| { | |
| "best_metric": 0.7777777777777778, | |
| "best_model_checkpoint": "CTMAE2_CS_V7_2/checkpoint-2600", | |
| "epoch": 49.016216216216215, | |
| "eval_steps": 500, | |
| "global_step": 12950, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0007722007722007722, | |
| "grad_norm": 2.6724629402160645, | |
| "learning_rate": 7.722007722007723e-08, | |
| "loss": 0.7116, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0015444015444015444, | |
| "grad_norm": 6.020544052124023, | |
| "learning_rate": 1.5444015444015445e-07, | |
| "loss": 0.6837, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0023166023166023165, | |
| "grad_norm": 6.438848972320557, | |
| "learning_rate": 2.3166023166023168e-07, | |
| "loss": 0.6963, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.003088803088803089, | |
| "grad_norm": 3.0015923976898193, | |
| "learning_rate": 3.088803088803089e-07, | |
| "loss": 0.6929, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.003861003861003861, | |
| "grad_norm": 3.25985050201416, | |
| "learning_rate": 3.8610038610038613e-07, | |
| "loss": 0.6872, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.004633204633204633, | |
| "grad_norm": 3.701803207397461, | |
| "learning_rate": 4.6332046332046336e-07, | |
| "loss": 0.6878, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.005405405405405406, | |
| "grad_norm": 3.4109742641448975, | |
| "learning_rate": 5.405405405405406e-07, | |
| "loss": 0.6827, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.006177606177606178, | |
| "grad_norm": 3.098447799682617, | |
| "learning_rate": 6.177606177606178e-07, | |
| "loss": 0.6798, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.0069498069498069494, | |
| "grad_norm": 5.643575668334961, | |
| "learning_rate": 6.94980694980695e-07, | |
| "loss": 0.6694, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.007722007722007722, | |
| "grad_norm": 5.473111152648926, | |
| "learning_rate": 7.722007722007723e-07, | |
| "loss": 0.658, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.008494208494208495, | |
| "grad_norm": 7.615536212921143, | |
| "learning_rate": 8.494208494208495e-07, | |
| "loss": 0.6749, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.009266409266409266, | |
| "grad_norm": 3.840909957885742, | |
| "learning_rate": 9.266409266409267e-07, | |
| "loss": 0.6897, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.010038610038610039, | |
| "grad_norm": 4.946235179901123, | |
| "learning_rate": 1.0038610038610038e-06, | |
| "loss": 0.6457, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.010810810810810811, | |
| "grad_norm": 4.3364386558532715, | |
| "learning_rate": 1.0810810810810812e-06, | |
| "loss": 0.5965, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.011583011583011582, | |
| "grad_norm": 5.371955394744873, | |
| "learning_rate": 1.1583011583011585e-06, | |
| "loss": 0.6401, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.012355212355212355, | |
| "grad_norm": 6.771674633026123, | |
| "learning_rate": 1.2355212355212356e-06, | |
| "loss": 0.6538, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.013127413127413128, | |
| "grad_norm": 8.420292854309082, | |
| "learning_rate": 1.3127413127413127e-06, | |
| "loss": 0.6751, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.013899613899613899, | |
| "grad_norm": 11.156353950500488, | |
| "learning_rate": 1.38996138996139e-06, | |
| "loss": 0.645, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.014671814671814672, | |
| "grad_norm": 6.941310405731201, | |
| "learning_rate": 1.4671814671814674e-06, | |
| "loss": 0.5783, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.015444015444015444, | |
| "grad_norm": 29.779354095458984, | |
| "learning_rate": 1.5444015444015445e-06, | |
| "loss": 0.5873, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.016216216216216217, | |
| "grad_norm": 36.21226501464844, | |
| "learning_rate": 1.6216216216216219e-06, | |
| "loss": 0.8262, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.01698841698841699, | |
| "grad_norm": 13.295347213745117, | |
| "learning_rate": 1.698841698841699e-06, | |
| "loss": 0.5922, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.01776061776061776, | |
| "grad_norm": 12.150893211364746, | |
| "learning_rate": 1.7760617760617763e-06, | |
| "loss": 0.5256, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.018532818532818532, | |
| "grad_norm": 18.2507381439209, | |
| "learning_rate": 1.8532818532818534e-06, | |
| "loss": 0.7013, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.019305019305019305, | |
| "grad_norm": 44.01077651977539, | |
| "learning_rate": 1.9305019305019305e-06, | |
| "loss": 0.6761, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.020077220077220077, | |
| "grad_norm": 8.128087997436523, | |
| "learning_rate": 2.0077220077220077e-06, | |
| "loss": 0.5325, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.020077220077220077, | |
| "eval_accuracy": 0.4666666666666667, | |
| "eval_loss": 0.8741418719291687, | |
| "eval_runtime": 14.8248, | |
| "eval_samples_per_second": 3.035, | |
| "eval_steps_per_second": 1.012, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.0007722007722009, | |
| "grad_norm": 8.959938049316406, | |
| "learning_rate": 2.084942084942085e-06, | |
| "loss": 0.6265, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.0015444015444015, | |
| "grad_norm": 14.598134994506836, | |
| "learning_rate": 2.1621621621621623e-06, | |
| "loss": 0.5794, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.0023166023166024, | |
| "grad_norm": 12.391850471496582, | |
| "learning_rate": 2.2393822393822394e-06, | |
| "loss": 0.9588, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.003088803088803, | |
| "grad_norm": 19.856307983398438, | |
| "learning_rate": 2.316602316602317e-06, | |
| "loss": 0.6347, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.0038610038610039, | |
| "grad_norm": 9.293577194213867, | |
| "learning_rate": 2.393822393822394e-06, | |
| "loss": 0.621, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.0046332046332047, | |
| "grad_norm": 18.37480354309082, | |
| "learning_rate": 2.4710424710424712e-06, | |
| "loss": 0.6988, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.0054054054054054, | |
| "grad_norm": 12.758193016052246, | |
| "learning_rate": 2.5482625482625484e-06, | |
| "loss": 0.6771, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.0061776061776062, | |
| "grad_norm": 11.605049133300781, | |
| "learning_rate": 2.6254826254826255e-06, | |
| "loss": 0.6083, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.0069498069498068, | |
| "grad_norm": 12.44294261932373, | |
| "learning_rate": 2.702702702702703e-06, | |
| "loss": 0.6117, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.0077220077220077, | |
| "grad_norm": 6.176090240478516, | |
| "learning_rate": 2.77992277992278e-06, | |
| "loss": 0.6549, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.0084942084942086, | |
| "grad_norm": 7.454275608062744, | |
| "learning_rate": 2.8571428571428573e-06, | |
| "loss": 0.6426, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.0092664092664092, | |
| "grad_norm": 8.811921119689941, | |
| "learning_rate": 2.934362934362935e-06, | |
| "loss": 0.5786, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.01003861003861, | |
| "grad_norm": 10.099629402160645, | |
| "learning_rate": 3.011583011583012e-06, | |
| "loss": 0.6792, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.0108108108108107, | |
| "grad_norm": 7.9834136962890625, | |
| "learning_rate": 3.088803088803089e-06, | |
| "loss": 0.6326, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.0115830115830116, | |
| "grad_norm": 22.69242286682129, | |
| "learning_rate": 3.166023166023166e-06, | |
| "loss": 0.6278, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.0123552123552124, | |
| "grad_norm": 26.15718650817871, | |
| "learning_rate": 3.2432432432432437e-06, | |
| "loss": 0.509, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.013127413127413, | |
| "grad_norm": 9.68459701538086, | |
| "learning_rate": 3.320463320463321e-06, | |
| "loss": 0.5917, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.013899613899614, | |
| "grad_norm": 11.585394859313965, | |
| "learning_rate": 3.397683397683398e-06, | |
| "loss": 0.5219, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.0146718146718148, | |
| "grad_norm": 25.82895851135254, | |
| "learning_rate": 3.4749034749034755e-06, | |
| "loss": 0.4272, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.0154440154440154, | |
| "grad_norm": 36.21267318725586, | |
| "learning_rate": 3.5521235521235526e-06, | |
| "loss": 0.7694, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.0162162162162163, | |
| "grad_norm": 9.973220825195312, | |
| "learning_rate": 3.6293436293436297e-06, | |
| "loss": 0.5405, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.016988416988417, | |
| "grad_norm": 16.958759307861328, | |
| "learning_rate": 3.706563706563707e-06, | |
| "loss": 0.5025, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.0177606177606178, | |
| "grad_norm": 8.369057655334473, | |
| "learning_rate": 3.7837837837837844e-06, | |
| "loss": 0.5491, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.0185328185328186, | |
| "grad_norm": 13.40450382232666, | |
| "learning_rate": 3.861003861003861e-06, | |
| "loss": 0.5724, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.0193050193050193, | |
| "grad_norm": 14.084481239318848, | |
| "learning_rate": 3.938223938223939e-06, | |
| "loss": 0.5336, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.0200772200772201, | |
| "grad_norm": 7.159802436828613, | |
| "learning_rate": 4.015444015444015e-06, | |
| "loss": 0.5015, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.0200772200772201, | |
| "eval_accuracy": 0.4444444444444444, | |
| "eval_loss": 0.9817712306976318, | |
| "eval_runtime": 14.2868, | |
| "eval_samples_per_second": 3.15, | |
| "eval_steps_per_second": 1.05, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 2.0007722007722006, | |
| "grad_norm": 11.340039253234863, | |
| "learning_rate": 4.092664092664093e-06, | |
| "loss": 0.6822, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 2.0015444015444017, | |
| "grad_norm": 22.847156524658203, | |
| "learning_rate": 4.16988416988417e-06, | |
| "loss": 0.5966, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 2.0023166023166024, | |
| "grad_norm": 16.122802734375, | |
| "learning_rate": 4.247104247104247e-06, | |
| "loss": 0.6605, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 2.003088803088803, | |
| "grad_norm": 15.669265747070312, | |
| "learning_rate": 4.324324324324325e-06, | |
| "loss": 0.6131, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 2.003861003861004, | |
| "grad_norm": 1.0841310024261475, | |
| "learning_rate": 4.401544401544402e-06, | |
| "loss": 0.3651, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 2.0046332046332047, | |
| "grad_norm": 13.917180061340332, | |
| "learning_rate": 4.478764478764479e-06, | |
| "loss": 1.0038, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 2.0054054054054054, | |
| "grad_norm": 8.559019088745117, | |
| "learning_rate": 4.5559845559845564e-06, | |
| "loss": 0.4637, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 2.006177606177606, | |
| "grad_norm": 14.26176643371582, | |
| "learning_rate": 4.633204633204634e-06, | |
| "loss": 0.4126, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.006949806949807, | |
| "grad_norm": 30.660659790039062, | |
| "learning_rate": 4.710424710424711e-06, | |
| "loss": 0.6691, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 2.0077220077220077, | |
| "grad_norm": 33.097259521484375, | |
| "learning_rate": 4.787644787644788e-06, | |
| "loss": 0.4737, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 2.0084942084942083, | |
| "grad_norm": 32.17298126220703, | |
| "learning_rate": 4.864864864864866e-06, | |
| "loss": 0.5081, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 2.0092664092664094, | |
| "grad_norm": 6.045112133026123, | |
| "learning_rate": 4.9420849420849425e-06, | |
| "loss": 0.7164, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 2.01003861003861, | |
| "grad_norm": 19.53457260131836, | |
| "learning_rate": 5.01930501930502e-06, | |
| "loss": 0.5659, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.0108108108108107, | |
| "grad_norm": 51.153682708740234, | |
| "learning_rate": 5.096525096525097e-06, | |
| "loss": 0.7365, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 2.011583011583012, | |
| "grad_norm": 9.247794151306152, | |
| "learning_rate": 5.173745173745173e-06, | |
| "loss": 0.5064, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 2.0123552123552124, | |
| "grad_norm": 15.04385757446289, | |
| "learning_rate": 5.250965250965251e-06, | |
| "loss": 0.5947, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 2.013127413127413, | |
| "grad_norm": 6.074428558349609, | |
| "learning_rate": 5.328185328185329e-06, | |
| "loss": 0.4695, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 2.0138996138996137, | |
| "grad_norm": 6.8704328536987305, | |
| "learning_rate": 5.405405405405406e-06, | |
| "loss": 0.8112, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.0146718146718148, | |
| "grad_norm": 24.314729690551758, | |
| "learning_rate": 5.4826254826254836e-06, | |
| "loss": 0.4951, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 2.0154440154440154, | |
| "grad_norm": 7.067862033843994, | |
| "learning_rate": 5.55984555984556e-06, | |
| "loss": 0.5792, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 2.016216216216216, | |
| "grad_norm": 32.30609893798828, | |
| "learning_rate": 5.637065637065637e-06, | |
| "loss": 0.969, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 2.016988416988417, | |
| "grad_norm": 10.73764705657959, | |
| "learning_rate": 5.7142857142857145e-06, | |
| "loss": 0.5096, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 2.0177606177606178, | |
| "grad_norm": 10.906487464904785, | |
| "learning_rate": 5.791505791505791e-06, | |
| "loss": 0.7254, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 2.0185328185328184, | |
| "grad_norm": 25.743921279907227, | |
| "learning_rate": 5.86872586872587e-06, | |
| "loss": 0.5756, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 2.0193050193050195, | |
| "grad_norm": 12.43097972869873, | |
| "learning_rate": 5.945945945945947e-06, | |
| "loss": 0.5903, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 2.02007722007722, | |
| "grad_norm": 45.06875228881836, | |
| "learning_rate": 6.023166023166024e-06, | |
| "loss": 0.5515, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 2.02007722007722, | |
| "eval_accuracy": 0.5111111111111111, | |
| "eval_loss": 0.7133760452270508, | |
| "eval_runtime": 12.995, | |
| "eval_samples_per_second": 3.463, | |
| "eval_steps_per_second": 1.154, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 3.0007722007722006, | |
| "grad_norm": 10.925649642944336, | |
| "learning_rate": 6.1003861003861005e-06, | |
| "loss": 0.7103, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 3.0015444015444017, | |
| "grad_norm": 9.422943115234375, | |
| "learning_rate": 6.177606177606178e-06, | |
| "loss": 0.3968, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 3.0023166023166024, | |
| "grad_norm": 24.057771682739258, | |
| "learning_rate": 6.254826254826255e-06, | |
| "loss": 0.4071, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 3.003088803088803, | |
| "grad_norm": 1.2915865182876587, | |
| "learning_rate": 6.332046332046332e-06, | |
| "loss": 0.7694, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 3.003861003861004, | |
| "grad_norm": 13.25047779083252, | |
| "learning_rate": 6.409266409266411e-06, | |
| "loss": 0.6309, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 3.0046332046332047, | |
| "grad_norm": 21.487367630004883, | |
| "learning_rate": 6.486486486486487e-06, | |
| "loss": 0.7567, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 3.0054054054054054, | |
| "grad_norm": 18.282329559326172, | |
| "learning_rate": 6.563706563706564e-06, | |
| "loss": 0.3432, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 3.006177606177606, | |
| "grad_norm": 19.872180938720703, | |
| "learning_rate": 6.640926640926642e-06, | |
| "loss": 0.6271, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 3.006949806949807, | |
| "grad_norm": 6.643412113189697, | |
| "learning_rate": 6.718146718146718e-06, | |
| "loss": 0.6144, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 3.0077220077220077, | |
| "grad_norm": 29.06749725341797, | |
| "learning_rate": 6.795366795366796e-06, | |
| "loss": 0.7109, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 3.0084942084942083, | |
| "grad_norm": 9.899932861328125, | |
| "learning_rate": 6.872586872586873e-06, | |
| "loss": 0.619, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 3.0092664092664094, | |
| "grad_norm": 7.574603080749512, | |
| "learning_rate": 6.949806949806951e-06, | |
| "loss": 0.5907, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 3.01003861003861, | |
| "grad_norm": 14.081249237060547, | |
| "learning_rate": 7.027027027027028e-06, | |
| "loss": 0.6675, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 3.0108108108108107, | |
| "grad_norm": 16.958438873291016, | |
| "learning_rate": 7.104247104247105e-06, | |
| "loss": 0.5028, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 3.011583011583012, | |
| "grad_norm": 57.52739715576172, | |
| "learning_rate": 7.181467181467182e-06, | |
| "loss": 0.4541, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 3.0123552123552124, | |
| "grad_norm": 9.852923393249512, | |
| "learning_rate": 7.2586872586872595e-06, | |
| "loss": 0.43, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 3.013127413127413, | |
| "grad_norm": 12.256267547607422, | |
| "learning_rate": 7.335907335907336e-06, | |
| "loss": 0.581, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 3.0138996138996137, | |
| "grad_norm": 7.483799457550049, | |
| "learning_rate": 7.413127413127414e-06, | |
| "loss": 0.667, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 3.0146718146718148, | |
| "grad_norm": 13.821316719055176, | |
| "learning_rate": 7.49034749034749e-06, | |
| "loss": 0.6437, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 3.0154440154440154, | |
| "grad_norm": 39.73973083496094, | |
| "learning_rate": 7.567567567567569e-06, | |
| "loss": 0.4757, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 3.016216216216216, | |
| "grad_norm": 7.891088008880615, | |
| "learning_rate": 7.644787644787645e-06, | |
| "loss": 0.5968, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 3.016988416988417, | |
| "grad_norm": 10.223976135253906, | |
| "learning_rate": 7.722007722007722e-06, | |
| "loss": 0.5044, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.0177606177606178, | |
| "grad_norm": 44.67567825317383, | |
| "learning_rate": 7.7992277992278e-06, | |
| "loss": 0.3877, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 3.0185328185328184, | |
| "grad_norm": 11.546664237976074, | |
| "learning_rate": 7.876447876447877e-06, | |
| "loss": 0.5934, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 3.0193050193050195, | |
| "grad_norm": 3.135157823562622, | |
| "learning_rate": 7.953667953667954e-06, | |
| "loss": 1.0837, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 3.02007722007722, | |
| "grad_norm": 53.9516716003418, | |
| "learning_rate": 8.03088803088803e-06, | |
| "loss": 0.3752, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 3.02007722007722, | |
| "eval_accuracy": 0.6222222222222222, | |
| "eval_loss": 0.7412399053573608, | |
| "eval_runtime": 13.1721, | |
| "eval_samples_per_second": 3.416, | |
| "eval_steps_per_second": 1.139, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 4.000772200772201, | |
| "grad_norm": 9.257226943969727, | |
| "learning_rate": 8.108108108108109e-06, | |
| "loss": 0.3859, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 4.001544401544401, | |
| "grad_norm": 20.57451057434082, | |
| "learning_rate": 8.185328185328186e-06, | |
| "loss": 0.6062, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 4.002316602316602, | |
| "grad_norm": 11.12028694152832, | |
| "learning_rate": 8.262548262548264e-06, | |
| "loss": 0.4765, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 4.003088803088803, | |
| "grad_norm": 73.99553680419922, | |
| "learning_rate": 8.33976833976834e-06, | |
| "loss": 0.681, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 4.003861003861004, | |
| "grad_norm": 2.5823240280151367, | |
| "learning_rate": 8.416988416988418e-06, | |
| "loss": 0.7646, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 4.004633204633205, | |
| "grad_norm": 25.63375473022461, | |
| "learning_rate": 8.494208494208494e-06, | |
| "loss": 0.2601, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 4.005405405405406, | |
| "grad_norm": 2.795551300048828, | |
| "learning_rate": 8.571428571428571e-06, | |
| "loss": 0.6665, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 4.006177606177606, | |
| "grad_norm": 5.9316205978393555, | |
| "learning_rate": 8.64864864864865e-06, | |
| "loss": 0.5213, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 4.006949806949807, | |
| "grad_norm": 18.198753356933594, | |
| "learning_rate": 8.725868725868728e-06, | |
| "loss": 0.798, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 4.007722007722008, | |
| "grad_norm": 62.04732131958008, | |
| "learning_rate": 8.803088803088804e-06, | |
| "loss": 0.4566, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 4.008494208494208, | |
| "grad_norm": 25.7177791595459, | |
| "learning_rate": 8.880308880308881e-06, | |
| "loss": 0.5294, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 4.009266409266409, | |
| "grad_norm": 13.888814926147461, | |
| "learning_rate": 8.957528957528958e-06, | |
| "loss": 0.4431, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 4.01003861003861, | |
| "grad_norm": 8.091224670410156, | |
| "learning_rate": 9.034749034749034e-06, | |
| "loss": 0.3993, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 4.010810810810811, | |
| "grad_norm": 1.0357552766799927, | |
| "learning_rate": 9.111969111969113e-06, | |
| "loss": 0.7643, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 4.011583011583012, | |
| "grad_norm": 59.91904830932617, | |
| "learning_rate": 9.189189189189191e-06, | |
| "loss": 0.8771, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 4.012355212355212, | |
| "grad_norm": 7.978598117828369, | |
| "learning_rate": 9.266409266409268e-06, | |
| "loss": 0.4757, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 4.013127413127413, | |
| "grad_norm": 35.72863006591797, | |
| "learning_rate": 9.343629343629345e-06, | |
| "loss": 0.5221, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 4.013899613899614, | |
| "grad_norm": 77.4381103515625, | |
| "learning_rate": 9.420849420849421e-06, | |
| "loss": 0.7294, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 4.014671814671814, | |
| "grad_norm": 43.887699127197266, | |
| "learning_rate": 9.498069498069498e-06, | |
| "loss": 0.7571, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 4.015444015444015, | |
| "grad_norm": 22.43756103515625, | |
| "learning_rate": 9.575289575289576e-06, | |
| "loss": 0.4187, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 4.0162162162162165, | |
| "grad_norm": 8.597156524658203, | |
| "learning_rate": 9.652509652509653e-06, | |
| "loss": 0.8185, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 4.016988416988417, | |
| "grad_norm": 28.265419006347656, | |
| "learning_rate": 9.729729729729732e-06, | |
| "loss": 0.3588, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 4.017760617760618, | |
| "grad_norm": 102.77274322509766, | |
| "learning_rate": 9.806949806949808e-06, | |
| "loss": 0.515, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 4.018532818532819, | |
| "grad_norm": 17.284040451049805, | |
| "learning_rate": 9.884169884169885e-06, | |
| "loss": 1.6301, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 4.019305019305019, | |
| "grad_norm": 39.904273986816406, | |
| "learning_rate": 9.961389961389962e-06, | |
| "loss": 0.5565, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 4.02007722007722, | |
| "grad_norm": 6.165562629699707, | |
| "learning_rate": 9.995709995709997e-06, | |
| "loss": 0.3128, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 4.02007722007722, | |
| "eval_accuracy": 0.6, | |
| "eval_loss": 0.8534404039382935, | |
| "eval_runtime": 13.8974, | |
| "eval_samples_per_second": 3.238, | |
| "eval_steps_per_second": 1.079, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 5.000772200772201, | |
| "grad_norm": 17.28194808959961, | |
| "learning_rate": 9.987129987129988e-06, | |
| "loss": 0.489, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 5.001544401544401, | |
| "grad_norm": 68.17495727539062, | |
| "learning_rate": 9.97854997854998e-06, | |
| "loss": 0.5224, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 5.002316602316602, | |
| "grad_norm": 104.81306457519531, | |
| "learning_rate": 9.96996996996997e-06, | |
| "loss": 0.3856, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 5.003088803088803, | |
| "grad_norm": 33.45752716064453, | |
| "learning_rate": 9.961389961389962e-06, | |
| "loss": 0.347, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 5.003861003861004, | |
| "grad_norm": 20.993886947631836, | |
| "learning_rate": 9.952809952809953e-06, | |
| "loss": 0.3714, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 5.004633204633205, | |
| "grad_norm": 15.487417221069336, | |
| "learning_rate": 9.944229944229946e-06, | |
| "loss": 0.9065, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 5.005405405405406, | |
| "grad_norm": 4.61458158493042, | |
| "learning_rate": 9.935649935649937e-06, | |
| "loss": 1.0333, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 5.006177606177606, | |
| "grad_norm": 8.850576400756836, | |
| "learning_rate": 9.927069927069928e-06, | |
| "loss": 0.4281, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 5.006949806949807, | |
| "grad_norm": 23.16834831237793, | |
| "learning_rate": 9.91848991848992e-06, | |
| "loss": 0.8578, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 5.007722007722008, | |
| "grad_norm": 7.291067600250244, | |
| "learning_rate": 9.90990990990991e-06, | |
| "loss": 0.3951, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 5.008494208494208, | |
| "grad_norm": 15.696223258972168, | |
| "learning_rate": 9.901329901329903e-06, | |
| "loss": 0.4542, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 5.009266409266409, | |
| "grad_norm": 43.79963302612305, | |
| "learning_rate": 9.892749892749894e-06, | |
| "loss": 0.8643, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 5.01003861003861, | |
| "grad_norm": 8.4163179397583, | |
| "learning_rate": 9.884169884169885e-06, | |
| "loss": 0.9005, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 5.010810810810811, | |
| "grad_norm": 13.300799369812012, | |
| "learning_rate": 9.875589875589876e-06, | |
| "loss": 0.5591, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 5.011583011583012, | |
| "grad_norm": 27.228687286376953, | |
| "learning_rate": 9.867009867009867e-06, | |
| "loss": 0.4523, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 5.012355212355212, | |
| "grad_norm": 3.9106037616729736, | |
| "learning_rate": 9.858429858429858e-06, | |
| "loss": 0.3319, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 5.013127413127413, | |
| "grad_norm": 86.4173583984375, | |
| "learning_rate": 9.849849849849851e-06, | |
| "loss": 0.3241, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 5.013899613899614, | |
| "grad_norm": 25.517906188964844, | |
| "learning_rate": 9.841269841269842e-06, | |
| "loss": 0.7146, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 5.014671814671814, | |
| "grad_norm": 9.686749458312988, | |
| "learning_rate": 9.832689832689833e-06, | |
| "loss": 0.7438, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 5.015444015444015, | |
| "grad_norm": 0.10342951864004135, | |
| "learning_rate": 9.824109824109826e-06, | |
| "loss": 0.4138, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 5.0162162162162165, | |
| "grad_norm": 0.23315098881721497, | |
| "learning_rate": 9.815529815529815e-06, | |
| "loss": 1.2718, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 5.016988416988417, | |
| "grad_norm": 15.688437461853027, | |
| "learning_rate": 9.806949806949808e-06, | |
| "loss": 0.5534, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 5.017760617760618, | |
| "grad_norm": 14.076128005981445, | |
| "learning_rate": 9.7983697983698e-06, | |
| "loss": 0.4031, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 5.018532818532819, | |
| "grad_norm": 13.882486343383789, | |
| "learning_rate": 9.78978978978979e-06, | |
| "loss": 0.3393, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 5.019305019305019, | |
| "grad_norm": 3.6558635234832764, | |
| "learning_rate": 9.781209781209782e-06, | |
| "loss": 0.5401, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 5.02007722007722, | |
| "grad_norm": 1.8778923749923706, | |
| "learning_rate": 9.772629772629774e-06, | |
| "loss": 0.7261, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 5.02007722007722, | |
| "eval_accuracy": 0.7111111111111111, | |
| "eval_loss": 0.700229287147522, | |
| "eval_runtime": 13.0303, | |
| "eval_samples_per_second": 3.453, | |
| "eval_steps_per_second": 1.151, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 6.000772200772201, | |
| "grad_norm": 0.08510206639766693, | |
| "learning_rate": 9.764049764049764e-06, | |
| "loss": 0.4059, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 6.001544401544401, | |
| "grad_norm": 0.5811074376106262, | |
| "learning_rate": 9.755469755469757e-06, | |
| "loss": 0.4122, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 6.002316602316602, | |
| "grad_norm": 0.5630064010620117, | |
| "learning_rate": 9.746889746889748e-06, | |
| "loss": 0.3503, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 6.003088803088803, | |
| "grad_norm": 47.41246032714844, | |
| "learning_rate": 9.738309738309739e-06, | |
| "loss": 0.8755, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 6.003861003861004, | |
| "grad_norm": 34.18187713623047, | |
| "learning_rate": 9.729729729729732e-06, | |
| "loss": 0.5853, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 6.004633204633205, | |
| "grad_norm": 37.81892776489258, | |
| "learning_rate": 9.721149721149723e-06, | |
| "loss": 0.7055, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 6.005405405405406, | |
| "grad_norm": 4.734610557556152, | |
| "learning_rate": 9.712569712569714e-06, | |
| "loss": 0.4654, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 6.006177606177606, | |
| "grad_norm": 12.163606643676758, | |
| "learning_rate": 9.703989703989705e-06, | |
| "loss": 0.826, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 6.006949806949807, | |
| "grad_norm": 6.117342948913574, | |
| "learning_rate": 9.695409695409696e-06, | |
| "loss": 0.5362, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 6.007722007722008, | |
| "grad_norm": 44.437744140625, | |
| "learning_rate": 9.686829686829687e-06, | |
| "loss": 0.6832, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 6.008494208494208, | |
| "grad_norm": 1.0263047218322754, | |
| "learning_rate": 9.67824967824968e-06, | |
| "loss": 0.6571, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 6.009266409266409, | |
| "grad_norm": 44.683998107910156, | |
| "learning_rate": 9.669669669669671e-06, | |
| "loss": 0.6011, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 6.01003861003861, | |
| "grad_norm": 99.36419677734375, | |
| "learning_rate": 9.661089661089662e-06, | |
| "loss": 0.554, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 6.010810810810811, | |
| "grad_norm": 63.681400299072266, | |
| "learning_rate": 9.652509652509653e-06, | |
| "loss": 0.6625, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 6.011583011583012, | |
| "grad_norm": 6.034546375274658, | |
| "learning_rate": 9.643929643929644e-06, | |
| "loss": 0.7592, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 6.012355212355212, | |
| "grad_norm": 32.6746940612793, | |
| "learning_rate": 9.635349635349635e-06, | |
| "loss": 0.7335, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 6.013127413127413, | |
| "grad_norm": 25.542985916137695, | |
| "learning_rate": 9.626769626769628e-06, | |
| "loss": 0.4995, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 6.013899613899614, | |
| "grad_norm": 35.672698974609375, | |
| "learning_rate": 9.61818961818962e-06, | |
| "loss": 0.623, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 6.014671814671814, | |
| "grad_norm": 0.8925091624259949, | |
| "learning_rate": 9.60960960960961e-06, | |
| "loss": 0.5084, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 6.015444015444015, | |
| "grad_norm": 0.9361721277236938, | |
| "learning_rate": 9.601029601029601e-06, | |
| "loss": 0.375, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 6.0162162162162165, | |
| "grad_norm": 80.69046783447266, | |
| "learning_rate": 9.592449592449593e-06, | |
| "loss": 0.326, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 6.016988416988417, | |
| "grad_norm": 123.89753723144531, | |
| "learning_rate": 9.583869583869585e-06, | |
| "loss": 0.6367, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 6.017760617760618, | |
| "grad_norm": 0.18116500973701477, | |
| "learning_rate": 9.575289575289576e-06, | |
| "loss": 0.565, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 6.018532818532819, | |
| "grad_norm": 33.38082504272461, | |
| "learning_rate": 9.566709566709568e-06, | |
| "loss": 0.9349, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 6.019305019305019, | |
| "grad_norm": 51.91962814331055, | |
| "learning_rate": 9.558129558129559e-06, | |
| "loss": 0.6956, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 6.02007722007722, | |
| "grad_norm": 4.865780830383301, | |
| "learning_rate": 9.54954954954955e-06, | |
| "loss": 0.4644, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 6.02007722007722, | |
| "eval_accuracy": 0.7111111111111111, | |
| "eval_loss": 0.6549674272537231, | |
| "eval_runtime": 13.794, | |
| "eval_samples_per_second": 3.262, | |
| "eval_steps_per_second": 1.087, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 7.000772200772201, | |
| "grad_norm": 20.322017669677734, | |
| "learning_rate": 9.540969540969541e-06, | |
| "loss": 0.3135, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 7.001544401544401, | |
| "grad_norm": 1.3213744163513184, | |
| "learning_rate": 9.532389532389534e-06, | |
| "loss": 0.539, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 7.002316602316602, | |
| "grad_norm": 31.4146728515625, | |
| "learning_rate": 9.523809523809525e-06, | |
| "loss": 0.2655, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 7.003088803088803, | |
| "grad_norm": 0.04433909431099892, | |
| "learning_rate": 9.515229515229516e-06, | |
| "loss": 0.8404, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 7.003861003861004, | |
| "grad_norm": 109.51227569580078, | |
| "learning_rate": 9.506649506649509e-06, | |
| "loss": 0.4611, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 7.004633204633205, | |
| "grad_norm": 0.7903939485549927, | |
| "learning_rate": 9.498069498069498e-06, | |
| "loss": 0.8119, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 7.005405405405406, | |
| "grad_norm": 6.808485507965088, | |
| "learning_rate": 9.489489489489491e-06, | |
| "loss": 0.4345, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 7.006177606177606, | |
| "grad_norm": 11.219454765319824, | |
| "learning_rate": 9.480909480909482e-06, | |
| "loss": 0.5567, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 7.006949806949807, | |
| "grad_norm": 13.435988426208496, | |
| "learning_rate": 9.472329472329473e-06, | |
| "loss": 0.6326, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 7.007722007722008, | |
| "grad_norm": 58.40050506591797, | |
| "learning_rate": 9.463749463749464e-06, | |
| "loss": 0.6078, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 7.008494208494208, | |
| "grad_norm": 9.723285675048828, | |
| "learning_rate": 9.455169455169457e-06, | |
| "loss": 0.3013, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 7.009266409266409, | |
| "grad_norm": 21.477807998657227, | |
| "learning_rate": 9.446589446589446e-06, | |
| "loss": 0.4369, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 7.01003861003861, | |
| "grad_norm": 34.23115158081055, | |
| "learning_rate": 9.43800943800944e-06, | |
| "loss": 0.8515, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 7.010810810810811, | |
| "grad_norm": 16.622177124023438, | |
| "learning_rate": 9.42942942942943e-06, | |
| "loss": 0.5567, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 7.011583011583012, | |
| "grad_norm": 19.550809860229492, | |
| "learning_rate": 9.420849420849421e-06, | |
| "loss": 0.272, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 7.012355212355212, | |
| "grad_norm": 16.057876586914062, | |
| "learning_rate": 9.412269412269412e-06, | |
| "loss": 2.3903, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 7.013127413127413, | |
| "grad_norm": 13.52114200592041, | |
| "learning_rate": 9.403689403689405e-06, | |
| "loss": 0.7999, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 7.013899613899614, | |
| "grad_norm": 9.787277221679688, | |
| "learning_rate": 9.395109395109396e-06, | |
| "loss": 0.4997, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 7.014671814671814, | |
| "grad_norm": 6.769506454467773, | |
| "learning_rate": 9.386529386529387e-06, | |
| "loss": 0.5747, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 7.015444015444015, | |
| "grad_norm": 23.32634925842285, | |
| "learning_rate": 9.377949377949379e-06, | |
| "loss": 0.6525, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 7.0162162162162165, | |
| "grad_norm": 0.12107282876968384, | |
| "learning_rate": 9.36936936936937e-06, | |
| "loss": 0.4083, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 7.016988416988417, | |
| "grad_norm": 32.930877685546875, | |
| "learning_rate": 9.360789360789362e-06, | |
| "loss": 0.6903, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 7.017760617760618, | |
| "grad_norm": 0.47700074315071106, | |
| "learning_rate": 9.352209352209352e-06, | |
| "loss": 0.3165, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 7.018532818532819, | |
| "grad_norm": 17.217500686645508, | |
| "learning_rate": 9.343629343629345e-06, | |
| "loss": 0.866, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 7.019305019305019, | |
| "grad_norm": 3.1114206314086914, | |
| "learning_rate": 9.335049335049336e-06, | |
| "loss": 0.5437, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 7.02007722007722, | |
| "grad_norm": 0.07721736282110214, | |
| "learning_rate": 9.326469326469327e-06, | |
| "loss": 0.7702, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 7.02007722007722, | |
| "eval_accuracy": 0.7333333333333333, | |
| "eval_loss": 0.6853463649749756, | |
| "eval_runtime": 13.0358, | |
| "eval_samples_per_second": 3.452, | |
| "eval_steps_per_second": 1.151, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 8.0007722007722, | |
| "grad_norm": 3.543567657470703, | |
| "learning_rate": 9.317889317889318e-06, | |
| "loss": 0.51, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 8.001544401544402, | |
| "grad_norm": 1.479987621307373, | |
| "learning_rate": 9.30930930930931e-06, | |
| "loss": 0.3237, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 8.002316602316602, | |
| "grad_norm": 23.15900993347168, | |
| "learning_rate": 9.300729300729302e-06, | |
| "loss": 0.8069, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 8.003088803088803, | |
| "grad_norm": 19.172855377197266, | |
| "learning_rate": 9.292149292149293e-06, | |
| "loss": 0.6838, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 8.003861003861005, | |
| "grad_norm": 36.1146240234375, | |
| "learning_rate": 9.283569283569284e-06, | |
| "loss": 0.644, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 8.004633204633205, | |
| "grad_norm": 18.696842193603516, | |
| "learning_rate": 9.274989274989275e-06, | |
| "loss": 0.4751, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 8.005405405405405, | |
| "grad_norm": 12.750860214233398, | |
| "learning_rate": 9.266409266409268e-06, | |
| "loss": 0.6119, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 8.006177606177607, | |
| "grad_norm": 40.49456787109375, | |
| "learning_rate": 9.257829257829259e-06, | |
| "loss": 0.7243, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 8.006949806949807, | |
| "grad_norm": 8.020015716552734, | |
| "learning_rate": 9.24924924924925e-06, | |
| "loss": 0.3259, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 8.007722007722007, | |
| "grad_norm": 30.134857177734375, | |
| "learning_rate": 9.240669240669241e-06, | |
| "loss": 0.3053, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 8.00849420849421, | |
| "grad_norm": 35.37698745727539, | |
| "learning_rate": 9.232089232089232e-06, | |
| "loss": 1.0682, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 8.00926640926641, | |
| "grad_norm": 42.585323333740234, | |
| "learning_rate": 9.223509223509223e-06, | |
| "loss": 0.5787, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 8.01003861003861, | |
| "grad_norm": 26.13615608215332, | |
| "learning_rate": 9.214929214929216e-06, | |
| "loss": 0.6936, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 8.010810810810812, | |
| "grad_norm": 40.68016815185547, | |
| "learning_rate": 9.206349206349207e-06, | |
| "loss": 0.4108, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 8.011583011583012, | |
| "grad_norm": 0.34476640820503235, | |
| "learning_rate": 9.197769197769198e-06, | |
| "loss": 0.5791, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 8.012355212355212, | |
| "grad_norm": 0.125451922416687, | |
| "learning_rate": 9.189189189189191e-06, | |
| "loss": 0.6043, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 8.013127413127414, | |
| "grad_norm": 46.2276611328125, | |
| "learning_rate": 9.18060918060918e-06, | |
| "loss": 0.8779, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 8.013899613899614, | |
| "grad_norm": 2.52577543258667, | |
| "learning_rate": 9.172029172029173e-06, | |
| "loss": 0.4187, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 8.014671814671814, | |
| "grad_norm": 3.3626627922058105, | |
| "learning_rate": 9.163449163449165e-06, | |
| "loss": 0.4761, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 8.015444015444016, | |
| "grad_norm": 27.020002365112305, | |
| "learning_rate": 9.154869154869156e-06, | |
| "loss": 0.3704, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 8.016216216216216, | |
| "grad_norm": 0.6957538723945618, | |
| "learning_rate": 9.146289146289147e-06, | |
| "loss": 0.4882, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 8.016988416988417, | |
| "grad_norm": 3.1972832679748535, | |
| "learning_rate": 9.137709137709138e-06, | |
| "loss": 0.801, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 8.017760617760617, | |
| "grad_norm": 38.638031005859375, | |
| "learning_rate": 9.129129129129129e-06, | |
| "loss": 0.3989, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 8.018532818532819, | |
| "grad_norm": 60.1053352355957, | |
| "learning_rate": 9.120549120549122e-06, | |
| "loss": 0.6521, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 8.019305019305019, | |
| "grad_norm": 9.336408615112305, | |
| "learning_rate": 9.111969111969113e-06, | |
| "loss": 0.3019, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 8.02007722007722, | |
| "grad_norm": 0.277237206697464, | |
| "learning_rate": 9.103389103389104e-06, | |
| "loss": 0.4589, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 8.02007722007722, | |
| "eval_accuracy": 0.7555555555555555, | |
| "eval_loss": 0.7447373867034912, | |
| "eval_runtime": 13.042, | |
| "eval_samples_per_second": 3.45, | |
| "eval_steps_per_second": 1.15, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 9.0007722007722, | |
| "grad_norm": 26.95766258239746, | |
| "learning_rate": 9.094809094809095e-06, | |
| "loss": 0.2924, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 9.001544401544402, | |
| "grad_norm": 0.3911036550998688, | |
| "learning_rate": 9.086229086229086e-06, | |
| "loss": 0.2926, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 9.002316602316602, | |
| "grad_norm": 15.294562339782715, | |
| "learning_rate": 9.077649077649079e-06, | |
| "loss": 0.6503, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 9.003088803088803, | |
| "grad_norm": 0.5500653982162476, | |
| "learning_rate": 9.06906906906907e-06, | |
| "loss": 0.4895, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 9.003861003861005, | |
| "grad_norm": 68.16986846923828, | |
| "learning_rate": 9.060489060489061e-06, | |
| "loss": 0.8309, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 9.004633204633205, | |
| "grad_norm": 30.308658599853516, | |
| "learning_rate": 9.051909051909052e-06, | |
| "loss": 0.4314, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 9.005405405405405, | |
| "grad_norm": 16.645893096923828, | |
| "learning_rate": 9.043329043329045e-06, | |
| "loss": 0.6478, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 9.006177606177607, | |
| "grad_norm": 114.66397857666016, | |
| "learning_rate": 9.034749034749034e-06, | |
| "loss": 0.6211, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 9.006949806949807, | |
| "grad_norm": 24.798070907592773, | |
| "learning_rate": 9.026169026169027e-06, | |
| "loss": 0.3774, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 9.007722007722007, | |
| "grad_norm": 17.036094665527344, | |
| "learning_rate": 9.017589017589018e-06, | |
| "loss": 0.684, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 9.00849420849421, | |
| "grad_norm": 22.78154945373535, | |
| "learning_rate": 9.00900900900901e-06, | |
| "loss": 0.5217, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 9.00926640926641, | |
| "grad_norm": 17.106143951416016, | |
| "learning_rate": 9.000429000429e-06, | |
| "loss": 0.4613, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 9.01003861003861, | |
| "grad_norm": 89.09002685546875, | |
| "learning_rate": 8.991848991848993e-06, | |
| "loss": 0.6451, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 9.010810810810812, | |
| "grad_norm": 51.08162307739258, | |
| "learning_rate": 8.983268983268984e-06, | |
| "loss": 0.4516, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 9.011583011583012, | |
| "grad_norm": 17.878747940063477, | |
| "learning_rate": 8.974688974688976e-06, | |
| "loss": 0.3903, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 9.012355212355212, | |
| "grad_norm": 9.77619743347168, | |
| "learning_rate": 8.966108966108967e-06, | |
| "loss": 0.6817, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 9.013127413127414, | |
| "grad_norm": 9.895012855529785, | |
| "learning_rate": 8.957528957528958e-06, | |
| "loss": 0.5136, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 9.013899613899614, | |
| "grad_norm": 1.2748266458511353, | |
| "learning_rate": 8.94894894894895e-06, | |
| "loss": 0.6643, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 9.014671814671814, | |
| "grad_norm": 40.85063552856445, | |
| "learning_rate": 8.940368940368942e-06, | |
| "loss": 0.7568, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 9.015444015444016, | |
| "grad_norm": 28.490476608276367, | |
| "learning_rate": 8.931788931788933e-06, | |
| "loss": 0.4918, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 9.016216216216216, | |
| "grad_norm": 18.73026466369629, | |
| "learning_rate": 8.923208923208924e-06, | |
| "loss": 0.4231, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 9.016988416988417, | |
| "grad_norm": 6.791349411010742, | |
| "learning_rate": 8.914628914628915e-06, | |
| "loss": 0.4267, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 9.017760617760617, | |
| "grad_norm": 0.5822336673736572, | |
| "learning_rate": 8.906048906048906e-06, | |
| "loss": 0.2161, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 9.018532818532819, | |
| "grad_norm": 8.814290046691895, | |
| "learning_rate": 8.897468897468899e-06, | |
| "loss": 0.3644, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 9.019305019305019, | |
| "grad_norm": 45.50792694091797, | |
| "learning_rate": 8.888888888888888e-06, | |
| "loss": 0.5233, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 9.02007722007722, | |
| "grad_norm": 0.14075778424739838, | |
| "learning_rate": 8.880308880308881e-06, | |
| "loss": 0.5026, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 9.02007722007722, | |
| "eval_accuracy": 0.7777777777777778, | |
| "eval_loss": 0.742344081401825, | |
| "eval_runtime": 11.6183, | |
| "eval_samples_per_second": 3.873, | |
| "eval_steps_per_second": 1.291, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 10.0007722007722, | |
| "grad_norm": 0.050024695694446564, | |
| "learning_rate": 8.871728871728872e-06, | |
| "loss": 0.5309, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 10.001544401544402, | |
| "grad_norm": 36.7778205871582, | |
| "learning_rate": 8.863148863148863e-06, | |
| "loss": 0.5166, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 10.002316602316602, | |
| "grad_norm": 1.0572352409362793, | |
| "learning_rate": 8.854568854568856e-06, | |
| "loss": 0.4621, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 10.003088803088803, | |
| "grad_norm": 34.59428024291992, | |
| "learning_rate": 8.845988845988847e-06, | |
| "loss": 1.0248, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 10.003861003861005, | |
| "grad_norm": 35.09554672241211, | |
| "learning_rate": 8.837408837408838e-06, | |
| "loss": 0.6444, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 10.004633204633205, | |
| "grad_norm": 6.455074310302734, | |
| "learning_rate": 8.82882882882883e-06, | |
| "loss": 0.2533, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 10.005405405405405, | |
| "grad_norm": 0.7161849141120911, | |
| "learning_rate": 8.82024882024882e-06, | |
| "loss": 0.9718, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 10.006177606177607, | |
| "grad_norm": 24.07243537902832, | |
| "learning_rate": 8.811668811668812e-06, | |
| "loss": 0.6386, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 10.006949806949807, | |
| "grad_norm": 1.7033041715621948, | |
| "learning_rate": 8.803088803088804e-06, | |
| "loss": 0.536, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 10.007722007722007, | |
| "grad_norm": 40.123138427734375, | |
| "learning_rate": 8.794508794508795e-06, | |
| "loss": 0.7141, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 10.00849420849421, | |
| "grad_norm": 27.220144271850586, | |
| "learning_rate": 8.785928785928787e-06, | |
| "loss": 0.5531, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 10.00926640926641, | |
| "grad_norm": 12.67369556427002, | |
| "learning_rate": 8.777348777348778e-06, | |
| "loss": 0.3725, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 10.01003861003861, | |
| "grad_norm": 3.601840019226074, | |
| "learning_rate": 8.768768768768769e-06, | |
| "loss": 0.5135, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 10.010810810810812, | |
| "grad_norm": 91.34984588623047, | |
| "learning_rate": 8.760188760188762e-06, | |
| "loss": 0.2448, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 10.011583011583012, | |
| "grad_norm": 89.07131958007812, | |
| "learning_rate": 8.751608751608753e-06, | |
| "loss": 0.2126, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 10.012355212355212, | |
| "grad_norm": 0.4536842405796051, | |
| "learning_rate": 8.743028743028744e-06, | |
| "loss": 0.1193, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 10.013127413127414, | |
| "grad_norm": 150.1531219482422, | |
| "learning_rate": 8.734448734448735e-06, | |
| "loss": 0.857, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 10.013899613899614, | |
| "grad_norm": 167.5815887451172, | |
| "learning_rate": 8.725868725868728e-06, | |
| "loss": 0.6876, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 10.014671814671814, | |
| "grad_norm": 0.09778659045696259, | |
| "learning_rate": 8.717288717288717e-06, | |
| "loss": 0.5219, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 10.015444015444016, | |
| "grad_norm": 55.60439682006836, | |
| "learning_rate": 8.70870870870871e-06, | |
| "loss": 0.825, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 10.016216216216216, | |
| "grad_norm": 115.7952880859375, | |
| "learning_rate": 8.700128700128701e-06, | |
| "loss": 0.6771, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 10.016988416988417, | |
| "grad_norm": 76.89239501953125, | |
| "learning_rate": 8.691548691548692e-06, | |
| "loss": 0.3557, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 10.017760617760617, | |
| "grad_norm": 7.334883689880371, | |
| "learning_rate": 8.682968682968683e-06, | |
| "loss": 0.3834, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 10.018532818532819, | |
| "grad_norm": 26.45685386657715, | |
| "learning_rate": 8.674388674388674e-06, | |
| "loss": 0.3072, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 10.019305019305019, | |
| "grad_norm": 16.094499588012695, | |
| "learning_rate": 8.665808665808665e-06, | |
| "loss": 0.9027, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 10.02007722007722, | |
| "grad_norm": 0.5503826141357422, | |
| "learning_rate": 8.657228657228658e-06, | |
| "loss": 0.5612, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 10.02007722007722, | |
| "eval_accuracy": 0.6888888888888889, | |
| "eval_loss": 0.8798052668571472, | |
| "eval_runtime": 11.5437, | |
| "eval_samples_per_second": 3.898, | |
| "eval_steps_per_second": 1.299, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 11.0007722007722, | |
| "grad_norm": 38.276878356933594, | |
| "learning_rate": 8.64864864864865e-06, | |
| "loss": 0.5818, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 11.001544401544402, | |
| "grad_norm": 0.12790539860725403, | |
| "learning_rate": 8.64006864006864e-06, | |
| "loss": 0.1793, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 11.002316602316602, | |
| "grad_norm": 6.592782020568848, | |
| "learning_rate": 8.631488631488633e-06, | |
| "loss": 0.7483, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 11.003088803088803, | |
| "grad_norm": 77.4202651977539, | |
| "learning_rate": 8.622908622908623e-06, | |
| "loss": 0.6152, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 11.003861003861005, | |
| "grad_norm": 0.5944340825080872, | |
| "learning_rate": 8.614328614328615e-06, | |
| "loss": 0.337, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 11.004633204633205, | |
| "grad_norm": 0.9054240584373474, | |
| "learning_rate": 8.605748605748607e-06, | |
| "loss": 0.1523, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 11.005405405405405, | |
| "grad_norm": 75.37146759033203, | |
| "learning_rate": 8.597168597168598e-06, | |
| "loss": 0.5706, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 11.006177606177607, | |
| "grad_norm": 68.32598876953125, | |
| "learning_rate": 8.588588588588589e-06, | |
| "loss": 0.7453, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 11.006949806949807, | |
| "grad_norm": 86.55941772460938, | |
| "learning_rate": 8.580008580008582e-06, | |
| "loss": 0.6767, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 11.007722007722007, | |
| "grad_norm": 0.5787020921707153, | |
| "learning_rate": 8.571428571428571e-06, | |
| "loss": 0.5304, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 11.00849420849421, | |
| "grad_norm": 1.052687406539917, | |
| "learning_rate": 8.562848562848564e-06, | |
| "loss": 0.4398, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 11.00926640926641, | |
| "grad_norm": 100.41271209716797, | |
| "learning_rate": 8.554268554268555e-06, | |
| "loss": 0.4632, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 11.01003861003861, | |
| "grad_norm": 12.1259183883667, | |
| "learning_rate": 8.545688545688546e-06, | |
| "loss": 0.6563, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 11.010810810810812, | |
| "grad_norm": 0.3354950249195099, | |
| "learning_rate": 8.537108537108539e-06, | |
| "loss": 0.5515, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 11.011583011583012, | |
| "grad_norm": 0.6421205997467041, | |
| "learning_rate": 8.52852852852853e-06, | |
| "loss": 0.0537, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 11.012355212355212, | |
| "grad_norm": 0.13986188173294067, | |
| "learning_rate": 8.519948519948521e-06, | |
| "loss": 0.8416, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 11.013127413127414, | |
| "grad_norm": 70.42572784423828, | |
| "learning_rate": 8.511368511368512e-06, | |
| "loss": 0.7627, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 11.013899613899614, | |
| "grad_norm": 3.2118000984191895, | |
| "learning_rate": 8.502788502788503e-06, | |
| "loss": 0.4206, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 11.014671814671814, | |
| "grad_norm": 175.60455322265625, | |
| "learning_rate": 8.494208494208494e-06, | |
| "loss": 0.8047, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 11.015444015444016, | |
| "grad_norm": 0.3263649344444275, | |
| "learning_rate": 8.485628485628487e-06, | |
| "loss": 0.5895, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 11.016216216216216, | |
| "grad_norm": 0.11546669155359268, | |
| "learning_rate": 8.477048477048478e-06, | |
| "loss": 0.7429, | |
| "step": 3070 | |
| }, | |
| { | |
| "epoch": 11.016988416988417, | |
| "grad_norm": 10.180035591125488, | |
| "learning_rate": 8.46846846846847e-06, | |
| "loss": 1.0899, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 11.017760617760617, | |
| "grad_norm": 0.3308240473270416, | |
| "learning_rate": 8.45988845988846e-06, | |
| "loss": 0.3317, | |
| "step": 3090 | |
| }, | |
| { | |
| "epoch": 11.018532818532819, | |
| "grad_norm": 11.988396644592285, | |
| "learning_rate": 8.451308451308451e-06, | |
| "loss": 0.9498, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 11.019305019305019, | |
| "grad_norm": 1.0947588682174683, | |
| "learning_rate": 8.442728442728444e-06, | |
| "loss": 0.4188, | |
| "step": 3110 | |
| }, | |
| { | |
| "epoch": 11.02007722007722, | |
| "grad_norm": 0.16421347856521606, | |
| "learning_rate": 8.434148434148435e-06, | |
| "loss": 0.536, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 11.02007722007722, | |
| "eval_accuracy": 0.4888888888888889, | |
| "eval_loss": 1.5358043909072876, | |
| "eval_runtime": 12.3592, | |
| "eval_samples_per_second": 3.641, | |
| "eval_steps_per_second": 1.214, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 12.0007722007722, | |
| "grad_norm": 46.494842529296875, | |
| "learning_rate": 8.425568425568426e-06, | |
| "loss": 0.306, | |
| "step": 3130 | |
| }, | |
| { | |
| "epoch": 12.001544401544402, | |
| "grad_norm": 15.413254737854004, | |
| "learning_rate": 8.416988416988418e-06, | |
| "loss": 0.3972, | |
| "step": 3140 | |
| }, | |
| { | |
| "epoch": 12.002316602316602, | |
| "grad_norm": 29.484281539916992, | |
| "learning_rate": 8.408408408408409e-06, | |
| "loss": 0.5977, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 12.003088803088803, | |
| "grad_norm": 45.659481048583984, | |
| "learning_rate": 8.3998283998284e-06, | |
| "loss": 0.5106, | |
| "step": 3160 | |
| }, | |
| { | |
| "epoch": 12.003861003861005, | |
| "grad_norm": 115.90227508544922, | |
| "learning_rate": 8.391248391248393e-06, | |
| "loss": 0.653, | |
| "step": 3170 | |
| }, | |
| { | |
| "epoch": 12.004633204633205, | |
| "grad_norm": 0.9203261733055115, | |
| "learning_rate": 8.382668382668384e-06, | |
| "loss": 0.479, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 12.005405405405405, | |
| "grad_norm": 34.71575927734375, | |
| "learning_rate": 8.374088374088375e-06, | |
| "loss": 0.8072, | |
| "step": 3190 | |
| }, | |
| { | |
| "epoch": 12.006177606177607, | |
| "grad_norm": 32.08717346191406, | |
| "learning_rate": 8.365508365508366e-06, | |
| "loss": 0.5189, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 12.006949806949807, | |
| "grad_norm": 0.04885344207286835, | |
| "learning_rate": 8.356928356928357e-06, | |
| "loss": 0.5292, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 12.007722007722007, | |
| "grad_norm": 108.91436004638672, | |
| "learning_rate": 8.348348348348348e-06, | |
| "loss": 0.6934, | |
| "step": 3220 | |
| }, | |
| { | |
| "epoch": 12.00849420849421, | |
| "grad_norm": 7.798171520233154, | |
| "learning_rate": 8.33976833976834e-06, | |
| "loss": 0.6352, | |
| "step": 3230 | |
| }, | |
| { | |
| "epoch": 12.00926640926641, | |
| "grad_norm": 0.5554088950157166, | |
| "learning_rate": 8.331188331188332e-06, | |
| "loss": 0.3745, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 12.01003861003861, | |
| "grad_norm": 85.23207092285156, | |
| "learning_rate": 8.322608322608323e-06, | |
| "loss": 0.3618, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 12.010810810810812, | |
| "grad_norm": 0.41603294014930725, | |
| "learning_rate": 8.314028314028316e-06, | |
| "loss": 0.4017, | |
| "step": 3260 | |
| }, | |
| { | |
| "epoch": 12.011583011583012, | |
| "grad_norm": 43.02106857299805, | |
| "learning_rate": 8.305448305448305e-06, | |
| "loss": 0.7779, | |
| "step": 3270 | |
| }, | |
| { | |
| "epoch": 12.012355212355212, | |
| "grad_norm": 15.187088012695312, | |
| "learning_rate": 8.296868296868298e-06, | |
| "loss": 0.3938, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 12.013127413127414, | |
| "grad_norm": 0.11593673378229141, | |
| "learning_rate": 8.288288288288289e-06, | |
| "loss": 0.4054, | |
| "step": 3290 | |
| }, | |
| { | |
| "epoch": 12.013899613899614, | |
| "grad_norm": 44.22844696044922, | |
| "learning_rate": 8.27970827970828e-06, | |
| "loss": 0.9659, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 12.014671814671814, | |
| "grad_norm": 39.142581939697266, | |
| "learning_rate": 8.271128271128271e-06, | |
| "loss": 0.1711, | |
| "step": 3310 | |
| }, | |
| { | |
| "epoch": 12.015444015444016, | |
| "grad_norm": 29.718122482299805, | |
| "learning_rate": 8.262548262548264e-06, | |
| "loss": 0.3803, | |
| "step": 3320 | |
| }, | |
| { | |
| "epoch": 12.016216216216216, | |
| "grad_norm": 2.7277615070343018, | |
| "learning_rate": 8.253968253968254e-06, | |
| "loss": 0.6516, | |
| "step": 3330 | |
| }, | |
| { | |
| "epoch": 12.016988416988417, | |
| "grad_norm": 9.546491622924805, | |
| "learning_rate": 8.245388245388246e-06, | |
| "loss": 1.0321, | |
| "step": 3340 | |
| }, | |
| { | |
| "epoch": 12.017760617760617, | |
| "grad_norm": 30.349594116210938, | |
| "learning_rate": 8.236808236808237e-06, | |
| "loss": 0.739, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 12.018532818532819, | |
| "grad_norm": 13.018799781799316, | |
| "learning_rate": 8.228228228228229e-06, | |
| "loss": 0.2881, | |
| "step": 3360 | |
| }, | |
| { | |
| "epoch": 12.019305019305019, | |
| "grad_norm": 1.475889801979065, | |
| "learning_rate": 8.219648219648221e-06, | |
| "loss": 0.6082, | |
| "step": 3370 | |
| }, | |
| { | |
| "epoch": 12.02007722007722, | |
| "grad_norm": 33.09438705444336, | |
| "learning_rate": 8.211068211068212e-06, | |
| "loss": 0.6695, | |
| "step": 3380 | |
| }, | |
| { | |
| "epoch": 12.02007722007722, | |
| "eval_accuracy": 0.6666666666666666, | |
| "eval_loss": 1.3352181911468506, | |
| "eval_runtime": 11.6323, | |
| "eval_samples_per_second": 3.869, | |
| "eval_steps_per_second": 1.29, | |
| "step": 3380 | |
| }, | |
| { | |
| "epoch": 13.0007722007722, | |
| "grad_norm": 0.033746566623449326, | |
| "learning_rate": 8.202488202488204e-06, | |
| "loss": 0.421, | |
| "step": 3390 | |
| }, | |
| { | |
| "epoch": 13.001544401544402, | |
| "grad_norm": 0.36186423897743225, | |
| "learning_rate": 8.193908193908195e-06, | |
| "loss": 0.7725, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 13.002316602316602, | |
| "grad_norm": 0.7419402599334717, | |
| "learning_rate": 8.185328185328186e-06, | |
| "loss": 0.2973, | |
| "step": 3410 | |
| }, | |
| { | |
| "epoch": 13.003088803088803, | |
| "grad_norm": 16.69759750366211, | |
| "learning_rate": 8.176748176748177e-06, | |
| "loss": 0.6846, | |
| "step": 3420 | |
| }, | |
| { | |
| "epoch": 13.003861003861005, | |
| "grad_norm": 1.2755367755889893, | |
| "learning_rate": 8.16816816816817e-06, | |
| "loss": 0.4318, | |
| "step": 3430 | |
| }, | |
| { | |
| "epoch": 13.004633204633205, | |
| "grad_norm": 10.125988006591797, | |
| "learning_rate": 8.159588159588159e-06, | |
| "loss": 0.5978, | |
| "step": 3440 | |
| }, | |
| { | |
| "epoch": 13.005405405405405, | |
| "grad_norm": 0.27345284819602966, | |
| "learning_rate": 8.151008151008152e-06, | |
| "loss": 0.7275, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 13.006177606177607, | |
| "grad_norm": 2.0094382762908936, | |
| "learning_rate": 8.142428142428143e-06, | |
| "loss": 0.3855, | |
| "step": 3460 | |
| }, | |
| { | |
| "epoch": 13.006949806949807, | |
| "grad_norm": 30.24265480041504, | |
| "learning_rate": 8.133848133848134e-06, | |
| "loss": 0.3675, | |
| "step": 3470 | |
| }, | |
| { | |
| "epoch": 13.007722007722007, | |
| "grad_norm": 28.360801696777344, | |
| "learning_rate": 8.125268125268127e-06, | |
| "loss": 0.8046, | |
| "step": 3480 | |
| }, | |
| { | |
| "epoch": 13.00849420849421, | |
| "grad_norm": 0.12912389636039734, | |
| "learning_rate": 8.116688116688118e-06, | |
| "loss": 0.7357, | |
| "step": 3490 | |
| }, | |
| { | |
| "epoch": 13.00926640926641, | |
| "grad_norm": 0.8967646956443787, | |
| "learning_rate": 8.108108108108109e-06, | |
| "loss": 0.5318, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 13.01003861003861, | |
| "grad_norm": 84.27301788330078, | |
| "learning_rate": 8.0995280995281e-06, | |
| "loss": 0.764, | |
| "step": 3510 | |
| }, | |
| { | |
| "epoch": 13.010810810810812, | |
| "grad_norm": 10.367392539978027, | |
| "learning_rate": 8.090948090948091e-06, | |
| "loss": 0.4441, | |
| "step": 3520 | |
| }, | |
| { | |
| "epoch": 13.011583011583012, | |
| "grad_norm": 77.54997253417969, | |
| "learning_rate": 8.082368082368082e-06, | |
| "loss": 0.3571, | |
| "step": 3530 | |
| }, | |
| { | |
| "epoch": 13.012355212355212, | |
| "grad_norm": 0.0930776447057724, | |
| "learning_rate": 8.073788073788075e-06, | |
| "loss": 0.7569, | |
| "step": 3540 | |
| }, | |
| { | |
| "epoch": 13.013127413127414, | |
| "grad_norm": 27.489704132080078, | |
| "learning_rate": 8.065208065208066e-06, | |
| "loss": 0.2051, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 13.013899613899614, | |
| "grad_norm": 7.215950965881348, | |
| "learning_rate": 8.056628056628057e-06, | |
| "loss": 0.4319, | |
| "step": 3560 | |
| }, | |
| { | |
| "epoch": 13.014671814671814, | |
| "grad_norm": 2.232342481613159, | |
| "learning_rate": 8.048048048048048e-06, | |
| "loss": 0.6153, | |
| "step": 3570 | |
| }, | |
| { | |
| "epoch": 13.015444015444016, | |
| "grad_norm": 8.222293853759766, | |
| "learning_rate": 8.03946803946804e-06, | |
| "loss": 0.7993, | |
| "step": 3580 | |
| }, | |
| { | |
| "epoch": 13.016216216216216, | |
| "grad_norm": 0.9838555455207825, | |
| "learning_rate": 8.03088803088803e-06, | |
| "loss": 0.5731, | |
| "step": 3590 | |
| }, | |
| { | |
| "epoch": 13.016988416988417, | |
| "grad_norm": 21.45343589782715, | |
| "learning_rate": 8.022308022308023e-06, | |
| "loss": 0.7791, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 13.017760617760617, | |
| "grad_norm": 0.30541619658470154, | |
| "learning_rate": 8.013728013728015e-06, | |
| "loss": 0.3412, | |
| "step": 3610 | |
| }, | |
| { | |
| "epoch": 13.018532818532819, | |
| "grad_norm": 0.06337114423513412, | |
| "learning_rate": 8.005148005148006e-06, | |
| "loss": 0.2777, | |
| "step": 3620 | |
| }, | |
| { | |
| "epoch": 13.019305019305019, | |
| "grad_norm": 0.14662937819957733, | |
| "learning_rate": 7.996567996567998e-06, | |
| "loss": 0.7978, | |
| "step": 3630 | |
| }, | |
| { | |
| "epoch": 13.02007722007722, | |
| "grad_norm": 9.782637596130371, | |
| "learning_rate": 7.987987987987988e-06, | |
| "loss": 0.2699, | |
| "step": 3640 | |
| }, | |
| { | |
| "epoch": 13.02007722007722, | |
| "eval_accuracy": 0.7333333333333333, | |
| "eval_loss": 1.1053425073623657, | |
| "eval_runtime": 13.0668, | |
| "eval_samples_per_second": 3.444, | |
| "eval_steps_per_second": 1.148, | |
| "step": 3640 | |
| }, | |
| { | |
| "epoch": 14.0007722007722, | |
| "grad_norm": 0.04395415633916855, | |
| "learning_rate": 7.97940797940798e-06, | |
| "loss": 0.3761, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 14.001544401544402, | |
| "grad_norm": 0.18629683554172516, | |
| "learning_rate": 7.970827970827972e-06, | |
| "loss": 0.5233, | |
| "step": 3660 | |
| }, | |
| { | |
| "epoch": 14.002316602316602, | |
| "grad_norm": 0.24980448186397552, | |
| "learning_rate": 7.962247962247963e-06, | |
| "loss": 0.4489, | |
| "step": 3670 | |
| }, | |
| { | |
| "epoch": 14.003088803088803, | |
| "grad_norm": 0.09628362208604813, | |
| "learning_rate": 7.953667953667954e-06, | |
| "loss": 0.5638, | |
| "step": 3680 | |
| }, | |
| { | |
| "epoch": 14.003861003861005, | |
| "grad_norm": 0.016532419249415398, | |
| "learning_rate": 7.945087945087945e-06, | |
| "loss": 0.8614, | |
| "step": 3690 | |
| }, | |
| { | |
| "epoch": 14.004633204633205, | |
| "grad_norm": 137.2128448486328, | |
| "learning_rate": 7.936507936507936e-06, | |
| "loss": 0.2668, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 14.005405405405405, | |
| "grad_norm": 14.007122039794922, | |
| "learning_rate": 7.927927927927929e-06, | |
| "loss": 0.5763, | |
| "step": 3710 | |
| }, | |
| { | |
| "epoch": 14.006177606177607, | |
| "grad_norm": 0.029359988868236542, | |
| "learning_rate": 7.91934791934792e-06, | |
| "loss": 0.551, | |
| "step": 3720 | |
| }, | |
| { | |
| "epoch": 14.006949806949807, | |
| "grad_norm": 1.142870545387268, | |
| "learning_rate": 7.910767910767911e-06, | |
| "loss": 0.3888, | |
| "step": 3730 | |
| }, | |
| { | |
| "epoch": 14.007722007722007, | |
| "grad_norm": 22.303936004638672, | |
| "learning_rate": 7.902187902187904e-06, | |
| "loss": 0.8488, | |
| "step": 3740 | |
| }, | |
| { | |
| "epoch": 14.00849420849421, | |
| "grad_norm": 10.081164360046387, | |
| "learning_rate": 7.893607893607893e-06, | |
| "loss": 0.3886, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 14.00926640926641, | |
| "grad_norm": 56.892921447753906, | |
| "learning_rate": 7.885027885027886e-06, | |
| "loss": 0.7013, | |
| "step": 3760 | |
| }, | |
| { | |
| "epoch": 14.01003861003861, | |
| "grad_norm": 57.699581146240234, | |
| "learning_rate": 7.876447876447877e-06, | |
| "loss": 0.4642, | |
| "step": 3770 | |
| }, | |
| { | |
| "epoch": 14.010810810810812, | |
| "grad_norm": 78.83262634277344, | |
| "learning_rate": 7.867867867867868e-06, | |
| "loss": 0.5118, | |
| "step": 3780 | |
| }, | |
| { | |
| "epoch": 14.011583011583012, | |
| "grad_norm": 1.0107831954956055, | |
| "learning_rate": 7.85928785928786e-06, | |
| "loss": 0.6178, | |
| "step": 3790 | |
| }, | |
| { | |
| "epoch": 14.012355212355212, | |
| "grad_norm": 0.07490326464176178, | |
| "learning_rate": 7.850707850707852e-06, | |
| "loss": 0.7406, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 14.013127413127414, | |
| "grad_norm": 33.424617767333984, | |
| "learning_rate": 7.842127842127842e-06, | |
| "loss": 0.4972, | |
| "step": 3810 | |
| }, | |
| { | |
| "epoch": 14.013899613899614, | |
| "grad_norm": 0.40565818548202515, | |
| "learning_rate": 7.833547833547834e-06, | |
| "loss": 0.4828, | |
| "step": 3820 | |
| }, | |
| { | |
| "epoch": 14.014671814671814, | |
| "grad_norm": 0.5566960573196411, | |
| "learning_rate": 7.824967824967826e-06, | |
| "loss": 0.2992, | |
| "step": 3830 | |
| }, | |
| { | |
| "epoch": 14.015444015444016, | |
| "grad_norm": 149.54197692871094, | |
| "learning_rate": 7.816387816387817e-06, | |
| "loss": 0.8926, | |
| "step": 3840 | |
| }, | |
| { | |
| "epoch": 14.016216216216216, | |
| "grad_norm": 2.9310715198516846, | |
| "learning_rate": 7.807807807807808e-06, | |
| "loss": 0.8425, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 14.016988416988417, | |
| "grad_norm": 58.41656494140625, | |
| "learning_rate": 7.7992277992278e-06, | |
| "loss": 0.8109, | |
| "step": 3860 | |
| }, | |
| { | |
| "epoch": 14.017760617760617, | |
| "grad_norm": 7.718285083770752, | |
| "learning_rate": 7.790647790647792e-06, | |
| "loss": 0.4207, | |
| "step": 3870 | |
| }, | |
| { | |
| "epoch": 14.018532818532819, | |
| "grad_norm": 0.1663883924484253, | |
| "learning_rate": 7.782067782067783e-06, | |
| "loss": 0.2474, | |
| "step": 3880 | |
| }, | |
| { | |
| "epoch": 14.019305019305019, | |
| "grad_norm": 0.2051921933889389, | |
| "learning_rate": 7.773487773487774e-06, | |
| "loss": 0.3523, | |
| "step": 3890 | |
| }, | |
| { | |
| "epoch": 14.02007722007722, | |
| "grad_norm": 0.03690739721059799, | |
| "learning_rate": 7.764907764907765e-06, | |
| "loss": 0.5277, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 14.02007722007722, | |
| "eval_accuracy": 0.7111111111111111, | |
| "eval_loss": 0.9907832741737366, | |
| "eval_runtime": 13.0331, | |
| "eval_samples_per_second": 3.453, | |
| "eval_steps_per_second": 1.151, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 15.0007722007722, | |
| "grad_norm": 7.357924938201904, | |
| "learning_rate": 7.756327756327758e-06, | |
| "loss": 0.2465, | |
| "step": 3910 | |
| }, | |
| { | |
| "epoch": 15.001544401544402, | |
| "grad_norm": 104.24243927001953, | |
| "learning_rate": 7.747747747747749e-06, | |
| "loss": 0.4626, | |
| "step": 3920 | |
| }, | |
| { | |
| "epoch": 15.002316602316602, | |
| "grad_norm": 47.60615921020508, | |
| "learning_rate": 7.73916773916774e-06, | |
| "loss": 0.4113, | |
| "step": 3930 | |
| }, | |
| { | |
| "epoch": 15.003088803088803, | |
| "grad_norm": 140.14810180664062, | |
| "learning_rate": 7.730587730587731e-06, | |
| "loss": 0.3187, | |
| "step": 3940 | |
| }, | |
| { | |
| "epoch": 15.003861003861005, | |
| "grad_norm": 148.16998291015625, | |
| "learning_rate": 7.722007722007722e-06, | |
| "loss": 0.4277, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 15.004633204633205, | |
| "grad_norm": 6.010843753814697, | |
| "learning_rate": 7.713427713427713e-06, | |
| "loss": 0.3613, | |
| "step": 3960 | |
| }, | |
| { | |
| "epoch": 15.005405405405405, | |
| "grad_norm": 1.0568628311157227, | |
| "learning_rate": 7.704847704847706e-06, | |
| "loss": 0.2409, | |
| "step": 3970 | |
| }, | |
| { | |
| "epoch": 15.006177606177607, | |
| "grad_norm": 1.002126693725586, | |
| "learning_rate": 7.696267696267697e-06, | |
| "loss": 0.1332, | |
| "step": 3980 | |
| }, | |
| { | |
| "epoch": 15.006949806949807, | |
| "grad_norm": 0.5772402286529541, | |
| "learning_rate": 7.687687687687688e-06, | |
| "loss": 0.1781, | |
| "step": 3990 | |
| }, | |
| { | |
| "epoch": 15.007722007722007, | |
| "grad_norm": 6.310009002685547, | |
| "learning_rate": 7.67910767910768e-06, | |
| "loss": 0.3342, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 15.00849420849421, | |
| "grad_norm": 69.40351867675781, | |
| "learning_rate": 7.67052767052767e-06, | |
| "loss": 0.7218, | |
| "step": 4010 | |
| }, | |
| { | |
| "epoch": 15.00926640926641, | |
| "grad_norm": 0.10429790616035461, | |
| "learning_rate": 7.661947661947663e-06, | |
| "loss": 0.369, | |
| "step": 4020 | |
| }, | |
| { | |
| "epoch": 15.01003861003861, | |
| "grad_norm": 129.76885986328125, | |
| "learning_rate": 7.653367653367654e-06, | |
| "loss": 1.0676, | |
| "step": 4030 | |
| }, | |
| { | |
| "epoch": 15.010810810810812, | |
| "grad_norm": 95.47364807128906, | |
| "learning_rate": 7.644787644787645e-06, | |
| "loss": 0.6629, | |
| "step": 4040 | |
| }, | |
| { | |
| "epoch": 15.011583011583012, | |
| "grad_norm": 0.16719844937324524, | |
| "learning_rate": 7.636207636207637e-06, | |
| "loss": 0.3401, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 15.012355212355212, | |
| "grad_norm": 5.297786712646484, | |
| "learning_rate": 7.6276276276276285e-06, | |
| "loss": 1.0181, | |
| "step": 4060 | |
| }, | |
| { | |
| "epoch": 15.013127413127414, | |
| "grad_norm": 20.62660789489746, | |
| "learning_rate": 7.61904761904762e-06, | |
| "loss": 0.8109, | |
| "step": 4070 | |
| }, | |
| { | |
| "epoch": 15.013899613899614, | |
| "grad_norm": 0.046302299946546555, | |
| "learning_rate": 7.610467610467612e-06, | |
| "loss": 0.655, | |
| "step": 4080 | |
| }, | |
| { | |
| "epoch": 15.014671814671814, | |
| "grad_norm": 9.246990203857422, | |
| "learning_rate": 7.601887601887602e-06, | |
| "loss": 0.3383, | |
| "step": 4090 | |
| }, | |
| { | |
| "epoch": 15.015444015444016, | |
| "grad_norm": 0.6101518273353577, | |
| "learning_rate": 7.593307593307594e-06, | |
| "loss": 0.7927, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 15.016216216216216, | |
| "grad_norm": 1.2248023748397827, | |
| "learning_rate": 7.584727584727586e-06, | |
| "loss": 0.3503, | |
| "step": 4110 | |
| }, | |
| { | |
| "epoch": 15.016988416988417, | |
| "grad_norm": 0.7881536483764648, | |
| "learning_rate": 7.576147576147577e-06, | |
| "loss": 0.2756, | |
| "step": 4120 | |
| }, | |
| { | |
| "epoch": 15.017760617760617, | |
| "grad_norm": 69.70541381835938, | |
| "learning_rate": 7.567567567567569e-06, | |
| "loss": 0.3127, | |
| "step": 4130 | |
| }, | |
| { | |
| "epoch": 15.018532818532819, | |
| "grad_norm": 0.18171033263206482, | |
| "learning_rate": 7.558987558987559e-06, | |
| "loss": 0.4698, | |
| "step": 4140 | |
| }, | |
| { | |
| "epoch": 15.019305019305019, | |
| "grad_norm": 134.1101531982422, | |
| "learning_rate": 7.550407550407551e-06, | |
| "loss": 0.3204, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 15.02007722007722, | |
| "grad_norm": 114.70227813720703, | |
| "learning_rate": 7.541827541827542e-06, | |
| "loss": 0.7975, | |
| "step": 4160 | |
| }, | |
| { | |
| "epoch": 15.02007722007722, | |
| "eval_accuracy": 0.6666666666666666, | |
| "eval_loss": 1.0846061706542969, | |
| "eval_runtime": 13.9438, | |
| "eval_samples_per_second": 3.227, | |
| "eval_steps_per_second": 1.076, | |
| "step": 4160 | |
| }, | |
| { | |
| "epoch": 16.0007722007722, | |
| "grad_norm": 11.493532180786133, | |
| "learning_rate": 7.533247533247534e-06, | |
| "loss": 0.8145, | |
| "step": 4170 | |
| }, | |
| { | |
| "epoch": 16.0015444015444, | |
| "grad_norm": 0.4535379409790039, | |
| "learning_rate": 7.524667524667525e-06, | |
| "loss": 0.3171, | |
| "step": 4180 | |
| }, | |
| { | |
| "epoch": 16.0023166023166, | |
| "grad_norm": 37.140682220458984, | |
| "learning_rate": 7.516087516087517e-06, | |
| "loss": 0.5947, | |
| "step": 4190 | |
| }, | |
| { | |
| "epoch": 16.003088803088804, | |
| "grad_norm": 0.18892982602119446, | |
| "learning_rate": 7.507507507507507e-06, | |
| "loss": 0.2023, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 16.003861003861005, | |
| "grad_norm": 0.2875947952270508, | |
| "learning_rate": 7.498927498927499e-06, | |
| "loss": 0.337, | |
| "step": 4210 | |
| }, | |
| { | |
| "epoch": 16.004633204633205, | |
| "grad_norm": 123.1329574584961, | |
| "learning_rate": 7.49034749034749e-06, | |
| "loss": 0.631, | |
| "step": 4220 | |
| }, | |
| { | |
| "epoch": 16.005405405405405, | |
| "grad_norm": 66.25767517089844, | |
| "learning_rate": 7.481767481767482e-06, | |
| "loss": 0.3625, | |
| "step": 4230 | |
| }, | |
| { | |
| "epoch": 16.006177606177605, | |
| "grad_norm": 31.93522071838379, | |
| "learning_rate": 7.473187473187474e-06, | |
| "loss": 0.5928, | |
| "step": 4240 | |
| }, | |
| { | |
| "epoch": 16.006949806949805, | |
| "grad_norm": 0.35881680250167847, | |
| "learning_rate": 7.464607464607465e-06, | |
| "loss": 0.7209, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 16.00772200772201, | |
| "grad_norm": 106.6112060546875, | |
| "learning_rate": 7.456027456027457e-06, | |
| "loss": 0.2844, | |
| "step": 4260 | |
| }, | |
| { | |
| "epoch": 16.00849420849421, | |
| "grad_norm": 0.10399631410837173, | |
| "learning_rate": 7.447447447447448e-06, | |
| "loss": 0.2302, | |
| "step": 4270 | |
| }, | |
| { | |
| "epoch": 16.00926640926641, | |
| "grad_norm": 10.694889068603516, | |
| "learning_rate": 7.4388674388674395e-06, | |
| "loss": 0.5468, | |
| "step": 4280 | |
| }, | |
| { | |
| "epoch": 16.01003861003861, | |
| "grad_norm": 15.967513084411621, | |
| "learning_rate": 7.430287430287431e-06, | |
| "loss": 0.8236, | |
| "step": 4290 | |
| }, | |
| { | |
| "epoch": 16.01081081081081, | |
| "grad_norm": 73.18832397460938, | |
| "learning_rate": 7.421707421707423e-06, | |
| "loss": 0.5327, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 16.01158301158301, | |
| "grad_norm": 17.472103118896484, | |
| "learning_rate": 7.413127413127414e-06, | |
| "loss": 0.4371, | |
| "step": 4310 | |
| }, | |
| { | |
| "epoch": 16.012355212355214, | |
| "grad_norm": 0.6622916460037231, | |
| "learning_rate": 7.404547404547406e-06, | |
| "loss": 0.1643, | |
| "step": 4320 | |
| }, | |
| { | |
| "epoch": 16.013127413127414, | |
| "grad_norm": 0.12445707619190216, | |
| "learning_rate": 7.395967395967396e-06, | |
| "loss": 0.037, | |
| "step": 4330 | |
| }, | |
| { | |
| "epoch": 16.013899613899614, | |
| "grad_norm": 68.56678009033203, | |
| "learning_rate": 7.387387387387388e-06, | |
| "loss": 0.3456, | |
| "step": 4340 | |
| }, | |
| { | |
| "epoch": 16.014671814671814, | |
| "grad_norm": 0.490728497505188, | |
| "learning_rate": 7.37880737880738e-06, | |
| "loss": 0.4691, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 16.015444015444015, | |
| "grad_norm": 0.11592353880405426, | |
| "learning_rate": 7.370227370227371e-06, | |
| "loss": 0.3818, | |
| "step": 4360 | |
| }, | |
| { | |
| "epoch": 16.016216216216215, | |
| "grad_norm": 0.130640909075737, | |
| "learning_rate": 7.361647361647363e-06, | |
| "loss": 0.404, | |
| "step": 4370 | |
| }, | |
| { | |
| "epoch": 16.01698841698842, | |
| "grad_norm": 1.6968755722045898, | |
| "learning_rate": 7.353067353067354e-06, | |
| "loss": 0.3531, | |
| "step": 4380 | |
| }, | |
| { | |
| "epoch": 16.01776061776062, | |
| "grad_norm": 0.15732808411121368, | |
| "learning_rate": 7.344487344487345e-06, | |
| "loss": 0.9261, | |
| "step": 4390 | |
| }, | |
| { | |
| "epoch": 16.01853281853282, | |
| "grad_norm": 0.8805422782897949, | |
| "learning_rate": 7.335907335907336e-06, | |
| "loss": 0.6919, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 16.01930501930502, | |
| "grad_norm": 0.9305623173713684, | |
| "learning_rate": 7.327327327327328e-06, | |
| "loss": 0.1012, | |
| "step": 4410 | |
| }, | |
| { | |
| "epoch": 16.02007722007722, | |
| "grad_norm": 118.25983428955078, | |
| "learning_rate": 7.318747318747319e-06, | |
| "loss": 0.5766, | |
| "step": 4420 | |
| }, | |
| { | |
| "epoch": 16.02007722007722, | |
| "eval_accuracy": 0.7333333333333333, | |
| "eval_loss": 0.9611772298812866, | |
| "eval_runtime": 11.7122, | |
| "eval_samples_per_second": 3.842, | |
| "eval_steps_per_second": 1.281, | |
| "step": 4420 | |
| }, | |
| { | |
| "epoch": 17.0007722007722, | |
| "grad_norm": 0.2061794251203537, | |
| "learning_rate": 7.310167310167311e-06, | |
| "loss": 0.6818, | |
| "step": 4430 | |
| }, | |
| { | |
| "epoch": 17.0015444015444, | |
| "grad_norm": 89.11824035644531, | |
| "learning_rate": 7.301587301587301e-06, | |
| "loss": 0.2661, | |
| "step": 4440 | |
| }, | |
| { | |
| "epoch": 17.0023166023166, | |
| "grad_norm": 1.2090812921524048, | |
| "learning_rate": 7.293007293007293e-06, | |
| "loss": 0.5385, | |
| "step": 4450 | |
| }, | |
| { | |
| "epoch": 17.003088803088804, | |
| "grad_norm": 10.95218276977539, | |
| "learning_rate": 7.2844272844272845e-06, | |
| "loss": 0.3814, | |
| "step": 4460 | |
| }, | |
| { | |
| "epoch": 17.003861003861005, | |
| "grad_norm": 0.34872862696647644, | |
| "learning_rate": 7.275847275847276e-06, | |
| "loss": 0.319, | |
| "step": 4470 | |
| }, | |
| { | |
| "epoch": 17.004633204633205, | |
| "grad_norm": 0.4813756048679352, | |
| "learning_rate": 7.267267267267268e-06, | |
| "loss": 0.526, | |
| "step": 4480 | |
| }, | |
| { | |
| "epoch": 17.005405405405405, | |
| "grad_norm": 0.2315344214439392, | |
| "learning_rate": 7.2586872586872595e-06, | |
| "loss": 0.4553, | |
| "step": 4490 | |
| }, | |
| { | |
| "epoch": 17.006177606177605, | |
| "grad_norm": 19.774232864379883, | |
| "learning_rate": 7.250107250107251e-06, | |
| "loss": 0.2821, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 17.006949806949805, | |
| "grad_norm": 0.24049396812915802, | |
| "learning_rate": 7.241527241527242e-06, | |
| "loss": 0.4244, | |
| "step": 4510 | |
| }, | |
| { | |
| "epoch": 17.00772200772201, | |
| "grad_norm": 0.1969604790210724, | |
| "learning_rate": 7.232947232947234e-06, | |
| "loss": 0.1505, | |
| "step": 4520 | |
| }, | |
| { | |
| "epoch": 17.00849420849421, | |
| "grad_norm": 61.09734344482422, | |
| "learning_rate": 7.224367224367225e-06, | |
| "loss": 0.542, | |
| "step": 4530 | |
| }, | |
| { | |
| "epoch": 17.00926640926641, | |
| "grad_norm": 1.5801252126693726, | |
| "learning_rate": 7.215787215787217e-06, | |
| "loss": 0.2881, | |
| "step": 4540 | |
| }, | |
| { | |
| "epoch": 17.01003861003861, | |
| "grad_norm": 0.6441929936408997, | |
| "learning_rate": 7.207207207207208e-06, | |
| "loss": 0.8682, | |
| "step": 4550 | |
| }, | |
| { | |
| "epoch": 17.01081081081081, | |
| "grad_norm": 0.10384261608123779, | |
| "learning_rate": 7.1986271986272e-06, | |
| "loss": 0.314, | |
| "step": 4560 | |
| }, | |
| { | |
| "epoch": 17.01158301158301, | |
| "grad_norm": 1.5747077465057373, | |
| "learning_rate": 7.19004719004719e-06, | |
| "loss": 0.1259, | |
| "step": 4570 | |
| }, | |
| { | |
| "epoch": 17.012355212355214, | |
| "grad_norm": 47.53998947143555, | |
| "learning_rate": 7.181467181467182e-06, | |
| "loss": 0.7676, | |
| "step": 4580 | |
| }, | |
| { | |
| "epoch": 17.013127413127414, | |
| "grad_norm": 0.08766768872737885, | |
| "learning_rate": 7.172887172887173e-06, | |
| "loss": 0.1305, | |
| "step": 4590 | |
| }, | |
| { | |
| "epoch": 17.013899613899614, | |
| "grad_norm": 2.643843650817871, | |
| "learning_rate": 7.164307164307165e-06, | |
| "loss": 0.3499, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 17.014671814671814, | |
| "grad_norm": 0.15104711055755615, | |
| "learning_rate": 7.155727155727157e-06, | |
| "loss": 0.1806, | |
| "step": 4610 | |
| }, | |
| { | |
| "epoch": 17.015444015444015, | |
| "grad_norm": 0.07185631990432739, | |
| "learning_rate": 7.147147147147148e-06, | |
| "loss": 0.178, | |
| "step": 4620 | |
| }, | |
| { | |
| "epoch": 17.016216216216215, | |
| "grad_norm": 12.274628639221191, | |
| "learning_rate": 7.13856713856714e-06, | |
| "loss": 0.8695, | |
| "step": 4630 | |
| }, | |
| { | |
| "epoch": 17.01698841698842, | |
| "grad_norm": 234.01995849609375, | |
| "learning_rate": 7.12998712998713e-06, | |
| "loss": 0.7621, | |
| "step": 4640 | |
| }, | |
| { | |
| "epoch": 17.01776061776062, | |
| "grad_norm": 0.27887094020843506, | |
| "learning_rate": 7.121407121407122e-06, | |
| "loss": 0.7544, | |
| "step": 4650 | |
| }, | |
| { | |
| "epoch": 17.01853281853282, | |
| "grad_norm": 22.927310943603516, | |
| "learning_rate": 7.112827112827113e-06, | |
| "loss": 0.4503, | |
| "step": 4660 | |
| }, | |
| { | |
| "epoch": 17.01930501930502, | |
| "grad_norm": 0.3981897234916687, | |
| "learning_rate": 7.104247104247105e-06, | |
| "loss": 0.8743, | |
| "step": 4670 | |
| }, | |
| { | |
| "epoch": 17.02007722007722, | |
| "grad_norm": 0.04332108423113823, | |
| "learning_rate": 7.0956670956670955e-06, | |
| "loss": 0.3323, | |
| "step": 4680 | |
| }, | |
| { | |
| "epoch": 17.02007722007722, | |
| "eval_accuracy": 0.6888888888888889, | |
| "eval_loss": 1.1610815525054932, | |
| "eval_runtime": 13.5549, | |
| "eval_samples_per_second": 3.32, | |
| "eval_steps_per_second": 1.107, | |
| "step": 4680 | |
| }, | |
| { | |
| "epoch": 18.0007722007722, | |
| "grad_norm": 0.15380901098251343, | |
| "learning_rate": 7.087087087087087e-06, | |
| "loss": 0.5581, | |
| "step": 4690 | |
| }, | |
| { | |
| "epoch": 18.0015444015444, | |
| "grad_norm": 3.311131238937378, | |
| "learning_rate": 7.0785070785070785e-06, | |
| "loss": 0.5238, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 18.0023166023166, | |
| "grad_norm": 0.010178559459745884, | |
| "learning_rate": 7.0699270699270705e-06, | |
| "loss": 0.438, | |
| "step": 4710 | |
| }, | |
| { | |
| "epoch": 18.003088803088804, | |
| "grad_norm": 0.11447062343358994, | |
| "learning_rate": 7.0613470613470616e-06, | |
| "loss": 0.2677, | |
| "step": 4720 | |
| }, | |
| { | |
| "epoch": 18.003861003861005, | |
| "grad_norm": 35.86017608642578, | |
| "learning_rate": 7.0527670527670535e-06, | |
| "loss": 0.4821, | |
| "step": 4730 | |
| }, | |
| { | |
| "epoch": 18.004633204633205, | |
| "grad_norm": 0.34406280517578125, | |
| "learning_rate": 7.0441870441870455e-06, | |
| "loss": 0.0562, | |
| "step": 4740 | |
| }, | |
| { | |
| "epoch": 18.005405405405405, | |
| "grad_norm": 0.14499299228191376, | |
| "learning_rate": 7.035607035607036e-06, | |
| "loss": 0.5351, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 18.006177606177605, | |
| "grad_norm": 0.05108552426099777, | |
| "learning_rate": 7.027027027027028e-06, | |
| "loss": 0.3214, | |
| "step": 4760 | |
| }, | |
| { | |
| "epoch": 18.006949806949805, | |
| "grad_norm": 0.02121145650744438, | |
| "learning_rate": 7.018447018447019e-06, | |
| "loss": 0.5145, | |
| "step": 4770 | |
| }, | |
| { | |
| "epoch": 18.00772200772201, | |
| "grad_norm": 0.0515630841255188, | |
| "learning_rate": 7.009867009867011e-06, | |
| "loss": 0.281, | |
| "step": 4780 | |
| }, | |
| { | |
| "epoch": 18.00849420849421, | |
| "grad_norm": 0.12710994482040405, | |
| "learning_rate": 7.001287001287002e-06, | |
| "loss": 0.8898, | |
| "step": 4790 | |
| }, | |
| { | |
| "epoch": 18.00926640926641, | |
| "grad_norm": 0.12529367208480835, | |
| "learning_rate": 6.992706992706994e-06, | |
| "loss": 0.6286, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 18.01003861003861, | |
| "grad_norm": 42.525726318359375, | |
| "learning_rate": 6.984126984126984e-06, | |
| "loss": 0.2252, | |
| "step": 4810 | |
| }, | |
| { | |
| "epoch": 18.01081081081081, | |
| "grad_norm": 0.06859328597784042, | |
| "learning_rate": 6.975546975546976e-06, | |
| "loss": 0.2308, | |
| "step": 4820 | |
| }, | |
| { | |
| "epoch": 18.01158301158301, | |
| "grad_norm": 21.27465057373047, | |
| "learning_rate": 6.966966966966967e-06, | |
| "loss": 0.5888, | |
| "step": 4830 | |
| }, | |
| { | |
| "epoch": 18.012355212355214, | |
| "grad_norm": 0.022731417790055275, | |
| "learning_rate": 6.958386958386959e-06, | |
| "loss": 0.5071, | |
| "step": 4840 | |
| }, | |
| { | |
| "epoch": 18.013127413127414, | |
| "grad_norm": 0.6973882913589478, | |
| "learning_rate": 6.949806949806951e-06, | |
| "loss": 0.1836, | |
| "step": 4850 | |
| }, | |
| { | |
| "epoch": 18.013899613899614, | |
| "grad_norm": 0.375298410654068, | |
| "learning_rate": 6.941226941226942e-06, | |
| "loss": 0.568, | |
| "step": 4860 | |
| }, | |
| { | |
| "epoch": 18.014671814671814, | |
| "grad_norm": 0.13478554785251617, | |
| "learning_rate": 6.932646932646934e-06, | |
| "loss": 0.2527, | |
| "step": 4870 | |
| }, | |
| { | |
| "epoch": 18.015444015444015, | |
| "grad_norm": 77.43624877929688, | |
| "learning_rate": 6.924066924066924e-06, | |
| "loss": 0.338, | |
| "step": 4880 | |
| }, | |
| { | |
| "epoch": 18.016216216216215, | |
| "grad_norm": 33.197998046875, | |
| "learning_rate": 6.915486915486916e-06, | |
| "loss": 0.4993, | |
| "step": 4890 | |
| }, | |
| { | |
| "epoch": 18.01698841698842, | |
| "grad_norm": 1.216697096824646, | |
| "learning_rate": 6.906906906906907e-06, | |
| "loss": 0.502, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 18.01776061776062, | |
| "grad_norm": 15.660502433776855, | |
| "learning_rate": 6.898326898326899e-06, | |
| "loss": 0.6187, | |
| "step": 4910 | |
| }, | |
| { | |
| "epoch": 18.01853281853282, | |
| "grad_norm": 0.01838286779820919, | |
| "learning_rate": 6.88974688974689e-06, | |
| "loss": 0.3636, | |
| "step": 4920 | |
| }, | |
| { | |
| "epoch": 18.01930501930502, | |
| "grad_norm": 157.7162628173828, | |
| "learning_rate": 6.881166881166882e-06, | |
| "loss": 0.1696, | |
| "step": 4930 | |
| }, | |
| { | |
| "epoch": 18.02007722007722, | |
| "grad_norm": 0.8771587610244751, | |
| "learning_rate": 6.872586872586873e-06, | |
| "loss": 0.7162, | |
| "step": 4940 | |
| }, | |
| { | |
| "epoch": 18.02007722007722, | |
| "eval_accuracy": 0.7111111111111111, | |
| "eval_loss": 1.30552339553833, | |
| "eval_runtime": 12.9627, | |
| "eval_samples_per_second": 3.472, | |
| "eval_steps_per_second": 1.157, | |
| "step": 4940 | |
| }, | |
| { | |
| "epoch": 19.0007722007722, | |
| "grad_norm": 0.13848955929279327, | |
| "learning_rate": 6.8640068640068645e-06, | |
| "loss": 0.2741, | |
| "step": 4950 | |
| }, | |
| { | |
| "epoch": 19.0015444015444, | |
| "grad_norm": 0.023964807391166687, | |
| "learning_rate": 6.855426855426856e-06, | |
| "loss": 0.2638, | |
| "step": 4960 | |
| }, | |
| { | |
| "epoch": 19.0023166023166, | |
| "grad_norm": 4.055501461029053, | |
| "learning_rate": 6.846846846846848e-06, | |
| "loss": 0.3041, | |
| "step": 4970 | |
| }, | |
| { | |
| "epoch": 19.003088803088804, | |
| "grad_norm": 0.14882507920265198, | |
| "learning_rate": 6.8382668382668395e-06, | |
| "loss": 0.3066, | |
| "step": 4980 | |
| }, | |
| { | |
| "epoch": 19.003861003861005, | |
| "grad_norm": 1.0661811828613281, | |
| "learning_rate": 6.82968682968683e-06, | |
| "loss": 0.1638, | |
| "step": 4990 | |
| }, | |
| { | |
| "epoch": 19.004633204633205, | |
| "grad_norm": 0.04383436217904091, | |
| "learning_rate": 6.821106821106822e-06, | |
| "loss": 0.3109, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 19.005405405405405, | |
| "grad_norm": 43.136775970458984, | |
| "learning_rate": 6.812526812526813e-06, | |
| "loss": 0.7934, | |
| "step": 5010 | |
| }, | |
| { | |
| "epoch": 19.006177606177605, | |
| "grad_norm": 0.18534782528877258, | |
| "learning_rate": 6.803946803946805e-06, | |
| "loss": 0.4056, | |
| "step": 5020 | |
| }, | |
| { | |
| "epoch": 19.006949806949805, | |
| "grad_norm": 4.773721218109131, | |
| "learning_rate": 6.795366795366796e-06, | |
| "loss": 0.3571, | |
| "step": 5030 | |
| }, | |
| { | |
| "epoch": 19.00772200772201, | |
| "grad_norm": 117.23564910888672, | |
| "learning_rate": 6.786786786786788e-06, | |
| "loss": 0.1148, | |
| "step": 5040 | |
| }, | |
| { | |
| "epoch": 19.00849420849421, | |
| "grad_norm": 0.11472011357545853, | |
| "learning_rate": 6.778206778206778e-06, | |
| "loss": 0.3655, | |
| "step": 5050 | |
| }, | |
| { | |
| "epoch": 19.00926640926641, | |
| "grad_norm": 0.10049240291118622, | |
| "learning_rate": 6.76962676962677e-06, | |
| "loss": 0.0786, | |
| "step": 5060 | |
| }, | |
| { | |
| "epoch": 19.01003861003861, | |
| "grad_norm": 1.114931344985962, | |
| "learning_rate": 6.761046761046761e-06, | |
| "loss": 0.5683, | |
| "step": 5070 | |
| }, | |
| { | |
| "epoch": 19.01081081081081, | |
| "grad_norm": 0.0499207079410553, | |
| "learning_rate": 6.752466752466753e-06, | |
| "loss": 0.1673, | |
| "step": 5080 | |
| }, | |
| { | |
| "epoch": 19.01158301158301, | |
| "grad_norm": 0.04504622146487236, | |
| "learning_rate": 6.743886743886744e-06, | |
| "loss": 0.0325, | |
| "step": 5090 | |
| }, | |
| { | |
| "epoch": 19.012355212355214, | |
| "grad_norm": 24.717525482177734, | |
| "learning_rate": 6.735306735306736e-06, | |
| "loss": 0.2299, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 19.013127413127414, | |
| "grad_norm": 0.004078052006661892, | |
| "learning_rate": 6.726726726726728e-06, | |
| "loss": 0.4058, | |
| "step": 5110 | |
| }, | |
| { | |
| "epoch": 19.013899613899614, | |
| "grad_norm": 28.717302322387695, | |
| "learning_rate": 6.718146718146718e-06, | |
| "loss": 1.0464, | |
| "step": 5120 | |
| }, | |
| { | |
| "epoch": 19.014671814671814, | |
| "grad_norm": 0.1303737759590149, | |
| "learning_rate": 6.70956670956671e-06, | |
| "loss": 0.053, | |
| "step": 5130 | |
| }, | |
| { | |
| "epoch": 19.015444015444015, | |
| "grad_norm": 23.893720626831055, | |
| "learning_rate": 6.700986700986701e-06, | |
| "loss": 0.8631, | |
| "step": 5140 | |
| }, | |
| { | |
| "epoch": 19.016216216216215, | |
| "grad_norm": 0.1701466143131256, | |
| "learning_rate": 6.692406692406693e-06, | |
| "loss": 0.3543, | |
| "step": 5150 | |
| }, | |
| { | |
| "epoch": 19.01698841698842, | |
| "grad_norm": 118.48575592041016, | |
| "learning_rate": 6.6838266838266844e-06, | |
| "loss": 0.5097, | |
| "step": 5160 | |
| }, | |
| { | |
| "epoch": 19.01776061776062, | |
| "grad_norm": 58.45772933959961, | |
| "learning_rate": 6.675246675246676e-06, | |
| "loss": 0.183, | |
| "step": 5170 | |
| }, | |
| { | |
| "epoch": 19.01853281853282, | |
| "grad_norm": 0.15131786465644836, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 0.1063, | |
| "step": 5180 | |
| }, | |
| { | |
| "epoch": 19.01930501930502, | |
| "grad_norm": 0.42406681180000305, | |
| "learning_rate": 6.658086658086659e-06, | |
| "loss": 0.7002, | |
| "step": 5190 | |
| }, | |
| { | |
| "epoch": 19.02007722007722, | |
| "grad_norm": 0.004914190154522657, | |
| "learning_rate": 6.64950664950665e-06, | |
| "loss": 0.1248, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 19.02007722007722, | |
| "eval_accuracy": 0.6666666666666666, | |
| "eval_loss": 1.5836811065673828, | |
| "eval_runtime": 12.2459, | |
| "eval_samples_per_second": 3.675, | |
| "eval_steps_per_second": 1.225, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 20.0007722007722, | |
| "grad_norm": 113.19306182861328, | |
| "learning_rate": 6.640926640926642e-06, | |
| "loss": 0.2273, | |
| "step": 5210 | |
| }, | |
| { | |
| "epoch": 20.0015444015444, | |
| "grad_norm": 0.14228223264217377, | |
| "learning_rate": 6.632346632346633e-06, | |
| "loss": 0.2361, | |
| "step": 5220 | |
| }, | |
| { | |
| "epoch": 20.0023166023166, | |
| "grad_norm": 0.030990980565547943, | |
| "learning_rate": 6.623766623766624e-06, | |
| "loss": 0.3815, | |
| "step": 5230 | |
| }, | |
| { | |
| "epoch": 20.003088803088804, | |
| "grad_norm": 67.11844635009766, | |
| "learning_rate": 6.615186615186616e-06, | |
| "loss": 0.1815, | |
| "step": 5240 | |
| }, | |
| { | |
| "epoch": 20.003861003861005, | |
| "grad_norm": 1.2835884094238281, | |
| "learning_rate": 6.606606606606607e-06, | |
| "loss": 0.1885, | |
| "step": 5250 | |
| }, | |
| { | |
| "epoch": 20.004633204633205, | |
| "grad_norm": 182.08621215820312, | |
| "learning_rate": 6.598026598026599e-06, | |
| "loss": 0.2653, | |
| "step": 5260 | |
| }, | |
| { | |
| "epoch": 20.005405405405405, | |
| "grad_norm": 0.08672849088907242, | |
| "learning_rate": 6.58944658944659e-06, | |
| "loss": 0.323, | |
| "step": 5270 | |
| }, | |
| { | |
| "epoch": 20.006177606177605, | |
| "grad_norm": 0.03671255707740784, | |
| "learning_rate": 6.580866580866582e-06, | |
| "loss": 0.0026, | |
| "step": 5280 | |
| }, | |
| { | |
| "epoch": 20.006949806949805, | |
| "grad_norm": 12.600577354431152, | |
| "learning_rate": 6.572286572286572e-06, | |
| "loss": 0.3475, | |
| "step": 5290 | |
| }, | |
| { | |
| "epoch": 20.00772200772201, | |
| "grad_norm": 14.476115226745605, | |
| "learning_rate": 6.563706563706564e-06, | |
| "loss": 0.8815, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 20.00849420849421, | |
| "grad_norm": 0.035546738654375076, | |
| "learning_rate": 6.555126555126555e-06, | |
| "loss": 0.5063, | |
| "step": 5310 | |
| }, | |
| { | |
| "epoch": 20.00926640926641, | |
| "grad_norm": 0.6188809275627136, | |
| "learning_rate": 6.546546546546547e-06, | |
| "loss": 0.3486, | |
| "step": 5320 | |
| }, | |
| { | |
| "epoch": 20.01003861003861, | |
| "grad_norm": 0.2754557728767395, | |
| "learning_rate": 6.537966537966538e-06, | |
| "loss": 0.1505, | |
| "step": 5330 | |
| }, | |
| { | |
| "epoch": 20.01081081081081, | |
| "grad_norm": 25.426105499267578, | |
| "learning_rate": 6.52938652938653e-06, | |
| "loss": 0.5716, | |
| "step": 5340 | |
| }, | |
| { | |
| "epoch": 20.01158301158301, | |
| "grad_norm": 0.03355805203318596, | |
| "learning_rate": 6.520806520806522e-06, | |
| "loss": 0.3115, | |
| "step": 5350 | |
| }, | |
| { | |
| "epoch": 20.012355212355214, | |
| "grad_norm": 0.5074798464775085, | |
| "learning_rate": 6.512226512226512e-06, | |
| "loss": 0.3633, | |
| "step": 5360 | |
| }, | |
| { | |
| "epoch": 20.013127413127414, | |
| "grad_norm": 5.739364147186279, | |
| "learning_rate": 6.503646503646504e-06, | |
| "loss": 0.5785, | |
| "step": 5370 | |
| }, | |
| { | |
| "epoch": 20.013899613899614, | |
| "grad_norm": 7.886667728424072, | |
| "learning_rate": 6.4950664950664955e-06, | |
| "loss": 0.8187, | |
| "step": 5380 | |
| }, | |
| { | |
| "epoch": 20.014671814671814, | |
| "grad_norm": 0.18780578672885895, | |
| "learning_rate": 6.486486486486487e-06, | |
| "loss": 0.7403, | |
| "step": 5390 | |
| }, | |
| { | |
| "epoch": 20.015444015444015, | |
| "grad_norm": 197.27720642089844, | |
| "learning_rate": 6.4779064779064785e-06, | |
| "loss": 0.1305, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 20.016216216216215, | |
| "grad_norm": 107.21308135986328, | |
| "learning_rate": 6.4693264693264705e-06, | |
| "loss": 0.0233, | |
| "step": 5410 | |
| }, | |
| { | |
| "epoch": 20.01698841698842, | |
| "grad_norm": 15.643366813659668, | |
| "learning_rate": 6.460746460746461e-06, | |
| "loss": 0.4351, | |
| "step": 5420 | |
| }, | |
| { | |
| "epoch": 20.01776061776062, | |
| "grad_norm": 0.6989626884460449, | |
| "learning_rate": 6.452166452166453e-06, | |
| "loss": 0.1247, | |
| "step": 5430 | |
| }, | |
| { | |
| "epoch": 20.01853281853282, | |
| "grad_norm": 404.39361572265625, | |
| "learning_rate": 6.443586443586444e-06, | |
| "loss": 0.1101, | |
| "step": 5440 | |
| }, | |
| { | |
| "epoch": 20.01930501930502, | |
| "grad_norm": 0.03856265917420387, | |
| "learning_rate": 6.435006435006436e-06, | |
| "loss": 0.2845, | |
| "step": 5450 | |
| }, | |
| { | |
| "epoch": 20.02007722007722, | |
| "grad_norm": 0.12517881393432617, | |
| "learning_rate": 6.426426426426427e-06, | |
| "loss": 0.1746, | |
| "step": 5460 | |
| }, | |
| { | |
| "epoch": 20.02007722007722, | |
| "eval_accuracy": 0.7333333333333333, | |
| "eval_loss": 1.2694180011749268, | |
| "eval_runtime": 12.6449, | |
| "eval_samples_per_second": 3.559, | |
| "eval_steps_per_second": 1.186, | |
| "step": 5460 | |
| }, | |
| { | |
| "epoch": 21.0007722007722, | |
| "grad_norm": 0.5627447962760925, | |
| "learning_rate": 6.417846417846419e-06, | |
| "loss": 0.5971, | |
| "step": 5470 | |
| }, | |
| { | |
| "epoch": 21.0015444015444, | |
| "grad_norm": 209.05210876464844, | |
| "learning_rate": 6.409266409266411e-06, | |
| "loss": 1.1831, | |
| "step": 5480 | |
| }, | |
| { | |
| "epoch": 21.0023166023166, | |
| "grad_norm": 0.19635902345180511, | |
| "learning_rate": 6.400686400686401e-06, | |
| "loss": 0.5821, | |
| "step": 5490 | |
| }, | |
| { | |
| "epoch": 21.003088803088804, | |
| "grad_norm": 0.13137683272361755, | |
| "learning_rate": 6.392106392106393e-06, | |
| "loss": 0.557, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 21.003861003861005, | |
| "grad_norm": 0.09237836301326752, | |
| "learning_rate": 6.383526383526384e-06, | |
| "loss": 0.2781, | |
| "step": 5510 | |
| }, | |
| { | |
| "epoch": 21.004633204633205, | |
| "grad_norm": 0.08199318498373032, | |
| "learning_rate": 6.374946374946376e-06, | |
| "loss": 0.0898, | |
| "step": 5520 | |
| }, | |
| { | |
| "epoch": 21.005405405405405, | |
| "grad_norm": 54.22182846069336, | |
| "learning_rate": 6.366366366366366e-06, | |
| "loss": 0.2855, | |
| "step": 5530 | |
| }, | |
| { | |
| "epoch": 21.006177606177605, | |
| "grad_norm": 0.5029569268226624, | |
| "learning_rate": 6.357786357786358e-06, | |
| "loss": 0.0304, | |
| "step": 5540 | |
| }, | |
| { | |
| "epoch": 21.006949806949805, | |
| "grad_norm": 177.83555603027344, | |
| "learning_rate": 6.349206349206349e-06, | |
| "loss": 0.3167, | |
| "step": 5550 | |
| }, | |
| { | |
| "epoch": 21.00772200772201, | |
| "grad_norm": 0.42930805683135986, | |
| "learning_rate": 6.340626340626341e-06, | |
| "loss": 0.2099, | |
| "step": 5560 | |
| }, | |
| { | |
| "epoch": 21.00849420849421, | |
| "grad_norm": 0.050365786999464035, | |
| "learning_rate": 6.332046332046332e-06, | |
| "loss": 0.3829, | |
| "step": 5570 | |
| }, | |
| { | |
| "epoch": 21.00926640926641, | |
| "grad_norm": 0.05553288757801056, | |
| "learning_rate": 6.323466323466324e-06, | |
| "loss": 0.8343, | |
| "step": 5580 | |
| }, | |
| { | |
| "epoch": 21.01003861003861, | |
| "grad_norm": 0.05720970034599304, | |
| "learning_rate": 6.3148863148863145e-06, | |
| "loss": 0.1574, | |
| "step": 5590 | |
| }, | |
| { | |
| "epoch": 21.01081081081081, | |
| "grad_norm": 2.832240104675293, | |
| "learning_rate": 6.3063063063063065e-06, | |
| "loss": 0.409, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 21.01158301158301, | |
| "grad_norm": 158.55467224121094, | |
| "learning_rate": 6.297726297726298e-06, | |
| "loss": 0.5644, | |
| "step": 5610 | |
| }, | |
| { | |
| "epoch": 21.012355212355214, | |
| "grad_norm": 0.13720889389514923, | |
| "learning_rate": 6.2891462891462895e-06, | |
| "loss": 0.6388, | |
| "step": 5620 | |
| }, | |
| { | |
| "epoch": 21.013127413127414, | |
| "grad_norm": 64.92716217041016, | |
| "learning_rate": 6.2805662805662815e-06, | |
| "loss": 0.6628, | |
| "step": 5630 | |
| }, | |
| { | |
| "epoch": 21.013899613899614, | |
| "grad_norm": 17.310565948486328, | |
| "learning_rate": 6.2719862719862726e-06, | |
| "loss": 0.5685, | |
| "step": 5640 | |
| }, | |
| { | |
| "epoch": 21.014671814671814, | |
| "grad_norm": 0.2614205479621887, | |
| "learning_rate": 6.2634062634062645e-06, | |
| "loss": 0.0937, | |
| "step": 5650 | |
| }, | |
| { | |
| "epoch": 21.015444015444015, | |
| "grad_norm": 11.924259185791016, | |
| "learning_rate": 6.254826254826255e-06, | |
| "loss": 0.3255, | |
| "step": 5660 | |
| }, | |
| { | |
| "epoch": 21.016216216216215, | |
| "grad_norm": 0.06924501061439514, | |
| "learning_rate": 6.246246246246247e-06, | |
| "loss": 0.2687, | |
| "step": 5670 | |
| }, | |
| { | |
| "epoch": 21.01698841698842, | |
| "grad_norm": 0.0833396166563034, | |
| "learning_rate": 6.237666237666238e-06, | |
| "loss": 0.1404, | |
| "step": 5680 | |
| }, | |
| { | |
| "epoch": 21.01776061776062, | |
| "grad_norm": 0.17966850101947784, | |
| "learning_rate": 6.22908622908623e-06, | |
| "loss": 0.1596, | |
| "step": 5690 | |
| }, | |
| { | |
| "epoch": 21.01853281853282, | |
| "grad_norm": 0.0691114291548729, | |
| "learning_rate": 6.220506220506221e-06, | |
| "loss": 0.2709, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 21.01930501930502, | |
| "grad_norm": 41.90225601196289, | |
| "learning_rate": 6.211926211926213e-06, | |
| "loss": 0.4854, | |
| "step": 5710 | |
| }, | |
| { | |
| "epoch": 21.02007722007722, | |
| "grad_norm": 6.455798149108887, | |
| "learning_rate": 6.203346203346203e-06, | |
| "loss": 0.1973, | |
| "step": 5720 | |
| }, | |
| { | |
| "epoch": 21.02007722007722, | |
| "eval_accuracy": 0.7777777777777778, | |
| "eval_loss": 1.1572198867797852, | |
| "eval_runtime": 13.795, | |
| "eval_samples_per_second": 3.262, | |
| "eval_steps_per_second": 1.087, | |
| "step": 5720 | |
| }, | |
| { | |
| "epoch": 22.0007722007722, | |
| "grad_norm": 3.293867349624634, | |
| "learning_rate": 6.194766194766195e-06, | |
| "loss": 0.2579, | |
| "step": 5730 | |
| }, | |
| { | |
| "epoch": 22.0015444015444, | |
| "grad_norm": 0.11658459901809692, | |
| "learning_rate": 6.186186186186187e-06, | |
| "loss": 0.4474, | |
| "step": 5740 | |
| }, | |
| { | |
| "epoch": 22.0023166023166, | |
| "grad_norm": 0.0438237190246582, | |
| "learning_rate": 6.177606177606178e-06, | |
| "loss": 0.2652, | |
| "step": 5750 | |
| }, | |
| { | |
| "epoch": 22.003088803088804, | |
| "grad_norm": 182.23719787597656, | |
| "learning_rate": 6.16902616902617e-06, | |
| "loss": 0.6564, | |
| "step": 5760 | |
| }, | |
| { | |
| "epoch": 22.003861003861005, | |
| "grad_norm": 10.42225170135498, | |
| "learning_rate": 6.160446160446161e-06, | |
| "loss": 0.3219, | |
| "step": 5770 | |
| }, | |
| { | |
| "epoch": 22.004633204633205, | |
| "grad_norm": 68.25077819824219, | |
| "learning_rate": 6.151866151866153e-06, | |
| "loss": 0.5195, | |
| "step": 5780 | |
| }, | |
| { | |
| "epoch": 22.005405405405405, | |
| "grad_norm": 0.22111119329929352, | |
| "learning_rate": 6.143286143286143e-06, | |
| "loss": 0.1278, | |
| "step": 5790 | |
| }, | |
| { | |
| "epoch": 22.006177606177605, | |
| "grad_norm": 119.94930267333984, | |
| "learning_rate": 6.134706134706135e-06, | |
| "loss": 0.3084, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 22.006949806949805, | |
| "grad_norm": 0.29699742794036865, | |
| "learning_rate": 6.126126126126126e-06, | |
| "loss": 0.276, | |
| "step": 5810 | |
| }, | |
| { | |
| "epoch": 22.00772200772201, | |
| "grad_norm": 0.22920526564121246, | |
| "learning_rate": 6.117546117546118e-06, | |
| "loss": 0.3415, | |
| "step": 5820 | |
| }, | |
| { | |
| "epoch": 22.00849420849421, | |
| "grad_norm": 93.11829376220703, | |
| "learning_rate": 6.108966108966109e-06, | |
| "loss": 0.4063, | |
| "step": 5830 | |
| }, | |
| { | |
| "epoch": 22.00926640926641, | |
| "grad_norm": 0.09231970459222794, | |
| "learning_rate": 6.1003861003861005e-06, | |
| "loss": 0.2913, | |
| "step": 5840 | |
| }, | |
| { | |
| "epoch": 22.01003861003861, | |
| "grad_norm": 0.13953036069869995, | |
| "learning_rate": 6.0918060918060925e-06, | |
| "loss": 0.2231, | |
| "step": 5850 | |
| }, | |
| { | |
| "epoch": 22.01081081081081, | |
| "grad_norm": 9.176348686218262, | |
| "learning_rate": 6.083226083226084e-06, | |
| "loss": 0.1031, | |
| "step": 5860 | |
| }, | |
| { | |
| "epoch": 22.01158301158301, | |
| "grad_norm": 0.002297205151990056, | |
| "learning_rate": 6.0746460746460755e-06, | |
| "loss": 0.3377, | |
| "step": 5870 | |
| }, | |
| { | |
| "epoch": 22.012355212355214, | |
| "grad_norm": 0.011434354819357395, | |
| "learning_rate": 6.066066066066067e-06, | |
| "loss": 0.1273, | |
| "step": 5880 | |
| }, | |
| { | |
| "epoch": 22.013127413127414, | |
| "grad_norm": 0.0924556627869606, | |
| "learning_rate": 6.057486057486059e-06, | |
| "loss": 0.1457, | |
| "step": 5890 | |
| }, | |
| { | |
| "epoch": 22.013899613899614, | |
| "grad_norm": 0.3709995448589325, | |
| "learning_rate": 6.048906048906049e-06, | |
| "loss": 0.126, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 22.014671814671814, | |
| "grad_norm": 0.08549296110868454, | |
| "learning_rate": 6.040326040326041e-06, | |
| "loss": 0.1878, | |
| "step": 5910 | |
| }, | |
| { | |
| "epoch": 22.015444015444015, | |
| "grad_norm": 100.20759582519531, | |
| "learning_rate": 6.031746031746032e-06, | |
| "loss": 0.5293, | |
| "step": 5920 | |
| }, | |
| { | |
| "epoch": 22.016216216216215, | |
| "grad_norm": 0.04931352287530899, | |
| "learning_rate": 6.023166023166024e-06, | |
| "loss": 0.2636, | |
| "step": 5930 | |
| }, | |
| { | |
| "epoch": 22.01698841698842, | |
| "grad_norm": 0.4091545343399048, | |
| "learning_rate": 6.014586014586015e-06, | |
| "loss": 0.4039, | |
| "step": 5940 | |
| }, | |
| { | |
| "epoch": 22.01776061776062, | |
| "grad_norm": 0.03857075050473213, | |
| "learning_rate": 6.006006006006007e-06, | |
| "loss": 0.387, | |
| "step": 5950 | |
| }, | |
| { | |
| "epoch": 22.01853281853282, | |
| "grad_norm": 91.95096588134766, | |
| "learning_rate": 5.997425997425997e-06, | |
| "loss": 0.3726, | |
| "step": 5960 | |
| }, | |
| { | |
| "epoch": 22.01930501930502, | |
| "grad_norm": 0.522257387638092, | |
| "learning_rate": 5.988845988845989e-06, | |
| "loss": 0.256, | |
| "step": 5970 | |
| }, | |
| { | |
| "epoch": 22.02007722007722, | |
| "grad_norm": 0.06971725821495056, | |
| "learning_rate": 5.980265980265981e-06, | |
| "loss": 0.7038, | |
| "step": 5980 | |
| }, | |
| { | |
| "epoch": 22.02007722007722, | |
| "eval_accuracy": 0.7111111111111111, | |
| "eval_loss": 1.4034613370895386, | |
| "eval_runtime": 13.0086, | |
| "eval_samples_per_second": 3.459, | |
| "eval_steps_per_second": 1.153, | |
| "step": 5980 | |
| }, | |
| { | |
| "epoch": 23.0007722007722, | |
| "grad_norm": 0.025312229990959167, | |
| "learning_rate": 5.971685971685972e-06, | |
| "loss": 0.2679, | |
| "step": 5990 | |
| }, | |
| { | |
| "epoch": 23.0015444015444, | |
| "grad_norm": 4.412787437438965, | |
| "learning_rate": 5.963105963105964e-06, | |
| "loss": 1.0691, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 23.0023166023166, | |
| "grad_norm": 90.64250183105469, | |
| "learning_rate": 5.954525954525955e-06, | |
| "loss": 0.5154, | |
| "step": 6010 | |
| }, | |
| { | |
| "epoch": 23.003088803088804, | |
| "grad_norm": 0.0309459138661623, | |
| "learning_rate": 5.945945945945947e-06, | |
| "loss": 0.1943, | |
| "step": 6020 | |
| }, | |
| { | |
| "epoch": 23.003861003861005, | |
| "grad_norm": 9.064765930175781, | |
| "learning_rate": 5.937365937365937e-06, | |
| "loss": 0.2176, | |
| "step": 6030 | |
| }, | |
| { | |
| "epoch": 23.004633204633205, | |
| "grad_norm": 261.7420654296875, | |
| "learning_rate": 5.928785928785929e-06, | |
| "loss": 0.3884, | |
| "step": 6040 | |
| }, | |
| { | |
| "epoch": 23.005405405405405, | |
| "grad_norm": 0.10809492319822311, | |
| "learning_rate": 5.9202059202059204e-06, | |
| "loss": 0.4808, | |
| "step": 6050 | |
| }, | |
| { | |
| "epoch": 23.006177606177605, | |
| "grad_norm": 0.49106425046920776, | |
| "learning_rate": 5.911625911625912e-06, | |
| "loss": 0.2767, | |
| "step": 6060 | |
| }, | |
| { | |
| "epoch": 23.006949806949805, | |
| "grad_norm": 0.0949685275554657, | |
| "learning_rate": 5.9030459030459035e-06, | |
| "loss": 0.2362, | |
| "step": 6070 | |
| }, | |
| { | |
| "epoch": 23.00772200772201, | |
| "grad_norm": 0.03749189153313637, | |
| "learning_rate": 5.894465894465895e-06, | |
| "loss": 0.3946, | |
| "step": 6080 | |
| }, | |
| { | |
| "epoch": 23.00849420849421, | |
| "grad_norm": 0.4452950358390808, | |
| "learning_rate": 5.885885885885886e-06, | |
| "loss": 0.282, | |
| "step": 6090 | |
| }, | |
| { | |
| "epoch": 23.00926640926641, | |
| "grad_norm": 0.19385001063346863, | |
| "learning_rate": 5.877305877305878e-06, | |
| "loss": 0.7796, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 23.01003861003861, | |
| "grad_norm": 23.24810791015625, | |
| "learning_rate": 5.86872586872587e-06, | |
| "loss": 0.2623, | |
| "step": 6110 | |
| }, | |
| { | |
| "epoch": 23.01081081081081, | |
| "grad_norm": 20.898603439331055, | |
| "learning_rate": 5.860145860145861e-06, | |
| "loss": 0.2286, | |
| "step": 6120 | |
| }, | |
| { | |
| "epoch": 23.01158301158301, | |
| "grad_norm": 4.817591190338135, | |
| "learning_rate": 5.851565851565853e-06, | |
| "loss": 0.1913, | |
| "step": 6130 | |
| }, | |
| { | |
| "epoch": 23.012355212355214, | |
| "grad_norm": 57.4672737121582, | |
| "learning_rate": 5.842985842985843e-06, | |
| "loss": 0.5206, | |
| "step": 6140 | |
| }, | |
| { | |
| "epoch": 23.013127413127414, | |
| "grad_norm": 0.08692280203104019, | |
| "learning_rate": 5.834405834405835e-06, | |
| "loss": 0.168, | |
| "step": 6150 | |
| }, | |
| { | |
| "epoch": 23.013899613899614, | |
| "grad_norm": 0.025257082656025887, | |
| "learning_rate": 5.825825825825826e-06, | |
| "loss": 0.8649, | |
| "step": 6160 | |
| }, | |
| { | |
| "epoch": 23.014671814671814, | |
| "grad_norm": 0.3466915190219879, | |
| "learning_rate": 5.817245817245818e-06, | |
| "loss": 0.3208, | |
| "step": 6170 | |
| }, | |
| { | |
| "epoch": 23.015444015444015, | |
| "grad_norm": 0.011916655115783215, | |
| "learning_rate": 5.808665808665809e-06, | |
| "loss": 0.3571, | |
| "step": 6180 | |
| }, | |
| { | |
| "epoch": 23.016216216216215, | |
| "grad_norm": 0.740023672580719, | |
| "learning_rate": 5.800085800085801e-06, | |
| "loss": 0.3148, | |
| "step": 6190 | |
| }, | |
| { | |
| "epoch": 23.01698841698842, | |
| "grad_norm": 0.10393180698156357, | |
| "learning_rate": 5.791505791505791e-06, | |
| "loss": 0.0048, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 23.01776061776062, | |
| "grad_norm": 0.09348829090595245, | |
| "learning_rate": 5.782925782925783e-06, | |
| "loss": 0.1214, | |
| "step": 6210 | |
| }, | |
| { | |
| "epoch": 23.01853281853282, | |
| "grad_norm": 11.198524475097656, | |
| "learning_rate": 5.774345774345775e-06, | |
| "loss": 0.5352, | |
| "step": 6220 | |
| }, | |
| { | |
| "epoch": 23.01930501930502, | |
| "grad_norm": 0.02058136835694313, | |
| "learning_rate": 5.765765765765766e-06, | |
| "loss": 0.0152, | |
| "step": 6230 | |
| }, | |
| { | |
| "epoch": 23.02007722007722, | |
| "grad_norm": 0.2556219696998596, | |
| "learning_rate": 5.757185757185758e-06, | |
| "loss": 0.3939, | |
| "step": 6240 | |
| }, | |
| { | |
| "epoch": 23.02007722007722, | |
| "eval_accuracy": 0.6666666666666666, | |
| "eval_loss": 1.6774938106536865, | |
| "eval_runtime": 11.8897, | |
| "eval_samples_per_second": 3.785, | |
| "eval_steps_per_second": 1.262, | |
| "step": 6240 | |
| }, | |
| { | |
| "epoch": 24.0007722007722, | |
| "grad_norm": 66.1806411743164, | |
| "learning_rate": 5.748605748605749e-06, | |
| "loss": 0.3863, | |
| "step": 6250 | |
| }, | |
| { | |
| "epoch": 24.0015444015444, | |
| "grad_norm": 0.03881999850273132, | |
| "learning_rate": 5.740025740025741e-06, | |
| "loss": 0.3546, | |
| "step": 6260 | |
| }, | |
| { | |
| "epoch": 24.0023166023166, | |
| "grad_norm": 17.894153594970703, | |
| "learning_rate": 5.7314457314457315e-06, | |
| "loss": 0.366, | |
| "step": 6270 | |
| }, | |
| { | |
| "epoch": 24.003088803088804, | |
| "grad_norm": 0.07766028493642807, | |
| "learning_rate": 5.722865722865723e-06, | |
| "loss": 0.5358, | |
| "step": 6280 | |
| }, | |
| { | |
| "epoch": 24.003861003861005, | |
| "grad_norm": 0.7994852662086487, | |
| "learning_rate": 5.7142857142857145e-06, | |
| "loss": 0.3454, | |
| "step": 6290 | |
| }, | |
| { | |
| "epoch": 24.004633204633205, | |
| "grad_norm": 0.5819251537322998, | |
| "learning_rate": 5.7057057057057065e-06, | |
| "loss": 0.004, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 24.005405405405405, | |
| "grad_norm": 2.2003848552703857, | |
| "learning_rate": 5.6971256971256976e-06, | |
| "loss": 0.1608, | |
| "step": 6310 | |
| }, | |
| { | |
| "epoch": 24.006177606177605, | |
| "grad_norm": 0.14731518924236298, | |
| "learning_rate": 5.6885456885456895e-06, | |
| "loss": 0.031, | |
| "step": 6320 | |
| }, | |
| { | |
| "epoch": 24.006949806949805, | |
| "grad_norm": 0.06956463307142258, | |
| "learning_rate": 5.67996567996568e-06, | |
| "loss": 0.2076, | |
| "step": 6330 | |
| }, | |
| { | |
| "epoch": 24.00772200772201, | |
| "grad_norm": 0.060266412794589996, | |
| "learning_rate": 5.671385671385672e-06, | |
| "loss": 0.5671, | |
| "step": 6340 | |
| }, | |
| { | |
| "epoch": 24.00849420849421, | |
| "grad_norm": 16.026199340820312, | |
| "learning_rate": 5.662805662805664e-06, | |
| "loss": 0.4175, | |
| "step": 6350 | |
| }, | |
| { | |
| "epoch": 24.00926640926641, | |
| "grad_norm": 0.04196026548743248, | |
| "learning_rate": 5.654225654225655e-06, | |
| "loss": 0.5757, | |
| "step": 6360 | |
| }, | |
| { | |
| "epoch": 24.01003861003861, | |
| "grad_norm": 0.09439310431480408, | |
| "learning_rate": 5.645645645645647e-06, | |
| "loss": 0.075, | |
| "step": 6370 | |
| }, | |
| { | |
| "epoch": 24.01081081081081, | |
| "grad_norm": 0.08069531619548798, | |
| "learning_rate": 5.637065637065637e-06, | |
| "loss": 0.198, | |
| "step": 6380 | |
| }, | |
| { | |
| "epoch": 24.01158301158301, | |
| "grad_norm": 0.04372420907020569, | |
| "learning_rate": 5.628485628485629e-06, | |
| "loss": 0.1694, | |
| "step": 6390 | |
| }, | |
| { | |
| "epoch": 24.012355212355214, | |
| "grad_norm": 6.671726703643799, | |
| "learning_rate": 5.61990561990562e-06, | |
| "loss": 0.1794, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 24.013127413127414, | |
| "grad_norm": 3.417841911315918, | |
| "learning_rate": 5.611325611325612e-06, | |
| "loss": 0.0045, | |
| "step": 6410 | |
| }, | |
| { | |
| "epoch": 24.013899613899614, | |
| "grad_norm": 0.07374245673418045, | |
| "learning_rate": 5.602745602745603e-06, | |
| "loss": 0.3517, | |
| "step": 6420 | |
| }, | |
| { | |
| "epoch": 24.014671814671814, | |
| "grad_norm": 136.1538848876953, | |
| "learning_rate": 5.594165594165595e-06, | |
| "loss": 0.4944, | |
| "step": 6430 | |
| }, | |
| { | |
| "epoch": 24.015444015444015, | |
| "grad_norm": 0.022647704929113388, | |
| "learning_rate": 5.585585585585585e-06, | |
| "loss": 0.564, | |
| "step": 6440 | |
| }, | |
| { | |
| "epoch": 24.016216216216215, | |
| "grad_norm": 41.52387619018555, | |
| "learning_rate": 5.577005577005577e-06, | |
| "loss": 0.7174, | |
| "step": 6450 | |
| }, | |
| { | |
| "epoch": 24.01698841698842, | |
| "grad_norm": 0.14816536009311676, | |
| "learning_rate": 5.568425568425568e-06, | |
| "loss": 0.152, | |
| "step": 6460 | |
| }, | |
| { | |
| "epoch": 24.01776061776062, | |
| "grad_norm": 3.488743543624878, | |
| "learning_rate": 5.55984555984556e-06, | |
| "loss": 0.2765, | |
| "step": 6470 | |
| }, | |
| { | |
| "epoch": 24.01853281853282, | |
| "grad_norm": 0.18953141570091248, | |
| "learning_rate": 5.551265551265552e-06, | |
| "loss": 0.4315, | |
| "step": 6480 | |
| }, | |
| { | |
| "epoch": 24.01930501930502, | |
| "grad_norm": 0.027093812823295593, | |
| "learning_rate": 5.542685542685543e-06, | |
| "loss": 0.1277, | |
| "step": 6490 | |
| }, | |
| { | |
| "epoch": 24.02007722007722, | |
| "grad_norm": 0.24826771020889282, | |
| "learning_rate": 5.534105534105535e-06, | |
| "loss": 0.4129, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 24.02007722007722, | |
| "eval_accuracy": 0.6888888888888889, | |
| "eval_loss": 1.6039283275604248, | |
| "eval_runtime": 11.6661, | |
| "eval_samples_per_second": 3.857, | |
| "eval_steps_per_second": 1.286, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 25.0007722007722, | |
| "grad_norm": 0.07161914557218552, | |
| "learning_rate": 5.5255255255255255e-06, | |
| "loss": 0.1366, | |
| "step": 6510 | |
| }, | |
| { | |
| "epoch": 25.0015444015444, | |
| "grad_norm": 125.4383544921875, | |
| "learning_rate": 5.5169455169455175e-06, | |
| "loss": 0.1942, | |
| "step": 6520 | |
| }, | |
| { | |
| "epoch": 25.0023166023166, | |
| "grad_norm": 0.08652602881193161, | |
| "learning_rate": 5.5083655083655086e-06, | |
| "loss": 0.1576, | |
| "step": 6530 | |
| }, | |
| { | |
| "epoch": 25.003088803088804, | |
| "grad_norm": 0.08510206639766693, | |
| "learning_rate": 5.4997854997855005e-06, | |
| "loss": 0.2385, | |
| "step": 6540 | |
| }, | |
| { | |
| "epoch": 25.003861003861005, | |
| "grad_norm": 0.060757704079151154, | |
| "learning_rate": 5.491205491205492e-06, | |
| "loss": 0.3652, | |
| "step": 6550 | |
| }, | |
| { | |
| "epoch": 25.004633204633205, | |
| "grad_norm": 53.286048889160156, | |
| "learning_rate": 5.4826254826254836e-06, | |
| "loss": 0.5818, | |
| "step": 6560 | |
| }, | |
| { | |
| "epoch": 25.005405405405405, | |
| "grad_norm": 0.04687732830643654, | |
| "learning_rate": 5.474045474045474e-06, | |
| "loss": 0.5724, | |
| "step": 6570 | |
| }, | |
| { | |
| "epoch": 25.006177606177605, | |
| "grad_norm": 0.12820780277252197, | |
| "learning_rate": 5.465465465465466e-06, | |
| "loss": 0.5346, | |
| "step": 6580 | |
| }, | |
| { | |
| "epoch": 25.006949806949805, | |
| "grad_norm": 128.59718322753906, | |
| "learning_rate": 5.456885456885457e-06, | |
| "loss": 0.5187, | |
| "step": 6590 | |
| }, | |
| { | |
| "epoch": 25.00772200772201, | |
| "grad_norm": 15.195361137390137, | |
| "learning_rate": 5.448305448305449e-06, | |
| "loss": 0.0123, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 25.00849420849421, | |
| "grad_norm": 0.21481989324092865, | |
| "learning_rate": 5.439725439725441e-06, | |
| "loss": 0.1878, | |
| "step": 6610 | |
| }, | |
| { | |
| "epoch": 25.00926640926641, | |
| "grad_norm": 1.0520018339157104, | |
| "learning_rate": 5.431145431145432e-06, | |
| "loss": 0.1593, | |
| "step": 6620 | |
| }, | |
| { | |
| "epoch": 25.01003861003861, | |
| "grad_norm": 0.05230041593313217, | |
| "learning_rate": 5.422565422565423e-06, | |
| "loss": 0.1364, | |
| "step": 6630 | |
| }, | |
| { | |
| "epoch": 25.01081081081081, | |
| "grad_norm": 0.03155992925167084, | |
| "learning_rate": 5.413985413985414e-06, | |
| "loss": 0.0715, | |
| "step": 6640 | |
| }, | |
| { | |
| "epoch": 25.01158301158301, | |
| "grad_norm": 25.422481536865234, | |
| "learning_rate": 5.405405405405406e-06, | |
| "loss": 0.3257, | |
| "step": 6650 | |
| }, | |
| { | |
| "epoch": 25.012355212355214, | |
| "grad_norm": 0.06870962679386139, | |
| "learning_rate": 5.396825396825397e-06, | |
| "loss": 0.1611, | |
| "step": 6660 | |
| }, | |
| { | |
| "epoch": 25.013127413127414, | |
| "grad_norm": 0.7361809611320496, | |
| "learning_rate": 5.388245388245389e-06, | |
| "loss": 0.0661, | |
| "step": 6670 | |
| }, | |
| { | |
| "epoch": 25.013899613899614, | |
| "grad_norm": 0.007542428094893694, | |
| "learning_rate": 5.379665379665379e-06, | |
| "loss": 0.303, | |
| "step": 6680 | |
| }, | |
| { | |
| "epoch": 25.014671814671814, | |
| "grad_norm": 0.03758428990840912, | |
| "learning_rate": 5.371085371085371e-06, | |
| "loss": 0.3519, | |
| "step": 6690 | |
| }, | |
| { | |
| "epoch": 25.015444015444015, | |
| "grad_norm": 4.279178142547607, | |
| "learning_rate": 5.362505362505362e-06, | |
| "loss": 0.2035, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 25.016216216216215, | |
| "grad_norm": 0.024570725858211517, | |
| "learning_rate": 5.353925353925354e-06, | |
| "loss": 0.544, | |
| "step": 6710 | |
| }, | |
| { | |
| "epoch": 25.01698841698842, | |
| "grad_norm": 0.11677242070436478, | |
| "learning_rate": 5.345345345345346e-06, | |
| "loss": 0.2069, | |
| "step": 6720 | |
| }, | |
| { | |
| "epoch": 25.01776061776062, | |
| "grad_norm": 56.570030212402344, | |
| "learning_rate": 5.336765336765337e-06, | |
| "loss": 0.37, | |
| "step": 6730 | |
| }, | |
| { | |
| "epoch": 25.01853281853282, | |
| "grad_norm": 111.80818176269531, | |
| "learning_rate": 5.328185328185329e-06, | |
| "loss": 0.1691, | |
| "step": 6740 | |
| }, | |
| { | |
| "epoch": 25.01930501930502, | |
| "grad_norm": 19.810441970825195, | |
| "learning_rate": 5.31960531960532e-06, | |
| "loss": 0.6511, | |
| "step": 6750 | |
| }, | |
| { | |
| "epoch": 25.02007722007722, | |
| "grad_norm": 0.07480132579803467, | |
| "learning_rate": 5.3110253110253115e-06, | |
| "loss": 0.2852, | |
| "step": 6760 | |
| }, | |
| { | |
| "epoch": 25.02007722007722, | |
| "eval_accuracy": 0.7111111111111111, | |
| "eval_loss": 1.676919937133789, | |
| "eval_runtime": 11.755, | |
| "eval_samples_per_second": 3.828, | |
| "eval_steps_per_second": 1.276, | |
| "step": 6760 | |
| }, | |
| { | |
| "epoch": 26.0007722007722, | |
| "grad_norm": 7.0793681144714355, | |
| "learning_rate": 5.302445302445303e-06, | |
| "loss": 0.3598, | |
| "step": 6770 | |
| }, | |
| { | |
| "epoch": 26.0015444015444, | |
| "grad_norm": 0.04627354443073273, | |
| "learning_rate": 5.293865293865295e-06, | |
| "loss": 0.2332, | |
| "step": 6780 | |
| }, | |
| { | |
| "epoch": 26.0023166023166, | |
| "grad_norm": 0.015025601722300053, | |
| "learning_rate": 5.285285285285286e-06, | |
| "loss": 0.3468, | |
| "step": 6790 | |
| }, | |
| { | |
| "epoch": 26.003088803088804, | |
| "grad_norm": 0.9374003410339355, | |
| "learning_rate": 5.276705276705278e-06, | |
| "loss": 0.2451, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 26.003861003861005, | |
| "grad_norm": 0.06715880334377289, | |
| "learning_rate": 5.268125268125268e-06, | |
| "loss": 0.0179, | |
| "step": 6810 | |
| }, | |
| { | |
| "epoch": 26.004633204633205, | |
| "grad_norm": 0.1493833363056183, | |
| "learning_rate": 5.25954525954526e-06, | |
| "loss": 0.4544, | |
| "step": 6820 | |
| }, | |
| { | |
| "epoch": 26.005405405405405, | |
| "grad_norm": 0.008493917062878609, | |
| "learning_rate": 5.250965250965251e-06, | |
| "loss": 0.4215, | |
| "step": 6830 | |
| }, | |
| { | |
| "epoch": 26.006177606177605, | |
| "grad_norm": 0.24284197390079498, | |
| "learning_rate": 5.242385242385243e-06, | |
| "loss": 0.2307, | |
| "step": 6840 | |
| }, | |
| { | |
| "epoch": 26.006949806949805, | |
| "grad_norm": 1.8147085905075073, | |
| "learning_rate": 5.233805233805235e-06, | |
| "loss": 0.1164, | |
| "step": 6850 | |
| }, | |
| { | |
| "epoch": 26.00772200772201, | |
| "grad_norm": 0.02225414104759693, | |
| "learning_rate": 5.225225225225226e-06, | |
| "loss": 0.6091, | |
| "step": 6860 | |
| }, | |
| { | |
| "epoch": 26.00849420849421, | |
| "grad_norm": 0.03931829333305359, | |
| "learning_rate": 5.216645216645218e-06, | |
| "loss": 0.4782, | |
| "step": 6870 | |
| }, | |
| { | |
| "epoch": 26.00926640926641, | |
| "grad_norm": 0.9832797050476074, | |
| "learning_rate": 5.208065208065208e-06, | |
| "loss": 0.107, | |
| "step": 6880 | |
| }, | |
| { | |
| "epoch": 26.01003861003861, | |
| "grad_norm": 0.5841915607452393, | |
| "learning_rate": 5.1994851994852e-06, | |
| "loss": 0.0516, | |
| "step": 6890 | |
| }, | |
| { | |
| "epoch": 26.01081081081081, | |
| "grad_norm": 0.2941983640193939, | |
| "learning_rate": 5.190905190905191e-06, | |
| "loss": 0.4148, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 26.01158301158301, | |
| "grad_norm": 0.057785652577877045, | |
| "learning_rate": 5.182325182325183e-06, | |
| "loss": 0.1971, | |
| "step": 6910 | |
| }, | |
| { | |
| "epoch": 26.012355212355214, | |
| "grad_norm": 0.07821908593177795, | |
| "learning_rate": 5.173745173745173e-06, | |
| "loss": 0.1446, | |
| "step": 6920 | |
| }, | |
| { | |
| "epoch": 26.013127413127414, | |
| "grad_norm": 76.94812774658203, | |
| "learning_rate": 5.165165165165165e-06, | |
| "loss": 0.5563, | |
| "step": 6930 | |
| }, | |
| { | |
| "epoch": 26.013899613899614, | |
| "grad_norm": 0.0389854796230793, | |
| "learning_rate": 5.1565851565851565e-06, | |
| "loss": 0.2554, | |
| "step": 6940 | |
| }, | |
| { | |
| "epoch": 26.014671814671814, | |
| "grad_norm": 0.08624967187643051, | |
| "learning_rate": 5.148005148005148e-06, | |
| "loss": 0.0047, | |
| "step": 6950 | |
| }, | |
| { | |
| "epoch": 26.015444015444015, | |
| "grad_norm": 0.022646430879831314, | |
| "learning_rate": 5.1394251394251395e-06, | |
| "loss": 0.3356, | |
| "step": 6960 | |
| }, | |
| { | |
| "epoch": 26.016216216216215, | |
| "grad_norm": 0.004841935355216265, | |
| "learning_rate": 5.1308451308451314e-06, | |
| "loss": 0.0198, | |
| "step": 6970 | |
| }, | |
| { | |
| "epoch": 26.01698841698842, | |
| "grad_norm": 118.31085968017578, | |
| "learning_rate": 5.122265122265123e-06, | |
| "loss": 0.3741, | |
| "step": 6980 | |
| }, | |
| { | |
| "epoch": 26.01776061776062, | |
| "grad_norm": 95.70652770996094, | |
| "learning_rate": 5.113685113685114e-06, | |
| "loss": 0.224, | |
| "step": 6990 | |
| }, | |
| { | |
| "epoch": 26.01853281853282, | |
| "grad_norm": 114.44501495361328, | |
| "learning_rate": 5.105105105105106e-06, | |
| "loss": 0.2583, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 26.01930501930502, | |
| "grad_norm": 197.85206604003906, | |
| "learning_rate": 5.096525096525097e-06, | |
| "loss": 0.0754, | |
| "step": 7010 | |
| }, | |
| { | |
| "epoch": 26.02007722007722, | |
| "grad_norm": 0.006557074375450611, | |
| "learning_rate": 5.087945087945089e-06, | |
| "loss": 0.3278, | |
| "step": 7020 | |
| }, | |
| { | |
| "epoch": 26.02007722007722, | |
| "eval_accuracy": 0.6888888888888889, | |
| "eval_loss": 1.9128968715667725, | |
| "eval_runtime": 11.5823, | |
| "eval_samples_per_second": 3.885, | |
| "eval_steps_per_second": 1.295, | |
| "step": 7020 | |
| }, | |
| { | |
| "epoch": 27.0007722007722, | |
| "grad_norm": 0.02128458209335804, | |
| "learning_rate": 5.07936507936508e-06, | |
| "loss": 0.896, | |
| "step": 7030 | |
| }, | |
| { | |
| "epoch": 27.0015444015444, | |
| "grad_norm": 1.866234540939331, | |
| "learning_rate": 5.070785070785072e-06, | |
| "loss": 0.139, | |
| "step": 7040 | |
| }, | |
| { | |
| "epoch": 27.0023166023166, | |
| "grad_norm": 121.14530944824219, | |
| "learning_rate": 5.062205062205062e-06, | |
| "loss": 0.1566, | |
| "step": 7050 | |
| }, | |
| { | |
| "epoch": 27.003088803088804, | |
| "grad_norm": 0.4306340515613556, | |
| "learning_rate": 5.053625053625054e-06, | |
| "loss": 0.1125, | |
| "step": 7060 | |
| }, | |
| { | |
| "epoch": 27.003861003861005, | |
| "grad_norm": 0.0931638553738594, | |
| "learning_rate": 5.045045045045045e-06, | |
| "loss": 0.0045, | |
| "step": 7070 | |
| }, | |
| { | |
| "epoch": 27.004633204633205, | |
| "grad_norm": 21.92576789855957, | |
| "learning_rate": 5.036465036465037e-06, | |
| "loss": 0.5565, | |
| "step": 7080 | |
| }, | |
| { | |
| "epoch": 27.005405405405405, | |
| "grad_norm": 0.08851654082536697, | |
| "learning_rate": 5.027885027885028e-06, | |
| "loss": 0.0472, | |
| "step": 7090 | |
| }, | |
| { | |
| "epoch": 27.006177606177605, | |
| "grad_norm": 0.02390648052096367, | |
| "learning_rate": 5.01930501930502e-06, | |
| "loss": 0.3594, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 27.006949806949805, | |
| "grad_norm": 0.02931853011250496, | |
| "learning_rate": 5.010725010725012e-06, | |
| "loss": 0.285, | |
| "step": 7110 | |
| }, | |
| { | |
| "epoch": 27.00772200772201, | |
| "grad_norm": 0.044821541756391525, | |
| "learning_rate": 5.002145002145002e-06, | |
| "loss": 0.4182, | |
| "step": 7120 | |
| }, | |
| { | |
| "epoch": 27.00849420849421, | |
| "grad_norm": 0.0022613268811255693, | |
| "learning_rate": 4.993564993564994e-06, | |
| "loss": 0.2255, | |
| "step": 7130 | |
| }, | |
| { | |
| "epoch": 27.00926640926641, | |
| "grad_norm": 0.052004750818014145, | |
| "learning_rate": 4.984984984984985e-06, | |
| "loss": 0.0132, | |
| "step": 7140 | |
| }, | |
| { | |
| "epoch": 27.01003861003861, | |
| "grad_norm": 0.07019229978322983, | |
| "learning_rate": 4.976404976404976e-06, | |
| "loss": 0.1932, | |
| "step": 7150 | |
| }, | |
| { | |
| "epoch": 27.01081081081081, | |
| "grad_norm": 2.3248353004455566, | |
| "learning_rate": 4.967824967824968e-06, | |
| "loss": 0.3329, | |
| "step": 7160 | |
| }, | |
| { | |
| "epoch": 27.01158301158301, | |
| "grad_norm": 0.016531068831682205, | |
| "learning_rate": 4.95924495924496e-06, | |
| "loss": 0.2348, | |
| "step": 7170 | |
| }, | |
| { | |
| "epoch": 27.012355212355214, | |
| "grad_norm": 0.010612938553094864, | |
| "learning_rate": 4.950664950664951e-06, | |
| "loss": 0.0597, | |
| "step": 7180 | |
| }, | |
| { | |
| "epoch": 27.013127413127414, | |
| "grad_norm": 0.01717378944158554, | |
| "learning_rate": 4.9420849420849425e-06, | |
| "loss": 0.171, | |
| "step": 7190 | |
| }, | |
| { | |
| "epoch": 27.013899613899614, | |
| "grad_norm": 0.027986228466033936, | |
| "learning_rate": 4.9335049335049336e-06, | |
| "loss": 0.3064, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 27.014671814671814, | |
| "grad_norm": 0.02567068673670292, | |
| "learning_rate": 4.9249249249249255e-06, | |
| "loss": 0.1591, | |
| "step": 7210 | |
| }, | |
| { | |
| "epoch": 27.015444015444015, | |
| "grad_norm": 0.11277701705694199, | |
| "learning_rate": 4.916344916344917e-06, | |
| "loss": 0.0005, | |
| "step": 7220 | |
| }, | |
| { | |
| "epoch": 27.016216216216215, | |
| "grad_norm": 0.00369657832197845, | |
| "learning_rate": 4.907764907764908e-06, | |
| "loss": 0.2153, | |
| "step": 7230 | |
| }, | |
| { | |
| "epoch": 27.01698841698842, | |
| "grad_norm": 0.036675579845905304, | |
| "learning_rate": 4.8991848991849e-06, | |
| "loss": 0.2176, | |
| "step": 7240 | |
| }, | |
| { | |
| "epoch": 27.01776061776062, | |
| "grad_norm": 0.026838406920433044, | |
| "learning_rate": 4.890604890604891e-06, | |
| "loss": 0.5287, | |
| "step": 7250 | |
| }, | |
| { | |
| "epoch": 27.01853281853282, | |
| "grad_norm": 0.8378888964653015, | |
| "learning_rate": 4.882024882024882e-06, | |
| "loss": 0.0196, | |
| "step": 7260 | |
| }, | |
| { | |
| "epoch": 27.01930501930502, | |
| "grad_norm": 0.03345318138599396, | |
| "learning_rate": 4.873444873444874e-06, | |
| "loss": 0.1936, | |
| "step": 7270 | |
| }, | |
| { | |
| "epoch": 27.02007722007722, | |
| "grad_norm": 0.004973672796040773, | |
| "learning_rate": 4.864864864864866e-06, | |
| "loss": 0.7677, | |
| "step": 7280 | |
| }, | |
| { | |
| "epoch": 27.02007722007722, | |
| "eval_accuracy": 0.6666666666666666, | |
| "eval_loss": 1.83974027633667, | |
| "eval_runtime": 11.7937, | |
| "eval_samples_per_second": 3.816, | |
| "eval_steps_per_second": 1.272, | |
| "step": 7280 | |
| }, | |
| { | |
| "epoch": 28.0007722007722, | |
| "grad_norm": 0.023714346811175346, | |
| "learning_rate": 4.856284856284857e-06, | |
| "loss": 0.1018, | |
| "step": 7290 | |
| }, | |
| { | |
| "epoch": 28.0015444015444, | |
| "grad_norm": 122.97506713867188, | |
| "learning_rate": 4.847704847704848e-06, | |
| "loss": 0.219, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 28.0023166023166, | |
| "grad_norm": 0.14896957576274872, | |
| "learning_rate": 4.83912483912484e-06, | |
| "loss": 0.1888, | |
| "step": 7310 | |
| }, | |
| { | |
| "epoch": 28.003088803088804, | |
| "grad_norm": 0.05097675696015358, | |
| "learning_rate": 4.830544830544831e-06, | |
| "loss": 0.276, | |
| "step": 7320 | |
| }, | |
| { | |
| "epoch": 28.003861003861005, | |
| "grad_norm": 0.01336723007261753, | |
| "learning_rate": 4.821964821964822e-06, | |
| "loss": 0.0007, | |
| "step": 7330 | |
| }, | |
| { | |
| "epoch": 28.004633204633205, | |
| "grad_norm": 2.9897027015686035, | |
| "learning_rate": 4.813384813384814e-06, | |
| "loss": 0.0017, | |
| "step": 7340 | |
| }, | |
| { | |
| "epoch": 28.005405405405405, | |
| "grad_norm": 310.45367431640625, | |
| "learning_rate": 4.804804804804805e-06, | |
| "loss": 0.2582, | |
| "step": 7350 | |
| }, | |
| { | |
| "epoch": 28.006177606177605, | |
| "grad_norm": 12.623224258422852, | |
| "learning_rate": 4.796224796224796e-06, | |
| "loss": 0.2349, | |
| "step": 7360 | |
| }, | |
| { | |
| "epoch": 28.006949806949805, | |
| "grad_norm": 323.9424743652344, | |
| "learning_rate": 4.787644787644788e-06, | |
| "loss": 0.4081, | |
| "step": 7370 | |
| }, | |
| { | |
| "epoch": 28.00772200772201, | |
| "grad_norm": 0.05448540300130844, | |
| "learning_rate": 4.779064779064779e-06, | |
| "loss": 0.5078, | |
| "step": 7380 | |
| }, | |
| { | |
| "epoch": 28.00849420849421, | |
| "grad_norm": 105.11054229736328, | |
| "learning_rate": 4.7704847704847704e-06, | |
| "loss": 0.4487, | |
| "step": 7390 | |
| }, | |
| { | |
| "epoch": 28.00926640926641, | |
| "grad_norm": 0.3405386209487915, | |
| "learning_rate": 4.761904761904762e-06, | |
| "loss": 0.6769, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 28.01003861003861, | |
| "grad_norm": 0.48794102668762207, | |
| "learning_rate": 4.753324753324754e-06, | |
| "loss": 0.1995, | |
| "step": 7410 | |
| }, | |
| { | |
| "epoch": 28.01081081081081, | |
| "grad_norm": 0.042529307305812836, | |
| "learning_rate": 4.7447447447447454e-06, | |
| "loss": 0.2511, | |
| "step": 7420 | |
| }, | |
| { | |
| "epoch": 28.01158301158301, | |
| "grad_norm": 0.014477269724011421, | |
| "learning_rate": 4.7361647361647365e-06, | |
| "loss": 0.0833, | |
| "step": 7430 | |
| }, | |
| { | |
| "epoch": 28.012355212355214, | |
| "grad_norm": 0.003762512933462858, | |
| "learning_rate": 4.7275847275847285e-06, | |
| "loss": 0.2294, | |
| "step": 7440 | |
| }, | |
| { | |
| "epoch": 28.013127413127414, | |
| "grad_norm": 157.7808380126953, | |
| "learning_rate": 4.71900471900472e-06, | |
| "loss": 0.3185, | |
| "step": 7450 | |
| }, | |
| { | |
| "epoch": 28.013899613899614, | |
| "grad_norm": 93.19122314453125, | |
| "learning_rate": 4.710424710424711e-06, | |
| "loss": 0.4783, | |
| "step": 7460 | |
| }, | |
| { | |
| "epoch": 28.014671814671814, | |
| "grad_norm": 0.0037978780455887318, | |
| "learning_rate": 4.701844701844703e-06, | |
| "loss": 0.0011, | |
| "step": 7470 | |
| }, | |
| { | |
| "epoch": 28.015444015444015, | |
| "grad_norm": 0.003815456759184599, | |
| "learning_rate": 4.693264693264694e-06, | |
| "loss": 0.071, | |
| "step": 7480 | |
| }, | |
| { | |
| "epoch": 28.016216216216215, | |
| "grad_norm": 0.29880446195602417, | |
| "learning_rate": 4.684684684684685e-06, | |
| "loss": 0.0157, | |
| "step": 7490 | |
| }, | |
| { | |
| "epoch": 28.01698841698842, | |
| "grad_norm": 151.8013458251953, | |
| "learning_rate": 4.676104676104676e-06, | |
| "loss": 0.3594, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 28.01776061776062, | |
| "grad_norm": 0.013940732926130295, | |
| "learning_rate": 4.667524667524668e-06, | |
| "loss": 0.0044, | |
| "step": 7510 | |
| }, | |
| { | |
| "epoch": 28.01853281853282, | |
| "grad_norm": 0.23434804379940033, | |
| "learning_rate": 4.658944658944659e-06, | |
| "loss": 0.1779, | |
| "step": 7520 | |
| }, | |
| { | |
| "epoch": 28.01930501930502, | |
| "grad_norm": 0.02195580303668976, | |
| "learning_rate": 4.650364650364651e-06, | |
| "loss": 0.2855, | |
| "step": 7530 | |
| }, | |
| { | |
| "epoch": 28.02007722007722, | |
| "grad_norm": 0.017229193821549416, | |
| "learning_rate": 4.641784641784642e-06, | |
| "loss": 0.0122, | |
| "step": 7540 | |
| }, | |
| { | |
| "epoch": 28.02007722007722, | |
| "eval_accuracy": 0.6888888888888889, | |
| "eval_loss": 2.029585838317871, | |
| "eval_runtime": 12.7497, | |
| "eval_samples_per_second": 3.529, | |
| "eval_steps_per_second": 1.176, | |
| "step": 7540 | |
| }, | |
| { | |
| "epoch": 29.0007722007722, | |
| "grad_norm": 0.01609969511628151, | |
| "learning_rate": 4.633204633204634e-06, | |
| "loss": 0.1734, | |
| "step": 7550 | |
| }, | |
| { | |
| "epoch": 29.0015444015444, | |
| "grad_norm": 131.36447143554688, | |
| "learning_rate": 4.624624624624625e-06, | |
| "loss": 0.0932, | |
| "step": 7560 | |
| }, | |
| { | |
| "epoch": 29.0023166023166, | |
| "grad_norm": 0.007553355302661657, | |
| "learning_rate": 4.616044616044616e-06, | |
| "loss": 0.233, | |
| "step": 7570 | |
| }, | |
| { | |
| "epoch": 29.003088803088804, | |
| "grad_norm": 29.221181869506836, | |
| "learning_rate": 4.607464607464608e-06, | |
| "loss": 0.0419, | |
| "step": 7580 | |
| }, | |
| { | |
| "epoch": 29.003861003861005, | |
| "grad_norm": 3.5382421016693115, | |
| "learning_rate": 4.598884598884599e-06, | |
| "loss": 0.2079, | |
| "step": 7590 | |
| }, | |
| { | |
| "epoch": 29.004633204633205, | |
| "grad_norm": 28.57002830505371, | |
| "learning_rate": 4.59030459030459e-06, | |
| "loss": 0.2519, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 29.005405405405405, | |
| "grad_norm": 0.129736989736557, | |
| "learning_rate": 4.581724581724582e-06, | |
| "loss": 0.0076, | |
| "step": 7610 | |
| }, | |
| { | |
| "epoch": 29.006177606177605, | |
| "grad_norm": 2.590891122817993, | |
| "learning_rate": 4.573144573144573e-06, | |
| "loss": 0.4462, | |
| "step": 7620 | |
| }, | |
| { | |
| "epoch": 29.006949806949805, | |
| "grad_norm": 0.006880035623908043, | |
| "learning_rate": 4.5645645645645645e-06, | |
| "loss": 0.0013, | |
| "step": 7630 | |
| }, | |
| { | |
| "epoch": 29.00772200772201, | |
| "grad_norm": 0.006625477224588394, | |
| "learning_rate": 4.5559845559845564e-06, | |
| "loss": 0.6628, | |
| "step": 7640 | |
| }, | |
| { | |
| "epoch": 29.00849420849421, | |
| "grad_norm": 0.3407871425151825, | |
| "learning_rate": 4.5474045474045475e-06, | |
| "loss": 0.0684, | |
| "step": 7650 | |
| }, | |
| { | |
| "epoch": 29.00926640926641, | |
| "grad_norm": 0.2093740552663803, | |
| "learning_rate": 4.5388245388245395e-06, | |
| "loss": 0.1559, | |
| "step": 7660 | |
| }, | |
| { | |
| "epoch": 29.01003861003861, | |
| "grad_norm": 0.5102840662002563, | |
| "learning_rate": 4.530244530244531e-06, | |
| "loss": 0.3277, | |
| "step": 7670 | |
| }, | |
| { | |
| "epoch": 29.01081081081081, | |
| "grad_norm": 0.02635369263589382, | |
| "learning_rate": 4.5216645216645225e-06, | |
| "loss": 0.2345, | |
| "step": 7680 | |
| }, | |
| { | |
| "epoch": 29.01158301158301, | |
| "grad_norm": 47.610321044921875, | |
| "learning_rate": 4.513084513084514e-06, | |
| "loss": 0.5266, | |
| "step": 7690 | |
| }, | |
| { | |
| "epoch": 29.012355212355214, | |
| "grad_norm": 0.03327294811606407, | |
| "learning_rate": 4.504504504504505e-06, | |
| "loss": 0.1346, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 29.013127413127414, | |
| "grad_norm": 0.08322642743587494, | |
| "learning_rate": 4.495924495924497e-06, | |
| "loss": 0.2012, | |
| "step": 7710 | |
| }, | |
| { | |
| "epoch": 29.013899613899614, | |
| "grad_norm": 1.1271024942398071, | |
| "learning_rate": 4.487344487344488e-06, | |
| "loss": 0.239, | |
| "step": 7720 | |
| }, | |
| { | |
| "epoch": 29.014671814671814, | |
| "grad_norm": 0.010516617447137833, | |
| "learning_rate": 4.478764478764479e-06, | |
| "loss": 0.3092, | |
| "step": 7730 | |
| }, | |
| { | |
| "epoch": 29.015444015444015, | |
| "grad_norm": 39.980411529541016, | |
| "learning_rate": 4.470184470184471e-06, | |
| "loss": 0.2006, | |
| "step": 7740 | |
| }, | |
| { | |
| "epoch": 29.016216216216215, | |
| "grad_norm": 13.700182914733887, | |
| "learning_rate": 4.461604461604462e-06, | |
| "loss": 0.215, | |
| "step": 7750 | |
| }, | |
| { | |
| "epoch": 29.01698841698842, | |
| "grad_norm": 7.643761157989502, | |
| "learning_rate": 4.453024453024453e-06, | |
| "loss": 0.2452, | |
| "step": 7760 | |
| }, | |
| { | |
| "epoch": 29.01776061776062, | |
| "grad_norm": 0.015428297221660614, | |
| "learning_rate": 4.444444444444444e-06, | |
| "loss": 0.004, | |
| "step": 7770 | |
| }, | |
| { | |
| "epoch": 29.01853281853282, | |
| "grad_norm": 0.02670716680586338, | |
| "learning_rate": 4.435864435864436e-06, | |
| "loss": 0.0358, | |
| "step": 7780 | |
| }, | |
| { | |
| "epoch": 29.01930501930502, | |
| "grad_norm": 0.052240751683712006, | |
| "learning_rate": 4.427284427284428e-06, | |
| "loss": 0.2065, | |
| "step": 7790 | |
| }, | |
| { | |
| "epoch": 29.02007722007722, | |
| "grad_norm": 0.0033651816193014383, | |
| "learning_rate": 4.418704418704419e-06, | |
| "loss": 0.3014, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 29.02007722007722, | |
| "eval_accuracy": 0.6, | |
| "eval_loss": 2.4233789443969727, | |
| "eval_runtime": 11.8236, | |
| "eval_samples_per_second": 3.806, | |
| "eval_steps_per_second": 1.269, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 30.0007722007722, | |
| "grad_norm": 259.00091552734375, | |
| "learning_rate": 4.41012441012441e-06, | |
| "loss": 0.5366, | |
| "step": 7810 | |
| }, | |
| { | |
| "epoch": 30.0015444015444, | |
| "grad_norm": 0.02561926282942295, | |
| "learning_rate": 4.401544401544402e-06, | |
| "loss": 0.257, | |
| "step": 7820 | |
| }, | |
| { | |
| "epoch": 30.0023166023166, | |
| "grad_norm": 6.1988606452941895, | |
| "learning_rate": 4.392964392964393e-06, | |
| "loss": 0.1662, | |
| "step": 7830 | |
| }, | |
| { | |
| "epoch": 30.003088803088804, | |
| "grad_norm": 37.069583892822266, | |
| "learning_rate": 4.384384384384384e-06, | |
| "loss": 0.3467, | |
| "step": 7840 | |
| }, | |
| { | |
| "epoch": 30.003861003861005, | |
| "grad_norm": 1.0843921899795532, | |
| "learning_rate": 4.375804375804376e-06, | |
| "loss": 0.0114, | |
| "step": 7850 | |
| }, | |
| { | |
| "epoch": 30.004633204633205, | |
| "grad_norm": 0.07318487763404846, | |
| "learning_rate": 4.3672243672243675e-06, | |
| "loss": 0.1679, | |
| "step": 7860 | |
| }, | |
| { | |
| "epoch": 30.005405405405405, | |
| "grad_norm": 7.664827823638916, | |
| "learning_rate": 4.3586443586443586e-06, | |
| "loss": 0.1834, | |
| "step": 7870 | |
| }, | |
| { | |
| "epoch": 30.006177606177605, | |
| "grad_norm": 84.84963989257812, | |
| "learning_rate": 4.3500643500643505e-06, | |
| "loss": 0.6496, | |
| "step": 7880 | |
| }, | |
| { | |
| "epoch": 30.006949806949805, | |
| "grad_norm": 0.12111800163984299, | |
| "learning_rate": 4.341484341484342e-06, | |
| "loss": 0.0015, | |
| "step": 7890 | |
| }, | |
| { | |
| "epoch": 30.00772200772201, | |
| "grad_norm": 0.15672890841960907, | |
| "learning_rate": 4.332904332904333e-06, | |
| "loss": 1.4159, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 30.00849420849421, | |
| "grad_norm": 144.9275665283203, | |
| "learning_rate": 4.324324324324325e-06, | |
| "loss": 0.5313, | |
| "step": 7910 | |
| }, | |
| { | |
| "epoch": 30.00926640926641, | |
| "grad_norm": 0.24197010695934296, | |
| "learning_rate": 4.315744315744317e-06, | |
| "loss": 0.6925, | |
| "step": 7920 | |
| }, | |
| { | |
| "epoch": 30.01003861003861, | |
| "grad_norm": 178.34043884277344, | |
| "learning_rate": 4.307164307164308e-06, | |
| "loss": 0.0522, | |
| "step": 7930 | |
| }, | |
| { | |
| "epoch": 30.01081081081081, | |
| "grad_norm": 0.059441838413476944, | |
| "learning_rate": 4.298584298584299e-06, | |
| "loss": 0.0813, | |
| "step": 7940 | |
| }, | |
| { | |
| "epoch": 30.01158301158301, | |
| "grad_norm": 0.12464194744825363, | |
| "learning_rate": 4.290004290004291e-06, | |
| "loss": 0.0013, | |
| "step": 7950 | |
| }, | |
| { | |
| "epoch": 30.012355212355214, | |
| "grad_norm": 0.02124338410794735, | |
| "learning_rate": 4.281424281424282e-06, | |
| "loss": 0.1996, | |
| "step": 7960 | |
| }, | |
| { | |
| "epoch": 30.013127413127414, | |
| "grad_norm": 0.03584703058004379, | |
| "learning_rate": 4.272844272844273e-06, | |
| "loss": 0.1586, | |
| "step": 7970 | |
| }, | |
| { | |
| "epoch": 30.013899613899614, | |
| "grad_norm": 0.07277407497167587, | |
| "learning_rate": 4.264264264264265e-06, | |
| "loss": 0.1546, | |
| "step": 7980 | |
| }, | |
| { | |
| "epoch": 30.014671814671814, | |
| "grad_norm": 0.16193543374538422, | |
| "learning_rate": 4.255684255684256e-06, | |
| "loss": 0.4482, | |
| "step": 7990 | |
| }, | |
| { | |
| "epoch": 30.015444015444015, | |
| "grad_norm": 0.02647043950855732, | |
| "learning_rate": 4.247104247104247e-06, | |
| "loss": 0.1865, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 30.016216216216215, | |
| "grad_norm": 0.6456356048583984, | |
| "learning_rate": 4.238524238524239e-06, | |
| "loss": 0.1314, | |
| "step": 8010 | |
| }, | |
| { | |
| "epoch": 30.01698841698842, | |
| "grad_norm": 133.39508056640625, | |
| "learning_rate": 4.22994422994423e-06, | |
| "loss": 0.3308, | |
| "step": 8020 | |
| }, | |
| { | |
| "epoch": 30.01776061776062, | |
| "grad_norm": 0.02105424925684929, | |
| "learning_rate": 4.221364221364222e-06, | |
| "loss": 0.4071, | |
| "step": 8030 | |
| }, | |
| { | |
| "epoch": 30.01853281853282, | |
| "grad_norm": 0.94647616147995, | |
| "learning_rate": 4.212784212784213e-06, | |
| "loss": 0.5776, | |
| "step": 8040 | |
| }, | |
| { | |
| "epoch": 30.01930501930502, | |
| "grad_norm": 4.219485282897949, | |
| "learning_rate": 4.204204204204204e-06, | |
| "loss": 0.2465, | |
| "step": 8050 | |
| }, | |
| { | |
| "epoch": 30.02007722007722, | |
| "grad_norm": 0.12528596818447113, | |
| "learning_rate": 4.195624195624196e-06, | |
| "loss": 0.3567, | |
| "step": 8060 | |
| }, | |
| { | |
| "epoch": 30.02007722007722, | |
| "eval_accuracy": 0.7111111111111111, | |
| "eval_loss": 1.7570387125015259, | |
| "eval_runtime": 11.787, | |
| "eval_samples_per_second": 3.818, | |
| "eval_steps_per_second": 1.273, | |
| "step": 8060 | |
| }, | |
| { | |
| "epoch": 31.0007722007722, | |
| "grad_norm": 0.003974312916398048, | |
| "learning_rate": 4.187044187044187e-06, | |
| "loss": 0.1881, | |
| "step": 8070 | |
| }, | |
| { | |
| "epoch": 31.0015444015444, | |
| "grad_norm": 0.21857279539108276, | |
| "learning_rate": 4.1784641784641785e-06, | |
| "loss": 0.0016, | |
| "step": 8080 | |
| }, | |
| { | |
| "epoch": 31.0023166023166, | |
| "grad_norm": 0.013827860355377197, | |
| "learning_rate": 4.16988416988417e-06, | |
| "loss": 0.1681, | |
| "step": 8090 | |
| }, | |
| { | |
| "epoch": 31.003088803088804, | |
| "grad_norm": 0.16716662049293518, | |
| "learning_rate": 4.1613041613041615e-06, | |
| "loss": 0.4215, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 31.003861003861005, | |
| "grad_norm": 58.75962448120117, | |
| "learning_rate": 4.152724152724153e-06, | |
| "loss": 0.03, | |
| "step": 8110 | |
| }, | |
| { | |
| "epoch": 31.004633204633205, | |
| "grad_norm": 70.25969696044922, | |
| "learning_rate": 4.1441441441441446e-06, | |
| "loss": 0.1636, | |
| "step": 8120 | |
| }, | |
| { | |
| "epoch": 31.005405405405405, | |
| "grad_norm": 0.07677017897367477, | |
| "learning_rate": 4.135564135564136e-06, | |
| "loss": 0.2356, | |
| "step": 8130 | |
| }, | |
| { | |
| "epoch": 31.006177606177605, | |
| "grad_norm": 0.009610922075808048, | |
| "learning_rate": 4.126984126984127e-06, | |
| "loss": 0.0007, | |
| "step": 8140 | |
| }, | |
| { | |
| "epoch": 31.006949806949805, | |
| "grad_norm": 0.0572146512567997, | |
| "learning_rate": 4.118404118404119e-06, | |
| "loss": 0.3287, | |
| "step": 8150 | |
| }, | |
| { | |
| "epoch": 31.00772200772201, | |
| "grad_norm": 0.07425564527511597, | |
| "learning_rate": 4.109824109824111e-06, | |
| "loss": 0.0168, | |
| "step": 8160 | |
| }, | |
| { | |
| "epoch": 31.00849420849421, | |
| "grad_norm": 0.026679538190364838, | |
| "learning_rate": 4.101244101244102e-06, | |
| "loss": 0.1668, | |
| "step": 8170 | |
| }, | |
| { | |
| "epoch": 31.00926640926641, | |
| "grad_norm": 0.0018737409263849258, | |
| "learning_rate": 4.092664092664093e-06, | |
| "loss": 0.0066, | |
| "step": 8180 | |
| }, | |
| { | |
| "epoch": 31.01003861003861, | |
| "grad_norm": 396.3301696777344, | |
| "learning_rate": 4.084084084084085e-06, | |
| "loss": 0.3928, | |
| "step": 8190 | |
| }, | |
| { | |
| "epoch": 31.01081081081081, | |
| "grad_norm": 0.6448129415512085, | |
| "learning_rate": 4.075504075504076e-06, | |
| "loss": 0.2184, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 31.01158301158301, | |
| "grad_norm": 0.034431375563144684, | |
| "learning_rate": 4.066924066924067e-06, | |
| "loss": 0.0284, | |
| "step": 8210 | |
| }, | |
| { | |
| "epoch": 31.012355212355214, | |
| "grad_norm": 0.013946445658802986, | |
| "learning_rate": 4.058344058344059e-06, | |
| "loss": 0.001, | |
| "step": 8220 | |
| }, | |
| { | |
| "epoch": 31.013127413127414, | |
| "grad_norm": 0.015018833801150322, | |
| "learning_rate": 4.04976404976405e-06, | |
| "loss": 0.0798, | |
| "step": 8230 | |
| }, | |
| { | |
| "epoch": 31.013899613899614, | |
| "grad_norm": 0.08934608846902847, | |
| "learning_rate": 4.041184041184041e-06, | |
| "loss": 0.0119, | |
| "step": 8240 | |
| }, | |
| { | |
| "epoch": 31.014671814671814, | |
| "grad_norm": 0.25796929001808167, | |
| "learning_rate": 4.032604032604033e-06, | |
| "loss": 0.1963, | |
| "step": 8250 | |
| }, | |
| { | |
| "epoch": 31.015444015444015, | |
| "grad_norm": 0.0014374108286574483, | |
| "learning_rate": 4.024024024024024e-06, | |
| "loss": 0.2168, | |
| "step": 8260 | |
| }, | |
| { | |
| "epoch": 31.016216216216215, | |
| "grad_norm": 0.03616923466324806, | |
| "learning_rate": 4.015444015444015e-06, | |
| "loss": 0.0099, | |
| "step": 8270 | |
| }, | |
| { | |
| "epoch": 31.01698841698842, | |
| "grad_norm": 0.26436764001846313, | |
| "learning_rate": 4.006864006864007e-06, | |
| "loss": 0.2602, | |
| "step": 8280 | |
| }, | |
| { | |
| "epoch": 31.01776061776062, | |
| "grad_norm": 199.53219604492188, | |
| "learning_rate": 3.998283998283999e-06, | |
| "loss": 0.6799, | |
| "step": 8290 | |
| }, | |
| { | |
| "epoch": 31.01853281853282, | |
| "grad_norm": 0.08120285719633102, | |
| "learning_rate": 3.98970398970399e-06, | |
| "loss": 0.4377, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 31.01930501930502, | |
| "grad_norm": 0.0741524025797844, | |
| "learning_rate": 3.9811239811239814e-06, | |
| "loss": 0.3345, | |
| "step": 8310 | |
| }, | |
| { | |
| "epoch": 31.02007722007722, | |
| "grad_norm": 0.04211797937750816, | |
| "learning_rate": 3.9725439725439725e-06, | |
| "loss": 0.0334, | |
| "step": 8320 | |
| }, | |
| { | |
| "epoch": 31.02007722007722, | |
| "eval_accuracy": 0.7111111111111111, | |
| "eval_loss": 2.034332513809204, | |
| "eval_runtime": 12.4294, | |
| "eval_samples_per_second": 3.62, | |
| "eval_steps_per_second": 1.207, | |
| "step": 8320 | |
| }, | |
| { | |
| "epoch": 32.000772200772204, | |
| "grad_norm": 0.12727074325084686, | |
| "learning_rate": 3.9639639639639645e-06, | |
| "loss": 0.0861, | |
| "step": 8330 | |
| }, | |
| { | |
| "epoch": 32.0015444015444, | |
| "grad_norm": 0.040023840963840485, | |
| "learning_rate": 3.955383955383956e-06, | |
| "loss": 0.0052, | |
| "step": 8340 | |
| }, | |
| { | |
| "epoch": 32.002316602316604, | |
| "grad_norm": 0.040361110121011734, | |
| "learning_rate": 3.946803946803947e-06, | |
| "loss": 0.3076, | |
| "step": 8350 | |
| }, | |
| { | |
| "epoch": 32.0030888030888, | |
| "grad_norm": 0.04353674501180649, | |
| "learning_rate": 3.938223938223939e-06, | |
| "loss": 0.0807, | |
| "step": 8360 | |
| }, | |
| { | |
| "epoch": 32.003861003861005, | |
| "grad_norm": 0.010240713134407997, | |
| "learning_rate": 3.92964392964393e-06, | |
| "loss": 0.2488, | |
| "step": 8370 | |
| }, | |
| { | |
| "epoch": 32.0046332046332, | |
| "grad_norm": 0.0010769086657091975, | |
| "learning_rate": 3.921063921063921e-06, | |
| "loss": 0.1591, | |
| "step": 8380 | |
| }, | |
| { | |
| "epoch": 32.005405405405405, | |
| "grad_norm": 0.037674229592084885, | |
| "learning_rate": 3.912483912483913e-06, | |
| "loss": 0.0437, | |
| "step": 8390 | |
| }, | |
| { | |
| "epoch": 32.00617760617761, | |
| "grad_norm": 0.005637090187519789, | |
| "learning_rate": 3.903903903903904e-06, | |
| "loss": 0.0042, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 32.006949806949805, | |
| "grad_norm": 0.07691396027803421, | |
| "learning_rate": 3.895323895323896e-06, | |
| "loss": 0.763, | |
| "step": 8410 | |
| }, | |
| { | |
| "epoch": 32.00772200772201, | |
| "grad_norm": 0.07949383556842804, | |
| "learning_rate": 3.886743886743887e-06, | |
| "loss": 0.006, | |
| "step": 8420 | |
| }, | |
| { | |
| "epoch": 32.008494208494206, | |
| "grad_norm": 0.019564570859074593, | |
| "learning_rate": 3.878163878163879e-06, | |
| "loss": 0.0007, | |
| "step": 8430 | |
| }, | |
| { | |
| "epoch": 32.00926640926641, | |
| "grad_norm": 0.03286294266581535, | |
| "learning_rate": 3.86958386958387e-06, | |
| "loss": 0.2045, | |
| "step": 8440 | |
| }, | |
| { | |
| "epoch": 32.01003861003861, | |
| "grad_norm": 0.004159833304584026, | |
| "learning_rate": 3.861003861003861e-06, | |
| "loss": 0.2489, | |
| "step": 8450 | |
| }, | |
| { | |
| "epoch": 32.01081081081081, | |
| "grad_norm": 0.001372206723317504, | |
| "learning_rate": 3.852423852423853e-06, | |
| "loss": 0.0011, | |
| "step": 8460 | |
| }, | |
| { | |
| "epoch": 32.011583011583014, | |
| "grad_norm": 0.06818258762359619, | |
| "learning_rate": 3.843843843843844e-06, | |
| "loss": 0.0446, | |
| "step": 8470 | |
| }, | |
| { | |
| "epoch": 32.01235521235521, | |
| "grad_norm": 0.010435489006340504, | |
| "learning_rate": 3.835263835263835e-06, | |
| "loss": 0.0007, | |
| "step": 8480 | |
| }, | |
| { | |
| "epoch": 32.013127413127414, | |
| "grad_norm": 0.09285730123519897, | |
| "learning_rate": 3.826683826683827e-06, | |
| "loss": 0.0655, | |
| "step": 8490 | |
| }, | |
| { | |
| "epoch": 32.01389961389961, | |
| "grad_norm": 0.007980815134942532, | |
| "learning_rate": 3.818103818103818e-06, | |
| "loss": 0.2478, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 32.014671814671814, | |
| "grad_norm": 268.385009765625, | |
| "learning_rate": 3.80952380952381e-06, | |
| "loss": 0.1007, | |
| "step": 8510 | |
| }, | |
| { | |
| "epoch": 32.01544401544402, | |
| "grad_norm": 4.248262882232666, | |
| "learning_rate": 3.800943800943801e-06, | |
| "loss": 0.29, | |
| "step": 8520 | |
| }, | |
| { | |
| "epoch": 32.016216216216215, | |
| "grad_norm": 0.01907108724117279, | |
| "learning_rate": 3.792363792363793e-06, | |
| "loss": 0.0009, | |
| "step": 8530 | |
| }, | |
| { | |
| "epoch": 32.01698841698842, | |
| "grad_norm": 5.199569225311279, | |
| "learning_rate": 3.7837837837837844e-06, | |
| "loss": 0.4102, | |
| "step": 8540 | |
| }, | |
| { | |
| "epoch": 32.017760617760615, | |
| "grad_norm": 0.012655259110033512, | |
| "learning_rate": 3.7752037752037755e-06, | |
| "loss": 0.3693, | |
| "step": 8550 | |
| }, | |
| { | |
| "epoch": 32.01853281853282, | |
| "grad_norm": 0.011323108337819576, | |
| "learning_rate": 3.766623766623767e-06, | |
| "loss": 0.0059, | |
| "step": 8560 | |
| }, | |
| { | |
| "epoch": 32.01930501930502, | |
| "grad_norm": 0.11602991819381714, | |
| "learning_rate": 3.7580437580437585e-06, | |
| "loss": 0.267, | |
| "step": 8570 | |
| }, | |
| { | |
| "epoch": 32.02007722007722, | |
| "grad_norm": 0.0008211358217522502, | |
| "learning_rate": 3.7494637494637496e-06, | |
| "loss": 0.0043, | |
| "step": 8580 | |
| }, | |
| { | |
| "epoch": 32.02007722007722, | |
| "eval_accuracy": 0.7333333333333333, | |
| "eval_loss": 1.8095158338546753, | |
| "eval_runtime": 11.6106, | |
| "eval_samples_per_second": 3.876, | |
| "eval_steps_per_second": 1.292, | |
| "step": 8580 | |
| }, | |
| { | |
| "epoch": 33.000772200772204, | |
| "grad_norm": 15.456527709960938, | |
| "learning_rate": 3.740883740883741e-06, | |
| "loss": 0.8774, | |
| "step": 8590 | |
| }, | |
| { | |
| "epoch": 33.0015444015444, | |
| "grad_norm": 0.003137608291581273, | |
| "learning_rate": 3.7323037323037327e-06, | |
| "loss": 0.0006, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 33.002316602316604, | |
| "grad_norm": 0.005284165497869253, | |
| "learning_rate": 3.723723723723724e-06, | |
| "loss": 0.0122, | |
| "step": 8610 | |
| }, | |
| { | |
| "epoch": 33.0030888030888, | |
| "grad_norm": 0.0019698645919561386, | |
| "learning_rate": 3.7151437151437153e-06, | |
| "loss": 0.4956, | |
| "step": 8620 | |
| }, | |
| { | |
| "epoch": 33.003861003861005, | |
| "grad_norm": 0.050644420087337494, | |
| "learning_rate": 3.706563706563707e-06, | |
| "loss": 0.0333, | |
| "step": 8630 | |
| }, | |
| { | |
| "epoch": 33.0046332046332, | |
| "grad_norm": 0.004142272751778364, | |
| "learning_rate": 3.697983697983698e-06, | |
| "loss": 0.0269, | |
| "step": 8640 | |
| }, | |
| { | |
| "epoch": 33.005405405405405, | |
| "grad_norm": 0.009355615824460983, | |
| "learning_rate": 3.68940368940369e-06, | |
| "loss": 0.4925, | |
| "step": 8650 | |
| }, | |
| { | |
| "epoch": 33.00617760617761, | |
| "grad_norm": 0.29982513189315796, | |
| "learning_rate": 3.6808236808236814e-06, | |
| "loss": 0.0305, | |
| "step": 8660 | |
| }, | |
| { | |
| "epoch": 33.006949806949805, | |
| "grad_norm": 0.11938207596540451, | |
| "learning_rate": 3.6722436722436725e-06, | |
| "loss": 0.1785, | |
| "step": 8670 | |
| }, | |
| { | |
| "epoch": 33.00772200772201, | |
| "grad_norm": 0.45936429500579834, | |
| "learning_rate": 3.663663663663664e-06, | |
| "loss": 0.1534, | |
| "step": 8680 | |
| }, | |
| { | |
| "epoch": 33.008494208494206, | |
| "grad_norm": 0.004370039328932762, | |
| "learning_rate": 3.6550836550836556e-06, | |
| "loss": 0.259, | |
| "step": 8690 | |
| }, | |
| { | |
| "epoch": 33.00926640926641, | |
| "grad_norm": 0.05320889130234718, | |
| "learning_rate": 3.6465036465036467e-06, | |
| "loss": 0.0851, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 33.01003861003861, | |
| "grad_norm": 0.026938198134303093, | |
| "learning_rate": 3.637923637923638e-06, | |
| "loss": 0.4497, | |
| "step": 8710 | |
| }, | |
| { | |
| "epoch": 33.01081081081081, | |
| "grad_norm": 0.06186371669173241, | |
| "learning_rate": 3.6293436293436297e-06, | |
| "loss": 0.0011, | |
| "step": 8720 | |
| }, | |
| { | |
| "epoch": 33.011583011583014, | |
| "grad_norm": 0.11341995745897293, | |
| "learning_rate": 3.620763620763621e-06, | |
| "loss": 0.1477, | |
| "step": 8730 | |
| }, | |
| { | |
| "epoch": 33.01235521235521, | |
| "grad_norm": 0.019785890355706215, | |
| "learning_rate": 3.6121836121836124e-06, | |
| "loss": 0.0007, | |
| "step": 8740 | |
| }, | |
| { | |
| "epoch": 33.013127413127414, | |
| "grad_norm": 0.03372610732913017, | |
| "learning_rate": 3.603603603603604e-06, | |
| "loss": 0.0823, | |
| "step": 8750 | |
| }, | |
| { | |
| "epoch": 33.01389961389961, | |
| "grad_norm": 0.011712102219462395, | |
| "learning_rate": 3.595023595023595e-06, | |
| "loss": 0.0005, | |
| "step": 8760 | |
| }, | |
| { | |
| "epoch": 33.014671814671814, | |
| "grad_norm": 0.004663593135774136, | |
| "learning_rate": 3.5864435864435865e-06, | |
| "loss": 0.0046, | |
| "step": 8770 | |
| }, | |
| { | |
| "epoch": 33.01544401544402, | |
| "grad_norm": 0.007854433730244637, | |
| "learning_rate": 3.5778635778635785e-06, | |
| "loss": 0.2609, | |
| "step": 8780 | |
| }, | |
| { | |
| "epoch": 33.016216216216215, | |
| "grad_norm": 350.70220947265625, | |
| "learning_rate": 3.56928356928357e-06, | |
| "loss": 0.3562, | |
| "step": 8790 | |
| }, | |
| { | |
| "epoch": 33.01698841698842, | |
| "grad_norm": 0.017881860956549644, | |
| "learning_rate": 3.560703560703561e-06, | |
| "loss": 0.4138, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 33.017760617760615, | |
| "grad_norm": 0.02991180308163166, | |
| "learning_rate": 3.5521235521235526e-06, | |
| "loss": 0.0012, | |
| "step": 8810 | |
| }, | |
| { | |
| "epoch": 33.01853281853282, | |
| "grad_norm": 0.011691785417497158, | |
| "learning_rate": 3.5435435435435437e-06, | |
| "loss": 0.0353, | |
| "step": 8820 | |
| }, | |
| { | |
| "epoch": 33.01930501930502, | |
| "grad_norm": 0.011075496673583984, | |
| "learning_rate": 3.5349635349635352e-06, | |
| "loss": 0.0004, | |
| "step": 8830 | |
| }, | |
| { | |
| "epoch": 33.02007722007722, | |
| "grad_norm": 0.0012854086235165596, | |
| "learning_rate": 3.5263835263835268e-06, | |
| "loss": 0.0119, | |
| "step": 8840 | |
| }, | |
| { | |
| "epoch": 33.02007722007722, | |
| "eval_accuracy": 0.7555555555555555, | |
| "eval_loss": 1.6489903926849365, | |
| "eval_runtime": 11.5598, | |
| "eval_samples_per_second": 3.893, | |
| "eval_steps_per_second": 1.298, | |
| "step": 8840 | |
| }, | |
| { | |
| "epoch": 34.000772200772204, | |
| "grad_norm": 0.006392328068614006, | |
| "learning_rate": 3.517803517803518e-06, | |
| "loss": 0.0007, | |
| "step": 8850 | |
| }, | |
| { | |
| "epoch": 34.0015444015444, | |
| "grad_norm": 0.005775567144155502, | |
| "learning_rate": 3.5092235092235094e-06, | |
| "loss": 0.669, | |
| "step": 8860 | |
| }, | |
| { | |
| "epoch": 34.002316602316604, | |
| "grad_norm": 23.32708168029785, | |
| "learning_rate": 3.500643500643501e-06, | |
| "loss": 0.5035, | |
| "step": 8870 | |
| }, | |
| { | |
| "epoch": 34.0030888030888, | |
| "grad_norm": 0.010441070422530174, | |
| "learning_rate": 3.492063492063492e-06, | |
| "loss": 0.5568, | |
| "step": 8880 | |
| }, | |
| { | |
| "epoch": 34.003861003861005, | |
| "grad_norm": 0.021087242290377617, | |
| "learning_rate": 3.4834834834834835e-06, | |
| "loss": 0.0004, | |
| "step": 8890 | |
| }, | |
| { | |
| "epoch": 34.0046332046332, | |
| "grad_norm": 0.017975032329559326, | |
| "learning_rate": 3.4749034749034755e-06, | |
| "loss": 0.0058, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 34.005405405405405, | |
| "grad_norm": 0.027708925306797028, | |
| "learning_rate": 3.466323466323467e-06, | |
| "loss": 0.115, | |
| "step": 8910 | |
| }, | |
| { | |
| "epoch": 34.00617760617761, | |
| "grad_norm": 2.0851829051971436, | |
| "learning_rate": 3.457743457743458e-06, | |
| "loss": 0.2488, | |
| "step": 8920 | |
| }, | |
| { | |
| "epoch": 34.006949806949805, | |
| "grad_norm": 0.029099296778440475, | |
| "learning_rate": 3.4491634491634496e-06, | |
| "loss": 0.2251, | |
| "step": 8930 | |
| }, | |
| { | |
| "epoch": 34.00772200772201, | |
| "grad_norm": 0.021991893649101257, | |
| "learning_rate": 3.440583440583441e-06, | |
| "loss": 0.1571, | |
| "step": 8940 | |
| }, | |
| { | |
| "epoch": 34.008494208494206, | |
| "grad_norm": 0.06672017276287079, | |
| "learning_rate": 3.4320034320034323e-06, | |
| "loss": 0.0018, | |
| "step": 8950 | |
| }, | |
| { | |
| "epoch": 34.00926640926641, | |
| "grad_norm": 0.015447171404957771, | |
| "learning_rate": 3.423423423423424e-06, | |
| "loss": 0.1852, | |
| "step": 8960 | |
| }, | |
| { | |
| "epoch": 34.01003861003861, | |
| "grad_norm": 0.023431655019521713, | |
| "learning_rate": 3.414843414843415e-06, | |
| "loss": 0.3665, | |
| "step": 8970 | |
| }, | |
| { | |
| "epoch": 34.01081081081081, | |
| "grad_norm": 0.014014458283782005, | |
| "learning_rate": 3.4062634062634064e-06, | |
| "loss": 0.1518, | |
| "step": 8980 | |
| }, | |
| { | |
| "epoch": 34.011583011583014, | |
| "grad_norm": 0.015809480100870132, | |
| "learning_rate": 3.397683397683398e-06, | |
| "loss": 0.0686, | |
| "step": 8990 | |
| }, | |
| { | |
| "epoch": 34.01235521235521, | |
| "grad_norm": 0.018740158528089523, | |
| "learning_rate": 3.389103389103389e-06, | |
| "loss": 0.0008, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 34.013127413127414, | |
| "grad_norm": 71.28443908691406, | |
| "learning_rate": 3.3805233805233806e-06, | |
| "loss": 0.2821, | |
| "step": 9010 | |
| }, | |
| { | |
| "epoch": 34.01389961389961, | |
| "grad_norm": 0.03172178566455841, | |
| "learning_rate": 3.371943371943372e-06, | |
| "loss": 0.1551, | |
| "step": 9020 | |
| }, | |
| { | |
| "epoch": 34.014671814671814, | |
| "grad_norm": 0.09424544870853424, | |
| "learning_rate": 3.363363363363364e-06, | |
| "loss": 0.001, | |
| "step": 9030 | |
| }, | |
| { | |
| "epoch": 34.01544401544402, | |
| "grad_norm": 0.0008962831343524158, | |
| "learning_rate": 3.354783354783355e-06, | |
| "loss": 0.1171, | |
| "step": 9040 | |
| }, | |
| { | |
| "epoch": 34.016216216216215, | |
| "grad_norm": 172.4552764892578, | |
| "learning_rate": 3.3462033462033467e-06, | |
| "loss": 0.1849, | |
| "step": 9050 | |
| }, | |
| { | |
| "epoch": 34.01698841698842, | |
| "grad_norm": 0.0054281530901789665, | |
| "learning_rate": 3.337623337623338e-06, | |
| "loss": 0.1935, | |
| "step": 9060 | |
| }, | |
| { | |
| "epoch": 34.017760617760615, | |
| "grad_norm": 39.82855987548828, | |
| "learning_rate": 3.3290433290433293e-06, | |
| "loss": 0.3482, | |
| "step": 9070 | |
| }, | |
| { | |
| "epoch": 34.01853281853282, | |
| "grad_norm": 0.02182273380458355, | |
| "learning_rate": 3.320463320463321e-06, | |
| "loss": 0.0011, | |
| "step": 9080 | |
| }, | |
| { | |
| "epoch": 34.01930501930502, | |
| "grad_norm": 0.0020787569228559732, | |
| "learning_rate": 3.311883311883312e-06, | |
| "loss": 0.4725, | |
| "step": 9090 | |
| }, | |
| { | |
| "epoch": 34.02007722007722, | |
| "grad_norm": 0.002378948265686631, | |
| "learning_rate": 3.3033033033033035e-06, | |
| "loss": 0.7503, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 34.02007722007722, | |
| "eval_accuracy": 0.6888888888888889, | |
| "eval_loss": 1.9143760204315186, | |
| "eval_runtime": 11.5549, | |
| "eval_samples_per_second": 3.894, | |
| "eval_steps_per_second": 1.298, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 35.000772200772204, | |
| "grad_norm": 0.024031909182667732, | |
| "learning_rate": 3.294723294723295e-06, | |
| "loss": 0.2538, | |
| "step": 9110 | |
| }, | |
| { | |
| "epoch": 35.0015444015444, | |
| "grad_norm": 0.2591168284416199, | |
| "learning_rate": 3.286143286143286e-06, | |
| "loss": 0.0076, | |
| "step": 9120 | |
| }, | |
| { | |
| "epoch": 35.002316602316604, | |
| "grad_norm": 0.015992112457752228, | |
| "learning_rate": 3.2775632775632776e-06, | |
| "loss": 0.2147, | |
| "step": 9130 | |
| }, | |
| { | |
| "epoch": 35.0030888030888, | |
| "grad_norm": 0.01742151565849781, | |
| "learning_rate": 3.268983268983269e-06, | |
| "loss": 0.2486, | |
| "step": 9140 | |
| }, | |
| { | |
| "epoch": 35.003861003861005, | |
| "grad_norm": 0.025499412789940834, | |
| "learning_rate": 3.260403260403261e-06, | |
| "loss": 0.0021, | |
| "step": 9150 | |
| }, | |
| { | |
| "epoch": 35.0046332046332, | |
| "grad_norm": 0.019917473196983337, | |
| "learning_rate": 3.251823251823252e-06, | |
| "loss": 0.1993, | |
| "step": 9160 | |
| }, | |
| { | |
| "epoch": 35.005405405405405, | |
| "grad_norm": 0.07677433639764786, | |
| "learning_rate": 3.2432432432432437e-06, | |
| "loss": 0.0006, | |
| "step": 9170 | |
| }, | |
| { | |
| "epoch": 35.00617760617761, | |
| "grad_norm": 3.4880127906799316, | |
| "learning_rate": 3.2346632346632352e-06, | |
| "loss": 0.0735, | |
| "step": 9180 | |
| }, | |
| { | |
| "epoch": 35.006949806949805, | |
| "grad_norm": 0.0009284795960411429, | |
| "learning_rate": 3.2260832260832263e-06, | |
| "loss": 0.2833, | |
| "step": 9190 | |
| }, | |
| { | |
| "epoch": 35.00772200772201, | |
| "grad_norm": 0.032943326979875565, | |
| "learning_rate": 3.217503217503218e-06, | |
| "loss": 0.2312, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 35.008494208494206, | |
| "grad_norm": 30.327800750732422, | |
| "learning_rate": 3.2089232089232094e-06, | |
| "loss": 0.4013, | |
| "step": 9210 | |
| }, | |
| { | |
| "epoch": 35.00926640926641, | |
| "grad_norm": 25.810949325561523, | |
| "learning_rate": 3.2003432003432005e-06, | |
| "loss": 0.0023, | |
| "step": 9220 | |
| }, | |
| { | |
| "epoch": 35.01003861003861, | |
| "grad_norm": 0.04931256175041199, | |
| "learning_rate": 3.191763191763192e-06, | |
| "loss": 0.1492, | |
| "step": 9230 | |
| }, | |
| { | |
| "epoch": 35.01081081081081, | |
| "grad_norm": 0.13327378034591675, | |
| "learning_rate": 3.183183183183183e-06, | |
| "loss": 0.1499, | |
| "step": 9240 | |
| }, | |
| { | |
| "epoch": 35.011583011583014, | |
| "grad_norm": 0.0011244782945141196, | |
| "learning_rate": 3.1746031746031746e-06, | |
| "loss": 0.002, | |
| "step": 9250 | |
| }, | |
| { | |
| "epoch": 35.01235521235521, | |
| "grad_norm": 0.0651550218462944, | |
| "learning_rate": 3.166023166023166e-06, | |
| "loss": 0.1949, | |
| "step": 9260 | |
| }, | |
| { | |
| "epoch": 35.013127413127414, | |
| "grad_norm": 0.05204087868332863, | |
| "learning_rate": 3.1574431574431573e-06, | |
| "loss": 0.2826, | |
| "step": 9270 | |
| }, | |
| { | |
| "epoch": 35.01389961389961, | |
| "grad_norm": 0.3644317388534546, | |
| "learning_rate": 3.148863148863149e-06, | |
| "loss": 0.0004, | |
| "step": 9280 | |
| }, | |
| { | |
| "epoch": 35.014671814671814, | |
| "grad_norm": 0.010621222667396069, | |
| "learning_rate": 3.1402831402831407e-06, | |
| "loss": 0.0004, | |
| "step": 9290 | |
| }, | |
| { | |
| "epoch": 35.01544401544402, | |
| "grad_norm": 0.006138904020190239, | |
| "learning_rate": 3.1317031317031323e-06, | |
| "loss": 0.2331, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 35.016216216216215, | |
| "grad_norm": 0.00993597973138094, | |
| "learning_rate": 3.1231231231231234e-06, | |
| "loss": 0.0003, | |
| "step": 9310 | |
| }, | |
| { | |
| "epoch": 35.01698841698842, | |
| "grad_norm": 0.0026269257068634033, | |
| "learning_rate": 3.114543114543115e-06, | |
| "loss": 0.0004, | |
| "step": 9320 | |
| }, | |
| { | |
| "epoch": 35.017760617760615, | |
| "grad_norm": 0.018364176154136658, | |
| "learning_rate": 3.1059631059631064e-06, | |
| "loss": 0.2907, | |
| "step": 9330 | |
| }, | |
| { | |
| "epoch": 35.01853281853282, | |
| "grad_norm": 0.15101493895053864, | |
| "learning_rate": 3.0973830973830975e-06, | |
| "loss": 0.3308, | |
| "step": 9340 | |
| }, | |
| { | |
| "epoch": 35.01930501930502, | |
| "grad_norm": 31.66868782043457, | |
| "learning_rate": 3.088803088803089e-06, | |
| "loss": 0.4634, | |
| "step": 9350 | |
| }, | |
| { | |
| "epoch": 35.02007722007722, | |
| "grad_norm": 0.12729884684085846, | |
| "learning_rate": 3.0802230802230806e-06, | |
| "loss": 0.2105, | |
| "step": 9360 | |
| }, | |
| { | |
| "epoch": 35.02007722007722, | |
| "eval_accuracy": 0.7333333333333333, | |
| "eval_loss": 1.8402609825134277, | |
| "eval_runtime": 12.2345, | |
| "eval_samples_per_second": 3.678, | |
| "eval_steps_per_second": 1.226, | |
| "step": 9360 | |
| }, | |
| { | |
| "epoch": 36.000772200772204, | |
| "grad_norm": 0.0169172715395689, | |
| "learning_rate": 3.0716430716430717e-06, | |
| "loss": 0.0842, | |
| "step": 9370 | |
| }, | |
| { | |
| "epoch": 36.0015444015444, | |
| "grad_norm": 228.06277465820312, | |
| "learning_rate": 3.063063063063063e-06, | |
| "loss": 0.3631, | |
| "step": 9380 | |
| }, | |
| { | |
| "epoch": 36.002316602316604, | |
| "grad_norm": 0.5954656004905701, | |
| "learning_rate": 3.0544830544830543e-06, | |
| "loss": 0.0006, | |
| "step": 9390 | |
| }, | |
| { | |
| "epoch": 36.0030888030888, | |
| "grad_norm": 0.026740454137325287, | |
| "learning_rate": 3.0459030459030462e-06, | |
| "loss": 0.2786, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 36.003861003861005, | |
| "grad_norm": 0.010428531095385551, | |
| "learning_rate": 3.0373230373230378e-06, | |
| "loss": 0.0411, | |
| "step": 9410 | |
| }, | |
| { | |
| "epoch": 36.0046332046332, | |
| "grad_norm": 0.29002007842063904, | |
| "learning_rate": 3.0287430287430293e-06, | |
| "loss": 0.0214, | |
| "step": 9420 | |
| }, | |
| { | |
| "epoch": 36.005405405405405, | |
| "grad_norm": 0.8424767255783081, | |
| "learning_rate": 3.0201630201630204e-06, | |
| "loss": 0.0016, | |
| "step": 9430 | |
| }, | |
| { | |
| "epoch": 36.00617760617761, | |
| "grad_norm": 0.014938040636479855, | |
| "learning_rate": 3.011583011583012e-06, | |
| "loss": 0.0003, | |
| "step": 9440 | |
| }, | |
| { | |
| "epoch": 36.006949806949805, | |
| "grad_norm": 0.014550106599926949, | |
| "learning_rate": 3.0030030030030034e-06, | |
| "loss": 0.5971, | |
| "step": 9450 | |
| }, | |
| { | |
| "epoch": 36.00772200772201, | |
| "grad_norm": 66.45316314697266, | |
| "learning_rate": 2.9944229944229945e-06, | |
| "loss": 0.1367, | |
| "step": 9460 | |
| }, | |
| { | |
| "epoch": 36.008494208494206, | |
| "grad_norm": 0.00797193218022585, | |
| "learning_rate": 2.985842985842986e-06, | |
| "loss": 0.1833, | |
| "step": 9470 | |
| }, | |
| { | |
| "epoch": 36.00926640926641, | |
| "grad_norm": 0.021717458963394165, | |
| "learning_rate": 2.9772629772629776e-06, | |
| "loss": 0.0004, | |
| "step": 9480 | |
| }, | |
| { | |
| "epoch": 36.01003861003861, | |
| "grad_norm": 98.05574035644531, | |
| "learning_rate": 2.9686829686829687e-06, | |
| "loss": 0.2001, | |
| "step": 9490 | |
| }, | |
| { | |
| "epoch": 36.01081081081081, | |
| "grad_norm": 163.5323486328125, | |
| "learning_rate": 2.9601029601029602e-06, | |
| "loss": 0.1822, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 36.011583011583014, | |
| "grad_norm": 0.009607790037989616, | |
| "learning_rate": 2.9515229515229517e-06, | |
| "loss": 0.0003, | |
| "step": 9510 | |
| }, | |
| { | |
| "epoch": 36.01235521235521, | |
| "grad_norm": 0.006017378065735102, | |
| "learning_rate": 2.942942942942943e-06, | |
| "loss": 0.0002, | |
| "step": 9520 | |
| }, | |
| { | |
| "epoch": 36.013127413127414, | |
| "grad_norm": 0.007440345361828804, | |
| "learning_rate": 2.934362934362935e-06, | |
| "loss": 0.0864, | |
| "step": 9530 | |
| }, | |
| { | |
| "epoch": 36.01389961389961, | |
| "grad_norm": 0.013611230067908764, | |
| "learning_rate": 2.9257829257829263e-06, | |
| "loss": 0.4486, | |
| "step": 9540 | |
| }, | |
| { | |
| "epoch": 36.014671814671814, | |
| "grad_norm": 0.0026765179354697466, | |
| "learning_rate": 2.9172029172029174e-06, | |
| "loss": 0.0003, | |
| "step": 9550 | |
| }, | |
| { | |
| "epoch": 36.01544401544402, | |
| "grad_norm": 0.029210234060883522, | |
| "learning_rate": 2.908622908622909e-06, | |
| "loss": 0.0003, | |
| "step": 9560 | |
| }, | |
| { | |
| "epoch": 36.016216216216215, | |
| "grad_norm": 11.608271598815918, | |
| "learning_rate": 2.9000429000429005e-06, | |
| "loss": 0.7841, | |
| "step": 9570 | |
| }, | |
| { | |
| "epoch": 36.01698841698842, | |
| "grad_norm": 0.0011675491696223617, | |
| "learning_rate": 2.8914628914628916e-06, | |
| "loss": 0.2966, | |
| "step": 9580 | |
| }, | |
| { | |
| "epoch": 36.017760617760615, | |
| "grad_norm": 46.25785827636719, | |
| "learning_rate": 2.882882882882883e-06, | |
| "loss": 0.5733, | |
| "step": 9590 | |
| }, | |
| { | |
| "epoch": 36.01853281853282, | |
| "grad_norm": 0.008772353641688824, | |
| "learning_rate": 2.8743028743028746e-06, | |
| "loss": 0.0949, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 36.01930501930502, | |
| "grad_norm": 3.046584129333496, | |
| "learning_rate": 2.8657228657228657e-06, | |
| "loss": 0.0009, | |
| "step": 9610 | |
| }, | |
| { | |
| "epoch": 36.02007722007722, | |
| "grad_norm": 0.05296040698885918, | |
| "learning_rate": 2.8571428571428573e-06, | |
| "loss": 0.003, | |
| "step": 9620 | |
| }, | |
| { | |
| "epoch": 36.02007722007722, | |
| "eval_accuracy": 0.7333333333333333, | |
| "eval_loss": 1.8769656419754028, | |
| "eval_runtime": 12.2555, | |
| "eval_samples_per_second": 3.672, | |
| "eval_steps_per_second": 1.224, | |
| "step": 9620 | |
| }, | |
| { | |
| "epoch": 37.000772200772204, | |
| "grad_norm": 0.017100481316447258, | |
| "learning_rate": 2.8485628485628488e-06, | |
| "loss": 0.0003, | |
| "step": 9630 | |
| }, | |
| { | |
| "epoch": 37.0015444015444, | |
| "grad_norm": 0.006559515371918678, | |
| "learning_rate": 2.83998283998284e-06, | |
| "loss": 0.0005, | |
| "step": 9640 | |
| }, | |
| { | |
| "epoch": 37.002316602316604, | |
| "grad_norm": 0.15823794901371002, | |
| "learning_rate": 2.831402831402832e-06, | |
| "loss": 0.2016, | |
| "step": 9650 | |
| }, | |
| { | |
| "epoch": 37.0030888030888, | |
| "grad_norm": 0.07216683030128479, | |
| "learning_rate": 2.8228228228228234e-06, | |
| "loss": 0.0141, | |
| "step": 9660 | |
| }, | |
| { | |
| "epoch": 37.003861003861005, | |
| "grad_norm": 6.186749458312988, | |
| "learning_rate": 2.8142428142428145e-06, | |
| "loss": 0.1256, | |
| "step": 9670 | |
| }, | |
| { | |
| "epoch": 37.0046332046332, | |
| "grad_norm": 0.005314590875059366, | |
| "learning_rate": 2.805662805662806e-06, | |
| "loss": 0.348, | |
| "step": 9680 | |
| }, | |
| { | |
| "epoch": 37.005405405405405, | |
| "grad_norm": 5.876717567443848, | |
| "learning_rate": 2.7970827970827975e-06, | |
| "loss": 0.2215, | |
| "step": 9690 | |
| }, | |
| { | |
| "epoch": 37.00617760617761, | |
| "grad_norm": 0.008445030078291893, | |
| "learning_rate": 2.7885027885027886e-06, | |
| "loss": 0.3307, | |
| "step": 9700 | |
| }, | |
| { | |
| "epoch": 37.006949806949805, | |
| "grad_norm": 0.07367029786109924, | |
| "learning_rate": 2.77992277992278e-06, | |
| "loss": 0.2974, | |
| "step": 9710 | |
| }, | |
| { | |
| "epoch": 37.00772200772201, | |
| "grad_norm": 0.016041463240981102, | |
| "learning_rate": 2.7713427713427717e-06, | |
| "loss": 0.3192, | |
| "step": 9720 | |
| }, | |
| { | |
| "epoch": 37.008494208494206, | |
| "grad_norm": 0.007501265965402126, | |
| "learning_rate": 2.7627627627627628e-06, | |
| "loss": 0.0006, | |
| "step": 9730 | |
| }, | |
| { | |
| "epoch": 37.00926640926641, | |
| "grad_norm": 0.02961740642786026, | |
| "learning_rate": 2.7541827541827543e-06, | |
| "loss": 0.0025, | |
| "step": 9740 | |
| }, | |
| { | |
| "epoch": 37.01003861003861, | |
| "grad_norm": 0.039396531879901886, | |
| "learning_rate": 2.745602745602746e-06, | |
| "loss": 0.1664, | |
| "step": 9750 | |
| }, | |
| { | |
| "epoch": 37.01081081081081, | |
| "grad_norm": 0.0014447583816945553, | |
| "learning_rate": 2.737022737022737e-06, | |
| "loss": 0.0036, | |
| "step": 9760 | |
| }, | |
| { | |
| "epoch": 37.011583011583014, | |
| "grad_norm": 0.005423523019999266, | |
| "learning_rate": 2.7284427284427284e-06, | |
| "loss": 0.0806, | |
| "step": 9770 | |
| }, | |
| { | |
| "epoch": 37.01235521235521, | |
| "grad_norm": 0.011083639226853848, | |
| "learning_rate": 2.7198627198627204e-06, | |
| "loss": 0.3024, | |
| "step": 9780 | |
| }, | |
| { | |
| "epoch": 37.013127413127414, | |
| "grad_norm": 1.4681493043899536, | |
| "learning_rate": 2.7112827112827115e-06, | |
| "loss": 0.0007, | |
| "step": 9790 | |
| }, | |
| { | |
| "epoch": 37.01389961389961, | |
| "grad_norm": 0.012962309643626213, | |
| "learning_rate": 2.702702702702703e-06, | |
| "loss": 0.0016, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 37.014671814671814, | |
| "grad_norm": 0.044124990701675415, | |
| "learning_rate": 2.6941226941226945e-06, | |
| "loss": 0.3081, | |
| "step": 9810 | |
| }, | |
| { | |
| "epoch": 37.01544401544402, | |
| "grad_norm": 0.019632622599601746, | |
| "learning_rate": 2.6855426855426856e-06, | |
| "loss": 0.1146, | |
| "step": 9820 | |
| }, | |
| { | |
| "epoch": 37.016216216216215, | |
| "grad_norm": 0.005031611770391464, | |
| "learning_rate": 2.676962676962677e-06, | |
| "loss": 0.0003, | |
| "step": 9830 | |
| }, | |
| { | |
| "epoch": 37.01698841698842, | |
| "grad_norm": 0.004604416433721781, | |
| "learning_rate": 2.6683826683826687e-06, | |
| "loss": 0.3262, | |
| "step": 9840 | |
| }, | |
| { | |
| "epoch": 37.017760617760615, | |
| "grad_norm": 0.0116556566208601, | |
| "learning_rate": 2.65980265980266e-06, | |
| "loss": 0.0015, | |
| "step": 9850 | |
| }, | |
| { | |
| "epoch": 37.01853281853282, | |
| "grad_norm": 0.030262470245361328, | |
| "learning_rate": 2.6512226512226513e-06, | |
| "loss": 0.2633, | |
| "step": 9860 | |
| }, | |
| { | |
| "epoch": 37.01930501930502, | |
| "grad_norm": 0.08174673467874527, | |
| "learning_rate": 2.642642642642643e-06, | |
| "loss": 0.0049, | |
| "step": 9870 | |
| }, | |
| { | |
| "epoch": 37.02007722007722, | |
| "grad_norm": 0.029679138213396072, | |
| "learning_rate": 2.634062634062634e-06, | |
| "loss": 0.1781, | |
| "step": 9880 | |
| }, | |
| { | |
| "epoch": 37.02007722007722, | |
| "eval_accuracy": 0.7333333333333333, | |
| "eval_loss": 1.863143801689148, | |
| "eval_runtime": 16.3783, | |
| "eval_samples_per_second": 2.748, | |
| "eval_steps_per_second": 0.916, | |
| "step": 9880 | |
| }, | |
| { | |
| "epoch": 38.000772200772204, | |
| "grad_norm": 0.009378015995025635, | |
| "learning_rate": 2.6254826254826255e-06, | |
| "loss": 0.0003, | |
| "step": 9890 | |
| }, | |
| { | |
| "epoch": 38.0015444015444, | |
| "grad_norm": 0.0015269636642187834, | |
| "learning_rate": 2.6169026169026174e-06, | |
| "loss": 0.0004, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 38.002316602316604, | |
| "grad_norm": 0.012068642303347588, | |
| "learning_rate": 2.608322608322609e-06, | |
| "loss": 0.3438, | |
| "step": 9910 | |
| }, | |
| { | |
| "epoch": 38.0030888030888, | |
| "grad_norm": 0.030750026926398277, | |
| "learning_rate": 2.5997425997426e-06, | |
| "loss": 0.001, | |
| "step": 9920 | |
| }, | |
| { | |
| "epoch": 38.003861003861005, | |
| "grad_norm": 0.016554566100239754, | |
| "learning_rate": 2.5911625911625916e-06, | |
| "loss": 0.082, | |
| "step": 9930 | |
| }, | |
| { | |
| "epoch": 38.0046332046332, | |
| "grad_norm": 0.012171209789812565, | |
| "learning_rate": 2.5825825825825827e-06, | |
| "loss": 0.2353, | |
| "step": 9940 | |
| }, | |
| { | |
| "epoch": 38.005405405405405, | |
| "grad_norm": 0.10157763212919235, | |
| "learning_rate": 2.574002574002574e-06, | |
| "loss": 0.3692, | |
| "step": 9950 | |
| }, | |
| { | |
| "epoch": 38.00617760617761, | |
| "grad_norm": 0.008098813705146313, | |
| "learning_rate": 2.5654225654225657e-06, | |
| "loss": 0.2255, | |
| "step": 9960 | |
| }, | |
| { | |
| "epoch": 38.006949806949805, | |
| "grad_norm": 0.31319352984428406, | |
| "learning_rate": 2.556842556842557e-06, | |
| "loss": 0.0032, | |
| "step": 9970 | |
| }, | |
| { | |
| "epoch": 38.00772200772201, | |
| "grad_norm": 17.32835578918457, | |
| "learning_rate": 2.5482625482625484e-06, | |
| "loss": 0.0031, | |
| "step": 9980 | |
| }, | |
| { | |
| "epoch": 38.008494208494206, | |
| "grad_norm": 0.004962326493114233, | |
| "learning_rate": 2.53968253968254e-06, | |
| "loss": 0.0882, | |
| "step": 9990 | |
| }, | |
| { | |
| "epoch": 38.00926640926641, | |
| "grad_norm": 0.007238659542053938, | |
| "learning_rate": 2.531102531102531e-06, | |
| "loss": 0.1908, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 38.01003861003861, | |
| "grad_norm": 0.003812581766396761, | |
| "learning_rate": 2.5225225225225225e-06, | |
| "loss": 0.0001, | |
| "step": 10010 | |
| }, | |
| { | |
| "epoch": 38.01081081081081, | |
| "grad_norm": 0.020626194775104523, | |
| "learning_rate": 2.513942513942514e-06, | |
| "loss": 0.2302, | |
| "step": 10020 | |
| }, | |
| { | |
| "epoch": 38.011583011583014, | |
| "grad_norm": 345.7996826171875, | |
| "learning_rate": 2.505362505362506e-06, | |
| "loss": 0.1319, | |
| "step": 10030 | |
| }, | |
| { | |
| "epoch": 38.01235521235521, | |
| "grad_norm": 0.0005508098402060568, | |
| "learning_rate": 2.496782496782497e-06, | |
| "loss": 0.0003, | |
| "step": 10040 | |
| }, | |
| { | |
| "epoch": 38.013127413127414, | |
| "grad_norm": 0.0054133799858391285, | |
| "learning_rate": 2.488202488202488e-06, | |
| "loss": 0.0004, | |
| "step": 10050 | |
| }, | |
| { | |
| "epoch": 38.01389961389961, | |
| "grad_norm": 0.025228584185242653, | |
| "learning_rate": 2.47962247962248e-06, | |
| "loss": 0.3296, | |
| "step": 10060 | |
| }, | |
| { | |
| "epoch": 38.014671814671814, | |
| "grad_norm": 0.04835504665970802, | |
| "learning_rate": 2.4710424710424712e-06, | |
| "loss": 0.0011, | |
| "step": 10070 | |
| }, | |
| { | |
| "epoch": 38.01544401544402, | |
| "grad_norm": 17.836328506469727, | |
| "learning_rate": 2.4624624624624628e-06, | |
| "loss": 0.0019, | |
| "step": 10080 | |
| }, | |
| { | |
| "epoch": 38.016216216216215, | |
| "grad_norm": 0.018395883962512016, | |
| "learning_rate": 2.453882453882454e-06, | |
| "loss": 0.0002, | |
| "step": 10090 | |
| }, | |
| { | |
| "epoch": 38.01698841698842, | |
| "grad_norm": 0.01676626317203045, | |
| "learning_rate": 2.4453024453024454e-06, | |
| "loss": 0.1144, | |
| "step": 10100 | |
| }, | |
| { | |
| "epoch": 38.017760617760615, | |
| "grad_norm": 0.007224744651466608, | |
| "learning_rate": 2.436722436722437e-06, | |
| "loss": 0.252, | |
| "step": 10110 | |
| }, | |
| { | |
| "epoch": 38.01853281853282, | |
| "grad_norm": 0.006745734717696905, | |
| "learning_rate": 2.4281424281424284e-06, | |
| "loss": 0.0003, | |
| "step": 10120 | |
| }, | |
| { | |
| "epoch": 38.01930501930502, | |
| "grad_norm": 0.012383128516376019, | |
| "learning_rate": 2.41956241956242e-06, | |
| "loss": 0.0005, | |
| "step": 10130 | |
| }, | |
| { | |
| "epoch": 38.02007722007722, | |
| "grad_norm": 0.010990212671458721, | |
| "learning_rate": 2.410982410982411e-06, | |
| "loss": 0.4092, | |
| "step": 10140 | |
| }, | |
| { | |
| "epoch": 38.02007722007722, | |
| "eval_accuracy": 0.7111111111111111, | |
| "eval_loss": 1.9994146823883057, | |
| "eval_runtime": 11.5839, | |
| "eval_samples_per_second": 3.885, | |
| "eval_steps_per_second": 1.295, | |
| "step": 10140 | |
| }, | |
| { | |
| "epoch": 39.000772200772204, | |
| "grad_norm": 0.0006115052383393049, | |
| "learning_rate": 2.4024024024024026e-06, | |
| "loss": 0.0001, | |
| "step": 10150 | |
| }, | |
| { | |
| "epoch": 39.0015444015444, | |
| "grad_norm": 0.01732112281024456, | |
| "learning_rate": 2.393822393822394e-06, | |
| "loss": 0.0002, | |
| "step": 10160 | |
| }, | |
| { | |
| "epoch": 39.002316602316604, | |
| "grad_norm": 388.23370361328125, | |
| "learning_rate": 2.3852423852423852e-06, | |
| "loss": 0.1257, | |
| "step": 10170 | |
| }, | |
| { | |
| "epoch": 39.0030888030888, | |
| "grad_norm": 0.04425964504480362, | |
| "learning_rate": 2.376662376662377e-06, | |
| "loss": 0.0003, | |
| "step": 10180 | |
| }, | |
| { | |
| "epoch": 39.003861003861005, | |
| "grad_norm": 0.02979341335594654, | |
| "learning_rate": 2.3680823680823683e-06, | |
| "loss": 0.0002, | |
| "step": 10190 | |
| }, | |
| { | |
| "epoch": 39.0046332046332, | |
| "grad_norm": 0.03975719213485718, | |
| "learning_rate": 2.35950235950236e-06, | |
| "loss": 0.4852, | |
| "step": 10200 | |
| }, | |
| { | |
| "epoch": 39.005405405405405, | |
| "grad_norm": 0.005834504030644894, | |
| "learning_rate": 2.3509223509223513e-06, | |
| "loss": 0.0002, | |
| "step": 10210 | |
| }, | |
| { | |
| "epoch": 39.00617760617761, | |
| "grad_norm": 0.0008963262080214918, | |
| "learning_rate": 2.3423423423423424e-06, | |
| "loss": 0.0004, | |
| "step": 10220 | |
| }, | |
| { | |
| "epoch": 39.006949806949805, | |
| "grad_norm": 0.2953839600086212, | |
| "learning_rate": 2.333762333762334e-06, | |
| "loss": 0.0002, | |
| "step": 10230 | |
| }, | |
| { | |
| "epoch": 39.00772200772201, | |
| "grad_norm": 0.005276846699416637, | |
| "learning_rate": 2.3251823251823255e-06, | |
| "loss": 0.0013, | |
| "step": 10240 | |
| }, | |
| { | |
| "epoch": 39.008494208494206, | |
| "grad_norm": 0.00223553623072803, | |
| "learning_rate": 2.316602316602317e-06, | |
| "loss": 0.2671, | |
| "step": 10250 | |
| }, | |
| { | |
| "epoch": 39.00926640926641, | |
| "grad_norm": 0.006985859479755163, | |
| "learning_rate": 2.308022308022308e-06, | |
| "loss": 0.3094, | |
| "step": 10260 | |
| }, | |
| { | |
| "epoch": 39.01003861003861, | |
| "grad_norm": 0.007334560621529818, | |
| "learning_rate": 2.2994422994422996e-06, | |
| "loss": 0.1669, | |
| "step": 10270 | |
| }, | |
| { | |
| "epoch": 39.01081081081081, | |
| "grad_norm": 0.02189444750547409, | |
| "learning_rate": 2.290862290862291e-06, | |
| "loss": 0.0002, | |
| "step": 10280 | |
| }, | |
| { | |
| "epoch": 39.011583011583014, | |
| "grad_norm": 0.002752017229795456, | |
| "learning_rate": 2.2822822822822822e-06, | |
| "loss": 0.0001, | |
| "step": 10290 | |
| }, | |
| { | |
| "epoch": 39.01235521235521, | |
| "grad_norm": 0.003848872147500515, | |
| "learning_rate": 2.2737022737022738e-06, | |
| "loss": 0.1079, | |
| "step": 10300 | |
| }, | |
| { | |
| "epoch": 39.013127413127414, | |
| "grad_norm": 0.022609086707234383, | |
| "learning_rate": 2.2651222651222653e-06, | |
| "loss": 0.0002, | |
| "step": 10310 | |
| }, | |
| { | |
| "epoch": 39.01389961389961, | |
| "grad_norm": 0.013301585800945759, | |
| "learning_rate": 2.256542256542257e-06, | |
| "loss": 0.1329, | |
| "step": 10320 | |
| }, | |
| { | |
| "epoch": 39.014671814671814, | |
| "grad_norm": 0.010520892217755318, | |
| "learning_rate": 2.2479622479622483e-06, | |
| "loss": 0.1651, | |
| "step": 10330 | |
| }, | |
| { | |
| "epoch": 39.01544401544402, | |
| "grad_norm": 0.01888640969991684, | |
| "learning_rate": 2.2393822393822394e-06, | |
| "loss": 0.0002, | |
| "step": 10340 | |
| }, | |
| { | |
| "epoch": 39.016216216216215, | |
| "grad_norm": 0.005222676321864128, | |
| "learning_rate": 2.230802230802231e-06, | |
| "loss": 0.0056, | |
| "step": 10350 | |
| }, | |
| { | |
| "epoch": 39.01698841698842, | |
| "grad_norm": 0.003466697409749031, | |
| "learning_rate": 2.222222222222222e-06, | |
| "loss": 0.1216, | |
| "step": 10360 | |
| }, | |
| { | |
| "epoch": 39.017760617760615, | |
| "grad_norm": 0.012701363302767277, | |
| "learning_rate": 2.213642213642214e-06, | |
| "loss": 0.2632, | |
| "step": 10370 | |
| }, | |
| { | |
| "epoch": 39.01853281853282, | |
| "grad_norm": 0.0034604999236762524, | |
| "learning_rate": 2.205062205062205e-06, | |
| "loss": 0.0002, | |
| "step": 10380 | |
| }, | |
| { | |
| "epoch": 39.01930501930502, | |
| "grad_norm": 0.22704817354679108, | |
| "learning_rate": 2.1964821964821967e-06, | |
| "loss": 0.1519, | |
| "step": 10390 | |
| }, | |
| { | |
| "epoch": 39.02007722007722, | |
| "grad_norm": 0.00676871370524168, | |
| "learning_rate": 2.187902187902188e-06, | |
| "loss": 0.232, | |
| "step": 10400 | |
| }, | |
| { | |
| "epoch": 39.02007722007722, | |
| "eval_accuracy": 0.6888888888888889, | |
| "eval_loss": 1.9918631315231323, | |
| "eval_runtime": 13.6157, | |
| "eval_samples_per_second": 3.305, | |
| "eval_steps_per_second": 1.102, | |
| "step": 10400 | |
| }, | |
| { | |
| "epoch": 40.000772200772204, | |
| "grad_norm": 0.0037049120292067528, | |
| "learning_rate": 2.1793221793221793e-06, | |
| "loss": 0.2593, | |
| "step": 10410 | |
| }, | |
| { | |
| "epoch": 40.0015444015444, | |
| "grad_norm": 0.007936818525195122, | |
| "learning_rate": 2.170742170742171e-06, | |
| "loss": 0.0004, | |
| "step": 10420 | |
| }, | |
| { | |
| "epoch": 40.002316602316604, | |
| "grad_norm": 0.008796462789177895, | |
| "learning_rate": 2.1621621621621623e-06, | |
| "loss": 0.0002, | |
| "step": 10430 | |
| }, | |
| { | |
| "epoch": 40.0030888030888, | |
| "grad_norm": 0.0008883124683052301, | |
| "learning_rate": 2.153582153582154e-06, | |
| "loss": 0.0003, | |
| "step": 10440 | |
| }, | |
| { | |
| "epoch": 40.003861003861005, | |
| "grad_norm": 0.021493464708328247, | |
| "learning_rate": 2.1450021450021454e-06, | |
| "loss": 0.0004, | |
| "step": 10450 | |
| }, | |
| { | |
| "epoch": 40.0046332046332, | |
| "grad_norm": 0.013466846197843552, | |
| "learning_rate": 2.1364221364221365e-06, | |
| "loss": 0.0003, | |
| "step": 10460 | |
| }, | |
| { | |
| "epoch": 40.005405405405405, | |
| "grad_norm": 0.0008845877600833774, | |
| "learning_rate": 2.127842127842128e-06, | |
| "loss": 0.0002, | |
| "step": 10470 | |
| }, | |
| { | |
| "epoch": 40.00617760617761, | |
| "grad_norm": 0.004395687952637672, | |
| "learning_rate": 2.1192621192621195e-06, | |
| "loss": 0.0388, | |
| "step": 10480 | |
| }, | |
| { | |
| "epoch": 40.006949806949805, | |
| "grad_norm": 0.0025481514167040586, | |
| "learning_rate": 2.110682110682111e-06, | |
| "loss": 0.0002, | |
| "step": 10490 | |
| }, | |
| { | |
| "epoch": 40.00772200772201, | |
| "grad_norm": 0.05023454874753952, | |
| "learning_rate": 2.102102102102102e-06, | |
| "loss": 0.0584, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 40.008494208494206, | |
| "grad_norm": 0.0015958435833454132, | |
| "learning_rate": 2.0935220935220937e-06, | |
| "loss": 0.0004, | |
| "step": 10510 | |
| }, | |
| { | |
| "epoch": 40.00926640926641, | |
| "grad_norm": 0.004534762352705002, | |
| "learning_rate": 2.084942084942085e-06, | |
| "loss": 0.001, | |
| "step": 10520 | |
| }, | |
| { | |
| "epoch": 40.01003861003861, | |
| "grad_norm": 0.6501943469047546, | |
| "learning_rate": 2.0763620763620763e-06, | |
| "loss": 0.0005, | |
| "step": 10530 | |
| }, | |
| { | |
| "epoch": 40.01081081081081, | |
| "grad_norm": 0.005923354998230934, | |
| "learning_rate": 2.067782067782068e-06, | |
| "loss": 0.4309, | |
| "step": 10540 | |
| }, | |
| { | |
| "epoch": 40.011583011583014, | |
| "grad_norm": 0.007845559157431126, | |
| "learning_rate": 2.0592020592020594e-06, | |
| "loss": 0.0478, | |
| "step": 10550 | |
| }, | |
| { | |
| "epoch": 40.01235521235521, | |
| "grad_norm": 0.0035400239285081625, | |
| "learning_rate": 2.050622050622051e-06, | |
| "loss": 0.0003, | |
| "step": 10560 | |
| }, | |
| { | |
| "epoch": 40.013127413127414, | |
| "grad_norm": 0.0006449856446124613, | |
| "learning_rate": 2.0420420420420424e-06, | |
| "loss": 0.1536, | |
| "step": 10570 | |
| }, | |
| { | |
| "epoch": 40.01389961389961, | |
| "grad_norm": 0.011894451454281807, | |
| "learning_rate": 2.0334620334620335e-06, | |
| "loss": 0.2582, | |
| "step": 10580 | |
| }, | |
| { | |
| "epoch": 40.014671814671814, | |
| "grad_norm": 0.35729271173477173, | |
| "learning_rate": 2.024882024882025e-06, | |
| "loss": 0.0004, | |
| "step": 10590 | |
| }, | |
| { | |
| "epoch": 40.01544401544402, | |
| "grad_norm": 410.946533203125, | |
| "learning_rate": 2.0163020163020166e-06, | |
| "loss": 0.3431, | |
| "step": 10600 | |
| }, | |
| { | |
| "epoch": 40.016216216216215, | |
| "grad_norm": 0.003937250003218651, | |
| "learning_rate": 2.0077220077220077e-06, | |
| "loss": 0.0001, | |
| "step": 10610 | |
| }, | |
| { | |
| "epoch": 40.01698841698842, | |
| "grad_norm": 38.59312438964844, | |
| "learning_rate": 1.9991419991419996e-06, | |
| "loss": 0.2471, | |
| "step": 10620 | |
| }, | |
| { | |
| "epoch": 40.017760617760615, | |
| "grad_norm": 2.9301681518554688, | |
| "learning_rate": 1.9905619905619907e-06, | |
| "loss": 0.0768, | |
| "step": 10630 | |
| }, | |
| { | |
| "epoch": 40.01853281853282, | |
| "grad_norm": 0.002492203377187252, | |
| "learning_rate": 1.9819819819819822e-06, | |
| "loss": 0.1803, | |
| "step": 10640 | |
| }, | |
| { | |
| "epoch": 40.01930501930502, | |
| "grad_norm": 0.02368726022541523, | |
| "learning_rate": 1.9734019734019733e-06, | |
| "loss": 0.1065, | |
| "step": 10650 | |
| }, | |
| { | |
| "epoch": 40.02007722007722, | |
| "grad_norm": 0.010710655711591244, | |
| "learning_rate": 1.964821964821965e-06, | |
| "loss": 0.2703, | |
| "step": 10660 | |
| }, | |
| { | |
| "epoch": 40.02007722007722, | |
| "eval_accuracy": 0.7111111111111111, | |
| "eval_loss": 2.10081148147583, | |
| "eval_runtime": 13.0367, | |
| "eval_samples_per_second": 3.452, | |
| "eval_steps_per_second": 1.151, | |
| "step": 10660 | |
| }, | |
| { | |
| "epoch": 41.000772200772204, | |
| "grad_norm": 0.003705637063831091, | |
| "learning_rate": 1.9562419562419564e-06, | |
| "loss": 0.4446, | |
| "step": 10670 | |
| }, | |
| { | |
| "epoch": 41.0015444015444, | |
| "grad_norm": 0.006835234817117453, | |
| "learning_rate": 1.947661947661948e-06, | |
| "loss": 0.1434, | |
| "step": 10680 | |
| }, | |
| { | |
| "epoch": 41.002316602316604, | |
| "grad_norm": 0.01129397563636303, | |
| "learning_rate": 1.9390819390819394e-06, | |
| "loss": 0.2595, | |
| "step": 10690 | |
| }, | |
| { | |
| "epoch": 41.0030888030888, | |
| "grad_norm": 0.11355545371770859, | |
| "learning_rate": 1.9305019305019305e-06, | |
| "loss": 0.0321, | |
| "step": 10700 | |
| }, | |
| { | |
| "epoch": 41.003861003861005, | |
| "grad_norm": 0.007223920896649361, | |
| "learning_rate": 1.921921921921922e-06, | |
| "loss": 0.0444, | |
| "step": 10710 | |
| }, | |
| { | |
| "epoch": 41.0046332046332, | |
| "grad_norm": 0.022122787311673164, | |
| "learning_rate": 1.9133419133419136e-06, | |
| "loss": 0.0556, | |
| "step": 10720 | |
| }, | |
| { | |
| "epoch": 41.005405405405405, | |
| "grad_norm": 0.0013790694065392017, | |
| "learning_rate": 1.904761904761905e-06, | |
| "loss": 0.0003, | |
| "step": 10730 | |
| }, | |
| { | |
| "epoch": 41.00617760617761, | |
| "grad_norm": 0.003318269969895482, | |
| "learning_rate": 1.8961818961818964e-06, | |
| "loss": 0.0006, | |
| "step": 10740 | |
| }, | |
| { | |
| "epoch": 41.006949806949805, | |
| "grad_norm": 0.06160283461213112, | |
| "learning_rate": 1.8876018876018877e-06, | |
| "loss": 0.1408, | |
| "step": 10750 | |
| }, | |
| { | |
| "epoch": 41.00772200772201, | |
| "grad_norm": 0.0031962695065885782, | |
| "learning_rate": 1.8790218790218793e-06, | |
| "loss": 0.345, | |
| "step": 10760 | |
| }, | |
| { | |
| "epoch": 41.008494208494206, | |
| "grad_norm": 0.008759942837059498, | |
| "learning_rate": 1.8704418704418706e-06, | |
| "loss": 0.0008, | |
| "step": 10770 | |
| }, | |
| { | |
| "epoch": 41.00926640926641, | |
| "grad_norm": 0.007954655215144157, | |
| "learning_rate": 1.861861861861862e-06, | |
| "loss": 0.0002, | |
| "step": 10780 | |
| }, | |
| { | |
| "epoch": 41.01003861003861, | |
| "grad_norm": 261.1466979980469, | |
| "learning_rate": 1.8532818532818534e-06, | |
| "loss": 0.2819, | |
| "step": 10790 | |
| }, | |
| { | |
| "epoch": 41.01081081081081, | |
| "grad_norm": 0.440922349691391, | |
| "learning_rate": 1.844701844701845e-06, | |
| "loss": 0.1343, | |
| "step": 10800 | |
| }, | |
| { | |
| "epoch": 41.011583011583014, | |
| "grad_norm": 0.02840583585202694, | |
| "learning_rate": 1.8361218361218363e-06, | |
| "loss": 0.0002, | |
| "step": 10810 | |
| }, | |
| { | |
| "epoch": 41.01235521235521, | |
| "grad_norm": 0.7776057720184326, | |
| "learning_rate": 1.8275418275418278e-06, | |
| "loss": 0.0344, | |
| "step": 10820 | |
| }, | |
| { | |
| "epoch": 41.013127413127414, | |
| "grad_norm": 0.0304377693682909, | |
| "learning_rate": 1.818961818961819e-06, | |
| "loss": 0.2329, | |
| "step": 10830 | |
| }, | |
| { | |
| "epoch": 41.01389961389961, | |
| "grad_norm": 0.03479522094130516, | |
| "learning_rate": 1.8103818103818104e-06, | |
| "loss": 0.1126, | |
| "step": 10840 | |
| }, | |
| { | |
| "epoch": 41.014671814671814, | |
| "grad_norm": 0.00659992964938283, | |
| "learning_rate": 1.801801801801802e-06, | |
| "loss": 0.0004, | |
| "step": 10850 | |
| }, | |
| { | |
| "epoch": 41.01544401544402, | |
| "grad_norm": 0.012349890545010567, | |
| "learning_rate": 1.7932217932217933e-06, | |
| "loss": 0.0134, | |
| "step": 10860 | |
| }, | |
| { | |
| "epoch": 41.016216216216215, | |
| "grad_norm": 0.017889924347400665, | |
| "learning_rate": 1.784641784641785e-06, | |
| "loss": 0.1972, | |
| "step": 10870 | |
| }, | |
| { | |
| "epoch": 41.01698841698842, | |
| "grad_norm": 0.07884511351585388, | |
| "learning_rate": 1.7760617760617763e-06, | |
| "loss": 0.0936, | |
| "step": 10880 | |
| }, | |
| { | |
| "epoch": 41.017760617760615, | |
| "grad_norm": 0.11204981803894043, | |
| "learning_rate": 1.7674817674817676e-06, | |
| "loss": 0.4485, | |
| "step": 10890 | |
| }, | |
| { | |
| "epoch": 41.01853281853282, | |
| "grad_norm": 0.005154389888048172, | |
| "learning_rate": 1.758901758901759e-06, | |
| "loss": 0.164, | |
| "step": 10900 | |
| }, | |
| { | |
| "epoch": 41.01930501930502, | |
| "grad_norm": 4.545119285583496, | |
| "learning_rate": 1.7503217503217505e-06, | |
| "loss": 0.1486, | |
| "step": 10910 | |
| }, | |
| { | |
| "epoch": 41.02007722007722, | |
| "grad_norm": 0.13940872251987457, | |
| "learning_rate": 1.7417417417417418e-06, | |
| "loss": 0.5169, | |
| "step": 10920 | |
| }, | |
| { | |
| "epoch": 41.02007722007722, | |
| "eval_accuracy": 0.6888888888888889, | |
| "eval_loss": 2.201923131942749, | |
| "eval_runtime": 11.5921, | |
| "eval_samples_per_second": 3.882, | |
| "eval_steps_per_second": 1.294, | |
| "step": 10920 | |
| }, | |
| { | |
| "epoch": 42.000772200772204, | |
| "grad_norm": 0.009260579943656921, | |
| "learning_rate": 1.7331617331617335e-06, | |
| "loss": 0.2545, | |
| "step": 10930 | |
| }, | |
| { | |
| "epoch": 42.0015444015444, | |
| "grad_norm": 0.01058141142129898, | |
| "learning_rate": 1.7245817245817248e-06, | |
| "loss": 0.0004, | |
| "step": 10940 | |
| }, | |
| { | |
| "epoch": 42.002316602316604, | |
| "grad_norm": 63.95749282836914, | |
| "learning_rate": 1.7160017160017161e-06, | |
| "loss": 0.1117, | |
| "step": 10950 | |
| }, | |
| { | |
| "epoch": 42.0030888030888, | |
| "grad_norm": 0.0007036329479888082, | |
| "learning_rate": 1.7074217074217074e-06, | |
| "loss": 0.0003, | |
| "step": 10960 | |
| }, | |
| { | |
| "epoch": 42.003861003861005, | |
| "grad_norm": 0.025101276114583015, | |
| "learning_rate": 1.698841698841699e-06, | |
| "loss": 0.1805, | |
| "step": 10970 | |
| }, | |
| { | |
| "epoch": 42.0046332046332, | |
| "grad_norm": 0.9655550122261047, | |
| "learning_rate": 1.6902616902616903e-06, | |
| "loss": 0.238, | |
| "step": 10980 | |
| }, | |
| { | |
| "epoch": 42.005405405405405, | |
| "grad_norm": 0.11792098730802536, | |
| "learning_rate": 1.681681681681682e-06, | |
| "loss": 0.2002, | |
| "step": 10990 | |
| }, | |
| { | |
| "epoch": 42.00617760617761, | |
| "grad_norm": 479.1819152832031, | |
| "learning_rate": 1.6731016731016733e-06, | |
| "loss": 0.1926, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 42.006949806949805, | |
| "grad_norm": 0.07615210115909576, | |
| "learning_rate": 1.6645216645216646e-06, | |
| "loss": 0.1492, | |
| "step": 11010 | |
| }, | |
| { | |
| "epoch": 42.00772200772201, | |
| "grad_norm": 0.004804663825780153, | |
| "learning_rate": 1.655941655941656e-06, | |
| "loss": 0.3703, | |
| "step": 11020 | |
| }, | |
| { | |
| "epoch": 42.008494208494206, | |
| "grad_norm": 0.006586713716387749, | |
| "learning_rate": 1.6473616473616475e-06, | |
| "loss": 0.0004, | |
| "step": 11030 | |
| }, | |
| { | |
| "epoch": 42.00926640926641, | |
| "grad_norm": 18.117666244506836, | |
| "learning_rate": 1.6387816387816388e-06, | |
| "loss": 0.007, | |
| "step": 11040 | |
| }, | |
| { | |
| "epoch": 42.01003861003861, | |
| "grad_norm": 0.028815440833568573, | |
| "learning_rate": 1.6302016302016305e-06, | |
| "loss": 0.2412, | |
| "step": 11050 | |
| }, | |
| { | |
| "epoch": 42.01081081081081, | |
| "grad_norm": 41.1140022277832, | |
| "learning_rate": 1.6216216216216219e-06, | |
| "loss": 0.2599, | |
| "step": 11060 | |
| }, | |
| { | |
| "epoch": 42.011583011583014, | |
| "grad_norm": 330.62469482421875, | |
| "learning_rate": 1.6130416130416132e-06, | |
| "loss": 0.2363, | |
| "step": 11070 | |
| }, | |
| { | |
| "epoch": 42.01235521235521, | |
| "grad_norm": 0.011965152807533741, | |
| "learning_rate": 1.6044616044616047e-06, | |
| "loss": 0.0008, | |
| "step": 11080 | |
| }, | |
| { | |
| "epoch": 42.013127413127414, | |
| "grad_norm": 81.2577896118164, | |
| "learning_rate": 1.595881595881596e-06, | |
| "loss": 0.0116, | |
| "step": 11090 | |
| }, | |
| { | |
| "epoch": 42.01389961389961, | |
| "grad_norm": 0.005599725525826216, | |
| "learning_rate": 1.5873015873015873e-06, | |
| "loss": 0.0003, | |
| "step": 11100 | |
| }, | |
| { | |
| "epoch": 42.014671814671814, | |
| "grad_norm": 0.03384312242269516, | |
| "learning_rate": 1.5787215787215786e-06, | |
| "loss": 0.1198, | |
| "step": 11110 | |
| }, | |
| { | |
| "epoch": 42.01544401544402, | |
| "grad_norm": 0.019553786143660545, | |
| "learning_rate": 1.5701415701415704e-06, | |
| "loss": 0.274, | |
| "step": 11120 | |
| }, | |
| { | |
| "epoch": 42.016216216216215, | |
| "grad_norm": 0.0044852448627352715, | |
| "learning_rate": 1.5615615615615617e-06, | |
| "loss": 0.2265, | |
| "step": 11130 | |
| }, | |
| { | |
| "epoch": 42.01698841698842, | |
| "grad_norm": 0.008225970901548862, | |
| "learning_rate": 1.5529815529815532e-06, | |
| "loss": 0.0055, | |
| "step": 11140 | |
| }, | |
| { | |
| "epoch": 42.017760617760615, | |
| "grad_norm": 0.006483301054686308, | |
| "learning_rate": 1.5444015444015445e-06, | |
| "loss": 0.0009, | |
| "step": 11150 | |
| }, | |
| { | |
| "epoch": 42.01853281853282, | |
| "grad_norm": 0.4023190140724182, | |
| "learning_rate": 1.5358215358215358e-06, | |
| "loss": 0.226, | |
| "step": 11160 | |
| }, | |
| { | |
| "epoch": 42.01930501930502, | |
| "grad_norm": 0.14850662648677826, | |
| "learning_rate": 1.5272415272415271e-06, | |
| "loss": 0.2314, | |
| "step": 11170 | |
| }, | |
| { | |
| "epoch": 42.02007722007722, | |
| "grad_norm": 81.3932113647461, | |
| "learning_rate": 1.5186615186615189e-06, | |
| "loss": 0.8418, | |
| "step": 11180 | |
| }, | |
| { | |
| "epoch": 42.02007722007722, | |
| "eval_accuracy": 0.6888888888888889, | |
| "eval_loss": 2.199957847595215, | |
| "eval_runtime": 11.6437, | |
| "eval_samples_per_second": 3.865, | |
| "eval_steps_per_second": 1.288, | |
| "step": 11180 | |
| }, | |
| { | |
| "epoch": 43.000772200772204, | |
| "grad_norm": 0.10664477944374084, | |
| "learning_rate": 1.5100815100815102e-06, | |
| "loss": 0.2151, | |
| "step": 11190 | |
| }, | |
| { | |
| "epoch": 43.0015444015444, | |
| "grad_norm": 0.012597335502505302, | |
| "learning_rate": 1.5015015015015017e-06, | |
| "loss": 0.0007, | |
| "step": 11200 | |
| }, | |
| { | |
| "epoch": 43.002316602316604, | |
| "grad_norm": 0.3459191620349884, | |
| "learning_rate": 1.492921492921493e-06, | |
| "loss": 0.0005, | |
| "step": 11210 | |
| }, | |
| { | |
| "epoch": 43.0030888030888, | |
| "grad_norm": 0.013864747248589993, | |
| "learning_rate": 1.4843414843414843e-06, | |
| "loss": 0.0008, | |
| "step": 11220 | |
| }, | |
| { | |
| "epoch": 43.003861003861005, | |
| "grad_norm": 0.10407616198062897, | |
| "learning_rate": 1.4757614757614759e-06, | |
| "loss": 0.2126, | |
| "step": 11230 | |
| }, | |
| { | |
| "epoch": 43.0046332046332, | |
| "grad_norm": 0.012360197491943836, | |
| "learning_rate": 1.4671814671814674e-06, | |
| "loss": 0.3865, | |
| "step": 11240 | |
| }, | |
| { | |
| "epoch": 43.005405405405405, | |
| "grad_norm": 0.012420023791491985, | |
| "learning_rate": 1.4586014586014587e-06, | |
| "loss": 0.0007, | |
| "step": 11250 | |
| }, | |
| { | |
| "epoch": 43.00617760617761, | |
| "grad_norm": 0.051371827721595764, | |
| "learning_rate": 1.4500214500214502e-06, | |
| "loss": 0.0005, | |
| "step": 11260 | |
| }, | |
| { | |
| "epoch": 43.006949806949805, | |
| "grad_norm": 0.029028650373220444, | |
| "learning_rate": 1.4414414414414416e-06, | |
| "loss": 0.0012, | |
| "step": 11270 | |
| }, | |
| { | |
| "epoch": 43.00772200772201, | |
| "grad_norm": 0.011852563358843327, | |
| "learning_rate": 1.4328614328614329e-06, | |
| "loss": 0.1394, | |
| "step": 11280 | |
| }, | |
| { | |
| "epoch": 43.008494208494206, | |
| "grad_norm": 0.9856188297271729, | |
| "learning_rate": 1.4242814242814244e-06, | |
| "loss": 0.1234, | |
| "step": 11290 | |
| }, | |
| { | |
| "epoch": 43.00926640926641, | |
| "grad_norm": 0.07063347846269608, | |
| "learning_rate": 1.415701415701416e-06, | |
| "loss": 0.1585, | |
| "step": 11300 | |
| }, | |
| { | |
| "epoch": 43.01003861003861, | |
| "grad_norm": 0.002462083473801613, | |
| "learning_rate": 1.4071214071214072e-06, | |
| "loss": 0.034, | |
| "step": 11310 | |
| }, | |
| { | |
| "epoch": 43.01081081081081, | |
| "grad_norm": 0.023495899513363838, | |
| "learning_rate": 1.3985413985413988e-06, | |
| "loss": 0.0026, | |
| "step": 11320 | |
| }, | |
| { | |
| "epoch": 43.011583011583014, | |
| "grad_norm": 0.0035925237461924553, | |
| "learning_rate": 1.38996138996139e-06, | |
| "loss": 0.0002, | |
| "step": 11330 | |
| }, | |
| { | |
| "epoch": 43.01235521235521, | |
| "grad_norm": 0.017117993906140327, | |
| "learning_rate": 1.3813813813813814e-06, | |
| "loss": 0.0005, | |
| "step": 11340 | |
| }, | |
| { | |
| "epoch": 43.013127413127414, | |
| "grad_norm": 0.05065334215760231, | |
| "learning_rate": 1.372801372801373e-06, | |
| "loss": 0.0002, | |
| "step": 11350 | |
| }, | |
| { | |
| "epoch": 43.01389961389961, | |
| "grad_norm": 0.21535246074199677, | |
| "learning_rate": 1.3642213642213642e-06, | |
| "loss": 0.0004, | |
| "step": 11360 | |
| }, | |
| { | |
| "epoch": 43.014671814671814, | |
| "grad_norm": 207.9502716064453, | |
| "learning_rate": 1.3556413556413557e-06, | |
| "loss": 0.1807, | |
| "step": 11370 | |
| }, | |
| { | |
| "epoch": 43.01544401544402, | |
| "grad_norm": 0.0034615122713148594, | |
| "learning_rate": 1.3470613470613473e-06, | |
| "loss": 0.1658, | |
| "step": 11380 | |
| }, | |
| { | |
| "epoch": 43.016216216216215, | |
| "grad_norm": 0.0042229765094816685, | |
| "learning_rate": 1.3384813384813386e-06, | |
| "loss": 0.0003, | |
| "step": 11390 | |
| }, | |
| { | |
| "epoch": 43.01698841698842, | |
| "grad_norm": 42.964447021484375, | |
| "learning_rate": 1.32990132990133e-06, | |
| "loss": 0.2509, | |
| "step": 11400 | |
| }, | |
| { | |
| "epoch": 43.017760617760615, | |
| "grad_norm": 0.018056800588965416, | |
| "learning_rate": 1.3213213213213214e-06, | |
| "loss": 0.1674, | |
| "step": 11410 | |
| }, | |
| { | |
| "epoch": 43.01853281853282, | |
| "grad_norm": 0.00773951830342412, | |
| "learning_rate": 1.3127413127413127e-06, | |
| "loss": 0.2136, | |
| "step": 11420 | |
| }, | |
| { | |
| "epoch": 43.01930501930502, | |
| "grad_norm": 0.03975954279303551, | |
| "learning_rate": 1.3041613041613045e-06, | |
| "loss": 0.0003, | |
| "step": 11430 | |
| }, | |
| { | |
| "epoch": 43.02007722007722, | |
| "grad_norm": 0.0015788354212418199, | |
| "learning_rate": 1.2955812955812958e-06, | |
| "loss": 0.0007, | |
| "step": 11440 | |
| }, | |
| { | |
| "epoch": 43.02007722007722, | |
| "eval_accuracy": 0.7111111111111111, | |
| "eval_loss": 2.0410959720611572, | |
| "eval_runtime": 11.6235, | |
| "eval_samples_per_second": 3.871, | |
| "eval_steps_per_second": 1.29, | |
| "step": 11440 | |
| }, | |
| { | |
| "epoch": 44.000772200772204, | |
| "grad_norm": 99.61504364013672, | |
| "learning_rate": 1.287001287001287e-06, | |
| "loss": 0.4639, | |
| "step": 11450 | |
| }, | |
| { | |
| "epoch": 44.0015444015444, | |
| "grad_norm": 0.004829080309718847, | |
| "learning_rate": 1.2784212784212784e-06, | |
| "loss": 0.2582, | |
| "step": 11460 | |
| }, | |
| { | |
| "epoch": 44.002316602316604, | |
| "grad_norm": 0.001873974921181798, | |
| "learning_rate": 1.26984126984127e-06, | |
| "loss": 0.0002, | |
| "step": 11470 | |
| }, | |
| { | |
| "epoch": 44.0030888030888, | |
| "grad_norm": 0.015835706144571304, | |
| "learning_rate": 1.2612612612612613e-06, | |
| "loss": 0.0005, | |
| "step": 11480 | |
| }, | |
| { | |
| "epoch": 44.003861003861005, | |
| "grad_norm": 0.0006254987674765289, | |
| "learning_rate": 1.252681252681253e-06, | |
| "loss": 0.2537, | |
| "step": 11490 | |
| }, | |
| { | |
| "epoch": 44.0046332046332, | |
| "grad_norm": 0.005784200504422188, | |
| "learning_rate": 1.244101244101244e-06, | |
| "loss": 0.1615, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 44.005405405405405, | |
| "grad_norm": 0.005181388929486275, | |
| "learning_rate": 1.2355212355212356e-06, | |
| "loss": 0.0002, | |
| "step": 11510 | |
| }, | |
| { | |
| "epoch": 44.00617760617761, | |
| "grad_norm": 0.019228698685765266, | |
| "learning_rate": 1.226941226941227e-06, | |
| "loss": 0.0914, | |
| "step": 11520 | |
| }, | |
| { | |
| "epoch": 44.006949806949805, | |
| "grad_norm": 0.0022733949590474367, | |
| "learning_rate": 1.2183612183612185e-06, | |
| "loss": 0.0053, | |
| "step": 11530 | |
| }, | |
| { | |
| "epoch": 44.00772200772201, | |
| "grad_norm": 0.02803795598447323, | |
| "learning_rate": 1.20978120978121e-06, | |
| "loss": 0.3945, | |
| "step": 11540 | |
| }, | |
| { | |
| "epoch": 44.008494208494206, | |
| "grad_norm": 0.006438721902668476, | |
| "learning_rate": 1.2012012012012013e-06, | |
| "loss": 0.0002, | |
| "step": 11550 | |
| }, | |
| { | |
| "epoch": 44.00926640926641, | |
| "grad_norm": 0.11115320026874542, | |
| "learning_rate": 1.1926211926211926e-06, | |
| "loss": 0.2235, | |
| "step": 11560 | |
| }, | |
| { | |
| "epoch": 44.01003861003861, | |
| "grad_norm": 0.009124440141022205, | |
| "learning_rate": 1.1840411840411841e-06, | |
| "loss": 0.0353, | |
| "step": 11570 | |
| }, | |
| { | |
| "epoch": 44.01081081081081, | |
| "grad_norm": 0.008701231330633163, | |
| "learning_rate": 1.1754611754611757e-06, | |
| "loss": 0.0008, | |
| "step": 11580 | |
| }, | |
| { | |
| "epoch": 44.011583011583014, | |
| "grad_norm": 0.012650973163545132, | |
| "learning_rate": 1.166881166881167e-06, | |
| "loss": 0.0002, | |
| "step": 11590 | |
| }, | |
| { | |
| "epoch": 44.01235521235521, | |
| "grad_norm": 0.01399000734090805, | |
| "learning_rate": 1.1583011583011585e-06, | |
| "loss": 0.0013, | |
| "step": 11600 | |
| }, | |
| { | |
| "epoch": 44.013127413127414, | |
| "grad_norm": 5.8696746826171875, | |
| "learning_rate": 1.1497211497211498e-06, | |
| "loss": 0.0831, | |
| "step": 11610 | |
| }, | |
| { | |
| "epoch": 44.01389961389961, | |
| "grad_norm": 0.004275842569768429, | |
| "learning_rate": 1.1411411411411411e-06, | |
| "loss": 0.0221, | |
| "step": 11620 | |
| }, | |
| { | |
| "epoch": 44.014671814671814, | |
| "grad_norm": 0.06453493237495422, | |
| "learning_rate": 1.1325611325611326e-06, | |
| "loss": 0.1346, | |
| "step": 11630 | |
| }, | |
| { | |
| "epoch": 44.01544401544402, | |
| "grad_norm": 0.009648929350078106, | |
| "learning_rate": 1.1239811239811242e-06, | |
| "loss": 0.0002, | |
| "step": 11640 | |
| }, | |
| { | |
| "epoch": 44.016216216216215, | |
| "grad_norm": 0.022414736449718475, | |
| "learning_rate": 1.1154011154011155e-06, | |
| "loss": 0.0001, | |
| "step": 11650 | |
| }, | |
| { | |
| "epoch": 44.01698841698842, | |
| "grad_norm": 0.004432376008480787, | |
| "learning_rate": 1.106821106821107e-06, | |
| "loss": 0.0003, | |
| "step": 11660 | |
| }, | |
| { | |
| "epoch": 44.017760617760615, | |
| "grad_norm": 0.007916075177490711, | |
| "learning_rate": 1.0982410982410983e-06, | |
| "loss": 0.2268, | |
| "step": 11670 | |
| }, | |
| { | |
| "epoch": 44.01853281853282, | |
| "grad_norm": 0.009835878387093544, | |
| "learning_rate": 1.0896610896610896e-06, | |
| "loss": 0.2344, | |
| "step": 11680 | |
| }, | |
| { | |
| "epoch": 44.01930501930502, | |
| "grad_norm": 0.0064972443506121635, | |
| "learning_rate": 1.0810810810810812e-06, | |
| "loss": 0.0004, | |
| "step": 11690 | |
| }, | |
| { | |
| "epoch": 44.02007722007722, | |
| "grad_norm": 0.0038432800211012363, | |
| "learning_rate": 1.0725010725010727e-06, | |
| "loss": 0.0001, | |
| "step": 11700 | |
| }, | |
| { | |
| "epoch": 44.02007722007722, | |
| "eval_accuracy": 0.6888888888888889, | |
| "eval_loss": 2.108095169067383, | |
| "eval_runtime": 11.673, | |
| "eval_samples_per_second": 3.855, | |
| "eval_steps_per_second": 1.285, | |
| "step": 11700 | |
| }, | |
| { | |
| "epoch": 45.000772200772204, | |
| "grad_norm": 0.003354213200509548, | |
| "learning_rate": 1.063921063921064e-06, | |
| "loss": 0.0873, | |
| "step": 11710 | |
| }, | |
| { | |
| "epoch": 45.0015444015444, | |
| "grad_norm": 0.23387674987316132, | |
| "learning_rate": 1.0553410553410555e-06, | |
| "loss": 0.1457, | |
| "step": 11720 | |
| }, | |
| { | |
| "epoch": 45.002316602316604, | |
| "grad_norm": 0.008299161680042744, | |
| "learning_rate": 1.0467610467610468e-06, | |
| "loss": 0.2289, | |
| "step": 11730 | |
| }, | |
| { | |
| "epoch": 45.0030888030888, | |
| "grad_norm": 0.0024218843318521976, | |
| "learning_rate": 1.0381810381810382e-06, | |
| "loss": 0.0002, | |
| "step": 11740 | |
| }, | |
| { | |
| "epoch": 45.003861003861005, | |
| "grad_norm": 0.0018756206845864654, | |
| "learning_rate": 1.0296010296010297e-06, | |
| "loss": 0.0014, | |
| "step": 11750 | |
| }, | |
| { | |
| "epoch": 45.0046332046332, | |
| "grad_norm": 0.005605476908385754, | |
| "learning_rate": 1.0210210210210212e-06, | |
| "loss": 0.226, | |
| "step": 11760 | |
| }, | |
| { | |
| "epoch": 45.005405405405405, | |
| "grad_norm": 0.011059476062655449, | |
| "learning_rate": 1.0124410124410125e-06, | |
| "loss": 0.0547, | |
| "step": 11770 | |
| }, | |
| { | |
| "epoch": 45.00617760617761, | |
| "grad_norm": 0.0044801486656069756, | |
| "learning_rate": 1.0038610038610038e-06, | |
| "loss": 0.0009, | |
| "step": 11780 | |
| }, | |
| { | |
| "epoch": 45.006949806949805, | |
| "grad_norm": 258.8080139160156, | |
| "learning_rate": 9.952809952809954e-07, | |
| "loss": 0.0366, | |
| "step": 11790 | |
| }, | |
| { | |
| "epoch": 45.00772200772201, | |
| "grad_norm": 0.002683807397261262, | |
| "learning_rate": 9.867009867009867e-07, | |
| "loss": 0.2216, | |
| "step": 11800 | |
| }, | |
| { | |
| "epoch": 45.008494208494206, | |
| "grad_norm": 0.005808249581605196, | |
| "learning_rate": 9.781209781209782e-07, | |
| "loss": 0.0002, | |
| "step": 11810 | |
| }, | |
| { | |
| "epoch": 45.00926640926641, | |
| "grad_norm": 45.80361557006836, | |
| "learning_rate": 9.695409695409697e-07, | |
| "loss": 0.2395, | |
| "step": 11820 | |
| }, | |
| { | |
| "epoch": 45.01003861003861, | |
| "grad_norm": 0.019828274846076965, | |
| "learning_rate": 9.60960960960961e-07, | |
| "loss": 0.0002, | |
| "step": 11830 | |
| }, | |
| { | |
| "epoch": 45.01081081081081, | |
| "grad_norm": 0.04362059012055397, | |
| "learning_rate": 9.523809523809525e-07, | |
| "loss": 0.033, | |
| "step": 11840 | |
| }, | |
| { | |
| "epoch": 45.011583011583014, | |
| "grad_norm": 0.003450556192547083, | |
| "learning_rate": 9.438009438009439e-07, | |
| "loss": 0.0145, | |
| "step": 11850 | |
| }, | |
| { | |
| "epoch": 45.01235521235521, | |
| "grad_norm": 0.0010107405250892043, | |
| "learning_rate": 9.352209352209353e-07, | |
| "loss": 0.0518, | |
| "step": 11860 | |
| }, | |
| { | |
| "epoch": 45.013127413127414, | |
| "grad_norm": 0.003476174082607031, | |
| "learning_rate": 9.266409266409267e-07, | |
| "loss": 0.0012, | |
| "step": 11870 | |
| }, | |
| { | |
| "epoch": 45.01389961389961, | |
| "grad_norm": 0.4040607511997223, | |
| "learning_rate": 9.180609180609181e-07, | |
| "loss": 0.0005, | |
| "step": 11880 | |
| }, | |
| { | |
| "epoch": 45.014671814671814, | |
| "grad_norm": 0.004233676008880138, | |
| "learning_rate": 9.094809094809096e-07, | |
| "loss": 0.0004, | |
| "step": 11890 | |
| }, | |
| { | |
| "epoch": 45.01544401544402, | |
| "grad_norm": 0.01031399890780449, | |
| "learning_rate": 9.00900900900901e-07, | |
| "loss": 0.0005, | |
| "step": 11900 | |
| }, | |
| { | |
| "epoch": 45.016216216216215, | |
| "grad_norm": 104.54281616210938, | |
| "learning_rate": 8.923208923208925e-07, | |
| "loss": 0.0774, | |
| "step": 11910 | |
| }, | |
| { | |
| "epoch": 45.01698841698842, | |
| "grad_norm": 0.0005152209778316319, | |
| "learning_rate": 8.837408837408838e-07, | |
| "loss": 0.0003, | |
| "step": 11920 | |
| }, | |
| { | |
| "epoch": 45.017760617760615, | |
| "grad_norm": 0.008937066420912743, | |
| "learning_rate": 8.751608751608752e-07, | |
| "loss": 0.0003, | |
| "step": 11930 | |
| }, | |
| { | |
| "epoch": 45.01853281853282, | |
| "grad_norm": 0.020448965951800346, | |
| "learning_rate": 8.665808665808668e-07, | |
| "loss": 0.1503, | |
| "step": 11940 | |
| }, | |
| { | |
| "epoch": 45.01930501930502, | |
| "grad_norm": 0.03731679543852806, | |
| "learning_rate": 8.580008580008581e-07, | |
| "loss": 0.0003, | |
| "step": 11950 | |
| }, | |
| { | |
| "epoch": 45.02007722007722, | |
| "grad_norm": 0.02011745609343052, | |
| "learning_rate": 8.494208494208495e-07, | |
| "loss": 0.0004, | |
| "step": 11960 | |
| }, | |
| { | |
| "epoch": 45.02007722007722, | |
| "eval_accuracy": 0.6888888888888889, | |
| "eval_loss": 2.182063341140747, | |
| "eval_runtime": 11.6012, | |
| "eval_samples_per_second": 3.879, | |
| "eval_steps_per_second": 1.293, | |
| "step": 11960 | |
| }, | |
| { | |
| "epoch": 46.000772200772204, | |
| "grad_norm": 0.003135395934805274, | |
| "learning_rate": 8.40840840840841e-07, | |
| "loss": 0.0008, | |
| "step": 11970 | |
| }, | |
| { | |
| "epoch": 46.0015444015444, | |
| "grad_norm": 0.002082926919683814, | |
| "learning_rate": 8.322608322608323e-07, | |
| "loss": 0.0002, | |
| "step": 11980 | |
| }, | |
| { | |
| "epoch": 46.002316602316604, | |
| "grad_norm": 0.04637220501899719, | |
| "learning_rate": 8.236808236808237e-07, | |
| "loss": 0.0022, | |
| "step": 11990 | |
| }, | |
| { | |
| "epoch": 46.0030888030888, | |
| "grad_norm": 0.00046908273361623287, | |
| "learning_rate": 8.151008151008153e-07, | |
| "loss": 0.0002, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 46.003861003861005, | |
| "grad_norm": 0.006788803730159998, | |
| "learning_rate": 8.065208065208066e-07, | |
| "loss": 0.2737, | |
| "step": 12010 | |
| }, | |
| { | |
| "epoch": 46.0046332046332, | |
| "grad_norm": 0.07136689126491547, | |
| "learning_rate": 7.97940797940798e-07, | |
| "loss": 0.0005, | |
| "step": 12020 | |
| }, | |
| { | |
| "epoch": 46.005405405405405, | |
| "grad_norm": 0.0029624011367559433, | |
| "learning_rate": 7.893607893607893e-07, | |
| "loss": 0.1519, | |
| "step": 12030 | |
| }, | |
| { | |
| "epoch": 46.00617760617761, | |
| "grad_norm": 0.001701715518720448, | |
| "learning_rate": 7.807807807807808e-07, | |
| "loss": 0.0001, | |
| "step": 12040 | |
| }, | |
| { | |
| "epoch": 46.006949806949805, | |
| "grad_norm": 0.0016193045303225517, | |
| "learning_rate": 7.722007722007723e-07, | |
| "loss": 0.047, | |
| "step": 12050 | |
| }, | |
| { | |
| "epoch": 46.00772200772201, | |
| "grad_norm": 0.0045381346717476845, | |
| "learning_rate": 7.636207636207636e-07, | |
| "loss": 0.0001, | |
| "step": 12060 | |
| }, | |
| { | |
| "epoch": 46.008494208494206, | |
| "grad_norm": 0.003102656453847885, | |
| "learning_rate": 7.550407550407551e-07, | |
| "loss": 0.3918, | |
| "step": 12070 | |
| }, | |
| { | |
| "epoch": 46.00926640926641, | |
| "grad_norm": 0.0017835380276665092, | |
| "learning_rate": 7.464607464607465e-07, | |
| "loss": 0.239, | |
| "step": 12080 | |
| }, | |
| { | |
| "epoch": 46.01003861003861, | |
| "grad_norm": 0.004557139705866575, | |
| "learning_rate": 7.378807378807379e-07, | |
| "loss": 0.1398, | |
| "step": 12090 | |
| }, | |
| { | |
| "epoch": 46.01081081081081, | |
| "grad_norm": 0.013430671766400337, | |
| "learning_rate": 7.293007293007294e-07, | |
| "loss": 0.0001, | |
| "step": 12100 | |
| }, | |
| { | |
| "epoch": 46.011583011583014, | |
| "grad_norm": 0.020614583045244217, | |
| "learning_rate": 7.207207207207208e-07, | |
| "loss": 0.2451, | |
| "step": 12110 | |
| }, | |
| { | |
| "epoch": 46.01235521235521, | |
| "grad_norm": 0.024003766477108, | |
| "learning_rate": 7.121407121407122e-07, | |
| "loss": 0.0441, | |
| "step": 12120 | |
| }, | |
| { | |
| "epoch": 46.013127413127414, | |
| "grad_norm": 0.007595748174935579, | |
| "learning_rate": 7.035607035607036e-07, | |
| "loss": 0.0862, | |
| "step": 12130 | |
| }, | |
| { | |
| "epoch": 46.01389961389961, | |
| "grad_norm": 0.2669449746608734, | |
| "learning_rate": 6.94980694980695e-07, | |
| "loss": 0.0165, | |
| "step": 12140 | |
| }, | |
| { | |
| "epoch": 46.014671814671814, | |
| "grad_norm": 0.004871792625635862, | |
| "learning_rate": 6.864006864006865e-07, | |
| "loss": 0.142, | |
| "step": 12150 | |
| }, | |
| { | |
| "epoch": 46.01544401544402, | |
| "grad_norm": 0.027607223019003868, | |
| "learning_rate": 6.778206778206779e-07, | |
| "loss": 0.0002, | |
| "step": 12160 | |
| }, | |
| { | |
| "epoch": 46.016216216216215, | |
| "grad_norm": 0.0004795089189428836, | |
| "learning_rate": 6.692406692406693e-07, | |
| "loss": 0.0916, | |
| "step": 12170 | |
| }, | |
| { | |
| "epoch": 46.01698841698842, | |
| "grad_norm": 0.0060905334539711475, | |
| "learning_rate": 6.606606606606607e-07, | |
| "loss": 0.3611, | |
| "step": 12180 | |
| }, | |
| { | |
| "epoch": 46.017760617760615, | |
| "grad_norm": 0.003681829432025552, | |
| "learning_rate": 6.520806520806522e-07, | |
| "loss": 0.0003, | |
| "step": 12190 | |
| }, | |
| { | |
| "epoch": 46.01853281853282, | |
| "grad_norm": 0.008164509199559689, | |
| "learning_rate": 6.435006435006435e-07, | |
| "loss": 0.0977, | |
| "step": 12200 | |
| }, | |
| { | |
| "epoch": 46.01930501930502, | |
| "grad_norm": 0.00053538748761639, | |
| "learning_rate": 6.34920634920635e-07, | |
| "loss": 0.0033, | |
| "step": 12210 | |
| }, | |
| { | |
| "epoch": 46.02007722007722, | |
| "grad_norm": 0.010476024821400642, | |
| "learning_rate": 6.263406263406265e-07, | |
| "loss": 0.0013, | |
| "step": 12220 | |
| }, | |
| { | |
| "epoch": 46.02007722007722, | |
| "eval_accuracy": 0.7111111111111111, | |
| "eval_loss": 2.131287097930908, | |
| "eval_runtime": 11.5156, | |
| "eval_samples_per_second": 3.908, | |
| "eval_steps_per_second": 1.303, | |
| "step": 12220 | |
| }, | |
| { | |
| "epoch": 47.000772200772204, | |
| "grad_norm": 0.0032295091077685356, | |
| "learning_rate": 6.177606177606178e-07, | |
| "loss": 0.0295, | |
| "step": 12230 | |
| }, | |
| { | |
| "epoch": 47.0015444015444, | |
| "grad_norm": 0.021106446161866188, | |
| "learning_rate": 6.091806091806092e-07, | |
| "loss": 0.048, | |
| "step": 12240 | |
| }, | |
| { | |
| "epoch": 47.002316602316604, | |
| "grad_norm": 0.002815277548506856, | |
| "learning_rate": 6.006006006006006e-07, | |
| "loss": 0.0007, | |
| "step": 12250 | |
| }, | |
| { | |
| "epoch": 47.0030888030888, | |
| "grad_norm": 0.001195929478853941, | |
| "learning_rate": 5.920205920205921e-07, | |
| "loss": 0.1884, | |
| "step": 12260 | |
| }, | |
| { | |
| "epoch": 47.003861003861005, | |
| "grad_norm": 80.01729583740234, | |
| "learning_rate": 5.834405834405835e-07, | |
| "loss": 0.0094, | |
| "step": 12270 | |
| }, | |
| { | |
| "epoch": 47.0046332046332, | |
| "grad_norm": 0.0021410963963717222, | |
| "learning_rate": 5.748605748605749e-07, | |
| "loss": 0.0002, | |
| "step": 12280 | |
| }, | |
| { | |
| "epoch": 47.005405405405405, | |
| "grad_norm": 0.8117086887359619, | |
| "learning_rate": 5.662805662805663e-07, | |
| "loss": 0.1675, | |
| "step": 12290 | |
| }, | |
| { | |
| "epoch": 47.00617760617761, | |
| "grad_norm": 0.008614442311227322, | |
| "learning_rate": 5.577005577005577e-07, | |
| "loss": 0.0004, | |
| "step": 12300 | |
| }, | |
| { | |
| "epoch": 47.006949806949805, | |
| "grad_norm": 0.005176466889679432, | |
| "learning_rate": 5.491205491205492e-07, | |
| "loss": 0.0003, | |
| "step": 12310 | |
| }, | |
| { | |
| "epoch": 47.00772200772201, | |
| "grad_norm": 0.0040098209865391254, | |
| "learning_rate": 5.405405405405406e-07, | |
| "loss": 0.1824, | |
| "step": 12320 | |
| }, | |
| { | |
| "epoch": 47.008494208494206, | |
| "grad_norm": 0.0035810640547424555, | |
| "learning_rate": 5.31960531960532e-07, | |
| "loss": 0.0001, | |
| "step": 12330 | |
| }, | |
| { | |
| "epoch": 47.00926640926641, | |
| "grad_norm": 0.043194353580474854, | |
| "learning_rate": 5.233805233805234e-07, | |
| "loss": 0.081, | |
| "step": 12340 | |
| }, | |
| { | |
| "epoch": 47.01003861003861, | |
| "grad_norm": 0.016690198332071304, | |
| "learning_rate": 5.148005148005148e-07, | |
| "loss": 0.0004, | |
| "step": 12350 | |
| }, | |
| { | |
| "epoch": 47.01081081081081, | |
| "grad_norm": 0.015656674280762672, | |
| "learning_rate": 5.062205062205063e-07, | |
| "loss": 0.0001, | |
| "step": 12360 | |
| }, | |
| { | |
| "epoch": 47.011583011583014, | |
| "grad_norm": 0.008801618590950966, | |
| "learning_rate": 4.976404976404977e-07, | |
| "loss": 0.0001, | |
| "step": 12370 | |
| }, | |
| { | |
| "epoch": 47.01235521235521, | |
| "grad_norm": 0.016372770071029663, | |
| "learning_rate": 4.890604890604891e-07, | |
| "loss": 0.0001, | |
| "step": 12380 | |
| }, | |
| { | |
| "epoch": 47.013127413127414, | |
| "grad_norm": 0.0037778830155730247, | |
| "learning_rate": 4.804804804804805e-07, | |
| "loss": 0.0001, | |
| "step": 12390 | |
| }, | |
| { | |
| "epoch": 47.01389961389961, | |
| "grad_norm": 0.09339027106761932, | |
| "learning_rate": 4.7190047190047194e-07, | |
| "loss": 0.0023, | |
| "step": 12400 | |
| }, | |
| { | |
| "epoch": 47.014671814671814, | |
| "grad_norm": 482.0824279785156, | |
| "learning_rate": 4.6332046332046336e-07, | |
| "loss": 0.0273, | |
| "step": 12410 | |
| }, | |
| { | |
| "epoch": 47.01544401544402, | |
| "grad_norm": 0.010961725376546383, | |
| "learning_rate": 4.547404547404548e-07, | |
| "loss": 0.0002, | |
| "step": 12420 | |
| }, | |
| { | |
| "epoch": 47.016216216216215, | |
| "grad_norm": 0.015132254920899868, | |
| "learning_rate": 4.4616044616044625e-07, | |
| "loss": 0.0001, | |
| "step": 12430 | |
| }, | |
| { | |
| "epoch": 47.01698841698842, | |
| "grad_norm": 0.0039853425696492195, | |
| "learning_rate": 4.375804375804376e-07, | |
| "loss": 0.0001, | |
| "step": 12440 | |
| }, | |
| { | |
| "epoch": 47.017760617760615, | |
| "grad_norm": 0.010092932730913162, | |
| "learning_rate": 4.2900042900042903e-07, | |
| "loss": 0.2261, | |
| "step": 12450 | |
| }, | |
| { | |
| "epoch": 47.01853281853282, | |
| "grad_norm": 0.007288198918104172, | |
| "learning_rate": 4.204204204204205e-07, | |
| "loss": 0.0001, | |
| "step": 12460 | |
| }, | |
| { | |
| "epoch": 47.01930501930502, | |
| "grad_norm": 0.0723380446434021, | |
| "learning_rate": 4.1184041184041187e-07, | |
| "loss": 0.0003, | |
| "step": 12470 | |
| }, | |
| { | |
| "epoch": 47.02007722007722, | |
| "grad_norm": 0.737527072429657, | |
| "learning_rate": 4.032604032604033e-07, | |
| "loss": 0.0003, | |
| "step": 12480 | |
| }, | |
| { | |
| "epoch": 47.02007722007722, | |
| "eval_accuracy": 0.7111111111111111, | |
| "eval_loss": 2.2112810611724854, | |
| "eval_runtime": 11.7955, | |
| "eval_samples_per_second": 3.815, | |
| "eval_steps_per_second": 1.272, | |
| "step": 12480 | |
| }, | |
| { | |
| "epoch": 48.000772200772204, | |
| "grad_norm": 0.018116367980837822, | |
| "learning_rate": 3.9468039468039466e-07, | |
| "loss": 0.0002, | |
| "step": 12490 | |
| }, | |
| { | |
| "epoch": 48.0015444015444, | |
| "grad_norm": 1.851183295249939, | |
| "learning_rate": 3.8610038610038613e-07, | |
| "loss": 0.0031, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 48.002316602316604, | |
| "grad_norm": 0.007373946253210306, | |
| "learning_rate": 3.7752037752037755e-07, | |
| "loss": 0.0001, | |
| "step": 12510 | |
| }, | |
| { | |
| "epoch": 48.0030888030888, | |
| "grad_norm": 0.031634677201509476, | |
| "learning_rate": 3.6894036894036897e-07, | |
| "loss": 0.0002, | |
| "step": 12520 | |
| }, | |
| { | |
| "epoch": 48.003861003861005, | |
| "grad_norm": 0.014068067073822021, | |
| "learning_rate": 3.603603603603604e-07, | |
| "loss": 0.0002, | |
| "step": 12530 | |
| }, | |
| { | |
| "epoch": 48.0046332046332, | |
| "grad_norm": 0.0021505013573914766, | |
| "learning_rate": 3.517803517803518e-07, | |
| "loss": 0.0002, | |
| "step": 12540 | |
| }, | |
| { | |
| "epoch": 48.005405405405405, | |
| "grad_norm": 0.007775390055030584, | |
| "learning_rate": 3.4320034320034323e-07, | |
| "loss": 0.0002, | |
| "step": 12550 | |
| }, | |
| { | |
| "epoch": 48.00617760617761, | |
| "grad_norm": 0.006418993230909109, | |
| "learning_rate": 3.3462033462033465e-07, | |
| "loss": 0.0001, | |
| "step": 12560 | |
| }, | |
| { | |
| "epoch": 48.006949806949805, | |
| "grad_norm": 0.0038523671682924032, | |
| "learning_rate": 3.260403260403261e-07, | |
| "loss": 0.0039, | |
| "step": 12570 | |
| }, | |
| { | |
| "epoch": 48.00772200772201, | |
| "grad_norm": 0.011759024113416672, | |
| "learning_rate": 3.174603174603175e-07, | |
| "loss": 0.0003, | |
| "step": 12580 | |
| }, | |
| { | |
| "epoch": 48.008494208494206, | |
| "grad_norm": 0.006618797779083252, | |
| "learning_rate": 3.088803088803089e-07, | |
| "loss": 0.0004, | |
| "step": 12590 | |
| }, | |
| { | |
| "epoch": 48.00926640926641, | |
| "grad_norm": 0.0052281878888607025, | |
| "learning_rate": 3.003003003003003e-07, | |
| "loss": 0.0001, | |
| "step": 12600 | |
| }, | |
| { | |
| "epoch": 48.01003861003861, | |
| "grad_norm": 0.026694072410464287, | |
| "learning_rate": 2.9172029172029174e-07, | |
| "loss": 0.0005, | |
| "step": 12610 | |
| }, | |
| { | |
| "epoch": 48.01081081081081, | |
| "grad_norm": 0.6997846364974976, | |
| "learning_rate": 2.8314028314028316e-07, | |
| "loss": 0.0002, | |
| "step": 12620 | |
| }, | |
| { | |
| "epoch": 48.011583011583014, | |
| "grad_norm": 0.003470169845968485, | |
| "learning_rate": 2.745602745602746e-07, | |
| "loss": 0.0016, | |
| "step": 12630 | |
| }, | |
| { | |
| "epoch": 48.01235521235521, | |
| "grad_norm": 1.1241451501846313, | |
| "learning_rate": 2.65980265980266e-07, | |
| "loss": 0.0005, | |
| "step": 12640 | |
| }, | |
| { | |
| "epoch": 48.013127413127414, | |
| "grad_norm": 0.0048046838492155075, | |
| "learning_rate": 2.574002574002574e-07, | |
| "loss": 0.0001, | |
| "step": 12650 | |
| }, | |
| { | |
| "epoch": 48.01389961389961, | |
| "grad_norm": 0.0028558846097439528, | |
| "learning_rate": 2.4882024882024884e-07, | |
| "loss": 0.0001, | |
| "step": 12660 | |
| }, | |
| { | |
| "epoch": 48.014671814671814, | |
| "grad_norm": 0.0014227797510102391, | |
| "learning_rate": 2.4024024024024026e-07, | |
| "loss": 0.0002, | |
| "step": 12670 | |
| }, | |
| { | |
| "epoch": 48.01544401544402, | |
| "grad_norm": 0.0037844879552721977, | |
| "learning_rate": 2.3166023166023168e-07, | |
| "loss": 0.017, | |
| "step": 12680 | |
| }, | |
| { | |
| "epoch": 48.016216216216215, | |
| "grad_norm": 0.001309237559325993, | |
| "learning_rate": 2.2308022308022312e-07, | |
| "loss": 0.0012, | |
| "step": 12690 | |
| }, | |
| { | |
| "epoch": 48.01698841698842, | |
| "grad_norm": 0.0035404835361987352, | |
| "learning_rate": 2.1450021450021452e-07, | |
| "loss": 0.0007, | |
| "step": 12700 | |
| }, | |
| { | |
| "epoch": 48.017760617760615, | |
| "grad_norm": 0.01660049706697464, | |
| "learning_rate": 2.0592020592020594e-07, | |
| "loss": 0.005, | |
| "step": 12710 | |
| }, | |
| { | |
| "epoch": 48.01853281853282, | |
| "grad_norm": 0.013520884327590466, | |
| "learning_rate": 1.9734019734019733e-07, | |
| "loss": 0.0122, | |
| "step": 12720 | |
| }, | |
| { | |
| "epoch": 48.01930501930502, | |
| "grad_norm": 0.0025345338508486748, | |
| "learning_rate": 1.8876018876018877e-07, | |
| "loss": 0.1069, | |
| "step": 12730 | |
| }, | |
| { | |
| "epoch": 48.02007722007722, | |
| "grad_norm": 101.20980072021484, | |
| "learning_rate": 1.801801801801802e-07, | |
| "loss": 0.83, | |
| "step": 12740 | |
| }, | |
| { | |
| "epoch": 48.02007722007722, | |
| "eval_accuracy": 0.6888888888888889, | |
| "eval_loss": 2.204815626144409, | |
| "eval_runtime": 11.3918, | |
| "eval_samples_per_second": 3.95, | |
| "eval_steps_per_second": 1.317, | |
| "step": 12740 | |
| }, | |
| { | |
| "epoch": 49.000772200772204, | |
| "grad_norm": 0.008942559361457825, | |
| "learning_rate": 1.7160017160017161e-07, | |
| "loss": 0.0002, | |
| "step": 12750 | |
| }, | |
| { | |
| "epoch": 49.0015444015444, | |
| "grad_norm": 0.005645198281854391, | |
| "learning_rate": 1.6302016302016306e-07, | |
| "loss": 0.0036, | |
| "step": 12760 | |
| }, | |
| { | |
| "epoch": 49.002316602316604, | |
| "grad_norm": 0.0036984614562243223, | |
| "learning_rate": 1.5444015444015445e-07, | |
| "loss": 0.0001, | |
| "step": 12770 | |
| }, | |
| { | |
| "epoch": 49.0030888030888, | |
| "grad_norm": 0.004046537913382053, | |
| "learning_rate": 1.4586014586014587e-07, | |
| "loss": 0.0001, | |
| "step": 12780 | |
| }, | |
| { | |
| "epoch": 49.003861003861005, | |
| "grad_norm": 0.0005429061129689217, | |
| "learning_rate": 1.372801372801373e-07, | |
| "loss": 0.0002, | |
| "step": 12790 | |
| }, | |
| { | |
| "epoch": 49.0046332046332, | |
| "grad_norm": 0.002891113283112645, | |
| "learning_rate": 1.287001287001287e-07, | |
| "loss": 0.0002, | |
| "step": 12800 | |
| }, | |
| { | |
| "epoch": 49.005405405405405, | |
| "grad_norm": 0.005318440031260252, | |
| "learning_rate": 1.2012012012012013e-07, | |
| "loss": 0.0001, | |
| "step": 12810 | |
| }, | |
| { | |
| "epoch": 49.00617760617761, | |
| "grad_norm": 0.0029130184557288885, | |
| "learning_rate": 1.1154011154011156e-07, | |
| "loss": 0.0002, | |
| "step": 12820 | |
| }, | |
| { | |
| "epoch": 49.006949806949805, | |
| "grad_norm": 0.7598032355308533, | |
| "learning_rate": 1.0296010296010297e-07, | |
| "loss": 0.0003, | |
| "step": 12830 | |
| }, | |
| { | |
| "epoch": 49.00772200772201, | |
| "grad_norm": 0.02352285198867321, | |
| "learning_rate": 9.438009438009439e-08, | |
| "loss": 0.0002, | |
| "step": 12840 | |
| }, | |
| { | |
| "epoch": 49.008494208494206, | |
| "grad_norm": 0.004137629177421331, | |
| "learning_rate": 8.580008580008581e-08, | |
| "loss": 0.0001, | |
| "step": 12850 | |
| }, | |
| { | |
| "epoch": 49.00926640926641, | |
| "grad_norm": 0.007847619242966175, | |
| "learning_rate": 7.722007722007723e-08, | |
| "loss": 0.0001, | |
| "step": 12860 | |
| }, | |
| { | |
| "epoch": 49.01003861003861, | |
| "grad_norm": 236.08055114746094, | |
| "learning_rate": 6.864006864006865e-08, | |
| "loss": 0.2852, | |
| "step": 12870 | |
| }, | |
| { | |
| "epoch": 49.01081081081081, | |
| "grad_norm": 0.004942096769809723, | |
| "learning_rate": 6.006006006006006e-08, | |
| "loss": 0.0001, | |
| "step": 12880 | |
| }, | |
| { | |
| "epoch": 49.011583011583014, | |
| "grad_norm": 0.009192113764584064, | |
| "learning_rate": 5.1480051480051484e-08, | |
| "loss": 0.2111, | |
| "step": 12890 | |
| }, | |
| { | |
| "epoch": 49.01235521235521, | |
| "grad_norm": 0.002397160744294524, | |
| "learning_rate": 4.2900042900042903e-08, | |
| "loss": 0.0611, | |
| "step": 12900 | |
| }, | |
| { | |
| "epoch": 49.013127413127414, | |
| "grad_norm": 0.016426341608166695, | |
| "learning_rate": 3.432003432003432e-08, | |
| "loss": 0.0002, | |
| "step": 12910 | |
| }, | |
| { | |
| "epoch": 49.01389961389961, | |
| "grad_norm": 0.37930360436439514, | |
| "learning_rate": 2.5740025740025742e-08, | |
| "loss": 0.0005, | |
| "step": 12920 | |
| }, | |
| { | |
| "epoch": 49.014671814671814, | |
| "grad_norm": 0.004664725624024868, | |
| "learning_rate": 1.716001716001716e-08, | |
| "loss": 0.0003, | |
| "step": 12930 | |
| }, | |
| { | |
| "epoch": 49.01544401544402, | |
| "grad_norm": 0.01354283094406128, | |
| "learning_rate": 8.58000858000858e-09, | |
| "loss": 0.0001, | |
| "step": 12940 | |
| }, | |
| { | |
| "epoch": 49.016216216216215, | |
| "grad_norm": 0.0005170258227735758, | |
| "learning_rate": 0.0, | |
| "loss": 0.0001, | |
| "step": 12950 | |
| }, | |
| { | |
| "epoch": 49.016216216216215, | |
| "eval_accuracy": 0.6888888888888889, | |
| "eval_loss": 2.199702739715576, | |
| "eval_runtime": 13.0793, | |
| "eval_samples_per_second": 3.441, | |
| "eval_steps_per_second": 1.147, | |
| "step": 12950 | |
| }, | |
| { | |
| "epoch": 49.016216216216215, | |
| "step": 12950, | |
| "total_flos": 1.7016105177038035e+20, | |
| "train_loss": 0.32982333966647276, | |
| "train_runtime": 21819.3604, | |
| "train_samples_per_second": 1.781, | |
| "train_steps_per_second": 0.594 | |
| }, | |
| { | |
| "epoch": 49.016216216216215, | |
| "eval_accuracy": 0.7777777777777778, | |
| "eval_loss": 0.7423439621925354, | |
| "eval_runtime": 11.4389, | |
| "eval_samples_per_second": 3.934, | |
| "eval_steps_per_second": 1.311, | |
| "step": 12950 | |
| }, | |
| { | |
| "epoch": 49.016216216216215, | |
| "eval_accuracy": 0.7777777777777778, | |
| "eval_loss": 0.7423441410064697, | |
| "eval_runtime": 11.3661, | |
| "eval_samples_per_second": 3.959, | |
| "eval_steps_per_second": 1.32, | |
| "step": 12950 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 12950, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 9223372036854775807, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.7016105177038035e+20, | |
| "train_batch_size": 3, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |