| "grad_norm": 0.98046875, | |
| "learning_rate": 8.256600148541339e-06, | |
| "loss": 0.689, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.28497017558598675, | |
| "grad_norm": 0.96484375, | |
| "learning_rate": 8.240309370924758e-06, | |
| "loss": 0.6683, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.2863143745274301, | |
| "grad_norm": 0.921875, | |
| "learning_rate": 8.223959071295492e-06, | |
| "loss": 0.6866, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.2876585734688734, | |
| "grad_norm": 0.94140625, | |
| "learning_rate": 8.207549549996083e-06, | |
| "loss": 0.6688, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.28900277241031674, | |
| "grad_norm": 0.91796875, | |
| "learning_rate": 8.191081108456922e-06, | |
| "loss": 0.6721, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.2903469713517601, | |
| "grad_norm": 0.89453125, | |
| "learning_rate": 8.174554049190726e-06, | |
| "loss": 0.6579, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.2916911702932034, | |
| "grad_norm": 0.9609375, | |
| "learning_rate": 8.157968675786971e-06, | |
| "loss": 0.6841, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.29303536923464674, | |
| "grad_norm": 0.93359375, | |
| "learning_rate": 8.141325292906325e-06, | |
| "loss": 0.6739, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.29437956817609007, | |
| "grad_norm": 0.875, | |
| "learning_rate": 8.124624206275041e-06, | |
| "loss": 0.6544, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.2957237671175334, | |
| "grad_norm": 0.97265625, | |
| "learning_rate": 8.107865722679347e-06, | |
| "loss": 0.6525, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.29706796605897673, | |
| "grad_norm": 0.890625, | |
| "learning_rate": 8.091050149959808e-06, | |
| "loss": 0.6394, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.29841216500042006, | |
| "grad_norm": 0.87890625, | |
| "learning_rate": 8.074177797005677e-06, | |
| "loss": 0.6674, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.2997563639418634, | |
| "grad_norm": 0.984375, | |
| "learning_rate": 8.057248973749216e-06, | |
| "loss": 0.6749, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.3011005628833067, | |
| "grad_norm": 1.0234375, | |
| "learning_rate": 8.040263991159996e-06, | |
| "loss": 0.7223, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.30244476182475005, | |
| "grad_norm": 0.859375, | |
| "learning_rate": 8.0232231612392e-06, | |
| "loss": 0.6861, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.3037889607661934, | |
| "grad_norm": 0.9140625, | |
| "learning_rate": 8.006126797013884e-06, | |
| "loss": 0.6883, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.3051331597076367, | |
| "grad_norm": 0.94140625, | |
| "learning_rate": 7.98897521253122e-06, | |
| "loss": 0.655, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.30647735864908004, | |
| "grad_norm": 0.8984375, | |
| "learning_rate": 7.971768722852741e-06, | |
| "loss": 0.6696, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.3078215575905234, | |
| "grad_norm": 0.875, | |
| "learning_rate": 7.954507644048544e-06, | |
| "loss": 0.6676, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.3091657565319667, | |
| "grad_norm": 0.8828125, | |
| "learning_rate": 7.937192293191485e-06, | |
| "loss": 0.6508, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.31050995547341004, | |
| "grad_norm": 0.9375, | |
| "learning_rate": 7.919822988351359e-06, | |
| "loss": 0.6676, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.3118541544148534, | |
| "grad_norm": 0.859375, | |
| "learning_rate": 7.902400048589051e-06, | |
| "loss": 0.6512, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.31319835335629675, | |
| "grad_norm": 0.94140625, | |
| "learning_rate": 7.884923793950684e-06, | |
| "loss": 0.6846, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.3145425522977401, | |
| "grad_norm": 0.8515625, | |
| "learning_rate": 7.86739454546173e-06, | |
| "loss": 0.6644, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.3158867512391834, | |
| "grad_norm": 0.8828125, | |
| "learning_rate": 7.849812625121122e-06, | |
| "loss": 0.6701, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.31723095018062675, | |
| "grad_norm": 0.94921875, | |
| "learning_rate": 7.832178355895327e-06, | |
| "loss": 0.6749, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.3185751491220701, | |
| "grad_norm": 0.87890625, | |
| "learning_rate": 7.81449206171243e-06, | |
| "loss": 0.6584, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.3199193480635134, | |
| "grad_norm": 0.890625, | |
| "learning_rate": 7.796754067456168e-06, | |
| "loss": 0.6685, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.32126354700495674, | |
| "grad_norm": 0.8671875, | |
| "learning_rate": 7.778964698959973e-06, | |
| "loss": 0.6659, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.32260774594640007, | |
| "grad_norm": 0.8671875, | |
| "learning_rate": 7.761124283000982e-06, | |
| "loss": 0.6647, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.3239519448878434, | |
| "grad_norm": 0.9140625, | |
| "learning_rate": 7.743233147294036e-06, | |
| "loss": 0.6793, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.32529614382928673, | |
| "grad_norm": 0.87109375, | |
| "learning_rate": 7.725291620485653e-06, | |
| "loss": 0.6633, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.32664034277073006, | |
| "grad_norm": 0.90625, | |
| "learning_rate": 7.707300032148004e-06, | |
| "loss": 0.6812, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.3279845417121734, | |
| "grad_norm": 0.88671875, | |
| "learning_rate": 7.689258712772851e-06, | |
| "loss": 0.6734, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.3293287406536167, | |
| "grad_norm": 0.859375, | |
| "learning_rate": 7.671167993765474e-06, | |
| "loss": 0.6775, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.33067293959506006, | |
| "grad_norm": 0.84375, | |
| "learning_rate": 7.653028207438588e-06, | |
| "loss": 0.6723, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.3320171385365034, | |
| "grad_norm": 0.90625, | |
| "learning_rate": 7.634839687006242e-06, | |
| "loss": 0.6814, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.3333613374779467, | |
| "grad_norm": 0.984375, | |
| "learning_rate": 7.616602766577683e-06, | |
| "loss": 0.6816, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.3333613374779467, | |
| "eval_loss": 2.7341208457946777, | |
| "eval_runtime": 5734.9058, | |
| "eval_samples_per_second": 3.861, | |
| "eval_steps_per_second": 0.483, | |
| "step": 248 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 743, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 248, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 5.854896016293102e+18, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |