| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.11466574934067195, | |
| "eval_steps": 500, | |
| "global_step": 500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.0, | |
| "loss": 7.2451, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.0, | |
| "loss": 7.3192, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 14.819024134660662, | |
| "learning_rate": 1.3333333333333334e-06, | |
| "loss": 7.2901, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 14.819024134660662, | |
| "learning_rate": 1.3333333333333334e-06, | |
| "loss": 7.4127, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 14.199093979418544, | |
| "learning_rate": 2.666666666666667e-06, | |
| "loss": 7.4321, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 10.185841979858534, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 7.1568, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 39.91704954913313, | |
| "learning_rate": 5.333333333333334e-06, | |
| "loss": 7.2837, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 18.930408398058322, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 7.2329, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 8.865495224195184, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 7.1651, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 35.166574749663724, | |
| "learning_rate": 9.333333333333334e-06, | |
| "loss": 7.1455, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 14.929376279495122, | |
| "learning_rate": 1.0666666666666667e-05, | |
| "loss": 7.2938, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 14.323589623470497, | |
| "learning_rate": 1.2e-05, | |
| "loss": 7.136, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 12.033497318446193, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 7.2873, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 12.713515670882014, | |
| "learning_rate": 1.4666666666666666e-05, | |
| "loss": 7.1778, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 16.40476702849706, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 7.3031, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 16.463860488428647, | |
| "learning_rate": 1.7333333333333336e-05, | |
| "loss": 6.958, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 11.114739607650908, | |
| "learning_rate": 1.866666666666667e-05, | |
| "loss": 7.328, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 8.560556168903155, | |
| "learning_rate": 2e-05, | |
| "loss": 6.8535, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 11.543963578227485, | |
| "learning_rate": 1.999979021001399e-05, | |
| "loss": 7.1039, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 13.024267824258278, | |
| "learning_rate": 1.999916084885832e-05, | |
| "loss": 7.03, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 24.347656967541983, | |
| "learning_rate": 1.9998111942939727e-05, | |
| "loss": 7.1369, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 13.978477853833892, | |
| "learning_rate": 1.9996643536268202e-05, | |
| "loss": 7.1052, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 9.01222616499256, | |
| "learning_rate": 1.9994755690455154e-05, | |
| "loss": 7.2016, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 8.338853855265398, | |
| "learning_rate": 1.99924484847108e-05, | |
| "loss": 7.0447, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 14.599634725835617, | |
| "learning_rate": 1.998972201584088e-05, | |
| "loss": 7.0037, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 8.332198683157735, | |
| "learning_rate": 1.9986576398242566e-05, | |
| "loss": 7.1784, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 14.23890995116397, | |
| "learning_rate": 1.9983011763899674e-05, | |
| "loss": 6.8922, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 8.923950492978154, | |
| "learning_rate": 1.997902826237712e-05, | |
| "loss": 6.9407, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 8.06378020962777, | |
| "learning_rate": 1.997462606081465e-05, | |
| "loss": 7.1025, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 11.24345210949927, | |
| "learning_rate": 1.9969805343919822e-05, | |
| "loss": 6.9762, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 8.092214633553768, | |
| "learning_rate": 1.9964566313960265e-05, | |
| "loss": 7.0578, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 12.87228727478449, | |
| "learning_rate": 1.995890919075519e-05, | |
| "loss": 6.8295, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 7.111093086376143, | |
| "learning_rate": 1.995283421166614e-05, | |
| "loss": 6.7973, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 8.632087862294554, | |
| "learning_rate": 1.9946341631587086e-05, | |
| "loss": 6.7752, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 8.994822315110934, | |
| "learning_rate": 1.9939431722933678e-05, | |
| "loss": 7.0197, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 7.486951908393647, | |
| "learning_rate": 1.9932104775631847e-05, | |
| "loss": 6.8283, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 7.4584761693908375, | |
| "learning_rate": 1.9924361097105624e-05, | |
| "loss": 6.8002, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 6.261343208572681, | |
| "learning_rate": 1.9916201012264255e-05, | |
| "loss": 6.6381, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 18.79112809499115, | |
| "learning_rate": 1.990762486348855e-05, | |
| "loss": 7.01, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 7.150915500939425, | |
| "learning_rate": 1.989863301061654e-05, | |
| "loss": 6.9012, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 6.738925101776216, | |
| "learning_rate": 1.9889225830928365e-05, | |
| "loss": 6.7694, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 6.944529789995812, | |
| "learning_rate": 1.987940371913044e-05, | |
| "loss": 6.799, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 7.556349878373683, | |
| "learning_rate": 1.9869167087338908e-05, | |
| "loss": 6.6428, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 6.611474483622349, | |
| "learning_rate": 1.9858516365062334e-05, | |
| "loss": 6.6243, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 6.916298787571706, | |
| "learning_rate": 1.9847451999183692e-05, | |
| "loss": 6.6087, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 8.457565509913838, | |
| "learning_rate": 1.9835974453941623e-05, | |
| "loss": 6.7551, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 8.540062078386342, | |
| "learning_rate": 1.9824084210910924e-05, | |
| "loss": 6.6972, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 7.4270052163791185, | |
| "learning_rate": 1.9811781768982392e-05, | |
| "loss": 6.9304, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 7.31190391936824, | |
| "learning_rate": 1.9799067644341844e-05, | |
| "loss": 6.7148, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 6.181290414236328, | |
| "learning_rate": 1.978594237044849e-05, | |
| "loss": 6.6086, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 6.059093545289046, | |
| "learning_rate": 1.977240649801253e-05, | |
| "loss": 6.5689, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 7.376855166703994, | |
| "learning_rate": 1.9758460594972068e-05, | |
| "loss": 6.5767, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 9.315509966882654, | |
| "learning_rate": 1.9744105246469264e-05, | |
| "loss": 6.601, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 7.110322978482923, | |
| "learning_rate": 1.9729341054825783e-05, | |
| "loss": 6.5206, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 8.348814181636438, | |
| "learning_rate": 1.9714168639517543e-05, | |
| "loss": 6.6781, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 7.889124746527842, | |
| "learning_rate": 1.9698588637148705e-05, | |
| "loss": 6.363, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 6.048373264000878, | |
| "learning_rate": 1.9682601701424958e-05, | |
| "loss": 6.5605, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 6.666923562210865, | |
| "learning_rate": 1.9666208503126115e-05, | |
| "loss": 6.5264, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 6.215144195698015, | |
| "learning_rate": 1.9649409730077934e-05, | |
| "loss": 6.7055, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 5.851763399842508, | |
| "learning_rate": 1.9632206087123296e-05, | |
| "loss": 6.3156, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 6.899818690987695, | |
| "learning_rate": 1.9614598296092603e-05, | |
| "loss": 6.5751, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 6.233556175977335, | |
| "learning_rate": 1.9596587095773496e-05, | |
| "loss": 6.3861, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 6.677665311671007, | |
| "learning_rate": 1.957817324187987e-05, | |
| "loss": 6.5699, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 6.655226097949778, | |
| "learning_rate": 1.9559357507020163e-05, | |
| "loss": 6.5021, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 7.47281811823693, | |
| "learning_rate": 1.9540140680664915e-05, | |
| "loss": 6.5324, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 5.6360351804666475, | |
| "learning_rate": 1.952052356911368e-05, | |
| "loss": 6.6096, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 5.4238841079756615, | |
| "learning_rate": 1.950050699546116e-05, | |
| "loss": 6.2913, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 6.691271836321062, | |
| "learning_rate": 1.9480091799562706e-05, | |
| "loss": 6.6404, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 5.779952770675111, | |
| "learning_rate": 1.9459278837999048e-05, | |
| "loss": 6.5131, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 5.75641781132488, | |
| "learning_rate": 1.9438068984040366e-05, | |
| "loss": 6.4102, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 5.473240945440776, | |
| "learning_rate": 1.9416463127609655e-05, | |
| "loss": 6.1943, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 5.639227300435135, | |
| "learning_rate": 1.9394462175245382e-05, | |
| "loss": 6.2661, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 6.359397311850186, | |
| "learning_rate": 1.937206705006344e-05, | |
| "loss": 6.3749, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 5.9168823793145835, | |
| "learning_rate": 1.9349278691718426e-05, | |
| "loss": 6.4857, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 5.934252295974776, | |
| "learning_rate": 1.9326098056364224e-05, | |
| "loss": 6.4401, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 5.184345891936555, | |
| "learning_rate": 1.9302526116613863e-05, | |
| "loss": 6.272, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 6.42618267064823, | |
| "learning_rate": 1.9278563861498726e-05, | |
| "loss": 6.4017, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 5.698764300135597, | |
| "learning_rate": 1.9254212296427043e-05, | |
| "loss": 6.422, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 5.515269918748001, | |
| "learning_rate": 1.922947244314172e-05, | |
| "loss": 6.4018, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 5.186220316136429, | |
| "learning_rate": 1.9204345339677442e-05, | |
| "loss": 6.4008, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 6.071075169984412, | |
| "learning_rate": 1.9178832040317153e-05, | |
| "loss": 6.236, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 5.211523270674776, | |
| "learning_rate": 1.91529336155478e-05, | |
| "loss": 6.2431, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 4.654030406820622, | |
| "learning_rate": 1.9126651152015404e-05, | |
| "loss": 6.1377, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 5.542618960137867, | |
| "learning_rate": 1.9099985752479505e-05, | |
| "loss": 6.3531, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 4.807919282457549, | |
| "learning_rate": 1.9072938535766864e-05, | |
| "loss": 6.2508, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 4.852730067283944, | |
| "learning_rate": 1.904551063672452e-05, | |
| "loss": 6.2348, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 4.925798203153842, | |
| "learning_rate": 1.9017703206172187e-05, | |
| "loss": 6.2765, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 5.0975569371666865, | |
| "learning_rate": 1.8989517410853956e-05, | |
| "loss": 6.2451, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 4.699489385610056, | |
| "learning_rate": 1.896095443338935e-05, | |
| "loss": 6.2578, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 5.660080834025456, | |
| "learning_rate": 1.8932015472223692e-05, | |
| "loss": 6.2843, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 4.9661954295884225, | |
| "learning_rate": 1.8902701741577844e-05, | |
| "loss": 6.1715, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 4.888648288627258, | |
| "learning_rate": 1.8873014471397225e-05, | |
| "loss": 6.2495, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 4.710120588181978, | |
| "learning_rate": 1.8842954907300236e-05, | |
| "loss": 5.9913, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 5.155284987315298, | |
| "learning_rate": 1.881252431052599e-05, | |
| "loss": 6.3151, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 5.013674756768704, | |
| "learning_rate": 1.8781723957881374e-05, | |
| "loss": 6.2286, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 4.011825376784628, | |
| "learning_rate": 1.87505551416875e-05, | |
| "loss": 6.2459, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 4.835636123417131, | |
| "learning_rate": 1.871901916972547e-05, | |
| "loss": 6.2033, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 5.171704259527532, | |
| "learning_rate": 1.8687117365181514e-05, | |
| "loss": 6.2608, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 5.181750952717401, | |
| "learning_rate": 1.865485106659145e-05, | |
| "loss": 6.0338, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 4.687033388755913, | |
| "learning_rate": 1.862222162778454e-05, | |
| "loss": 6.1161, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 5.061452310176261, | |
| "learning_rate": 1.85892304178267e-05, | |
| "loss": 6.1018, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 4.742767764167209, | |
| "learning_rate": 1.8555878820963014e-05, | |
| "loss": 6.2337, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 5.294660524784012, | |
| "learning_rate": 1.8522168236559693e-05, | |
| "loss": 6.2453, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 4.1375219719939915, | |
| "learning_rate": 1.8488100079045345e-05, | |
| "loss": 6.1595, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 4.537495039210601, | |
| "learning_rate": 1.8453675777851627e-05, | |
| "loss": 6.1449, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 5.402914196313115, | |
| "learning_rate": 1.8418896777353272e-05, | |
| "loss": 6.2237, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 4.1771424603623615, | |
| "learning_rate": 1.8383764536807486e-05, | |
| "loss": 6.1728, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 4.198851233644863, | |
| "learning_rate": 1.8348280530292712e-05, | |
| "loss": 6.0628, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 4.559214073625153, | |
| "learning_rate": 1.831244624664681e-05, | |
| "loss": 6.2743, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 3.97241616226669, | |
| "learning_rate": 1.827626318940454e-05, | |
| "loss": 6.1513, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 4.029464327696206, | |
| "learning_rate": 1.8239732876734525e-05, | |
| "loss": 6.1743, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 4.9362284691818905, | |
| "learning_rate": 1.8202856841375517e-05, | |
| "loss": 6.0945, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 5.525591589138074, | |
| "learning_rate": 1.816563663057211e-05, | |
| "loss": 6.2648, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 3.913078134221513, | |
| "learning_rate": 1.81280738060098e-05, | |
| "loss": 6.0767, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 4.539776653253226, | |
| "learning_rate": 1.8090169943749477e-05, | |
| "loss": 6.2008, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 7.07561003185471, | |
| "learning_rate": 1.8051926634161282e-05, | |
| "loss": 6.1817, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 4.4932679628420376, | |
| "learning_rate": 1.8013345481857903e-05, | |
| "loss": 6.0112, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 4.728104224718154, | |
| "learning_rate": 1.797442810562721e-05, | |
| "loss": 6.1642, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 5.010653662783029, | |
| "learning_rate": 1.793517613836437e-05, | |
| "loss": 6.0345, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 4.58293874507086, | |
| "learning_rate": 1.7895591227003316e-05, | |
| "loss": 6.0006, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 4.357889872637869, | |
| "learning_rate": 1.7855675032447648e-05, | |
| "loss": 6.0577, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 4.095191120284703, | |
| "learning_rate": 1.7815429229500946e-05, | |
| "loss": 5.9336, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 4.736888424569404, | |
| "learning_rate": 1.7774855506796497e-05, | |
| "loss": 5.961, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 9.32933189803368, | |
| "learning_rate": 1.7733955566726438e-05, | |
| "loss": 5.8729, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 4.352013234575603, | |
| "learning_rate": 1.7692731125370355e-05, | |
| "loss": 5.9631, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 4.0044514035329986, | |
| "learning_rate": 1.7651183912423228e-05, | |
| "loss": 6.0128, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 4.158183905120493, | |
| "learning_rate": 1.7609315671122912e-05, | |
| "loss": 6.0247, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 4.257943967689872, | |
| "learning_rate": 1.7567128158176955e-05, | |
| "loss": 6.2873, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 4.1763697860525655, | |
| "learning_rate": 1.7524623143688905e-05, | |
| "loss": 5.9392, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 3.9607955760999056, | |
| "learning_rate": 1.748180241108404e-05, | |
| "loss": 5.8487, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 4.0896919573576875, | |
| "learning_rate": 1.7438667757034547e-05, | |
| "loss": 6.0042, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 4.282944803020933, | |
| "learning_rate": 1.739522099138411e-05, | |
| "loss": 6.1652, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 4.13081574380951, | |
| "learning_rate": 1.7351463937072008e-05, | |
| "loss": 6.0233, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 4.96960839892131, | |
| "learning_rate": 1.7307398430056595e-05, | |
| "loss": 5.7853, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 4.459520148451556, | |
| "learning_rate": 1.72630263192383e-05, | |
| "loss": 5.9414, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 3.7882596771040893, | |
| "learning_rate": 1.7218349466382024e-05, | |
| "loss": 5.8131, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 4.207888044625232, | |
| "learning_rate": 1.7173369746039026e-05, | |
| "loss": 5.8724, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 3.6658140789192695, | |
| "learning_rate": 1.7128089045468294e-05, | |
| "loss": 5.9003, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 3.7722042701323493, | |
| "learning_rate": 1.7082509264557333e-05, | |
| "loss": 5.8756, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 4.8972909846842185, | |
| "learning_rate": 1.7036632315742464e-05, | |
| "loss": 6.0623, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 3.703243911637958, | |
| "learning_rate": 1.6990460123928577e-05, | |
| "loss": 5.9731, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 4.040351904365242, | |
| "learning_rate": 1.6943994626408365e-05, | |
| "loss": 6.0629, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 7.5014912670407865, | |
| "learning_rate": 1.6897237772781046e-05, | |
| "loss": 5.945, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 4.183607320000293, | |
| "learning_rate": 1.6850191524870548e-05, | |
| "loss": 5.906, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 4.441179437405616, | |
| "learning_rate": 1.6802857856643214e-05, | |
| "loss": 5.9791, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 3.5946652188893187, | |
| "learning_rate": 1.6755238754124965e-05, | |
| "loss": 5.8481, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 5.030328438406491, | |
| "learning_rate": 1.6707336215317968e-05, | |
| "loss": 6.047, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 4.048177117370896, | |
| "learning_rate": 1.665915225011681e-05, | |
| "loss": 5.9285, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 4.299948019967649, | |
| "learning_rate": 1.6610688880224178e-05, | |
| "loss": 5.8888, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 4.095923631849489, | |
| "learning_rate": 1.6561948139065997e-05, | |
| "loss": 6.2453, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 3.481920812397553, | |
| "learning_rate": 1.6512932071706153e-05, | |
| "loss": 6.0019, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 4.3296848974625135, | |
| "learning_rate": 1.646364273476067e-05, | |
| "loss": 5.8587, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 3.947321316567111, | |
| "learning_rate": 1.6414082196311402e-05, | |
| "loss": 5.928, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 4.7243925766736705, | |
| "learning_rate": 1.6364252535819284e-05, | |
| "loss": 5.9007, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 13.974458344598341, | |
| "learning_rate": 1.6314155844037074e-05, | |
| "loss": 5.9754, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 4.17123392978259, | |
| "learning_rate": 1.626379422292162e-05, | |
| "loss": 5.8421, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 4.043820315204433, | |
| "learning_rate": 1.6213169785545688e-05, | |
| "loss": 5.9519, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 3.526011441047756, | |
| "learning_rate": 1.6162284656009276e-05, | |
| "loss": 5.8076, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 3.8459147008292724, | |
| "learning_rate": 1.6111140969350504e-05, | |
| "loss": 5.7672, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 3.562828950103298, | |
| "learning_rate": 1.6059740871456035e-05, | |
| "loss": 5.898, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 4.431960095480538, | |
| "learning_rate": 1.6008086518971037e-05, | |
| "loss": 5.7521, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 3.6428660471557737, | |
| "learning_rate": 1.5956180079208684e-05, | |
| "loss": 5.953, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 3.767661352257826, | |
| "learning_rate": 1.5904023730059227e-05, | |
| "loss": 6.0916, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 3.660701500822884, | |
| "learning_rate": 1.5851619659898623e-05, | |
| "loss": 5.8085, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 3.6310726747437143, | |
| "learning_rate": 1.57989700674967e-05, | |
| "loss": 5.8102, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 4.0023266213419495, | |
| "learning_rate": 1.5746077161924905e-05, | |
| "loss": 5.8879, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 3.99237095864865, | |
| "learning_rate": 1.5692943162463628e-05, | |
| "loss": 5.8366, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 3.5149347856437703, | |
| "learning_rate": 1.5639570298509067e-05, | |
| "loss": 5.7649, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 3.301730499203444, | |
| "learning_rate": 1.5585960809479698e-05, | |
| "loss": 5.9754, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 3.462115038986119, | |
| "learning_rate": 1.5532116944722308e-05, | |
| "loss": 5.6883, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 4.004970215448472, | |
| "learning_rate": 1.547804096341763e-05, | |
| "loss": 5.8396, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 3.944647631447267, | |
| "learning_rate": 1.5423735134485537e-05, | |
| "loss": 5.9988, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 3.5900984593685723, | |
| "learning_rate": 1.536920173648984e-05, | |
| "loss": 5.7239, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 3.6285904810079574, | |
| "learning_rate": 1.5314443057542703e-05, | |
| "loss": 5.8721, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 3.8894726151488843, | |
| "learning_rate": 1.5259461395208628e-05, | |
| "loss": 5.8299, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 3.5924238969723072, | |
| "learning_rate": 1.5204259056408046e-05, | |
| "loss": 5.9516, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 5.538554836605968, | |
| "learning_rate": 1.5148838357320537e-05, | |
| "loss": 5.8963, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 3.560612079271743, | |
| "learning_rate": 1.5093201623287631e-05, | |
| "loss": 5.888, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 3.467820770014584, | |
| "learning_rate": 1.5037351188715265e-05, | |
| "loss": 5.7864, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 4.172673845567331, | |
| "learning_rate": 1.4981289396975818e-05, | |
| "loss": 5.8498, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 3.1779411060929816, | |
| "learning_rate": 1.4925018600309784e-05, | |
| "loss": 5.6926, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 3.782675685637412, | |
| "learning_rate": 1.4868541159727097e-05, | |
| "loss": 5.6674, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 3.7799393833784296, | |
| "learning_rate": 1.4811859444908053e-05, | |
| "loss": 5.8661, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 4.487305708943541, | |
| "learning_rate": 1.4754975834103877e-05, | |
| "loss": 5.6999, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 6.107952099039134, | |
| "learning_rate": 1.4697892714036959e-05, | |
| "loss": 5.9596, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 3.897845235029772, | |
| "learning_rate": 1.4640612479800686e-05, | |
| "loss": 5.8336, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 3.574618642064676, | |
| "learning_rate": 1.4583137534758968e-05, | |
| "loss": 5.678, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 3.2767605854680717, | |
| "learning_rate": 1.4525470290445392e-05, | |
| "loss": 5.7779, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 3.3494797594069343, | |
| "learning_rate": 1.4467613166462024e-05, | |
| "loss": 5.6117, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 3.4560187160140114, | |
| "learning_rate": 1.4409568590377918e-05, | |
| "loss": 5.9744, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 5.124637343157441, | |
| "learning_rate": 1.4351338997627233e-05, | |
| "loss": 5.8884, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 4.802869734676735, | |
| "learning_rate": 1.429292683140706e-05, | |
| "loss": 5.8834, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 4.830747863900574, | |
| "learning_rate": 1.4234334542574906e-05, | |
| "loss": 5.8741, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 3.0222266779531313, | |
| "learning_rate": 1.4175564589545853e-05, | |
| "loss": 5.7957, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 3.7517853962161003, | |
| "learning_rate": 1.411661943818944e-05, | |
| "loss": 5.773, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 3.481751429815316, | |
| "learning_rate": 1.4057501561726157e-05, | |
| "loss": 5.79, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 4.032738430652106, | |
| "learning_rate": 1.3998213440623691e-05, | |
| "loss": 5.8233, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 4.3048433576019045, | |
| "learning_rate": 1.3938757562492873e-05, | |
| "loss": 5.7213, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 3.42093480136207, | |
| "learning_rate": 1.3879136421983265e-05, | |
| "loss": 5.6734, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 3.372940690507083, | |
| "learning_rate": 1.3819352520678519e-05, | |
| "loss": 5.8204, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 3.866414604706453, | |
| "learning_rate": 1.3759408366991391e-05, | |
| "loss": 5.7685, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 3.0437497252782015, | |
| "learning_rate": 1.3699306476058523e-05, | |
| "loss": 5.9321, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 3.5806152918043153, | |
| "learning_rate": 1.3639049369634878e-05, | |
| "loss": 5.6817, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 3.652624509453939, | |
| "learning_rate": 1.357863957598796e-05, | |
| "loss": 5.6123, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 3.2444221493972063, | |
| "learning_rate": 1.3518079629791725e-05, | |
| "loss": 5.5523, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 7.46063417126853, | |
| "learning_rate": 1.345737207202023e-05, | |
| "loss": 5.6855, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 3.3218441227109787, | |
| "learning_rate": 1.3396519449841006e-05, | |
| "loss": 5.7366, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 4.5944315284742725, | |
| "learning_rate": 1.3335524316508208e-05, | |
| "loss": 5.7278, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 3.0069349104091603, | |
| "learning_rate": 1.3274389231255466e-05, | |
| "loss": 5.7355, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 3.9317190820703756, | |
| "learning_rate": 1.3213116759188525e-05, | |
| "loss": 5.6806, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 3.33285104772104, | |
| "learning_rate": 1.3151709471177589e-05, | |
| "loss": 5.7708, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 3.6380751570471332, | |
| "learning_rate": 1.3090169943749475e-05, | |
| "loss": 5.8772, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 3.742526748849025, | |
| "learning_rate": 1.3028500758979507e-05, | |
| "loss": 5.6913, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 3.3393134329516196, | |
| "learning_rate": 1.296670450438317e-05, | |
| "loss": 5.5817, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 3.887935011194972, | |
| "learning_rate": 1.2904783772807534e-05, | |
| "loss": 5.6364, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 3.5742955487496175, | |
| "learning_rate": 1.2842741162322487e-05, | |
| "loss": 5.6486, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 3.2697917373453245, | |
| "learning_rate": 1.2780579276111702e-05, | |
| "loss": 5.7063, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 3.401421083396212, | |
| "learning_rate": 1.2718300722363431e-05, | |
| "loss": 5.7316, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 3.6121121716149887, | |
| "learning_rate": 1.2655908114161053e-05, | |
| "loss": 5.7758, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 3.2222903292023406, | |
| "learning_rate": 1.2593404069373452e-05, | |
| "loss": 5.7089, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 3.025802281634462, | |
| "learning_rate": 1.2530791210545163e-05, | |
| "loss": 5.6953, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 3.4673520664847075, | |
| "learning_rate": 1.2468072164786342e-05, | |
| "loss": 5.8728, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 3.6806603608994735, | |
| "learning_rate": 1.2405249563662539e-05, | |
| "loss": 5.669, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 4.64700420741913, | |
| "learning_rate": 1.2342326043084268e-05, | |
| "loss": 5.7352, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 3.0790982186864504, | |
| "learning_rate": 1.2279304243196438e-05, | |
| "loss": 5.5999, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 3.24646862661049, | |
| "learning_rate": 1.2216186808267544e-05, | |
| "loss": 5.668, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 3.1831196277858864, | |
| "learning_rate": 1.215297638657875e-05, | |
| "loss": 5.7959, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 3.4421555634435035, | |
| "learning_rate": 1.2089675630312755e-05, | |
| "loss": 5.6636, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 2.833448990477623, | |
| "learning_rate": 1.2026287195442503e-05, | |
| "loss": 5.5742, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 3.221649072609624, | |
| "learning_rate": 1.1962813741619777e-05, | |
| "loss": 5.7345, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 4.5576181124301, | |
| "learning_rate": 1.189925793206357e-05, | |
| "loss": 5.8758, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 3.1198515608939252, | |
| "learning_rate": 1.1835622433448361e-05, | |
| "loss": 5.7026, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 3.2391737753303063, | |
| "learning_rate": 1.177190991579223e-05, | |
| "loss": 5.8245, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 3.0415855629042006, | |
| "learning_rate": 1.1708123052344803e-05, | |
| "loss": 5.7423, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 3.417436664351836, | |
| "learning_rate": 1.164426451947513e-05, | |
| "loss": 5.7224, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 3.0428028654941963, | |
| "learning_rate": 1.1580336996559343e-05, | |
| "loss": 5.5215, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 3.2412878897803843, | |
| "learning_rate": 1.151634316586828e-05, | |
| "loss": 5.6753, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 2.9715981798900173, | |
| "learning_rate": 1.1452285712454905e-05, | |
| "loss": 5.6633, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 3.3507364284363375, | |
| "learning_rate": 1.138816732404167e-05, | |
| "loss": 5.7531, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 3.695763982565934, | |
| "learning_rate": 1.1323990690907734e-05, | |
| "loss": 5.6431, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 3.442076843360669, | |
| "learning_rate": 1.1259758505776092e-05, | |
| "loss": 5.763, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 3.4221454818185797, | |
| "learning_rate": 1.119547346370059e-05, | |
| "loss": 5.7121, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 32.81366795375805, | |
| "learning_rate": 1.1131138261952845e-05, | |
| "loss": 5.7642, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 2.824713348661609, | |
| "learning_rate": 1.1066755599909065e-05, | |
| "loss": 5.5872, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 3.2122840787719356, | |
| "learning_rate": 1.1002328178936813e-05, | |
| "loss": 5.6909, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 2.914139469492162, | |
| "learning_rate": 1.0937858702281631e-05, | |
| "loss": 5.5502, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 3.372671623706637, | |
| "learning_rate": 1.087334987495364e-05, | |
| "loss": 5.638, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 2.9794994417982705, | |
| "learning_rate": 1.0808804403614044e-05, | |
| "loss": 5.7327, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 3.545219321353326, | |
| "learning_rate": 1.0744224996461541e-05, | |
| "loss": 5.773, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 3.0509122712537904, | |
| "learning_rate": 1.0679614363118718e-05, | |
| "loss": 5.4833, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 3.1328946212391573, | |
| "learning_rate": 1.061497521451835e-05, | |
| "loss": 5.5846, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 2.912951012800422, | |
| "learning_rate": 1.055031026278965e-05, | |
| "loss": 5.7325, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 3.190063789344824, | |
| "learning_rate": 1.0485622221144485e-05, | |
| "loss": 5.6504, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 4.059074907434253, | |
| "learning_rate": 1.0420913803763522e-05, | |
| "loss": 5.5256, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 3.2665977095269763, | |
| "learning_rate": 1.0356187725682359e-05, | |
| "loss": 5.6545, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 2.8731971309590723, | |
| "learning_rate": 1.0291446702677598e-05, | |
| "loss": 5.4238, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 5.960836799262505, | |
| "learning_rate": 1.02266934511529e-05, | |
| "loss": 5.601, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 3.479663130288781, | |
| "learning_rate": 1.0161930688025018e-05, | |
| "loss": 5.49, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 3.458943691824299, | |
| "learning_rate": 1.0097161130609774e-05, | |
| "loss": 5.7461, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 3.279222709074338, | |
| "learning_rate": 1.003238749650809e-05, | |
| "loss": 5.5439, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 5.311427762294207, | |
| "learning_rate": 9.967612503491915e-06, | |
| "loss": 5.5148, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 2.887421444069423, | |
| "learning_rate": 9.90283886939023e-06, | |
| "loss": 5.4186, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 4.401386334484841, | |
| "learning_rate": 9.838069311974986e-06, | |
| "loss": 5.6778, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 2.6023206995102135, | |
| "learning_rate": 9.773306548847102e-06, | |
| "loss": 5.6155, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 3.593734709591131, | |
| "learning_rate": 9.708553297322407e-06, | |
| "loss": 5.6777, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 3.477325665085784, | |
| "learning_rate": 9.643812274317644e-06, | |
| "loss": 5.5781, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 2.8887444330887266, | |
| "learning_rate": 9.579086196236483e-06, | |
| "loss": 5.7172, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 3.8360288485375484, | |
| "learning_rate": 9.514377778855521e-06, | |
| "loss": 5.5803, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 3.135597210733372, | |
| "learning_rate": 9.449689737210352e-06, | |
| "loss": 5.7742, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 3.0578215352523745, | |
| "learning_rate": 9.385024785481653e-06, | |
| "loss": 5.5618, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 2.925737489172214, | |
| "learning_rate": 9.320385636881283e-06, | |
| "loss": 5.6728, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 3.429796358511194, | |
| "learning_rate": 9.255775003538462e-06, | |
| "loss": 5.7857, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 3.1866968840925742, | |
| "learning_rate": 9.19119559638596e-06, | |
| "loss": 5.504, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 3.199919489689139, | |
| "learning_rate": 9.126650125046361e-06, | |
| "loss": 5.6548, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 6.644972124408221, | |
| "learning_rate": 9.062141297718372e-06, | |
| "loss": 5.7688, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 2.840251387463206, | |
| "learning_rate": 8.99767182106319e-06, | |
| "loss": 5.5356, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 2.847024195497787, | |
| "learning_rate": 8.933244400090937e-06, | |
| "loss": 5.538, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 4.300138770028974, | |
| "learning_rate": 8.868861738047158e-06, | |
| "loss": 5.5466, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 3.2598992072099806, | |
| "learning_rate": 8.804526536299413e-06, | |
| "loss": 5.5438, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 2.9223921302516205, | |
| "learning_rate": 8.740241494223911e-06, | |
| "loss": 5.5199, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 2.7228943907903944, | |
| "learning_rate": 8.676009309092273e-06, | |
| "loss": 5.316, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 3.151284881347976, | |
| "learning_rate": 8.611832675958335e-06, | |
| "loss": 5.605, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 2.971020945163525, | |
| "learning_rate": 8.5477142875451e-06, | |
| "loss": 5.5074, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 3.383598237834473, | |
| "learning_rate": 8.48365683413172e-06, | |
| "loss": 5.6454, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 3.1708795210472065, | |
| "learning_rate": 8.419663003440657e-06, | |
| "loss": 5.6293, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 3.0060809846941523, | |
| "learning_rate": 8.355735480524874e-06, | |
| "loss": 5.7011, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 3.082555191182687, | |
| "learning_rate": 8.291876947655197e-06, | |
| "loss": 5.5058, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 3.1290655800212988, | |
| "learning_rate": 8.228090084207773e-06, | |
| "loss": 5.6645, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 3.6512062285821107, | |
| "learning_rate": 8.16437756655164e-06, | |
| "loss": 5.632, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 5.580032474883606, | |
| "learning_rate": 8.100742067936432e-06, | |
| "loss": 5.6529, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 2.8279735010912828, | |
| "learning_rate": 8.037186258380226e-06, | |
| "loss": 5.653, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 3.0485365662315185, | |
| "learning_rate": 7.9737128045575e-06, | |
| "loss": 5.6271, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 2.9000686534385065, | |
| "learning_rate": 7.91032436968725e-06, | |
| "loss": 5.497, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 3.7477894288698823, | |
| "learning_rate": 7.847023613421251e-06, | |
| "loss": 5.6454, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 3.1406780416239988, | |
| "learning_rate": 7.78381319173246e-06, | |
| "loss": 5.702, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 2.8819040313896664, | |
| "learning_rate": 7.720695756803569e-06, | |
| "loss": 5.5487, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 3.1093053507455433, | |
| "learning_rate": 7.657673956915735e-06, | |
| "loss": 5.6263, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 3.532080902497016, | |
| "learning_rate": 7.594750436337467e-06, | |
| "loss": 5.7086, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 2.9271604989314417, | |
| "learning_rate": 7.531927835213657e-06, | |
| "loss": 5.5651, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 3.9573677159990344, | |
| "learning_rate": 7.469208789454838e-06, | |
| "loss": 5.6534, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 2.813452933446664, | |
| "learning_rate": 7.40659593062655e-06, | |
| "loss": 5.5948, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 3.64918124403021, | |
| "learning_rate": 7.344091885838949e-06, | |
| "loss": 5.833, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 8.447026127108531, | |
| "learning_rate": 7.2816992776365714e-06, | |
| "loss": 5.6345, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 3.487783377440865, | |
| "learning_rate": 7.219420723888301e-06, | |
| "loss": 5.5796, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 3.3746557421600545, | |
| "learning_rate": 7.157258837677514e-06, | |
| "loss": 5.4475, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 3.109170816456849, | |
| "learning_rate": 7.095216227192467e-06, | |
| "loss": 5.4312, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 3.5209993320764634, | |
| "learning_rate": 7.033295495616834e-06, | |
| "loss": 5.6655, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 3.7885132212833215, | |
| "learning_rate": 6.971499241020495e-06, | |
| "loss": 5.5506, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 2.8035280155793902, | |
| "learning_rate": 6.909830056250527e-06, | |
| "loss": 5.5899, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 3.3096990722045665, | |
| "learning_rate": 6.848290528822417e-06, | |
| "loss": 5.5894, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 3.1035770416615245, | |
| "learning_rate": 6.786883240811479e-06, | |
| "loss": 5.5137, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 2.934891742489131, | |
| "learning_rate": 6.725610768744535e-06, | |
| "loss": 5.5028, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 3.371896210185273, | |
| "learning_rate": 6.664475683491797e-06, | |
| "loss": 5.5001, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 2.8035512763165222, | |
| "learning_rate": 6.603480550158995e-06, | |
| "loss": 5.6237, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 4.14158466178092, | |
| "learning_rate": 6.542627927979772e-06, | |
| "loss": 5.6736, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 2.972559988613375, | |
| "learning_rate": 6.481920370208274e-06, | |
| "loss": 5.6447, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 3.411613112123063, | |
| "learning_rate": 6.421360424012039e-06, | |
| "loss": 5.7263, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 2.8412153634814308, | |
| "learning_rate": 6.360950630365126e-06, | |
| "loss": 5.5072, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 3.376692187351766, | |
| "learning_rate": 6.300693523941481e-06, | |
| "loss": 5.6328, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 2.950706203004661, | |
| "learning_rate": 6.2405916330086106e-06, | |
| "loss": 5.606, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 3.6164997018204152, | |
| "learning_rate": 6.180647479321484e-06, | |
| "loss": 5.7159, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 2.6776792490163492, | |
| "learning_rate": 6.120863578016736e-06, | |
| "loss": 5.4034, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 2.8823500476883783, | |
| "learning_rate": 6.061242437507131e-06, | |
| "loss": 5.5799, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 2.7310153497627305, | |
| "learning_rate": 6.00178655937631e-06, | |
| "loss": 5.4403, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 2.8043161109289105, | |
| "learning_rate": 5.942498438273849e-06, | |
| "loss": 5.5201, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 2.6257376922078577, | |
| "learning_rate": 5.8833805618105635e-06, | |
| "loss": 5.5161, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 3.100365890406591, | |
| "learning_rate": 5.82443541045415e-06, | |
| "loss": 5.4591, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 3.164224570414931, | |
| "learning_rate": 5.765665457425102e-06, | |
| "loss": 5.5032, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 2.8688071534260713, | |
| "learning_rate": 5.707073168592943e-06, | |
| "loss": 5.6875, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 2.9334355868792925, | |
| "learning_rate": 5.648661002372769e-06, | |
| "loss": 5.4918, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 2.618203487365493, | |
| "learning_rate": 5.590431409622081e-06, | |
| "loss": 5.6324, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 2.7468055145560473, | |
| "learning_rate": 5.5323868335379775e-06, | |
| "loss": 5.5696, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 2.945148993510563, | |
| "learning_rate": 5.4745297095546125e-06, | |
| "loss": 5.6865, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 4.27286874096374, | |
| "learning_rate": 5.416862465241033e-06, | |
| "loss": 5.574, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 3.075467269510141, | |
| "learning_rate": 5.359387520199317e-06, | |
| "loss": 5.4386, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 12.979333973830173, | |
| "learning_rate": 5.302107285963045e-06, | |
| "loss": 5.5171, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 2.946796474047621, | |
| "learning_rate": 5.245024165896126e-06, | |
| "loss": 5.4337, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 47.491660622837905, | |
| "learning_rate": 5.18814055509195e-06, | |
| "loss": 5.3093, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 2.6684688936423306, | |
| "learning_rate": 5.131458840272905e-06, | |
| "loss": 5.4504, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 2.9524813775023664, | |
| "learning_rate": 5.074981399690219e-06, | |
| "loss": 5.6183, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 3.071692317071865, | |
| "learning_rate": 5.018710603024187e-06, | |
| "loss": 5.8018, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 2.8943426237790604, | |
| "learning_rate": 4.9626488112847384e-06, | |
| "loss": 5.5124, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 3.643486862482141, | |
| "learning_rate": 4.9067983767123736e-06, | |
| "loss": 5.6324, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 3.0446879911014793, | |
| "learning_rate": 4.851161642679466e-06, | |
| "loss": 5.5722, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 3.0223614188577583, | |
| "learning_rate": 4.795740943591955e-06, | |
| "loss": 5.6997, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 3.139884095364713, | |
| "learning_rate": 4.740538604791371e-06, | |
| "loss": 5.6098, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 3.0355695744448905, | |
| "learning_rate": 4.685556942457296e-06, | |
| "loss": 5.5642, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 3.2148124769621145, | |
| "learning_rate": 4.630798263510162e-06, | |
| "loss": 5.4686, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 6.599100924296892, | |
| "learning_rate": 4.576264865514467e-06, | |
| "loss": 5.6283, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 3.176903096047862, | |
| "learning_rate": 4.521959036582372e-06, | |
| "loss": 5.6207, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 2.7882598761964217, | |
| "learning_rate": 4.467883055277696e-06, | |
| "loss": 5.4235, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 3.4794420663108894, | |
| "learning_rate": 4.414039190520308e-06, | |
| "loss": 5.5631, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 2.903622461402484, | |
| "learning_rate": 4.360429701490935e-06, | |
| "loss": 5.6318, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 7.699211582040187, | |
| "learning_rate": 4.307056837536373e-06, | |
| "loss": 5.7193, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 3.637017761567808, | |
| "learning_rate": 4.2539228380750955e-06, | |
| "loss": 5.6718, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 3.871928827914661, | |
| "learning_rate": 4.201029932503303e-06, | |
| "loss": 5.5911, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 3.373623961352284, | |
| "learning_rate": 4.14838034010138e-06, | |
| "loss": 5.6983, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 2.7039462572520025, | |
| "learning_rate": 4.095976269940777e-06, | |
| "loss": 5.4657, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 2.906259270274048, | |
| "learning_rate": 4.043819920791322e-06, | |
| "loss": 5.6059, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 2.80215901884808, | |
| "learning_rate": 3.991913481028965e-06, | |
| "loss": 5.3867, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 2.7611740391342763, | |
| "learning_rate": 3.940259128543967e-06, | |
| "loss": 5.6265, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 3.496220876703193, | |
| "learning_rate": 3.888859030649498e-06, | |
| "loss": 5.7369, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 3.069627191877829, | |
| "learning_rate": 3.837715343990727e-06, | |
| "loss": 5.6878, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 2.8861264690211628, | |
| "learning_rate": 3.7868302144543146e-06, | |
| "loss": 5.6452, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 2.973826787518843, | |
| "learning_rate": 3.736205777078381e-06, | |
| "loss": 5.4028, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 2.9173591380715704, | |
| "learning_rate": 3.685844155962931e-06, | |
| "loss": 5.5656, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 3.147705831696598, | |
| "learning_rate": 3.63574746418072e-06, | |
| "loss": 5.5783, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 3.072248624553244, | |
| "learning_rate": 3.585917803688603e-06, | |
| "loss": 5.539, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 2.7322222965861336, | |
| "learning_rate": 3.536357265239333e-06, | |
| "loss": 5.4976, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 2.658079903087022, | |
| "learning_rate": 3.487067928293848e-06, | |
| "loss": 5.4221, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 2.680370794084422, | |
| "learning_rate": 3.4380518609340076e-06, | |
| "loss": 5.5785, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 2.89315062272381, | |
| "learning_rate": 3.3893111197758276e-06, | |
| "loss": 5.4699, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 3.088114743299729, | |
| "learning_rate": 3.3408477498831917e-06, | |
| "loss": 5.5848, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 3.3744405750175783, | |
| "learning_rate": 3.2926637846820366e-06, | |
| "loss": 5.5855, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 3.0949682435479815, | |
| "learning_rate": 3.2447612458750365e-06, | |
| "loss": 5.5665, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 3.5298918745049064, | |
| "learning_rate": 3.197142143356787e-06, | |
| "loss": 5.4282, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 2.3415078853618483, | |
| "learning_rate": 3.1498084751294523e-06, | |
| "loss": 5.548, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 2.6530303000947377, | |
| "learning_rate": 3.1027622272189572e-06, | |
| "loss": 5.4614, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 2.6198793225005095, | |
| "learning_rate": 3.0560053735916372e-06, | |
| "loss": 5.3387, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 3.253733908459485, | |
| "learning_rate": 3.009539876071427e-06, | |
| "loss": 5.5892, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 2.893078721591448, | |
| "learning_rate": 2.9633676842575386e-06, | |
| "loss": 5.5098, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 3.2332247538663927, | |
| "learning_rate": 2.9174907354426696e-06, | |
| "loss": 5.611, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 3.717694710480861, | |
| "learning_rate": 2.8719109545317102e-06, | |
| "loss": 5.7531, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 5.973350020162763, | |
| "learning_rate": 2.8266302539609747e-06, | |
| "loss": 5.5053, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 3.106939977808064, | |
| "learning_rate": 2.78165053361798e-06, | |
| "loss": 5.5299, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 3.743314132309298, | |
| "learning_rate": 2.736973680761702e-06, | |
| "loss": 5.6108, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 3.020284676653917, | |
| "learning_rate": 2.692601569943407e-06, | |
| "loss": 5.5949, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 2.8688896218347466, | |
| "learning_rate": 2.648536062927999e-06, | |
| "loss": 5.6542, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 2.8774214420614292, | |
| "learning_rate": 2.604779008615895e-06, | |
| "loss": 5.6006, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 2.5974922903934443, | |
| "learning_rate": 2.5613322429654573e-06, | |
| "loss": 5.4603, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 3.1182016556730034, | |
| "learning_rate": 2.5181975889159615e-06, | |
| "loss": 5.4617, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 3.02296875164515, | |
| "learning_rate": 2.475376856311097e-06, | |
| "loss": 5.5758, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 3.1443470043442066, | |
| "learning_rate": 2.432871841823047e-06, | |
| "loss": 5.4938, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 5.21437085093465, | |
| "learning_rate": 2.390684328877089e-06, | |
| "loss": 5.543, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 3.0664362463556603, | |
| "learning_rate": 2.3488160875767717e-06, | |
| "loss": 5.6436, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 2.793841886787081, | |
| "learning_rate": 2.307268874629649e-06, | |
| "loss": 5.3901, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 2.6568460536060137, | |
| "learning_rate": 2.266044433273562e-06, | |
| "loss": 5.5585, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 3.3923148492013504, | |
| "learning_rate": 2.2251444932035094e-06, | |
| "loss": 5.3515, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 4.187296832849823, | |
| "learning_rate": 2.184570770499056e-06, | |
| "loss": 5.4722, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 2.8884881509409555, | |
| "learning_rate": 2.1443249675523536e-06, | |
| "loss": 5.537, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 3.051680492849398, | |
| "learning_rate": 2.1044087729966856e-06, | |
| "loss": 5.4369, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 2.8280027110122608, | |
| "learning_rate": 2.064823861635633e-06, | |
| "loss": 5.5705, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 3.4037253520135216, | |
| "learning_rate": 2.025571894372794e-06, | |
| "loss": 5.5584, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 4.319540102434969, | |
| "learning_rate": 1.9866545181421016e-06, | |
| "loss": 5.4076, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 2.808044157619595, | |
| "learning_rate": 1.9480733658387175e-06, | |
| "loss": 5.4129, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 2.7351469976881306, | |
| "learning_rate": 1.9098300562505266e-06, | |
| "loss": 5.6271, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 2.7165917157682737, | |
| "learning_rate": 1.8719261939902023e-06, | |
| "loss": 5.4525, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 3.040380201846354, | |
| "learning_rate": 1.8343633694278895e-06, | |
| "loss": 5.5268, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 2.8167171485840035, | |
| "learning_rate": 1.7971431586244814e-06, | |
| "loss": 5.4247, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 2.8317598164442632, | |
| "learning_rate": 1.7602671232654755e-06, | |
| "loss": 5.4478, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 2.8994461696949276, | |
| "learning_rate": 1.723736810595461e-06, | |
| "loss": 5.6048, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 2.848027250263409, | |
| "learning_rate": 1.687553753353195e-06, | |
| "loss": 5.6236, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 2.7979223820102126, | |
| "learning_rate": 1.6517194697072903e-06, | |
| "loss": 5.5997, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 4.406985290020591, | |
| "learning_rate": 1.6162354631925203e-06, | |
| "loss": 5.6047, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 3.2716585943989362, | |
| "learning_rate": 1.5811032226467304e-06, | |
| "loss": 5.5763, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 3.2119530719491967, | |
| "learning_rate": 1.5463242221483742e-06, | |
| "loss": 5.5576, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 2.7525960595562, | |
| "learning_rate": 1.511899920954656e-06, | |
| "loss": 5.6124, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 2.575626661151134, | |
| "learning_rate": 1.4778317634403082e-06, | |
| "loss": 5.5585, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 3.109772746844713, | |
| "learning_rate": 1.4441211790369892e-06, | |
| "loss": 5.6023, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 3.48770216871105, | |
| "learning_rate": 1.4107695821733026e-06, | |
| "loss": 5.6905, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 2.776399817120803, | |
| "learning_rate": 1.3777783722154603e-06, | |
| "loss": 5.5346, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 2.689995264190314, | |
| "learning_rate": 1.3451489334085555e-06, | |
| "loss": 5.776, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 2.9589020357519855, | |
| "learning_rate": 1.3128826348184886e-06, | |
| "loss": 5.5889, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 3.926701278155943, | |
| "learning_rate": 1.2809808302745298e-06, | |
| "loss": 5.4377, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 8.614907788515115, | |
| "learning_rate": 1.249444858312502e-06, | |
| "loss": 5.3811, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 2.844919169807355, | |
| "learning_rate": 1.218276042118629e-06, | |
| "loss": 5.3662, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 3.200788157807446, | |
| "learning_rate": 1.1874756894740137e-06, | |
| "loss": 5.5491, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 2.8493126527965655, | |
| "learning_rate": 1.1570450926997657e-06, | |
| "loss": 5.6191, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 5.199235991842511, | |
| "learning_rate": 1.1269855286027798e-06, | |
| "loss": 5.6003, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 2.577166179252903, | |
| "learning_rate": 1.0972982584221592e-06, | |
| "loss": 5.5332, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 2.5080571141871837, | |
| "learning_rate": 1.067984527776309e-06, | |
| "loss": 5.5226, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 2.9156280169840563, | |
| "learning_rate": 1.0390455666106547e-06, | |
| "loss": 5.4879, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 3.3439938592575094, | |
| "learning_rate": 1.010482589146048e-06, | |
| "loss": 5.5742, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 2.70223178081651, | |
| "learning_rate": 9.822967938278172e-07, | |
| "loss": 5.53, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 2.910070747579829, | |
| "learning_rate": 9.544893632754816e-07, | |
| "loss": 5.5037, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 2.859773242243526, | |
| "learning_rate": 9.270614642331377e-07, | |
| "loss": 5.5309, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 2.990603251521546, | |
| "learning_rate": 9.000142475204965e-07, | |
| "loss": 5.4596, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 11.13467016828935, | |
| "learning_rate": 8.733488479845997e-07, | |
| "loss": 5.4875, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 3.499523656417884, | |
| "learning_rate": 8.470663844522053e-07, | |
| "loss": 5.5061, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 3.2234243851466418, | |
| "learning_rate": 8.211679596828481e-07, | |
| "loss": 5.5508, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 3.689783836568065, | |
| "learning_rate": 7.956546603225601e-07, | |
| "loss": 5.4979, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 3.550751273965273, | |
| "learning_rate": 7.705275568582848e-07, | |
| "loss": 5.6264, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 2.721521824889205, | |
| "learning_rate": 7.457877035729588e-07, | |
| "loss": 5.5113, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 2.768024715912084, | |
| "learning_rate": 7.21436138501278e-07, | |
| "loss": 5.5389, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 2.6088969137232456, | |
| "learning_rate": 6.974738833861383e-07, | |
| "loss": 5.4611, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 2.693525018419924, | |
| "learning_rate": 6.739019436357774e-07, | |
| "loss": 5.5317, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 2.5671066935530367, | |
| "learning_rate": 6.507213082815745e-07, | |
| "loss": 5.5331, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 3.4210107522987996, | |
| "learning_rate": 6.279329499365649e-07, | |
| "loss": 5.6906, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 2.772985182537546, | |
| "learning_rate": 6.055378247546217e-07, | |
| "loss": 5.6865, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 3.298986027174084, | |
| "learning_rate": 5.835368723903456e-07, | |
| "loss": 5.5719, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 2.715794146885051, | |
| "learning_rate": 5.619310159596358e-07, | |
| "loss": 5.6766, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 3.0933470216619243, | |
| "learning_rate": 5.407211620009545e-07, | |
| "loss": 5.5423, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 4.143643936396307, | |
| "learning_rate": 5.199082004372958e-07, | |
| "loss": 5.3755, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 2.868440457281347, | |
| "learning_rate": 4.994930045388414e-07, | |
| "loss": 5.5415, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 2.9560966126847372, | |
| "learning_rate": 4.794764308863242e-07, | |
| "loss": 5.636, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 3.658181273470064, | |
| "learning_rate": 4.5985931933508757e-07, | |
| "loss": 5.6803, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 3.423368722688064, | |
| "learning_rate": 4.406424929798403e-07, | |
| "loss": 5.4089, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 3.0534636300165188, | |
| "learning_rate": 4.218267581201296e-07, | |
| "loss": 5.5947, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 4.345779139508837, | |
| "learning_rate": 4.034129042265067e-07, | |
| "loss": 5.5494, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 2.733287289238342, | |
| "learning_rate": 3.8540170390740097e-07, | |
| "loss": 5.3952, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 2.578794557649385, | |
| "learning_rate": 3.67793912876705e-07, | |
| "loss": 5.4544, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 6.462067447086064, | |
| "learning_rate": 3.5059026992206645e-07, | |
| "loss": 5.4354, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 2.6125800075217755, | |
| "learning_rate": 3.3379149687388866e-07, | |
| "loss": 5.3488, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 2.7049049281222506, | |
| "learning_rate": 3.1739829857504235e-07, | |
| "loss": 5.5727, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 3.620043130272865, | |
| "learning_rate": 3.0141136285129825e-07, | |
| "loss": 5.6706, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 3.0809997678670187, | |
| "learning_rate": 2.8583136048245697e-07, | |
| "loss": 5.7061, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 3.1616078894179482, | |
| "learning_rate": 2.706589451742181e-07, | |
| "loss": 5.6098, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 2.6779003366529275, | |
| "learning_rate": 2.5589475353073987e-07, | |
| "loss": 5.4707, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 6.707485770015231, | |
| "learning_rate": 2.4153940502793185e-07, | |
| "loss": 5.4335, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 3.1605981177662383, | |
| "learning_rate": 2.2759350198746978e-07, | |
| "loss": 5.5259, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 3.014186395148381, | |
| "learning_rate": 2.1405762955151178e-07, | |
| "loss": 5.6467, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 2.889242425776768, | |
| "learning_rate": 2.009323556581566e-07, | |
| "loss": 5.4662, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 3.095182172474526, | |
| "learning_rate": 1.8821823101760949e-07, | |
| "loss": 5.5991, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 2.8025822861133, | |
| "learning_rate": 1.7591578908907724e-07, | |
| "loss": 5.6508, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 3.052158040260495, | |
| "learning_rate": 1.6402554605838173e-07, | |
| "loss": 5.6078, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 3.0334351565612603, | |
| "learning_rate": 1.5254800081630828e-07, | |
| "loss": 5.6521, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 3.2390067488864203, | |
| "learning_rate": 1.4148363493766803e-07, | |
| "loss": 5.5932, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 2.6513576443759246, | |
| "learning_rate": 1.30832912661093e-07, | |
| "loss": 5.4863, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 2.8858856433117337, | |
| "learning_rate": 1.2059628086956044e-07, | |
| "loss": 5.6502, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 6.026494893009913, | |
| "learning_rate": 1.1077416907163573e-07, | |
| "loss": 5.6161, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 3.269750470433105, | |
| "learning_rate": 1.0136698938346012e-07, | |
| "loss": 5.6146, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 2.915176554260305, | |
| "learning_rate": 9.237513651145224e-08, | |
| "loss": 5.4934, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 2.703171232414342, | |
| "learning_rate": 8.379898773574924e-08, | |
| "loss": 5.4725, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 2.955963983229779, | |
| "learning_rate": 7.563890289437825e-08, | |
| "loss": 5.7339, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 3.089224348103812, | |
| "learning_rate": 6.78952243681541e-08, | |
| "loss": 5.5206, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 3.1319980991851524, | |
| "learning_rate": 6.056827706632185e-08, | |
| "loss": 5.3931, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 3.2761777556196376, | |
| "learning_rate": 5.365836841291439e-08, | |
| "loss": 5.4341, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 2.8804231914855665, | |
| "learning_rate": 4.716578833386054e-08, | |
| "loss": 5.3863, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 2.2873446601496052, | |
| "learning_rate": 4.109080924481479e-08, | |
| "loss": 5.4484, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 3.110318580535298, | |
| "learning_rate": 3.543368603973529e-08, | |
| "loss": 5.5471, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 3.791270347470137, | |
| "learning_rate": 3.019465608018024e-08, | |
| "loss": 5.5891, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 2.6716828750765056, | |
| "learning_rate": 2.537393918535358e-08, | |
| "loss": 5.5102, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 4.590128972359705, | |
| "learning_rate": 2.0971737622883515e-08, | |
| "loss": 5.6497, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 2.491730237188739, | |
| "learning_rate": 1.698823610032929e-08, | |
| "loss": 5.5464, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 2.541793206348526, | |
| "learning_rate": 1.3423601757436289e-08, | |
| "loss": 5.5446, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 4.43180997277326, | |
| "learning_rate": 1.0277984159122734e-08, | |
| "loss": 5.5588, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 2.7657573951340706, | |
| "learning_rate": 7.551515289203615e-09, | |
| "loss": 5.6594, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 2.8695545506323783, | |
| "learning_rate": 5.2443095448506674e-09, | |
| "loss": 5.625, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 3.165811767433061, | |
| "learning_rate": 3.3564637317984318e-09, | |
| "loss": 5.4195, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 2.9299117263430903, | |
| "learning_rate": 1.888057060274173e-09, | |
| "loss": 5.6383, | |
| "step": 500 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 500, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 25, | |
| "total_flos": 2968550903808.0, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |