| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.18066847335140018, | |
| "eval_steps": 500, | |
| "global_step": 100, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0018066847335140017, | |
| "grad_norm": 2.834148994241387, | |
| "learning_rate": 0.0, | |
| "loss": 0.8251, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0036133694670280035, | |
| "grad_norm": 2.8814188855606915, | |
| "learning_rate": 3.5714285714285716e-07, | |
| "loss": 0.8284, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.005420054200542005, | |
| "grad_norm": 2.8877082225775945, | |
| "learning_rate": 7.142857142857143e-07, | |
| "loss": 0.8432, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.007226738934056007, | |
| "grad_norm": 2.84706824634251, | |
| "learning_rate": 1.0714285714285714e-06, | |
| "loss": 0.8395, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.009033423667570008, | |
| "grad_norm": 2.763620649396407, | |
| "learning_rate": 1.4285714285714286e-06, | |
| "loss": 0.8291, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.01084010840108401, | |
| "grad_norm": 2.7250326443499935, | |
| "learning_rate": 1.7857142857142859e-06, | |
| "loss": 0.8289, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.012646793134598013, | |
| "grad_norm": 2.5921911890865355, | |
| "learning_rate": 2.1428571428571427e-06, | |
| "loss": 0.8081, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.014453477868112014, | |
| "grad_norm": 2.5976860246328948, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.8152, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.016260162601626018, | |
| "grad_norm": 2.1600249112189367, | |
| "learning_rate": 2.8571428571428573e-06, | |
| "loss": 0.8041, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.018066847335140017, | |
| "grad_norm": 2.05676893822907, | |
| "learning_rate": 3.2142857142857147e-06, | |
| "loss": 0.7978, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.01987353206865402, | |
| "grad_norm": 1.5227088422406245, | |
| "learning_rate": 3.5714285714285718e-06, | |
| "loss": 0.7766, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.02168021680216802, | |
| "grad_norm": 1.4305134864471831, | |
| "learning_rate": 3.928571428571429e-06, | |
| "loss": 0.7865, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.023486901535682024, | |
| "grad_norm": 1.3403916222492462, | |
| "learning_rate": 4.2857142857142855e-06, | |
| "loss": 0.7734, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.025293586269196026, | |
| "grad_norm": 1.2438356291232484, | |
| "learning_rate": 4.642857142857144e-06, | |
| "loss": 0.7668, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.02710027100271003, | |
| "grad_norm": 1.4470043220956499, | |
| "learning_rate": 5e-06, | |
| "loss": 0.7518, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.028906955736224028, | |
| "grad_norm": 1.6508105965347686, | |
| "learning_rate": 5.357142857142857e-06, | |
| "loss": 0.7428, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.03071364046973803, | |
| "grad_norm": 1.8197388085593902, | |
| "learning_rate": 5.7142857142857145e-06, | |
| "loss": 0.7504, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.032520325203252036, | |
| "grad_norm": 1.6628024493622566, | |
| "learning_rate": 6.071428571428571e-06, | |
| "loss": 0.7406, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.03432700993676603, | |
| "grad_norm": 1.3996342649224327, | |
| "learning_rate": 6.4285714285714295e-06, | |
| "loss": 0.7317, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.036133694670280034, | |
| "grad_norm": 0.8529424931984829, | |
| "learning_rate": 6.785714285714287e-06, | |
| "loss": 0.7029, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.037940379403794036, | |
| "grad_norm": 1.012307815845279, | |
| "learning_rate": 7.1428571428571436e-06, | |
| "loss": 0.7052, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.03974706413730804, | |
| "grad_norm": 0.996083810328283, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.6937, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.04155374887082204, | |
| "grad_norm": 0.9027655260544101, | |
| "learning_rate": 7.857142857142858e-06, | |
| "loss": 0.691, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.04336043360433604, | |
| "grad_norm": 1.9068544177754214, | |
| "learning_rate": 8.214285714285714e-06, | |
| "loss": 0.697, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.045167118337850046, | |
| "grad_norm": 0.7710305137202722, | |
| "learning_rate": 8.571428571428571e-06, | |
| "loss": 0.677, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.04697380307136405, | |
| "grad_norm": 0.6673976579110235, | |
| "learning_rate": 8.92857142857143e-06, | |
| "loss": 0.6911, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.04878048780487805, | |
| "grad_norm": 0.6017238871929963, | |
| "learning_rate": 9.285714285714288e-06, | |
| "loss": 0.667, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.05058717253839205, | |
| "grad_norm": 0.6176897562760387, | |
| "learning_rate": 9.642857142857144e-06, | |
| "loss": 0.6685, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.052393857271906055, | |
| "grad_norm": 0.5460140411641373, | |
| "learning_rate": 1e-05, | |
| "loss": 0.6788, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.05420054200542006, | |
| "grad_norm": 0.5024441002282133, | |
| "learning_rate": 9.999910480045805e-06, | |
| "loss": 0.6776, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.05600722673893405, | |
| "grad_norm": 0.5327568235598038, | |
| "learning_rate": 9.999641923388745e-06, | |
| "loss": 0.666, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.057813911472448055, | |
| "grad_norm": 0.45860832293770903, | |
| "learning_rate": 9.999194339645292e-06, | |
| "loss": 0.6537, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.05962059620596206, | |
| "grad_norm": 0.462200098378075, | |
| "learning_rate": 9.998567744842518e-06, | |
| "loss": 0.6639, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.06142728093947606, | |
| "grad_norm": 0.44108437546829815, | |
| "learning_rate": 9.997762161417517e-06, | |
| "loss": 0.6507, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.06323396567299007, | |
| "grad_norm": 0.4056783549136984, | |
| "learning_rate": 9.996777618216608e-06, | |
| "loss": 0.6559, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.06504065040650407, | |
| "grad_norm": 0.3860777271649062, | |
| "learning_rate": 9.995614150494293e-06, | |
| "loss": 0.6503, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.06684733514001806, | |
| "grad_norm": 0.434533435588928, | |
| "learning_rate": 9.994271799912004e-06, | |
| "loss": 0.6541, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.06865401987353206, | |
| "grad_norm": 0.4371779969673632, | |
| "learning_rate": 9.992750614536606e-06, | |
| "loss": 0.6469, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.07046070460704607, | |
| "grad_norm": 0.36457045763018364, | |
| "learning_rate": 9.991050648838676e-06, | |
| "loss": 0.6475, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.07226738934056007, | |
| "grad_norm": 0.322635334253873, | |
| "learning_rate": 9.989171963690556e-06, | |
| "loss": 0.6366, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.07407407407407407, | |
| "grad_norm": 0.35059303076021425, | |
| "learning_rate": 9.987114626364172e-06, | |
| "loss": 0.6431, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.07588075880758807, | |
| "grad_norm": 0.36612621878509766, | |
| "learning_rate": 9.984878710528615e-06, | |
| "loss": 0.6339, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.07768744354110207, | |
| "grad_norm": 0.3365894359412745, | |
| "learning_rate": 9.982464296247523e-06, | |
| "loss": 0.6343, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.07949412827461608, | |
| "grad_norm": 0.3035590675895417, | |
| "learning_rate": 9.979871469976197e-06, | |
| "loss": 0.6275, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.08130081300813008, | |
| "grad_norm": 0.2858077116353655, | |
| "learning_rate": 9.97710032455851e-06, | |
| "loss": 0.6344, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.08310749774164408, | |
| "grad_norm": 0.2752507660143639, | |
| "learning_rate": 9.974150959223591e-06, | |
| "loss": 0.6362, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.08491418247515808, | |
| "grad_norm": 0.3109343353140871, | |
| "learning_rate": 9.971023479582258e-06, | |
| "loss": 0.6389, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.08672086720867209, | |
| "grad_norm": 0.27841301256018586, | |
| "learning_rate": 9.967717997623245e-06, | |
| "loss": 0.6256, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.08852755194218609, | |
| "grad_norm": 0.3002752587221535, | |
| "learning_rate": 9.964234631709188e-06, | |
| "loss": 0.6316, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.09033423667570009, | |
| "grad_norm": 0.24995397436718791, | |
| "learning_rate": 9.960573506572391e-06, | |
| "loss": 0.6303, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.0921409214092141, | |
| "grad_norm": 0.2592033541282412, | |
| "learning_rate": 9.956734753310355e-06, | |
| "loss": 0.6199, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.0939476061427281, | |
| "grad_norm": 0.26267469782137715, | |
| "learning_rate": 9.952718509381086e-06, | |
| "loss": 0.6381, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.0957542908762421, | |
| "grad_norm": 0.2841156370165673, | |
| "learning_rate": 9.948524918598175e-06, | |
| "loss": 0.6223, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.0975609756097561, | |
| "grad_norm": 0.2602382902911376, | |
| "learning_rate": 9.944154131125643e-06, | |
| "loss": 0.613, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.0993676603432701, | |
| "grad_norm": 0.2548980055658011, | |
| "learning_rate": 9.93960630347257e-06, | |
| "loss": 0.6265, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.1011743450767841, | |
| "grad_norm": 0.23755160689461077, | |
| "learning_rate": 9.934881598487478e-06, | |
| "loss": 0.6318, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.10298102981029811, | |
| "grad_norm": 0.24890312911017654, | |
| "learning_rate": 9.929980185352525e-06, | |
| "loss": 0.6175, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.10478771454381211, | |
| "grad_norm": 0.2773541758032373, | |
| "learning_rate": 9.924902239577419e-06, | |
| "loss": 0.6253, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.10659439927732611, | |
| "grad_norm": 0.25242119733836627, | |
| "learning_rate": 9.91964794299315e-06, | |
| "loss": 0.611, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.10840108401084012, | |
| "grad_norm": 0.2684984552226452, | |
| "learning_rate": 9.914217483745472e-06, | |
| "loss": 0.6119, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.1102077687443541, | |
| "grad_norm": 0.2574590656590093, | |
| "learning_rate": 9.90861105628817e-06, | |
| "loss": 0.6159, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.1120144534778681, | |
| "grad_norm": 0.25603682816247697, | |
| "learning_rate": 9.902828861376101e-06, | |
| "loss": 0.621, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.11382113821138211, | |
| "grad_norm": 0.24731176189318566, | |
| "learning_rate": 9.896871106057989e-06, | |
| "loss": 0.6205, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.11562782294489611, | |
| "grad_norm": 0.27598163635496337, | |
| "learning_rate": 9.890738003669029e-06, | |
| "loss": 0.6188, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.11743450767841011, | |
| "grad_norm": 0.26447533420584596, | |
| "learning_rate": 9.884429773823238e-06, | |
| "loss": 0.6134, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.11924119241192412, | |
| "grad_norm": 0.2764756892440485, | |
| "learning_rate": 9.877946642405598e-06, | |
| "loss": 0.6153, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.12104787714543812, | |
| "grad_norm": 0.2878202327908946, | |
| "learning_rate": 9.871288841563956e-06, | |
| "loss": 0.6057, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.12285456187895212, | |
| "grad_norm": 0.25954176321509254, | |
| "learning_rate": 9.864456609700726e-06, | |
| "loss": 0.6213, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.12466124661246612, | |
| "grad_norm": 0.25912508931378314, | |
| "learning_rate": 9.857450191464337e-06, | |
| "loss": 0.6233, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.12646793134598014, | |
| "grad_norm": 0.2698506612904357, | |
| "learning_rate": 9.85026983774049e-06, | |
| "loss": 0.6285, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.12827461607949414, | |
| "grad_norm": 0.23787370000958719, | |
| "learning_rate": 9.842915805643156e-06, | |
| "loss": 0.5996, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.13008130081300814, | |
| "grad_norm": 0.21973413712743903, | |
| "learning_rate": 9.835388358505383e-06, | |
| "loss": 0.6171, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.13188798554652212, | |
| "grad_norm": 0.2542097049285353, | |
| "learning_rate": 9.827687765869859e-06, | |
| "loss": 0.6159, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.13369467028003612, | |
| "grad_norm": 0.28885915694073955, | |
| "learning_rate": 9.819814303479268e-06, | |
| "loss": 0.6081, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.13550135501355012, | |
| "grad_norm": 0.2732542861288398, | |
| "learning_rate": 9.811768253266401e-06, | |
| "loss": 0.606, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.13730803974706413, | |
| "grad_norm": 0.2639987596902754, | |
| "learning_rate": 9.803549903344081e-06, | |
| "loss": 0.6016, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.13911472448057813, | |
| "grad_norm": 0.2625743746400295, | |
| "learning_rate": 9.79515954799483e-06, | |
| "loss": 0.5963, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.14092140921409213, | |
| "grad_norm": 0.2636110963817243, | |
| "learning_rate": 9.786597487660336e-06, | |
| "loss": 0.6082, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.14272809394760613, | |
| "grad_norm": 0.2548650343445913, | |
| "learning_rate": 9.777864028930705e-06, | |
| "loss": 0.6173, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.14453477868112014, | |
| "grad_norm": 0.24198639697590743, | |
| "learning_rate": 9.768959484533461e-06, | |
| "loss": 0.6262, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.14634146341463414, | |
| "grad_norm": 0.23616390413838728, | |
| "learning_rate": 9.75988417332237e-06, | |
| "loss": 0.6087, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.14814814814814814, | |
| "grad_norm": 0.24740792444302148, | |
| "learning_rate": 9.750638420266008e-06, | |
| "loss": 0.6023, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.14995483288166214, | |
| "grad_norm": 0.2708089867070639, | |
| "learning_rate": 9.741222556436132e-06, | |
| "loss": 0.6133, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.15176151761517614, | |
| "grad_norm": 0.23019955730854466, | |
| "learning_rate": 9.731636918995821e-06, | |
| "loss": 0.606, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.15356820234869015, | |
| "grad_norm": 0.24984754231513484, | |
| "learning_rate": 9.721881851187406e-06, | |
| "loss": 0.6082, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.15537488708220415, | |
| "grad_norm": 0.2514290283088365, | |
| "learning_rate": 9.711957702320176e-06, | |
| "loss": 0.6082, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.15718157181571815, | |
| "grad_norm": 0.25050319285982814, | |
| "learning_rate": 9.701864827757868e-06, | |
| "loss": 0.6103, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.15898825654923215, | |
| "grad_norm": 0.27127901136367444, | |
| "learning_rate": 9.691603588905956e-06, | |
| "loss": 0.6146, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.16079494128274616, | |
| "grad_norm": 0.2538064222847297, | |
| "learning_rate": 9.681174353198687e-06, | |
| "loss": 0.6103, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.16260162601626016, | |
| "grad_norm": 0.23895346723279925, | |
| "learning_rate": 9.670577494085945e-06, | |
| "loss": 0.6034, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.16440831074977416, | |
| "grad_norm": 0.23015986092903712, | |
| "learning_rate": 9.659813391019867e-06, | |
| "loss": 0.6013, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.16621499548328816, | |
| "grad_norm": 0.23826711464547476, | |
| "learning_rate": 9.648882429441258e-06, | |
| "loss": 0.6048, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.16802168021680217, | |
| "grad_norm": 0.27338890022372714, | |
| "learning_rate": 9.637785000765789e-06, | |
| "loss": 0.6114, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.16982836495031617, | |
| "grad_norm": 0.3426444715164819, | |
| "learning_rate": 9.626521502369984e-06, | |
| "loss": 0.6104, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.17163504968383017, | |
| "grad_norm": 0.23367032192171455, | |
| "learning_rate": 9.615092337576987e-06, | |
| "loss": 0.6027, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.17344173441734417, | |
| "grad_norm": 0.25432897179920155, | |
| "learning_rate": 9.603497915642122e-06, | |
| "loss": 0.6016, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.17524841915085818, | |
| "grad_norm": 0.26462048665974897, | |
| "learning_rate": 9.591738651738235e-06, | |
| "loss": 0.6073, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.17705510388437218, | |
| "grad_norm": 0.24672431475599113, | |
| "learning_rate": 9.579814966940833e-06, | |
| "loss": 0.6013, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.17886178861788618, | |
| "grad_norm": 0.23213124494510456, | |
| "learning_rate": 9.567727288213005e-06, | |
| "loss": 0.6138, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.18066847335140018, | |
| "grad_norm": 0.22763203934971926, | |
| "learning_rate": 9.55547604839013e-06, | |
| "loss": 0.5873, | |
| "step": 100 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 553, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 115440686923776.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |