| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.19282219383451035, | |
| "eval_steps": 500, | |
| "global_step": 3000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0006427406461150345, | |
| "grad_norm": 0.458984375, | |
| "learning_rate": 1e-05, | |
| "loss": 3.0419, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.001285481292230069, | |
| "grad_norm": 8.8125, | |
| "learning_rate": 2e-05, | |
| "loss": 3.0641, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0019282219383451036, | |
| "grad_norm": 0.62109375, | |
| "learning_rate": 1.9999950454033063e-05, | |
| "loss": 3.2405, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.002570962584460138, | |
| "grad_norm": 0.5625, | |
| "learning_rate": 1.9999801816623205e-05, | |
| "loss": 2.8447, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.0032137032305751727, | |
| "grad_norm": 0.60546875, | |
| "learning_rate": 1.9999554089243305e-05, | |
| "loss": 2.9348, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.003856443876690207, | |
| "grad_norm": 0.69921875, | |
| "learning_rate": 1.9999207274348143e-05, | |
| "loss": 2.966, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.004499184522805242, | |
| "grad_norm": 2.9375, | |
| "learning_rate": 1.9998761375374376e-05, | |
| "loss": 2.8849, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.005141925168920276, | |
| "grad_norm": 0.953125, | |
| "learning_rate": 1.9998216396740497e-05, | |
| "loss": 2.8674, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.0057846658150353105, | |
| "grad_norm": 0.96484375, | |
| "learning_rate": 1.9997572343846814e-05, | |
| "loss": 2.789, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.006427406461150345, | |
| "grad_norm": 0.63671875, | |
| "learning_rate": 1.9996829223075363e-05, | |
| "loss": 2.8306, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.0070701471072653795, | |
| "grad_norm": 0.91796875, | |
| "learning_rate": 1.9995987041789876e-05, | |
| "loss": 2.5481, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.007712887753380414, | |
| "grad_norm": 0.84765625, | |
| "learning_rate": 1.999504580833569e-05, | |
| "loss": 2.6249, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.00835562839949545, | |
| "grad_norm": 0.54296875, | |
| "learning_rate": 1.9994005532039665e-05, | |
| "loss": 2.6078, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.008998369045610483, | |
| "grad_norm": 1.328125, | |
| "learning_rate": 1.9992866223210105e-05, | |
| "loss": 2.4607, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.009641109691725517, | |
| "grad_norm": 1.0703125, | |
| "learning_rate": 1.999162789313664e-05, | |
| "loss": 2.3672, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.010283850337840551, | |
| "grad_norm": 0.796875, | |
| "learning_rate": 1.9990290554090123e-05, | |
| "loss": 2.4135, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.010926590983955587, | |
| "grad_norm": 1.1484375, | |
| "learning_rate": 1.9988854219322507e-05, | |
| "loss": 2.5179, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.011569331630070621, | |
| "grad_norm": 0.55078125, | |
| "learning_rate": 1.9987318903066704e-05, | |
| "loss": 2.3391, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.012212072276185655, | |
| "grad_norm": 1.171875, | |
| "learning_rate": 1.9985684620536466e-05, | |
| "loss": 2.4168, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.01285481292230069, | |
| "grad_norm": 1.4609375, | |
| "learning_rate": 1.9983951387926216e-05, | |
| "loss": 2.2706, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.013497553568415725, | |
| "grad_norm": 0.68359375, | |
| "learning_rate": 1.998211922241088e-05, | |
| "loss": 2.2684, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.014140294214530759, | |
| "grad_norm": 1.1640625, | |
| "learning_rate": 1.9980188142145755e-05, | |
| "loss": 2.2022, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.014783034860645793, | |
| "grad_norm": 0.52734375, | |
| "learning_rate": 1.997815816626628e-05, | |
| "loss": 2.2821, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.015425775506760829, | |
| "grad_norm": 1.359375, | |
| "learning_rate": 1.9976029314887882e-05, | |
| "loss": 2.1055, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.016068516152875863, | |
| "grad_norm": 0.890625, | |
| "learning_rate": 1.9973801609105757e-05, | |
| "loss": 2.1481, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.0167112567989909, | |
| "grad_norm": 1.2265625, | |
| "learning_rate": 1.9971475070994675e-05, | |
| "loss": 2.1277, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.01735399744510593, | |
| "grad_norm": 0.6796875, | |
| "learning_rate": 1.9969049723608753e-05, | |
| "loss": 2.1143, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.017996738091220967, | |
| "grad_norm": 0.828125, | |
| "learning_rate": 1.9966525590981228e-05, | |
| "loss": 2.1824, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.018639478737336002, | |
| "grad_norm": 1.6171875, | |
| "learning_rate": 1.9963902698124212e-05, | |
| "loss": 2.1849, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.019282219383451035, | |
| "grad_norm": 0.96875, | |
| "learning_rate": 1.996118107102847e-05, | |
| "loss": 2.1996, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.01992496002956607, | |
| "grad_norm": 1.1171875, | |
| "learning_rate": 1.9958360736663117e-05, | |
| "loss": 2.1914, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.020567700675681103, | |
| "grad_norm": 0.62109375, | |
| "learning_rate": 1.99554417229754e-05, | |
| "loss": 2.0519, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.02121044132179614, | |
| "grad_norm": 0.890625, | |
| "learning_rate": 1.995242405889039e-05, | |
| "loss": 2.0804, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.021853181967911174, | |
| "grad_norm": 0.71875, | |
| "learning_rate": 1.99493077743107e-05, | |
| "loss": 2.0424, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.022495922614026206, | |
| "grad_norm": 1.4296875, | |
| "learning_rate": 1.99460929001162e-05, | |
| "loss": 2.1151, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.023138663260141242, | |
| "grad_norm": 0.7421875, | |
| "learning_rate": 1.9942779468163696e-05, | |
| "loss": 1.8499, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.023781403906256278, | |
| "grad_norm": 0.8671875, | |
| "learning_rate": 1.9939367511286635e-05, | |
| "loss": 2.1271, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.02442414455237131, | |
| "grad_norm": 0.75, | |
| "learning_rate": 1.993585706329475e-05, | |
| "loss": 2.0739, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.025066885198486346, | |
| "grad_norm": 1.3671875, | |
| "learning_rate": 1.9932248158973746e-05, | |
| "loss": 2.0379, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.02570962584460138, | |
| "grad_norm": 0.78515625, | |
| "learning_rate": 1.992854083408496e-05, | |
| "loss": 1.9566, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.026352366490716414, | |
| "grad_norm": 0.859375, | |
| "learning_rate": 1.992473512536499e-05, | |
| "loss": 2.0377, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.02699510713683145, | |
| "grad_norm": 0.81640625, | |
| "learning_rate": 1.992083107052534e-05, | |
| "loss": 1.9839, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.027637847782946486, | |
| "grad_norm": 0.80859375, | |
| "learning_rate": 1.9916828708252046e-05, | |
| "loss": 1.9477, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.028280588429061518, | |
| "grad_norm": 0.703125, | |
| "learning_rate": 1.9912728078205285e-05, | |
| "loss": 2.0033, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.028923329075176554, | |
| "grad_norm": 1.03125, | |
| "learning_rate": 1.9908529221018994e-05, | |
| "loss": 1.9196, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.029566069721291586, | |
| "grad_norm": 1.0546875, | |
| "learning_rate": 1.9904232178300465e-05, | |
| "loss": 1.9712, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.03020881036740662, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 1.9899836992629922e-05, | |
| "loss": 1.9251, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.030851551013521657, | |
| "grad_norm": 0.71484375, | |
| "learning_rate": 1.989534370756011e-05, | |
| "loss": 1.9695, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.03149429165963669, | |
| "grad_norm": 0.9765625, | |
| "learning_rate": 1.989075236761586e-05, | |
| "loss": 1.9208, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.032137032305751725, | |
| "grad_norm": 0.74609375, | |
| "learning_rate": 1.988606301829365e-05, | |
| "loss": 1.9474, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.03277977295186676, | |
| "grad_norm": 0.6796875, | |
| "learning_rate": 1.9881275706061146e-05, | |
| "loss": 1.8636, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.0334225135979818, | |
| "grad_norm": 0.9453125, | |
| "learning_rate": 1.987639047835675e-05, | |
| "loss": 1.9595, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.03406525424409683, | |
| "grad_norm": 0.9921875, | |
| "learning_rate": 1.987140738358913e-05, | |
| "loss": 1.8648, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.03470799489021186, | |
| "grad_norm": 1.453125, | |
| "learning_rate": 1.9866326471136738e-05, | |
| "loss": 1.9641, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.0353507355363269, | |
| "grad_norm": 1.1328125, | |
| "learning_rate": 1.986114779134731e-05, | |
| "loss": 1.9649, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.03599347618244193, | |
| "grad_norm": 0.921875, | |
| "learning_rate": 1.9855871395537395e-05, | |
| "loss": 1.9086, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.036636216828556965, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.9850497335991815e-05, | |
| "loss": 1.8802, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.037278957474672005, | |
| "grad_norm": 0.78125, | |
| "learning_rate": 1.984502566596317e-05, | |
| "loss": 1.8694, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.03792169812078704, | |
| "grad_norm": 0.984375, | |
| "learning_rate": 1.9839456439671288e-05, | |
| "loss": 1.9474, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.03856443876690207, | |
| "grad_norm": 0.87890625, | |
| "learning_rate": 1.9833789712302714e-05, | |
| "loss": 1.8633, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.03920717941301711, | |
| "grad_norm": 0.8515625, | |
| "learning_rate": 1.9828025540010144e-05, | |
| "loss": 1.9034, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.03984992005913214, | |
| "grad_norm": 0.77734375, | |
| "learning_rate": 1.9822163979911878e-05, | |
| "loss": 1.9021, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.04049266070524717, | |
| "grad_norm": 1.2421875, | |
| "learning_rate": 1.981620509009125e-05, | |
| "loss": 1.8945, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.041135401351362205, | |
| "grad_norm": 0.87109375, | |
| "learning_rate": 1.9810148929596047e-05, | |
| "loss": 1.9007, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.041778141997477244, | |
| "grad_norm": 0.85546875, | |
| "learning_rate": 1.9803995558437938e-05, | |
| "loss": 1.8548, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.04242088264359228, | |
| "grad_norm": 0.87890625, | |
| "learning_rate": 1.9797745037591866e-05, | |
| "loss": 1.952, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.04306362328970731, | |
| "grad_norm": 1.265625, | |
| "learning_rate": 1.979139742899545e-05, | |
| "loss": 1.8855, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.04370636393582235, | |
| "grad_norm": 0.7734375, | |
| "learning_rate": 1.9784952795548374e-05, | |
| "loss": 1.8523, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.04434910458193738, | |
| "grad_norm": 0.99609375, | |
| "learning_rate": 1.9778411201111755e-05, | |
| "loss": 1.808, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.04499184522805241, | |
| "grad_norm": 1.4296875, | |
| "learning_rate": 1.9771772710507516e-05, | |
| "loss": 1.8894, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.04563458587416745, | |
| "grad_norm": 1.0390625, | |
| "learning_rate": 1.976503738951775e-05, | |
| "loss": 1.8869, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.046277326520282484, | |
| "grad_norm": 1.1640625, | |
| "learning_rate": 1.9758205304884048e-05, | |
| "loss": 1.9048, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.04692006716639752, | |
| "grad_norm": 1.015625, | |
| "learning_rate": 1.9751276524306865e-05, | |
| "loss": 1.848, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.047562807812512556, | |
| "grad_norm": 1.0546875, | |
| "learning_rate": 1.974425111644482e-05, | |
| "loss": 1.8434, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.04820554845862759, | |
| "grad_norm": 0.8828125, | |
| "learning_rate": 1.9737129150914045e-05, | |
| "loss": 1.9363, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.04884828910474262, | |
| "grad_norm": 1.109375, | |
| "learning_rate": 1.972991069828747e-05, | |
| "loss": 1.8212, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.04949102975085766, | |
| "grad_norm": 1.0390625, | |
| "learning_rate": 1.9722595830094138e-05, | |
| "loss": 1.8591, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.05013377039697269, | |
| "grad_norm": 1.0, | |
| "learning_rate": 1.9715184618818493e-05, | |
| "loss": 1.9277, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.050776511043087724, | |
| "grad_norm": 0.921875, | |
| "learning_rate": 1.9707677137899662e-05, | |
| "loss": 1.8552, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.05141925168920276, | |
| "grad_norm": 0.89453125, | |
| "learning_rate": 1.9700073461730725e-05, | |
| "loss": 1.8265, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.052061992335317796, | |
| "grad_norm": 1.3125, | |
| "learning_rate": 1.969237366565798e-05, | |
| "loss": 1.7932, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.05270473298143283, | |
| "grad_norm": 0.76171875, | |
| "learning_rate": 1.9684577825980192e-05, | |
| "loss": 1.8122, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.05334747362754787, | |
| "grad_norm": 0.9296875, | |
| "learning_rate": 1.9676686019947852e-05, | |
| "loss": 1.8671, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.0539902142736629, | |
| "grad_norm": 0.9765625, | |
| "learning_rate": 1.9668698325762378e-05, | |
| "loss": 1.884, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.05463295491977793, | |
| "grad_norm": 1.0234375, | |
| "learning_rate": 1.9660614822575394e-05, | |
| "loss": 1.9472, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.05527569556589297, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 1.9652435590487878e-05, | |
| "loss": 1.7799, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.055918436212008, | |
| "grad_norm": 0.8125, | |
| "learning_rate": 1.964416071054944e-05, | |
| "loss": 1.8548, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.056561176858123036, | |
| "grad_norm": 1.15625, | |
| "learning_rate": 1.963579026475745e-05, | |
| "loss": 1.8213, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.05720391750423807, | |
| "grad_norm": 1.0078125, | |
| "learning_rate": 1.962732433605629e-05, | |
| "loss": 1.8675, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.05784665815035311, | |
| "grad_norm": 1.0625, | |
| "learning_rate": 1.961876300833647e-05, | |
| "loss": 1.8382, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.05848939879646814, | |
| "grad_norm": 1.3359375, | |
| "learning_rate": 1.9610106366433846e-05, | |
| "loss": 1.8516, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.05913213944258317, | |
| "grad_norm": 1.1796875, | |
| "learning_rate": 1.9601354496128765e-05, | |
| "loss": 1.8137, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.05977488008869821, | |
| "grad_norm": 0.8046875, | |
| "learning_rate": 1.9592507484145193e-05, | |
| "loss": 1.9144, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.06041762073481324, | |
| "grad_norm": 0.796875, | |
| "learning_rate": 1.9583565418149887e-05, | |
| "loss": 1.8284, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.061060361380928276, | |
| "grad_norm": 0.90234375, | |
| "learning_rate": 1.9574528386751507e-05, | |
| "loss": 1.7851, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.061703102027043315, | |
| "grad_norm": 0.8828125, | |
| "learning_rate": 1.9565396479499744e-05, | |
| "loss": 1.769, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.06234584267315835, | |
| "grad_norm": 0.921875, | |
| "learning_rate": 1.9556169786884436e-05, | |
| "loss": 1.8202, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.06298858331927339, | |
| "grad_norm": 1.015625, | |
| "learning_rate": 1.9546848400334658e-05, | |
| "loss": 1.8815, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.06363132396538841, | |
| "grad_norm": 0.82421875, | |
| "learning_rate": 1.953743241221784e-05, | |
| "loss": 1.866, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.06427406461150345, | |
| "grad_norm": 0.87109375, | |
| "learning_rate": 1.9527921915838827e-05, | |
| "loss": 1.8496, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.06491680525761849, | |
| "grad_norm": 1.6015625, | |
| "learning_rate": 1.9518317005438964e-05, | |
| "loss": 1.7597, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.06555954590373352, | |
| "grad_norm": 1.0625, | |
| "learning_rate": 1.9508617776195167e-05, | |
| "loss": 1.8705, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.06620228654984855, | |
| "grad_norm": 0.90234375, | |
| "learning_rate": 1.9498824324218973e-05, | |
| "loss": 1.787, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.0668450271959636, | |
| "grad_norm": 1.1328125, | |
| "learning_rate": 1.9488936746555593e-05, | |
| "loss": 1.7781, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.06748776784207862, | |
| "grad_norm": 0.8046875, | |
| "learning_rate": 1.9478955141182948e-05, | |
| "loss": 1.7506, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.06813050848819366, | |
| "grad_norm": 0.94921875, | |
| "learning_rate": 1.946887960701069e-05, | |
| "loss": 1.8555, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.0687732491343087, | |
| "grad_norm": 0.8515625, | |
| "learning_rate": 1.9458710243879242e-05, | |
| "loss": 1.7989, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.06941598978042372, | |
| "grad_norm": 1.109375, | |
| "learning_rate": 1.944844715255879e-05, | |
| "loss": 1.7939, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.07005873042653876, | |
| "grad_norm": 1.046875, | |
| "learning_rate": 1.9438090434748285e-05, | |
| "loss": 1.7889, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.0707014710726538, | |
| "grad_norm": 0.90625, | |
| "learning_rate": 1.942764019307445e-05, | |
| "loss": 1.7511, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.07134421171876883, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 1.9417096531090752e-05, | |
| "loss": 1.8022, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.07198695236488387, | |
| "grad_norm": 1.1640625, | |
| "learning_rate": 1.940645955327637e-05, | |
| "loss": 1.8737, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.0726296930109989, | |
| "grad_norm": 0.98046875, | |
| "learning_rate": 1.9395729365035185e-05, | |
| "loss": 1.764, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.07327243365711393, | |
| "grad_norm": 1.6171875, | |
| "learning_rate": 1.9384906072694703e-05, | |
| "loss": 1.8344, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.07391517430322897, | |
| "grad_norm": 1.1328125, | |
| "learning_rate": 1.9373989783505017e-05, | |
| "loss": 1.8173, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.07455791494934401, | |
| "grad_norm": 0.9375, | |
| "learning_rate": 1.936298060563775e-05, | |
| "loss": 1.8271, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.07520065559545903, | |
| "grad_norm": 0.83203125, | |
| "learning_rate": 1.9351878648184973e-05, | |
| "loss": 1.8369, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.07584339624157407, | |
| "grad_norm": 1.1328125, | |
| "learning_rate": 1.9340684021158133e-05, | |
| "loss": 1.8255, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.07648613688768911, | |
| "grad_norm": 1.125, | |
| "learning_rate": 1.932939683548695e-05, | |
| "loss": 1.7836, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.07712887753380414, | |
| "grad_norm": 0.80078125, | |
| "learning_rate": 1.9318017203018334e-05, | |
| "loss": 1.7955, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.07777161817991918, | |
| "grad_norm": 0.73828125, | |
| "learning_rate": 1.9306545236515264e-05, | |
| "loss": 1.748, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.07841435882603422, | |
| "grad_norm": 1.5078125, | |
| "learning_rate": 1.929498104965567e-05, | |
| "loss": 1.7522, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.07905709947214924, | |
| "grad_norm": 1.3828125, | |
| "learning_rate": 1.928332475703132e-05, | |
| "loss": 1.8157, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.07969984011826428, | |
| "grad_norm": 0.83984375, | |
| "learning_rate": 1.9271576474146667e-05, | |
| "loss": 1.7609, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.0803425807643793, | |
| "grad_norm": 0.84765625, | |
| "learning_rate": 1.9259736317417723e-05, | |
| "loss": 1.8088, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.08098532141049435, | |
| "grad_norm": 1.40625, | |
| "learning_rate": 1.9247804404170888e-05, | |
| "loss": 1.8767, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.08162806205660939, | |
| "grad_norm": 0.87890625, | |
| "learning_rate": 1.9235780852641798e-05, | |
| "loss": 1.7125, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.08227080270272441, | |
| "grad_norm": 1.078125, | |
| "learning_rate": 1.9223665781974154e-05, | |
| "loss": 1.8067, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.08291354334883945, | |
| "grad_norm": 0.890625, | |
| "learning_rate": 1.921145931221853e-05, | |
| "loss": 1.669, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.08355628399495449, | |
| "grad_norm": 1.0, | |
| "learning_rate": 1.9199161564331196e-05, | |
| "loss": 1.7822, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.08419902464106951, | |
| "grad_norm": 1.03125, | |
| "learning_rate": 1.9186772660172916e-05, | |
| "loss": 1.8901, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.08484176528718455, | |
| "grad_norm": 0.75390625, | |
| "learning_rate": 1.9174292722507735e-05, | |
| "loss": 1.7516, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.08548450593329959, | |
| "grad_norm": 1.1328125, | |
| "learning_rate": 1.916172187500177e-05, | |
| "loss": 1.8496, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.08612724657941462, | |
| "grad_norm": 1.046875, | |
| "learning_rate": 1.914906024222198e-05, | |
| "loss": 1.8261, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.08676998722552966, | |
| "grad_norm": 0.953125, | |
| "learning_rate": 1.913630794963493e-05, | |
| "loss": 1.7513, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.0874127278716447, | |
| "grad_norm": 0.73046875, | |
| "learning_rate": 1.9123465123605558e-05, | |
| "loss": 1.6671, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.08805546851775972, | |
| "grad_norm": 1.4921875, | |
| "learning_rate": 1.9110531891395906e-05, | |
| "loss": 1.8489, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.08869820916387476, | |
| "grad_norm": 0.67578125, | |
| "learning_rate": 1.9097508381163877e-05, | |
| "loss": 1.7958, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.0893409498099898, | |
| "grad_norm": 0.8359375, | |
| "learning_rate": 1.9084394721961956e-05, | |
| "loss": 1.7382, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.08998369045610483, | |
| "grad_norm": 1.296875, | |
| "learning_rate": 1.907119104373592e-05, | |
| "loss": 1.8107, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.09062643110221986, | |
| "grad_norm": 1.140625, | |
| "learning_rate": 1.9057897477323572e-05, | |
| "loss": 1.829, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.0912691717483349, | |
| "grad_norm": 1.078125, | |
| "learning_rate": 1.9044514154453434e-05, | |
| "loss": 1.7685, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.09191191239444993, | |
| "grad_norm": 0.890625, | |
| "learning_rate": 1.903104120774344e-05, | |
| "loss": 1.8001, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.09255465304056497, | |
| "grad_norm": 1.078125, | |
| "learning_rate": 1.9017478770699624e-05, | |
| "loss": 1.7632, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.09319739368668001, | |
| "grad_norm": 0.99609375, | |
| "learning_rate": 1.90038269777148e-05, | |
| "loss": 1.7482, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.09384013433279503, | |
| "grad_norm": 1.421875, | |
| "learning_rate": 1.899008596406722e-05, | |
| "loss": 1.8016, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.09448287497891007, | |
| "grad_norm": 0.96875, | |
| "learning_rate": 1.897625586591925e-05, | |
| "loss": 1.7998, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.09512561562502511, | |
| "grad_norm": 0.796875, | |
| "learning_rate": 1.896233682031601e-05, | |
| "loss": 1.7887, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.09576835627114014, | |
| "grad_norm": 1.0078125, | |
| "learning_rate": 1.8948328965184004e-05, | |
| "loss": 1.7226, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.09641109691725518, | |
| "grad_norm": 1.09375, | |
| "learning_rate": 1.8934232439329786e-05, | |
| "loss": 1.7656, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.09705383756337022, | |
| "grad_norm": 0.7734375, | |
| "learning_rate": 1.8920047382438554e-05, | |
| "loss": 1.7482, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.09769657820948524, | |
| "grad_norm": 0.63671875, | |
| "learning_rate": 1.890577393507278e-05, | |
| "loss": 1.7171, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.09833931885560028, | |
| "grad_norm": 0.8984375, | |
| "learning_rate": 1.8891412238670814e-05, | |
| "loss": 1.7527, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.09898205950171532, | |
| "grad_norm": 0.9453125, | |
| "learning_rate": 1.887696243554549e-05, | |
| "loss": 1.7824, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.09962480014783034, | |
| "grad_norm": 1.03125, | |
| "learning_rate": 1.8862424668882687e-05, | |
| "loss": 1.8317, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.10026754079394538, | |
| "grad_norm": 0.81640625, | |
| "learning_rate": 1.8847799082739957e-05, | |
| "loss": 1.785, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.10091028144006042, | |
| "grad_norm": 1.1171875, | |
| "learning_rate": 1.8833085822045062e-05, | |
| "loss": 1.7549, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 0.10155302208617545, | |
| "grad_norm": 1.109375, | |
| "learning_rate": 1.8818285032594537e-05, | |
| "loss": 1.803, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.10219576273229049, | |
| "grad_norm": 1.046875, | |
| "learning_rate": 1.8803396861052284e-05, | |
| "loss": 1.7646, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.10283850337840553, | |
| "grad_norm": 1.1171875, | |
| "learning_rate": 1.878842145494806e-05, | |
| "loss": 1.7383, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.10348124402452055, | |
| "grad_norm": 1.140625, | |
| "learning_rate": 1.8773358962676066e-05, | |
| "loss": 1.7851, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 0.10412398467063559, | |
| "grad_norm": 1.015625, | |
| "learning_rate": 1.8758209533493447e-05, | |
| "loss": 1.8107, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.10476672531675063, | |
| "grad_norm": 0.94140625, | |
| "learning_rate": 1.874297331751883e-05, | |
| "loss": 1.8237, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 0.10540946596286566, | |
| "grad_norm": 1.046875, | |
| "learning_rate": 1.8727650465730827e-05, | |
| "loss": 1.8021, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.1060522066089807, | |
| "grad_norm": 0.94140625, | |
| "learning_rate": 1.8712241129966532e-05, | |
| "loss": 1.7842, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.10669494725509573, | |
| "grad_norm": 1.0625, | |
| "learning_rate": 1.8696745462920046e-05, | |
| "loss": 1.7979, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.10733768790121076, | |
| "grad_norm": 1.234375, | |
| "learning_rate": 1.868116361814092e-05, | |
| "loss": 1.746, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 0.1079804285473258, | |
| "grad_norm": 0.91015625, | |
| "learning_rate": 1.8665495750032664e-05, | |
| "loss": 1.8202, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.10862316919344084, | |
| "grad_norm": 0.91796875, | |
| "learning_rate": 1.8649742013851225e-05, | |
| "loss": 1.7969, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 0.10926590983955586, | |
| "grad_norm": 0.859375, | |
| "learning_rate": 1.8633902565703415e-05, | |
| "loss": 1.7795, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.1099086504856709, | |
| "grad_norm": 1.5546875, | |
| "learning_rate": 1.861797756254539e-05, | |
| "loss": 1.7474, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 0.11055139113178594, | |
| "grad_norm": 1.0078125, | |
| "learning_rate": 1.8601967162181082e-05, | |
| "loss": 1.8134, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 0.11119413177790097, | |
| "grad_norm": 1.0, | |
| "learning_rate": 1.8585871523260653e-05, | |
| "loss": 1.7977, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 0.111836872424016, | |
| "grad_norm": 0.7734375, | |
| "learning_rate": 1.8569690805278894e-05, | |
| "loss": 1.7256, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.11247961307013103, | |
| "grad_norm": 0.8046875, | |
| "learning_rate": 1.8553425168573667e-05, | |
| "loss": 1.7866, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.11312235371624607, | |
| "grad_norm": 1.0234375, | |
| "learning_rate": 1.8537074774324318e-05, | |
| "loss": 1.7567, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 0.11376509436236111, | |
| "grad_norm": 1.328125, | |
| "learning_rate": 1.8520639784550068e-05, | |
| "loss": 1.769, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 0.11440783500847614, | |
| "grad_norm": 1.125, | |
| "learning_rate": 1.85041203621084e-05, | |
| "loss": 1.7404, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 0.11505057565459118, | |
| "grad_norm": 1.078125, | |
| "learning_rate": 1.848751667069347e-05, | |
| "loss": 1.785, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 0.11569331630070621, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 1.847082887483447e-05, | |
| "loss": 1.6875, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.11633605694682124, | |
| "grad_norm": 0.8828125, | |
| "learning_rate": 1.8454057139893992e-05, | |
| "loss": 1.7259, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 0.11697879759293628, | |
| "grad_norm": 1.0078125, | |
| "learning_rate": 1.84372016320664e-05, | |
| "loss": 1.7666, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 0.11762153823905132, | |
| "grad_norm": 1.0078125, | |
| "learning_rate": 1.842026251837619e-05, | |
| "loss": 1.7275, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 0.11826427888516634, | |
| "grad_norm": 1.2265625, | |
| "learning_rate": 1.8403239966676304e-05, | |
| "loss": 1.8194, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 0.11890701953128138, | |
| "grad_norm": 1.1328125, | |
| "learning_rate": 1.8386134145646505e-05, | |
| "loss": 1.7619, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.11954976017739642, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 1.836894522479168e-05, | |
| "loss": 1.761, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 0.12019250082351145, | |
| "grad_norm": 0.82421875, | |
| "learning_rate": 1.835167337444017e-05, | |
| "loss": 1.722, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 0.12083524146962649, | |
| "grad_norm": 0.9375, | |
| "learning_rate": 1.8334318765742078e-05, | |
| "loss": 1.7103, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 0.12147798211574153, | |
| "grad_norm": 1.5703125, | |
| "learning_rate": 1.8316881570667583e-05, | |
| "loss": 1.7891, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 0.12212072276185655, | |
| "grad_norm": 0.77734375, | |
| "learning_rate": 1.8299361962005218e-05, | |
| "loss": 1.7342, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.12276346340797159, | |
| "grad_norm": 1.125, | |
| "learning_rate": 1.8281760113360177e-05, | |
| "loss": 1.6901, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 0.12340620405408663, | |
| "grad_norm": 1.203125, | |
| "learning_rate": 1.8264076199152582e-05, | |
| "loss": 1.7208, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 0.12404894470020165, | |
| "grad_norm": 1.109375, | |
| "learning_rate": 1.8246310394615753e-05, | |
| "loss": 1.7616, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 0.1246916853463167, | |
| "grad_norm": 0.71484375, | |
| "learning_rate": 1.822846287579449e-05, | |
| "loss": 1.7558, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 0.12533442599243172, | |
| "grad_norm": 1.3125, | |
| "learning_rate": 1.8210533819543304e-05, | |
| "loss": 1.7703, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.12597716663854677, | |
| "grad_norm": 0.83203125, | |
| "learning_rate": 1.8192523403524685e-05, | |
| "loss": 1.7272, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 0.1266199072846618, | |
| "grad_norm": 0.71875, | |
| "learning_rate": 1.8174431806207326e-05, | |
| "loss": 1.7556, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 0.12726264793077682, | |
| "grad_norm": 1.2890625, | |
| "learning_rate": 1.815625920686436e-05, | |
| "loss": 1.7382, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 0.12790538857689188, | |
| "grad_norm": 0.93359375, | |
| "learning_rate": 1.813800578557159e-05, | |
| "loss": 1.754, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 0.1285481292230069, | |
| "grad_norm": 0.98046875, | |
| "learning_rate": 1.8119671723205708e-05, | |
| "loss": 1.8341, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.12919086986912193, | |
| "grad_norm": 0.921875, | |
| "learning_rate": 1.8101257201442468e-05, | |
| "loss": 1.8343, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 0.12983361051523698, | |
| "grad_norm": 0.95703125, | |
| "learning_rate": 1.8082762402754936e-05, | |
| "loss": 1.7762, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 0.130476351161352, | |
| "grad_norm": 0.8125, | |
| "learning_rate": 1.8064187510411646e-05, | |
| "loss": 1.7232, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 0.13111909180746703, | |
| "grad_norm": 1.15625, | |
| "learning_rate": 1.80455327084748e-05, | |
| "loss": 1.7593, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 0.13176183245358208, | |
| "grad_norm": 0.890625, | |
| "learning_rate": 1.8026798181798434e-05, | |
| "loss": 1.6904, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.1324045730996971, | |
| "grad_norm": 1.0703125, | |
| "learning_rate": 1.8007984116026604e-05, | |
| "loss": 1.7421, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 0.13304731374581213, | |
| "grad_norm": 1.078125, | |
| "learning_rate": 1.7989090697591517e-05, | |
| "loss": 1.7673, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 0.1336900543919272, | |
| "grad_norm": 0.83984375, | |
| "learning_rate": 1.7970118113711715e-05, | |
| "loss": 1.7927, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 0.1343327950380422, | |
| "grad_norm": 0.9375, | |
| "learning_rate": 1.7951066552390204e-05, | |
| "loss": 1.781, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 0.13497553568415724, | |
| "grad_norm": 1.3203125, | |
| "learning_rate": 1.7931936202412582e-05, | |
| "loss": 1.6528, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.1356182763302723, | |
| "grad_norm": 0.75390625, | |
| "learning_rate": 1.791272725334519e-05, | |
| "loss": 1.7846, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 0.13626101697638732, | |
| "grad_norm": 0.9765625, | |
| "learning_rate": 1.789343989553322e-05, | |
| "loss": 1.7236, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 0.13690375762250234, | |
| "grad_norm": 1.0546875, | |
| "learning_rate": 1.787407432009883e-05, | |
| "loss": 1.7879, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 0.1375464982686174, | |
| "grad_norm": 0.87109375, | |
| "learning_rate": 1.7854630718939254e-05, | |
| "loss": 1.6429, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 0.13818923891473242, | |
| "grad_norm": 1.4921875, | |
| "learning_rate": 1.7835109284724886e-05, | |
| "loss": 1.7904, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.13883197956084745, | |
| "grad_norm": 0.828125, | |
| "learning_rate": 1.7815510210897407e-05, | |
| "loss": 1.7258, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 0.1394747202069625, | |
| "grad_norm": 1.0703125, | |
| "learning_rate": 1.779583369166782e-05, | |
| "loss": 1.6941, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 0.14011746085307752, | |
| "grad_norm": 0.9375, | |
| "learning_rate": 1.7776079922014564e-05, | |
| "loss": 1.7171, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 0.14076020149919255, | |
| "grad_norm": 1.171875, | |
| "learning_rate": 1.7756249097681563e-05, | |
| "loss": 1.7273, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 0.1414029421453076, | |
| "grad_norm": 1.1640625, | |
| "learning_rate": 1.7736341415176286e-05, | |
| "loss": 1.722, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.14204568279142263, | |
| "grad_norm": 0.9140625, | |
| "learning_rate": 1.7716357071767812e-05, | |
| "loss": 1.711, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 0.14268842343753765, | |
| "grad_norm": 0.9140625, | |
| "learning_rate": 1.7696296265484863e-05, | |
| "loss": 1.7142, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 0.1433311640836527, | |
| "grad_norm": 0.96875, | |
| "learning_rate": 1.7676159195113853e-05, | |
| "loss": 1.7341, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 0.14397390472976773, | |
| "grad_norm": 0.9921875, | |
| "learning_rate": 1.7655946060196894e-05, | |
| "loss": 1.7583, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 0.14461664537588276, | |
| "grad_norm": 0.96875, | |
| "learning_rate": 1.7635657061029863e-05, | |
| "loss": 1.6996, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.1452593860219978, | |
| "grad_norm": 0.91015625, | |
| "learning_rate": 1.7615292398660372e-05, | |
| "loss": 1.7818, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 0.14590212666811284, | |
| "grad_norm": 1.546875, | |
| "learning_rate": 1.7594852274885795e-05, | |
| "loss": 1.7629, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 0.14654486731422786, | |
| "grad_norm": 1.5625, | |
| "learning_rate": 1.7574336892251277e-05, | |
| "loss": 1.7498, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 0.14718760796034291, | |
| "grad_norm": 1.3046875, | |
| "learning_rate": 1.7553746454047705e-05, | |
| "loss": 1.7655, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 0.14783034860645794, | |
| "grad_norm": 1.5234375, | |
| "learning_rate": 1.753308116430972e-05, | |
| "loss": 1.7429, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.14847308925257297, | |
| "grad_norm": 1.015625, | |
| "learning_rate": 1.751234122781367e-05, | |
| "loss": 1.7795, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 0.14911582989868802, | |
| "grad_norm": 1.40625, | |
| "learning_rate": 1.7491526850075593e-05, | |
| "loss": 1.7375, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 0.14975857054480304, | |
| "grad_norm": 0.88671875, | |
| "learning_rate": 1.7470638237349195e-05, | |
| "loss": 1.7663, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 0.15040131119091807, | |
| "grad_norm": 0.85546875, | |
| "learning_rate": 1.7449675596623765e-05, | |
| "loss": 1.7934, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 0.15104405183703312, | |
| "grad_norm": 1.015625, | |
| "learning_rate": 1.742863913562218e-05, | |
| "loss": 1.7493, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.15168679248314815, | |
| "grad_norm": 1.109375, | |
| "learning_rate": 1.7407529062798784e-05, | |
| "loss": 1.7814, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 0.15232953312926317, | |
| "grad_norm": 1.0, | |
| "learning_rate": 1.7386345587337382e-05, | |
| "loss": 1.7127, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 0.15297227377537823, | |
| "grad_norm": 1.125, | |
| "learning_rate": 1.7365088919149124e-05, | |
| "loss": 1.7743, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 0.15361501442149325, | |
| "grad_norm": 1.1796875, | |
| "learning_rate": 1.734375926887045e-05, | |
| "loss": 1.7204, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 0.15425775506760828, | |
| "grad_norm": 2.25, | |
| "learning_rate": 1.7322356847860985e-05, | |
| "loss": 1.8416, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.15490049571372333, | |
| "grad_norm": 1.046875, | |
| "learning_rate": 1.7300881868201456e-05, | |
| "loss": 1.7365, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 0.15554323635983835, | |
| "grad_norm": 1.140625, | |
| "learning_rate": 1.7279334542691596e-05, | |
| "loss": 1.7371, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 0.15618597700595338, | |
| "grad_norm": 0.921875, | |
| "learning_rate": 1.725771508484802e-05, | |
| "loss": 1.7034, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 0.15682871765206843, | |
| "grad_norm": 1.0234375, | |
| "learning_rate": 1.7236023708902113e-05, | |
| "loss": 1.7502, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 0.15747145829818346, | |
| "grad_norm": 0.6640625, | |
| "learning_rate": 1.7214260629797913e-05, | |
| "loss": 1.681, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.15811419894429848, | |
| "grad_norm": 0.953125, | |
| "learning_rate": 1.7192426063189982e-05, | |
| "loss": 1.6597, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 0.1587569395904135, | |
| "grad_norm": 1.0234375, | |
| "learning_rate": 1.7170520225441264e-05, | |
| "loss": 1.7344, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 0.15939968023652856, | |
| "grad_norm": 0.765625, | |
| "learning_rate": 1.714854333362094e-05, | |
| "loss": 1.7025, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 0.1600424208826436, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.712649560550228e-05, | |
| "loss": 1.7985, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 0.1606851615287586, | |
| "grad_norm": 1.03125, | |
| "learning_rate": 1.710437725956049e-05, | |
| "loss": 1.7754, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.16132790217487367, | |
| "grad_norm": 1.15625, | |
| "learning_rate": 1.7082188514970534e-05, | |
| "loss": 1.7558, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 0.1619706428209887, | |
| "grad_norm": 1.140625, | |
| "learning_rate": 1.705992959160497e-05, | |
| "loss": 1.6337, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 0.16261338346710372, | |
| "grad_norm": 1.0703125, | |
| "learning_rate": 1.7037600710031783e-05, | |
| "loss": 1.7362, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 0.16325612411321877, | |
| "grad_norm": 0.796875, | |
| "learning_rate": 1.701520209151217e-05, | |
| "loss": 1.7269, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 0.1638988647593338, | |
| "grad_norm": 0.859375, | |
| "learning_rate": 1.699273395799838e-05, | |
| "loss": 1.8314, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 0.16454160540544882, | |
| "grad_norm": 0.8671875, | |
| "learning_rate": 1.697019653213149e-05, | |
| "loss": 1.7012, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 0.16518434605156387, | |
| "grad_norm": 0.984375, | |
| "learning_rate": 1.694759003723921e-05, | |
| "loss": 1.7954, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 0.1658270866976789, | |
| "grad_norm": 0.9296875, | |
| "learning_rate": 1.692491469733367e-05, | |
| "loss": 1.6907, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 0.16646982734379392, | |
| "grad_norm": 1.0546875, | |
| "learning_rate": 1.6902170737109203e-05, | |
| "loss": 1.6991, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 0.16711256798990898, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 1.6879358381940103e-05, | |
| "loss": 1.715, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.167755308636024, | |
| "grad_norm": 1.0078125, | |
| "learning_rate": 1.6856477857878415e-05, | |
| "loss": 1.7482, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 0.16839804928213903, | |
| "grad_norm": 1.046875, | |
| "learning_rate": 1.683352939165167e-05, | |
| "loss": 1.7095, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 0.16904078992825408, | |
| "grad_norm": 0.8515625, | |
| "learning_rate": 1.6810513210660665e-05, | |
| "loss": 1.6142, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 0.1696835305743691, | |
| "grad_norm": 1.1171875, | |
| "learning_rate": 1.678742954297718e-05, | |
| "loss": 1.7428, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 0.17032627122048413, | |
| "grad_norm": 1.0703125, | |
| "learning_rate": 1.676427861734175e-05, | |
| "loss": 1.7764, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 0.17096901186659919, | |
| "grad_norm": 1.2421875, | |
| "learning_rate": 1.674106066316137e-05, | |
| "loss": 1.7345, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 0.1716117525127142, | |
| "grad_norm": 1.0546875, | |
| "learning_rate": 1.671777591050724e-05, | |
| "loss": 1.7397, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 0.17225449315882924, | |
| "grad_norm": 1.109375, | |
| "learning_rate": 1.669442459011248e-05, | |
| "loss": 1.7448, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 0.1728972338049443, | |
| "grad_norm": 0.984375, | |
| "learning_rate": 1.667100693336983e-05, | |
| "loss": 1.747, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 0.17353997445105931, | |
| "grad_norm": 1.0546875, | |
| "learning_rate": 1.664752317232939e-05, | |
| "loss": 1.7379, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.17418271509717434, | |
| "grad_norm": 0.984375, | |
| "learning_rate": 1.6623973539696283e-05, | |
| "loss": 1.8181, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 0.1748254557432894, | |
| "grad_norm": 1.1015625, | |
| "learning_rate": 1.6600358268828376e-05, | |
| "loss": 1.6882, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 0.17546819638940442, | |
| "grad_norm": 0.96484375, | |
| "learning_rate": 1.6576677593733955e-05, | |
| "loss": 1.7557, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 0.17611093703551944, | |
| "grad_norm": 1.125, | |
| "learning_rate": 1.6552931749069403e-05, | |
| "loss": 1.7477, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 0.1767536776816345, | |
| "grad_norm": 0.77734375, | |
| "learning_rate": 1.6529120970136893e-05, | |
| "loss": 1.7788, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 0.17739641832774952, | |
| "grad_norm": 0.8046875, | |
| "learning_rate": 1.6505245492882044e-05, | |
| "loss": 1.7058, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 0.17803915897386455, | |
| "grad_norm": 1.1171875, | |
| "learning_rate": 1.6481305553891563e-05, | |
| "loss": 1.8183, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 0.1786818996199796, | |
| "grad_norm": 0.93359375, | |
| "learning_rate": 1.6457301390390945e-05, | |
| "loss": 1.7238, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 0.17932464026609463, | |
| "grad_norm": 0.99609375, | |
| "learning_rate": 1.6433233240242085e-05, | |
| "loss": 1.7097, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 0.17996738091220965, | |
| "grad_norm": 1.015625, | |
| "learning_rate": 1.6409101341940937e-05, | |
| "loss": 1.7503, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.1806101215583247, | |
| "grad_norm": 1.0859375, | |
| "learning_rate": 1.638490593461515e-05, | |
| "loss": 1.7222, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 0.18125286220443973, | |
| "grad_norm": 0.984375, | |
| "learning_rate": 1.6360647258021698e-05, | |
| "loss": 1.67, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 0.18189560285055476, | |
| "grad_norm": 1.078125, | |
| "learning_rate": 1.633632555254449e-05, | |
| "loss": 1.741, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 0.1825383434966698, | |
| "grad_norm": 0.74609375, | |
| "learning_rate": 1.6311941059192015e-05, | |
| "loss": 1.698, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 0.18318108414278483, | |
| "grad_norm": 1.0234375, | |
| "learning_rate": 1.6287494019594928e-05, | |
| "loss": 1.7521, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 0.18382382478889986, | |
| "grad_norm": 1.3203125, | |
| "learning_rate": 1.626298467600368e-05, | |
| "loss": 1.7319, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 0.1844665654350149, | |
| "grad_norm": 1.1640625, | |
| "learning_rate": 1.623841327128609e-05, | |
| "loss": 1.7694, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 0.18510930608112994, | |
| "grad_norm": 1.203125, | |
| "learning_rate": 1.6213780048924964e-05, | |
| "loss": 1.774, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 0.18575204672724496, | |
| "grad_norm": 0.8203125, | |
| "learning_rate": 1.6189085253015656e-05, | |
| "loss": 1.7477, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 0.18639478737336002, | |
| "grad_norm": 0.85546875, | |
| "learning_rate": 1.616432912826369e-05, | |
| "loss": 1.6605, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.18703752801947504, | |
| "grad_norm": 1.046875, | |
| "learning_rate": 1.613951191998228e-05, | |
| "loss": 1.7154, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 0.18768026866559007, | |
| "grad_norm": 0.9453125, | |
| "learning_rate": 1.6114633874089955e-05, | |
| "loss": 1.7665, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 0.18832300931170512, | |
| "grad_norm": 1.2578125, | |
| "learning_rate": 1.608969523710807e-05, | |
| "loss": 1.6787, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 0.18896574995782014, | |
| "grad_norm": 1.1484375, | |
| "learning_rate": 1.6064696256158408e-05, | |
| "loss": 1.6682, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 0.18960849060393517, | |
| "grad_norm": 0.99609375, | |
| "learning_rate": 1.6039637178960704e-05, | |
| "loss": 1.7495, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 0.19025123125005022, | |
| "grad_norm": 1.0546875, | |
| "learning_rate": 1.6014518253830204e-05, | |
| "loss": 1.6987, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 0.19089397189616525, | |
| "grad_norm": 1.1640625, | |
| "learning_rate": 1.5989339729675187e-05, | |
| "loss": 1.6799, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 0.19153671254228027, | |
| "grad_norm": 1.1640625, | |
| "learning_rate": 1.5964101855994527e-05, | |
| "loss": 1.7866, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 0.19217945318839533, | |
| "grad_norm": 1.171875, | |
| "learning_rate": 1.593880488287519e-05, | |
| "loss": 1.8115, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 0.19282219383451035, | |
| "grad_norm": 0.94921875, | |
| "learning_rate": 1.5913449060989777e-05, | |
| "loss": 1.7679, | |
| "step": 3000 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 10000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 500, | |
| "total_flos": 1.138460822905815e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |