| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.20472595352485537, | |
| "eval_steps": 348, | |
| "global_step": 1044, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.00019609765663300324, | |
| "grad_norm": 20.33372688293457, | |
| "learning_rate": 2e-05, | |
| "loss": 3.0843, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.00019609765663300324, | |
| "eval_loss": 1.1017773151397705, | |
| "eval_runtime": 79.9135, | |
| "eval_samples_per_second": 26.879, | |
| "eval_steps_per_second": 13.44, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0003921953132660065, | |
| "grad_norm": 19.32895278930664, | |
| "learning_rate": 4e-05, | |
| "loss": 3.2221, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0005882929698990097, | |
| "grad_norm": 18.6882266998291, | |
| "learning_rate": 6e-05, | |
| "loss": 3.8951, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.000784390626532013, | |
| "grad_norm": 43.008060455322266, | |
| "learning_rate": 8e-05, | |
| "loss": 5.167, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0009804882831650162, | |
| "grad_norm": 21.642993927001953, | |
| "learning_rate": 0.0001, | |
| "loss": 3.1304, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0011765859397980193, | |
| "grad_norm": 29.79266929626465, | |
| "learning_rate": 0.00012, | |
| "loss": 4.5153, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0013726835964310226, | |
| "grad_norm": 25.503681182861328, | |
| "learning_rate": 0.00014, | |
| "loss": 3.8083, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.001568781253064026, | |
| "grad_norm": 32.35524368286133, | |
| "learning_rate": 0.00016, | |
| "loss": 4.253, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0017648789096970292, | |
| "grad_norm": 21.053390502929688, | |
| "learning_rate": 0.00018, | |
| "loss": 3.3757, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.0019609765663300325, | |
| "grad_norm": 25.7067928314209, | |
| "learning_rate": 0.0002, | |
| "loss": 3.2484, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0021570742229630358, | |
| "grad_norm": 22.57227897644043, | |
| "learning_rate": 0.00019999974049780868, | |
| "loss": 2.8378, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.0023531718795960386, | |
| "grad_norm": 19.06597900390625, | |
| "learning_rate": 0.00019999896199258152, | |
| "loss": 3.231, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.002549269536229042, | |
| "grad_norm": 17.590620040893555, | |
| "learning_rate": 0.000199997664488359, | |
| "loss": 2.2391, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.002745367192862045, | |
| "grad_norm": 8.627043724060059, | |
| "learning_rate": 0.00019999584799187522, | |
| "loss": 1.7095, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.0029414648494950485, | |
| "grad_norm": 21.60858917236328, | |
| "learning_rate": 0.0001999935125125579, | |
| "loss": 3.9299, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.003137562506128052, | |
| "grad_norm": 8.075380325317383, | |
| "learning_rate": 0.00019999065806252829, | |
| "loss": 1.7939, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.003333660162761055, | |
| "grad_norm": 11.393594741821289, | |
| "learning_rate": 0.00019998728465660105, | |
| "loss": 1.601, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.0035297578193940584, | |
| "grad_norm": 8.256339073181152, | |
| "learning_rate": 0.00019998339231228434, | |
| "loss": 3.1556, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.0037258554760270617, | |
| "grad_norm": 20.03615951538086, | |
| "learning_rate": 0.0001999789810497796, | |
| "loss": 2.0883, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.003921953132660065, | |
| "grad_norm": 10.166353225708008, | |
| "learning_rate": 0.0001999740508919815, | |
| "loss": 3.5616, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.004118050789293068, | |
| "grad_norm": 15.80553913116455, | |
| "learning_rate": 0.0001999686018644777, | |
| "loss": 3.0344, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.0043141484459260715, | |
| "grad_norm": 7.451974391937256, | |
| "learning_rate": 0.00019996263399554897, | |
| "loss": 2.1049, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.004510246102559075, | |
| "grad_norm": 5.434274673461914, | |
| "learning_rate": 0.00019995614731616875, | |
| "loss": 2.3178, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.004706343759192077, | |
| "grad_norm": 10.594315528869629, | |
| "learning_rate": 0.00019994914186000328, | |
| "loss": 1.7096, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.0049024414158250805, | |
| "grad_norm": 5.348718166351318, | |
| "learning_rate": 0.0001999416176634111, | |
| "loss": 2.695, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.005098539072458084, | |
| "grad_norm": 17.776073455810547, | |
| "learning_rate": 0.00019993357476544312, | |
| "loss": 1.7411, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.005294636729091087, | |
| "grad_norm": 10.051606178283691, | |
| "learning_rate": 0.0001999250132078424, | |
| "loss": 2.6161, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.00549073438572409, | |
| "grad_norm": 26.03020668029785, | |
| "learning_rate": 0.00019991593303504376, | |
| "loss": 3.3977, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.005686832042357094, | |
| "grad_norm": 10.213540077209473, | |
| "learning_rate": 0.00019990633429417363, | |
| "loss": 1.2442, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.005882929698990097, | |
| "grad_norm": 11.69288444519043, | |
| "learning_rate": 0.00019989621703505, | |
| "loss": 1.4702, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.0060790273556231, | |
| "grad_norm": 4.343452453613281, | |
| "learning_rate": 0.00019988558131018186, | |
| "loss": 1.0779, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.006275125012256104, | |
| "grad_norm": 9.106976509094238, | |
| "learning_rate": 0.00019987442717476906, | |
| "loss": 2.5887, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.006471222668889107, | |
| "grad_norm": 17.658370971679688, | |
| "learning_rate": 0.00019986275468670205, | |
| "loss": 2.2258, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.00666732032552211, | |
| "grad_norm": 6.7451090812683105, | |
| "learning_rate": 0.00019985056390656162, | |
| "loss": 1.7206, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.0068634179821551134, | |
| "grad_norm": 28.07065200805664, | |
| "learning_rate": 0.00019983785489761837, | |
| "loss": 2.7356, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.007059515638788117, | |
| "grad_norm": 11.387879371643066, | |
| "learning_rate": 0.00019982462772583266, | |
| "loss": 1.973, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.00725561329542112, | |
| "grad_norm": 9.64372444152832, | |
| "learning_rate": 0.00019981088245985408, | |
| "loss": 2.7339, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.007451710952054123, | |
| "grad_norm": 9.302544593811035, | |
| "learning_rate": 0.00019979661917102115, | |
| "loss": 1.7498, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.007647808608687127, | |
| "grad_norm": 15.064400672912598, | |
| "learning_rate": 0.000199781837933361, | |
| "loss": 3.0109, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.00784390626532013, | |
| "grad_norm": 7.281099319458008, | |
| "learning_rate": 0.00019976653882358884, | |
| "loss": 1.3118, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.008040003921953132, | |
| "grad_norm": 6.4474873542785645, | |
| "learning_rate": 0.0001997507219211078, | |
| "loss": 1.408, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.008236101578586136, | |
| "grad_norm": 13.101079940795898, | |
| "learning_rate": 0.00019973438730800822, | |
| "loss": 2.3367, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.008432199235219139, | |
| "grad_norm": 5.951049327850342, | |
| "learning_rate": 0.00019971753506906753, | |
| "loss": 0.9101, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.008628296891852143, | |
| "grad_norm": 11.212276458740234, | |
| "learning_rate": 0.00019970016529174947, | |
| "loss": 2.7058, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.008824394548485145, | |
| "grad_norm": 8.68136978149414, | |
| "learning_rate": 0.0001996822780662041, | |
| "loss": 2.0276, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.00902049220511815, | |
| "grad_norm": 17.70038414001465, | |
| "learning_rate": 0.00019966387348526683, | |
| "loss": 2.7989, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.009216589861751152, | |
| "grad_norm": 10.247598648071289, | |
| "learning_rate": 0.00019964495164445824, | |
| "loss": 1.9618, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.009412687518384154, | |
| "grad_norm": 10.378255844116211, | |
| "learning_rate": 0.0001996255126419835, | |
| "loss": 1.8003, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.009608785175017159, | |
| "grad_norm": 31.620820999145508, | |
| "learning_rate": 0.0001996055565787319, | |
| "loss": 2.8785, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.009804882831650161, | |
| "grad_norm": 9.976147651672363, | |
| "learning_rate": 0.0001995850835582763, | |
| "loss": 2.5605, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.010000980488283165, | |
| "grad_norm": 11.751899719238281, | |
| "learning_rate": 0.00019956409368687258, | |
| "loss": 2.7556, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.010197078144916168, | |
| "grad_norm": 15.828932762145996, | |
| "learning_rate": 0.000199542587073459, | |
| "loss": 2.7773, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.010393175801549172, | |
| "grad_norm": 10.772979736328125, | |
| "learning_rate": 0.00019952056382965597, | |
| "loss": 1.9553, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.010589273458182174, | |
| "grad_norm": 10.821427345275879, | |
| "learning_rate": 0.00019949802406976495, | |
| "loss": 1.8528, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.010785371114815178, | |
| "grad_norm": 7.228662490844727, | |
| "learning_rate": 0.00019947496791076837, | |
| "loss": 1.1844, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.01098146877144818, | |
| "grad_norm": 7.164773941040039, | |
| "learning_rate": 0.00019945139547232872, | |
| "loss": 1.0291, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.011177566428081185, | |
| "grad_norm": 13.927733421325684, | |
| "learning_rate": 0.0001994273068767879, | |
| "loss": 1.5417, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.011373664084714187, | |
| "grad_norm": 10.366493225097656, | |
| "learning_rate": 0.00019940270224916688, | |
| "loss": 1.5122, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.011569761741347192, | |
| "grad_norm": 11.2214994430542, | |
| "learning_rate": 0.00019937758171716468, | |
| "loss": 1.6003, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.011765859397980194, | |
| "grad_norm": 14.360090255737305, | |
| "learning_rate": 0.000199351945411158, | |
| "loss": 1.5651, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.011961957054613198, | |
| "grad_norm": 17.97150993347168, | |
| "learning_rate": 0.00019932579346420038, | |
| "loss": 1.6064, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.0121580547112462, | |
| "grad_norm": 10.190518379211426, | |
| "learning_rate": 0.00019929912601202151, | |
| "loss": 1.9151, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.012354152367879203, | |
| "grad_norm": 13.573248863220215, | |
| "learning_rate": 0.00019927194319302677, | |
| "loss": 4.0602, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.012550250024512207, | |
| "grad_norm": 16.919841766357422, | |
| "learning_rate": 0.00019924424514829606, | |
| "loss": 2.8292, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.01274634768114521, | |
| "grad_norm": 58.470252990722656, | |
| "learning_rate": 0.00019921603202158354, | |
| "loss": 1.9637, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.012942445337778214, | |
| "grad_norm": 18.334800720214844, | |
| "learning_rate": 0.00019918730395931649, | |
| "loss": 2.5609, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.013138542994411216, | |
| "grad_norm": 12.280759811401367, | |
| "learning_rate": 0.00019915806111059486, | |
| "loss": 1.2495, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.01333464065104422, | |
| "grad_norm": 8.015874862670898, | |
| "learning_rate": 0.0001991283036271903, | |
| "loss": 1.505, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.013530738307677223, | |
| "grad_norm": 7.713284969329834, | |
| "learning_rate": 0.0001990980316635455, | |
| "loss": 2.3898, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.013726835964310227, | |
| "grad_norm": 18.01800537109375, | |
| "learning_rate": 0.00019906724537677316, | |
| "loss": 3.0263, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.01392293362094323, | |
| "grad_norm": 21.270421981811523, | |
| "learning_rate": 0.00019903594492665558, | |
| "loss": 3.2547, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.014119031277576233, | |
| "grad_norm": 21.60205841064453, | |
| "learning_rate": 0.0001990041304756434, | |
| "loss": 2.577, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.014315128934209236, | |
| "grad_norm": 10.01419734954834, | |
| "learning_rate": 0.00019897180218885507, | |
| "loss": 1.9092, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.01451122659084224, | |
| "grad_norm": 14.10943603515625, | |
| "learning_rate": 0.00019893896023407578, | |
| "loss": 2.2377, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.014707324247475242, | |
| "grad_norm": 11.310667037963867, | |
| "learning_rate": 0.0001989056047817567, | |
| "loss": 1.6645, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.014903421904108247, | |
| "grad_norm": 6.586666107177734, | |
| "learning_rate": 0.0001988717360050141, | |
| "loss": 2.2651, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.015099519560741249, | |
| "grad_norm": 4.402716159820557, | |
| "learning_rate": 0.00019883735407962846, | |
| "loss": 1.3483, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.015295617217374253, | |
| "grad_norm": 9.384387016296387, | |
| "learning_rate": 0.00019880245918404342, | |
| "loss": 2.6391, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.015491714874007256, | |
| "grad_norm": 6.753894329071045, | |
| "learning_rate": 0.000198767051499365, | |
| "loss": 2.9391, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.01568781253064026, | |
| "grad_norm": 6.399787902832031, | |
| "learning_rate": 0.00019873113120936074, | |
| "loss": 3.7452, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.01588391018727326, | |
| "grad_norm": 8.880107879638672, | |
| "learning_rate": 0.00019869469850045842, | |
| "loss": 1.2771, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.016080007843906265, | |
| "grad_norm": 12.630661964416504, | |
| "learning_rate": 0.00019865775356174545, | |
| "loss": 2.2072, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.01627610550053927, | |
| "grad_norm": 7.974503993988037, | |
| "learning_rate": 0.00019862029658496762, | |
| "loss": 1.9795, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.016472203157172273, | |
| "grad_norm": 50.43594741821289, | |
| "learning_rate": 0.00019858232776452837, | |
| "loss": 1.5331, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.016668300813805274, | |
| "grad_norm": 7.273484230041504, | |
| "learning_rate": 0.00019854384729748746, | |
| "loss": 2.4005, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.016864398470438278, | |
| "grad_norm": 5.826492786407471, | |
| "learning_rate": 0.00019850485538356027, | |
| "loss": 2.1915, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.017060496127071282, | |
| "grad_norm": 9.881019592285156, | |
| "learning_rate": 0.0001984653522251165, | |
| "loss": 2.3309, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.017256593783704286, | |
| "grad_norm": 9.147713661193848, | |
| "learning_rate": 0.00019842533802717923, | |
| "loss": 1.1404, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.017452691440337287, | |
| "grad_norm": 13.98263931274414, | |
| "learning_rate": 0.00019838481299742398, | |
| "loss": 1.2166, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.01764878909697029, | |
| "grad_norm": 8.206791877746582, | |
| "learning_rate": 0.0001983437773461774, | |
| "loss": 2.6039, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.017844886753603295, | |
| "grad_norm": 10.445443153381348, | |
| "learning_rate": 0.00019830223128641637, | |
| "loss": 2.3554, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.0180409844102363, | |
| "grad_norm": 11.756292343139648, | |
| "learning_rate": 0.00019826017503376666, | |
| "loss": 1.7371, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.0182370820668693, | |
| "grad_norm": 7.509032249450684, | |
| "learning_rate": 0.00019821760880650214, | |
| "loss": 1.389, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.018433179723502304, | |
| "grad_norm": 8.619280815124512, | |
| "learning_rate": 0.00019817453282554333, | |
| "loss": 1.6818, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.01862927738013531, | |
| "grad_norm": 9.11640739440918, | |
| "learning_rate": 0.00019813094731445654, | |
| "loss": 1.631, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.01882537503676831, | |
| "grad_norm": 14.109521865844727, | |
| "learning_rate": 0.00019808685249945245, | |
| "loss": 2.0497, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.019021472693401313, | |
| "grad_norm": 10.804281234741211, | |
| "learning_rate": 0.00019804224860938506, | |
| "loss": 2.2364, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.019217570350034317, | |
| "grad_norm": 7.363731384277344, | |
| "learning_rate": 0.0001979971358757505, | |
| "loss": 1.0967, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.01941366800666732, | |
| "grad_norm": 15.269912719726562, | |
| "learning_rate": 0.0001979515145326859, | |
| "loss": 2.8752, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.019609765663300322, | |
| "grad_norm": 5.457535266876221, | |
| "learning_rate": 0.000197905384816968, | |
| "loss": 1.7098, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.019805863319933326, | |
| "grad_norm": 4.689967632293701, | |
| "learning_rate": 0.00019785874696801202, | |
| "loss": 2.2133, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.02000196097656633, | |
| "grad_norm": 10.993409156799316, | |
| "learning_rate": 0.00019781160122787046, | |
| "loss": 2.314, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.020198058633199335, | |
| "grad_norm": 8.199251174926758, | |
| "learning_rate": 0.00019776394784123177, | |
| "loss": 2.5164, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.020394156289832335, | |
| "grad_norm": 15.144885063171387, | |
| "learning_rate": 0.00019771578705541916, | |
| "loss": 2.0058, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.02059025394646534, | |
| "grad_norm": 5.252450466156006, | |
| "learning_rate": 0.00019766711912038915, | |
| "loss": 1.7012, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.020786351603098344, | |
| "grad_norm": 8.265049934387207, | |
| "learning_rate": 0.0001976179442887305, | |
| "loss": 1.8646, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.020982449259731348, | |
| "grad_norm": 8.365408897399902, | |
| "learning_rate": 0.00019756826281566272, | |
| "loss": 1.9615, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.02117854691636435, | |
| "grad_norm": 7.514213562011719, | |
| "learning_rate": 0.00019751807495903484, | |
| "loss": 1.4897, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.021374644572997353, | |
| "grad_norm": 15.234655380249023, | |
| "learning_rate": 0.00019746738097932407, | |
| "loss": 2.0467, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.021570742229630357, | |
| "grad_norm": 6.856448650360107, | |
| "learning_rate": 0.0001974161811396343, | |
| "loss": 1.4492, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.021766839886263357, | |
| "grad_norm": 7.893224716186523, | |
| "learning_rate": 0.00019736447570569503, | |
| "loss": 1.919, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.02196293754289636, | |
| "grad_norm": 8.966511726379395, | |
| "learning_rate": 0.0001973122649458597, | |
| "loss": 2.4484, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.022159035199529366, | |
| "grad_norm": 7.631579875946045, | |
| "learning_rate": 0.00019725954913110442, | |
| "loss": 1.4992, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.02235513285616237, | |
| "grad_norm": 7.418518543243408, | |
| "learning_rate": 0.0001972063285350266, | |
| "loss": 0.8401, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.02255123051279537, | |
| "grad_norm": 7.739930629730225, | |
| "learning_rate": 0.00019715260343384347, | |
| "loss": 2.0713, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.022747328169428375, | |
| "grad_norm": 6.441893100738525, | |
| "learning_rate": 0.00019709837410639063, | |
| "loss": 1.4438, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.02294342582606138, | |
| "grad_norm": 6.008083820343018, | |
| "learning_rate": 0.0001970436408341207, | |
| "loss": 1.3503, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.023139523482694383, | |
| "grad_norm": 7.100820541381836, | |
| "learning_rate": 0.00019698840390110176, | |
| "loss": 1.4726, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.023335621139327384, | |
| "grad_norm": 10.213756561279297, | |
| "learning_rate": 0.0001969326635940159, | |
| "loss": 0.8107, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.023531718795960388, | |
| "grad_norm": 5.251387119293213, | |
| "learning_rate": 0.00019687642020215775, | |
| "loss": 1.5542, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.023727816452593392, | |
| "grad_norm": 6.100740432739258, | |
| "learning_rate": 0.00019681967401743297, | |
| "loss": 1.2512, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.023923914109226396, | |
| "grad_norm": 7.356696128845215, | |
| "learning_rate": 0.00019676242533435678, | |
| "loss": 2.4725, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.024120011765859397, | |
| "grad_norm": 11.542431831359863, | |
| "learning_rate": 0.00019670467445005233, | |
| "loss": 3.0307, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.0243161094224924, | |
| "grad_norm": 12.166086196899414, | |
| "learning_rate": 0.00019664642166424928, | |
| "loss": 1.2784, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.024512207079125405, | |
| "grad_norm": 5.222433090209961, | |
| "learning_rate": 0.00019658766727928206, | |
| "loss": 1.1759, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.024708304735758406, | |
| "grad_norm": 4.77174711227417, | |
| "learning_rate": 0.00019652841160008858, | |
| "loss": 1.1041, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.02490440239239141, | |
| "grad_norm": 4.879274368286133, | |
| "learning_rate": 0.0001964686549342084, | |
| "loss": 2.6326, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.025100500049024414, | |
| "grad_norm": 14.171689987182617, | |
| "learning_rate": 0.00019640839759178116, | |
| "loss": 3.4144, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.02529659770565742, | |
| "grad_norm": 7.598373889923096, | |
| "learning_rate": 0.00019634763988554522, | |
| "loss": 2.0596, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.02549269536229042, | |
| "grad_norm": 6.88770866394043, | |
| "learning_rate": 0.00019628638213083565, | |
| "loss": 1.4691, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.025688793018923423, | |
| "grad_norm": 7.128096580505371, | |
| "learning_rate": 0.00019622462464558295, | |
| "loss": 1.3307, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.025884890675556427, | |
| "grad_norm": 6.430881500244141, | |
| "learning_rate": 0.00019616236775031113, | |
| "loss": 0.9491, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.02608098833218943, | |
| "grad_norm": 9.912070274353027, | |
| "learning_rate": 0.00019609961176813624, | |
| "loss": 2.5006, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.026277085988822432, | |
| "grad_norm": 8.550467491149902, | |
| "learning_rate": 0.0001960363570247645, | |
| "loss": 2.4952, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.026473183645455436, | |
| "grad_norm": 4.201476573944092, | |
| "learning_rate": 0.0001959726038484909, | |
| "loss": 0.9033, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.02666928130208844, | |
| "grad_norm": 5.774847984313965, | |
| "learning_rate": 0.00019590835257019714, | |
| "loss": 2.1291, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.026865378958721445, | |
| "grad_norm": 8.179195404052734, | |
| "learning_rate": 0.00019584360352335023, | |
| "loss": 2.7527, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.027061476615354445, | |
| "grad_norm": 15.658841133117676, | |
| "learning_rate": 0.0001957783570440005, | |
| "loss": 1.8304, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.02725757427198745, | |
| "grad_norm": 5.7399163246154785, | |
| "learning_rate": 0.0001957126134707801, | |
| "loss": 1.7071, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.027453671928620454, | |
| "grad_norm": 5.0817389488220215, | |
| "learning_rate": 0.00019564637314490108, | |
| "loss": 1.8933, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.027649769585253458, | |
| "grad_norm": 5.634946346282959, | |
| "learning_rate": 0.0001955796364101535, | |
| "loss": 1.7343, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.02784586724188646, | |
| "grad_norm": 6.406938552856445, | |
| "learning_rate": 0.00019551240361290407, | |
| "loss": 2.3013, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.028041964898519463, | |
| "grad_norm": 8.239458084106445, | |
| "learning_rate": 0.00019544467510209388, | |
| "loss": 1.2177, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.028238062555152467, | |
| "grad_norm": 11.887965202331543, | |
| "learning_rate": 0.0001953764512292369, | |
| "loss": 2.4312, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.028434160211785468, | |
| "grad_norm": 7.482359409332275, | |
| "learning_rate": 0.00019530773234841803, | |
| "loss": 1.1083, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.028630257868418472, | |
| "grad_norm": 8.86729621887207, | |
| "learning_rate": 0.00019523851881629126, | |
| "loss": 1.6451, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.028826355525051476, | |
| "grad_norm": 7.395509719848633, | |
| "learning_rate": 0.0001951688109920778, | |
| "loss": 1.31, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.02902245318168448, | |
| "grad_norm": 4.955163955688477, | |
| "learning_rate": 0.00019509860923756442, | |
| "loss": 2.5206, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.02921855083831748, | |
| "grad_norm": 5.034746170043945, | |
| "learning_rate": 0.00019502791391710125, | |
| "loss": 0.9336, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.029414648494950485, | |
| "grad_norm": 12.375234603881836, | |
| "learning_rate": 0.00019495672539760007, | |
| "loss": 2.1276, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.02961074615158349, | |
| "grad_norm": 5.832932949066162, | |
| "learning_rate": 0.00019488504404853248, | |
| "loss": 1.3252, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.029806843808216493, | |
| "grad_norm": 5.934417724609375, | |
| "learning_rate": 0.00019481287024192775, | |
| "loss": 1.5907, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.030002941464849494, | |
| "grad_norm": 9.238896369934082, | |
| "learning_rate": 0.00019474020435237117, | |
| "loss": 1.1184, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.030199039121482498, | |
| "grad_norm": 9.787931442260742, | |
| "learning_rate": 0.00019466704675700185, | |
| "loss": 1.4931, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.030395136778115502, | |
| "grad_norm": 7.260796070098877, | |
| "learning_rate": 0.00019459339783551094, | |
| "loss": 0.8924, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.030591234434748506, | |
| "grad_norm": 8.712836265563965, | |
| "learning_rate": 0.00019451925797013954, | |
| "loss": 1.586, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.030787332091381507, | |
| "grad_norm": 11.15104866027832, | |
| "learning_rate": 0.00019444462754567682, | |
| "loss": 1.5007, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.03098342974801451, | |
| "grad_norm": 7.158255100250244, | |
| "learning_rate": 0.00019436950694945798, | |
| "loss": 2.4118, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.031179527404647515, | |
| "grad_norm": 11.58385944366455, | |
| "learning_rate": 0.00019429389657136213, | |
| "loss": 2.1638, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.03137562506128052, | |
| "grad_norm": 7.469117641448975, | |
| "learning_rate": 0.00019421779680381054, | |
| "loss": 3.0682, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.031571722717913524, | |
| "grad_norm": 10.78966999053955, | |
| "learning_rate": 0.00019414120804176426, | |
| "loss": 1.1822, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.03176782037454652, | |
| "grad_norm": 9.68694019317627, | |
| "learning_rate": 0.00019406413068272238, | |
| "loss": 2.5351, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.031963918031179525, | |
| "grad_norm": 11.67428970336914, | |
| "learning_rate": 0.00019398656512671972, | |
| "loss": 1.9244, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.03216001568781253, | |
| "grad_norm": 12.72513198852539, | |
| "learning_rate": 0.00019390851177632497, | |
| "loss": 3.2138, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.03235611334444553, | |
| "grad_norm": 8.345921516418457, | |
| "learning_rate": 0.00019382997103663838, | |
| "loss": 2.6435, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.03255221100107854, | |
| "grad_norm": 7.740304470062256, | |
| "learning_rate": 0.0001937509433152899, | |
| "loss": 0.8189, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.03274830865771154, | |
| "grad_norm": 9.329862594604492, | |
| "learning_rate": 0.0001936714290224368, | |
| "loss": 1.4106, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.032944406314344546, | |
| "grad_norm": 7.179844379425049, | |
| "learning_rate": 0.00019359142857076176, | |
| "loss": 1.8125, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.03314050397097755, | |
| "grad_norm": 7.835447311401367, | |
| "learning_rate": 0.00019351094237547066, | |
| "loss": 1.6617, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.03333660162761055, | |
| "grad_norm": 6.018518924713135, | |
| "learning_rate": 0.0001934299708542904, | |
| "loss": 2.4333, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.03353269928424355, | |
| "grad_norm": 8.176468849182129, | |
| "learning_rate": 0.00019334851442746664, | |
| "loss": 2.5915, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.033728796940876556, | |
| "grad_norm": 8.241739273071289, | |
| "learning_rate": 0.00019326657351776186, | |
| "loss": 1.666, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.03392489459750956, | |
| "grad_norm": 8.064835548400879, | |
| "learning_rate": 0.000193184148550453, | |
| "loss": 1.477, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.034120992254142564, | |
| "grad_norm": 5.790217399597168, | |
| "learning_rate": 0.00019310123995332917, | |
| "loss": 0.7703, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.03431708991077557, | |
| "grad_norm": 9.38430118560791, | |
| "learning_rate": 0.00019301784815668974, | |
| "loss": 1.5785, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.03451318756740857, | |
| "grad_norm": 8.252826690673828, | |
| "learning_rate": 0.00019293397359334167, | |
| "loss": 2.1462, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.03470928522404157, | |
| "grad_norm": 12.65652847290039, | |
| "learning_rate": 0.00019284961669859766, | |
| "loss": 1.3009, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.034905382880674574, | |
| "grad_norm": 6.8490753173828125, | |
| "learning_rate": 0.00019276477791027374, | |
| "loss": 2.4905, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.03510148053730758, | |
| "grad_norm": 4.2581048011779785, | |
| "learning_rate": 0.0001926794576686869, | |
| "loss": 0.9042, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.03529757819394058, | |
| "grad_norm": 6.415445327758789, | |
| "learning_rate": 0.0001925936564166529, | |
| "loss": 2.238, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.035493675850573586, | |
| "grad_norm": 13.620756149291992, | |
| "learning_rate": 0.00019250737459948405, | |
| "loss": 1.5966, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.03568977350720659, | |
| "grad_norm": 10.609662055969238, | |
| "learning_rate": 0.00019242061266498675, | |
| "loss": 1.081, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.035885871163839594, | |
| "grad_norm": 8.404073715209961, | |
| "learning_rate": 0.00019233337106345925, | |
| "loss": 1.849, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.0360819688204726, | |
| "grad_norm": 5.560455322265625, | |
| "learning_rate": 0.00019224565024768926, | |
| "loss": 1.4533, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.036278066477105596, | |
| "grad_norm": 7.896220684051514, | |
| "learning_rate": 0.00019215745067295169, | |
| "loss": 2.482, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.0364741641337386, | |
| "grad_norm": 9.554024696350098, | |
| "learning_rate": 0.00019206877279700612, | |
| "loss": 1.9367, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.036670261790371604, | |
| "grad_norm": 3.333113193511963, | |
| "learning_rate": 0.00019197961708009473, | |
| "loss": 1.1477, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.03686635944700461, | |
| "grad_norm": 9.468240737915039, | |
| "learning_rate": 0.00019188998398493953, | |
| "loss": 1.0849, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.03706245710363761, | |
| "grad_norm": 10.807921409606934, | |
| "learning_rate": 0.00019179987397674022, | |
| "loss": 2.0192, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.03725855476027062, | |
| "grad_norm": 7.14724588394165, | |
| "learning_rate": 0.0001917092875231717, | |
| "loss": 2.1502, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.03745465241690362, | |
| "grad_norm": 12.262707710266113, | |
| "learning_rate": 0.00019161822509438162, | |
| "loss": 2.423, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.03765075007353662, | |
| "grad_norm": 35.0489387512207, | |
| "learning_rate": 0.000191526687162988, | |
| "loss": 2.5959, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.03784684773016962, | |
| "grad_norm": 6.615735054016113, | |
| "learning_rate": 0.0001914346742040767, | |
| "loss": 1.7733, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.038042945386802626, | |
| "grad_norm": 4.537426471710205, | |
| "learning_rate": 0.00019134218669519896, | |
| "loss": 1.0028, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.03823904304343563, | |
| "grad_norm": 5.247801303863525, | |
| "learning_rate": 0.00019124922511636912, | |
| "loss": 0.8412, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.038435140700068635, | |
| "grad_norm": 6.2183918952941895, | |
| "learning_rate": 0.00019115578995006173, | |
| "loss": 1.7212, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.03863123835670164, | |
| "grad_norm": 9.330825805664062, | |
| "learning_rate": 0.00019106188168120948, | |
| "loss": 1.5341, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.03882733601333464, | |
| "grad_norm": 9.86260986328125, | |
| "learning_rate": 0.00019096750079720037, | |
| "loss": 2.765, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.03902343366996765, | |
| "grad_norm": 10.341052055358887, | |
| "learning_rate": 0.00019087264778787534, | |
| "loss": 1.9024, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.039219531326600644, | |
| "grad_norm": 9.549159049987793, | |
| "learning_rate": 0.00019077732314552566, | |
| "loss": 1.2644, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.03941562898323365, | |
| "grad_norm": 5.25094747543335, | |
| "learning_rate": 0.00019068152736489036, | |
| "loss": 1.334, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.03961172663986665, | |
| "grad_norm": 7.197662830352783, | |
| "learning_rate": 0.00019058526094315378, | |
| "loss": 1.9093, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.03980782429649966, | |
| "grad_norm": 8.476766586303711, | |
| "learning_rate": 0.0001904885243799429, | |
| "loss": 1.477, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.04000392195313266, | |
| "grad_norm": 8.232537269592285, | |
| "learning_rate": 0.00019039131817732462, | |
| "loss": 1.4013, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.040200019609765665, | |
| "grad_norm": 15.687997817993164, | |
| "learning_rate": 0.0001902936428398035, | |
| "loss": 1.6772, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.04039611726639867, | |
| "grad_norm": 7.573246479034424, | |
| "learning_rate": 0.00019019549887431877, | |
| "loss": 1.5007, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.040592214923031666, | |
| "grad_norm": 11.531679153442383, | |
| "learning_rate": 0.0001900968867902419, | |
| "loss": 2.6798, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.04078831257966467, | |
| "grad_norm": 6.225399494171143, | |
| "learning_rate": 0.00018999780709937398, | |
| "loss": 1.3078, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.040984410236297675, | |
| "grad_norm": 10.358306884765625, | |
| "learning_rate": 0.0001898982603159429, | |
| "loss": 1.7353, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.04118050789293068, | |
| "grad_norm": 8.146821975708008, | |
| "learning_rate": 0.00018979824695660087, | |
| "loss": 1.415, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.04137660554956368, | |
| "grad_norm": 4.390834808349609, | |
| "learning_rate": 0.00018969776754042156, | |
| "loss": 1.7612, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.04157270320619669, | |
| "grad_norm": 7.958174228668213, | |
| "learning_rate": 0.0001895968225888976, | |
| "loss": 2.6614, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.04176880086282969, | |
| "grad_norm": 9.981225967407227, | |
| "learning_rate": 0.00018949541262593762, | |
| "loss": 2.0158, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.041964898519462696, | |
| "grad_norm": 4.456605911254883, | |
| "learning_rate": 0.00018939353817786387, | |
| "loss": 1.0621, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.04216099617609569, | |
| "grad_norm": 7.546274662017822, | |
| "learning_rate": 0.00018929119977340917, | |
| "loss": 1.7333, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.0423570938327287, | |
| "grad_norm": 11.629569053649902, | |
| "learning_rate": 0.0001891883979437143, | |
| "loss": 1.4268, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.0425531914893617, | |
| "grad_norm": 17.710948944091797, | |
| "learning_rate": 0.00018908513322232528, | |
| "loss": 2.8701, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.042749289145994705, | |
| "grad_norm": 6.267049789428711, | |
| "learning_rate": 0.00018898140614519054, | |
| "loss": 1.6313, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.04294538680262771, | |
| "grad_norm": 4.971591949462891, | |
| "learning_rate": 0.00018887721725065814, | |
| "loss": 2.0962, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.043141484459260714, | |
| "grad_norm": 5.603585243225098, | |
| "learning_rate": 0.00018877256707947306, | |
| "loss": 0.6683, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.04333758211589372, | |
| "grad_norm": 6.029137134552002, | |
| "learning_rate": 0.00018866745617477423, | |
| "loss": 1.5375, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.043533679772526715, | |
| "grad_norm": 7.4105143547058105, | |
| "learning_rate": 0.00018856188508209183, | |
| "loss": 1.9524, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.04372977742915972, | |
| "grad_norm": 8.321500778198242, | |
| "learning_rate": 0.00018845585434934452, | |
| "loss": 2.1109, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.04392587508579272, | |
| "grad_norm": 9.238992691040039, | |
| "learning_rate": 0.00018834936452683638, | |
| "loss": 1.4247, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.04412197274242573, | |
| "grad_norm": 5.125700950622559, | |
| "learning_rate": 0.00018824241616725434, | |
| "loss": 1.1266, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.04431807039905873, | |
| "grad_norm": 7.538069725036621, | |
| "learning_rate": 0.000188135009825665, | |
| "loss": 2.1554, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.044514168055691736, | |
| "grad_norm": 8.309137344360352, | |
| "learning_rate": 0.00018802714605951199, | |
| "loss": 1.1435, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.04471026571232474, | |
| "grad_norm": 22.02942657470703, | |
| "learning_rate": 0.00018791882542861302, | |
| "loss": 1.8154, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.044906363368957744, | |
| "grad_norm": 7.017299652099609, | |
| "learning_rate": 0.0001878100484951569, | |
| "loss": 1.4998, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.04510246102559074, | |
| "grad_norm": 18.39406394958496, | |
| "learning_rate": 0.00018770081582370068, | |
| "loss": 2.1662, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.045298558682223745, | |
| "grad_norm": 9.11802864074707, | |
| "learning_rate": 0.0001875911279811667, | |
| "loss": 0.7446, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.04549465633885675, | |
| "grad_norm": 7.193735122680664, | |
| "learning_rate": 0.00018748098553683968, | |
| "loss": 1.9472, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.045690753995489754, | |
| "grad_norm": 23.407245635986328, | |
| "learning_rate": 0.0001873703890623637, | |
| "loss": 2.1782, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.04588685165212276, | |
| "grad_norm": 6.547053813934326, | |
| "learning_rate": 0.00018725933913173938, | |
| "loss": 1.9687, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.04608294930875576, | |
| "grad_norm": 10.576699256896973, | |
| "learning_rate": 0.00018714783632132068, | |
| "loss": 1.8832, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.046279046965388766, | |
| "grad_norm": 5.852027416229248, | |
| "learning_rate": 0.00018703588120981207, | |
| "loss": 1.8932, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.04647514462202176, | |
| "grad_norm": 7.023755073547363, | |
| "learning_rate": 0.00018692347437826548, | |
| "loss": 3.7953, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.04667124227865477, | |
| "grad_norm": 13.61612606048584, | |
| "learning_rate": 0.00018681061641007737, | |
| "loss": 1.9077, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.04686733993528777, | |
| "grad_norm": 5.3344526290893555, | |
| "learning_rate": 0.0001866973078909854, | |
| "loss": 1.4342, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.047063437591920776, | |
| "grad_norm": 38.80408477783203, | |
| "learning_rate": 0.00018658354940906586, | |
| "loss": 2.3665, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.04725953524855378, | |
| "grad_norm": 9.670344352722168, | |
| "learning_rate": 0.00018646934155473022, | |
| "loss": 0.9006, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.047455632905186784, | |
| "grad_norm": 5.1102495193481445, | |
| "learning_rate": 0.00018635468492072228, | |
| "loss": 1.2289, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.04765173056181979, | |
| "grad_norm": 9.1209077835083, | |
| "learning_rate": 0.00018623958010211493, | |
| "loss": 1.6009, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.04784782821845279, | |
| "grad_norm": 16.793027877807617, | |
| "learning_rate": 0.0001861240276963073, | |
| "loss": 0.94, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.04804392587508579, | |
| "grad_norm": 6.90054988861084, | |
| "learning_rate": 0.00018600802830302134, | |
| "loss": 1.559, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.048240023531718794, | |
| "grad_norm": 13.111268043518066, | |
| "learning_rate": 0.0001858915825242991, | |
| "loss": 2.1186, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.0484361211883518, | |
| "grad_norm": 6.356579780578613, | |
| "learning_rate": 0.00018577469096449925, | |
| "loss": 1.6653, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.0486322188449848, | |
| "grad_norm": 9.505541801452637, | |
| "learning_rate": 0.00018565735423029404, | |
| "loss": 0.9774, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.048828316501617806, | |
| "grad_norm": 8.927581787109375, | |
| "learning_rate": 0.00018553957293066632, | |
| "loss": 2.6455, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.04902441415825081, | |
| "grad_norm": 7.568793773651123, | |
| "learning_rate": 0.00018542134767690616, | |
| "loss": 1.1464, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.049220511814883815, | |
| "grad_norm": 7.632232189178467, | |
| "learning_rate": 0.00018530267908260784, | |
| "loss": 1.2671, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.04941660947151681, | |
| "grad_norm": 4.4279561042785645, | |
| "learning_rate": 0.00018518356776366657, | |
| "loss": 2.0384, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.049612707128149816, | |
| "grad_norm": 10.818602561950684, | |
| "learning_rate": 0.00018506401433827528, | |
| "loss": 1.0559, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.04980880478478282, | |
| "grad_norm": 5.57148551940918, | |
| "learning_rate": 0.00018494401942692153, | |
| "loss": 0.9603, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.050004902441415824, | |
| "grad_norm": 11.1985502243042, | |
| "learning_rate": 0.00018482358365238413, | |
| "loss": 2.4928, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.05020100009804883, | |
| "grad_norm": 4.890799522399902, | |
| "learning_rate": 0.00018470270763973004, | |
| "loss": 1.4034, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.05039709775468183, | |
| "grad_norm": 6.2078680992126465, | |
| "learning_rate": 0.00018458139201631108, | |
| "loss": 1.782, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.05059319541131484, | |
| "grad_norm": 24.89278221130371, | |
| "learning_rate": 0.00018445963741176065, | |
| "loss": 3.7879, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.05078929306794784, | |
| "grad_norm": 5.363570213317871, | |
| "learning_rate": 0.00018433744445799045, | |
| "loss": 1.4292, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.05098539072458084, | |
| "grad_norm": 7.669764041900635, | |
| "learning_rate": 0.0001842148137891873, | |
| "loss": 2.0483, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.05118148838121384, | |
| "grad_norm": 5.229150295257568, | |
| "learning_rate": 0.00018409174604180976, | |
| "loss": 3.2863, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.05137758603784685, | |
| "grad_norm": 5.850373268127441, | |
| "learning_rate": 0.0001839682418545848, | |
| "loss": 1.8197, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.05157368369447985, | |
| "grad_norm": 7.138283729553223, | |
| "learning_rate": 0.00018384430186850454, | |
| "loss": 2.7101, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.051769781351112855, | |
| "grad_norm": 10.918169975280762, | |
| "learning_rate": 0.000183719926726823, | |
| "loss": 1.8243, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.05196587900774586, | |
| "grad_norm": 9.205517768859863, | |
| "learning_rate": 0.00018359511707505258, | |
| "loss": 1.4992, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.05216197666437886, | |
| "grad_norm": 8.567139625549316, | |
| "learning_rate": 0.00018346987356096086, | |
| "loss": 1.051, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.05235807432101187, | |
| "grad_norm": 10.313075065612793, | |
| "learning_rate": 0.00018334419683456717, | |
| "loss": 2.6062, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.052554171977644865, | |
| "grad_norm": 7.515801906585693, | |
| "learning_rate": 0.0001832180875481392, | |
| "loss": 1.266, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.05275026963427787, | |
| "grad_norm": 5.345809459686279, | |
| "learning_rate": 0.00018309154635618965, | |
| "loss": 1.2526, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.05294636729091087, | |
| "grad_norm": 13.568882942199707, | |
| "learning_rate": 0.00018296457391547296, | |
| "loss": 2.5183, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.05314246494754388, | |
| "grad_norm": 10.022235870361328, | |
| "learning_rate": 0.00018283717088498155, | |
| "loss": 2.2774, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.05333856260417688, | |
| "grad_norm": 6.537176132202148, | |
| "learning_rate": 0.0001827093379259428, | |
| "loss": 1.4989, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.053534660260809885, | |
| "grad_norm": 17.213987350463867, | |
| "learning_rate": 0.00018258107570181533, | |
| "loss": 2.4885, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.05373075791744289, | |
| "grad_norm": 6.48647403717041, | |
| "learning_rate": 0.00018245238487828573, | |
| "loss": 1.2309, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.05392685557407589, | |
| "grad_norm": 5.479822158813477, | |
| "learning_rate": 0.000182323266123265, | |
| "loss": 1.8959, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.05412295323070889, | |
| "grad_norm": 7.716124534606934, | |
| "learning_rate": 0.00018219372010688515, | |
| "loss": 1.8321, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.054319050887341895, | |
| "grad_norm": 9.968965530395508, | |
| "learning_rate": 0.00018206374750149567, | |
| "loss": 4.1652, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.0545151485439749, | |
| "grad_norm": 6.009235382080078, | |
| "learning_rate": 0.00018193334898166007, | |
| "loss": 0.8178, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.0547112462006079, | |
| "grad_norm": 8.031886100769043, | |
| "learning_rate": 0.00018180252522415242, | |
| "loss": 1.783, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.05490734385724091, | |
| "grad_norm": 5.5589680671691895, | |
| "learning_rate": 0.00018167127690795368, | |
| "loss": 1.3049, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.05510344151387391, | |
| "grad_norm": 5.04995059967041, | |
| "learning_rate": 0.0001815396047142485, | |
| "loss": 0.8962, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.055299539170506916, | |
| "grad_norm": 5.3526692390441895, | |
| "learning_rate": 0.0001814075093264212, | |
| "loss": 1.201, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.05549563682713991, | |
| "grad_norm": 11.980429649353027, | |
| "learning_rate": 0.00018127499143005268, | |
| "loss": 0.6955, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.05569173448377292, | |
| "grad_norm": 38.28229904174805, | |
| "learning_rate": 0.00018114205171291663, | |
| "loss": 1.7335, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.05588783214040592, | |
| "grad_norm": 6.15138053894043, | |
| "learning_rate": 0.000181008690864976, | |
| "loss": 1.2766, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.056083929797038926, | |
| "grad_norm": 7.846836566925049, | |
| "learning_rate": 0.00018087490957837944, | |
| "loss": 1.155, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.05628002745367193, | |
| "grad_norm": 7.675628185272217, | |
| "learning_rate": 0.00018074070854745772, | |
| "loss": 1.6129, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.056476125110304934, | |
| "grad_norm": 12.245649337768555, | |
| "learning_rate": 0.00018060608846872005, | |
| "loss": 1.7585, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.05667222276693794, | |
| "grad_norm": 10.520101547241211, | |
| "learning_rate": 0.00018047105004085053, | |
| "loss": 1.9265, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.056868320423570935, | |
| "grad_norm": 7.400151252746582, | |
| "learning_rate": 0.00018033559396470454, | |
| "loss": 1.4189, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.05706441808020394, | |
| "grad_norm": 12.058060646057129, | |
| "learning_rate": 0.00018019972094330503, | |
| "loss": 2.3312, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.057260515736836944, | |
| "grad_norm": 5.313794136047363, | |
| "learning_rate": 0.00018006343168183893, | |
| "loss": 2.0051, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.05745661339346995, | |
| "grad_norm": 11.182997703552246, | |
| "learning_rate": 0.0001799267268876535, | |
| "loss": 1.4779, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.05765271105010295, | |
| "grad_norm": 16.24866485595703, | |
| "learning_rate": 0.0001797896072702526, | |
| "loss": 2.4689, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.057848808706735956, | |
| "grad_norm": 7.471411228179932, | |
| "learning_rate": 0.00017965207354129307, | |
| "loss": 3.0599, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.05804490636336896, | |
| "grad_norm": 7.715878486633301, | |
| "learning_rate": 0.00017951412641458098, | |
| "loss": 0.8256, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.058241004020001964, | |
| "grad_norm": 22.084482192993164, | |
| "learning_rate": 0.000179375766606068, | |
| "loss": 2.457, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.05843710167663496, | |
| "grad_norm": 8.041847229003906, | |
| "learning_rate": 0.00017923699483384753, | |
| "loss": 1.5642, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.058633199333267966, | |
| "grad_norm": 12.814888000488281, | |
| "learning_rate": 0.00017909781181815117, | |
| "loss": 1.5129, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.05882929698990097, | |
| "grad_norm": 9.216371536254883, | |
| "learning_rate": 0.0001789582182813449, | |
| "loss": 2.0632, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.059025394646533974, | |
| "grad_norm": 12.80371379852295, | |
| "learning_rate": 0.00017881821494792528, | |
| "loss": 2.8705, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.05922149230316698, | |
| "grad_norm": 7.234943389892578, | |
| "learning_rate": 0.00017867780254451576, | |
| "loss": 2.6664, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.05941758995979998, | |
| "grad_norm": 11.168726921081543, | |
| "learning_rate": 0.00017853698179986282, | |
| "loss": 1.347, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.059613687616432987, | |
| "grad_norm": 19.369266510009766, | |
| "learning_rate": 0.00017839575344483238, | |
| "loss": 2.68, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.059809785273065984, | |
| "grad_norm": 7.1730570793151855, | |
| "learning_rate": 0.0001782541182124057, | |
| "loss": 2.3908, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.06000588292969899, | |
| "grad_norm": 7.243929862976074, | |
| "learning_rate": 0.0001781120768376759, | |
| "loss": 1.0056, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.06020198058633199, | |
| "grad_norm": 7.748988628387451, | |
| "learning_rate": 0.00017796963005784394, | |
| "loss": 2.1776, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.060398078242964996, | |
| "grad_norm": 13.446945190429688, | |
| "learning_rate": 0.0001778267786122148, | |
| "loss": 2.3275, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.060594175899598, | |
| "grad_norm": 10.720627784729004, | |
| "learning_rate": 0.0001776835232421938, | |
| "loss": 1.046, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.060790273556231005, | |
| "grad_norm": 11.274985313415527, | |
| "learning_rate": 0.00017753986469128257, | |
| "loss": 2.4269, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.06098637121286401, | |
| "grad_norm": 8.671335220336914, | |
| "learning_rate": 0.00017739580370507532, | |
| "loss": 2.1488, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.06118246886949701, | |
| "grad_norm": 8.375978469848633, | |
| "learning_rate": 0.0001772513410312548, | |
| "loss": 1.8458, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.06137856652613001, | |
| "grad_norm": 11.178112983703613, | |
| "learning_rate": 0.00017710647741958868, | |
| "loss": 2.7169, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.061574664182763014, | |
| "grad_norm": 8.29799747467041, | |
| "learning_rate": 0.00017696121362192544, | |
| "loss": 1.455, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.06177076183939602, | |
| "grad_norm": 6.712766647338867, | |
| "learning_rate": 0.00017681555039219054, | |
| "loss": 1.2604, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.06196685949602902, | |
| "grad_norm": 7.891608238220215, | |
| "learning_rate": 0.00017666948848638257, | |
| "loss": 2.1795, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.06216295715266203, | |
| "grad_norm": 5.039219379425049, | |
| "learning_rate": 0.00017652302866256916, | |
| "loss": 0.9069, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.06235905480929503, | |
| "grad_norm": 9.421103477478027, | |
| "learning_rate": 0.00017637617168088325, | |
| "loss": 2.4256, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.06255515246592804, | |
| "grad_norm": 4.435902118682861, | |
| "learning_rate": 0.000176228918303519, | |
| "loss": 1.9269, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.06275125012256104, | |
| "grad_norm": 10.938987731933594, | |
| "learning_rate": 0.00017608126929472795, | |
| "loss": 1.4649, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.06294734777919404, | |
| "grad_norm": 6.332970142364502, | |
| "learning_rate": 0.00017593322542081485, | |
| "loss": 2.0089, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.06314344543582705, | |
| "grad_norm": 6.731532573699951, | |
| "learning_rate": 0.00017578478745013392, | |
| "loss": 2.4046, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.06333954309246005, | |
| "grad_norm": 8.772012710571289, | |
| "learning_rate": 0.00017563595615308474, | |
| "loss": 1.4935, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.06353564074909304, | |
| "grad_norm": 5.693745136260986, | |
| "learning_rate": 0.00017548673230210823, | |
| "loss": 1.848, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.06373173840572605, | |
| "grad_norm": 15.056157112121582, | |
| "learning_rate": 0.0001753371166716828, | |
| "loss": 1.4598, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.06392783606235905, | |
| "grad_norm": 9.370506286621094, | |
| "learning_rate": 0.00017518711003832002, | |
| "loss": 1.4809, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.06412393371899205, | |
| "grad_norm": 19.398839950561523, | |
| "learning_rate": 0.000175036713180561, | |
| "loss": 1.0093, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.06432003137562506, | |
| "grad_norm": 4.393742084503174, | |
| "learning_rate": 0.00017488592687897193, | |
| "loss": 0.817, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.06451612903225806, | |
| "grad_norm": 6.7713799476623535, | |
| "learning_rate": 0.00017473475191614037, | |
| "loss": 2.1701, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.06471222668889107, | |
| "grad_norm": 5.920267581939697, | |
| "learning_rate": 0.00017458318907667098, | |
| "loss": 3.3491, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.06490832434552407, | |
| "grad_norm": 15.095996856689453, | |
| "learning_rate": 0.0001744312391471816, | |
| "loss": 1.7637, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.06510442200215708, | |
| "grad_norm": 9.470211029052734, | |
| "learning_rate": 0.00017427890291629893, | |
| "loss": 2.7744, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.06530051965879008, | |
| "grad_norm": 9.082067489624023, | |
| "learning_rate": 0.00017412618117465477, | |
| "loss": 3.1791, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.06549661731542308, | |
| "grad_norm": 5.174635410308838, | |
| "learning_rate": 0.0001739730747148816, | |
| "loss": 1.2189, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.06569271497205609, | |
| "grad_norm": 5.053405284881592, | |
| "learning_rate": 0.00017381958433160865, | |
| "loss": 1.7119, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.06588881262868909, | |
| "grad_norm": 5.771046161651611, | |
| "learning_rate": 0.0001736657108214578, | |
| "loss": 1.4188, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.0660849102853221, | |
| "grad_norm": 8.400517463684082, | |
| "learning_rate": 0.00017351145498303925, | |
| "loss": 2.3167, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.0662810079419551, | |
| "grad_norm": 4.6646728515625, | |
| "learning_rate": 0.0001733568176169476, | |
| "loss": 1.2102, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.06647710559858809, | |
| "grad_norm": 8.288646697998047, | |
| "learning_rate": 0.0001732017995257575, | |
| "loss": 2.4803, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.0666732032552211, | |
| "grad_norm": 10.970074653625488, | |
| "learning_rate": 0.00017304640151401967, | |
| "loss": 2.5839, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.0668693009118541, | |
| "grad_norm": 6.0125732421875, | |
| "learning_rate": 0.00017289062438825665, | |
| "loss": 1.5807, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.0670653985684871, | |
| "grad_norm": 5.844028472900391, | |
| "learning_rate": 0.0001727344689569585, | |
| "loss": 3.34, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.06726149622512011, | |
| "grad_norm": 7.1026387214660645, | |
| "learning_rate": 0.00017257793603057871, | |
| "loss": 1.4347, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.06745759388175311, | |
| "grad_norm": 9.198262214660645, | |
| "learning_rate": 0.00017242102642153016, | |
| "loss": 1.834, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.06765369153838612, | |
| "grad_norm": 5.76854133605957, | |
| "learning_rate": 0.00017226374094418044, | |
| "loss": 0.9294, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.06784978919501912, | |
| "grad_norm": 10.319186210632324, | |
| "learning_rate": 0.0001721060804148482, | |
| "loss": 2.0088, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.06804588685165212, | |
| "grad_norm": 22.298240661621094, | |
| "learning_rate": 0.00017194804565179842, | |
| "loss": 2.6901, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.06824198450828513, | |
| "grad_norm": 11.38401985168457, | |
| "learning_rate": 0.00017178963747523847, | |
| "loss": 2.6342, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.06824198450828513, | |
| "eval_loss": 0.4400941729545593, | |
| "eval_runtime": 78.7276, | |
| "eval_samples_per_second": 27.284, | |
| "eval_steps_per_second": 13.642, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.06843808216491813, | |
| "grad_norm": 7.237430095672607, | |
| "learning_rate": 0.00017163085670731371, | |
| "loss": 1.6659, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.06863417982155114, | |
| "grad_norm": 10.189397811889648, | |
| "learning_rate": 0.00017147170417210333, | |
| "loss": 1.5962, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.06883027747818414, | |
| "grad_norm": 11.650528907775879, | |
| "learning_rate": 0.00017131218069561593, | |
| "loss": 2.8224, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.06902637513481714, | |
| "grad_norm": 5.0007147789001465, | |
| "learning_rate": 0.00017115228710578534, | |
| "loss": 1.8055, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.06922247279145015, | |
| "grad_norm": 4.950777530670166, | |
| "learning_rate": 0.0001709920242324663, | |
| "loss": 1.1707, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.06941857044808314, | |
| "grad_norm": 11.658537864685059, | |
| "learning_rate": 0.0001708313929074302, | |
| "loss": 1.6848, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.06961466810471614, | |
| "grad_norm": 6.6815009117126465, | |
| "learning_rate": 0.00017067039396436058, | |
| "loss": 1.6768, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.06981076576134915, | |
| "grad_norm": 7.522084712982178, | |
| "learning_rate": 0.00017050902823884903, | |
| "loss": 1.2074, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.07000686341798215, | |
| "grad_norm": 9.470911026000977, | |
| "learning_rate": 0.00017034729656839078, | |
| "loss": 0.8594, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.07020296107461516, | |
| "grad_norm": 4.798830032348633, | |
| "learning_rate": 0.00017018519979238023, | |
| "loss": 1.6142, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.07039905873124816, | |
| "grad_norm": 4.626184940338135, | |
| "learning_rate": 0.0001700227387521068, | |
| "loss": 1.6397, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.07059515638788116, | |
| "grad_norm": 6.68535041809082, | |
| "learning_rate": 0.00016985991429075036, | |
| "loss": 0.623, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.07079125404451417, | |
| "grad_norm": 3.028799533843994, | |
| "learning_rate": 0.00016969672725337706, | |
| "loss": 0.7686, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.07098735170114717, | |
| "grad_norm": 14.592923164367676, | |
| "learning_rate": 0.00016953317848693474, | |
| "loss": 1.5171, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.07118344935778018, | |
| "grad_norm": 9.847641944885254, | |
| "learning_rate": 0.00016936926884024864, | |
| "loss": 2.5365, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.07137954701441318, | |
| "grad_norm": 5.795401573181152, | |
| "learning_rate": 0.00016920499916401707, | |
| "loss": 1.2201, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.07157564467104618, | |
| "grad_norm": 10.178254127502441, | |
| "learning_rate": 0.0001690403703108068, | |
| "loss": 2.3089, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.07177174232767919, | |
| "grad_norm": 5.105298042297363, | |
| "learning_rate": 0.00016887538313504883, | |
| "loss": 1.3503, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.07196783998431219, | |
| "grad_norm": 5.165500164031982, | |
| "learning_rate": 0.00016871003849303382, | |
| "loss": 1.9088, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.0721639376409452, | |
| "grad_norm": 6.532622337341309, | |
| "learning_rate": 0.0001685443372429077, | |
| "loss": 1.2754, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.07236003529757819, | |
| "grad_norm": 6.758758068084717, | |
| "learning_rate": 0.0001683782802446672, | |
| "loss": 1.6721, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.07255613295421119, | |
| "grad_norm": 4.5956902503967285, | |
| "learning_rate": 0.0001682118683601555, | |
| "loss": 1.7976, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.0727522306108442, | |
| "grad_norm": 8.587203025817871, | |
| "learning_rate": 0.00016804510245305745, | |
| "loss": 1.7933, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.0729483282674772, | |
| "grad_norm": 8.018026351928711, | |
| "learning_rate": 0.00016787798338889552, | |
| "loss": 2.4964, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.0731444259241102, | |
| "grad_norm": 8.505922317504883, | |
| "learning_rate": 0.00016771051203502493, | |
| "loss": 2.9766, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.07334052358074321, | |
| "grad_norm": 7.051482200622559, | |
| "learning_rate": 0.00016754268926062938, | |
| "loss": 1.9882, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.07353662123737621, | |
| "grad_norm": 4.620705604553223, | |
| "learning_rate": 0.00016737451593671636, | |
| "loss": 2.3198, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.07373271889400922, | |
| "grad_norm": 8.614477157592773, | |
| "learning_rate": 0.00016720599293611286, | |
| "loss": 2.0611, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.07392881655064222, | |
| "grad_norm": 5.840315818786621, | |
| "learning_rate": 0.0001670371211334606, | |
| "loss": 1.5961, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.07412491420727522, | |
| "grad_norm": 6.045500755310059, | |
| "learning_rate": 0.00016686790140521164, | |
| "loss": 1.5132, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.07432101186390823, | |
| "grad_norm": 11.521757125854492, | |
| "learning_rate": 0.0001666983346296238, | |
| "loss": 2.0003, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.07451710952054123, | |
| "grad_norm": 7.7568230628967285, | |
| "learning_rate": 0.0001665284216867561, | |
| "loss": 2.4289, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.07471320717717424, | |
| "grad_norm": 9.928805351257324, | |
| "learning_rate": 0.0001663581634584641, | |
| "loss": 2.2129, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.07490930483380724, | |
| "grad_norm": 6.011744022369385, | |
| "learning_rate": 0.00016618756082839554, | |
| "loss": 1.8396, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.07510540249044025, | |
| "grad_norm": 7.308403968811035, | |
| "learning_rate": 0.0001660166146819855, | |
| "loss": 0.9335, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.07530150014707324, | |
| "grad_norm": 10.000845909118652, | |
| "learning_rate": 0.0001658453259064519, | |
| "loss": 1.7884, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.07549759780370624, | |
| "grad_norm": 16.585180282592773, | |
| "learning_rate": 0.00016567369539079114, | |
| "loss": 3.0675, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.07569369546033924, | |
| "grad_norm": 10.055068969726562, | |
| "learning_rate": 0.000165501724025773, | |
| "loss": 2.0398, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.07588979311697225, | |
| "grad_norm": 9.56424331665039, | |
| "learning_rate": 0.0001653294127039365, | |
| "loss": 3.1253, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.07608589077360525, | |
| "grad_norm": 6.628775119781494, | |
| "learning_rate": 0.0001651567623195849, | |
| "loss": 0.869, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.07628198843023826, | |
| "grad_norm": 8.333356857299805, | |
| "learning_rate": 0.00016498377376878126, | |
| "loss": 2.1941, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.07647808608687126, | |
| "grad_norm": 6.76808500289917, | |
| "learning_rate": 0.0001648104479493437, | |
| "loss": 1.3802, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.07667418374350427, | |
| "grad_norm": 7.221822738647461, | |
| "learning_rate": 0.0001646367857608409, | |
| "loss": 1.0955, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.07687028140013727, | |
| "grad_norm": 31.469024658203125, | |
| "learning_rate": 0.00016446278810458716, | |
| "loss": 2.4294, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.07706637905677027, | |
| "grad_norm": 5.735899448394775, | |
| "learning_rate": 0.000164288455883638, | |
| "loss": 2.1574, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.07726247671340328, | |
| "grad_norm": 6.462769031524658, | |
| "learning_rate": 0.00016411379000278524, | |
| "loss": 1.8524, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.07745857437003628, | |
| "grad_norm": 12.341320037841797, | |
| "learning_rate": 0.00016393879136855248, | |
| "loss": 2.2439, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.07765467202666929, | |
| "grad_norm": 6.433187007904053, | |
| "learning_rate": 0.0001637634608891903, | |
| "loss": 1.2662, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.07785076968330229, | |
| "grad_norm": 5.184267997741699, | |
| "learning_rate": 0.00016358779947467158, | |
| "loss": 1.0905, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.0780468673399353, | |
| "grad_norm": 8.464557647705078, | |
| "learning_rate": 0.00016341180803668674, | |
| "loss": 2.0666, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.07824296499656828, | |
| "grad_norm": 9.498639106750488, | |
| "learning_rate": 0.00016323548748863907, | |
| "loss": 2.2394, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.07843906265320129, | |
| "grad_norm": 7.747230052947998, | |
| "learning_rate": 0.00016305883874563994, | |
| "loss": 1.4313, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.07863516030983429, | |
| "grad_norm": 8.759115219116211, | |
| "learning_rate": 0.0001628818627245041, | |
| "loss": 1.6428, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.0788312579664673, | |
| "grad_norm": 5.250945568084717, | |
| "learning_rate": 0.00016270456034374474, | |
| "loss": 0.7777, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.0790273556231003, | |
| "grad_norm": 9.001906394958496, | |
| "learning_rate": 0.00016252693252356916, | |
| "loss": 1.678, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.0792234532797333, | |
| "grad_norm": 7.177880764007568, | |
| "learning_rate": 0.00016234898018587337, | |
| "loss": 1.4463, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.07941955093636631, | |
| "grad_norm": 10.035067558288574, | |
| "learning_rate": 0.00016217070425423788, | |
| "loss": 2.3035, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.07961564859299931, | |
| "grad_norm": 4.974843978881836, | |
| "learning_rate": 0.0001619921056539226, | |
| "loss": 1.2633, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.07981174624963232, | |
| "grad_norm": 7.599184036254883, | |
| "learning_rate": 0.00016181318531186206, | |
| "loss": 0.8719, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.08000784390626532, | |
| "grad_norm": 11.066450119018555, | |
| "learning_rate": 0.0001616339441566607, | |
| "loss": 2.2938, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.08020394156289833, | |
| "grad_norm": 6.441056251525879, | |
| "learning_rate": 0.000161454383118588, | |
| "loss": 0.9166, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.08040003921953133, | |
| "grad_norm": 9.670071601867676, | |
| "learning_rate": 0.00016127450312957353, | |
| "loss": 2.2514, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.08059613687616433, | |
| "grad_norm": 7.870420455932617, | |
| "learning_rate": 0.00016109430512320237, | |
| "loss": 1.417, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.08079223453279734, | |
| "grad_norm": 8.569132804870605, | |
| "learning_rate": 0.00016091379003471007, | |
| "loss": 1.6958, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.08098833218943034, | |
| "grad_norm": 4.8630900382995605, | |
| "learning_rate": 0.00016073295880097784, | |
| "loss": 1.8919, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.08118442984606333, | |
| "grad_norm": 10.894848823547363, | |
| "learning_rate": 0.0001605518123605277, | |
| "loss": 1.7899, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.08138052750269634, | |
| "grad_norm": 6.055443286895752, | |
| "learning_rate": 0.00016037035165351768, | |
| "loss": 1.9432, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.08157662515932934, | |
| "grad_norm": 7.106583118438721, | |
| "learning_rate": 0.0001601885776217367, | |
| "loss": 1.736, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.08177272281596235, | |
| "grad_norm": 7.7677764892578125, | |
| "learning_rate": 0.00016000649120860003, | |
| "loss": 2.4425, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.08196882047259535, | |
| "grad_norm": 5.940224647521973, | |
| "learning_rate": 0.00015982409335914407, | |
| "loss": 1.0954, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.08216491812922835, | |
| "grad_norm": 6.025548458099365, | |
| "learning_rate": 0.00015964138502002175, | |
| "loss": 0.93, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.08236101578586136, | |
| "grad_norm": 7.528570652008057, | |
| "learning_rate": 0.00015945836713949726, | |
| "loss": 1.8381, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.08255711344249436, | |
| "grad_norm": 15.305673599243164, | |
| "learning_rate": 0.00015927504066744148, | |
| "loss": 1.7413, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.08275321109912737, | |
| "grad_norm": 4.707043647766113, | |
| "learning_rate": 0.0001590914065553268, | |
| "loss": 1.4186, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.08294930875576037, | |
| "grad_norm": 5.808046817779541, | |
| "learning_rate": 0.00015890746575622231, | |
| "loss": 1.095, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.08314540641239337, | |
| "grad_norm": 6.294839859008789, | |
| "learning_rate": 0.00015872321922478884, | |
| "loss": 1.918, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.08334150406902638, | |
| "grad_norm": 4.934250354766846, | |
| "learning_rate": 0.00015853866791727396, | |
| "loss": 1.2578, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.08353760172565938, | |
| "grad_norm": 8.470808982849121, | |
| "learning_rate": 0.00015835381279150705, | |
| "loss": 2.0231, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.08373369938229239, | |
| "grad_norm": 18.526151657104492, | |
| "learning_rate": 0.00015816865480689426, | |
| "loss": 4.1691, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.08392979703892539, | |
| "grad_norm": 7.958248615264893, | |
| "learning_rate": 0.0001579831949244137, | |
| "loss": 1.066, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.08412589469555838, | |
| "grad_norm": 6.250977516174316, | |
| "learning_rate": 0.00015779743410661033, | |
| "loss": 0.8756, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.08432199235219139, | |
| "grad_norm": 4.739164352416992, | |
| "learning_rate": 0.00015761137331759084, | |
| "loss": 1.3986, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.08451809000882439, | |
| "grad_norm": 10.723671913146973, | |
| "learning_rate": 0.00015742501352301893, | |
| "loss": 2.3962, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.0847141876654574, | |
| "grad_norm": 5.097165584564209, | |
| "learning_rate": 0.00015723835569011007, | |
| "loss": 3.1665, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.0849102853220904, | |
| "grad_norm": 7.0065388679504395, | |
| "learning_rate": 0.00015705140078762665, | |
| "loss": 1.4634, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.0851063829787234, | |
| "grad_norm": 6.814738750457764, | |
| "learning_rate": 0.00015686414978587277, | |
| "loss": 0.9439, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.0853024806353564, | |
| "grad_norm": 8.381726264953613, | |
| "learning_rate": 0.0001566766036566893, | |
| "loss": 1.6046, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.08549857829198941, | |
| "grad_norm": 8.249631881713867, | |
| "learning_rate": 0.00015648876337344896, | |
| "loss": 1.6886, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.08569467594862241, | |
| "grad_norm": 9.020310401916504, | |
| "learning_rate": 0.00015630062991105098, | |
| "loss": 2.5881, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.08589077360525542, | |
| "grad_norm": 3.999058723449707, | |
| "learning_rate": 0.0001561122042459163, | |
| "loss": 1.8363, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.08608687126188842, | |
| "grad_norm": 7.487119674682617, | |
| "learning_rate": 0.00015592348735598237, | |
| "loss": 1.4798, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.08628296891852143, | |
| "grad_norm": 4.982146739959717, | |
| "learning_rate": 0.00015573448022069815, | |
| "loss": 2.7132, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.08647906657515443, | |
| "grad_norm": 3.5383474826812744, | |
| "learning_rate": 0.00015554518382101892, | |
| "loss": 0.6142, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.08667516423178744, | |
| "grad_norm": 3.881410598754883, | |
| "learning_rate": 0.00015535559913940126, | |
| "loss": 0.8708, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.08687126188842044, | |
| "grad_norm": 5.159698963165283, | |
| "learning_rate": 0.00015516572715979806, | |
| "loss": 2.32, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.08706735954505343, | |
| "grad_norm": 9.29107666015625, | |
| "learning_rate": 0.00015497556886765316, | |
| "loss": 2.5744, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.08726345720168643, | |
| "grad_norm": 8.70083999633789, | |
| "learning_rate": 0.00015478512524989645, | |
| "loss": 0.8412, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.08745955485831944, | |
| "grad_norm": 6.164897918701172, | |
| "learning_rate": 0.00015459439729493865, | |
| "loss": 1.4857, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.08765565251495244, | |
| "grad_norm": 4.213920593261719, | |
| "learning_rate": 0.00015440338599266622, | |
| "loss": 1.2077, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.08785175017158545, | |
| "grad_norm": 14.897852897644043, | |
| "learning_rate": 0.00015421209233443617, | |
| "loss": 1.815, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.08804784782821845, | |
| "grad_norm": 6.773970127105713, | |
| "learning_rate": 0.00015402051731307093, | |
| "loss": 2.9485, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 0.08824394548485145, | |
| "grad_norm": 5.569252014160156, | |
| "learning_rate": 0.0001538286619228533, | |
| "loss": 1.1505, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.08844004314148446, | |
| "grad_norm": 5.61276388168335, | |
| "learning_rate": 0.0001536365271595212, | |
| "loss": 1.2065, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 0.08863614079811746, | |
| "grad_norm": 3.9499528408050537, | |
| "learning_rate": 0.00015344411402026245, | |
| "loss": 1.3547, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 0.08883223845475047, | |
| "grad_norm": 11.833647727966309, | |
| "learning_rate": 0.00015325142350370967, | |
| "loss": 2.8476, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 0.08902833611138347, | |
| "grad_norm": 6.281819820404053, | |
| "learning_rate": 0.00015305845660993503, | |
| "loss": 1.0563, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 0.08922443376801648, | |
| "grad_norm": 9.338071823120117, | |
| "learning_rate": 0.00015286521434044526, | |
| "loss": 1.9847, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.08942053142464948, | |
| "grad_norm": 12.811955451965332, | |
| "learning_rate": 0.0001526716976981761, | |
| "loss": 1.2632, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.08961662908128248, | |
| "grad_norm": 5.077617168426514, | |
| "learning_rate": 0.0001524779076874875, | |
| "loss": 0.9728, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 0.08981272673791549, | |
| "grad_norm": 5.802744388580322, | |
| "learning_rate": 0.0001522838453141581, | |
| "loss": 0.9826, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 0.09000882439454848, | |
| "grad_norm": 6.643836498260498, | |
| "learning_rate": 0.00015208951158538004, | |
| "loss": 1.3868, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 0.09020492205118148, | |
| "grad_norm": 8.100834846496582, | |
| "learning_rate": 0.000151894907509754, | |
| "loss": 1.5572, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.09040101970781449, | |
| "grad_norm": 8.695015907287598, | |
| "learning_rate": 0.00015170003409728356, | |
| "loss": 1.7788, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 0.09059711736444749, | |
| "grad_norm": 11.81883430480957, | |
| "learning_rate": 0.00015150489235937035, | |
| "loss": 1.7457, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.0907932150210805, | |
| "grad_norm": 8.165855407714844, | |
| "learning_rate": 0.00015130948330880847, | |
| "loss": 1.3886, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 0.0909893126777135, | |
| "grad_norm": 22.797456741333008, | |
| "learning_rate": 0.00015111380795977954, | |
| "loss": 1.8671, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 0.0911854103343465, | |
| "grad_norm": 5.275067329406738, | |
| "learning_rate": 0.00015091786732784716, | |
| "loss": 0.7186, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.09138150799097951, | |
| "grad_norm": 8.38714599609375, | |
| "learning_rate": 0.00015072166242995175, | |
| "loss": 1.7764, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 0.09157760564761251, | |
| "grad_norm": 8.916413307189941, | |
| "learning_rate": 0.0001505251942844054, | |
| "loss": 1.5589, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 0.09177370330424552, | |
| "grad_norm": 5.007913112640381, | |
| "learning_rate": 0.00015032846391088635, | |
| "loss": 0.6688, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 0.09196980096087852, | |
| "grad_norm": 10.685876846313477, | |
| "learning_rate": 0.0001501314723304339, | |
| "loss": 2.2118, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 0.09216589861751152, | |
| "grad_norm": 7.626431465148926, | |
| "learning_rate": 0.00014993422056544295, | |
| "loss": 1.1849, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.09236199627414453, | |
| "grad_norm": 5.862109184265137, | |
| "learning_rate": 0.00014973670963965883, | |
| "loss": 1.4594, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 0.09255809393077753, | |
| "grad_norm": 5.194091320037842, | |
| "learning_rate": 0.00014953894057817188, | |
| "loss": 1.2243, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 0.09275419158741054, | |
| "grad_norm": 7.806075572967529, | |
| "learning_rate": 0.0001493409144074122, | |
| "loss": 1.2365, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 0.09295028924404353, | |
| "grad_norm": 5.004404067993164, | |
| "learning_rate": 0.00014914263215514431, | |
| "loss": 1.6081, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 0.09314638690067653, | |
| "grad_norm": 5.245405197143555, | |
| "learning_rate": 0.00014894409485046177, | |
| "loss": 2.2626, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.09334248455730954, | |
| "grad_norm": 6.017452716827393, | |
| "learning_rate": 0.00014874530352378194, | |
| "loss": 3.0452, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 0.09353858221394254, | |
| "grad_norm": 8.504718780517578, | |
| "learning_rate": 0.00014854625920684042, | |
| "loss": 1.4529, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 0.09373467987057554, | |
| "grad_norm": 7.850302696228027, | |
| "learning_rate": 0.00014834696293268603, | |
| "loss": 1.5511, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 0.09393077752720855, | |
| "grad_norm": 6.846382141113281, | |
| "learning_rate": 0.00014814741573567514, | |
| "loss": 2.0178, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 0.09412687518384155, | |
| "grad_norm": 8.834015846252441, | |
| "learning_rate": 0.00014794761865146648, | |
| "loss": 1.6438, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.09432297284047456, | |
| "grad_norm": 7.127365589141846, | |
| "learning_rate": 0.00014774757271701557, | |
| "loss": 0.5588, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 0.09451907049710756, | |
| "grad_norm": 13.859881401062012, | |
| "learning_rate": 0.00014754727897056967, | |
| "loss": 2.9212, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 0.09471516815374056, | |
| "grad_norm": 5.914462089538574, | |
| "learning_rate": 0.0001473467384516621, | |
| "loss": 1.334, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 0.09491126581037357, | |
| "grad_norm": 3.947435140609741, | |
| "learning_rate": 0.0001471459522011069, | |
| "loss": 1.5284, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 0.09510736346700657, | |
| "grad_norm": 7.171452522277832, | |
| "learning_rate": 0.00014694492126099353, | |
| "loss": 1.7676, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.09530346112363958, | |
| "grad_norm": 7.364241600036621, | |
| "learning_rate": 0.0001467436466746814, | |
| "loss": 1.7551, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 0.09549955878027258, | |
| "grad_norm": 12.173360824584961, | |
| "learning_rate": 0.0001465421294867944, | |
| "loss": 2.5703, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 0.09569565643690559, | |
| "grad_norm": 14.59663200378418, | |
| "learning_rate": 0.00014634037074321557, | |
| "loss": 1.7594, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 0.09589175409353858, | |
| "grad_norm": 5.692611217498779, | |
| "learning_rate": 0.00014613837149108163, | |
| "loss": 2.5973, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 0.09608785175017158, | |
| "grad_norm": 6.9588470458984375, | |
| "learning_rate": 0.00014593613277877758, | |
| "loss": 2.6119, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.09628394940680458, | |
| "grad_norm": 14.370820999145508, | |
| "learning_rate": 0.0001457336556559312, | |
| "loss": 1.9495, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 0.09648004706343759, | |
| "grad_norm": 8.259851455688477, | |
| "learning_rate": 0.0001455309411734076, | |
| "loss": 2.2663, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 0.09667614472007059, | |
| "grad_norm": 11.025699615478516, | |
| "learning_rate": 0.00014532799038330385, | |
| "loss": 2.2184, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 0.0968722423767036, | |
| "grad_norm": 6.014841079711914, | |
| "learning_rate": 0.00014512480433894343, | |
| "loss": 2.3436, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 0.0970683400333366, | |
| "grad_norm": 11.881712913513184, | |
| "learning_rate": 0.00014492138409487085, | |
| "loss": 0.7869, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.0972644376899696, | |
| "grad_norm": 4.720561504364014, | |
| "learning_rate": 0.00014471773070684599, | |
| "loss": 2.417, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 0.09746053534660261, | |
| "grad_norm": 11.112417221069336, | |
| "learning_rate": 0.00014451384523183903, | |
| "loss": 2.155, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 0.09765663300323561, | |
| "grad_norm": 5.968791961669922, | |
| "learning_rate": 0.0001443097287280244, | |
| "loss": 0.8096, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 0.09785273065986862, | |
| "grad_norm": 7.543048858642578, | |
| "learning_rate": 0.0001441053822547757, | |
| "loss": 1.7723, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 0.09804882831650162, | |
| "grad_norm": 7.892668724060059, | |
| "learning_rate": 0.00014390080687266013, | |
| "loss": 1.4027, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.09824492597313463, | |
| "grad_norm": 7.690077304840088, | |
| "learning_rate": 0.00014369600364343285, | |
| "loss": 1.901, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 0.09844102362976763, | |
| "grad_norm": 4.714064598083496, | |
| "learning_rate": 0.00014349097363003163, | |
| "loss": 1.6198, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 0.09863712128640063, | |
| "grad_norm": 4.764143466949463, | |
| "learning_rate": 0.0001432857178965712, | |
| "loss": 1.3231, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 0.09883321894303362, | |
| "grad_norm": 9.042250633239746, | |
| "learning_rate": 0.00014308023750833783, | |
| "loss": 2.1337, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 0.09902931659966663, | |
| "grad_norm": 7.619399547576904, | |
| "learning_rate": 0.00014287453353178372, | |
| "loss": 1.1525, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.09922541425629963, | |
| "grad_norm": 10.186138153076172, | |
| "learning_rate": 0.00014266860703452156, | |
| "loss": 2.5917, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 0.09942151191293264, | |
| "grad_norm": 6.784362316131592, | |
| "learning_rate": 0.00014246245908531882, | |
| "loss": 2.1194, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 0.09961760956956564, | |
| "grad_norm": 46.69243621826172, | |
| "learning_rate": 0.0001422560907540925, | |
| "loss": 2.0196, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 0.09981370722619864, | |
| "grad_norm": 5.98226261138916, | |
| "learning_rate": 0.00014204950311190318, | |
| "loss": 2.131, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 0.10000980488283165, | |
| "grad_norm": 19.974212646484375, | |
| "learning_rate": 0.00014184269723094988, | |
| "loss": 1.9812, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.10020590253946465, | |
| "grad_norm": 6.612239360809326, | |
| "learning_rate": 0.00014163567418456406, | |
| "loss": 1.2688, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 0.10040200019609766, | |
| "grad_norm": 5.024367332458496, | |
| "learning_rate": 0.0001414284350472045, | |
| "loss": 1.0559, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 0.10059809785273066, | |
| "grad_norm": 6.517631530761719, | |
| "learning_rate": 0.00014122098089445142, | |
| "loss": 1.2611, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 0.10079419550936367, | |
| "grad_norm": 7.715580463409424, | |
| "learning_rate": 0.0001410133128030009, | |
| "loss": 1.0419, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 0.10099029316599667, | |
| "grad_norm": 5.8916120529174805, | |
| "learning_rate": 0.00014080543185065943, | |
| "loss": 0.938, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.10118639082262967, | |
| "grad_norm": 5.587378978729248, | |
| "learning_rate": 0.0001405973391163383, | |
| "loss": 1.4988, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 0.10138248847926268, | |
| "grad_norm": 7.234095096588135, | |
| "learning_rate": 0.0001403890356800479, | |
| "loss": 2.0476, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 0.10157858613589568, | |
| "grad_norm": 5.781263828277588, | |
| "learning_rate": 0.00014018052262289223, | |
| "loss": 1.9736, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 0.10177468379252867, | |
| "grad_norm": 7.789620399475098, | |
| "learning_rate": 0.0001399718010270632, | |
| "loss": 1.2162, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 0.10197078144916168, | |
| "grad_norm": 2.692659378051758, | |
| "learning_rate": 0.00013976287197583494, | |
| "loss": 0.391, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.10216687910579468, | |
| "grad_norm": 7.2949676513671875, | |
| "learning_rate": 0.0001395537365535585, | |
| "loss": 1.2454, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 0.10236297676242768, | |
| "grad_norm": 14.400397300720215, | |
| "learning_rate": 0.00013934439584565583, | |
| "loss": 1.844, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 0.10255907441906069, | |
| "grad_norm": 5.972201347351074, | |
| "learning_rate": 0.0001391348509386144, | |
| "loss": 1.2853, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 0.1027551720756937, | |
| "grad_norm": 8.810315132141113, | |
| "learning_rate": 0.00013892510291998146, | |
| "loss": 2.7965, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 0.1029512697323267, | |
| "grad_norm": 8.922815322875977, | |
| "learning_rate": 0.00013871515287835839, | |
| "loss": 1.2606, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.1031473673889597, | |
| "grad_norm": 7.5427021980285645, | |
| "learning_rate": 0.00013850500190339514, | |
| "loss": 1.0504, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 0.1033434650455927, | |
| "grad_norm": 6.646986484527588, | |
| "learning_rate": 0.00013829465108578445, | |
| "loss": 1.1522, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 0.10353956270222571, | |
| "grad_norm": 8.387070655822754, | |
| "learning_rate": 0.0001380841015172563, | |
| "loss": 1.4298, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 0.10373566035885871, | |
| "grad_norm": 21.383251190185547, | |
| "learning_rate": 0.0001378733542905722, | |
| "loss": 1.9162, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 0.10393175801549172, | |
| "grad_norm": 5.776891231536865, | |
| "learning_rate": 0.00013766241049951948, | |
| "loss": 1.2837, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.10412785567212472, | |
| "grad_norm": 4.5066094398498535, | |
| "learning_rate": 0.00013745127123890565, | |
| "loss": 1.5433, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 0.10432395332875773, | |
| "grad_norm": 6.012543201446533, | |
| "learning_rate": 0.00013723993760455272, | |
| "loss": 1.3534, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 0.10452005098539073, | |
| "grad_norm": 8.182510375976562, | |
| "learning_rate": 0.0001370284106932915, | |
| "loss": 1.5422, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 0.10471614864202373, | |
| "grad_norm": 12.530463218688965, | |
| "learning_rate": 0.00013681669160295597, | |
| "loss": 1.651, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 0.10491224629865673, | |
| "grad_norm": 8.226496696472168, | |
| "learning_rate": 0.00013660478143237746, | |
| "loss": 1.8992, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.10510834395528973, | |
| "grad_norm": 4.896340847015381, | |
| "learning_rate": 0.00013639268128137907, | |
| "loss": 2.2539, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 0.10530444161192273, | |
| "grad_norm": 4.573820114135742, | |
| "learning_rate": 0.00013618039225076986, | |
| "loss": 2.4075, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 0.10550053926855574, | |
| "grad_norm": 4.920849323272705, | |
| "learning_rate": 0.0001359679154423392, | |
| "loss": 2.2481, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 0.10569663692518874, | |
| "grad_norm": 6.206282138824463, | |
| "learning_rate": 0.00013575525195885107, | |
| "loss": 2.5488, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 0.10589273458182175, | |
| "grad_norm": 9.469644546508789, | |
| "learning_rate": 0.0001355424029040382, | |
| "loss": 2.0894, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.10608883223845475, | |
| "grad_norm": 5.231925010681152, | |
| "learning_rate": 0.00013532936938259656, | |
| "loss": 1.2463, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 0.10628492989508775, | |
| "grad_norm": 6.779435157775879, | |
| "learning_rate": 0.0001351161525001795, | |
| "loss": 1.8802, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 0.10648102755172076, | |
| "grad_norm": 7.8035197257995605, | |
| "learning_rate": 0.00013490275336339188, | |
| "loss": 1.1796, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 0.10667712520835376, | |
| "grad_norm": 5.93556022644043, | |
| "learning_rate": 0.00013468917307978467, | |
| "loss": 1.1082, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 0.10687322286498677, | |
| "grad_norm": 8.9798002243042, | |
| "learning_rate": 0.00013447541275784887, | |
| "loss": 1.6289, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.10706932052161977, | |
| "grad_norm": 4.270269393920898, | |
| "learning_rate": 0.00013426147350700996, | |
| "loss": 3.2414, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 0.10726541817825277, | |
| "grad_norm": 7.755331516265869, | |
| "learning_rate": 0.00013404735643762192, | |
| "loss": 3.5263, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 0.10746151583488578, | |
| "grad_norm": 3.364109754562378, | |
| "learning_rate": 0.0001338330626609618, | |
| "loss": 1.1847, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 0.10765761349151878, | |
| "grad_norm": 7.73459529876709, | |
| "learning_rate": 0.0001336185932892237, | |
| "loss": 2.0363, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 0.10785371114815177, | |
| "grad_norm": 7.035179138183594, | |
| "learning_rate": 0.000133403949435513, | |
| "loss": 1.3317, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.10804980880478478, | |
| "grad_norm": 5.30495023727417, | |
| "learning_rate": 0.00013318913221384076, | |
| "loss": 1.3776, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 0.10824590646141778, | |
| "grad_norm": 6.3127593994140625, | |
| "learning_rate": 0.00013297414273911784, | |
| "loss": 0.9931, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 0.10844200411805079, | |
| "grad_norm": 3.763789653778076, | |
| "learning_rate": 0.00013275898212714889, | |
| "loss": 0.9685, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 0.10863810177468379, | |
| "grad_norm": 9.935264587402344, | |
| "learning_rate": 0.00013254365149462699, | |
| "loss": 1.1951, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 0.1088341994313168, | |
| "grad_norm": 6.151608943939209, | |
| "learning_rate": 0.00013232815195912754, | |
| "loss": 1.9978, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.1090302970879498, | |
| "grad_norm": 7.497893333435059, | |
| "learning_rate": 0.00013211248463910262, | |
| "loss": 2.1158, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 0.1092263947445828, | |
| "grad_norm": 4.703649520874023, | |
| "learning_rate": 0.00013189665065387507, | |
| "loss": 2.9424, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 0.1094224924012158, | |
| "grad_norm": 6.602818489074707, | |
| "learning_rate": 0.00013168065112363264, | |
| "loss": 1.9994, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 0.10961859005784881, | |
| "grad_norm": 5.672735691070557, | |
| "learning_rate": 0.00013146448716942245, | |
| "loss": 1.6569, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 0.10981468771448182, | |
| "grad_norm": 6.252768039703369, | |
| "learning_rate": 0.0001312481599131449, | |
| "loss": 1.0263, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.11001078537111482, | |
| "grad_norm": 5.062610626220703, | |
| "learning_rate": 0.00013103167047754784, | |
| "loss": 1.8557, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 0.11020688302774782, | |
| "grad_norm": 17.327529907226562, | |
| "learning_rate": 0.000130815019986221, | |
| "loss": 2.2288, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 0.11040298068438083, | |
| "grad_norm": 9.787447929382324, | |
| "learning_rate": 0.00013059820956358998, | |
| "loss": 2.8692, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 0.11059907834101383, | |
| "grad_norm": 17.163454055786133, | |
| "learning_rate": 0.00013038124033491025, | |
| "loss": 2.3984, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 0.11079517599764682, | |
| "grad_norm": 7.947646617889404, | |
| "learning_rate": 0.00013016411342626168, | |
| "loss": 1.2144, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.11099127365427983, | |
| "grad_norm": 8.151859283447266, | |
| "learning_rate": 0.00012994682996454247, | |
| "loss": 1.7198, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 0.11118737131091283, | |
| "grad_norm": 8.358308792114258, | |
| "learning_rate": 0.00012972939107746325, | |
| "loss": 1.1448, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 0.11138346896754583, | |
| "grad_norm": 5.084352970123291, | |
| "learning_rate": 0.0001295117978935414, | |
| "loss": 2.3229, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 0.11157956662417884, | |
| "grad_norm": 4.846659183502197, | |
| "learning_rate": 0.0001292940515420951, | |
| "loss": 1.3941, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 0.11177566428081184, | |
| "grad_norm": 5.432335376739502, | |
| "learning_rate": 0.0001290761531532374, | |
| "loss": 1.9803, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.11197176193744485, | |
| "grad_norm": 5.953497886657715, | |
| "learning_rate": 0.00012885810385787055, | |
| "loss": 1.0619, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 0.11216785959407785, | |
| "grad_norm": 4.770197868347168, | |
| "learning_rate": 0.00012863990478767994, | |
| "loss": 2.2999, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 0.11236395725071086, | |
| "grad_norm": 5.719841957092285, | |
| "learning_rate": 0.00012842155707512825, | |
| "loss": 2.0115, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 0.11256005490734386, | |
| "grad_norm": 6.13926362991333, | |
| "learning_rate": 0.00012820306185344976, | |
| "loss": 1.7, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 0.11275615256397686, | |
| "grad_norm": 14.388799667358398, | |
| "learning_rate": 0.0001279844202566442, | |
| "loss": 1.4477, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.11295225022060987, | |
| "grad_norm": 5.510779857635498, | |
| "learning_rate": 0.00012776563341947104, | |
| "loss": 1.145, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 0.11314834787724287, | |
| "grad_norm": 7.694248199462891, | |
| "learning_rate": 0.00012754670247744354, | |
| "loss": 2.0622, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 0.11334444553387588, | |
| "grad_norm": 10.04172420501709, | |
| "learning_rate": 0.0001273276285668229, | |
| "loss": 2.1074, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 0.11354054319050888, | |
| "grad_norm": 4.613293170928955, | |
| "learning_rate": 0.00012710841282461238, | |
| "loss": 1.3278, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 0.11373664084714187, | |
| "grad_norm": 7.878372669219971, | |
| "learning_rate": 0.0001268890563885512, | |
| "loss": 3.3353, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.11393273850377487, | |
| "grad_norm": 4.252342224121094, | |
| "learning_rate": 0.0001266695603971089, | |
| "loss": 1.2478, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 0.11412883616040788, | |
| "grad_norm": 13.550718307495117, | |
| "learning_rate": 0.0001264499259894793, | |
| "loss": 2.3346, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 0.11432493381704088, | |
| "grad_norm": 5.674976348876953, | |
| "learning_rate": 0.0001262301543055746, | |
| "loss": 1.9865, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 0.11452103147367389, | |
| "grad_norm": 14.854180335998535, | |
| "learning_rate": 0.0001260102464860195, | |
| "loss": 2.1993, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 0.11471712913030689, | |
| "grad_norm": 5.227717399597168, | |
| "learning_rate": 0.0001257902036721452, | |
| "loss": 1.3025, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.1149132267869399, | |
| "grad_norm": 9.518735885620117, | |
| "learning_rate": 0.00012557002700598353, | |
| "loss": 1.9782, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 0.1151093244435729, | |
| "grad_norm": 15.842093467712402, | |
| "learning_rate": 0.00012534971763026104, | |
| "loss": 2.2285, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 0.1153054221002059, | |
| "grad_norm": 7.949627876281738, | |
| "learning_rate": 0.00012512927668839304, | |
| "loss": 1.0101, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 0.11550151975683891, | |
| "grad_norm": 10.285954475402832, | |
| "learning_rate": 0.00012490870532447774, | |
| "loss": 2.2186, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 0.11569761741347191, | |
| "grad_norm": 13.825008392333984, | |
| "learning_rate": 0.00012468800468329013, | |
| "loss": 1.9117, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.11589371507010492, | |
| "grad_norm": 4.062519073486328, | |
| "learning_rate": 0.00012446717591027624, | |
| "loss": 1.3969, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 0.11608981272673792, | |
| "grad_norm": 13.4169282913208, | |
| "learning_rate": 0.00012424622015154703, | |
| "loss": 1.6026, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 0.11628591038337092, | |
| "grad_norm": 5.054627895355225, | |
| "learning_rate": 0.0001240251385538726, | |
| "loss": 3.3848, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 0.11648200804000393, | |
| "grad_norm": 5.519535541534424, | |
| "learning_rate": 0.00012380393226467615, | |
| "loss": 2.0605, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 0.11667810569663692, | |
| "grad_norm": 8.747673034667969, | |
| "learning_rate": 0.000123582602432028, | |
| "loss": 1.945, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.11687420335326992, | |
| "grad_norm": 7.695986747741699, | |
| "learning_rate": 0.0001233611502046397, | |
| "loss": 1.2818, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 0.11707030100990293, | |
| "grad_norm": 6.775882720947266, | |
| "learning_rate": 0.000123139576731858, | |
| "loss": 1.5774, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 0.11726639866653593, | |
| "grad_norm": 6.141412258148193, | |
| "learning_rate": 0.00012291788316365888, | |
| "loss": 2.0431, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 0.11746249632316894, | |
| "grad_norm": 6.908384323120117, | |
| "learning_rate": 0.00012269607065064177, | |
| "loss": 2.298, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 0.11765859397980194, | |
| "grad_norm": 6.526587963104248, | |
| "learning_rate": 0.0001224741403440233, | |
| "loss": 0.7667, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.11785469163643494, | |
| "grad_norm": 6.505580902099609, | |
| "learning_rate": 0.00012225209339563145, | |
| "loss": 2.2598, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 0.11805078929306795, | |
| "grad_norm": 5.94541072845459, | |
| "learning_rate": 0.00012202993095789966, | |
| "loss": 2.5347, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 0.11824688694970095, | |
| "grad_norm": 4.825118064880371, | |
| "learning_rate": 0.00012180765418386068, | |
| "loss": 1.3592, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 0.11844298460633396, | |
| "grad_norm": 10.260912895202637, | |
| "learning_rate": 0.00012158526422714076, | |
| "loss": 1.9816, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 0.11863908226296696, | |
| "grad_norm": 4.47437047958374, | |
| "learning_rate": 0.00012136276224195348, | |
| "loss": 1.3715, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.11883517991959996, | |
| "grad_norm": 5.319314002990723, | |
| "learning_rate": 0.00012114014938309393, | |
| "loss": 1.7164, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 0.11903127757623297, | |
| "grad_norm": 9.467094421386719, | |
| "learning_rate": 0.00012091742680593254, | |
| "loss": 1.711, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 0.11922737523286597, | |
| "grad_norm": 15.36571979522705, | |
| "learning_rate": 0.0001206945956664093, | |
| "loss": 2.1494, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 0.11942347288949898, | |
| "grad_norm": 5.132857322692871, | |
| "learning_rate": 0.00012047165712102759, | |
| "loss": 1.0723, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 0.11961957054613197, | |
| "grad_norm": 7.701923847198486, | |
| "learning_rate": 0.00012024861232684823, | |
| "loss": 0.6081, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.11981566820276497, | |
| "grad_norm": 8.608162879943848, | |
| "learning_rate": 0.00012002546244148345, | |
| "loss": 2.7694, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 0.12001176585939798, | |
| "grad_norm": 6.971923351287842, | |
| "learning_rate": 0.00011980220862309097, | |
| "loss": 2.4804, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 0.12020786351603098, | |
| "grad_norm": 6.421287536621094, | |
| "learning_rate": 0.00011957885203036785, | |
| "loss": 1.1601, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 0.12040396117266398, | |
| "grad_norm": 5.901442527770996, | |
| "learning_rate": 0.00011935539382254459, | |
| "loss": 1.1592, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 0.12060005882929699, | |
| "grad_norm": 8.312341690063477, | |
| "learning_rate": 0.00011913183515937916, | |
| "loss": 1.4096, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.12079615648592999, | |
| "grad_norm": 9.224618911743164, | |
| "learning_rate": 0.00011890817720115075, | |
| "loss": 1.8728, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 0.120992254142563, | |
| "grad_norm": 6.559041500091553, | |
| "learning_rate": 0.00011868442110865399, | |
| "loss": 1.4327, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 0.121188351799196, | |
| "grad_norm": 6.139137268066406, | |
| "learning_rate": 0.0001184605680431928, | |
| "loss": 2.3143, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 0.121384449455829, | |
| "grad_norm": 7.806881427764893, | |
| "learning_rate": 0.0001182366191665744, | |
| "loss": 2.0893, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 0.12158054711246201, | |
| "grad_norm": 5.146418571472168, | |
| "learning_rate": 0.00011801257564110329, | |
| "loss": 0.9106, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.12177664476909501, | |
| "grad_norm": 10.43996810913086, | |
| "learning_rate": 0.00011778843862957514, | |
| "loss": 1.5653, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 0.12197274242572802, | |
| "grad_norm": 9.192870140075684, | |
| "learning_rate": 0.0001175642092952709, | |
| "loss": 1.097, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 0.12216884008236102, | |
| "grad_norm": 12.34306526184082, | |
| "learning_rate": 0.00011733988880195068, | |
| "loss": 0.8703, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 0.12236493773899403, | |
| "grad_norm": 6.297984600067139, | |
| "learning_rate": 0.00011711547831384761, | |
| "loss": 0.8209, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 0.12256103539562702, | |
| "grad_norm": 6.844716548919678, | |
| "learning_rate": 0.00011689097899566198, | |
| "loss": 1.2469, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.12275713305226002, | |
| "grad_norm": 4.610165119171143, | |
| "learning_rate": 0.00011666639201255506, | |
| "loss": 1.6045, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 0.12295323070889302, | |
| "grad_norm": 8.704841613769531, | |
| "learning_rate": 0.00011644171853014319, | |
| "loss": 1.5989, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 0.12314932836552603, | |
| "grad_norm": 14.035907745361328, | |
| "learning_rate": 0.00011621695971449154, | |
| "loss": 2.3428, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 0.12334542602215903, | |
| "grad_norm": 8.880375862121582, | |
| "learning_rate": 0.00011599211673210826, | |
| "loss": 1.6628, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 0.12354152367879204, | |
| "grad_norm": 5.904176235198975, | |
| "learning_rate": 0.00011576719074993827, | |
| "loss": 1.8985, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.12373762133542504, | |
| "grad_norm": 5.038250923156738, | |
| "learning_rate": 0.00011554218293535725, | |
| "loss": 1.8648, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 0.12393371899205805, | |
| "grad_norm": 4.206717014312744, | |
| "learning_rate": 0.00011531709445616563, | |
| "loss": 1.7799, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 0.12412981664869105, | |
| "grad_norm": 9.580304145812988, | |
| "learning_rate": 0.00011509192648058249, | |
| "loss": 1.484, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 0.12432591430532405, | |
| "grad_norm": 5.086032867431641, | |
| "learning_rate": 0.00011486668017723949, | |
| "loss": 1.2848, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 0.12452201196195706, | |
| "grad_norm": 3.946556568145752, | |
| "learning_rate": 0.00011464135671517482, | |
| "loss": 0.6885, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.12471810961859006, | |
| "grad_norm": 5.43235969543457, | |
| "learning_rate": 0.0001144159572638271, | |
| "loss": 1.1324, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 0.12491420727522307, | |
| "grad_norm": 14.442760467529297, | |
| "learning_rate": 0.00011419048299302939, | |
| "loss": 2.5022, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 0.12511030493185607, | |
| "grad_norm": 5.514584541320801, | |
| "learning_rate": 0.00011396493507300303, | |
| "loss": 0.714, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 0.12530640258848907, | |
| "grad_norm": 9.007026672363281, | |
| "learning_rate": 0.00011373931467435159, | |
| "loss": 1.894, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 0.12550250024512208, | |
| "grad_norm": 4.5296196937561035, | |
| "learning_rate": 0.00011351362296805485, | |
| "loss": 1.8764, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.12569859790175508, | |
| "grad_norm": 6.608786106109619, | |
| "learning_rate": 0.00011328786112546269, | |
| "loss": 1.8515, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 0.1258946955583881, | |
| "grad_norm": 5.0649261474609375, | |
| "learning_rate": 0.00011306203031828886, | |
| "loss": 1.1955, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 0.1260907932150211, | |
| "grad_norm": 7.887217998504639, | |
| "learning_rate": 0.00011283613171860525, | |
| "loss": 1.9566, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 0.1262868908716541, | |
| "grad_norm": 3.997429609298706, | |
| "learning_rate": 0.00011261016649883545, | |
| "loss": 0.4142, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 0.1264829885282871, | |
| "grad_norm": 6.192634582519531, | |
| "learning_rate": 0.0001123841358317489, | |
| "loss": 0.7429, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.1266790861849201, | |
| "grad_norm": 10.252041816711426, | |
| "learning_rate": 0.00011215804089045459, | |
| "loss": 1.5821, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 0.12687518384155308, | |
| "grad_norm": 15.608290672302246, | |
| "learning_rate": 0.00011193188284839517, | |
| "loss": 2.9049, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 0.12707128149818608, | |
| "grad_norm": 7.4965314865112305, | |
| "learning_rate": 0.00011170566287934088, | |
| "loss": 1.7288, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 0.1272673791548191, | |
| "grad_norm": 13.268543243408203, | |
| "learning_rate": 0.00011147938215738323, | |
| "loss": 2.6662, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 0.1274634768114521, | |
| "grad_norm": 6.431972503662109, | |
| "learning_rate": 0.00011125304185692907, | |
| "loss": 2.416, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.1276595744680851, | |
| "grad_norm": 5.82267951965332, | |
| "learning_rate": 0.0001110266431526945, | |
| "loss": 1.0295, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 0.1278556721247181, | |
| "grad_norm": 8.612664222717285, | |
| "learning_rate": 0.00011080018721969871, | |
| "loss": 2.5151, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 0.1280517697813511, | |
| "grad_norm": 10.103599548339844, | |
| "learning_rate": 0.00011057367523325792, | |
| "loss": 2.4646, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 0.1282478674379841, | |
| "grad_norm": 6.955473899841309, | |
| "learning_rate": 0.00011034710836897921, | |
| "loss": 1.3256, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 0.1284439650946171, | |
| "grad_norm": 6.604331016540527, | |
| "learning_rate": 0.00011012048780275463, | |
| "loss": 1.1124, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.12864006275125012, | |
| "grad_norm": 7.2269511222839355, | |
| "learning_rate": 0.00010989381471075482, | |
| "loss": 2.1239, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 0.12883616040788312, | |
| "grad_norm": 7.591269016265869, | |
| "learning_rate": 0.00010966709026942303, | |
| "loss": 1.9995, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 0.12903225806451613, | |
| "grad_norm": 5.342310428619385, | |
| "learning_rate": 0.00010944031565546906, | |
| "loss": 0.8718, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 0.12922835572114913, | |
| "grad_norm": 9.045146942138672, | |
| "learning_rate": 0.0001092134920458631, | |
| "loss": 0.9157, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 0.12942445337778213, | |
| "grad_norm": 4.300910472869873, | |
| "learning_rate": 0.00010898662061782965, | |
| "loss": 1.0317, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.12962055103441514, | |
| "grad_norm": 11.28966999053955, | |
| "learning_rate": 0.0001087597025488413, | |
| "loss": 0.7102, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 0.12981664869104814, | |
| "grad_norm": 7.143383502960205, | |
| "learning_rate": 0.00010853273901661285, | |
| "loss": 1.4524, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 0.13001274634768115, | |
| "grad_norm": 8.20251750946045, | |
| "learning_rate": 0.00010830573119909493, | |
| "loss": 1.825, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 0.13020884400431415, | |
| "grad_norm": 8.748811721801758, | |
| "learning_rate": 0.00010807868027446808, | |
| "loss": 0.6214, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 0.13040494166094715, | |
| "grad_norm": 6.7635040283203125, | |
| "learning_rate": 0.00010785158742113655, | |
| "loss": 1.158, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.13060103931758016, | |
| "grad_norm": 5.804599761962891, | |
| "learning_rate": 0.00010762445381772217, | |
| "loss": 1.2651, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 0.13079713697421316, | |
| "grad_norm": 6.046361923217773, | |
| "learning_rate": 0.00010739728064305834, | |
| "loss": 1.5022, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 0.13099323463084617, | |
| "grad_norm": 10.987192153930664, | |
| "learning_rate": 0.00010717006907618377, | |
| "loss": 1.5359, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 0.13118933228747917, | |
| "grad_norm": 8.694954872131348, | |
| "learning_rate": 0.00010694282029633647, | |
| "loss": 1.682, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 0.13138542994411218, | |
| "grad_norm": 15.458917617797852, | |
| "learning_rate": 0.00010671553548294753, | |
| "loss": 1.3832, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.13158152760074518, | |
| "grad_norm": 3.7156457901000977, | |
| "learning_rate": 0.00010648821581563513, | |
| "loss": 2.4507, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 0.13177762525737818, | |
| "grad_norm": 9.7117338180542, | |
| "learning_rate": 0.00010626086247419826, | |
| "loss": 2.3141, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 0.1319737229140112, | |
| "grad_norm": 7.740649700164795, | |
| "learning_rate": 0.00010603347663861079, | |
| "loss": 2.1845, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 0.1321698205706442, | |
| "grad_norm": 7.079351425170898, | |
| "learning_rate": 0.00010580605948901514, | |
| "loss": 1.5912, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 0.1323659182272772, | |
| "grad_norm": 3.896484613418579, | |
| "learning_rate": 0.00010557861220571625, | |
| "loss": 0.8481, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.1325620158839102, | |
| "grad_norm": 12.182394981384277, | |
| "learning_rate": 0.00010535113596917556, | |
| "loss": 2.1278, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 0.13275811354054318, | |
| "grad_norm": 7.2080230712890625, | |
| "learning_rate": 0.00010512363196000465, | |
| "loss": 1.8256, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 0.13295421119717618, | |
| "grad_norm": 7.081282615661621, | |
| "learning_rate": 0.00010489610135895933, | |
| "loss": 1.5351, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 0.13315030885380919, | |
| "grad_norm": 4.274497985839844, | |
| "learning_rate": 0.00010466854534693335, | |
| "loss": 1.3264, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 0.1333464065104422, | |
| "grad_norm": 6.364016532897949, | |
| "learning_rate": 0.00010444096510495243, | |
| "loss": 1.751, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.1335425041670752, | |
| "grad_norm": 6.791689872741699, | |
| "learning_rate": 0.00010421336181416797, | |
| "loss": 0.7944, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 0.1337386018237082, | |
| "grad_norm": 9.867424964904785, | |
| "learning_rate": 0.00010398573665585105, | |
| "loss": 1.6282, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 0.1339346994803412, | |
| "grad_norm": 12.289506912231445, | |
| "learning_rate": 0.0001037580908113862, | |
| "loss": 2.1015, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 0.1341307971369742, | |
| "grad_norm": 5.21271276473999, | |
| "learning_rate": 0.00010353042546226537, | |
| "loss": 1.0861, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 0.1343268947936072, | |
| "grad_norm": 6.344359874725342, | |
| "learning_rate": 0.00010330274179008161, | |
| "loss": 1.295, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.13452299245024021, | |
| "grad_norm": 4.4853196144104, | |
| "learning_rate": 0.00010307504097652323, | |
| "loss": 1.5746, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 0.13471909010687322, | |
| "grad_norm": 4.868480205535889, | |
| "learning_rate": 0.0001028473242033674, | |
| "loss": 0.7534, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 0.13491518776350622, | |
| "grad_norm": 33.27044677734375, | |
| "learning_rate": 0.00010261959265247419, | |
| "loss": 3.1026, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 0.13511128542013923, | |
| "grad_norm": 5.2553253173828125, | |
| "learning_rate": 0.0001023918475057803, | |
| "loss": 1.7087, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 0.13530738307677223, | |
| "grad_norm": 9.47342586517334, | |
| "learning_rate": 0.00010216408994529303, | |
| "loss": 1.2846, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.13550348073340523, | |
| "grad_norm": 5.372720241546631, | |
| "learning_rate": 0.00010193632115308411, | |
| "loss": 2.6163, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 0.13569957839003824, | |
| "grad_norm": 9.89535903930664, | |
| "learning_rate": 0.00010170854231128352, | |
| "loss": 2.3616, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 0.13589567604667124, | |
| "grad_norm": 8.19672966003418, | |
| "learning_rate": 0.00010148075460207347, | |
| "loss": 1.8404, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 0.13609177370330425, | |
| "grad_norm": 5.280313491821289, | |
| "learning_rate": 0.0001012529592076821, | |
| "loss": 2.2078, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 0.13628787135993725, | |
| "grad_norm": 7.142581462860107, | |
| "learning_rate": 0.00010102515731037758, | |
| "loss": 0.9421, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.13648396901657026, | |
| "grad_norm": 8.142248153686523, | |
| "learning_rate": 0.00010079735009246167, | |
| "loss": 0.83, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 0.13648396901657026, | |
| "eval_loss": 0.40602046251296997, | |
| "eval_runtime": 78.837, | |
| "eval_samples_per_second": 27.246, | |
| "eval_steps_per_second": 13.623, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 0.13668006667320326, | |
| "grad_norm": 3.3614909648895264, | |
| "learning_rate": 0.00010056953873626384, | |
| "loss": 0.7098, | |
| "step": 697 | |
| }, | |
| { | |
| "epoch": 0.13687616432983626, | |
| "grad_norm": 6.330589294433594, | |
| "learning_rate": 0.00010034172442413501, | |
| "loss": 1.97, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 0.13707226198646927, | |
| "grad_norm": 6.881412506103516, | |
| "learning_rate": 0.00010011390833844143, | |
| "loss": 2.2338, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 0.13726835964310227, | |
| "grad_norm": 6.3669114112854, | |
| "learning_rate": 9.988609166155859e-05, | |
| "loss": 1.2176, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.13746445729973528, | |
| "grad_norm": 4.456204414367676, | |
| "learning_rate": 9.965827557586498e-05, | |
| "loss": 2.5429, | |
| "step": 701 | |
| }, | |
| { | |
| "epoch": 0.13766055495636828, | |
| "grad_norm": 11.206175804138184, | |
| "learning_rate": 9.943046126373618e-05, | |
| "loss": 0.9134, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 0.13785665261300128, | |
| "grad_norm": 3.9151828289031982, | |
| "learning_rate": 9.920264990753837e-05, | |
| "loss": 0.8662, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 0.1380527502696343, | |
| "grad_norm": 9.173954963684082, | |
| "learning_rate": 9.897484268962243e-05, | |
| "loss": 1.7328, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 0.1382488479262673, | |
| "grad_norm": 9.788105010986328, | |
| "learning_rate": 9.874704079231791e-05, | |
| "loss": 1.5085, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.1384449455829003, | |
| "grad_norm": 4.705010414123535, | |
| "learning_rate": 9.851924539792656e-05, | |
| "loss": 1.7789, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 0.1386410432395333, | |
| "grad_norm": 9.228858947753906, | |
| "learning_rate": 9.82914576887165e-05, | |
| "loss": 1.3564, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 0.13883714089616628, | |
| "grad_norm": 13.22546100616455, | |
| "learning_rate": 9.806367884691594e-05, | |
| "loss": 1.752, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 0.13903323855279928, | |
| "grad_norm": 8.163661003112793, | |
| "learning_rate": 9.783591005470698e-05, | |
| "loss": 1.3811, | |
| "step": 709 | |
| }, | |
| { | |
| "epoch": 0.1392293362094323, | |
| "grad_norm": 8.644879341125488, | |
| "learning_rate": 9.760815249421973e-05, | |
| "loss": 1.8551, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.1394254338660653, | |
| "grad_norm": 4.906625270843506, | |
| "learning_rate": 9.738040734752582e-05, | |
| "loss": 1.3423, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 0.1396215315226983, | |
| "grad_norm": 7.070640563964844, | |
| "learning_rate": 9.715267579663262e-05, | |
| "loss": 1.1465, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 0.1398176291793313, | |
| "grad_norm": 4.564503192901611, | |
| "learning_rate": 9.692495902347678e-05, | |
| "loss": 1.167, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 0.1400137268359643, | |
| "grad_norm": 6.235091686248779, | |
| "learning_rate": 9.669725820991841e-05, | |
| "loss": 1.0448, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 0.1402098244925973, | |
| "grad_norm": 6.636250019073486, | |
| "learning_rate": 9.646957453773469e-05, | |
| "loss": 1.1572, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.1404059221492303, | |
| "grad_norm": 4.873953342437744, | |
| "learning_rate": 9.62419091886138e-05, | |
| "loss": 1.6995, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 0.14060201980586332, | |
| "grad_norm": 7.3861589431762695, | |
| "learning_rate": 9.601426334414898e-05, | |
| "loss": 1.6366, | |
| "step": 717 | |
| }, | |
| { | |
| "epoch": 0.14079811746249632, | |
| "grad_norm": 4.689702033996582, | |
| "learning_rate": 9.578663818583203e-05, | |
| "loss": 0.6136, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 0.14099421511912932, | |
| "grad_norm": 6.917242527008057, | |
| "learning_rate": 9.555903489504761e-05, | |
| "loss": 1.985, | |
| "step": 719 | |
| }, | |
| { | |
| "epoch": 0.14119031277576233, | |
| "grad_norm": 11.42885971069336, | |
| "learning_rate": 9.533145465306667e-05, | |
| "loss": 1.2503, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.14138641043239533, | |
| "grad_norm": 6.561046600341797, | |
| "learning_rate": 9.510389864104069e-05, | |
| "loss": 1.207, | |
| "step": 721 | |
| }, | |
| { | |
| "epoch": 0.14158250808902834, | |
| "grad_norm": 5.668408393859863, | |
| "learning_rate": 9.487636803999538e-05, | |
| "loss": 1.5323, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 0.14177860574566134, | |
| "grad_norm": 5.94567346572876, | |
| "learning_rate": 9.464886403082445e-05, | |
| "loss": 1.6155, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 0.14197470340229434, | |
| "grad_norm": 5.67940092086792, | |
| "learning_rate": 9.442138779428376e-05, | |
| "loss": 1.1803, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 0.14217080105892735, | |
| "grad_norm": 5.852919101715088, | |
| "learning_rate": 9.419394051098489e-05, | |
| "loss": 2.2628, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.14236689871556035, | |
| "grad_norm": 5.615955352783203, | |
| "learning_rate": 9.396652336138922e-05, | |
| "loss": 2.6917, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 0.14256299637219336, | |
| "grad_norm": 14.781683921813965, | |
| "learning_rate": 9.373913752580175e-05, | |
| "loss": 1.8918, | |
| "step": 727 | |
| }, | |
| { | |
| "epoch": 0.14275909402882636, | |
| "grad_norm": 6.98385763168335, | |
| "learning_rate": 9.351178418436488e-05, | |
| "loss": 0.8915, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 0.14295519168545937, | |
| "grad_norm": 11.97396469116211, | |
| "learning_rate": 9.328446451705249e-05, | |
| "loss": 1.3453, | |
| "step": 729 | |
| }, | |
| { | |
| "epoch": 0.14315128934209237, | |
| "grad_norm": 7.939299583435059, | |
| "learning_rate": 9.305717970366358e-05, | |
| "loss": 1.7643, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.14334738699872537, | |
| "grad_norm": 6.75924015045166, | |
| "learning_rate": 9.282993092381625e-05, | |
| "loss": 0.9198, | |
| "step": 731 | |
| }, | |
| { | |
| "epoch": 0.14354348465535838, | |
| "grad_norm": 7.962716579437256, | |
| "learning_rate": 9.260271935694168e-05, | |
| "loss": 1.5201, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 0.14373958231199138, | |
| "grad_norm": 6.077849864959717, | |
| "learning_rate": 9.237554618227785e-05, | |
| "loss": 1.1027, | |
| "step": 733 | |
| }, | |
| { | |
| "epoch": 0.14393567996862439, | |
| "grad_norm": 5.715548992156982, | |
| "learning_rate": 9.214841257886349e-05, | |
| "loss": 0.9475, | |
| "step": 734 | |
| }, | |
| { | |
| "epoch": 0.1441317776252574, | |
| "grad_norm": 8.250515937805176, | |
| "learning_rate": 9.192131972553191e-05, | |
| "loss": 1.5568, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.1443278752818904, | |
| "grad_norm": 4.075748443603516, | |
| "learning_rate": 9.16942688009051e-05, | |
| "loss": 1.1728, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 0.1445239729385234, | |
| "grad_norm": 6.287231922149658, | |
| "learning_rate": 9.146726098338719e-05, | |
| "loss": 0.9493, | |
| "step": 737 | |
| }, | |
| { | |
| "epoch": 0.14472007059515637, | |
| "grad_norm": 7.561791896820068, | |
| "learning_rate": 9.12402974511587e-05, | |
| "loss": 1.2801, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 0.14491616825178938, | |
| "grad_norm": 8.35126781463623, | |
| "learning_rate": 9.101337938217038e-05, | |
| "loss": 1.4908, | |
| "step": 739 | |
| }, | |
| { | |
| "epoch": 0.14511226590842238, | |
| "grad_norm": 7.66392707824707, | |
| "learning_rate": 9.078650795413692e-05, | |
| "loss": 2.3328, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.1453083635650554, | |
| "grad_norm": 8.209086418151855, | |
| "learning_rate": 9.055968434453097e-05, | |
| "loss": 1.1856, | |
| "step": 741 | |
| }, | |
| { | |
| "epoch": 0.1455044612216884, | |
| "grad_norm": 6.804962635040283, | |
| "learning_rate": 9.0332909730577e-05, | |
| "loss": 1.8494, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 0.1457005588783214, | |
| "grad_norm": 8.357268333435059, | |
| "learning_rate": 9.01061852892452e-05, | |
| "loss": 2.0554, | |
| "step": 743 | |
| }, | |
| { | |
| "epoch": 0.1458966565349544, | |
| "grad_norm": 9.071808815002441, | |
| "learning_rate": 8.98795121972454e-05, | |
| "loss": 1.455, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 0.1460927541915874, | |
| "grad_norm": 6.041928768157959, | |
| "learning_rate": 8.965289163102078e-05, | |
| "loss": 1.386, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.1462888518482204, | |
| "grad_norm": 9.787776947021484, | |
| "learning_rate": 8.942632476674211e-05, | |
| "loss": 1.4516, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 0.1464849495048534, | |
| "grad_norm": 11.039609909057617, | |
| "learning_rate": 8.919981278030133e-05, | |
| "loss": 2.5638, | |
| "step": 747 | |
| }, | |
| { | |
| "epoch": 0.14668104716148642, | |
| "grad_norm": 7.651856899261475, | |
| "learning_rate": 8.89733568473055e-05, | |
| "loss": 2.2337, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 0.14687714481811942, | |
| "grad_norm": 5.052631855010986, | |
| "learning_rate": 8.874695814307094e-05, | |
| "loss": 2.1706, | |
| "step": 749 | |
| }, | |
| { | |
| "epoch": 0.14707324247475242, | |
| "grad_norm": 4.033751964569092, | |
| "learning_rate": 8.852061784261678e-05, | |
| "loss": 0.7492, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.14726934013138543, | |
| "grad_norm": 7.171450138092041, | |
| "learning_rate": 8.829433712065914e-05, | |
| "loss": 1.1281, | |
| "step": 751 | |
| }, | |
| { | |
| "epoch": 0.14746543778801843, | |
| "grad_norm": 4.533056735992432, | |
| "learning_rate": 8.806811715160485e-05, | |
| "loss": 1.1391, | |
| "step": 752 | |
| }, | |
| { | |
| "epoch": 0.14766153544465144, | |
| "grad_norm": 6.083869934082031, | |
| "learning_rate": 8.784195910954545e-05, | |
| "loss": 1.0986, | |
| "step": 753 | |
| }, | |
| { | |
| "epoch": 0.14785763310128444, | |
| "grad_norm": 8.265074729919434, | |
| "learning_rate": 8.761586416825117e-05, | |
| "loss": 2.8251, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 0.14805373075791745, | |
| "grad_norm": 4.621420860290527, | |
| "learning_rate": 8.738983350116454e-05, | |
| "loss": 1.3493, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.14824982841455045, | |
| "grad_norm": 9.803274154663086, | |
| "learning_rate": 8.716386828139477e-05, | |
| "loss": 1.6575, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 0.14844592607118345, | |
| "grad_norm": 8.033260345458984, | |
| "learning_rate": 8.693796968171113e-05, | |
| "loss": 1.4908, | |
| "step": 757 | |
| }, | |
| { | |
| "epoch": 0.14864202372781646, | |
| "grad_norm": 10.312782287597656, | |
| "learning_rate": 8.671213887453735e-05, | |
| "loss": 1.0033, | |
| "step": 758 | |
| }, | |
| { | |
| "epoch": 0.14883812138444946, | |
| "grad_norm": 4.6735615730285645, | |
| "learning_rate": 8.648637703194516e-05, | |
| "loss": 2.4385, | |
| "step": 759 | |
| }, | |
| { | |
| "epoch": 0.14903421904108247, | |
| "grad_norm": 7.040375709533691, | |
| "learning_rate": 8.62606853256484e-05, | |
| "loss": 1.4345, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.14923031669771547, | |
| "grad_norm": 5.844077110290527, | |
| "learning_rate": 8.603506492699699e-05, | |
| "loss": 1.9578, | |
| "step": 761 | |
| }, | |
| { | |
| "epoch": 0.14942641435434847, | |
| "grad_norm": 7.8477606773376465, | |
| "learning_rate": 8.58095170069706e-05, | |
| "loss": 1.915, | |
| "step": 762 | |
| }, | |
| { | |
| "epoch": 0.14962251201098148, | |
| "grad_norm": 6.696018218994141, | |
| "learning_rate": 8.55840427361729e-05, | |
| "loss": 2.6032, | |
| "step": 763 | |
| }, | |
| { | |
| "epoch": 0.14981860966761448, | |
| "grad_norm": 6.304353713989258, | |
| "learning_rate": 8.535864328482523e-05, | |
| "loss": 1.0302, | |
| "step": 764 | |
| }, | |
| { | |
| "epoch": 0.1500147073242475, | |
| "grad_norm": 4.32615852355957, | |
| "learning_rate": 8.513331982276053e-05, | |
| "loss": 1.253, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.1502108049808805, | |
| "grad_norm": 5.056833744049072, | |
| "learning_rate": 8.490807351941753e-05, | |
| "loss": 0.9107, | |
| "step": 766 | |
| }, | |
| { | |
| "epoch": 0.1504069026375135, | |
| "grad_norm": 11.254875183105469, | |
| "learning_rate": 8.468290554383436e-05, | |
| "loss": 2.1077, | |
| "step": 767 | |
| }, | |
| { | |
| "epoch": 0.15060300029414647, | |
| "grad_norm": 5.245373249053955, | |
| "learning_rate": 8.445781706464277e-05, | |
| "loss": 0.7334, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 0.15079909795077948, | |
| "grad_norm": 6.346343994140625, | |
| "learning_rate": 8.423280925006178e-05, | |
| "loss": 2.2026, | |
| "step": 769 | |
| }, | |
| { | |
| "epoch": 0.15099519560741248, | |
| "grad_norm": 9.745604515075684, | |
| "learning_rate": 8.400788326789175e-05, | |
| "loss": 0.8139, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.15119129326404548, | |
| "grad_norm": 4.820820331573486, | |
| "learning_rate": 8.378304028550848e-05, | |
| "loss": 1.2438, | |
| "step": 771 | |
| }, | |
| { | |
| "epoch": 0.1513873909206785, | |
| "grad_norm": 7.597200393676758, | |
| "learning_rate": 8.355828146985684e-05, | |
| "loss": 2.3756, | |
| "step": 772 | |
| }, | |
| { | |
| "epoch": 0.1515834885773115, | |
| "grad_norm": 4.522254943847656, | |
| "learning_rate": 8.333360798744496e-05, | |
| "loss": 1.9562, | |
| "step": 773 | |
| }, | |
| { | |
| "epoch": 0.1517795862339445, | |
| "grad_norm": 11.164361000061035, | |
| "learning_rate": 8.310902100433809e-05, | |
| "loss": 2.1888, | |
| "step": 774 | |
| }, | |
| { | |
| "epoch": 0.1519756838905775, | |
| "grad_norm": 8.457398414611816, | |
| "learning_rate": 8.288452168615242e-05, | |
| "loss": 0.7737, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.1521717815472105, | |
| "grad_norm": 10.31162166595459, | |
| "learning_rate": 8.266011119804936e-05, | |
| "loss": 1.1301, | |
| "step": 776 | |
| }, | |
| { | |
| "epoch": 0.1523678792038435, | |
| "grad_norm": 6.120646953582764, | |
| "learning_rate": 8.243579070472909e-05, | |
| "loss": 0.6059, | |
| "step": 777 | |
| }, | |
| { | |
| "epoch": 0.1525639768604765, | |
| "grad_norm": 7.931275844573975, | |
| "learning_rate": 8.221156137042489e-05, | |
| "loss": 2.2832, | |
| "step": 778 | |
| }, | |
| { | |
| "epoch": 0.15276007451710952, | |
| "grad_norm": 8.665715217590332, | |
| "learning_rate": 8.198742435889674e-05, | |
| "loss": 1.5639, | |
| "step": 779 | |
| }, | |
| { | |
| "epoch": 0.15295617217374252, | |
| "grad_norm": 6.210934638977051, | |
| "learning_rate": 8.176338083342561e-05, | |
| "loss": 0.5992, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.15315226983037553, | |
| "grad_norm": 4.350475311279297, | |
| "learning_rate": 8.153943195680723e-05, | |
| "loss": 1.9462, | |
| "step": 781 | |
| }, | |
| { | |
| "epoch": 0.15334836748700853, | |
| "grad_norm": 10.62283992767334, | |
| "learning_rate": 8.131557889134602e-05, | |
| "loss": 2.5713, | |
| "step": 782 | |
| }, | |
| { | |
| "epoch": 0.15354446514364153, | |
| "grad_norm": 19.049753189086914, | |
| "learning_rate": 8.109182279884928e-05, | |
| "loss": 2.0163, | |
| "step": 783 | |
| }, | |
| { | |
| "epoch": 0.15374056280027454, | |
| "grad_norm": 6.098108768463135, | |
| "learning_rate": 8.086816484062085e-05, | |
| "loss": 0.9044, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 0.15393666045690754, | |
| "grad_norm": 4.7711567878723145, | |
| "learning_rate": 8.064460617745542e-05, | |
| "loss": 1.8375, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 0.15413275811354055, | |
| "grad_norm": 6.954524993896484, | |
| "learning_rate": 8.042114796963219e-05, | |
| "loss": 1.4588, | |
| "step": 786 | |
| }, | |
| { | |
| "epoch": 0.15432885577017355, | |
| "grad_norm": 4.3450422286987305, | |
| "learning_rate": 8.019779137690906e-05, | |
| "loss": 1.9173, | |
| "step": 787 | |
| }, | |
| { | |
| "epoch": 0.15452495342680655, | |
| "grad_norm": 6.4524688720703125, | |
| "learning_rate": 7.997453755851658e-05, | |
| "loss": 0.9777, | |
| "step": 788 | |
| }, | |
| { | |
| "epoch": 0.15472105108343956, | |
| "grad_norm": 4.7262139320373535, | |
| "learning_rate": 7.975138767315178e-05, | |
| "loss": 2.0938, | |
| "step": 789 | |
| }, | |
| { | |
| "epoch": 0.15491714874007256, | |
| "grad_norm": 6.220945835113525, | |
| "learning_rate": 7.952834287897242e-05, | |
| "loss": 1.1835, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.15511324639670557, | |
| "grad_norm": 12.306231498718262, | |
| "learning_rate": 7.930540433359071e-05, | |
| "loss": 1.2498, | |
| "step": 791 | |
| }, | |
| { | |
| "epoch": 0.15530934405333857, | |
| "grad_norm": 5.665555953979492, | |
| "learning_rate": 7.908257319406747e-05, | |
| "loss": 1.2846, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 0.15550544170997158, | |
| "grad_norm": 5.636224269866943, | |
| "learning_rate": 7.88598506169061e-05, | |
| "loss": 1.1359, | |
| "step": 793 | |
| }, | |
| { | |
| "epoch": 0.15570153936660458, | |
| "grad_norm": 7.562717914581299, | |
| "learning_rate": 7.863723775804651e-05, | |
| "loss": 1.3149, | |
| "step": 794 | |
| }, | |
| { | |
| "epoch": 0.15589763702323758, | |
| "grad_norm": 6.797637462615967, | |
| "learning_rate": 7.841473577285925e-05, | |
| "loss": 1.2934, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 0.1560937346798706, | |
| "grad_norm": 9.185527801513672, | |
| "learning_rate": 7.819234581613934e-05, | |
| "loss": 2.0876, | |
| "step": 796 | |
| }, | |
| { | |
| "epoch": 0.1562898323365036, | |
| "grad_norm": 5.824263095855713, | |
| "learning_rate": 7.797006904210035e-05, | |
| "loss": 1.3984, | |
| "step": 797 | |
| }, | |
| { | |
| "epoch": 0.15648592999313657, | |
| "grad_norm": 8.152481079101562, | |
| "learning_rate": 7.774790660436858e-05, | |
| "loss": 1.9248, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 0.15668202764976957, | |
| "grad_norm": 5.90303897857666, | |
| "learning_rate": 7.752585965597673e-05, | |
| "loss": 1.3629, | |
| "step": 799 | |
| }, | |
| { | |
| "epoch": 0.15687812530640258, | |
| "grad_norm": 7.13554573059082, | |
| "learning_rate": 7.730392934935825e-05, | |
| "loss": 1.6686, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.15707422296303558, | |
| "grad_norm": 14.757647514343262, | |
| "learning_rate": 7.708211683634112e-05, | |
| "loss": 3.0637, | |
| "step": 801 | |
| }, | |
| { | |
| "epoch": 0.15727032061966859, | |
| "grad_norm": 6.754796504974365, | |
| "learning_rate": 7.686042326814205e-05, | |
| "loss": 1.6339, | |
| "step": 802 | |
| }, | |
| { | |
| "epoch": 0.1574664182763016, | |
| "grad_norm": 4.844980716705322, | |
| "learning_rate": 7.663884979536035e-05, | |
| "loss": 1.799, | |
| "step": 803 | |
| }, | |
| { | |
| "epoch": 0.1576625159329346, | |
| "grad_norm": 7.784121990203857, | |
| "learning_rate": 7.641739756797202e-05, | |
| "loss": 1.4723, | |
| "step": 804 | |
| }, | |
| { | |
| "epoch": 0.1578586135895676, | |
| "grad_norm": 9.167228698730469, | |
| "learning_rate": 7.619606773532386e-05, | |
| "loss": 1.3234, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 0.1580547112462006, | |
| "grad_norm": 4.63117790222168, | |
| "learning_rate": 7.59748614461274e-05, | |
| "loss": 1.2373, | |
| "step": 806 | |
| }, | |
| { | |
| "epoch": 0.1582508089028336, | |
| "grad_norm": 6.5301690101623535, | |
| "learning_rate": 7.5753779848453e-05, | |
| "loss": 2.5172, | |
| "step": 807 | |
| }, | |
| { | |
| "epoch": 0.1584469065594666, | |
| "grad_norm": 6.791456699371338, | |
| "learning_rate": 7.553282408972382e-05, | |
| "loss": 1.563, | |
| "step": 808 | |
| }, | |
| { | |
| "epoch": 0.15864300421609961, | |
| "grad_norm": 5.697129249572754, | |
| "learning_rate": 7.531199531670988e-05, | |
| "loss": 1.6731, | |
| "step": 809 | |
| }, | |
| { | |
| "epoch": 0.15883910187273262, | |
| "grad_norm": 6.511412143707275, | |
| "learning_rate": 7.50912946755223e-05, | |
| "loss": 1.2208, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.15903519952936562, | |
| "grad_norm": 4.849149703979492, | |
| "learning_rate": 7.487072331160696e-05, | |
| "loss": 1.9081, | |
| "step": 811 | |
| }, | |
| { | |
| "epoch": 0.15923129718599863, | |
| "grad_norm": 2.844851016998291, | |
| "learning_rate": 7.465028236973897e-05, | |
| "loss": 1.2562, | |
| "step": 812 | |
| }, | |
| { | |
| "epoch": 0.15942739484263163, | |
| "grad_norm": 4.782979965209961, | |
| "learning_rate": 7.442997299401652e-05, | |
| "loss": 0.8038, | |
| "step": 813 | |
| }, | |
| { | |
| "epoch": 0.15962349249926464, | |
| "grad_norm": 13.652055740356445, | |
| "learning_rate": 7.420979632785483e-05, | |
| "loss": 1.3842, | |
| "step": 814 | |
| }, | |
| { | |
| "epoch": 0.15981959015589764, | |
| "grad_norm": 9.692163467407227, | |
| "learning_rate": 7.398975351398053e-05, | |
| "loss": 1.3089, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 0.16001568781253064, | |
| "grad_norm": 6.529005527496338, | |
| "learning_rate": 7.37698456944254e-05, | |
| "loss": 1.8166, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 0.16021178546916365, | |
| "grad_norm": 6.1307454109191895, | |
| "learning_rate": 7.355007401052072e-05, | |
| "loss": 1.6805, | |
| "step": 817 | |
| }, | |
| { | |
| "epoch": 0.16040788312579665, | |
| "grad_norm": 23.4924259185791, | |
| "learning_rate": 7.333043960289113e-05, | |
| "loss": 1.8122, | |
| "step": 818 | |
| }, | |
| { | |
| "epoch": 0.16060398078242966, | |
| "grad_norm": 5.680008411407471, | |
| "learning_rate": 7.311094361144881e-05, | |
| "loss": 1.7381, | |
| "step": 819 | |
| }, | |
| { | |
| "epoch": 0.16080007843906266, | |
| "grad_norm": 6.790538787841797, | |
| "learning_rate": 7.289158717538765e-05, | |
| "loss": 1.9142, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.16099617609569566, | |
| "grad_norm": 7.620028495788574, | |
| "learning_rate": 7.267237143317706e-05, | |
| "loss": 1.8411, | |
| "step": 821 | |
| }, | |
| { | |
| "epoch": 0.16119227375232867, | |
| "grad_norm": 5.989513397216797, | |
| "learning_rate": 7.245329752255647e-05, | |
| "loss": 1.7845, | |
| "step": 822 | |
| }, | |
| { | |
| "epoch": 0.16138837140896167, | |
| "grad_norm": 5.91652774810791, | |
| "learning_rate": 7.223436658052898e-05, | |
| "loss": 1.7267, | |
| "step": 823 | |
| }, | |
| { | |
| "epoch": 0.16158446906559468, | |
| "grad_norm": 6.713134765625, | |
| "learning_rate": 7.201557974335583e-05, | |
| "loss": 1.6799, | |
| "step": 824 | |
| }, | |
| { | |
| "epoch": 0.16178056672222768, | |
| "grad_norm": 5.243997097015381, | |
| "learning_rate": 7.179693814655026e-05, | |
| "loss": 1.7857, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.16197666437886069, | |
| "grad_norm": 4.856655120849609, | |
| "learning_rate": 7.157844292487173e-05, | |
| "loss": 1.0967, | |
| "step": 826 | |
| }, | |
| { | |
| "epoch": 0.1621727620354937, | |
| "grad_norm": 4.94920015335083, | |
| "learning_rate": 7.13600952123201e-05, | |
| "loss": 1.4571, | |
| "step": 827 | |
| }, | |
| { | |
| "epoch": 0.16236885969212667, | |
| "grad_norm": 5.146086692810059, | |
| "learning_rate": 7.114189614212944e-05, | |
| "loss": 0.6922, | |
| "step": 828 | |
| }, | |
| { | |
| "epoch": 0.16256495734875967, | |
| "grad_norm": 8.812291145324707, | |
| "learning_rate": 7.092384684676262e-05, | |
| "loss": 1.2531, | |
| "step": 829 | |
| }, | |
| { | |
| "epoch": 0.16276105500539267, | |
| "grad_norm": 6.924501419067383, | |
| "learning_rate": 7.070594845790497e-05, | |
| "loss": 1.3011, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.16295715266202568, | |
| "grad_norm": 10.452120780944824, | |
| "learning_rate": 7.048820210645862e-05, | |
| "loss": 1.1114, | |
| "step": 831 | |
| }, | |
| { | |
| "epoch": 0.16315325031865868, | |
| "grad_norm": 8.212396621704102, | |
| "learning_rate": 7.027060892253679e-05, | |
| "loss": 0.9207, | |
| "step": 832 | |
| }, | |
| { | |
| "epoch": 0.1633493479752917, | |
| "grad_norm": 17.70586585998535, | |
| "learning_rate": 7.005317003545754e-05, | |
| "loss": 1.3019, | |
| "step": 833 | |
| }, | |
| { | |
| "epoch": 0.1635454456319247, | |
| "grad_norm": 12.174421310424805, | |
| "learning_rate": 6.983588657373833e-05, | |
| "loss": 2.1275, | |
| "step": 834 | |
| }, | |
| { | |
| "epoch": 0.1637415432885577, | |
| "grad_norm": 3.9426016807556152, | |
| "learning_rate": 6.96187596650898e-05, | |
| "loss": 2.4098, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 0.1639376409451907, | |
| "grad_norm": 6.121764183044434, | |
| "learning_rate": 6.940179043641005e-05, | |
| "loss": 1.0717, | |
| "step": 836 | |
| }, | |
| { | |
| "epoch": 0.1641337386018237, | |
| "grad_norm": 4.13683557510376, | |
| "learning_rate": 6.918498001377901e-05, | |
| "loss": 1.4934, | |
| "step": 837 | |
| }, | |
| { | |
| "epoch": 0.1643298362584567, | |
| "grad_norm": 4.907169818878174, | |
| "learning_rate": 6.896832952245217e-05, | |
| "loss": 1.9002, | |
| "step": 838 | |
| }, | |
| { | |
| "epoch": 0.1645259339150897, | |
| "grad_norm": 8.611984252929688, | |
| "learning_rate": 6.875184008685514e-05, | |
| "loss": 1.1805, | |
| "step": 839 | |
| }, | |
| { | |
| "epoch": 0.16472203157172272, | |
| "grad_norm": 8.626840591430664, | |
| "learning_rate": 6.853551283057757e-05, | |
| "loss": 1.1437, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.16491812922835572, | |
| "grad_norm": 6.1718878746032715, | |
| "learning_rate": 6.831934887636738e-05, | |
| "loss": 1.5465, | |
| "step": 841 | |
| }, | |
| { | |
| "epoch": 0.16511422688498872, | |
| "grad_norm": 6.6393961906433105, | |
| "learning_rate": 6.8103349346125e-05, | |
| "loss": 1.0682, | |
| "step": 842 | |
| }, | |
| { | |
| "epoch": 0.16531032454162173, | |
| "grad_norm": 5.233375072479248, | |
| "learning_rate": 6.788751536089739e-05, | |
| "loss": 1.8218, | |
| "step": 843 | |
| }, | |
| { | |
| "epoch": 0.16550642219825473, | |
| "grad_norm": 9.505565643310547, | |
| "learning_rate": 6.767184804087248e-05, | |
| "loss": 1.3462, | |
| "step": 844 | |
| }, | |
| { | |
| "epoch": 0.16570251985488774, | |
| "grad_norm": 5.554503440856934, | |
| "learning_rate": 6.745634850537302e-05, | |
| "loss": 0.7193, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 0.16589861751152074, | |
| "grad_norm": 6.603649139404297, | |
| "learning_rate": 6.724101787285114e-05, | |
| "loss": 1.3164, | |
| "step": 846 | |
| }, | |
| { | |
| "epoch": 0.16609471516815374, | |
| "grad_norm": 10.690621376037598, | |
| "learning_rate": 6.702585726088222e-05, | |
| "loss": 3.3145, | |
| "step": 847 | |
| }, | |
| { | |
| "epoch": 0.16629081282478675, | |
| "grad_norm": 17.521020889282227, | |
| "learning_rate": 6.681086778615922e-05, | |
| "loss": 2.3221, | |
| "step": 848 | |
| }, | |
| { | |
| "epoch": 0.16648691048141975, | |
| "grad_norm": 11.0389986038208, | |
| "learning_rate": 6.659605056448702e-05, | |
| "loss": 1.2903, | |
| "step": 849 | |
| }, | |
| { | |
| "epoch": 0.16668300813805276, | |
| "grad_norm": 5.6815900802612305, | |
| "learning_rate": 6.638140671077633e-05, | |
| "loss": 1.1651, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.16687910579468576, | |
| "grad_norm": 14.61434555053711, | |
| "learning_rate": 6.616693733903823e-05, | |
| "loss": 2.5921, | |
| "step": 851 | |
| }, | |
| { | |
| "epoch": 0.16707520345131877, | |
| "grad_norm": 6.983891487121582, | |
| "learning_rate": 6.595264356237812e-05, | |
| "loss": 1.2114, | |
| "step": 852 | |
| }, | |
| { | |
| "epoch": 0.16727130110795177, | |
| "grad_norm": 4.762997627258301, | |
| "learning_rate": 6.573852649299009e-05, | |
| "loss": 0.8655, | |
| "step": 853 | |
| }, | |
| { | |
| "epoch": 0.16746739876458477, | |
| "grad_norm": 8.16657829284668, | |
| "learning_rate": 6.552458724215114e-05, | |
| "loss": 1.584, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 0.16766349642121778, | |
| "grad_norm": 3.753600597381592, | |
| "learning_rate": 6.531082692021532e-05, | |
| "loss": 0.7213, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 0.16785959407785078, | |
| "grad_norm": 20.525659561157227, | |
| "learning_rate": 6.509724663660813e-05, | |
| "loss": 3.2134, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 0.1680556917344838, | |
| "grad_norm": 4.803929805755615, | |
| "learning_rate": 6.488384749982053e-05, | |
| "loss": 1.0283, | |
| "step": 857 | |
| }, | |
| { | |
| "epoch": 0.16825178939111676, | |
| "grad_norm": 4.031068325042725, | |
| "learning_rate": 6.467063061740345e-05, | |
| "loss": 0.86, | |
| "step": 858 | |
| }, | |
| { | |
| "epoch": 0.16844788704774977, | |
| "grad_norm": 8.304176330566406, | |
| "learning_rate": 6.445759709596182e-05, | |
| "loss": 0.7242, | |
| "step": 859 | |
| }, | |
| { | |
| "epoch": 0.16864398470438277, | |
| "grad_norm": 5.160337448120117, | |
| "learning_rate": 6.424474804114895e-05, | |
| "loss": 1.7592, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.16884008236101578, | |
| "grad_norm": 6.138307571411133, | |
| "learning_rate": 6.403208455766081e-05, | |
| "loss": 2.6926, | |
| "step": 861 | |
| }, | |
| { | |
| "epoch": 0.16903618001764878, | |
| "grad_norm": 3.878741979598999, | |
| "learning_rate": 6.381960774923017e-05, | |
| "loss": 1.2294, | |
| "step": 862 | |
| }, | |
| { | |
| "epoch": 0.16923227767428178, | |
| "grad_norm": 4.720643043518066, | |
| "learning_rate": 6.360731871862093e-05, | |
| "loss": 1.7562, | |
| "step": 863 | |
| }, | |
| { | |
| "epoch": 0.1694283753309148, | |
| "grad_norm": 4.653330326080322, | |
| "learning_rate": 6.339521856762254e-05, | |
| "loss": 1.2522, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 0.1696244729875478, | |
| "grad_norm": 6.003101348876953, | |
| "learning_rate": 6.318330839704405e-05, | |
| "loss": 1.4006, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 0.1698205706441808, | |
| "grad_norm": 5.075412273406982, | |
| "learning_rate": 6.297158930670852e-05, | |
| "loss": 0.7885, | |
| "step": 866 | |
| }, | |
| { | |
| "epoch": 0.1700166683008138, | |
| "grad_norm": 4.062684059143066, | |
| "learning_rate": 6.276006239544729e-05, | |
| "loss": 1.0201, | |
| "step": 867 | |
| }, | |
| { | |
| "epoch": 0.1702127659574468, | |
| "grad_norm": 5.678539752960205, | |
| "learning_rate": 6.254872876109438e-05, | |
| "loss": 1.1822, | |
| "step": 868 | |
| }, | |
| { | |
| "epoch": 0.1704088636140798, | |
| "grad_norm": 4.655857563018799, | |
| "learning_rate": 6.233758950048056e-05, | |
| "loss": 1.4193, | |
| "step": 869 | |
| }, | |
| { | |
| "epoch": 0.1706049612707128, | |
| "grad_norm": 18.205698013305664, | |
| "learning_rate": 6.21266457094278e-05, | |
| "loss": 0.7067, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.17080105892734582, | |
| "grad_norm": 8.73768424987793, | |
| "learning_rate": 6.191589848274368e-05, | |
| "loss": 1.9048, | |
| "step": 871 | |
| }, | |
| { | |
| "epoch": 0.17099715658397882, | |
| "grad_norm": 5.723482131958008, | |
| "learning_rate": 6.170534891421556e-05, | |
| "loss": 0.8121, | |
| "step": 872 | |
| }, | |
| { | |
| "epoch": 0.17119325424061183, | |
| "grad_norm": 6.16123104095459, | |
| "learning_rate": 6.149499809660488e-05, | |
| "loss": 2.5358, | |
| "step": 873 | |
| }, | |
| { | |
| "epoch": 0.17138935189724483, | |
| "grad_norm": 6.115005016326904, | |
| "learning_rate": 6.128484712164164e-05, | |
| "loss": 1.6489, | |
| "step": 874 | |
| }, | |
| { | |
| "epoch": 0.17158544955387783, | |
| "grad_norm": 9.149282455444336, | |
| "learning_rate": 6.107489708001855e-05, | |
| "loss": 1.7977, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.17178154721051084, | |
| "grad_norm": 7.190993309020996, | |
| "learning_rate": 6.086514906138563e-05, | |
| "loss": 1.3526, | |
| "step": 876 | |
| }, | |
| { | |
| "epoch": 0.17197764486714384, | |
| "grad_norm": 8.292972564697266, | |
| "learning_rate": 6.065560415434417e-05, | |
| "loss": 1.8477, | |
| "step": 877 | |
| }, | |
| { | |
| "epoch": 0.17217374252377685, | |
| "grad_norm": 8.858931541442871, | |
| "learning_rate": 6.044626344644151e-05, | |
| "loss": 1.3605, | |
| "step": 878 | |
| }, | |
| { | |
| "epoch": 0.17236984018040985, | |
| "grad_norm": 6.186944961547852, | |
| "learning_rate": 6.023712802416508e-05, | |
| "loss": 1.0484, | |
| "step": 879 | |
| }, | |
| { | |
| "epoch": 0.17256593783704285, | |
| "grad_norm": 9.702549934387207, | |
| "learning_rate": 6.0028198972936836e-05, | |
| "loss": 1.3224, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.17276203549367586, | |
| "grad_norm": 12.195408821105957, | |
| "learning_rate": 5.981947737710779e-05, | |
| "loss": 1.8586, | |
| "step": 881 | |
| }, | |
| { | |
| "epoch": 0.17295813315030886, | |
| "grad_norm": 6.8144025802612305, | |
| "learning_rate": 5.9610964319952104e-05, | |
| "loss": 1.5967, | |
| "step": 882 | |
| }, | |
| { | |
| "epoch": 0.17315423080694187, | |
| "grad_norm": 10.37683391571045, | |
| "learning_rate": 5.940266088366173e-05, | |
| "loss": 1.7278, | |
| "step": 883 | |
| }, | |
| { | |
| "epoch": 0.17335032846357487, | |
| "grad_norm": 5.85037899017334, | |
| "learning_rate": 5.919456814934061e-05, | |
| "loss": 1.3165, | |
| "step": 884 | |
| }, | |
| { | |
| "epoch": 0.17354642612020788, | |
| "grad_norm": 7.210351943969727, | |
| "learning_rate": 5.8986687196999135e-05, | |
| "loss": 1.1156, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 0.17374252377684088, | |
| "grad_norm": 5.840338706970215, | |
| "learning_rate": 5.877901910554862e-05, | |
| "loss": 1.4444, | |
| "step": 886 | |
| }, | |
| { | |
| "epoch": 0.17393862143347388, | |
| "grad_norm": 3.731435775756836, | |
| "learning_rate": 5.8571564952795475e-05, | |
| "loss": 0.9172, | |
| "step": 887 | |
| }, | |
| { | |
| "epoch": 0.17413471909010686, | |
| "grad_norm": 13.559743881225586, | |
| "learning_rate": 5.8364325815435916e-05, | |
| "loss": 2.1312, | |
| "step": 888 | |
| }, | |
| { | |
| "epoch": 0.17433081674673986, | |
| "grad_norm": 5.5318193435668945, | |
| "learning_rate": 5.815730276905014e-05, | |
| "loss": 1.735, | |
| "step": 889 | |
| }, | |
| { | |
| "epoch": 0.17452691440337287, | |
| "grad_norm": 7.40202522277832, | |
| "learning_rate": 5.7950496888096795e-05, | |
| "loss": 1.42, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.17472301206000587, | |
| "grad_norm": 3.6923038959503174, | |
| "learning_rate": 5.7743909245907535e-05, | |
| "loss": 1.095, | |
| "step": 891 | |
| }, | |
| { | |
| "epoch": 0.17491910971663888, | |
| "grad_norm": 5.6912713050842285, | |
| "learning_rate": 5.753754091468115e-05, | |
| "loss": 2.2399, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 0.17511520737327188, | |
| "grad_norm": 3.8538315296173096, | |
| "learning_rate": 5.7331392965478493e-05, | |
| "loss": 1.2043, | |
| "step": 893 | |
| }, | |
| { | |
| "epoch": 0.17531130502990488, | |
| "grad_norm": 5.220914840698242, | |
| "learning_rate": 5.712546646821627e-05, | |
| "loss": 1.0669, | |
| "step": 894 | |
| }, | |
| { | |
| "epoch": 0.1755074026865379, | |
| "grad_norm": 5.886853218078613, | |
| "learning_rate": 5.6919762491662164e-05, | |
| "loss": 1.3425, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 0.1757035003431709, | |
| "grad_norm": 7.820699214935303, | |
| "learning_rate": 5.671428210342884e-05, | |
| "loss": 1.6338, | |
| "step": 896 | |
| }, | |
| { | |
| "epoch": 0.1758995979998039, | |
| "grad_norm": 9.36957836151123, | |
| "learning_rate": 5.650902636996837e-05, | |
| "loss": 1.1894, | |
| "step": 897 | |
| }, | |
| { | |
| "epoch": 0.1760956956564369, | |
| "grad_norm": 3.9427287578582764, | |
| "learning_rate": 5.6303996356567177e-05, | |
| "loss": 1.179, | |
| "step": 898 | |
| }, | |
| { | |
| "epoch": 0.1762917933130699, | |
| "grad_norm": 8.343499183654785, | |
| "learning_rate": 5.6099193127339864e-05, | |
| "loss": 1.9635, | |
| "step": 899 | |
| }, | |
| { | |
| "epoch": 0.1764878909697029, | |
| "grad_norm": 7.9742279052734375, | |
| "learning_rate": 5.589461774522433e-05, | |
| "loss": 1.6162, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.1766839886263359, | |
| "grad_norm": 4.933746337890625, | |
| "learning_rate": 5.5690271271975644e-05, | |
| "loss": 1.0064, | |
| "step": 901 | |
| }, | |
| { | |
| "epoch": 0.17688008628296892, | |
| "grad_norm": 6.986325740814209, | |
| "learning_rate": 5.548615476816097e-05, | |
| "loss": 0.7052, | |
| "step": 902 | |
| }, | |
| { | |
| "epoch": 0.17707618393960192, | |
| "grad_norm": 5.324771881103516, | |
| "learning_rate": 5.528226929315401e-05, | |
| "loss": 1.5658, | |
| "step": 903 | |
| }, | |
| { | |
| "epoch": 0.17727228159623493, | |
| "grad_norm": 3.9526147842407227, | |
| "learning_rate": 5.507861590512916e-05, | |
| "loss": 1.0183, | |
| "step": 904 | |
| }, | |
| { | |
| "epoch": 0.17746837925286793, | |
| "grad_norm": 8.242249488830566, | |
| "learning_rate": 5.48751956610566e-05, | |
| "loss": 1.9991, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 0.17766447690950093, | |
| "grad_norm": 6.519176483154297, | |
| "learning_rate": 5.467200961669619e-05, | |
| "loss": 1.2312, | |
| "step": 906 | |
| }, | |
| { | |
| "epoch": 0.17786057456613394, | |
| "grad_norm": 3.49365496635437, | |
| "learning_rate": 5.446905882659243e-05, | |
| "loss": 0.5029, | |
| "step": 907 | |
| }, | |
| { | |
| "epoch": 0.17805667222276694, | |
| "grad_norm": 5.808468818664551, | |
| "learning_rate": 5.426634434406883e-05, | |
| "loss": 1.1932, | |
| "step": 908 | |
| }, | |
| { | |
| "epoch": 0.17825276987939995, | |
| "grad_norm": 6.1657328605651855, | |
| "learning_rate": 5.40638672212224e-05, | |
| "loss": 1.0127, | |
| "step": 909 | |
| }, | |
| { | |
| "epoch": 0.17844886753603295, | |
| "grad_norm": 8.090532302856445, | |
| "learning_rate": 5.3861628508918384e-05, | |
| "loss": 1.0638, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.17864496519266596, | |
| "grad_norm": 6.026912689208984, | |
| "learning_rate": 5.3659629256784424e-05, | |
| "loss": 1.1598, | |
| "step": 911 | |
| }, | |
| { | |
| "epoch": 0.17884106284929896, | |
| "grad_norm": 7.103152751922607, | |
| "learning_rate": 5.345787051320564e-05, | |
| "loss": 1.9001, | |
| "step": 912 | |
| }, | |
| { | |
| "epoch": 0.17903716050593196, | |
| "grad_norm": 4.5661163330078125, | |
| "learning_rate": 5.325635332531864e-05, | |
| "loss": 1.2711, | |
| "step": 913 | |
| }, | |
| { | |
| "epoch": 0.17923325816256497, | |
| "grad_norm": 6.035799980163574, | |
| "learning_rate": 5.305507873900649e-05, | |
| "loss": 2.3649, | |
| "step": 914 | |
| }, | |
| { | |
| "epoch": 0.17942935581919797, | |
| "grad_norm": 6.231631278991699, | |
| "learning_rate": 5.2854047798893125e-05, | |
| "loss": 2.7058, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 0.17962545347583098, | |
| "grad_norm": 9.146753311157227, | |
| "learning_rate": 5.26532615483379e-05, | |
| "loss": 1.1388, | |
| "step": 916 | |
| }, | |
| { | |
| "epoch": 0.17982155113246398, | |
| "grad_norm": 8.76220417022705, | |
| "learning_rate": 5.245272102943034e-05, | |
| "loss": 1.0848, | |
| "step": 917 | |
| }, | |
| { | |
| "epoch": 0.18001764878909696, | |
| "grad_norm": 6.711489200592041, | |
| "learning_rate": 5.225242728298445e-05, | |
| "loss": 1.0747, | |
| "step": 918 | |
| }, | |
| { | |
| "epoch": 0.18021374644572996, | |
| "grad_norm": 4.66898775100708, | |
| "learning_rate": 5.2052381348533564e-05, | |
| "loss": 0.6995, | |
| "step": 919 | |
| }, | |
| { | |
| "epoch": 0.18040984410236297, | |
| "grad_norm": 6.504391193389893, | |
| "learning_rate": 5.1852584264324866e-05, | |
| "loss": 1.5352, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.18060594175899597, | |
| "grad_norm": 6.724686145782471, | |
| "learning_rate": 5.165303706731397e-05, | |
| "loss": 1.2562, | |
| "step": 921 | |
| }, | |
| { | |
| "epoch": 0.18080203941562897, | |
| "grad_norm": 7.192296981811523, | |
| "learning_rate": 5.1453740793159586e-05, | |
| "loss": 1.0158, | |
| "step": 922 | |
| }, | |
| { | |
| "epoch": 0.18099813707226198, | |
| "grad_norm": 5.173651218414307, | |
| "learning_rate": 5.12546964762181e-05, | |
| "loss": 1.3406, | |
| "step": 923 | |
| }, | |
| { | |
| "epoch": 0.18119423472889498, | |
| "grad_norm": 8.315585136413574, | |
| "learning_rate": 5.105590514953824e-05, | |
| "loss": 2.0057, | |
| "step": 924 | |
| }, | |
| { | |
| "epoch": 0.18139033238552799, | |
| "grad_norm": 4.940390110015869, | |
| "learning_rate": 5.085736784485571e-05, | |
| "loss": 1.6468, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.181586430042161, | |
| "grad_norm": 6.491610527038574, | |
| "learning_rate": 5.065908559258782e-05, | |
| "loss": 2.6567, | |
| "step": 926 | |
| }, | |
| { | |
| "epoch": 0.181782527698794, | |
| "grad_norm": 8.9893798828125, | |
| "learning_rate": 5.046105942182815e-05, | |
| "loss": 1.2412, | |
| "step": 927 | |
| }, | |
| { | |
| "epoch": 0.181978625355427, | |
| "grad_norm": 8.103008270263672, | |
| "learning_rate": 5.026329036034119e-05, | |
| "loss": 2.1319, | |
| "step": 928 | |
| }, | |
| { | |
| "epoch": 0.18217472301206, | |
| "grad_norm": 8.022109985351562, | |
| "learning_rate": 5.006577943455706e-05, | |
| "loss": 1.4119, | |
| "step": 929 | |
| }, | |
| { | |
| "epoch": 0.182370820668693, | |
| "grad_norm": 12.814486503601074, | |
| "learning_rate": 4.9868527669566113e-05, | |
| "loss": 1.9388, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.182566918325326, | |
| "grad_norm": 7.568852424621582, | |
| "learning_rate": 4.967153608911366e-05, | |
| "loss": 1.2114, | |
| "step": 931 | |
| }, | |
| { | |
| "epoch": 0.18276301598195901, | |
| "grad_norm": 3.7613790035247803, | |
| "learning_rate": 4.947480571559462e-05, | |
| "loss": 1.2221, | |
| "step": 932 | |
| }, | |
| { | |
| "epoch": 0.18295911363859202, | |
| "grad_norm": 8.301228523254395, | |
| "learning_rate": 4.927833757004826e-05, | |
| "loss": 1.2867, | |
| "step": 933 | |
| }, | |
| { | |
| "epoch": 0.18315521129522502, | |
| "grad_norm": 5.2592244148254395, | |
| "learning_rate": 4.908213267215287e-05, | |
| "loss": 2.0937, | |
| "step": 934 | |
| }, | |
| { | |
| "epoch": 0.18335130895185803, | |
| "grad_norm": 6.635796546936035, | |
| "learning_rate": 4.888619204022047e-05, | |
| "loss": 1.9856, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 0.18354740660849103, | |
| "grad_norm": 5.507410049438477, | |
| "learning_rate": 4.869051669119154e-05, | |
| "loss": 1.2305, | |
| "step": 936 | |
| }, | |
| { | |
| "epoch": 0.18374350426512404, | |
| "grad_norm": 8.553701400756836, | |
| "learning_rate": 4.8495107640629675e-05, | |
| "loss": 1.6872, | |
| "step": 937 | |
| }, | |
| { | |
| "epoch": 0.18393960192175704, | |
| "grad_norm": 7.652828693389893, | |
| "learning_rate": 4.829996590271646e-05, | |
| "loss": 1.4061, | |
| "step": 938 | |
| }, | |
| { | |
| "epoch": 0.18413569957839004, | |
| "grad_norm": 6.595338821411133, | |
| "learning_rate": 4.810509249024604e-05, | |
| "loss": 1.2953, | |
| "step": 939 | |
| }, | |
| { | |
| "epoch": 0.18433179723502305, | |
| "grad_norm": 4.507275581359863, | |
| "learning_rate": 4.7910488414619967e-05, | |
| "loss": 0.95, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.18452789489165605, | |
| "grad_norm": 7.35803747177124, | |
| "learning_rate": 4.7716154685841944e-05, | |
| "loss": 1.7958, | |
| "step": 941 | |
| }, | |
| { | |
| "epoch": 0.18472399254828906, | |
| "grad_norm": 5.53951358795166, | |
| "learning_rate": 4.752209231251251e-05, | |
| "loss": 2.8191, | |
| "step": 942 | |
| }, | |
| { | |
| "epoch": 0.18492009020492206, | |
| "grad_norm": 4.738945960998535, | |
| "learning_rate": 4.73283023018239e-05, | |
| "loss": 1.7728, | |
| "step": 943 | |
| }, | |
| { | |
| "epoch": 0.18511618786155506, | |
| "grad_norm": 5.192052364349365, | |
| "learning_rate": 4.713478565955478e-05, | |
| "loss": 0.8087, | |
| "step": 944 | |
| }, | |
| { | |
| "epoch": 0.18531228551818807, | |
| "grad_norm": 5.9963507652282715, | |
| "learning_rate": 4.694154339006501e-05, | |
| "loss": 1.1223, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 0.18550838317482107, | |
| "grad_norm": 14.023632049560547, | |
| "learning_rate": 4.6748576496290356e-05, | |
| "loss": 3.0858, | |
| "step": 946 | |
| }, | |
| { | |
| "epoch": 0.18570448083145408, | |
| "grad_norm": 6.268606662750244, | |
| "learning_rate": 4.655588597973754e-05, | |
| "loss": 0.8988, | |
| "step": 947 | |
| }, | |
| { | |
| "epoch": 0.18590057848808705, | |
| "grad_norm": 7.719700336456299, | |
| "learning_rate": 4.636347284047877e-05, | |
| "loss": 1.5117, | |
| "step": 948 | |
| }, | |
| { | |
| "epoch": 0.18609667614472006, | |
| "grad_norm": 4.05161714553833, | |
| "learning_rate": 4.617133807714666e-05, | |
| "loss": 0.7625, | |
| "step": 949 | |
| }, | |
| { | |
| "epoch": 0.18629277380135306, | |
| "grad_norm": 5.103604793548584, | |
| "learning_rate": 4.59794826869291e-05, | |
| "loss": 0.8914, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.18648887145798607, | |
| "grad_norm": 7.7601847648620605, | |
| "learning_rate": 4.578790766556386e-05, | |
| "loss": 0.8747, | |
| "step": 951 | |
| }, | |
| { | |
| "epoch": 0.18668496911461907, | |
| "grad_norm": 12.439096450805664, | |
| "learning_rate": 4.559661400733383e-05, | |
| "loss": 2.2439, | |
| "step": 952 | |
| }, | |
| { | |
| "epoch": 0.18688106677125207, | |
| "grad_norm": 7.002286911010742, | |
| "learning_rate": 4.5405602705061345e-05, | |
| "loss": 1.4917, | |
| "step": 953 | |
| }, | |
| { | |
| "epoch": 0.18707716442788508, | |
| "grad_norm": 5.178001880645752, | |
| "learning_rate": 4.521487475010354e-05, | |
| "loss": 1.5737, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 0.18727326208451808, | |
| "grad_norm": 9.555988311767578, | |
| "learning_rate": 4.502443113234688e-05, | |
| "loss": 1.7495, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 0.1874693597411511, | |
| "grad_norm": 8.907003402709961, | |
| "learning_rate": 4.483427284020194e-05, | |
| "loss": 1.6053, | |
| "step": 956 | |
| }, | |
| { | |
| "epoch": 0.1876654573977841, | |
| "grad_norm": 5.562190532684326, | |
| "learning_rate": 4.464440086059878e-05, | |
| "loss": 1.1854, | |
| "step": 957 | |
| }, | |
| { | |
| "epoch": 0.1878615550544171, | |
| "grad_norm": 6.078771591186523, | |
| "learning_rate": 4.4454816178981115e-05, | |
| "loss": 1.3774, | |
| "step": 958 | |
| }, | |
| { | |
| "epoch": 0.1880576527110501, | |
| "grad_norm": 12.021577835083008, | |
| "learning_rate": 4.426551977930191e-05, | |
| "loss": 1.6445, | |
| "step": 959 | |
| }, | |
| { | |
| "epoch": 0.1882537503676831, | |
| "grad_norm": 7.406763553619385, | |
| "learning_rate": 4.407651264401763e-05, | |
| "loss": 2.0784, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.1884498480243161, | |
| "grad_norm": 3.961423635482788, | |
| "learning_rate": 4.38877957540837e-05, | |
| "loss": 1.5526, | |
| "step": 961 | |
| }, | |
| { | |
| "epoch": 0.1886459456809491, | |
| "grad_norm": 7.410959243774414, | |
| "learning_rate": 4.3699370088949066e-05, | |
| "loss": 2.0247, | |
| "step": 962 | |
| }, | |
| { | |
| "epoch": 0.18884204333758212, | |
| "grad_norm": 5.237401485443115, | |
| "learning_rate": 4.3511236626551047e-05, | |
| "loss": 2.199, | |
| "step": 963 | |
| }, | |
| { | |
| "epoch": 0.18903814099421512, | |
| "grad_norm": 4.068509578704834, | |
| "learning_rate": 4.3323396343310715e-05, | |
| "loss": 2.3318, | |
| "step": 964 | |
| }, | |
| { | |
| "epoch": 0.18923423865084812, | |
| "grad_norm": 5.523316383361816, | |
| "learning_rate": 4.313585021412724e-05, | |
| "loss": 1.3474, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 0.18943033630748113, | |
| "grad_norm": 5.052401542663574, | |
| "learning_rate": 4.294859921237339e-05, | |
| "loss": 1.1039, | |
| "step": 966 | |
| }, | |
| { | |
| "epoch": 0.18962643396411413, | |
| "grad_norm": 8.468581199645996, | |
| "learning_rate": 4.2761644309889946e-05, | |
| "loss": 2.9943, | |
| "step": 967 | |
| }, | |
| { | |
| "epoch": 0.18982253162074714, | |
| "grad_norm": 4.66163444519043, | |
| "learning_rate": 4.257498647698107e-05, | |
| "loss": 2.0152, | |
| "step": 968 | |
| }, | |
| { | |
| "epoch": 0.19001862927738014, | |
| "grad_norm": 4.58842658996582, | |
| "learning_rate": 4.2388626682409194e-05, | |
| "loss": 1.2759, | |
| "step": 969 | |
| }, | |
| { | |
| "epoch": 0.19021472693401315, | |
| "grad_norm": 3.5642545223236084, | |
| "learning_rate": 4.220256589338968e-05, | |
| "loss": 0.9877, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.19041082459064615, | |
| "grad_norm": 6.506111145019531, | |
| "learning_rate": 4.201680507558631e-05, | |
| "loss": 1.7171, | |
| "step": 971 | |
| }, | |
| { | |
| "epoch": 0.19060692224727915, | |
| "grad_norm": 4.93696928024292, | |
| "learning_rate": 4.183134519310576e-05, | |
| "loss": 0.8798, | |
| "step": 972 | |
| }, | |
| { | |
| "epoch": 0.19080301990391216, | |
| "grad_norm": 12.539702415466309, | |
| "learning_rate": 4.1646187208493005e-05, | |
| "loss": 3.0454, | |
| "step": 973 | |
| }, | |
| { | |
| "epoch": 0.19099911756054516, | |
| "grad_norm": 9.189221382141113, | |
| "learning_rate": 4.146133208272608e-05, | |
| "loss": 1.9269, | |
| "step": 974 | |
| }, | |
| { | |
| "epoch": 0.19119521521717817, | |
| "grad_norm": 5.362178802490234, | |
| "learning_rate": 4.1276780775211156e-05, | |
| "loss": 1.7086, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 0.19139131287381117, | |
| "grad_norm": 4.587559700012207, | |
| "learning_rate": 4.109253424377772e-05, | |
| "loss": 0.7529, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 0.19158741053044417, | |
| "grad_norm": 7.922691345214844, | |
| "learning_rate": 4.090859344467325e-05, | |
| "loss": 2.6641, | |
| "step": 977 | |
| }, | |
| { | |
| "epoch": 0.19178350818707715, | |
| "grad_norm": 6.759201526641846, | |
| "learning_rate": 4.072495933255857e-05, | |
| "loss": 1.3197, | |
| "step": 978 | |
| }, | |
| { | |
| "epoch": 0.19197960584371015, | |
| "grad_norm": 5.635506629943848, | |
| "learning_rate": 4.054163286050276e-05, | |
| "loss": 0.8712, | |
| "step": 979 | |
| }, | |
| { | |
| "epoch": 0.19217570350034316, | |
| "grad_norm": 4.1558637619018555, | |
| "learning_rate": 4.035861497997828e-05, | |
| "loss": 2.3436, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.19237180115697616, | |
| "grad_norm": 9.821369171142578, | |
| "learning_rate": 4.017590664085593e-05, | |
| "loss": 1.6202, | |
| "step": 981 | |
| }, | |
| { | |
| "epoch": 0.19256789881360917, | |
| "grad_norm": 7.816912651062012, | |
| "learning_rate": 3.999350879139997e-05, | |
| "loss": 1.1312, | |
| "step": 982 | |
| }, | |
| { | |
| "epoch": 0.19276399647024217, | |
| "grad_norm": 5.585190296173096, | |
| "learning_rate": 3.981142237826332e-05, | |
| "loss": 1.9218, | |
| "step": 983 | |
| }, | |
| { | |
| "epoch": 0.19296009412687518, | |
| "grad_norm": 4.785935878753662, | |
| "learning_rate": 3.962964834648236e-05, | |
| "loss": 0.8693, | |
| "step": 984 | |
| }, | |
| { | |
| "epoch": 0.19315619178350818, | |
| "grad_norm": 5.3974385261535645, | |
| "learning_rate": 3.944818763947231e-05, | |
| "loss": 1.1065, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 0.19335228944014118, | |
| "grad_norm": 9.581192016601562, | |
| "learning_rate": 3.926704119902219e-05, | |
| "loss": 0.867, | |
| "step": 986 | |
| }, | |
| { | |
| "epoch": 0.1935483870967742, | |
| "grad_norm": 10.229225158691406, | |
| "learning_rate": 3.9086209965289965e-05, | |
| "loss": 2.8435, | |
| "step": 987 | |
| }, | |
| { | |
| "epoch": 0.1937444847534072, | |
| "grad_norm": 7.257898807525635, | |
| "learning_rate": 3.890569487679766e-05, | |
| "loss": 2.4263, | |
| "step": 988 | |
| }, | |
| { | |
| "epoch": 0.1939405824100402, | |
| "grad_norm": 5.026792526245117, | |
| "learning_rate": 3.87254968704265e-05, | |
| "loss": 1.2831, | |
| "step": 989 | |
| }, | |
| { | |
| "epoch": 0.1941366800666732, | |
| "grad_norm": 5.230473041534424, | |
| "learning_rate": 3.854561688141205e-05, | |
| "loss": 3.3858, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.1943327777233062, | |
| "grad_norm": 6.340061187744141, | |
| "learning_rate": 3.836605584333931e-05, | |
| "loss": 1.5421, | |
| "step": 991 | |
| }, | |
| { | |
| "epoch": 0.1945288753799392, | |
| "grad_norm": 7.890949726104736, | |
| "learning_rate": 3.818681468813794e-05, | |
| "loss": 1.7882, | |
| "step": 992 | |
| }, | |
| { | |
| "epoch": 0.1947249730365722, | |
| "grad_norm": 4.434299945831299, | |
| "learning_rate": 3.800789434607741e-05, | |
| "loss": 1.274, | |
| "step": 993 | |
| }, | |
| { | |
| "epoch": 0.19492107069320522, | |
| "grad_norm": 5.311686038970947, | |
| "learning_rate": 3.782929574576213e-05, | |
| "loss": 1.7605, | |
| "step": 994 | |
| }, | |
| { | |
| "epoch": 0.19511716834983822, | |
| "grad_norm": 22.297441482543945, | |
| "learning_rate": 3.7651019814126654e-05, | |
| "loss": 1.9094, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 0.19531326600647123, | |
| "grad_norm": 4.7996826171875, | |
| "learning_rate": 3.747306747643089e-05, | |
| "loss": 1.7196, | |
| "step": 996 | |
| }, | |
| { | |
| "epoch": 0.19550936366310423, | |
| "grad_norm": 12.06876277923584, | |
| "learning_rate": 3.729543965625526e-05, | |
| "loss": 1.2351, | |
| "step": 997 | |
| }, | |
| { | |
| "epoch": 0.19570546131973723, | |
| "grad_norm": 5.477703094482422, | |
| "learning_rate": 3.711813727549594e-05, | |
| "loss": 1.3366, | |
| "step": 998 | |
| }, | |
| { | |
| "epoch": 0.19590155897637024, | |
| "grad_norm": 4.249327659606934, | |
| "learning_rate": 3.694116125436007e-05, | |
| "loss": 1.9066, | |
| "step": 999 | |
| }, | |
| { | |
| "epoch": 0.19609765663300324, | |
| "grad_norm": 7.70191764831543, | |
| "learning_rate": 3.6764512511360935e-05, | |
| "loss": 1.3611, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.19629375428963625, | |
| "grad_norm": 5.822709560394287, | |
| "learning_rate": 3.658819196331327e-05, | |
| "loss": 2.3203, | |
| "step": 1001 | |
| }, | |
| { | |
| "epoch": 0.19648985194626925, | |
| "grad_norm": 4.531265735626221, | |
| "learning_rate": 3.6412200525328435e-05, | |
| "loss": 0.7644, | |
| "step": 1002 | |
| }, | |
| { | |
| "epoch": 0.19668594960290225, | |
| "grad_norm": 4.714295387268066, | |
| "learning_rate": 3.623653911080971e-05, | |
| "loss": 1.6555, | |
| "step": 1003 | |
| }, | |
| { | |
| "epoch": 0.19688204725953526, | |
| "grad_norm": 7.089095592498779, | |
| "learning_rate": 3.606120863144753e-05, | |
| "loss": 1.6242, | |
| "step": 1004 | |
| }, | |
| { | |
| "epoch": 0.19707814491616826, | |
| "grad_norm": 2.268071174621582, | |
| "learning_rate": 3.588620999721477e-05, | |
| "loss": 0.3747, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 0.19727424257280127, | |
| "grad_norm": 9.244414329528809, | |
| "learning_rate": 3.571154411636203e-05, | |
| "loss": 1.8298, | |
| "step": 1006 | |
| }, | |
| { | |
| "epoch": 0.19747034022943427, | |
| "grad_norm": 4.8095808029174805, | |
| "learning_rate": 3.5537211895412846e-05, | |
| "loss": 0.8923, | |
| "step": 1007 | |
| }, | |
| { | |
| "epoch": 0.19766643788606725, | |
| "grad_norm": 7.700345516204834, | |
| "learning_rate": 3.536321423915913e-05, | |
| "loss": 0.691, | |
| "step": 1008 | |
| }, | |
| { | |
| "epoch": 0.19786253554270025, | |
| "grad_norm": 7.553009510040283, | |
| "learning_rate": 3.518955205065632e-05, | |
| "loss": 1.1634, | |
| "step": 1009 | |
| }, | |
| { | |
| "epoch": 0.19805863319933326, | |
| "grad_norm": 5.901485443115234, | |
| "learning_rate": 3.5016226231218774e-05, | |
| "loss": 2.3853, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.19825473085596626, | |
| "grad_norm": 5.342306137084961, | |
| "learning_rate": 3.4843237680415156e-05, | |
| "loss": 0.9747, | |
| "step": 1011 | |
| }, | |
| { | |
| "epoch": 0.19845082851259926, | |
| "grad_norm": 5.793070316314697, | |
| "learning_rate": 3.46705872960635e-05, | |
| "loss": 0.7877, | |
| "step": 1012 | |
| }, | |
| { | |
| "epoch": 0.19864692616923227, | |
| "grad_norm": 9.105026245117188, | |
| "learning_rate": 3.449827597422698e-05, | |
| "loss": 1.5736, | |
| "step": 1013 | |
| }, | |
| { | |
| "epoch": 0.19884302382586527, | |
| "grad_norm": 7.359753608703613, | |
| "learning_rate": 3.432630460920887e-05, | |
| "loss": 1.4213, | |
| "step": 1014 | |
| }, | |
| { | |
| "epoch": 0.19903912148249828, | |
| "grad_norm": 4.728425025939941, | |
| "learning_rate": 3.415467409354809e-05, | |
| "loss": 0.7005, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 0.19923521913913128, | |
| "grad_norm": 4.326775074005127, | |
| "learning_rate": 3.398338531801457e-05, | |
| "loss": 1.0779, | |
| "step": 1016 | |
| }, | |
| { | |
| "epoch": 0.19943131679576429, | |
| "grad_norm": 5.717733383178711, | |
| "learning_rate": 3.381243917160448e-05, | |
| "loss": 2.7467, | |
| "step": 1017 | |
| }, | |
| { | |
| "epoch": 0.1996274144523973, | |
| "grad_norm": 12.189995765686035, | |
| "learning_rate": 3.364183654153592e-05, | |
| "loss": 1.4596, | |
| "step": 1018 | |
| }, | |
| { | |
| "epoch": 0.1998235121090303, | |
| "grad_norm": 5.9200239181518555, | |
| "learning_rate": 3.3471578313243903e-05, | |
| "loss": 1.8764, | |
| "step": 1019 | |
| }, | |
| { | |
| "epoch": 0.2000196097656633, | |
| "grad_norm": 11.550288200378418, | |
| "learning_rate": 3.330166537037618e-05, | |
| "loss": 1.3655, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.2002157074222963, | |
| "grad_norm": 4.954752445220947, | |
| "learning_rate": 3.313209859478839e-05, | |
| "loss": 1.2426, | |
| "step": 1021 | |
| }, | |
| { | |
| "epoch": 0.2004118050789293, | |
| "grad_norm": 11.455079078674316, | |
| "learning_rate": 3.296287886653941e-05, | |
| "loss": 2.2854, | |
| "step": 1022 | |
| }, | |
| { | |
| "epoch": 0.2006079027355623, | |
| "grad_norm": 6.088212966918945, | |
| "learning_rate": 3.2794007063887186e-05, | |
| "loss": 1.3675, | |
| "step": 1023 | |
| }, | |
| { | |
| "epoch": 0.20080400039219531, | |
| "grad_norm": 7.098971366882324, | |
| "learning_rate": 3.262548406328365e-05, | |
| "loss": 1.0983, | |
| "step": 1024 | |
| }, | |
| { | |
| "epoch": 0.20100009804882832, | |
| "grad_norm": 3.8312485218048096, | |
| "learning_rate": 3.245731073937068e-05, | |
| "loss": 1.7243, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 0.20119619570546132, | |
| "grad_norm": 3.807969570159912, | |
| "learning_rate": 3.2289487964975076e-05, | |
| "loss": 1.0291, | |
| "step": 1026 | |
| }, | |
| { | |
| "epoch": 0.20139229336209433, | |
| "grad_norm": 4.957508087158203, | |
| "learning_rate": 3.212201661110449e-05, | |
| "loss": 1.1555, | |
| "step": 1027 | |
| }, | |
| { | |
| "epoch": 0.20158839101872733, | |
| "grad_norm": 10.736882209777832, | |
| "learning_rate": 3.1954897546942584e-05, | |
| "loss": 1.9095, | |
| "step": 1028 | |
| }, | |
| { | |
| "epoch": 0.20178448867536034, | |
| "grad_norm": 9.901065826416016, | |
| "learning_rate": 3.1788131639844534e-05, | |
| "loss": 2.2914, | |
| "step": 1029 | |
| }, | |
| { | |
| "epoch": 0.20198058633199334, | |
| "grad_norm": 5.642101764678955, | |
| "learning_rate": 3.162171975533282e-05, | |
| "loss": 1.3295, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.20217668398862634, | |
| "grad_norm": 6.677104949951172, | |
| "learning_rate": 3.1455662757092306e-05, | |
| "loss": 1.3554, | |
| "step": 1031 | |
| }, | |
| { | |
| "epoch": 0.20237278164525935, | |
| "grad_norm": 5.27210807800293, | |
| "learning_rate": 3.1289961506966214e-05, | |
| "loss": 1.0698, | |
| "step": 1032 | |
| }, | |
| { | |
| "epoch": 0.20256887930189235, | |
| "grad_norm": 6.362321853637695, | |
| "learning_rate": 3.11246168649512e-05, | |
| "loss": 1.135, | |
| "step": 1033 | |
| }, | |
| { | |
| "epoch": 0.20276497695852536, | |
| "grad_norm": 4.513217449188232, | |
| "learning_rate": 3.095962968919319e-05, | |
| "loss": 1.9138, | |
| "step": 1034 | |
| }, | |
| { | |
| "epoch": 0.20296107461515836, | |
| "grad_norm": 16.53858184814453, | |
| "learning_rate": 3.079500083598297e-05, | |
| "loss": 2.4364, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 0.20315717227179136, | |
| "grad_norm": 5.629922389984131, | |
| "learning_rate": 3.063073115975136e-05, | |
| "loss": 1.7899, | |
| "step": 1036 | |
| }, | |
| { | |
| "epoch": 0.20335326992842437, | |
| "grad_norm": 9.979622840881348, | |
| "learning_rate": 3.0466821513065314e-05, | |
| "loss": 1.7134, | |
| "step": 1037 | |
| }, | |
| { | |
| "epoch": 0.20354936758505734, | |
| "grad_norm": 8.400696754455566, | |
| "learning_rate": 3.030327274662298e-05, | |
| "loss": 1.6557, | |
| "step": 1038 | |
| }, | |
| { | |
| "epoch": 0.20374546524169035, | |
| "grad_norm": 9.457781791687012, | |
| "learning_rate": 3.0140085709249667e-05, | |
| "loss": 2.3906, | |
| "step": 1039 | |
| }, | |
| { | |
| "epoch": 0.20394156289832335, | |
| "grad_norm": 4.818408489227295, | |
| "learning_rate": 2.997726124789324e-05, | |
| "loss": 1.0997, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.20413766055495636, | |
| "grad_norm": 7.07374382019043, | |
| "learning_rate": 2.9814800207619774e-05, | |
| "loss": 1.441, | |
| "step": 1041 | |
| }, | |
| { | |
| "epoch": 0.20433375821158936, | |
| "grad_norm": 9.976203918457031, | |
| "learning_rate": 2.9652703431609263e-05, | |
| "loss": 2.2031, | |
| "step": 1042 | |
| }, | |
| { | |
| "epoch": 0.20452985586822237, | |
| "grad_norm": 8.051301002502441, | |
| "learning_rate": 2.9490971761151e-05, | |
| "loss": 1.2837, | |
| "step": 1043 | |
| }, | |
| { | |
| "epoch": 0.20472595352485537, | |
| "grad_norm": 3.8618202209472656, | |
| "learning_rate": 2.9329606035639458e-05, | |
| "loss": 1.5356, | |
| "step": 1044 | |
| }, | |
| { | |
| "epoch": 0.20472595352485537, | |
| "eval_loss": 0.3870697021484375, | |
| "eval_runtime": 78.8825, | |
| "eval_samples_per_second": 27.23, | |
| "eval_steps_per_second": 13.615, | |
| "step": 1044 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 1389, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 348, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.242105399144612e+17, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |