| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.36133694670280037, | |
| "eval_steps": 500, | |
| "global_step": 200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0018066847335140017, | |
| "grad_norm": 2.6597820350894357, | |
| "learning_rate": 0.0, | |
| "loss": 0.8251, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0036133694670280035, | |
| "grad_norm": 2.7266510344811494, | |
| "learning_rate": 1.0714285714285714e-06, | |
| "loss": 0.8284, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.005420054200542005, | |
| "grad_norm": 2.7200407943823968, | |
| "learning_rate": 2.1428571428571427e-06, | |
| "loss": 0.8429, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.007226738934056007, | |
| "grad_norm": 2.6398267415203076, | |
| "learning_rate": 3.2142857142857143e-06, | |
| "loss": 0.8381, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.009033423667570008, | |
| "grad_norm": 2.450058379558589, | |
| "learning_rate": 4.2857142857142855e-06, | |
| "loss": 0.8245, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.01084010840108401, | |
| "grad_norm": 1.9536840033460428, | |
| "learning_rate": 5.357142857142857e-06, | |
| "loss": 0.814, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.012646793134598013, | |
| "grad_norm": 1.2180340367347295, | |
| "learning_rate": 6.428571428571429e-06, | |
| "loss": 0.7845, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.014453477868112014, | |
| "grad_norm": 1.1514973199014618, | |
| "learning_rate": 7.5e-06, | |
| "loss": 0.7874, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.016260162601626018, | |
| "grad_norm": 1.1427251731602237, | |
| "learning_rate": 8.571428571428571e-06, | |
| "loss": 0.7793, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.018066847335140017, | |
| "grad_norm": 2.4537001853666673, | |
| "learning_rate": 9.642857142857144e-06, | |
| "loss": 0.7624, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.01987353206865402, | |
| "grad_norm": 2.474412722550665, | |
| "learning_rate": 1.0714285714285714e-05, | |
| "loss": 0.7491, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.02168021680216802, | |
| "grad_norm": 2.333456673377541, | |
| "learning_rate": 1.1785714285714286e-05, | |
| "loss": 0.762, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.023486901535682024, | |
| "grad_norm": 1.6274020508818596, | |
| "learning_rate": 1.2857142857142857e-05, | |
| "loss": 0.7368, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.025293586269196026, | |
| "grad_norm": 1.563869927563279, | |
| "learning_rate": 1.3928571428571429e-05, | |
| "loss": 0.7378, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.02710027100271003, | |
| "grad_norm": 1.253626680340628, | |
| "learning_rate": 1.5e-05, | |
| "loss": 0.7271, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.028906955736224028, | |
| "grad_norm": 0.9176330164857595, | |
| "learning_rate": 1.6071428571428572e-05, | |
| "loss": 0.7131, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.03071364046973803, | |
| "grad_norm": 0.7185343991261351, | |
| "learning_rate": 1.7142857142857142e-05, | |
| "loss": 0.7131, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.032520325203252036, | |
| "grad_norm": 0.7884609645105337, | |
| "learning_rate": 1.8214285714285712e-05, | |
| "loss": 0.7045, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.03432700993676603, | |
| "grad_norm": 0.7399830971167167, | |
| "learning_rate": 1.928571428571429e-05, | |
| "loss": 0.6995, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.036133694670280034, | |
| "grad_norm": 0.7376328500886337, | |
| "learning_rate": 2.0357142857142858e-05, | |
| "loss": 0.6779, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.037940379403794036, | |
| "grad_norm": 0.6439873145292982, | |
| "learning_rate": 2.1428571428571428e-05, | |
| "loss": 0.6834, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.03974706413730804, | |
| "grad_norm": 0.5790542088592184, | |
| "learning_rate": 2.25e-05, | |
| "loss": 0.6721, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.04155374887082204, | |
| "grad_norm": 0.5191304250088061, | |
| "learning_rate": 2.357142857142857e-05, | |
| "loss": 0.6707, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.04336043360433604, | |
| "grad_norm": 0.4488946670945292, | |
| "learning_rate": 2.464285714285714e-05, | |
| "loss": 0.676, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.045167118337850046, | |
| "grad_norm": 0.4798647764154148, | |
| "learning_rate": 2.5714285714285714e-05, | |
| "loss": 0.6585, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.04697380307136405, | |
| "grad_norm": 0.4204330566493729, | |
| "learning_rate": 2.6785714285714288e-05, | |
| "loss": 0.673, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.04878048780487805, | |
| "grad_norm": 0.47130112684574876, | |
| "learning_rate": 2.7857142857142858e-05, | |
| "loss": 0.6516, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.05058717253839205, | |
| "grad_norm": 0.4362815754533529, | |
| "learning_rate": 2.892857142857143e-05, | |
| "loss": 0.6545, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.052393857271906055, | |
| "grad_norm": 0.43822648560890193, | |
| "learning_rate": 3e-05, | |
| "loss": 0.6647, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.05420054200542006, | |
| "grad_norm": 0.3853311574389301, | |
| "learning_rate": 2.9999731440137413e-05, | |
| "loss": 0.6644, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.05600722673893405, | |
| "grad_norm": 0.4221785516753897, | |
| "learning_rate": 2.9998925770166232e-05, | |
| "loss": 0.6534, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.057813911472448055, | |
| "grad_norm": 0.3980570059327311, | |
| "learning_rate": 2.9997583018935875e-05, | |
| "loss": 0.6416, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.05962059620596206, | |
| "grad_norm": 0.35571676272289743, | |
| "learning_rate": 2.9995703234527553e-05, | |
| "loss": 0.6513, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.06142728093947606, | |
| "grad_norm": 0.38024957472516396, | |
| "learning_rate": 2.999328648425255e-05, | |
| "loss": 0.6385, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.06323396567299007, | |
| "grad_norm": 0.2924858728569485, | |
| "learning_rate": 2.999033285464982e-05, | |
| "loss": 0.644, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.06504065040650407, | |
| "grad_norm": 0.3482798299572746, | |
| "learning_rate": 2.9986842451482876e-05, | |
| "loss": 0.6387, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.06684733514001806, | |
| "grad_norm": 0.3705337140725527, | |
| "learning_rate": 2.9982815399736008e-05, | |
| "loss": 0.6432, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.06865401987353206, | |
| "grad_norm": 0.3364721163164359, | |
| "learning_rate": 2.9978251843609816e-05, | |
| "loss": 0.6351, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.07046070460704607, | |
| "grad_norm": 0.3272631660096153, | |
| "learning_rate": 2.9973151946516027e-05, | |
| "loss": 0.6364, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.07226738934056007, | |
| "grad_norm": 0.29708127282462954, | |
| "learning_rate": 2.996751589107167e-05, | |
| "loss": 0.6252, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.07407407407407407, | |
| "grad_norm": 0.24921077158970448, | |
| "learning_rate": 2.9961343879092512e-05, | |
| "loss": 0.6308, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.07588075880758807, | |
| "grad_norm": 0.3479569257927118, | |
| "learning_rate": 2.9954636131585845e-05, | |
| "loss": 0.6234, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.07768744354110207, | |
| "grad_norm": 0.28917511643628596, | |
| "learning_rate": 2.9947392888742566e-05, | |
| "loss": 0.6233, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.07949412827461608, | |
| "grad_norm": 0.34737822266968965, | |
| "learning_rate": 2.993961440992859e-05, | |
| "loss": 0.6171, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.08130081300813008, | |
| "grad_norm": 0.3247689228102166, | |
| "learning_rate": 2.993130097367553e-05, | |
| "loss": 0.6238, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.08310749774164408, | |
| "grad_norm": 0.30719902123424325, | |
| "learning_rate": 2.9922452877670775e-05, | |
| "loss": 0.6252, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.08491418247515808, | |
| "grad_norm": 0.3175141153740525, | |
| "learning_rate": 2.991307043874677e-05, | |
| "loss": 0.6281, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.08672086720867209, | |
| "grad_norm": 0.31244488771705314, | |
| "learning_rate": 2.9903153992869734e-05, | |
| "loss": 0.6148, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.08852755194218609, | |
| "grad_norm": 0.3429131319420423, | |
| "learning_rate": 2.989270389512756e-05, | |
| "loss": 0.6211, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.09033423667570009, | |
| "grad_norm": 0.3104062918234721, | |
| "learning_rate": 2.988172051971717e-05, | |
| "loss": 0.6204, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.0921409214092141, | |
| "grad_norm": 0.31461103786584654, | |
| "learning_rate": 2.9870204259931062e-05, | |
| "loss": 0.6097, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.0939476061427281, | |
| "grad_norm": 0.3151625720082357, | |
| "learning_rate": 2.9858155528143256e-05, | |
| "loss": 0.6268, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.0957542908762421, | |
| "grad_norm": 0.2860126481493585, | |
| "learning_rate": 2.9845574755794522e-05, | |
| "loss": 0.6125, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.0975609756097561, | |
| "grad_norm": 0.28841121592431124, | |
| "learning_rate": 2.9832462393376926e-05, | |
| "loss": 0.6018, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.0993676603432701, | |
| "grad_norm": 0.23534181927847642, | |
| "learning_rate": 2.9818818910417706e-05, | |
| "loss": 0.6169, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.1011743450767841, | |
| "grad_norm": 0.28661471046317555, | |
| "learning_rate": 2.9804644795462437e-05, | |
| "loss": 0.6212, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.10298102981029811, | |
| "grad_norm": 0.2443857675436309, | |
| "learning_rate": 2.9789940556057574e-05, | |
| "loss": 0.6065, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.10478771454381211, | |
| "grad_norm": 0.292418070185767, | |
| "learning_rate": 2.9774706718732255e-05, | |
| "loss": 0.6152, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.10659439927732611, | |
| "grad_norm": 0.24742182455944323, | |
| "learning_rate": 2.9758943828979444e-05, | |
| "loss": 0.6009, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.10840108401084012, | |
| "grad_norm": 0.25007211477600044, | |
| "learning_rate": 2.9742652451236414e-05, | |
| "loss": 0.6023, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.1102077687443541, | |
| "grad_norm": 0.24539117288225054, | |
| "learning_rate": 2.972583316886451e-05, | |
| "loss": 0.6059, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.1120144534778681, | |
| "grad_norm": 0.2544907417129908, | |
| "learning_rate": 2.9708486584128303e-05, | |
| "loss": 0.6107, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.11382113821138211, | |
| "grad_norm": 0.25269030392560393, | |
| "learning_rate": 2.9690613318173966e-05, | |
| "loss": 0.611, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.11562782294489611, | |
| "grad_norm": 0.20651691272156247, | |
| "learning_rate": 2.9672214011007087e-05, | |
| "loss": 0.6091, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.11743450767841011, | |
| "grad_norm": 0.25759031036445135, | |
| "learning_rate": 2.9653289321469715e-05, | |
| "loss": 0.6031, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.11924119241192412, | |
| "grad_norm": 0.27004948462624506, | |
| "learning_rate": 2.9633839927216793e-05, | |
| "loss": 0.6061, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.12104787714543812, | |
| "grad_norm": 0.23899477664304233, | |
| "learning_rate": 2.9613866524691867e-05, | |
| "loss": 0.5958, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.12285456187895212, | |
| "grad_norm": 0.3109356502387666, | |
| "learning_rate": 2.9593369829102173e-05, | |
| "loss": 0.6119, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.12466124661246612, | |
| "grad_norm": 0.23973937288759506, | |
| "learning_rate": 2.957235057439301e-05, | |
| "loss": 0.6135, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.12646793134598014, | |
| "grad_norm": 0.29402451520285133, | |
| "learning_rate": 2.955080951322147e-05, | |
| "loss": 0.6178, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.12827461607949414, | |
| "grad_norm": 0.29085498534508775, | |
| "learning_rate": 2.9528747416929467e-05, | |
| "loss": 0.5907, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.13008130081300814, | |
| "grad_norm": 0.25477514255751904, | |
| "learning_rate": 2.9506165075516148e-05, | |
| "loss": 0.6073, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.13188798554652212, | |
| "grad_norm": 0.28366229733780235, | |
| "learning_rate": 2.9483063297609577e-05, | |
| "loss": 0.6063, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.13369467028003612, | |
| "grad_norm": 0.26317649422081363, | |
| "learning_rate": 2.9459442910437798e-05, | |
| "loss": 0.598, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.13550135501355012, | |
| "grad_norm": 0.28039599919104496, | |
| "learning_rate": 2.94353047597992e-05, | |
| "loss": 0.5971, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.13730803974706413, | |
| "grad_norm": 0.271745198157929, | |
| "learning_rate": 2.941064971003224e-05, | |
| "loss": 0.593, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.13911472448057813, | |
| "grad_norm": 0.24998502389958926, | |
| "learning_rate": 2.9385478643984484e-05, | |
| "loss": 0.5878, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.14092140921409213, | |
| "grad_norm": 0.30322559816019845, | |
| "learning_rate": 2.9359792462981007e-05, | |
| "loss": 0.599, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.14272809394760613, | |
| "grad_norm": 0.29351194680487536, | |
| "learning_rate": 2.9333592086792113e-05, | |
| "loss": 0.6075, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.14453477868112014, | |
| "grad_norm": 0.2670107767188747, | |
| "learning_rate": 2.9306878453600382e-05, | |
| "loss": 0.6161, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.14634146341463414, | |
| "grad_norm": 0.2927366904918854, | |
| "learning_rate": 2.9279652519967105e-05, | |
| "loss": 0.6004, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.14814814814814814, | |
| "grad_norm": 0.28777141201555767, | |
| "learning_rate": 2.9251915260798024e-05, | |
| "loss": 0.593, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.14995483288166214, | |
| "grad_norm": 0.282262421377073, | |
| "learning_rate": 2.9223667669308395e-05, | |
| "loss": 0.6038, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.15176151761517614, | |
| "grad_norm": 0.24811154699761442, | |
| "learning_rate": 2.9194910756987464e-05, | |
| "loss": 0.5976, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.15356820234869015, | |
| "grad_norm": 0.27670190484589513, | |
| "learning_rate": 2.9165645553562215e-05, | |
| "loss": 0.5997, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.15537488708220415, | |
| "grad_norm": 0.24338487756299684, | |
| "learning_rate": 2.9135873106960525e-05, | |
| "loss": 0.5994, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.15718157181571815, | |
| "grad_norm": 0.3040191021283788, | |
| "learning_rate": 2.9105594483273603e-05, | |
| "loss": 0.6012, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.15898825654923215, | |
| "grad_norm": 0.26703924304748566, | |
| "learning_rate": 2.9074810766717865e-05, | |
| "loss": 0.6058, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.16079494128274616, | |
| "grad_norm": 0.2694312104706441, | |
| "learning_rate": 2.904352305959606e-05, | |
| "loss": 0.602, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.16260162601626016, | |
| "grad_norm": 0.29517928394037635, | |
| "learning_rate": 2.9011732482257835e-05, | |
| "loss": 0.5944, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.16440831074977416, | |
| "grad_norm": 0.25118801401605734, | |
| "learning_rate": 2.89794401730596e-05, | |
| "loss": 0.5923, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.16621499548328816, | |
| "grad_norm": 0.3348680034328015, | |
| "learning_rate": 2.894664728832377e-05, | |
| "loss": 0.5958, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.16802168021680217, | |
| "grad_norm": 0.25945936645525364, | |
| "learning_rate": 2.8913355002297367e-05, | |
| "loss": 0.6025, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.16982836495031617, | |
| "grad_norm": 0.31672412557275315, | |
| "learning_rate": 2.887956450710995e-05, | |
| "loss": 0.602, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.17163504968383017, | |
| "grad_norm": 0.2631935513788724, | |
| "learning_rate": 2.8845277012730963e-05, | |
| "loss": 0.5939, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.17344173441734417, | |
| "grad_norm": 0.28309967073656406, | |
| "learning_rate": 2.8810493746926364e-05, | |
| "loss": 0.5932, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.17524841915085818, | |
| "grad_norm": 0.28939178950900774, | |
| "learning_rate": 2.87752159552147e-05, | |
| "loss": 0.5993, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.17705510388437218, | |
| "grad_norm": 0.3021445789644071, | |
| "learning_rate": 2.87394449008225e-05, | |
| "loss": 0.5924, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.17886178861788618, | |
| "grad_norm": 0.2648161341085879, | |
| "learning_rate": 2.8703181864639013e-05, | |
| "loss": 0.6058, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.18066847335140018, | |
| "grad_norm": 0.28484263640616314, | |
| "learning_rate": 2.8666428145170385e-05, | |
| "loss": 0.5789, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.18247515808491419, | |
| "grad_norm": 0.2611846886724786, | |
| "learning_rate": 2.8629185058493116e-05, | |
| "loss": 0.5949, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.1842818428184282, | |
| "grad_norm": 0.33149916071305763, | |
| "learning_rate": 2.8591453938206985e-05, | |
| "loss": 0.6134, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.1860885275519422, | |
| "grad_norm": 0.28402400837841096, | |
| "learning_rate": 2.8553236135387247e-05, | |
| "loss": 0.5914, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.1878952122854562, | |
| "grad_norm": 0.28825320782588004, | |
| "learning_rate": 2.8514533018536286e-05, | |
| "loss": 0.5897, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.1897018970189702, | |
| "grad_norm": 0.2901395812522203, | |
| "learning_rate": 2.8475345973534605e-05, | |
| "loss": 0.5822, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.1915085817524842, | |
| "grad_norm": 0.25034051990006584, | |
| "learning_rate": 2.8435676403591193e-05, | |
| "loss": 0.5846, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.1933152664859982, | |
| "grad_norm": 0.27868357694305945, | |
| "learning_rate": 2.8395525729193284e-05, | |
| "loss": 0.5907, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.1951219512195122, | |
| "grad_norm": 0.257742952703987, | |
| "learning_rate": 2.835489538805548e-05, | |
| "loss": 0.5888, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.1969286359530262, | |
| "grad_norm": 0.26397326166674545, | |
| "learning_rate": 2.8313786835068314e-05, | |
| "loss": 0.5976, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.1987353206865402, | |
| "grad_norm": 0.2721484510112116, | |
| "learning_rate": 2.8272201542246077e-05, | |
| "loss": 0.5792, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.2005420054200542, | |
| "grad_norm": 0.2374425057236497, | |
| "learning_rate": 2.8230140998674185e-05, | |
| "loss": 0.599, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.2023486901535682, | |
| "grad_norm": 0.24242218904343407, | |
| "learning_rate": 2.8187606710455807e-05, | |
| "loss": 0.5921, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.2041553748870822, | |
| "grad_norm": 0.2742052397132371, | |
| "learning_rate": 2.8144600200657953e-05, | |
| "loss": 0.5965, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.20596205962059622, | |
| "grad_norm": 0.2240931438132582, | |
| "learning_rate": 2.8101123009256946e-05, | |
| "loss": 0.5792, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.20776874435411022, | |
| "grad_norm": 0.2763456467888191, | |
| "learning_rate": 2.8057176693083253e-05, | |
| "loss": 0.5958, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.20957542908762422, | |
| "grad_norm": 0.2435434093260049, | |
| "learning_rate": 2.8012762825765763e-05, | |
| "loss": 0.5891, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.21138211382113822, | |
| "grad_norm": 0.2769370185736096, | |
| "learning_rate": 2.7967882997675424e-05, | |
| "loss": 0.6012, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.21318879855465223, | |
| "grad_norm": 0.2556991024086139, | |
| "learning_rate": 2.7922538815868287e-05, | |
| "loss": 0.5879, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.21499548328816623, | |
| "grad_norm": 0.2913165244476914, | |
| "learning_rate": 2.7876731904027994e-05, | |
| "loss": 0.5784, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.21680216802168023, | |
| "grad_norm": 0.29037318268167045, | |
| "learning_rate": 2.78304639024076e-05, | |
| "loss": 0.5968, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.2186088527551942, | |
| "grad_norm": 0.24340049924612678, | |
| "learning_rate": 2.7783736467770863e-05, | |
| "loss": 0.5927, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.2204155374887082, | |
| "grad_norm": 0.29825692316659325, | |
| "learning_rate": 2.7736551273332908e-05, | |
| "loss": 0.5972, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.2222222222222222, | |
| "grad_norm": 0.2593832259683122, | |
| "learning_rate": 2.7688910008700305e-05, | |
| "loss": 0.5876, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.2240289069557362, | |
| "grad_norm": 0.27547771011759425, | |
| "learning_rate": 2.764081437981059e-05, | |
| "loss": 0.5742, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.22583559168925021, | |
| "grad_norm": 0.27158658907185634, | |
| "learning_rate": 2.7592266108871158e-05, | |
| "loss": 0.5812, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.22764227642276422, | |
| "grad_norm": 0.2715497634460973, | |
| "learning_rate": 2.754326693429761e-05, | |
| "loss": 0.5931, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.22944896115627822, | |
| "grad_norm": 0.24142985625388147, | |
| "learning_rate": 2.7493818610651493e-05, | |
| "loss": 0.5859, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.23125564588979222, | |
| "grad_norm": 0.2530474955317245, | |
| "learning_rate": 2.744392290857747e-05, | |
| "loss": 0.5857, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.23306233062330622, | |
| "grad_norm": 0.2567731311347638, | |
| "learning_rate": 2.7393581614739924e-05, | |
| "loss": 0.5803, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.23486901535682023, | |
| "grad_norm": 0.2628073110459229, | |
| "learning_rate": 2.7342796531758984e-05, | |
| "loss": 0.5987, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.23667570009033423, | |
| "grad_norm": 0.23510376935644756, | |
| "learning_rate": 2.729156947814598e-05, | |
| "loss": 0.5926, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.23848238482384823, | |
| "grad_norm": 0.23576954058784624, | |
| "learning_rate": 2.7239902288238297e-05, | |
| "loss": 0.5721, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.24028906955736223, | |
| "grad_norm": 0.24964231366212492, | |
| "learning_rate": 2.7187796812133733e-05, | |
| "loss": 0.5873, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.24209575429087624, | |
| "grad_norm": 0.23911643013915115, | |
| "learning_rate": 2.7135254915624213e-05, | |
| "loss": 0.5878, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.24390243902439024, | |
| "grad_norm": 0.23601189225311667, | |
| "learning_rate": 2.708227848012901e-05, | |
| "loss": 0.585, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.24570912375790424, | |
| "grad_norm": 0.28061149806949875, | |
| "learning_rate": 2.7028869402627357e-05, | |
| "loss": 0.5889, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.24751580849141824, | |
| "grad_norm": 0.2537605160128264, | |
| "learning_rate": 2.6975029595590523e-05, | |
| "loss": 0.5726, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.24932249322493225, | |
| "grad_norm": 0.2631279788827005, | |
| "learning_rate": 2.6920760986913332e-05, | |
| "loss": 0.5901, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.25112917795844625, | |
| "grad_norm": 0.24684546421512338, | |
| "learning_rate": 2.6866065519845124e-05, | |
| "loss": 0.5932, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.2529358626919603, | |
| "grad_norm": 0.25108587954115535, | |
| "learning_rate": 2.681094515292018e-05, | |
| "loss": 0.5927, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.25474254742547425, | |
| "grad_norm": 0.2759732073497199, | |
| "learning_rate": 2.6755401859887598e-05, | |
| "loss": 0.573, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.2565492321589883, | |
| "grad_norm": 0.2927986426856055, | |
| "learning_rate": 2.6699437629640595e-05, | |
| "loss": 0.5934, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.25835591689250226, | |
| "grad_norm": 0.2915811926744858, | |
| "learning_rate": 2.6643054466145297e-05, | |
| "loss": 0.5846, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.2601626016260163, | |
| "grad_norm": 0.25639674407942853, | |
| "learning_rate": 2.6586254388368995e-05, | |
| "loss": 0.5733, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.26196928635953026, | |
| "grad_norm": 0.25466281147535325, | |
| "learning_rate": 2.652903943020783e-05, | |
| "loss": 0.5862, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.26377597109304424, | |
| "grad_norm": 0.24151461009286013, | |
| "learning_rate": 2.647141164041398e-05, | |
| "loss": 0.5904, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.26558265582655827, | |
| "grad_norm": 0.259822033974557, | |
| "learning_rate": 2.641337308252228e-05, | |
| "loss": 0.581, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.26738934056007224, | |
| "grad_norm": 0.24532475377494284, | |
| "learning_rate": 2.6354925834776346e-05, | |
| "loss": 0.5817, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.26919602529358627, | |
| "grad_norm": 0.2555788655192275, | |
| "learning_rate": 2.6296071990054167e-05, | |
| "loss": 0.579, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.27100271002710025, | |
| "grad_norm": 0.2534339164884998, | |
| "learning_rate": 2.6236813655793123e-05, | |
| "loss": 0.5886, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.2728093947606143, | |
| "grad_norm": 0.25366511050468843, | |
| "learning_rate": 2.617715295391457e-05, | |
| "loss": 0.5707, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.27461607949412825, | |
| "grad_norm": 0.26118391438335414, | |
| "learning_rate": 2.6117092020747824e-05, | |
| "loss": 0.5779, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.2764227642276423, | |
| "grad_norm": 0.31049055796662123, | |
| "learning_rate": 2.6056633006953677e-05, | |
| "loss": 0.5783, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.27822944896115626, | |
| "grad_norm": 0.2611788471953917, | |
| "learning_rate": 2.5995778077447393e-05, | |
| "loss": 0.5823, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.2800361336946703, | |
| "grad_norm": 0.3259190805002546, | |
| "learning_rate": 2.5934529411321174e-05, | |
| "loss": 0.5743, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.28184281842818426, | |
| "grad_norm": 0.2222204544644696, | |
| "learning_rate": 2.587288920176613e-05, | |
| "loss": 0.575, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.2836495031616983, | |
| "grad_norm": 0.29042132848143276, | |
| "learning_rate": 2.581085965599375e-05, | |
| "loss": 0.5877, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.28545618789521227, | |
| "grad_norm": 0.2577824036794116, | |
| "learning_rate": 2.5748442995156882e-05, | |
| "loss": 0.5854, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.2872628726287263, | |
| "grad_norm": 0.24041412490024394, | |
| "learning_rate": 2.5685641454270172e-05, | |
| "loss": 0.566, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.28906955736224027, | |
| "grad_norm": 0.33137613211077116, | |
| "learning_rate": 2.5622457282130046e-05, | |
| "loss": 0.5753, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.2908762420957543, | |
| "grad_norm": 0.2252657452973701, | |
| "learning_rate": 2.5558892741234173e-05, | |
| "loss": 0.5827, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.2926829268292683, | |
| "grad_norm": 0.353225035311219, | |
| "learning_rate": 2.5494950107700482e-05, | |
| "loss": 0.5692, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.2944896115627823, | |
| "grad_norm": 0.2549710140850951, | |
| "learning_rate": 2.5430631671185616e-05, | |
| "loss": 0.5775, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.2962962962962963, | |
| "grad_norm": 0.320108207545845, | |
| "learning_rate": 2.5365939734802973e-05, | |
| "loss": 0.5871, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.2981029810298103, | |
| "grad_norm": 0.27118178617195304, | |
| "learning_rate": 2.5300876615040223e-05, | |
| "loss": 0.5777, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.2999096657633243, | |
| "grad_norm": 0.2732065296628375, | |
| "learning_rate": 2.523544464167637e-05, | |
| "loss": 0.5842, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.3017163504968383, | |
| "grad_norm": 0.2440418264084237, | |
| "learning_rate": 2.5169646157698313e-05, | |
| "loss": 0.583, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.3035230352303523, | |
| "grad_norm": 0.2400624724182379, | |
| "learning_rate": 2.5103483519216964e-05, | |
| "loss": 0.5813, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.3053297199638663, | |
| "grad_norm": 0.25905283588713374, | |
| "learning_rate": 2.5036959095382875e-05, | |
| "loss": 0.574, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.3071364046973803, | |
| "grad_norm": 0.2591290671871167, | |
| "learning_rate": 2.4970075268301388e-05, | |
| "loss": 0.5877, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.3089430894308943, | |
| "grad_norm": 0.2235510897395504, | |
| "learning_rate": 2.4902834432947353e-05, | |
| "loss": 0.5706, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.3107497741644083, | |
| "grad_norm": 0.23760061910249497, | |
| "learning_rate": 2.4835238997079382e-05, | |
| "loss": 0.5771, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.31255645889792233, | |
| "grad_norm": 0.2617951363502666, | |
| "learning_rate": 2.4767291381153603e-05, | |
| "loss": 0.5807, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.3143631436314363, | |
| "grad_norm": 0.23091611449448254, | |
| "learning_rate": 2.4698994018236994e-05, | |
| "loss": 0.5721, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.31616982836495033, | |
| "grad_norm": 0.25262572503013164, | |
| "learning_rate": 2.4630349353920284e-05, | |
| "loss": 0.5767, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.3179765130984643, | |
| "grad_norm": 0.23055115622552902, | |
| "learning_rate": 2.4561359846230346e-05, | |
| "loss": 0.5906, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.31978319783197834, | |
| "grad_norm": 0.21968694104114944, | |
| "learning_rate": 2.4492027965542217e-05, | |
| "loss": 0.5657, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.3215898825654923, | |
| "grad_norm": 0.2525014182918365, | |
| "learning_rate": 2.44223561944906e-05, | |
| "loss": 0.5724, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.32339656729900634, | |
| "grad_norm": 0.2407328065060556, | |
| "learning_rate": 2.4352347027881003e-05, | |
| "loss": 0.5825, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.3252032520325203, | |
| "grad_norm": 0.2547387307288275, | |
| "learning_rate": 2.4282002972600382e-05, | |
| "loss": 0.5908, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.32700993676603435, | |
| "grad_norm": 0.2263495601766316, | |
| "learning_rate": 2.4211326547527377e-05, | |
| "loss": 0.5688, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.3288166214995483, | |
| "grad_norm": 0.2789402543570071, | |
| "learning_rate": 2.4140320283442125e-05, | |
| "loss": 0.579, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.33062330623306235, | |
| "grad_norm": 0.2567309540458022, | |
| "learning_rate": 2.4068986722935625e-05, | |
| "loss": 0.5893, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.3324299909665763, | |
| "grad_norm": 0.24182513399667138, | |
| "learning_rate": 2.3997328420318705e-05, | |
| "loss": 0.5858, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.33423667570009036, | |
| "grad_norm": 0.2675738732389291, | |
| "learning_rate": 2.3925347941530556e-05, | |
| "loss": 0.5822, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.33604336043360433, | |
| "grad_norm": 0.23428952659343405, | |
| "learning_rate": 2.3853047864046843e-05, | |
| "loss": 0.5779, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.33785004516711836, | |
| "grad_norm": 0.26339154899807604, | |
| "learning_rate": 2.3780430776787413e-05, | |
| "loss": 0.5747, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.33965672990063234, | |
| "grad_norm": 0.21349849308750005, | |
| "learning_rate": 2.3707499280023604e-05, | |
| "loss": 0.5747, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.34146341463414637, | |
| "grad_norm": 0.26554041110868815, | |
| "learning_rate": 2.3634255985285104e-05, | |
| "loss": 0.5825, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.34327009936766034, | |
| "grad_norm": 0.23538164196617103, | |
| "learning_rate": 2.356070351526648e-05, | |
| "loss": 0.5708, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.34507678410117437, | |
| "grad_norm": 0.24516795740455555, | |
| "learning_rate": 2.348684450373322e-05, | |
| "loss": 0.5691, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.34688346883468835, | |
| "grad_norm": 0.2320939063694949, | |
| "learning_rate": 2.3412681595427467e-05, | |
| "loss": 0.5761, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.3486901535682023, | |
| "grad_norm": 0.24367157013907606, | |
| "learning_rate": 2.3338217445973268e-05, | |
| "loss": 0.5776, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.35049683830171635, | |
| "grad_norm": 0.2383952654545809, | |
| "learning_rate": 2.3263454721781537e-05, | |
| "loss": 0.5793, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.3523035230352303, | |
| "grad_norm": 0.21970860044862534, | |
| "learning_rate": 2.318839609995453e-05, | |
| "loss": 0.5625, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.35411020776874436, | |
| "grad_norm": 0.22393266445368756, | |
| "learning_rate": 2.3113044268189995e-05, | |
| "loss": 0.5758, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.35591689250225833, | |
| "grad_norm": 0.20965253283535495, | |
| "learning_rate": 2.303740192468495e-05, | |
| "loss": 0.5754, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.35772357723577236, | |
| "grad_norm": 0.2134213495454822, | |
| "learning_rate": 2.2961471778039045e-05, | |
| "loss": 0.5521, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.35953026196928634, | |
| "grad_norm": 0.2077804399624621, | |
| "learning_rate": 2.288525654715757e-05, | |
| "loss": 0.5758, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.36133694670280037, | |
| "grad_norm": 0.21089298206975676, | |
| "learning_rate": 2.280875896115413e-05, | |
| "loss": 0.5645, | |
| "step": 200 | |
| } | |
| ], | |
| "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": 231524290789376.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |