| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 2825, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0017699115044247787, | |
| "grad_norm": 5.395828211145053, | |
| "learning_rate": 7.06713780918728e-08, | |
| "loss": 0.8769, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.008849557522123894, | |
| "grad_norm": 5.078223153245108, | |
| "learning_rate": 3.53356890459364e-07, | |
| "loss": 0.8957, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.017699115044247787, | |
| "grad_norm": 4.3462255239787675, | |
| "learning_rate": 7.06713780918728e-07, | |
| "loss": 0.8679, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.02654867256637168, | |
| "grad_norm": 2.7342805368137206, | |
| "learning_rate": 1.060070671378092e-06, | |
| "loss": 0.849, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.035398230088495575, | |
| "grad_norm": 2.245119548796511, | |
| "learning_rate": 1.413427561837456e-06, | |
| "loss": 0.8632, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.04424778761061947, | |
| "grad_norm": 1.9300825255667216, | |
| "learning_rate": 1.76678445229682e-06, | |
| "loss": 0.8287, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.05309734513274336, | |
| "grad_norm": 2.0548676134373025, | |
| "learning_rate": 2.120141342756184e-06, | |
| "loss": 0.8216, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.061946902654867256, | |
| "grad_norm": 2.107520072503226, | |
| "learning_rate": 2.473498233215548e-06, | |
| "loss": 0.8336, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.07079646017699115, | |
| "grad_norm": 2.012421705138888, | |
| "learning_rate": 2.826855123674912e-06, | |
| "loss": 0.8106, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.07964601769911504, | |
| "grad_norm": 2.0477120819165138, | |
| "learning_rate": 3.1802120141342757e-06, | |
| "loss": 0.8035, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.08849557522123894, | |
| "grad_norm": 1.94045311291632, | |
| "learning_rate": 3.53356890459364e-06, | |
| "loss": 0.8212, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.09734513274336283, | |
| "grad_norm": 1.9781807437344778, | |
| "learning_rate": 3.886925795053004e-06, | |
| "loss": 0.8167, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.10619469026548672, | |
| "grad_norm": 2.1588067674308955, | |
| "learning_rate": 4.240282685512368e-06, | |
| "loss": 0.8159, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.11504424778761062, | |
| "grad_norm": 2.0574710215669936, | |
| "learning_rate": 4.593639575971732e-06, | |
| "loss": 0.7809, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.12389380530973451, | |
| "grad_norm": 2.0440859301429692, | |
| "learning_rate": 4.946996466431096e-06, | |
| "loss": 0.7735, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.13274336283185842, | |
| "grad_norm": 2.1450366358677857, | |
| "learning_rate": 5.300353356890459e-06, | |
| "loss": 0.775, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.1415929203539823, | |
| "grad_norm": 2.0239595981615444, | |
| "learning_rate": 5.653710247349824e-06, | |
| "loss": 0.7739, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.1504424778761062, | |
| "grad_norm": 2.248989868043356, | |
| "learning_rate": 6.0070671378091885e-06, | |
| "loss": 0.7629, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.1592920353982301, | |
| "grad_norm": 2.113620356465933, | |
| "learning_rate": 6.360424028268551e-06, | |
| "loss": 0.7617, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.168141592920354, | |
| "grad_norm": 2.1063718938021987, | |
| "learning_rate": 6.713780918727916e-06, | |
| "loss": 0.7741, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.17699115044247787, | |
| "grad_norm": 2.0994149998808074, | |
| "learning_rate": 7.06713780918728e-06, | |
| "loss": 0.7657, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.18584070796460178, | |
| "grad_norm": 2.0741285650614785, | |
| "learning_rate": 7.420494699646644e-06, | |
| "loss": 0.7564, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.19469026548672566, | |
| "grad_norm": 1.8812346063525685, | |
| "learning_rate": 7.773851590106007e-06, | |
| "loss": 0.764, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.20353982300884957, | |
| "grad_norm": 1.6875596687855905, | |
| "learning_rate": 8.127208480565372e-06, | |
| "loss": 0.7395, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.21238938053097345, | |
| "grad_norm": 1.7946446429435303, | |
| "learning_rate": 8.480565371024736e-06, | |
| "loss": 0.7507, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.22123893805309736, | |
| "grad_norm": 1.5175436501935706, | |
| "learning_rate": 8.8339222614841e-06, | |
| "loss": 0.7197, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.23008849557522124, | |
| "grad_norm": 1.5846428467972697, | |
| "learning_rate": 9.187279151943464e-06, | |
| "loss": 0.7422, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.23893805309734514, | |
| "grad_norm": 1.9811600560411595, | |
| "learning_rate": 9.540636042402828e-06, | |
| "loss": 0.7305, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.24778761061946902, | |
| "grad_norm": 2.0603096951143822, | |
| "learning_rate": 9.893992932862191e-06, | |
| "loss": 0.7266, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.25663716814159293, | |
| "grad_norm": 1.79227194384631, | |
| "learning_rate": 1.0247349823321556e-05, | |
| "loss": 0.7273, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.26548672566371684, | |
| "grad_norm": 1.935925432369111, | |
| "learning_rate": 1.0600706713780919e-05, | |
| "loss": 0.7337, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.2743362831858407, | |
| "grad_norm": 1.6735245424795926, | |
| "learning_rate": 1.0954063604240283e-05, | |
| "loss": 0.726, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.2831858407079646, | |
| "grad_norm": 1.6786739654407825, | |
| "learning_rate": 1.1307420494699648e-05, | |
| "loss": 0.7442, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.2920353982300885, | |
| "grad_norm": 1.4647520145743216, | |
| "learning_rate": 1.1660777385159012e-05, | |
| "loss": 0.7104, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.3008849557522124, | |
| "grad_norm": 1.6975078748459098, | |
| "learning_rate": 1.2014134275618377e-05, | |
| "loss": 0.7202, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.30973451327433627, | |
| "grad_norm": 2.1939246563522925, | |
| "learning_rate": 1.2367491166077738e-05, | |
| "loss": 0.7135, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.3185840707964602, | |
| "grad_norm": 1.7970007273551845, | |
| "learning_rate": 1.2720848056537103e-05, | |
| "loss": 0.7404, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.3274336283185841, | |
| "grad_norm": 2.2094978373729, | |
| "learning_rate": 1.3074204946996467e-05, | |
| "loss": 0.7395, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.336283185840708, | |
| "grad_norm": 1.6148149490160606, | |
| "learning_rate": 1.3427561837455832e-05, | |
| "loss": 0.7236, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.34513274336283184, | |
| "grad_norm": 1.5716018292237066, | |
| "learning_rate": 1.3780918727915195e-05, | |
| "loss": 0.7176, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.35398230088495575, | |
| "grad_norm": 1.56394291086092, | |
| "learning_rate": 1.413427561837456e-05, | |
| "loss": 0.7077, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.36283185840707965, | |
| "grad_norm": 1.5944555462298586, | |
| "learning_rate": 1.4487632508833924e-05, | |
| "loss": 0.732, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.37168141592920356, | |
| "grad_norm": 3.0644131086145716, | |
| "learning_rate": 1.4840989399293289e-05, | |
| "loss": 0.7226, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.3805309734513274, | |
| "grad_norm": 48.89141497644312, | |
| "learning_rate": 1.519434628975265e-05, | |
| "loss": 0.9165, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.3893805309734513, | |
| "grad_norm": 63.82971563129597, | |
| "learning_rate": 1.5547703180212014e-05, | |
| "loss": 2.8435, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.39823008849557523, | |
| "grad_norm": 63.166528581176436, | |
| "learning_rate": 1.590106007067138e-05, | |
| "loss": 1.8489, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.40707964601769914, | |
| "grad_norm": 11.532232434183182, | |
| "learning_rate": 1.6254416961130744e-05, | |
| "loss": 1.1145, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.415929203539823, | |
| "grad_norm": 8.320428887121134, | |
| "learning_rate": 1.6607773851590106e-05, | |
| "loss": 1.0432, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.4247787610619469, | |
| "grad_norm": 5.49695556302703, | |
| "learning_rate": 1.6961130742049473e-05, | |
| "loss": 0.9249, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.4336283185840708, | |
| "grad_norm": 3.8086780643307954, | |
| "learning_rate": 1.7314487632508836e-05, | |
| "loss": 0.901, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.4424778761061947, | |
| "grad_norm": 3.0530888165680254, | |
| "learning_rate": 1.76678445229682e-05, | |
| "loss": 0.8318, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.45132743362831856, | |
| "grad_norm": 2.051374080224868, | |
| "learning_rate": 1.802120141342756e-05, | |
| "loss": 0.8123, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.46017699115044247, | |
| "grad_norm": 1.9860592805013697, | |
| "learning_rate": 1.8374558303886928e-05, | |
| "loss": 0.8145, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.4690265486725664, | |
| "grad_norm": 1.9585131132246008, | |
| "learning_rate": 1.872791519434629e-05, | |
| "loss": 0.8015, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.4778761061946903, | |
| "grad_norm": 2.0807728265740937, | |
| "learning_rate": 1.9081272084805657e-05, | |
| "loss": 0.7572, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.48672566371681414, | |
| "grad_norm": 8.726925488702674, | |
| "learning_rate": 1.943462897526502e-05, | |
| "loss": 0.7968, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.49557522123893805, | |
| "grad_norm": 3.0506762781221926, | |
| "learning_rate": 1.9787985865724383e-05, | |
| "loss": 0.7859, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.504424778761062, | |
| "grad_norm": 20.477250938697214, | |
| "learning_rate": 1.999996945230629e-05, | |
| "loss": 0.7726, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.5132743362831859, | |
| "grad_norm": 18.602610432760407, | |
| "learning_rate": 1.9999625792895357e-05, | |
| "loss": 0.7615, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.5221238938053098, | |
| "grad_norm": 1.9360445735897074, | |
| "learning_rate": 1.9998900302622567e-05, | |
| "loss": 0.7346, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.5309734513274337, | |
| "grad_norm": 1.9907733223798811, | |
| "learning_rate": 1.9997793009190403e-05, | |
| "loss": 0.7257, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.5398230088495575, | |
| "grad_norm": 2.649430391779402, | |
| "learning_rate": 1.999630395488034e-05, | |
| "loss": 0.7398, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.5486725663716814, | |
| "grad_norm": 2.295494291640341, | |
| "learning_rate": 1.9994433196551183e-05, | |
| "loss": 0.7404, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.5575221238938053, | |
| "grad_norm": 1.6045394414529988, | |
| "learning_rate": 1.9992180805636936e-05, | |
| "loss": 0.7157, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.5663716814159292, | |
| "grad_norm": 1.7545617453249185, | |
| "learning_rate": 1.998954686814406e-05, | |
| "loss": 0.722, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.5752212389380531, | |
| "grad_norm": 2.20743008915746, | |
| "learning_rate": 1.998653148464817e-05, | |
| "loss": 0.7133, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.584070796460177, | |
| "grad_norm": 1.8768620308499198, | |
| "learning_rate": 1.9983134770290232e-05, | |
| "loss": 0.7247, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.5929203539823009, | |
| "grad_norm": 2.4610134324640396, | |
| "learning_rate": 1.9979356854772128e-05, | |
| "loss": 0.6939, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.6017699115044248, | |
| "grad_norm": 1.5013573865061423, | |
| "learning_rate": 1.997519788235174e-05, | |
| "loss": 0.7184, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.6106194690265486, | |
| "grad_norm": 1.607524878532932, | |
| "learning_rate": 1.9970658011837404e-05, | |
| "loss": 0.7206, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.6194690265486725, | |
| "grad_norm": 1.57447888430762, | |
| "learning_rate": 1.996573741658188e-05, | |
| "loss": 0.7082, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.6283185840707964, | |
| "grad_norm": 1.5581830775430778, | |
| "learning_rate": 1.9960436284475712e-05, | |
| "loss": 0.6727, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.6371681415929203, | |
| "grad_norm": 1.5066025396716325, | |
| "learning_rate": 1.9954754817940054e-05, | |
| "loss": 0.708, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.6460176991150443, | |
| "grad_norm": 1.4817421328262548, | |
| "learning_rate": 1.994869323391895e-05, | |
| "loss": 0.6863, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.6548672566371682, | |
| "grad_norm": 1.4055734979132297, | |
| "learning_rate": 1.9942251763871056e-05, | |
| "loss": 0.7108, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.6637168141592921, | |
| "grad_norm": 1.391086495332054, | |
| "learning_rate": 1.9935430653760772e-05, | |
| "loss": 0.6608, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.672566371681416, | |
| "grad_norm": 1.5585496911243557, | |
| "learning_rate": 1.9928230164048885e-05, | |
| "loss": 0.6968, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.6814159292035398, | |
| "grad_norm": 1.5660850244776638, | |
| "learning_rate": 1.99206505696826e-05, | |
| "loss": 0.6892, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.6902654867256637, | |
| "grad_norm": 1.6030826428695462, | |
| "learning_rate": 1.9912692160085054e-05, | |
| "loss": 0.707, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.6991150442477876, | |
| "grad_norm": 1.4696166964471262, | |
| "learning_rate": 1.990435523914426e-05, | |
| "loss": 0.6773, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.7079646017699115, | |
| "grad_norm": 1.4674837525767894, | |
| "learning_rate": 1.9895640125201498e-05, | |
| "loss": 0.7007, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.7168141592920354, | |
| "grad_norm": 1.6349741965597524, | |
| "learning_rate": 1.988654715103917e-05, | |
| "loss": 0.6884, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.7256637168141593, | |
| "grad_norm": 1.6317810785917048, | |
| "learning_rate": 1.9877076663868084e-05, | |
| "loss": 0.6761, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.7345132743362832, | |
| "grad_norm": 1.7519745537508327, | |
| "learning_rate": 1.9867229025314204e-05, | |
| "loss": 0.6843, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.7433628318584071, | |
| "grad_norm": 1.707801199033859, | |
| "learning_rate": 1.9857004611404825e-05, | |
| "loss": 0.6735, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.7522123893805309, | |
| "grad_norm": 1.687330084394135, | |
| "learning_rate": 1.984640381255424e-05, | |
| "loss": 0.6661, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.7610619469026548, | |
| "grad_norm": 1.7043833380846305, | |
| "learning_rate": 1.9835427033548807e-05, | |
| "loss": 0.6794, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.7699115044247787, | |
| "grad_norm": 1.5675248939760222, | |
| "learning_rate": 1.982407469353152e-05, | |
| "loss": 0.6864, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.7787610619469026, | |
| "grad_norm": 1.4561869480484766, | |
| "learning_rate": 1.9812347225985966e-05, | |
| "loss": 0.657, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.7876106194690266, | |
| "grad_norm": 1.3651968406362616, | |
| "learning_rate": 1.9800245078719814e-05, | |
| "loss": 0.6725, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.7964601769911505, | |
| "grad_norm": 1.7068501163645904, | |
| "learning_rate": 1.9787768713847685e-05, | |
| "loss": 0.6907, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.8053097345132744, | |
| "grad_norm": 1.8186666681748165, | |
| "learning_rate": 1.9774918607773524e-05, | |
| "loss": 0.6666, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.8141592920353983, | |
| "grad_norm": 52.762622436160925, | |
| "learning_rate": 1.9761695251172398e-05, | |
| "loss": 0.6903, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.8230088495575221, | |
| "grad_norm": 2.785296160404902, | |
| "learning_rate": 1.9748099148971766e-05, | |
| "loss": 0.682, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.831858407079646, | |
| "grad_norm": 1.8158247940222614, | |
| "learning_rate": 1.97341308203322e-05, | |
| "loss": 0.6654, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.8407079646017699, | |
| "grad_norm": 3.2136675982919427, | |
| "learning_rate": 1.9719790798627555e-05, | |
| "loss": 0.6875, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.8495575221238938, | |
| "grad_norm": 1.9905381473001953, | |
| "learning_rate": 1.9705079631424605e-05, | |
| "loss": 0.6785, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.8584070796460177, | |
| "grad_norm": 2.788997906081037, | |
| "learning_rate": 1.9689997880462134e-05, | |
| "loss": 0.6614, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.8672566371681416, | |
| "grad_norm": 1.947320603121196, | |
| "learning_rate": 1.9674546121629495e-05, | |
| "loss": 0.6612, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.8761061946902655, | |
| "grad_norm": 16.64653233306791, | |
| "learning_rate": 1.9658724944944597e-05, | |
| "loss": 0.6755, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.8849557522123894, | |
| "grad_norm": 2.445435643330624, | |
| "learning_rate": 1.964253495453141e-05, | |
| "loss": 0.6489, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.8938053097345132, | |
| "grad_norm": 1.8384904388034906, | |
| "learning_rate": 1.9625976768596862e-05, | |
| "loss": 0.6832, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.9026548672566371, | |
| "grad_norm": 1.7969557638918616, | |
| "learning_rate": 1.9609051019407254e-05, | |
| "loss": 0.6624, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.911504424778761, | |
| "grad_norm": 1.7456317363852023, | |
| "learning_rate": 1.9591758353264106e-05, | |
| "loss": 0.6573, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.9203539823008849, | |
| "grad_norm": 1.4813878883475884, | |
| "learning_rate": 1.9574099430479498e-05, | |
| "loss": 0.659, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.9292035398230089, | |
| "grad_norm": 4.494324308598284, | |
| "learning_rate": 1.9556074925350826e-05, | |
| "loss": 0.6811, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.9380530973451328, | |
| "grad_norm": 7.919549938124173, | |
| "learning_rate": 1.9537685526135088e-05, | |
| "loss": 0.6812, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.9469026548672567, | |
| "grad_norm": 15.115841458620135, | |
| "learning_rate": 1.951893193502256e-05, | |
| "loss": 0.6774, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.9557522123893806, | |
| "grad_norm": 5.495425735869417, | |
| "learning_rate": 1.9499814868110035e-05, | |
| "loss": 0.6889, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.9646017699115044, | |
| "grad_norm": 12.34480108912751, | |
| "learning_rate": 1.9480335055373444e-05, | |
| "loss": 0.689, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.9734513274336283, | |
| "grad_norm": 15.091035487093771, | |
| "learning_rate": 1.9460493240639985e-05, | |
| "loss": 0.6907, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.9823008849557522, | |
| "grad_norm": 2.6410011813854815, | |
| "learning_rate": 1.9440290181559737e-05, | |
| "loss": 0.6728, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.9911504424778761, | |
| "grad_norm": 3.807048913327981, | |
| "learning_rate": 1.9419726649576707e-05, | |
| "loss": 0.6699, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 11.71551051866065, | |
| "learning_rate": 1.93988034298994e-05, | |
| "loss": 0.6823, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.6282660961151123, | |
| "eval_runtime": 346.0891, | |
| "eval_samples_per_second": 21.731, | |
| "eval_steps_per_second": 0.341, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 1.008849557522124, | |
| "grad_norm": 5.445947511608255, | |
| "learning_rate": 1.9377521321470806e-05, | |
| "loss": 0.5764, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.0176991150442478, | |
| "grad_norm": 5.441443896405208, | |
| "learning_rate": 1.935588113693792e-05, | |
| "loss": 0.5922, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 1.0265486725663717, | |
| "grad_norm": 4.764716514583245, | |
| "learning_rate": 1.9333883702620692e-05, | |
| "loss": 0.5688, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.0353982300884956, | |
| "grad_norm": 6.234212565235311, | |
| "learning_rate": 1.9311529858480488e-05, | |
| "loss": 0.5627, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 1.0442477876106195, | |
| "grad_norm": 3.275772180265634, | |
| "learning_rate": 1.9288820458088004e-05, | |
| "loss": 0.5405, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.0530973451327434, | |
| "grad_norm": 2.4494316038791752, | |
| "learning_rate": 1.926575636859068e-05, | |
| "loss": 0.5353, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 1.0619469026548674, | |
| "grad_norm": 1.964088050973173, | |
| "learning_rate": 1.924233847067959e-05, | |
| "loss": 0.5409, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.0707964601769913, | |
| "grad_norm": 1.7637636165900412, | |
| "learning_rate": 1.9218567658555813e-05, | |
| "loss": 0.5477, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 1.079646017699115, | |
| "grad_norm": 2.3214954304344433, | |
| "learning_rate": 1.919444483989628e-05, | |
| "loss": 0.5519, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.0884955752212389, | |
| "grad_norm": 2.0536428402479134, | |
| "learning_rate": 1.9169970935819123e-05, | |
| "loss": 0.5218, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 1.0973451327433628, | |
| "grad_norm": 1.670923852239301, | |
| "learning_rate": 1.9145146880848505e-05, | |
| "loss": 0.531, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.1061946902654867, | |
| "grad_norm": 1.9728407984682095, | |
| "learning_rate": 1.9119973622878928e-05, | |
| "loss": 0.5402, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 1.1150442477876106, | |
| "grad_norm": 8.142996728944224, | |
| "learning_rate": 1.9094452123139034e-05, | |
| "loss": 0.5656, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.1238938053097345, | |
| "grad_norm": 1.890805938276718, | |
| "learning_rate": 1.9068583356154917e-05, | |
| "loss": 0.539, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 1.1327433628318584, | |
| "grad_norm": 2.1078644912267968, | |
| "learning_rate": 1.9042368309712906e-05, | |
| "loss": 0.5461, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.1415929203539823, | |
| "grad_norm": 1.8971067802854567, | |
| "learning_rate": 1.9015807984821827e-05, | |
| "loss": 0.5494, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 1.1504424778761062, | |
| "grad_norm": 27.881164380475635, | |
| "learning_rate": 1.8988903395674814e-05, | |
| "loss": 0.535, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.1592920353982301, | |
| "grad_norm": 3.4980726529367576, | |
| "learning_rate": 1.8961655569610557e-05, | |
| "loss": 0.531, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 1.168141592920354, | |
| "grad_norm": 2.2255039588879066, | |
| "learning_rate": 1.8934065547074077e-05, | |
| "loss": 0.5369, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.176991150442478, | |
| "grad_norm": 1.7665125865767044, | |
| "learning_rate": 1.8906134381577008e-05, | |
| "loss": 0.5231, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 1.1858407079646018, | |
| "grad_norm": 1.7187220901255802, | |
| "learning_rate": 1.887786313965736e-05, | |
| "loss": 0.5205, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.1946902654867257, | |
| "grad_norm": 1.793384142575199, | |
| "learning_rate": 1.8849252900838795e-05, | |
| "loss": 0.5307, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 1.2035398230088497, | |
| "grad_norm": 1.789651406311183, | |
| "learning_rate": 1.8820304757589406e-05, | |
| "loss": 0.5259, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.2123893805309733, | |
| "grad_norm": 1.9237494570492129, | |
| "learning_rate": 1.8791019815280015e-05, | |
| "loss": 0.5262, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 1.2212389380530975, | |
| "grad_norm": 1.6656226502649325, | |
| "learning_rate": 1.8761399192141933e-05, | |
| "loss": 0.5681, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.2300884955752212, | |
| "grad_norm": 1.8325987162210478, | |
| "learning_rate": 1.8731444019224296e-05, | |
| "loss": 0.5373, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 1.238938053097345, | |
| "grad_norm": 1.772645490498923, | |
| "learning_rate": 1.8701155440350854e-05, | |
| "loss": 0.5274, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.247787610619469, | |
| "grad_norm": 1.7076814876614839, | |
| "learning_rate": 1.8670534612076304e-05, | |
| "loss": 0.5345, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 1.2566371681415929, | |
| "grad_norm": 2.8366257516212925, | |
| "learning_rate": 1.863958270364213e-05, | |
| "loss": 0.5448, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.2654867256637168, | |
| "grad_norm": 1.6438063307669566, | |
| "learning_rate": 1.8608300896931935e-05, | |
| "loss": 0.5345, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 1.2743362831858407, | |
| "grad_norm": 7.910992161903767, | |
| "learning_rate": 1.857669038642635e-05, | |
| "loss": 0.5771, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.2831858407079646, | |
| "grad_norm": 5.320203779573794, | |
| "learning_rate": 1.8544752379157383e-05, | |
| "loss": 0.5889, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 1.2920353982300885, | |
| "grad_norm": 4.200987040185538, | |
| "learning_rate": 1.851248809466236e-05, | |
| "loss": 0.5572, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.3008849557522124, | |
| "grad_norm": 3.6844001075774564, | |
| "learning_rate": 1.847989876493733e-05, | |
| "loss": 0.5729, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 1.3097345132743363, | |
| "grad_norm": 2.6959513168479003, | |
| "learning_rate": 1.8446985634390056e-05, | |
| "loss": 0.5438, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.3185840707964602, | |
| "grad_norm": 2.0110192321282074, | |
| "learning_rate": 1.841374995979246e-05, | |
| "loss": 0.5346, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 1.3274336283185841, | |
| "grad_norm": 2.4149221674637142, | |
| "learning_rate": 1.8380193010232664e-05, | |
| "loss": 0.5443, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.336283185840708, | |
| "grad_norm": 3.216072321253876, | |
| "learning_rate": 1.834631606706651e-05, | |
| "loss": 0.5388, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 1.3451327433628317, | |
| "grad_norm": 1.7562091306971943, | |
| "learning_rate": 1.831212042386865e-05, | |
| "loss": 0.5332, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.3539823008849559, | |
| "grad_norm": 1.7681274800133804, | |
| "learning_rate": 1.8277607386383134e-05, | |
| "loss": 0.5531, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 1.3628318584070795, | |
| "grad_norm": 1.6184444020633955, | |
| "learning_rate": 1.8242778272473566e-05, | |
| "loss": 0.5288, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.3716814159292037, | |
| "grad_norm": 1.8584096368775243, | |
| "learning_rate": 1.8207634412072765e-05, | |
| "loss": 0.5134, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 1.3805309734513274, | |
| "grad_norm": 2.3984599706556504, | |
| "learning_rate": 1.8172177147132e-05, | |
| "loss": 0.5293, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.3893805309734513, | |
| "grad_norm": 1.6184621514028006, | |
| "learning_rate": 1.8136407831569748e-05, | |
| "loss": 0.5332, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 1.3982300884955752, | |
| "grad_norm": 1.659497791050273, | |
| "learning_rate": 1.8100327831219968e-05, | |
| "loss": 0.5499, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.407079646017699, | |
| "grad_norm": 2.080085747337152, | |
| "learning_rate": 1.806393852377998e-05, | |
| "loss": 0.5373, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 1.415929203539823, | |
| "grad_norm": 1.7412321009147458, | |
| "learning_rate": 1.802724129875784e-05, | |
| "loss": 0.5237, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.424778761061947, | |
| "grad_norm": 1.5117367250487712, | |
| "learning_rate": 1.7990237557419298e-05, | |
| "loss": 0.5212, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 1.4336283185840708, | |
| "grad_norm": 1.5677828815765256, | |
| "learning_rate": 1.7952928712734266e-05, | |
| "loss": 0.5293, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.4424778761061947, | |
| "grad_norm": 1.6191863845989973, | |
| "learning_rate": 1.791531618932289e-05, | |
| "loss": 0.5108, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 1.4513274336283186, | |
| "grad_norm": 1.6356183497375685, | |
| "learning_rate": 1.7877401423401134e-05, | |
| "loss": 0.535, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.4601769911504425, | |
| "grad_norm": 2.120698903964094, | |
| "learning_rate": 1.7839185862725953e-05, | |
| "loss": 0.5276, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 1.4690265486725664, | |
| "grad_norm": 1.7162725245546222, | |
| "learning_rate": 1.7800670966539997e-05, | |
| "loss": 0.5157, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.4778761061946903, | |
| "grad_norm": 1.8342947704435748, | |
| "learning_rate": 1.7761858205515904e-05, | |
| "loss": 0.503, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 1.4867256637168142, | |
| "grad_norm": 1.6080172248542548, | |
| "learning_rate": 1.7722749061700122e-05, | |
| "loss": 0.5164, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.495575221238938, | |
| "grad_norm": 1.7310863566366472, | |
| "learning_rate": 1.7683345028456357e-05, | |
| "loss": 0.5144, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 1.504424778761062, | |
| "grad_norm": 1.5259860237803888, | |
| "learning_rate": 1.7643647610408507e-05, | |
| "loss": 0.5144, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.5132743362831858, | |
| "grad_norm": 1.949693136998924, | |
| "learning_rate": 1.760365832338322e-05, | |
| "loss": 0.5208, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 1.5221238938053099, | |
| "grad_norm": 1.4606753283332923, | |
| "learning_rate": 1.7563378694352038e-05, | |
| "loss": 0.514, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.5309734513274336, | |
| "grad_norm": 1.500515219738256, | |
| "learning_rate": 1.752281026137306e-05, | |
| "loss": 0.5105, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 1.5398230088495575, | |
| "grad_norm": 1.6809093610034818, | |
| "learning_rate": 1.7481954573532233e-05, | |
| "loss": 0.5246, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.5486725663716814, | |
| "grad_norm": 1.8505913851080076, | |
| "learning_rate": 1.7440813190884177e-05, | |
| "loss": 0.5263, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 1.5575221238938053, | |
| "grad_norm": 1.5042921112971175, | |
| "learning_rate": 1.7399387684392643e-05, | |
| "loss": 0.5078, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.5663716814159292, | |
| "grad_norm": 1.7603274810661258, | |
| "learning_rate": 1.7357679635870504e-05, | |
| "loss": 0.5152, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 1.575221238938053, | |
| "grad_norm": 1.9907861198097643, | |
| "learning_rate": 1.731569063791937e-05, | |
| "loss": 0.517, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.584070796460177, | |
| "grad_norm": 2.287444737461443, | |
| "learning_rate": 1.727342229386877e-05, | |
| "loss": 0.5118, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 1.592920353982301, | |
| "grad_norm": 1.7114791811335306, | |
| "learning_rate": 1.723087621771492e-05, | |
| "loss": 0.512, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.6017699115044248, | |
| "grad_norm": 1.6387744190074889, | |
| "learning_rate": 1.718805403405911e-05, | |
| "loss": 0.5151, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 1.6106194690265485, | |
| "grad_norm": 1.9126014391813266, | |
| "learning_rate": 1.7144957378045656e-05, | |
| "loss": 0.5072, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 1.6194690265486726, | |
| "grad_norm": 1.5534492075843847, | |
| "learning_rate": 1.7101587895299463e-05, | |
| "loss": 0.5139, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 1.6283185840707963, | |
| "grad_norm": 3.324022735746321, | |
| "learning_rate": 1.7057947241863207e-05, | |
| "loss": 0.486, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 1.6371681415929205, | |
| "grad_norm": 2.5161147424579413, | |
| "learning_rate": 1.7014037084134076e-05, | |
| "loss": 0.5127, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 1.6460176991150441, | |
| "grad_norm": 2.6183686325273814, | |
| "learning_rate": 1.696985909880015e-05, | |
| "loss": 0.5103, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 1.6548672566371683, | |
| "grad_norm": 1.8311730816272584, | |
| "learning_rate": 1.692541497277637e-05, | |
| "loss": 0.51, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 1.663716814159292, | |
| "grad_norm": 2.077861904241967, | |
| "learning_rate": 1.6880706403140146e-05, | |
| "loss": 0.5082, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 1.672566371681416, | |
| "grad_norm": 2.0643493814936282, | |
| "learning_rate": 1.6835735097066524e-05, | |
| "loss": 0.5199, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 1.6814159292035398, | |
| "grad_norm": 1.5582651497341313, | |
| "learning_rate": 1.6790502771763018e-05, | |
| "loss": 0.5014, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.6902654867256637, | |
| "grad_norm": 1.6655179770008597, | |
| "learning_rate": 1.6745011154404037e-05, | |
| "loss": 0.4854, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 1.6991150442477876, | |
| "grad_norm": 1.4915522872459333, | |
| "learning_rate": 1.669926198206493e-05, | |
| "loss": 0.5132, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 1.7079646017699115, | |
| "grad_norm": 1.6252968757056179, | |
| "learning_rate": 1.6653257001655652e-05, | |
| "loss": 0.5016, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 1.7168141592920354, | |
| "grad_norm": 2.4097499030817096, | |
| "learning_rate": 1.6606997969854087e-05, | |
| "loss": 0.5227, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 1.7256637168141593, | |
| "grad_norm": 1.631370053795557, | |
| "learning_rate": 1.6560486653038916e-05, | |
| "loss": 0.5119, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 1.7345132743362832, | |
| "grad_norm": 2.442759512892503, | |
| "learning_rate": 1.6513724827222225e-05, | |
| "loss": 0.4912, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 1.7433628318584071, | |
| "grad_norm": 12.222832820637523, | |
| "learning_rate": 1.6466714277981656e-05, | |
| "loss": 0.5224, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 1.752212389380531, | |
| "grad_norm": 1.699557253240237, | |
| "learning_rate": 1.641945680039223e-05, | |
| "loss": 0.52, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 1.7610619469026547, | |
| "grad_norm": 1.6614165390235756, | |
| "learning_rate": 1.6371954198957823e-05, | |
| "loss": 0.5118, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 1.7699115044247788, | |
| "grad_norm": 5.006976593399911, | |
| "learning_rate": 1.6324208287542228e-05, | |
| "loss": 0.4785, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.7787610619469025, | |
| "grad_norm": 2.501687669574672, | |
| "learning_rate": 1.6276220889299918e-05, | |
| "loss": 0.494, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 1.7876106194690267, | |
| "grad_norm": 1.593500017019138, | |
| "learning_rate": 1.622799383660643e-05, | |
| "loss": 0.5184, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 1.7964601769911503, | |
| "grad_norm": 1.8062411504712435, | |
| "learning_rate": 1.617952897098839e-05, | |
| "loss": 0.4905, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 1.8053097345132745, | |
| "grad_norm": 1.7405963191869815, | |
| "learning_rate": 1.6130828143053173e-05, | |
| "loss": 0.4826, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 1.8141592920353982, | |
| "grad_norm": 2.046715193192915, | |
| "learning_rate": 1.6081893212418292e-05, | |
| "loss": 0.4923, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 1.823008849557522, | |
| "grad_norm": 2.016184225830357, | |
| "learning_rate": 1.6032726047640336e-05, | |
| "loss": 0.5014, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 1.831858407079646, | |
| "grad_norm": 35.09896301727333, | |
| "learning_rate": 1.5983328526143653e-05, | |
| "loss": 0.4711, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 1.8407079646017699, | |
| "grad_norm": 2.493655284729721, | |
| "learning_rate": 1.5933702534148648e-05, | |
| "loss": 0.5138, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 1.8495575221238938, | |
| "grad_norm": 5.918180177995943, | |
| "learning_rate": 1.588384996659976e-05, | |
| "loss": 0.5389, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 1.8584070796460177, | |
| "grad_norm": 20.200037687607217, | |
| "learning_rate": 1.583377272709311e-05, | |
| "loss": 0.5038, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.8672566371681416, | |
| "grad_norm": 11.981522736747214, | |
| "learning_rate": 1.5783472727803796e-05, | |
| "loss": 0.5098, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 1.8761061946902655, | |
| "grad_norm": 2.493450240622104, | |
| "learning_rate": 1.5732951889412905e-05, | |
| "loss": 0.5068, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 1.8849557522123894, | |
| "grad_norm": 84.35037035076192, | |
| "learning_rate": 1.5682212141034153e-05, | |
| "loss": 0.5365, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 1.893805309734513, | |
| "grad_norm": 5.882151543248406, | |
| "learning_rate": 1.5631255420140225e-05, | |
| "loss": 0.5275, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 1.9026548672566372, | |
| "grad_norm": 4.244653752423409, | |
| "learning_rate": 1.55800836724888e-05, | |
| "loss": 0.5221, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 1.911504424778761, | |
| "grad_norm": 2.318077082885742, | |
| "learning_rate": 1.5528698852048247e-05, | |
| "loss": 0.5034, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 1.920353982300885, | |
| "grad_norm": 1.9732014714384287, | |
| "learning_rate": 1.547710292092301e-05, | |
| "loss": 0.522, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 1.9292035398230087, | |
| "grad_norm": 1.650645415727171, | |
| "learning_rate": 1.5425297849278714e-05, | |
| "loss": 0.511, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 1.9380530973451329, | |
| "grad_norm": 1.8379847429115603, | |
| "learning_rate": 1.5373285615266884e-05, | |
| "loss": 0.5234, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 1.9469026548672566, | |
| "grad_norm": 1.8823141679209345, | |
| "learning_rate": 1.5321068204949465e-05, | |
| "loss": 0.494, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.9557522123893807, | |
| "grad_norm": 1.7345679128452123, | |
| "learning_rate": 1.526864761222294e-05, | |
| "loss": 0.4964, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 1.9646017699115044, | |
| "grad_norm": 2.076946865200806, | |
| "learning_rate": 1.5216025838742226e-05, | |
| "loss": 0.487, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 1.9734513274336283, | |
| "grad_norm": 1.6316830069632373, | |
| "learning_rate": 1.5163204893844223e-05, | |
| "loss": 0.4799, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 1.9823008849557522, | |
| "grad_norm": 1.5027611303212294, | |
| "learning_rate": 1.5110186794471105e-05, | |
| "loss": 0.5015, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 1.991150442477876, | |
| "grad_norm": 1.4539381914768303, | |
| "learning_rate": 1.505697356509328e-05, | |
| "loss": 0.4817, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.6088305066129869, | |
| "learning_rate": 1.5003567237632113e-05, | |
| "loss": 0.4922, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.385861873626709, | |
| "eval_runtime": 342.0399, | |
| "eval_samples_per_second": 21.989, | |
| "eval_steps_per_second": 0.345, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 2.0088495575221237, | |
| "grad_norm": 2.7761717129790804, | |
| "learning_rate": 1.4949969851382315e-05, | |
| "loss": 0.3518, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 2.017699115044248, | |
| "grad_norm": 1.7619273167046978, | |
| "learning_rate": 1.4896183452934087e-05, | |
| "loss": 0.3277, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 2.0265486725663715, | |
| "grad_norm": 1.7864597766278332, | |
| "learning_rate": 1.4842210096094984e-05, | |
| "loss": 0.3257, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 2.0353982300884956, | |
| "grad_norm": 2.7271140490759724, | |
| "learning_rate": 1.478805184181145e-05, | |
| "loss": 0.3358, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 2.0442477876106193, | |
| "grad_norm": 1.6882000707315932, | |
| "learning_rate": 1.4733710758090175e-05, | |
| "loss": 0.3295, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 2.0530973451327434, | |
| "grad_norm": 1.6306905404465186, | |
| "learning_rate": 1.4679188919919076e-05, | |
| "loss": 0.3355, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 2.061946902654867, | |
| "grad_norm": 1.7785685190803522, | |
| "learning_rate": 1.4624488409188116e-05, | |
| "loss": 0.329, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 2.0707964601769913, | |
| "grad_norm": 1.7355960995540745, | |
| "learning_rate": 1.4569611314609767e-05, | |
| "loss": 0.3384, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 2.079646017699115, | |
| "grad_norm": 4.525568032764957, | |
| "learning_rate": 1.4514559731639273e-05, | |
| "loss": 0.3318, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 2.088495575221239, | |
| "grad_norm": 28.29868813913778, | |
| "learning_rate": 1.4459335762394637e-05, | |
| "loss": 0.3307, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 2.0973451327433628, | |
| "grad_norm": 3.34159347381774, | |
| "learning_rate": 1.4403941515576344e-05, | |
| "loss": 0.331, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 2.106194690265487, | |
| "grad_norm": 2.279005617470207, | |
| "learning_rate": 1.434837910638685e-05, | |
| "loss": 0.3506, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 2.1150442477876106, | |
| "grad_norm": 1.818174492390053, | |
| "learning_rate": 1.42926506564498e-05, | |
| "loss": 0.335, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 2.1238938053097347, | |
| "grad_norm": 2.21280984926424, | |
| "learning_rate": 1.4236758293729034e-05, | |
| "loss": 0.3383, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.1327433628318584, | |
| "grad_norm": 2.6931991751982154, | |
| "learning_rate": 1.4180704152447322e-05, | |
| "loss": 0.3431, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 2.1415929203539825, | |
| "grad_norm": 3.1061095288545504, | |
| "learning_rate": 1.4124490373004864e-05, | |
| "loss": 0.3485, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 2.150442477876106, | |
| "grad_norm": 1.8692302359437503, | |
| "learning_rate": 1.4068119101897568e-05, | |
| "loss": 0.3482, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 2.15929203539823, | |
| "grad_norm": 4.590488061615715, | |
| "learning_rate": 1.4011592491635088e-05, | |
| "loss": 0.3349, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 2.168141592920354, | |
| "grad_norm": 1.7406001344065656, | |
| "learning_rate": 1.3954912700658626e-05, | |
| "loss": 0.33, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 2.1769911504424777, | |
| "grad_norm": 2.147923934228603, | |
| "learning_rate": 1.389808189325851e-05, | |
| "loss": 0.3384, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 2.185840707964602, | |
| "grad_norm": 1.980447202723778, | |
| "learning_rate": 1.3841102239491567e-05, | |
| "loss": 0.3409, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 2.1946902654867255, | |
| "grad_norm": 1.754785548742135, | |
| "learning_rate": 1.3783975915098244e-05, | |
| "loss": 0.3267, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 2.2035398230088497, | |
| "grad_norm": 1.685034715782438, | |
| "learning_rate": 1.3726705101419538e-05, | |
| "loss": 0.3173, | |
| "step": 1245 | |
| }, | |
| { | |
| "epoch": 2.2123893805309733, | |
| "grad_norm": 1.697282116117874, | |
| "learning_rate": 1.3669291985313695e-05, | |
| "loss": 0.3422, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 2.2212389380530975, | |
| "grad_norm": 1.7593945507869428, | |
| "learning_rate": 1.3611738759072712e-05, | |
| "loss": 0.33, | |
| "step": 1255 | |
| }, | |
| { | |
| "epoch": 2.230088495575221, | |
| "grad_norm": 1.7917883090230355, | |
| "learning_rate": 1.3554047620338629e-05, | |
| "loss": 0.3305, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 2.2389380530973453, | |
| "grad_norm": 1.8630367840502837, | |
| "learning_rate": 1.3496220772019597e-05, | |
| "loss": 0.3331, | |
| "step": 1265 | |
| }, | |
| { | |
| "epoch": 2.247787610619469, | |
| "grad_norm": 1.4789663707909015, | |
| "learning_rate": 1.3438260422205779e-05, | |
| "loss": 0.3388, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 2.256637168141593, | |
| "grad_norm": 1.605944505641843, | |
| "learning_rate": 1.3380168784085028e-05, | |
| "loss": 0.3366, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 2.265486725663717, | |
| "grad_norm": 1.6963565599437993, | |
| "learning_rate": 1.3321948075858377e-05, | |
| "loss": 0.3563, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 2.274336283185841, | |
| "grad_norm": 1.6228304534956426, | |
| "learning_rate": 1.3263600520655333e-05, | |
| "loss": 0.3365, | |
| "step": 1285 | |
| }, | |
| { | |
| "epoch": 2.2831858407079646, | |
| "grad_norm": 1.76469997194976, | |
| "learning_rate": 1.3205128346449003e-05, | |
| "loss": 0.3443, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 2.2920353982300883, | |
| "grad_norm": 1.645969241835665, | |
| "learning_rate": 1.3146533785970997e-05, | |
| "loss": 0.3288, | |
| "step": 1295 | |
| }, | |
| { | |
| "epoch": 2.3008849557522124, | |
| "grad_norm": 1.6481237108795372, | |
| "learning_rate": 1.3087819076626201e-05, | |
| "loss": 0.3314, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.309734513274336, | |
| "grad_norm": 1.672673680905006, | |
| "learning_rate": 1.3028986460407312e-05, | |
| "loss": 0.3142, | |
| "step": 1305 | |
| }, | |
| { | |
| "epoch": 2.3185840707964602, | |
| "grad_norm": 1.6126633305212483, | |
| "learning_rate": 1.297003818380926e-05, | |
| "loss": 0.3331, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 2.327433628318584, | |
| "grad_norm": 1.6511744854665646, | |
| "learning_rate": 1.2910976497743389e-05, | |
| "loss": 0.321, | |
| "step": 1315 | |
| }, | |
| { | |
| "epoch": 2.336283185840708, | |
| "grad_norm": 1.6614647687728916, | |
| "learning_rate": 1.2851803657451554e-05, | |
| "loss": 0.34, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 2.3451327433628317, | |
| "grad_norm": 1.5556961404816416, | |
| "learning_rate": 1.2792521922419958e-05, | |
| "loss": 0.3378, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 2.353982300884956, | |
| "grad_norm": 1.607457047378207, | |
| "learning_rate": 1.2733133556292914e-05, | |
| "loss": 0.3277, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 2.3628318584070795, | |
| "grad_norm": 1.6195875099518586, | |
| "learning_rate": 1.2673640826786378e-05, | |
| "loss": 0.3268, | |
| "step": 1335 | |
| }, | |
| { | |
| "epoch": 2.3716814159292037, | |
| "grad_norm": 1.5352915106052365, | |
| "learning_rate": 1.2614046005601377e-05, | |
| "loss": 0.3186, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 2.3805309734513274, | |
| "grad_norm": 1.5797503971889433, | |
| "learning_rate": 1.2554351368337262e-05, | |
| "loss": 0.3344, | |
| "step": 1345 | |
| }, | |
| { | |
| "epoch": 2.3893805309734515, | |
| "grad_norm": 2.5943284467596253, | |
| "learning_rate": 1.2494559194404809e-05, | |
| "loss": 0.3468, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 2.398230088495575, | |
| "grad_norm": 1.7257039415600963, | |
| "learning_rate": 1.2434671766939184e-05, | |
| "loss": 0.3348, | |
| "step": 1355 | |
| }, | |
| { | |
| "epoch": 2.4070796460176993, | |
| "grad_norm": 1.892133636249101, | |
| "learning_rate": 1.2374691372712761e-05, | |
| "loss": 0.3276, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 2.415929203539823, | |
| "grad_norm": 1.959589712843592, | |
| "learning_rate": 1.2314620302047818e-05, | |
| "loss": 0.3273, | |
| "step": 1365 | |
| }, | |
| { | |
| "epoch": 2.4247787610619467, | |
| "grad_norm": 1.5186200162272043, | |
| "learning_rate": 1.2254460848729046e-05, | |
| "loss": 0.3274, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 2.433628318584071, | |
| "grad_norm": 1.5473486484026677, | |
| "learning_rate": 1.2194215309916005e-05, | |
| "loss": 0.3443, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 2.442477876106195, | |
| "grad_norm": 1.6097900368392104, | |
| "learning_rate": 1.2133885986055379e-05, | |
| "loss": 0.3179, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 2.4513274336283186, | |
| "grad_norm": 1.59719321690748, | |
| "learning_rate": 1.2073475180793144e-05, | |
| "loss": 0.324, | |
| "step": 1385 | |
| }, | |
| { | |
| "epoch": 2.4601769911504423, | |
| "grad_norm": 1.657560283923696, | |
| "learning_rate": 1.2012985200886602e-05, | |
| "loss": 0.3279, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 2.4690265486725664, | |
| "grad_norm": 1.6312886292352506, | |
| "learning_rate": 1.1952418356116309e-05, | |
| "loss": 0.342, | |
| "step": 1395 | |
| }, | |
| { | |
| "epoch": 2.47787610619469, | |
| "grad_norm": 1.673413598360019, | |
| "learning_rate": 1.1891776959197854e-05, | |
| "loss": 0.3325, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.4867256637168142, | |
| "grad_norm": 1.5641257680673357, | |
| "learning_rate": 1.1831063325693578e-05, | |
| "loss": 0.33, | |
| "step": 1405 | |
| }, | |
| { | |
| "epoch": 2.495575221238938, | |
| "grad_norm": 1.557582800520083, | |
| "learning_rate": 1.1770279773924133e-05, | |
| "loss": 0.3229, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 2.504424778761062, | |
| "grad_norm": 2.0076002123451033, | |
| "learning_rate": 1.1709428624879971e-05, | |
| "loss": 0.338, | |
| "step": 1415 | |
| }, | |
| { | |
| "epoch": 2.5132743362831858, | |
| "grad_norm": 1.7747271993836249, | |
| "learning_rate": 1.1648512202132705e-05, | |
| "loss": 0.3312, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 2.52212389380531, | |
| "grad_norm": 1.7345249327456702, | |
| "learning_rate": 1.15875328317464e-05, | |
| "loss": 0.3324, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 2.5309734513274336, | |
| "grad_norm": 1.5526595051928562, | |
| "learning_rate": 1.1526492842188746e-05, | |
| "loss": 0.3183, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 2.5398230088495577, | |
| "grad_norm": 1.7188069365366003, | |
| "learning_rate": 1.1465394564242142e-05, | |
| "loss": 0.3382, | |
| "step": 1435 | |
| }, | |
| { | |
| "epoch": 2.5486725663716814, | |
| "grad_norm": 1.5515396732118238, | |
| "learning_rate": 1.1404240330914706e-05, | |
| "loss": 0.3214, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 2.557522123893805, | |
| "grad_norm": 1.7016705926250035, | |
| "learning_rate": 1.1343032477351183e-05, | |
| "loss": 0.341, | |
| "step": 1445 | |
| }, | |
| { | |
| "epoch": 2.566371681415929, | |
| "grad_norm": 1.700802699390158, | |
| "learning_rate": 1.128177334074377e-05, | |
| "loss": 0.3206, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 2.5752212389380533, | |
| "grad_norm": 1.5960912094984059, | |
| "learning_rate": 1.122046526024291e-05, | |
| "loss": 0.3155, | |
| "step": 1455 | |
| }, | |
| { | |
| "epoch": 2.584070796460177, | |
| "grad_norm": 1.7339788975369348, | |
| "learning_rate": 1.1159110576867915e-05, | |
| "loss": 0.3239, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 2.5929203539823007, | |
| "grad_norm": 1.6059957404958596, | |
| "learning_rate": 1.1097711633417623e-05, | |
| "loss": 0.3221, | |
| "step": 1465 | |
| }, | |
| { | |
| "epoch": 2.601769911504425, | |
| "grad_norm": 1.8427719621309488, | |
| "learning_rate": 1.1036270774380906e-05, | |
| "loss": 0.3304, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 2.6106194690265485, | |
| "grad_norm": 1.631329628154416, | |
| "learning_rate": 1.0974790345847187e-05, | |
| "loss": 0.3202, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 2.6194690265486726, | |
| "grad_norm": 1.588519061268432, | |
| "learning_rate": 1.0913272695416807e-05, | |
| "loss": 0.3262, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 2.6283185840707963, | |
| "grad_norm": 1.6832971532585765, | |
| "learning_rate": 1.085172017211142e-05, | |
| "loss": 0.343, | |
| "step": 1485 | |
| }, | |
| { | |
| "epoch": 2.6371681415929205, | |
| "grad_norm": 1.6201470424053184, | |
| "learning_rate": 1.0790135126284275e-05, | |
| "loss": 0.3173, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 2.646017699115044, | |
| "grad_norm": 1.6068035539370134, | |
| "learning_rate": 1.072851990953049e-05, | |
| "loss": 0.3302, | |
| "step": 1495 | |
| }, | |
| { | |
| "epoch": 2.6548672566371683, | |
| "grad_norm": 1.6527582221176542, | |
| "learning_rate": 1.0666876874597235e-05, | |
| "loss": 0.317, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.663716814159292, | |
| "grad_norm": 1.6872940097146771, | |
| "learning_rate": 1.0605208375293905e-05, | |
| "loss": 0.3327, | |
| "step": 1505 | |
| }, | |
| { | |
| "epoch": 2.672566371681416, | |
| "grad_norm": 1.9993336890377256, | |
| "learning_rate": 1.0543516766402245e-05, | |
| "loss": 0.327, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 2.6814159292035398, | |
| "grad_norm": 1.8925758992460178, | |
| "learning_rate": 1.0481804403586421e-05, | |
| "loss": 0.3232, | |
| "step": 1515 | |
| }, | |
| { | |
| "epoch": 2.6902654867256635, | |
| "grad_norm": 1.7506417492910218, | |
| "learning_rate": 1.0420073643303085e-05, | |
| "loss": 0.3236, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 2.6991150442477876, | |
| "grad_norm": 1.7254760531768936, | |
| "learning_rate": 1.0358326842711383e-05, | |
| "loss": 0.3376, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 2.7079646017699117, | |
| "grad_norm": 1.6056971874044912, | |
| "learning_rate": 1.0296566359582951e-05, | |
| "loss": 0.3197, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 2.7168141592920354, | |
| "grad_norm": 1.692264180914051, | |
| "learning_rate": 1.023479455221189e-05, | |
| "loss": 0.317, | |
| "step": 1535 | |
| }, | |
| { | |
| "epoch": 2.725663716814159, | |
| "grad_norm": 1.755215116123095, | |
| "learning_rate": 1.0173013779324714e-05, | |
| "loss": 0.3309, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 2.734513274336283, | |
| "grad_norm": 1.6225703848243365, | |
| "learning_rate": 1.0111226399990267e-05, | |
| "loss": 0.3247, | |
| "step": 1545 | |
| }, | |
| { | |
| "epoch": 2.7433628318584073, | |
| "grad_norm": 1.574502935573898, | |
| "learning_rate": 1.0049434773529678e-05, | |
| "loss": 0.3193, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 2.752212389380531, | |
| "grad_norm": 1.642771315319423, | |
| "learning_rate": 9.98764125942623e-06, | |
| "loss": 0.3304, | |
| "step": 1555 | |
| }, | |
| { | |
| "epoch": 2.7610619469026547, | |
| "grad_norm": 1.5731646251204943, | |
| "learning_rate": 9.9258482172353e-06, | |
| "loss": 0.3438, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 2.769911504424779, | |
| "grad_norm": 1.8184099244247223, | |
| "learning_rate": 9.864058006494237e-06, | |
| "loss": 0.3278, | |
| "step": 1565 | |
| }, | |
| { | |
| "epoch": 2.7787610619469025, | |
| "grad_norm": 1.9834453880133907, | |
| "learning_rate": 9.80227298663227e-06, | |
| "loss": 0.3305, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 2.7876106194690267, | |
| "grad_norm": 1.9264046453250971, | |
| "learning_rate": 9.740495516880428e-06, | |
| "loss": 0.3158, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 2.7964601769911503, | |
| "grad_norm": 2.1903532218624857, | |
| "learning_rate": 9.678727956181438e-06, | |
| "loss": 0.3267, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 2.8053097345132745, | |
| "grad_norm": 1.8158272144124583, | |
| "learning_rate": 9.616972663099648e-06, | |
| "loss": 0.342, | |
| "step": 1585 | |
| }, | |
| { | |
| "epoch": 2.814159292035398, | |
| "grad_norm": 2.7692218225078724, | |
| "learning_rate": 9.55523199573098e-06, | |
| "loss": 0.3258, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 2.823008849557522, | |
| "grad_norm": 1.6127123082950938, | |
| "learning_rate": 9.493508311612874e-06, | |
| "loss": 0.3214, | |
| "step": 1595 | |
| }, | |
| { | |
| "epoch": 2.831858407079646, | |
| "grad_norm": 1.7692693875404215, | |
| "learning_rate": 9.431803967634284e-06, | |
| "loss": 0.337, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.84070796460177, | |
| "grad_norm": 1.5722273068604184, | |
| "learning_rate": 9.370121319945657e-06, | |
| "loss": 0.3354, | |
| "step": 1605 | |
| }, | |
| { | |
| "epoch": 2.849557522123894, | |
| "grad_norm": 1.5455898969077757, | |
| "learning_rate": 9.308462723868987e-06, | |
| "loss": 0.3203, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 2.8584070796460175, | |
| "grad_norm": 1.6331654733208454, | |
| "learning_rate": 9.246830533807857e-06, | |
| "loss": 0.3215, | |
| "step": 1615 | |
| }, | |
| { | |
| "epoch": 2.8672566371681416, | |
| "grad_norm": 1.6276033261489797, | |
| "learning_rate": 9.185227103157573e-06, | |
| "loss": 0.3152, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 2.8761061946902657, | |
| "grad_norm": 1.6614797915917234, | |
| "learning_rate": 9.12365478421525e-06, | |
| "loss": 0.3214, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 2.8849557522123894, | |
| "grad_norm": 1.732426911950826, | |
| "learning_rate": 9.062115928090036e-06, | |
| "loss": 0.3068, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 2.893805309734513, | |
| "grad_norm": 1.518118162492201, | |
| "learning_rate": 9.000612884613306e-06, | |
| "loss": 0.3126, | |
| "step": 1635 | |
| }, | |
| { | |
| "epoch": 2.9026548672566372, | |
| "grad_norm": 1.6702030494722016, | |
| "learning_rate": 8.939148002248954e-06, | |
| "loss": 0.3348, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 2.911504424778761, | |
| "grad_norm": 1.7443283818272917, | |
| "learning_rate": 8.877723628003703e-06, | |
| "loss": 0.3266, | |
| "step": 1645 | |
| }, | |
| { | |
| "epoch": 2.920353982300885, | |
| "grad_norm": 1.6887892210021154, | |
| "learning_rate": 8.816342107337501e-06, | |
| "loss": 0.331, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 2.9292035398230087, | |
| "grad_norm": 1.6444996183942215, | |
| "learning_rate": 8.755005784073948e-06, | |
| "loss": 0.3078, | |
| "step": 1655 | |
| }, | |
| { | |
| "epoch": 2.938053097345133, | |
| "grad_norm": 1.5829832005281133, | |
| "learning_rate": 8.693717000310801e-06, | |
| "loss": 0.3071, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 2.9469026548672566, | |
| "grad_norm": 1.7157739006910784, | |
| "learning_rate": 8.632478096330559e-06, | |
| "loss": 0.3255, | |
| "step": 1665 | |
| }, | |
| { | |
| "epoch": 2.9557522123893807, | |
| "grad_norm": 1.5977542612636495, | |
| "learning_rate": 8.571291410511063e-06, | |
| "loss": 0.3176, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 2.9646017699115044, | |
| "grad_norm": 1.6999676018113206, | |
| "learning_rate": 8.510159279236244e-06, | |
| "loss": 0.3275, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 2.9734513274336285, | |
| "grad_norm": 1.8724555229847881, | |
| "learning_rate": 8.449084036806893e-06, | |
| "loss": 0.3201, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 2.982300884955752, | |
| "grad_norm": 1.980909859882966, | |
| "learning_rate": 8.388068015351521e-06, | |
| "loss": 0.3105, | |
| "step": 1685 | |
| }, | |
| { | |
| "epoch": 2.991150442477876, | |
| "grad_norm": 1.797933699630095, | |
| "learning_rate": 8.327113544737325e-06, | |
| "loss": 0.3207, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 1.6572798098507155, | |
| "learning_rate": 8.2662229524812e-06, | |
| "loss": 0.3003, | |
| "step": 1695 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 0.235044464468956, | |
| "eval_runtime": 341.1312, | |
| "eval_samples_per_second": 22.047, | |
| "eval_steps_per_second": 0.346, | |
| "step": 1695 | |
| }, | |
| { | |
| "epoch": 3.0088495575221237, | |
| "grad_norm": 2.353752696189, | |
| "learning_rate": 8.205398563660886e-06, | |
| "loss": 0.179, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 3.017699115044248, | |
| "grad_norm": 2.3653722927763288, | |
| "learning_rate": 8.144642700826182e-06, | |
| "loss": 0.1704, | |
| "step": 1705 | |
| }, | |
| { | |
| "epoch": 3.0265486725663715, | |
| "grad_norm": 1.9259007577930276, | |
| "learning_rate": 8.08395768391024e-06, | |
| "loss": 0.1707, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 3.0353982300884956, | |
| "grad_norm": 1.893459900305888, | |
| "learning_rate": 8.02334583014101e-06, | |
| "loss": 0.1675, | |
| "step": 1715 | |
| }, | |
| { | |
| "epoch": 3.0442477876106193, | |
| "grad_norm": 2.0717584069847037, | |
| "learning_rate": 7.96280945395273e-06, | |
| "loss": 0.1839, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 3.0530973451327434, | |
| "grad_norm": 2.114700678563183, | |
| "learning_rate": 7.902350866897573e-06, | |
| "loss": 0.1793, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 3.061946902654867, | |
| "grad_norm": 1.8731510987373088, | |
| "learning_rate": 7.841972377557366e-06, | |
| "loss": 0.1846, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 3.0707964601769913, | |
| "grad_norm": 1.722257787492872, | |
| "learning_rate": 7.78167629145545e-06, | |
| "loss": 0.1697, | |
| "step": 1735 | |
| }, | |
| { | |
| "epoch": 3.079646017699115, | |
| "grad_norm": 1.705359298929866, | |
| "learning_rate": 7.721464910968628e-06, | |
| "loss": 0.1687, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 3.088495575221239, | |
| "grad_norm": 1.8344884220833366, | |
| "learning_rate": 7.661340535239266e-06, | |
| "loss": 0.1724, | |
| "step": 1745 | |
| }, | |
| { | |
| "epoch": 3.0973451327433628, | |
| "grad_norm": 1.7566442049461333, | |
| "learning_rate": 7.6013054600875005e-06, | |
| "loss": 0.1754, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 3.106194690265487, | |
| "grad_norm": 1.6129096564380034, | |
| "learning_rate": 7.541361977923564e-06, | |
| "loss": 0.1667, | |
| "step": 1755 | |
| }, | |
| { | |
| "epoch": 3.1150442477876106, | |
| "grad_norm": 1.7068252140891824, | |
| "learning_rate": 7.481512377660251e-06, | |
| "loss": 0.1667, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 3.1238938053097347, | |
| "grad_norm": 1.6531864536593857, | |
| "learning_rate": 7.421758944625528e-06, | |
| "loss": 0.1785, | |
| "step": 1765 | |
| }, | |
| { | |
| "epoch": 3.1327433628318584, | |
| "grad_norm": 1.821254758412697, | |
| "learning_rate": 7.362103960475258e-06, | |
| "loss": 0.1698, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 3.1415929203539825, | |
| "grad_norm": 1.7780006102159944, | |
| "learning_rate": 7.302549703106084e-06, | |
| "loss": 0.1828, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 3.150442477876106, | |
| "grad_norm": 2.39713947955002, | |
| "learning_rate": 7.243098446568442e-06, | |
| "loss": 0.1736, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 3.15929203539823, | |
| "grad_norm": 1.6290436678999984, | |
| "learning_rate": 7.183752460979737e-06, | |
| "loss": 0.1699, | |
| "step": 1785 | |
| }, | |
| { | |
| "epoch": 3.168141592920354, | |
| "grad_norm": 1.582870471036582, | |
| "learning_rate": 7.124514012437645e-06, | |
| "loss": 0.1718, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 3.1769911504424777, | |
| "grad_norm": 1.5585723392749353, | |
| "learning_rate": 7.065385362933603e-06, | |
| "loss": 0.166, | |
| "step": 1795 | |
| }, | |
| { | |
| "epoch": 3.185840707964602, | |
| "grad_norm": 1.6356629585103508, | |
| "learning_rate": 7.006368770266421e-06, | |
| "loss": 0.1738, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 3.1946902654867255, | |
| "grad_norm": 1.8092852534302832, | |
| "learning_rate": 6.947466487956067e-06, | |
| "loss": 0.184, | |
| "step": 1805 | |
| }, | |
| { | |
| "epoch": 3.2035398230088497, | |
| "grad_norm": 1.763331264313237, | |
| "learning_rate": 6.88868076515763e-06, | |
| "loss": 0.1747, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 3.2123893805309733, | |
| "grad_norm": 1.6618660989856646, | |
| "learning_rate": 6.83001384657543e-06, | |
| "loss": 0.1753, | |
| "step": 1815 | |
| }, | |
| { | |
| "epoch": 3.2212389380530975, | |
| "grad_norm": 1.587869179848222, | |
| "learning_rate": 6.7714679723772996e-06, | |
| "loss": 0.177, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 3.230088495575221, | |
| "grad_norm": 1.7012507664894645, | |
| "learning_rate": 6.713045378109058e-06, | |
| "loss": 0.182, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 3.2389380530973453, | |
| "grad_norm": 3.092871482624404, | |
| "learning_rate": 6.654748294609137e-06, | |
| "loss": 0.1749, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 3.247787610619469, | |
| "grad_norm": 5.246672357838822, | |
| "learning_rate": 6.596578947923395e-06, | |
| "loss": 0.1852, | |
| "step": 1835 | |
| }, | |
| { | |
| "epoch": 3.256637168141593, | |
| "grad_norm": 2.6149168080932492, | |
| "learning_rate": 6.538539559220141e-06, | |
| "loss": 0.1717, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 3.265486725663717, | |
| "grad_norm": 2.298194261929981, | |
| "learning_rate": 6.480632344705274e-06, | |
| "loss": 0.1827, | |
| "step": 1845 | |
| }, | |
| { | |
| "epoch": 3.274336283185841, | |
| "grad_norm": 2.0137609080864083, | |
| "learning_rate": 6.422859515537709e-06, | |
| "loss": 0.1783, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 3.2831858407079646, | |
| "grad_norm": 1.9067096587439358, | |
| "learning_rate": 6.365223277744907e-06, | |
| "loss": 0.1762, | |
| "step": 1855 | |
| }, | |
| { | |
| "epoch": 3.2920353982300883, | |
| "grad_norm": 1.749958280136199, | |
| "learning_rate": 6.3077258321386604e-06, | |
| "loss": 0.1666, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 3.3008849557522124, | |
| "grad_norm": 1.6753892133330817, | |
| "learning_rate": 6.25036937423105e-06, | |
| "loss": 0.1817, | |
| "step": 1865 | |
| }, | |
| { | |
| "epoch": 3.309734513274336, | |
| "grad_norm": 1.770861004923346, | |
| "learning_rate": 6.1931560941506055e-06, | |
| "loss": 0.1753, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 3.3185840707964602, | |
| "grad_norm": 1.6941989909031387, | |
| "learning_rate": 6.136088176558683e-06, | |
| "loss": 0.1683, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 3.327433628318584, | |
| "grad_norm": 1.7818744676278702, | |
| "learning_rate": 6.07916780056604e-06, | |
| "loss": 0.1819, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 3.336283185840708, | |
| "grad_norm": 1.7169221713713938, | |
| "learning_rate": 6.022397139649636e-06, | |
| "loss": 0.1753, | |
| "step": 1885 | |
| }, | |
| { | |
| "epoch": 3.3451327433628317, | |
| "grad_norm": 1.7383379208379126, | |
| "learning_rate": 5.96577836156963e-06, | |
| "loss": 0.1772, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 3.353982300884956, | |
| "grad_norm": 10.593484473821267, | |
| "learning_rate": 5.9093136282866014e-06, | |
| "loss": 0.1776, | |
| "step": 1895 | |
| }, | |
| { | |
| "epoch": 3.3628318584070795, | |
| "grad_norm": 7.892581375758647, | |
| "learning_rate": 5.853005095879015e-06, | |
| "loss": 0.177, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 3.3716814159292037, | |
| "grad_norm": 3.632404735066819, | |
| "learning_rate": 5.796854914460873e-06, | |
| "loss": 0.1819, | |
| "step": 1905 | |
| }, | |
| { | |
| "epoch": 3.3805309734513274, | |
| "grad_norm": 2.48312406396527, | |
| "learning_rate": 5.740865228099621e-06, | |
| "loss": 0.1765, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 3.3893805309734515, | |
| "grad_norm": 2.2660397170129993, | |
| "learning_rate": 5.68503817473429e-06, | |
| "loss": 0.1833, | |
| "step": 1915 | |
| }, | |
| { | |
| "epoch": 3.398230088495575, | |
| "grad_norm": 1.9157388795853385, | |
| "learning_rate": 5.629375886093835e-06, | |
| "loss": 0.1735, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 3.4070796460176993, | |
| "grad_norm": 1.9732798667035623, | |
| "learning_rate": 5.573880487615755e-06, | |
| "loss": 0.1776, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 3.415929203539823, | |
| "grad_norm": 1.694802471034398, | |
| "learning_rate": 5.518554098364932e-06, | |
| "loss": 0.1723, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 3.4247787610619467, | |
| "grad_norm": 1.607717285267641, | |
| "learning_rate": 5.463398830952714e-06, | |
| "loss": 0.1699, | |
| "step": 1935 | |
| }, | |
| { | |
| "epoch": 3.433628318584071, | |
| "grad_norm": 1.742187593317827, | |
| "learning_rate": 5.408416791456239e-06, | |
| "loss": 0.1829, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 3.442477876106195, | |
| "grad_norm": 1.6295575263043325, | |
| "learning_rate": 5.3536100793380234e-06, | |
| "loss": 0.168, | |
| "step": 1945 | |
| }, | |
| { | |
| "epoch": 3.4513274336283186, | |
| "grad_norm": 1.5116611834425375, | |
| "learning_rate": 5.298980787365785e-06, | |
| "loss": 0.1733, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 3.4601769911504423, | |
| "grad_norm": 1.7614935769749696, | |
| "learning_rate": 5.244531001532558e-06, | |
| "loss": 0.1639, | |
| "step": 1955 | |
| }, | |
| { | |
| "epoch": 3.4690265486725664, | |
| "grad_norm": 1.4882500944214951, | |
| "learning_rate": 5.190262800977007e-06, | |
| "loss": 0.1623, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 3.47787610619469, | |
| "grad_norm": 1.7919227523392, | |
| "learning_rate": 5.136178257904048e-06, | |
| "loss": 0.1793, | |
| "step": 1965 | |
| }, | |
| { | |
| "epoch": 3.4867256637168142, | |
| "grad_norm": 1.5720819445780152, | |
| "learning_rate": 5.082279437505739e-06, | |
| "loss": 0.1814, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 3.495575221238938, | |
| "grad_norm": 1.777272679047606, | |
| "learning_rate": 5.028568397882397e-06, | |
| "loss": 0.1732, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 3.504424778761062, | |
| "grad_norm": 1.661227110456758, | |
| "learning_rate": 4.975047189964027e-06, | |
| "loss": 0.1681, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 3.5132743362831858, | |
| "grad_norm": 1.5442922066155755, | |
| "learning_rate": 4.921717857431997e-06, | |
| "loss": 0.165, | |
| "step": 1985 | |
| }, | |
| { | |
| "epoch": 3.52212389380531, | |
| "grad_norm": 1.7475725868687078, | |
| "learning_rate": 4.868582436641006e-06, | |
| "loss": 0.1654, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 3.5309734513274336, | |
| "grad_norm": 1.6819734436503684, | |
| "learning_rate": 4.81564295654134e-06, | |
| "loss": 0.1689, | |
| "step": 1995 | |
| }, | |
| { | |
| "epoch": 3.5398230088495577, | |
| "grad_norm": 1.5921158221510299, | |
| "learning_rate": 4.762901438601368e-06, | |
| "loss": 0.1712, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.5486725663716814, | |
| "grad_norm": 1.7221909524561103, | |
| "learning_rate": 4.710359896730379e-06, | |
| "loss": 0.1761, | |
| "step": 2005 | |
| }, | |
| { | |
| "epoch": 3.557522123893805, | |
| "grad_norm": 1.6159991650918777, | |
| "learning_rate": 4.658020337201666e-06, | |
| "loss": 0.1779, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 3.566371681415929, | |
| "grad_norm": 1.6617020529449875, | |
| "learning_rate": 4.6058847585759335e-06, | |
| "loss": 0.1805, | |
| "step": 2015 | |
| }, | |
| { | |
| "epoch": 3.5752212389380533, | |
| "grad_norm": 1.6141307718438942, | |
| "learning_rate": 4.5539551516249735e-06, | |
| "loss": 0.181, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 3.584070796460177, | |
| "grad_norm": 1.613159926564658, | |
| "learning_rate": 4.502233499255641e-06, | |
| "loss": 0.1812, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 3.5929203539823007, | |
| "grad_norm": 1.5123746738195785, | |
| "learning_rate": 4.450721776434152e-06, | |
| "loss": 0.1737, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 3.601769911504425, | |
| "grad_norm": 1.59229310691719, | |
| "learning_rate": 4.399421950110657e-06, | |
| "loss": 0.156, | |
| "step": 2035 | |
| }, | |
| { | |
| "epoch": 3.6106194690265485, | |
| "grad_norm": 1.5956185168711787, | |
| "learning_rate": 4.348335979144158e-06, | |
| "loss": 0.1739, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 3.6194690265486726, | |
| "grad_norm": 1.5890271518636088, | |
| "learning_rate": 4.297465814227678e-06, | |
| "loss": 0.1682, | |
| "step": 2045 | |
| }, | |
| { | |
| "epoch": 3.6283185840707963, | |
| "grad_norm": 1.5754061195856448, | |
| "learning_rate": 4.2468133978137945e-06, | |
| "loss": 0.172, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 3.6371681415929205, | |
| "grad_norm": 1.5165865020871168, | |
| "learning_rate": 4.196380664040468e-06, | |
| "loss": 0.1514, | |
| "step": 2055 | |
| }, | |
| { | |
| "epoch": 3.646017699115044, | |
| "grad_norm": 1.47915973391418, | |
| "learning_rate": 4.146169538657185e-06, | |
| "loss": 0.1685, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 3.6548672566371683, | |
| "grad_norm": 1.6368988601716563, | |
| "learning_rate": 4.096181938951419e-06, | |
| "loss": 0.163, | |
| "step": 2065 | |
| }, | |
| { | |
| "epoch": 3.663716814159292, | |
| "grad_norm": 1.6225383306108707, | |
| "learning_rate": 4.046419773675421e-06, | |
| "loss": 0.1709, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 3.672566371681416, | |
| "grad_norm": 1.4634676872930676, | |
| "learning_rate": 3.9968849429733396e-06, | |
| "loss": 0.1617, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 3.6814159292035398, | |
| "grad_norm": 1.6544557711131658, | |
| "learning_rate": 3.94757933830867e-06, | |
| "loss": 0.1675, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 3.6902654867256635, | |
| "grad_norm": 1.6512592824076686, | |
| "learning_rate": 3.898504842392017e-06, | |
| "loss": 0.1722, | |
| "step": 2085 | |
| }, | |
| { | |
| "epoch": 3.6991150442477876, | |
| "grad_norm": 1.5886810130145228, | |
| "learning_rate": 3.849663329109206e-06, | |
| "loss": 0.1726, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 3.7079646017699117, | |
| "grad_norm": 1.631842537467322, | |
| "learning_rate": 3.801056663449737e-06, | |
| "loss": 0.1598, | |
| "step": 2095 | |
| }, | |
| { | |
| "epoch": 3.7168141592920354, | |
| "grad_norm": 1.5134893226587467, | |
| "learning_rate": 3.7526867014355685e-06, | |
| "loss": 0.1647, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 3.725663716814159, | |
| "grad_norm": 1.5640824884747317, | |
| "learning_rate": 3.70455529005025e-06, | |
| "loss": 0.1739, | |
| "step": 2105 | |
| }, | |
| { | |
| "epoch": 3.734513274336283, | |
| "grad_norm": 1.50638388767751, | |
| "learning_rate": 3.6566642671683806e-06, | |
| "loss": 0.1644, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 3.7433628318584073, | |
| "grad_norm": 1.6707015790458888, | |
| "learning_rate": 3.6090154614854432e-06, | |
| "loss": 0.1623, | |
| "step": 2115 | |
| }, | |
| { | |
| "epoch": 3.752212389380531, | |
| "grad_norm": 1.6269953969553939, | |
| "learning_rate": 3.561610692447982e-06, | |
| "loss": 0.1603, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 3.7610619469026547, | |
| "grad_norm": 1.6173678776254214, | |
| "learning_rate": 3.514451770184113e-06, | |
| "loss": 0.1751, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 3.769911504424779, | |
| "grad_norm": 2.901697583641324, | |
| "learning_rate": 3.467540495434415e-06, | |
| "loss": 0.1599, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 3.7787610619469025, | |
| "grad_norm": 1.6837282238012816, | |
| "learning_rate": 3.420878659483161e-06, | |
| "loss": 0.1636, | |
| "step": 2135 | |
| }, | |
| { | |
| "epoch": 3.7876106194690267, | |
| "grad_norm": 1.4175543694096393, | |
| "learning_rate": 3.374468044089937e-06, | |
| "loss": 0.1649, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 3.7964601769911503, | |
| "grad_norm": 1.7233110576016455, | |
| "learning_rate": 3.328310421421579e-06, | |
| "loss": 0.1641, | |
| "step": 2145 | |
| }, | |
| { | |
| "epoch": 3.8053097345132745, | |
| "grad_norm": 1.5605391560510777, | |
| "learning_rate": 3.2824075539845334e-06, | |
| "loss": 0.1705, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 3.814159292035398, | |
| "grad_norm": 1.5377565176020294, | |
| "learning_rate": 3.2367611945575308e-06, | |
| "loss": 0.1539, | |
| "step": 2155 | |
| }, | |
| { | |
| "epoch": 3.823008849557522, | |
| "grad_norm": 1.5868929794341793, | |
| "learning_rate": 3.191373086124666e-06, | |
| "loss": 0.1709, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 3.831858407079646, | |
| "grad_norm": 1.5622595706222757, | |
| "learning_rate": 3.1462449618088576e-06, | |
| "loss": 0.1559, | |
| "step": 2165 | |
| }, | |
| { | |
| "epoch": 3.84070796460177, | |
| "grad_norm": 1.7270166919786822, | |
| "learning_rate": 3.1013785448056454e-06, | |
| "loss": 0.1748, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 3.849557522123894, | |
| "grad_norm": 1.5748266845975978, | |
| "learning_rate": 3.0567755483174043e-06, | |
| "loss": 0.1726, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 3.8584070796460175, | |
| "grad_norm": 1.5844961174820655, | |
| "learning_rate": 3.0124376754879305e-06, | |
| "loss": 0.1586, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 3.8672566371681416, | |
| "grad_norm": 2.1519156556967776, | |
| "learning_rate": 2.968366619337394e-06, | |
| "loss": 0.1619, | |
| "step": 2185 | |
| }, | |
| { | |
| "epoch": 3.8761061946902657, | |
| "grad_norm": 1.5353376420507088, | |
| "learning_rate": 2.9245640626977012e-06, | |
| "loss": 0.1582, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 3.8849557522123894, | |
| "grad_norm": 1.604740201034412, | |
| "learning_rate": 2.881031678148244e-06, | |
| "loss": 0.1632, | |
| "step": 2195 | |
| }, | |
| { | |
| "epoch": 3.893805309734513, | |
| "grad_norm": 1.5866339698891518, | |
| "learning_rate": 2.837771127952007e-06, | |
| "loss": 0.1564, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 3.9026548672566372, | |
| "grad_norm": 2.295835726400093, | |
| "learning_rate": 2.7947840639921308e-06, | |
| "loss": 0.163, | |
| "step": 2205 | |
| }, | |
| { | |
| "epoch": 3.911504424778761, | |
| "grad_norm": 1.5757417684634734, | |
| "learning_rate": 2.7520721277088023e-06, | |
| "loss": 0.1596, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 3.920353982300885, | |
| "grad_norm": 1.8301488972667728, | |
| "learning_rate": 2.709636950036597e-06, | |
| "loss": 0.1609, | |
| "step": 2215 | |
| }, | |
| { | |
| "epoch": 3.9292035398230087, | |
| "grad_norm": 1.5193689628299787, | |
| "learning_rate": 2.6674801513421945e-06, | |
| "loss": 0.166, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 3.938053097345133, | |
| "grad_norm": 1.514470929679295, | |
| "learning_rate": 2.6256033413625136e-06, | |
| "loss": 0.1579, | |
| "step": 2225 | |
| }, | |
| { | |
| "epoch": 3.9469026548672566, | |
| "grad_norm": 1.5711723708738545, | |
| "learning_rate": 2.584008119143234e-06, | |
| "loss": 0.1636, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 3.9557522123893807, | |
| "grad_norm": 1.5856863391487659, | |
| "learning_rate": 2.5426960729777496e-06, | |
| "loss": 0.1656, | |
| "step": 2235 | |
| }, | |
| { | |
| "epoch": 3.9646017699115044, | |
| "grad_norm": 1.511859910396652, | |
| "learning_rate": 2.5016687803465033e-06, | |
| "loss": 0.1583, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 3.9734513274336285, | |
| "grad_norm": 1.5002228905494757, | |
| "learning_rate": 2.460927807856778e-06, | |
| "loss": 0.1602, | |
| "step": 2245 | |
| }, | |
| { | |
| "epoch": 3.982300884955752, | |
| "grad_norm": 1.4446988734661728, | |
| "learning_rate": 2.4204747111828463e-06, | |
| "loss": 0.1587, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 3.991150442477876, | |
| "grad_norm": 1.549269756232167, | |
| "learning_rate": 2.3803110350065884e-06, | |
| "loss": 0.1696, | |
| "step": 2255 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 1.641037232857028, | |
| "learning_rate": 2.3404383129585018e-06, | |
| "loss": 0.1776, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 0.1632937490940094, | |
| "eval_runtime": 341.9824, | |
| "eval_samples_per_second": 21.992, | |
| "eval_steps_per_second": 0.345, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 4.008849557522124, | |
| "grad_norm": 2.475219520944848, | |
| "learning_rate": 2.3008580675591462e-06, | |
| "loss": 0.0861, | |
| "step": 2265 | |
| }, | |
| { | |
| "epoch": 4.017699115044247, | |
| "grad_norm": 1.729540631610469, | |
| "learning_rate": 2.2615718101609986e-06, | |
| "loss": 0.0782, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 4.0265486725663715, | |
| "grad_norm": 1.467284022386594, | |
| "learning_rate": 2.222581040890741e-06, | |
| "loss": 0.0792, | |
| "step": 2275 | |
| }, | |
| { | |
| "epoch": 4.035398230088496, | |
| "grad_norm": 1.4386951051586159, | |
| "learning_rate": 2.183887248591996e-06, | |
| "loss": 0.0799, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 4.04424778761062, | |
| "grad_norm": 1.3926757930334368, | |
| "learning_rate": 2.1454919107684615e-06, | |
| "loss": 0.0824, | |
| "step": 2285 | |
| }, | |
| { | |
| "epoch": 4.053097345132743, | |
| "grad_norm": 1.3315563224559441, | |
| "learning_rate": 2.107396493527489e-06, | |
| "loss": 0.0765, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 4.061946902654867, | |
| "grad_norm": 1.36195332075, | |
| "learning_rate": 2.069602451524114e-06, | |
| "loss": 0.0752, | |
| "step": 2295 | |
| }, | |
| { | |
| "epoch": 4.070796460176991, | |
| "grad_norm": 1.350156993277736, | |
| "learning_rate": 2.0321112279055e-06, | |
| "loss": 0.0791, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 4.079646017699115, | |
| "grad_norm": 1.448441545074319, | |
| "learning_rate": 1.9949242542558466e-06, | |
| "loss": 0.0737, | |
| "step": 2305 | |
| }, | |
| { | |
| "epoch": 4.088495575221239, | |
| "grad_norm": 1.3217457960916599, | |
| "learning_rate": 1.9580429505417054e-06, | |
| "loss": 0.083, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 4.097345132743363, | |
| "grad_norm": 1.33752273930371, | |
| "learning_rate": 1.9214687250577766e-06, | |
| "loss": 0.0833, | |
| "step": 2315 | |
| }, | |
| { | |
| "epoch": 4.106194690265487, | |
| "grad_norm": 1.3522253645938511, | |
| "learning_rate": 1.8852029743731203e-06, | |
| "loss": 0.0778, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 4.115044247787611, | |
| "grad_norm": 1.3575033220346027, | |
| "learning_rate": 1.8492470832778442e-06, | |
| "loss": 0.0812, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 4.123893805309734, | |
| "grad_norm": 1.2823684094843621, | |
| "learning_rate": 1.8136024247302152e-06, | |
| "loss": 0.0763, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 4.132743362831858, | |
| "grad_norm": 1.3765176805961667, | |
| "learning_rate": 1.7782703598042327e-06, | |
| "loss": 0.0758, | |
| "step": 2335 | |
| }, | |
| { | |
| "epoch": 4.1415929203539825, | |
| "grad_norm": 1.2766052485378403, | |
| "learning_rate": 1.7432522376376637e-06, | |
| "loss": 0.0824, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 4.150442477876107, | |
| "grad_norm": 1.3056775652487742, | |
| "learning_rate": 1.7085493953805187e-06, | |
| "loss": 0.0788, | |
| "step": 2345 | |
| }, | |
| { | |
| "epoch": 4.15929203539823, | |
| "grad_norm": 1.2765927958394436, | |
| "learning_rate": 1.6741631581440066e-06, | |
| "loss": 0.0792, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 4.168141592920354, | |
| "grad_norm": 1.2224525666208792, | |
| "learning_rate": 1.6400948389499194e-06, | |
| "loss": 0.0767, | |
| "step": 2355 | |
| }, | |
| { | |
| "epoch": 4.176991150442478, | |
| "grad_norm": 1.3375837067028162, | |
| "learning_rate": 1.6063457386805004e-06, | |
| "loss": 0.0734, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 4.185840707964601, | |
| "grad_norm": 1.336866928045873, | |
| "learning_rate": 1.572917146028783e-06, | |
| "loss": 0.0812, | |
| "step": 2365 | |
| }, | |
| { | |
| "epoch": 4.1946902654867255, | |
| "grad_norm": 1.3309435734104655, | |
| "learning_rate": 1.539810337449369e-06, | |
| "loss": 0.079, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 4.20353982300885, | |
| "grad_norm": 1.32678570412564, | |
| "learning_rate": 1.507026577109686e-06, | |
| "loss": 0.0796, | |
| "step": 2375 | |
| }, | |
| { | |
| "epoch": 4.212389380530974, | |
| "grad_norm": 1.3156430424303722, | |
| "learning_rate": 1.4745671168417265e-06, | |
| "loss": 0.0777, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 4.221238938053097, | |
| "grad_norm": 1.2973545964944124, | |
| "learning_rate": 1.442433196094236e-06, | |
| "loss": 0.0827, | |
| "step": 2385 | |
| }, | |
| { | |
| "epoch": 4.230088495575221, | |
| "grad_norm": 1.311408962321446, | |
| "learning_rate": 1.4106260418854033e-06, | |
| "loss": 0.0775, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 4.238938053097345, | |
| "grad_norm": 1.3792578224833796, | |
| "learning_rate": 1.379146868755985e-06, | |
| "loss": 0.0804, | |
| "step": 2395 | |
| }, | |
| { | |
| "epoch": 4.247787610619469, | |
| "grad_norm": 1.2720416544972974, | |
| "learning_rate": 1.3479968787229402e-06, | |
| "loss": 0.0811, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 4.256637168141593, | |
| "grad_norm": 1.2865117421029972, | |
| "learning_rate": 1.3171772612335332e-06, | |
| "loss": 0.076, | |
| "step": 2405 | |
| }, | |
| { | |
| "epoch": 4.265486725663717, | |
| "grad_norm": 1.3809883790356827, | |
| "learning_rate": 1.2866891931199132e-06, | |
| "loss": 0.0797, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 4.274336283185841, | |
| "grad_norm": 1.339742153054439, | |
| "learning_rate": 1.2565338385541792e-06, | |
| "loss": 0.0773, | |
| "step": 2415 | |
| }, | |
| { | |
| "epoch": 4.283185840707965, | |
| "grad_norm": 1.369494397621055, | |
| "learning_rate": 1.2267123490039201e-06, | |
| "loss": 0.0803, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 4.292035398230088, | |
| "grad_norm": 1.2994483999857558, | |
| "learning_rate": 1.1972258631882527e-06, | |
| "loss": 0.076, | |
| "step": 2425 | |
| }, | |
| { | |
| "epoch": 4.300884955752212, | |
| "grad_norm": 1.3824019299588173, | |
| "learning_rate": 1.168075507034341e-06, | |
| "loss": 0.0779, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 4.3097345132743365, | |
| "grad_norm": 1.3842282775051018, | |
| "learning_rate": 1.1392623936343994e-06, | |
| "loss": 0.08, | |
| "step": 2435 | |
| }, | |
| { | |
| "epoch": 4.31858407079646, | |
| "grad_norm": 1.359855304821783, | |
| "learning_rate": 1.110787623203189e-06, | |
| "loss": 0.08, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 4.327433628318584, | |
| "grad_norm": 1.3510428064785265, | |
| "learning_rate": 1.0826522830360087e-06, | |
| "loss": 0.0814, | |
| "step": 2445 | |
| }, | |
| { | |
| "epoch": 4.336283185840708, | |
| "grad_norm": 1.3365703898748194, | |
| "learning_rate": 1.0548574474671835e-06, | |
| "loss": 0.0791, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 4.345132743362832, | |
| "grad_norm": 1.391541582481427, | |
| "learning_rate": 1.027404177829031e-06, | |
| "loss": 0.0827, | |
| "step": 2455 | |
| }, | |
| { | |
| "epoch": 4.353982300884955, | |
| "grad_norm": 1.324819487757073, | |
| "learning_rate": 1.0002935224113387e-06, | |
| "loss": 0.0796, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 4.3628318584070795, | |
| "grad_norm": 1.403599449784847, | |
| "learning_rate": 9.735265164213349e-07, | |
| "loss": 0.0806, | |
| "step": 2465 | |
| }, | |
| { | |
| "epoch": 4.371681415929204, | |
| "grad_norm": 1.254012872145482, | |
| "learning_rate": 9.471041819441673e-07, | |
| "loss": 0.0762, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 4.380530973451328, | |
| "grad_norm": 1.3727040572468259, | |
| "learning_rate": 9.210275279038638e-07, | |
| "loss": 0.0773, | |
| "step": 2475 | |
| }, | |
| { | |
| "epoch": 4.389380530973451, | |
| "grad_norm": 1.269212054809803, | |
| "learning_rate": 8.952975500248129e-07, | |
| "loss": 0.0789, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 4.398230088495575, | |
| "grad_norm": 1.2901290730207529, | |
| "learning_rate": 8.69915230793742e-07, | |
| "loss": 0.075, | |
| "step": 2485 | |
| }, | |
| { | |
| "epoch": 4.407079646017699, | |
| "grad_norm": 1.2318338508628874, | |
| "learning_rate": 8.448815394222043e-07, | |
| "loss": 0.0813, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 4.415929203539823, | |
| "grad_norm": 1.2925512164200812, | |
| "learning_rate": 8.20197431809564e-07, | |
| "loss": 0.0755, | |
| "step": 2495 | |
| }, | |
| { | |
| "epoch": 4.424778761061947, | |
| "grad_norm": 1.3632740857909407, | |
| "learning_rate": 7.958638505065031e-07, | |
| "loss": 0.077, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 4.433628318584071, | |
| "grad_norm": 1.2727197193668072, | |
| "learning_rate": 7.718817246790222e-07, | |
| "loss": 0.0756, | |
| "step": 2505 | |
| }, | |
| { | |
| "epoch": 4.442477876106195, | |
| "grad_norm": 1.283160067311808, | |
| "learning_rate": 7.48251970072964e-07, | |
| "loss": 0.0771, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 4.451327433628318, | |
| "grad_norm": 1.3266085812679953, | |
| "learning_rate": 7.249754889790539e-07, | |
| "loss": 0.0779, | |
| "step": 2515 | |
| }, | |
| { | |
| "epoch": 4.460176991150442, | |
| "grad_norm": 1.3553958194021836, | |
| "learning_rate": 7.020531701984334e-07, | |
| "loss": 0.0815, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 4.469026548672566, | |
| "grad_norm": 1.4498354656139365, | |
| "learning_rate": 6.794858890087275e-07, | |
| "loss": 0.0804, | |
| "step": 2525 | |
| }, | |
| { | |
| "epoch": 4.477876106194691, | |
| "grad_norm": 1.33445017208115, | |
| "learning_rate": 6.572745071306286e-07, | |
| "loss": 0.0825, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 4.486725663716814, | |
| "grad_norm": 1.2094739845924918, | |
| "learning_rate": 6.3541987269498e-07, | |
| "loss": 0.0724, | |
| "step": 2535 | |
| }, | |
| { | |
| "epoch": 4.495575221238938, | |
| "grad_norm": 1.2915737532727858, | |
| "learning_rate": 6.139228202104008e-07, | |
| "loss": 0.0745, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 4.504424778761062, | |
| "grad_norm": 1.3914894835662084, | |
| "learning_rate": 5.927841705314175e-07, | |
| "loss": 0.0796, | |
| "step": 2545 | |
| }, | |
| { | |
| "epoch": 4.513274336283186, | |
| "grad_norm": 1.292394593993158, | |
| "learning_rate": 5.720047308271149e-07, | |
| "loss": 0.078, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 4.522123893805309, | |
| "grad_norm": 1.3661686769110382, | |
| "learning_rate": 5.515852945503241e-07, | |
| "loss": 0.0811, | |
| "step": 2555 | |
| }, | |
| { | |
| "epoch": 4.530973451327434, | |
| "grad_norm": 1.2989196151190672, | |
| "learning_rate": 5.315266414073161e-07, | |
| "loss": 0.077, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 4.539823008849558, | |
| "grad_norm": 1.415979853747286, | |
| "learning_rate": 5.118295373280335e-07, | |
| "loss": 0.0812, | |
| "step": 2565 | |
| }, | |
| { | |
| "epoch": 4.548672566371682, | |
| "grad_norm": 1.3239654771249467, | |
| "learning_rate": 4.924947344368448e-07, | |
| "loss": 0.079, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 4.557522123893805, | |
| "grad_norm": 1.4440576212335767, | |
| "learning_rate": 4.7352297102382317e-07, | |
| "loss": 0.0747, | |
| "step": 2575 | |
| }, | |
| { | |
| "epoch": 4.566371681415929, | |
| "grad_norm": 1.3092000545668983, | |
| "learning_rate": 4.549149715165546e-07, | |
| "loss": 0.0754, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 4.575221238938053, | |
| "grad_norm": 1.2120991458734485, | |
| "learning_rate": 4.3667144645247463e-07, | |
| "loss": 0.0782, | |
| "step": 2585 | |
| }, | |
| { | |
| "epoch": 4.584070796460177, | |
| "grad_norm": 1.2333368517726013, | |
| "learning_rate": 4.187930924517436e-07, | |
| "loss": 0.076, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 4.592920353982301, | |
| "grad_norm": 1.2863679695017827, | |
| "learning_rate": 4.012805921906393e-07, | |
| "loss": 0.0751, | |
| "step": 2595 | |
| }, | |
| { | |
| "epoch": 4.601769911504425, | |
| "grad_norm": 1.3354897817488147, | |
| "learning_rate": 3.8413461437549203e-07, | |
| "loss": 0.0774, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 4.610619469026549, | |
| "grad_norm": 1.3165459456119635, | |
| "learning_rate": 3.673558137171496e-07, | |
| "loss": 0.0758, | |
| "step": 2605 | |
| }, | |
| { | |
| "epoch": 4.619469026548672, | |
| "grad_norm": 1.448724213864104, | |
| "learning_rate": 3.5094483090597706e-07, | |
| "loss": 0.0772, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 4.628318584070796, | |
| "grad_norm": 1.401376773315023, | |
| "learning_rate": 3.3490229258739794e-07, | |
| "loss": 0.0799, | |
| "step": 2615 | |
| }, | |
| { | |
| "epoch": 4.6371681415929205, | |
| "grad_norm": 1.3907713234287105, | |
| "learning_rate": 3.1922881133795827e-07, | |
| "loss": 0.0784, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 4.646017699115045, | |
| "grad_norm": 1.2597928306220052, | |
| "learning_rate": 3.0392498564193685e-07, | |
| "loss": 0.076, | |
| "step": 2625 | |
| }, | |
| { | |
| "epoch": 4.654867256637168, | |
| "grad_norm": 1.3365628143767583, | |
| "learning_rate": 2.889913998684979e-07, | |
| "loss": 0.0772, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 4.663716814159292, | |
| "grad_norm": 1.284544234891874, | |
| "learning_rate": 2.744286242493721e-07, | |
| "loss": 0.0782, | |
| "step": 2635 | |
| }, | |
| { | |
| "epoch": 4.672566371681416, | |
| "grad_norm": 1.2551721985478679, | |
| "learning_rate": 2.602372148570864e-07, | |
| "loss": 0.0811, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 4.68141592920354, | |
| "grad_norm": 1.2553558965086657, | |
| "learning_rate": 2.4641771358372537e-07, | |
| "loss": 0.0807, | |
| "step": 2645 | |
| }, | |
| { | |
| "epoch": 4.6902654867256635, | |
| "grad_norm": 1.2348755466860069, | |
| "learning_rate": 2.329706481202443e-07, | |
| "loss": 0.0801, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 4.699115044247788, | |
| "grad_norm": 1.1907724591359092, | |
| "learning_rate": 2.1989653193631667e-07, | |
| "loss": 0.0738, | |
| "step": 2655 | |
| }, | |
| { | |
| "epoch": 4.707964601769912, | |
| "grad_norm": 1.3510071207779715, | |
| "learning_rate": 2.0719586426072858e-07, | |
| "loss": 0.084, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 4.716814159292035, | |
| "grad_norm": 1.295370917721196, | |
| "learning_rate": 1.9486913006231846e-07, | |
| "loss": 0.0776, | |
| "step": 2665 | |
| }, | |
| { | |
| "epoch": 4.725663716814159, | |
| "grad_norm": 1.3413694479416487, | |
| "learning_rate": 1.8291680003145074e-07, | |
| "loss": 0.0797, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 4.734513274336283, | |
| "grad_norm": 1.4581808262991165, | |
| "learning_rate": 1.7133933056205366e-07, | |
| "loss": 0.0804, | |
| "step": 2675 | |
| }, | |
| { | |
| "epoch": 4.743362831858407, | |
| "grad_norm": 1.2741635529778859, | |
| "learning_rate": 1.601371637341864e-07, | |
| "loss": 0.0752, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 4.752212389380531, | |
| "grad_norm": 1.2782881439793918, | |
| "learning_rate": 1.49310727297155e-07, | |
| "loss": 0.0774, | |
| "step": 2685 | |
| }, | |
| { | |
| "epoch": 4.761061946902655, | |
| "grad_norm": 1.3120371018641195, | |
| "learning_rate": 1.3886043465318522e-07, | |
| "loss": 0.0813, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 4.769911504424779, | |
| "grad_norm": 1.1988340814149363, | |
| "learning_rate": 1.2878668484163303e-07, | |
| "loss": 0.0777, | |
| "step": 2695 | |
| }, | |
| { | |
| "epoch": 4.778761061946903, | |
| "grad_norm": 1.321759520377253, | |
| "learning_rate": 1.1908986252375243e-07, | |
| "loss": 0.078, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 4.787610619469026, | |
| "grad_norm": 1.4194270019959467, | |
| "learning_rate": 1.097703379679993e-07, | |
| "loss": 0.0775, | |
| "step": 2705 | |
| }, | |
| { | |
| "epoch": 4.79646017699115, | |
| "grad_norm": 1.2788598484535139, | |
| "learning_rate": 1.0082846703590055e-07, | |
| "loss": 0.0765, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 4.8053097345132745, | |
| "grad_norm": 1.3148204824038185, | |
| "learning_rate": 9.226459116846054e-08, | |
| "loss": 0.0751, | |
| "step": 2715 | |
| }, | |
| { | |
| "epoch": 4.814159292035399, | |
| "grad_norm": 1.344867899591603, | |
| "learning_rate": 8.407903737312929e-08, | |
| "loss": 0.0808, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 4.823008849557522, | |
| "grad_norm": 1.278318794310315, | |
| "learning_rate": 7.627211821130576e-08, | |
| "loss": 0.0756, | |
| "step": 2725 | |
| }, | |
| { | |
| "epoch": 4.831858407079646, | |
| "grad_norm": 1.2101543880815988, | |
| "learning_rate": 6.884413178641414e-08, | |
| "loss": 0.0728, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 4.84070796460177, | |
| "grad_norm": 1.4344243105198062, | |
| "learning_rate": 6.179536173251399e-08, | |
| "loss": 0.0844, | |
| "step": 2735 | |
| }, | |
| { | |
| "epoch": 4.849557522123893, | |
| "grad_norm": 1.3276414122745406, | |
| "learning_rate": 5.5126077203471186e-08, | |
| "loss": 0.0808, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 4.8584070796460175, | |
| "grad_norm": 1.3506069083893197, | |
| "learning_rate": 4.883653286268164e-08, | |
| "loss": 0.0833, | |
| "step": 2745 | |
| }, | |
| { | |
| "epoch": 4.867256637168142, | |
| "grad_norm": 1.3682014190493492, | |
| "learning_rate": 4.292696887334691e-08, | |
| "loss": 0.0771, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 4.876106194690266, | |
| "grad_norm": 1.2828761325152347, | |
| "learning_rate": 3.7397610889300384e-08, | |
| "loss": 0.0801, | |
| "step": 2755 | |
| }, | |
| { | |
| "epoch": 4.88495575221239, | |
| "grad_norm": 1.3118235861689838, | |
| "learning_rate": 3.224867004639642e-08, | |
| "loss": 0.0775, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 4.893805309734513, | |
| "grad_norm": 1.2599575243275205, | |
| "learning_rate": 2.7480342954444572e-08, | |
| "loss": 0.0771, | |
| "step": 2765 | |
| }, | |
| { | |
| "epoch": 4.902654867256637, | |
| "grad_norm": 1.2689219424528377, | |
| "learning_rate": 2.309281168970223e-08, | |
| "loss": 0.0745, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 4.911504424778761, | |
| "grad_norm": 1.3933156024565598, | |
| "learning_rate": 1.9086243787922453e-08, | |
| "loss": 0.0754, | |
| "step": 2775 | |
| }, | |
| { | |
| "epoch": 4.920353982300885, | |
| "grad_norm": 1.321697338465418, | |
| "learning_rate": 1.5460792237960154e-08, | |
| "loss": 0.0757, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 4.929203539823009, | |
| "grad_norm": 1.2224130906716606, | |
| "learning_rate": 1.2216595475921245e-08, | |
| "loss": 0.0804, | |
| "step": 2785 | |
| }, | |
| { | |
| "epoch": 4.938053097345133, | |
| "grad_norm": 1.3113541450877073, | |
| "learning_rate": 9.353777379889073e-09, | |
| "loss": 0.0774, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 4.946902654867257, | |
| "grad_norm": 1.3170038581923675, | |
| "learning_rate": 6.8724472651815474e-09, | |
| "loss": 0.0793, | |
| "step": 2795 | |
| }, | |
| { | |
| "epoch": 4.95575221238938, | |
| "grad_norm": 1.2713773209438923, | |
| "learning_rate": 4.772699880187804e-09, | |
| "loss": 0.0763, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 4.964601769911504, | |
| "grad_norm": 1.201702049473955, | |
| "learning_rate": 3.054615402743322e-09, | |
| "loss": 0.0733, | |
| "step": 2805 | |
| }, | |
| { | |
| "epoch": 4.9734513274336285, | |
| "grad_norm": 1.272942179409837, | |
| "learning_rate": 1.7182594370701577e-09, | |
| "loss": 0.0772, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 4.982300884955752, | |
| "grad_norm": 1.480916856542288, | |
| "learning_rate": 7.636830112733862e-10, | |
| "loss": 0.0793, | |
| "step": 2815 | |
| }, | |
| { | |
| "epoch": 4.991150442477876, | |
| "grad_norm": 1.328167260480697, | |
| "learning_rate": 1.9092257538932956e-10, | |
| "loss": 0.0787, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 1.2170466610329072, | |
| "learning_rate": 0.0, | |
| "loss": 0.0793, | |
| "step": 2825 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 0.1581079065799713, | |
| "eval_runtime": 341.788, | |
| "eval_samples_per_second": 22.005, | |
| "eval_steps_per_second": 0.345, | |
| "step": 2825 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 2825, | |
| "total_flos": 2365990109184000.0, | |
| "train_loss": 0.3740393664457102, | |
| "train_runtime": 65777.1761, | |
| "train_samples_per_second": 5.497, | |
| "train_steps_per_second": 0.043 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 2825, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 100, | |
| "total_flos": 2365990109184000.0, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |