| { |
| "best_metric": 1.4238797426223755, |
| "best_model_checkpoint": "tinybert_train_book_v2/checkpoint-350000", |
| "epoch": 25.0, |
| "eval_steps": 10000, |
| "global_step": 355850, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.0351271603203597, |
| "grad_norm": 2.383054733276367, |
| "learning_rate": 5e-06, |
| "loss": 9.5651, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.0702543206407194, |
| "grad_norm": 0.9246260523796082, |
| "learning_rate": 1e-05, |
| "loss": 7.4554, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.1053814809610791, |
| "grad_norm": 0.49956014752388, |
| "learning_rate": 1.5e-05, |
| "loss": 6.2188, |
| "step": 1500 |
| }, |
| { |
| "epoch": 0.1405086412814388, |
| "grad_norm": 0.5331165194511414, |
| "learning_rate": 2e-05, |
| "loss": 6.0207, |
| "step": 2000 |
| }, |
| { |
| "epoch": 0.1756358016017985, |
| "grad_norm": 0.666080892086029, |
| "learning_rate": 2.5e-05, |
| "loss": 5.9325, |
| "step": 2500 |
| }, |
| { |
| "epoch": 0.2107629619221582, |
| "grad_norm": 0.5235181450843811, |
| "learning_rate": 3e-05, |
| "loss": 5.8734, |
| "step": 3000 |
| }, |
| { |
| "epoch": 0.2458901222425179, |
| "grad_norm": 0.5165570378303528, |
| "learning_rate": 3.5e-05, |
| "loss": 5.8359, |
| "step": 3500 |
| }, |
| { |
| "epoch": 0.2810172825628776, |
| "grad_norm": 0.5053095817565918, |
| "learning_rate": 4e-05, |
| "loss": 5.8055, |
| "step": 4000 |
| }, |
| { |
| "epoch": 0.31614444288323734, |
| "grad_norm": 0.5578732490539551, |
| "learning_rate": 4.5e-05, |
| "loss": 5.7797, |
| "step": 4500 |
| }, |
| { |
| "epoch": 0.351271603203597, |
| "grad_norm": 0.48547250032424927, |
| "learning_rate": 5e-05, |
| "loss": 5.7586, |
| "step": 5000 |
| }, |
| { |
| "epoch": 0.38639876352395675, |
| "grad_norm": 0.5520783066749573, |
| "learning_rate": 5.500000000000001e-05, |
| "loss": 5.7352, |
| "step": 5500 |
| }, |
| { |
| "epoch": 0.4215259238443164, |
| "grad_norm": 0.5271908640861511, |
| "learning_rate": 6e-05, |
| "loss": 5.7235, |
| "step": 6000 |
| }, |
| { |
| "epoch": 0.45665308416467615, |
| "grad_norm": 0.5117993950843811, |
| "learning_rate": 6.500000000000001e-05, |
| "loss": 5.7078, |
| "step": 6500 |
| }, |
| { |
| "epoch": 0.4917802444850358, |
| "grad_norm": 0.4803953468799591, |
| "learning_rate": 7e-05, |
| "loss": 5.6977, |
| "step": 7000 |
| }, |
| { |
| "epoch": 0.5269074048053956, |
| "grad_norm": 0.4752044975757599, |
| "learning_rate": 7.500000000000001e-05, |
| "loss": 5.6854, |
| "step": 7500 |
| }, |
| { |
| "epoch": 0.5620345651257552, |
| "grad_norm": 0.6038092970848083, |
| "learning_rate": 8e-05, |
| "loss": 5.6738, |
| "step": 8000 |
| }, |
| { |
| "epoch": 0.5971617254461149, |
| "grad_norm": 0.5730042457580566, |
| "learning_rate": 8.5e-05, |
| "loss": 5.6634, |
| "step": 8500 |
| }, |
| { |
| "epoch": 0.6322888857664747, |
| "grad_norm": 0.822317898273468, |
| "learning_rate": 9e-05, |
| "loss": 5.6513, |
| "step": 9000 |
| }, |
| { |
| "epoch": 0.6674160460868344, |
| "grad_norm": 0.46957486867904663, |
| "learning_rate": 9.5e-05, |
| "loss": 5.647, |
| "step": 9500 |
| }, |
| { |
| "epoch": 0.702543206407194, |
| "grad_norm": 0.46289733052253723, |
| "learning_rate": 0.0001, |
| "loss": 5.6396, |
| "step": 10000 |
| }, |
| { |
| "epoch": 0.702543206407194, |
| "eval_accuracy": 0.16447421236542026, |
| "eval_loss": 5.4784393310546875, |
| "eval_runtime": 120.9916, |
| "eval_samples_per_second": 997.64, |
| "eval_steps_per_second": 6.24, |
| "step": 10000 |
| }, |
| { |
| "epoch": 0.7376703667275537, |
| "grad_norm": 0.4592319130897522, |
| "learning_rate": 9.985542865404077e-05, |
| "loss": 5.632, |
| "step": 10500 |
| }, |
| { |
| "epoch": 0.7727975270479135, |
| "grad_norm": 0.44356584548950195, |
| "learning_rate": 9.971085730808155e-05, |
| "loss": 5.6253, |
| "step": 11000 |
| }, |
| { |
| "epoch": 0.8079246873682732, |
| "grad_norm": 0.45771583914756775, |
| "learning_rate": 9.956628596212231e-05, |
| "loss": 5.6223, |
| "step": 11500 |
| }, |
| { |
| "epoch": 0.8430518476886328, |
| "grad_norm": 0.627301812171936, |
| "learning_rate": 9.942171461616309e-05, |
| "loss": 5.6142, |
| "step": 12000 |
| }, |
| { |
| "epoch": 0.8781790080089925, |
| "grad_norm": 0.5001237392425537, |
| "learning_rate": 9.927714327020385e-05, |
| "loss": 5.6105, |
| "step": 12500 |
| }, |
| { |
| "epoch": 0.9133061683293523, |
| "grad_norm": 0.5627796053886414, |
| "learning_rate": 9.913257192424462e-05, |
| "loss": 5.6017, |
| "step": 13000 |
| }, |
| { |
| "epoch": 0.948433328649712, |
| "grad_norm": 0.5000184178352356, |
| "learning_rate": 9.89880005782854e-05, |
| "loss": 5.5963, |
| "step": 13500 |
| }, |
| { |
| "epoch": 0.9835604889700716, |
| "grad_norm": 0.4531259834766388, |
| "learning_rate": 9.884342923232616e-05, |
| "loss": 5.5942, |
| "step": 14000 |
| }, |
| { |
| "epoch": 1.0186876492904313, |
| "grad_norm": 0.4834921658039093, |
| "learning_rate": 9.869885788636692e-05, |
| "loss": 5.588, |
| "step": 14500 |
| }, |
| { |
| "epoch": 1.053814809610791, |
| "grad_norm": 0.5159989595413208, |
| "learning_rate": 9.855428654040769e-05, |
| "loss": 5.5856, |
| "step": 15000 |
| }, |
| { |
| "epoch": 1.0889419699311507, |
| "grad_norm": 0.4877914786338806, |
| "learning_rate": 9.840971519444846e-05, |
| "loss": 5.5793, |
| "step": 15500 |
| }, |
| { |
| "epoch": 1.1240691302515105, |
| "grad_norm": 0.4312230944633484, |
| "learning_rate": 9.826514384848924e-05, |
| "loss": 5.5762, |
| "step": 16000 |
| }, |
| { |
| "epoch": 1.1591962905718702, |
| "grad_norm": 0.44624242186546326, |
| "learning_rate": 9.812057250253e-05, |
| "loss": 5.5725, |
| "step": 16500 |
| }, |
| { |
| "epoch": 1.1943234508922298, |
| "grad_norm": 0.44468972086906433, |
| "learning_rate": 9.797600115657077e-05, |
| "loss": 5.5701, |
| "step": 17000 |
| }, |
| { |
| "epoch": 1.2294506112125896, |
| "grad_norm": 0.44398409128189087, |
| "learning_rate": 9.783142981061155e-05, |
| "loss": 5.5632, |
| "step": 17500 |
| }, |
| { |
| "epoch": 1.2645777715329491, |
| "grad_norm": 0.46841973066329956, |
| "learning_rate": 9.768685846465231e-05, |
| "loss": 5.5622, |
| "step": 18000 |
| }, |
| { |
| "epoch": 1.299704931853309, |
| "grad_norm": 0.41625532507896423, |
| "learning_rate": 9.754228711869307e-05, |
| "loss": 5.5522, |
| "step": 18500 |
| }, |
| { |
| "epoch": 1.3348320921736687, |
| "grad_norm": 0.4070848822593689, |
| "learning_rate": 9.739771577273384e-05, |
| "loss": 5.5558, |
| "step": 19000 |
| }, |
| { |
| "epoch": 1.3699592524940285, |
| "grad_norm": 0.4079872965812683, |
| "learning_rate": 9.725314442677463e-05, |
| "loss": 5.5551, |
| "step": 19500 |
| }, |
| { |
| "epoch": 1.405086412814388, |
| "grad_norm": 0.3937780559062958, |
| "learning_rate": 9.71085730808154e-05, |
| "loss": 5.5513, |
| "step": 20000 |
| }, |
| { |
| "epoch": 1.405086412814388, |
| "eval_accuracy": 0.16595895902858707, |
| "eval_loss": 5.408265590667725, |
| "eval_runtime": 122.0231, |
| "eval_samples_per_second": 989.206, |
| "eval_steps_per_second": 6.187, |
| "step": 20000 |
| }, |
| { |
| "epoch": 1.4402135731347478, |
| "grad_norm": 0.4092162251472473, |
| "learning_rate": 9.696400173485616e-05, |
| "loss": 5.5471, |
| "step": 20500 |
| }, |
| { |
| "epoch": 1.4753407334551074, |
| "grad_norm": 0.7277459502220154, |
| "learning_rate": 9.681943038889692e-05, |
| "loss": 5.516, |
| "step": 21000 |
| }, |
| { |
| "epoch": 1.5104678937754672, |
| "grad_norm": 1.4790245294570923, |
| "learning_rate": 9.66748590429377e-05, |
| "loss": 5.4136, |
| "step": 21500 |
| }, |
| { |
| "epoch": 1.545595054095827, |
| "grad_norm": 1.2697235345840454, |
| "learning_rate": 9.653028769697846e-05, |
| "loss": 5.2823, |
| "step": 22000 |
| }, |
| { |
| "epoch": 1.5807222144161865, |
| "grad_norm": 3.272162675857544, |
| "learning_rate": 9.638571635101923e-05, |
| "loss": 4.9266, |
| "step": 22500 |
| }, |
| { |
| "epoch": 1.6158493747365463, |
| "grad_norm": 2.815300941467285, |
| "learning_rate": 9.624114500506e-05, |
| "loss": 4.4687, |
| "step": 23000 |
| }, |
| { |
| "epoch": 1.650976535056906, |
| "grad_norm": 2.189101219177246, |
| "learning_rate": 9.609657365910077e-05, |
| "loss": 4.1717, |
| "step": 23500 |
| }, |
| { |
| "epoch": 1.6861036953772657, |
| "grad_norm": 2.5990891456604004, |
| "learning_rate": 9.595200231314155e-05, |
| "loss": 3.8671, |
| "step": 24000 |
| }, |
| { |
| "epoch": 1.7212308556976255, |
| "grad_norm": 1.462751865386963, |
| "learning_rate": 9.580743096718231e-05, |
| "loss": 3.6371, |
| "step": 24500 |
| }, |
| { |
| "epoch": 1.7563580160179852, |
| "grad_norm": 1.3992371559143066, |
| "learning_rate": 9.566285962122307e-05, |
| "loss": 3.4862, |
| "step": 25000 |
| }, |
| { |
| "epoch": 1.7914851763383448, |
| "grad_norm": 1.383026123046875, |
| "learning_rate": 9.551828827526385e-05, |
| "loss": 3.3789, |
| "step": 25500 |
| }, |
| { |
| "epoch": 1.8266123366587044, |
| "grad_norm": 1.2738844156265259, |
| "learning_rate": 9.537371692930462e-05, |
| "loss": 3.2885, |
| "step": 26000 |
| }, |
| { |
| "epoch": 1.8617394969790642, |
| "grad_norm": 1.3613308668136597, |
| "learning_rate": 9.522914558334538e-05, |
| "loss": 3.2159, |
| "step": 26500 |
| }, |
| { |
| "epoch": 1.896866657299424, |
| "grad_norm": 1.2374885082244873, |
| "learning_rate": 9.508457423738616e-05, |
| "loss": 3.1581, |
| "step": 27000 |
| }, |
| { |
| "epoch": 1.9319938176197837, |
| "grad_norm": 1.2320979833602905, |
| "learning_rate": 9.494000289142692e-05, |
| "loss": 3.1081, |
| "step": 27500 |
| }, |
| { |
| "epoch": 1.9671209779401433, |
| "grad_norm": 1.1193397045135498, |
| "learning_rate": 9.47954315454677e-05, |
| "loss": 3.0601, |
| "step": 28000 |
| }, |
| { |
| "epoch": 2.002248138260503, |
| "grad_norm": 1.2231374979019165, |
| "learning_rate": 9.465086019950846e-05, |
| "loss": 3.0291, |
| "step": 28500 |
| }, |
| { |
| "epoch": 2.0373752985808626, |
| "grad_norm": 1.1425702571868896, |
| "learning_rate": 9.450628885354923e-05, |
| "loss": 2.9847, |
| "step": 29000 |
| }, |
| { |
| "epoch": 2.0725024589012224, |
| "grad_norm": 1.176235318183899, |
| "learning_rate": 9.436171750758999e-05, |
| "loss": 2.952, |
| "step": 29500 |
| }, |
| { |
| "epoch": 2.107629619221582, |
| "grad_norm": 1.0708513259887695, |
| "learning_rate": 9.421714616163077e-05, |
| "loss": 2.9289, |
| "step": 30000 |
| }, |
| { |
| "epoch": 2.107629619221582, |
| "eval_accuracy": 0.5214166097696011, |
| "eval_loss": 2.5487782955169678, |
| "eval_runtime": 122.6181, |
| "eval_samples_per_second": 984.406, |
| "eval_steps_per_second": 6.157, |
| "step": 30000 |
| }, |
| { |
| "epoch": 2.142756779541942, |
| "grad_norm": 1.0883877277374268, |
| "learning_rate": 9.407257481567155e-05, |
| "loss": 2.8998, |
| "step": 30500 |
| }, |
| { |
| "epoch": 2.1778839398623013, |
| "grad_norm": 1.073864221572876, |
| "learning_rate": 9.392800346971231e-05, |
| "loss": 2.8706, |
| "step": 31000 |
| }, |
| { |
| "epoch": 2.213011100182661, |
| "grad_norm": 1.1122660636901855, |
| "learning_rate": 9.378343212375307e-05, |
| "loss": 2.8513, |
| "step": 31500 |
| }, |
| { |
| "epoch": 2.248138260503021, |
| "grad_norm": 1.056282877922058, |
| "learning_rate": 9.363886077779385e-05, |
| "loss": 2.8277, |
| "step": 32000 |
| }, |
| { |
| "epoch": 2.2832654208233807, |
| "grad_norm": 1.092287540435791, |
| "learning_rate": 9.349428943183462e-05, |
| "loss": 2.8084, |
| "step": 32500 |
| }, |
| { |
| "epoch": 2.3183925811437405, |
| "grad_norm": 1.0690412521362305, |
| "learning_rate": 9.334971808587538e-05, |
| "loss": 2.7854, |
| "step": 33000 |
| }, |
| { |
| "epoch": 2.3535197414641003, |
| "grad_norm": 1.0466443300247192, |
| "learning_rate": 9.320514673991614e-05, |
| "loss": 2.7653, |
| "step": 33500 |
| }, |
| { |
| "epoch": 2.3886469017844596, |
| "grad_norm": 1.04971444606781, |
| "learning_rate": 9.306057539395692e-05, |
| "loss": 2.7481, |
| "step": 34000 |
| }, |
| { |
| "epoch": 2.4237740621048194, |
| "grad_norm": 1.0288169384002686, |
| "learning_rate": 9.29160040479977e-05, |
| "loss": 2.7276, |
| "step": 34500 |
| }, |
| { |
| "epoch": 2.458901222425179, |
| "grad_norm": 1.020531415939331, |
| "learning_rate": 9.277143270203846e-05, |
| "loss": 2.7115, |
| "step": 35000 |
| }, |
| { |
| "epoch": 2.494028382745539, |
| "grad_norm": 1.0094306468963623, |
| "learning_rate": 9.262686135607923e-05, |
| "loss": 2.6973, |
| "step": 35500 |
| }, |
| { |
| "epoch": 2.5291555430658983, |
| "grad_norm": 1.018281102180481, |
| "learning_rate": 9.248229001012e-05, |
| "loss": 2.681, |
| "step": 36000 |
| }, |
| { |
| "epoch": 2.564282703386258, |
| "grad_norm": 1.0176887512207031, |
| "learning_rate": 9.233771866416077e-05, |
| "loss": 2.6646, |
| "step": 36500 |
| }, |
| { |
| "epoch": 2.599409863706618, |
| "grad_norm": 0.971754789352417, |
| "learning_rate": 9.219314731820153e-05, |
| "loss": 2.6496, |
| "step": 37000 |
| }, |
| { |
| "epoch": 2.6345370240269776, |
| "grad_norm": 0.9924113154411316, |
| "learning_rate": 9.20485759722423e-05, |
| "loss": 2.6362, |
| "step": 37500 |
| }, |
| { |
| "epoch": 2.6696641843473374, |
| "grad_norm": 0.9813032150268555, |
| "learning_rate": 9.190400462628307e-05, |
| "loss": 2.6223, |
| "step": 38000 |
| }, |
| { |
| "epoch": 2.704791344667697, |
| "grad_norm": 0.9945317506790161, |
| "learning_rate": 9.175943328032385e-05, |
| "loss": 2.609, |
| "step": 38500 |
| }, |
| { |
| "epoch": 2.739918504988057, |
| "grad_norm": 0.9665996432304382, |
| "learning_rate": 9.161486193436461e-05, |
| "loss": 2.5943, |
| "step": 39000 |
| }, |
| { |
| "epoch": 2.7750456653084163, |
| "grad_norm": 0.9895161390304565, |
| "learning_rate": 9.147029058840538e-05, |
| "loss": 2.5829, |
| "step": 39500 |
| }, |
| { |
| "epoch": 2.810172825628776, |
| "grad_norm": 0.9730009436607361, |
| "learning_rate": 9.132571924244616e-05, |
| "loss": 2.5686, |
| "step": 40000 |
| }, |
| { |
| "epoch": 2.810172825628776, |
| "eval_accuracy": 0.5650218670053577, |
| "eval_loss": 2.2268664836883545, |
| "eval_runtime": 123.9655, |
| "eval_samples_per_second": 973.706, |
| "eval_steps_per_second": 6.09, |
| "step": 40000 |
| }, |
| { |
| "epoch": 2.845299985949136, |
| "grad_norm": 0.9833718538284302, |
| "learning_rate": 9.118114789648692e-05, |
| "loss": 2.5534, |
| "step": 40500 |
| }, |
| { |
| "epoch": 2.8804271462694957, |
| "grad_norm": 0.9579553604125977, |
| "learning_rate": 9.103657655052768e-05, |
| "loss": 2.5412, |
| "step": 41000 |
| }, |
| { |
| "epoch": 2.915554306589855, |
| "grad_norm": 0.9241520762443542, |
| "learning_rate": 9.089200520456845e-05, |
| "loss": 2.5296, |
| "step": 41500 |
| }, |
| { |
| "epoch": 2.950681466910215, |
| "grad_norm": 0.9512749314308167, |
| "learning_rate": 9.074743385860923e-05, |
| "loss": 2.5204, |
| "step": 42000 |
| }, |
| { |
| "epoch": 2.9858086272305746, |
| "grad_norm": 0.9125353097915649, |
| "learning_rate": 9.060286251265e-05, |
| "loss": 2.5089, |
| "step": 42500 |
| }, |
| { |
| "epoch": 3.0209357875509344, |
| "grad_norm": 0.9652732014656067, |
| "learning_rate": 9.045829116669077e-05, |
| "loss": 2.4886, |
| "step": 43000 |
| }, |
| { |
| "epoch": 3.056062947871294, |
| "grad_norm": 0.9841357469558716, |
| "learning_rate": 9.031371982073153e-05, |
| "loss": 2.4832, |
| "step": 43500 |
| }, |
| { |
| "epoch": 3.091190108191654, |
| "grad_norm": 0.9314492344856262, |
| "learning_rate": 9.016914847477231e-05, |
| "loss": 2.4752, |
| "step": 44000 |
| }, |
| { |
| "epoch": 3.1263172685120133, |
| "grad_norm": 0.935992419719696, |
| "learning_rate": 9.002457712881307e-05, |
| "loss": 2.4617, |
| "step": 44500 |
| }, |
| { |
| "epoch": 3.161444428832373, |
| "grad_norm": 0.9312289357185364, |
| "learning_rate": 8.988000578285384e-05, |
| "loss": 2.452, |
| "step": 45000 |
| }, |
| { |
| "epoch": 3.196571589152733, |
| "grad_norm": 0.9356395602226257, |
| "learning_rate": 8.973543443689461e-05, |
| "loss": 2.4386, |
| "step": 45500 |
| }, |
| { |
| "epoch": 3.2316987494730927, |
| "grad_norm": 0.9319062232971191, |
| "learning_rate": 8.959086309093538e-05, |
| "loss": 2.4312, |
| "step": 46000 |
| }, |
| { |
| "epoch": 3.2668259097934524, |
| "grad_norm": 0.9284427165985107, |
| "learning_rate": 8.944629174497616e-05, |
| "loss": 2.4191, |
| "step": 46500 |
| }, |
| { |
| "epoch": 3.301953070113812, |
| "grad_norm": 0.920961856842041, |
| "learning_rate": 8.930172039901692e-05, |
| "loss": 2.4094, |
| "step": 47000 |
| }, |
| { |
| "epoch": 3.3370802304341716, |
| "grad_norm": 0.8934434652328491, |
| "learning_rate": 8.915714905305768e-05, |
| "loss": 2.4003, |
| "step": 47500 |
| }, |
| { |
| "epoch": 3.3722073907545314, |
| "grad_norm": 0.934516966342926, |
| "learning_rate": 8.901257770709845e-05, |
| "loss": 2.3924, |
| "step": 48000 |
| }, |
| { |
| "epoch": 3.407334551074891, |
| "grad_norm": 0.8855811357498169, |
| "learning_rate": 8.886800636113923e-05, |
| "loss": 2.3796, |
| "step": 48500 |
| }, |
| { |
| "epoch": 3.442461711395251, |
| "grad_norm": 0.921901285648346, |
| "learning_rate": 8.872343501518e-05, |
| "loss": 2.3688, |
| "step": 49000 |
| }, |
| { |
| "epoch": 3.4775888717156107, |
| "grad_norm": 0.9006240963935852, |
| "learning_rate": 8.857886366922077e-05, |
| "loss": 2.3602, |
| "step": 49500 |
| }, |
| { |
| "epoch": 3.51271603203597, |
| "grad_norm": 0.9159087538719177, |
| "learning_rate": 8.843429232326153e-05, |
| "loss": 2.3542, |
| "step": 50000 |
| }, |
| { |
| "epoch": 3.51271603203597, |
| "eval_accuracy": 0.5945121298744974, |
| "eval_loss": 2.0268421173095703, |
| "eval_runtime": 122.7402, |
| "eval_samples_per_second": 983.427, |
| "eval_steps_per_second": 6.151, |
| "step": 50000 |
| }, |
| { |
| "epoch": 3.54784319235633, |
| "grad_norm": 0.9109728336334229, |
| "learning_rate": 8.828972097730231e-05, |
| "loss": 2.3482, |
| "step": 50500 |
| }, |
| { |
| "epoch": 3.5829703526766896, |
| "grad_norm": 0.9026636481285095, |
| "learning_rate": 8.814514963134307e-05, |
| "loss": 2.3357, |
| "step": 51000 |
| }, |
| { |
| "epoch": 3.6180975129970494, |
| "grad_norm": 0.8863571882247925, |
| "learning_rate": 8.800057828538384e-05, |
| "loss": 2.3316, |
| "step": 51500 |
| }, |
| { |
| "epoch": 3.6532246733174087, |
| "grad_norm": 0.9058268666267395, |
| "learning_rate": 8.78560069394246e-05, |
| "loss": 2.3224, |
| "step": 52000 |
| }, |
| { |
| "epoch": 3.6883518336377685, |
| "grad_norm": 0.894909679889679, |
| "learning_rate": 8.771143559346538e-05, |
| "loss": 2.311, |
| "step": 52500 |
| }, |
| { |
| "epoch": 3.7234789939581283, |
| "grad_norm": 0.9029729962348938, |
| "learning_rate": 8.756686424750616e-05, |
| "loss": 2.3077, |
| "step": 53000 |
| }, |
| { |
| "epoch": 3.758606154278488, |
| "grad_norm": 0.9087358117103577, |
| "learning_rate": 8.742229290154692e-05, |
| "loss": 2.2997, |
| "step": 53500 |
| }, |
| { |
| "epoch": 3.793733314598848, |
| "grad_norm": 0.8855478167533875, |
| "learning_rate": 8.727772155558768e-05, |
| "loss": 2.2866, |
| "step": 54000 |
| }, |
| { |
| "epoch": 3.8288604749192077, |
| "grad_norm": 0.891355574131012, |
| "learning_rate": 8.713315020962846e-05, |
| "loss": 2.2795, |
| "step": 54500 |
| }, |
| { |
| "epoch": 3.8639876352395675, |
| "grad_norm": 0.8804638385772705, |
| "learning_rate": 8.698857886366923e-05, |
| "loss": 2.2686, |
| "step": 55000 |
| }, |
| { |
| "epoch": 3.899114795559927, |
| "grad_norm": 0.8741514682769775, |
| "learning_rate": 8.684400751770999e-05, |
| "loss": 2.2626, |
| "step": 55500 |
| }, |
| { |
| "epoch": 3.9342419558802866, |
| "grad_norm": 0.8713488578796387, |
| "learning_rate": 8.669943617175075e-05, |
| "loss": 2.256, |
| "step": 56000 |
| }, |
| { |
| "epoch": 3.9693691162006464, |
| "grad_norm": 0.873836100101471, |
| "learning_rate": 8.655486482579153e-05, |
| "loss": 2.2516, |
| "step": 56500 |
| }, |
| { |
| "epoch": 4.004496276521006, |
| "grad_norm": 0.8814264535903931, |
| "learning_rate": 8.641029347983231e-05, |
| "loss": 2.2409, |
| "step": 57000 |
| }, |
| { |
| "epoch": 4.0396234368413655, |
| "grad_norm": 0.8913155794143677, |
| "learning_rate": 8.626572213387307e-05, |
| "loss": 2.2315, |
| "step": 57500 |
| }, |
| { |
| "epoch": 4.074750597161725, |
| "grad_norm": 0.8890813589096069, |
| "learning_rate": 8.612115078791384e-05, |
| "loss": 2.2267, |
| "step": 58000 |
| }, |
| { |
| "epoch": 4.109877757482085, |
| "grad_norm": 0.8791043162345886, |
| "learning_rate": 8.597657944195461e-05, |
| "loss": 2.221, |
| "step": 58500 |
| }, |
| { |
| "epoch": 4.145004917802445, |
| "grad_norm": 0.9051641225814819, |
| "learning_rate": 8.583200809599538e-05, |
| "loss": 2.2135, |
| "step": 59000 |
| }, |
| { |
| "epoch": 4.180132078122805, |
| "grad_norm": 0.8600994944572449, |
| "learning_rate": 8.568743675003614e-05, |
| "loss": 2.2129, |
| "step": 59500 |
| }, |
| { |
| "epoch": 4.215259238443164, |
| "grad_norm": 0.9008757472038269, |
| "learning_rate": 8.55428654040769e-05, |
| "loss": 2.2044, |
| "step": 60000 |
| }, |
| { |
| "epoch": 4.215259238443164, |
| "eval_accuracy": 0.6152191597308374, |
| "eval_loss": 1.8923770189285278, |
| "eval_runtime": 125.1362, |
| "eval_samples_per_second": 964.597, |
| "eval_steps_per_second": 6.033, |
| "step": 60000 |
| }, |
| { |
| "epoch": 4.250386398763524, |
| "grad_norm": 0.8716664910316467, |
| "learning_rate": 8.539829405811768e-05, |
| "loss": 2.201, |
| "step": 60500 |
| }, |
| { |
| "epoch": 4.285513559083884, |
| "grad_norm": 0.8732978701591492, |
| "learning_rate": 8.525372271215846e-05, |
| "loss": 2.1911, |
| "step": 61000 |
| }, |
| { |
| "epoch": 4.320640719404244, |
| "grad_norm": 0.8640414476394653, |
| "learning_rate": 8.510915136619923e-05, |
| "loss": 2.1848, |
| "step": 61500 |
| }, |
| { |
| "epoch": 4.355767879724603, |
| "grad_norm": 0.8764206171035767, |
| "learning_rate": 8.496458002023999e-05, |
| "loss": 2.1804, |
| "step": 62000 |
| }, |
| { |
| "epoch": 4.3908950400449624, |
| "grad_norm": 0.8793691992759705, |
| "learning_rate": 8.482000867428075e-05, |
| "loss": 2.1801, |
| "step": 62500 |
| }, |
| { |
| "epoch": 4.426022200365322, |
| "grad_norm": 0.8929078578948975, |
| "learning_rate": 8.467543732832153e-05, |
| "loss": 2.1742, |
| "step": 63000 |
| }, |
| { |
| "epoch": 4.461149360685682, |
| "grad_norm": 0.8832878470420837, |
| "learning_rate": 8.45308659823623e-05, |
| "loss": 2.1656, |
| "step": 63500 |
| }, |
| { |
| "epoch": 4.496276521006042, |
| "grad_norm": 0.9125290513038635, |
| "learning_rate": 8.438629463640307e-05, |
| "loss": 2.1626, |
| "step": 64000 |
| }, |
| { |
| "epoch": 4.531403681326402, |
| "grad_norm": 0.9044434428215027, |
| "learning_rate": 8.424172329044384e-05, |
| "loss": 2.1576, |
| "step": 64500 |
| }, |
| { |
| "epoch": 4.566530841646761, |
| "grad_norm": 0.8726450800895691, |
| "learning_rate": 8.409715194448461e-05, |
| "loss": 2.152, |
| "step": 65000 |
| }, |
| { |
| "epoch": 4.601658001967121, |
| "grad_norm": 0.8615670800209045, |
| "learning_rate": 8.395258059852538e-05, |
| "loss": 2.1476, |
| "step": 65500 |
| }, |
| { |
| "epoch": 4.636785162287481, |
| "grad_norm": 0.8794130086898804, |
| "learning_rate": 8.380800925256614e-05, |
| "loss": 2.1423, |
| "step": 66000 |
| }, |
| { |
| "epoch": 4.671912322607841, |
| "grad_norm": 0.8808785676956177, |
| "learning_rate": 8.36634379066069e-05, |
| "loss": 2.1402, |
| "step": 66500 |
| }, |
| { |
| "epoch": 4.7070394829282005, |
| "grad_norm": 0.9059625267982483, |
| "learning_rate": 8.351886656064768e-05, |
| "loss": 2.1322, |
| "step": 67000 |
| }, |
| { |
| "epoch": 4.742166643248559, |
| "grad_norm": 0.8811041712760925, |
| "learning_rate": 8.337429521468846e-05, |
| "loss": 2.1298, |
| "step": 67500 |
| }, |
| { |
| "epoch": 4.777293803568919, |
| "grad_norm": 0.8630144000053406, |
| "learning_rate": 8.322972386872923e-05, |
| "loss": 2.1267, |
| "step": 68000 |
| }, |
| { |
| "epoch": 4.812420963889279, |
| "grad_norm": 0.8777346014976501, |
| "learning_rate": 8.308515252276999e-05, |
| "loss": 2.1232, |
| "step": 68500 |
| }, |
| { |
| "epoch": 4.847548124209639, |
| "grad_norm": 0.8725584745407104, |
| "learning_rate": 8.294058117681077e-05, |
| "loss": 2.1161, |
| "step": 69000 |
| }, |
| { |
| "epoch": 4.8826752845299985, |
| "grad_norm": 0.8755514025688171, |
| "learning_rate": 8.279600983085153e-05, |
| "loss": 2.1139, |
| "step": 69500 |
| }, |
| { |
| "epoch": 4.917802444850358, |
| "grad_norm": 0.8769282698631287, |
| "learning_rate": 8.26514384848923e-05, |
| "loss": 2.1102, |
| "step": 70000 |
| }, |
| { |
| "epoch": 4.917802444850358, |
| "eval_accuracy": 0.6296170722465991, |
| "eval_loss": 1.802460789680481, |
| "eval_runtime": 122.2219, |
| "eval_samples_per_second": 987.597, |
| "eval_steps_per_second": 6.177, |
| "step": 70000 |
| }, |
| { |
| "epoch": 4.952929605170718, |
| "grad_norm": 0.8601032495498657, |
| "learning_rate": 8.250686713893306e-05, |
| "loss": 2.1074, |
| "step": 70500 |
| }, |
| { |
| "epoch": 4.988056765491078, |
| "grad_norm": 0.863791286945343, |
| "learning_rate": 8.236229579297384e-05, |
| "loss": 2.1005, |
| "step": 71000 |
| }, |
| { |
| "epoch": 5.023183925811438, |
| "grad_norm": 0.9040274024009705, |
| "learning_rate": 8.221772444701461e-05, |
| "loss": 2.0994, |
| "step": 71500 |
| }, |
| { |
| "epoch": 5.0583110861317975, |
| "grad_norm": 0.8910251259803772, |
| "learning_rate": 8.207315310105538e-05, |
| "loss": 2.0904, |
| "step": 72000 |
| }, |
| { |
| "epoch": 5.093438246452157, |
| "grad_norm": 0.8713977336883545, |
| "learning_rate": 8.192858175509614e-05, |
| "loss": 2.0859, |
| "step": 72500 |
| }, |
| { |
| "epoch": 5.128565406772516, |
| "grad_norm": 0.8776283860206604, |
| "learning_rate": 8.178401040913692e-05, |
| "loss": 2.0818, |
| "step": 73000 |
| }, |
| { |
| "epoch": 5.163692567092876, |
| "grad_norm": 0.8538689017295837, |
| "learning_rate": 8.163943906317768e-05, |
| "loss": 2.0799, |
| "step": 73500 |
| }, |
| { |
| "epoch": 5.198819727413236, |
| "grad_norm": 0.874956488609314, |
| "learning_rate": 8.149486771721845e-05, |
| "loss": 2.0746, |
| "step": 74000 |
| }, |
| { |
| "epoch": 5.2339468877335955, |
| "grad_norm": 0.8646124601364136, |
| "learning_rate": 8.135029637125921e-05, |
| "loss": 2.0755, |
| "step": 74500 |
| }, |
| { |
| "epoch": 5.269074048053955, |
| "grad_norm": 0.8849400281906128, |
| "learning_rate": 8.120572502529999e-05, |
| "loss": 2.071, |
| "step": 75000 |
| }, |
| { |
| "epoch": 5.304201208374315, |
| "grad_norm": 0.8349267244338989, |
| "learning_rate": 8.106115367934077e-05, |
| "loss": 2.068, |
| "step": 75500 |
| }, |
| { |
| "epoch": 5.339328368694675, |
| "grad_norm": 0.8874001502990723, |
| "learning_rate": 8.091658233338153e-05, |
| "loss": 2.0635, |
| "step": 76000 |
| }, |
| { |
| "epoch": 5.374455529015035, |
| "grad_norm": 0.8925509452819824, |
| "learning_rate": 8.07720109874223e-05, |
| "loss": 2.0602, |
| "step": 76500 |
| }, |
| { |
| "epoch": 5.409582689335394, |
| "grad_norm": 0.8840634226799011, |
| "learning_rate": 8.062743964146306e-05, |
| "loss": 2.0552, |
| "step": 77000 |
| }, |
| { |
| "epoch": 5.444709849655754, |
| "grad_norm": 0.8974944353103638, |
| "learning_rate": 8.048286829550384e-05, |
| "loss": 2.0499, |
| "step": 77500 |
| }, |
| { |
| "epoch": 5.479837009976113, |
| "grad_norm": 0.8761953115463257, |
| "learning_rate": 8.03382969495446e-05, |
| "loss": 2.0505, |
| "step": 78000 |
| }, |
| { |
| "epoch": 5.514964170296473, |
| "grad_norm": 0.8673203587532043, |
| "learning_rate": 8.019372560358536e-05, |
| "loss": 2.0474, |
| "step": 78500 |
| }, |
| { |
| "epoch": 5.550091330616833, |
| "grad_norm": 0.8759644627571106, |
| "learning_rate": 8.004915425762614e-05, |
| "loss": 2.0434, |
| "step": 79000 |
| }, |
| { |
| "epoch": 5.5852184909371925, |
| "grad_norm": 0.8770958185195923, |
| "learning_rate": 7.990458291166692e-05, |
| "loss": 2.0414, |
| "step": 79500 |
| }, |
| { |
| "epoch": 5.620345651257552, |
| "grad_norm": 0.8764468431472778, |
| "learning_rate": 7.976001156570768e-05, |
| "loss": 2.0378, |
| "step": 80000 |
| }, |
| { |
| "epoch": 5.620345651257552, |
| "eval_accuracy": 0.6393084966209296, |
| "eval_loss": 1.739489197731018, |
| "eval_runtime": 124.0472, |
| "eval_samples_per_second": 973.065, |
| "eval_steps_per_second": 6.086, |
| "step": 80000 |
| }, |
| { |
| "epoch": 5.655472811577912, |
| "grad_norm": 0.9069517254829407, |
| "learning_rate": 7.961544021974845e-05, |
| "loss": 2.0333, |
| "step": 80500 |
| }, |
| { |
| "epoch": 5.690599971898272, |
| "grad_norm": 0.8785437941551208, |
| "learning_rate": 7.947086887378921e-05, |
| "loss": 2.0389, |
| "step": 81000 |
| }, |
| { |
| "epoch": 5.725727132218632, |
| "grad_norm": 0.8835466504096985, |
| "learning_rate": 7.932629752782999e-05, |
| "loss": 2.0321, |
| "step": 81500 |
| }, |
| { |
| "epoch": 5.760854292538991, |
| "grad_norm": 0.915365993976593, |
| "learning_rate": 7.918172618187075e-05, |
| "loss": 2.0268, |
| "step": 82000 |
| }, |
| { |
| "epoch": 5.795981452859351, |
| "grad_norm": 0.9009836316108704, |
| "learning_rate": 7.903715483591153e-05, |
| "loss": 2.0269, |
| "step": 82500 |
| }, |
| { |
| "epoch": 5.83110861317971, |
| "grad_norm": 0.8733998537063599, |
| "learning_rate": 7.88925834899523e-05, |
| "loss": 2.0236, |
| "step": 83000 |
| }, |
| { |
| "epoch": 5.866235773500071, |
| "grad_norm": 0.8742640614509583, |
| "learning_rate": 7.874801214399307e-05, |
| "loss": 2.0209, |
| "step": 83500 |
| }, |
| { |
| "epoch": 5.90136293382043, |
| "grad_norm": 0.8777474761009216, |
| "learning_rate": 7.860344079803384e-05, |
| "loss": 2.0179, |
| "step": 84000 |
| }, |
| { |
| "epoch": 5.936490094140789, |
| "grad_norm": 0.8741117119789124, |
| "learning_rate": 7.84588694520746e-05, |
| "loss": 2.0146, |
| "step": 84500 |
| }, |
| { |
| "epoch": 5.971617254461149, |
| "grad_norm": 0.9120887517929077, |
| "learning_rate": 7.831429810611536e-05, |
| "loss": 2.0125, |
| "step": 85000 |
| }, |
| { |
| "epoch": 6.006744414781509, |
| "grad_norm": 0.9081856608390808, |
| "learning_rate": 7.816972676015614e-05, |
| "loss": 2.011, |
| "step": 85500 |
| }, |
| { |
| "epoch": 6.041871575101869, |
| "grad_norm": 0.8779035210609436, |
| "learning_rate": 7.802515541419692e-05, |
| "loss": 2.0044, |
| "step": 86000 |
| }, |
| { |
| "epoch": 6.076998735422229, |
| "grad_norm": 0.8693752884864807, |
| "learning_rate": 7.788058406823768e-05, |
| "loss": 2.0062, |
| "step": 86500 |
| }, |
| { |
| "epoch": 6.112125895742588, |
| "grad_norm": 0.8722647428512573, |
| "learning_rate": 7.773601272227845e-05, |
| "loss": 1.9981, |
| "step": 87000 |
| }, |
| { |
| "epoch": 6.147253056062948, |
| "grad_norm": 0.9012197256088257, |
| "learning_rate": 7.759144137631922e-05, |
| "loss": 1.9965, |
| "step": 87500 |
| }, |
| { |
| "epoch": 6.182380216383308, |
| "grad_norm": 0.8717825412750244, |
| "learning_rate": 7.744687003035999e-05, |
| "loss": 1.9962, |
| "step": 88000 |
| }, |
| { |
| "epoch": 6.217507376703667, |
| "grad_norm": 0.8815474510192871, |
| "learning_rate": 7.730229868440075e-05, |
| "loss": 1.9943, |
| "step": 88500 |
| }, |
| { |
| "epoch": 6.252634537024027, |
| "grad_norm": 0.8928734660148621, |
| "learning_rate": 7.715772733844152e-05, |
| "loss": 1.9897, |
| "step": 89000 |
| }, |
| { |
| "epoch": 6.287761697344386, |
| "grad_norm": 0.9025377631187439, |
| "learning_rate": 7.70131559924823e-05, |
| "loss": 1.9892, |
| "step": 89500 |
| }, |
| { |
| "epoch": 6.322888857664746, |
| "grad_norm": 0.8946100473403931, |
| "learning_rate": 7.686858464652307e-05, |
| "loss": 1.9881, |
| "step": 90000 |
| }, |
| { |
| "epoch": 6.322888857664746, |
| "eval_accuracy": 0.646682600908374, |
| "eval_loss": 1.6945347785949707, |
| "eval_runtime": 122.5009, |
| "eval_samples_per_second": 985.348, |
| "eval_steps_per_second": 6.163, |
| "step": 90000 |
| }, |
| { |
| "epoch": 6.358016017985106, |
| "grad_norm": 0.8891465663909912, |
| "learning_rate": 7.672401330056384e-05, |
| "loss": 1.9863, |
| "step": 90500 |
| }, |
| { |
| "epoch": 6.393143178305466, |
| "grad_norm": 0.8933587074279785, |
| "learning_rate": 7.65794419546046e-05, |
| "loss": 1.9834, |
| "step": 91000 |
| }, |
| { |
| "epoch": 6.4282703386258255, |
| "grad_norm": 0.8788536787033081, |
| "learning_rate": 7.643487060864536e-05, |
| "loss": 1.98, |
| "step": 91500 |
| }, |
| { |
| "epoch": 6.463397498946185, |
| "grad_norm": 0.9046235084533691, |
| "learning_rate": 7.629029926268614e-05, |
| "loss": 1.979, |
| "step": 92000 |
| }, |
| { |
| "epoch": 6.498524659266545, |
| "grad_norm": 0.8782976269721985, |
| "learning_rate": 7.61457279167269e-05, |
| "loss": 1.9777, |
| "step": 92500 |
| }, |
| { |
| "epoch": 6.533651819586905, |
| "grad_norm": 0.8855561017990112, |
| "learning_rate": 7.600115657076767e-05, |
| "loss": 1.9775, |
| "step": 93000 |
| }, |
| { |
| "epoch": 6.568778979907265, |
| "grad_norm": 0.9085721373558044, |
| "learning_rate": 7.585658522480845e-05, |
| "loss": 1.9734, |
| "step": 93500 |
| }, |
| { |
| "epoch": 6.603906140227624, |
| "grad_norm": 0.8760172128677368, |
| "learning_rate": 7.571201387884922e-05, |
| "loss": 1.9699, |
| "step": 94000 |
| }, |
| { |
| "epoch": 6.639033300547983, |
| "grad_norm": 0.8842225670814514, |
| "learning_rate": 7.556744253288999e-05, |
| "loss": 1.9678, |
| "step": 94500 |
| }, |
| { |
| "epoch": 6.674160460868343, |
| "grad_norm": 0.8828365802764893, |
| "learning_rate": 7.542287118693075e-05, |
| "loss": 1.9653, |
| "step": 95000 |
| }, |
| { |
| "epoch": 6.709287621188703, |
| "grad_norm": 0.8868061900138855, |
| "learning_rate": 7.527829984097152e-05, |
| "loss": 1.9639, |
| "step": 95500 |
| }, |
| { |
| "epoch": 6.744414781509063, |
| "grad_norm": 0.8875888586044312, |
| "learning_rate": 7.51337284950123e-05, |
| "loss": 1.9688, |
| "step": 96000 |
| }, |
| { |
| "epoch": 6.7795419418294225, |
| "grad_norm": 0.8854140043258667, |
| "learning_rate": 7.498915714905306e-05, |
| "loss": 1.9596, |
| "step": 96500 |
| }, |
| { |
| "epoch": 6.814669102149782, |
| "grad_norm": 0.9050253033638, |
| "learning_rate": 7.484458580309382e-05, |
| "loss": 1.9608, |
| "step": 97000 |
| }, |
| { |
| "epoch": 6.849796262470142, |
| "grad_norm": 0.8908172845840454, |
| "learning_rate": 7.47000144571346e-05, |
| "loss": 1.9533, |
| "step": 97500 |
| }, |
| { |
| "epoch": 6.884923422790502, |
| "grad_norm": 0.8731047511100769, |
| "learning_rate": 7.455544311117538e-05, |
| "loss": 1.9546, |
| "step": 98000 |
| }, |
| { |
| "epoch": 6.920050583110862, |
| "grad_norm": 0.8872097730636597, |
| "learning_rate": 7.441087176521614e-05, |
| "loss": 1.9516, |
| "step": 98500 |
| }, |
| { |
| "epoch": 6.955177743431221, |
| "grad_norm": 0.8854101300239563, |
| "learning_rate": 7.42663004192569e-05, |
| "loss": 1.9514, |
| "step": 99000 |
| }, |
| { |
| "epoch": 6.99030490375158, |
| "grad_norm": 0.8745369911193848, |
| "learning_rate": 7.412172907329767e-05, |
| "loss": 1.9463, |
| "step": 99500 |
| }, |
| { |
| "epoch": 7.02543206407194, |
| "grad_norm": 0.8900884985923767, |
| "learning_rate": 7.397715772733845e-05, |
| "loss": 1.9469, |
| "step": 100000 |
| }, |
| { |
| "epoch": 7.02543206407194, |
| "eval_accuracy": 0.6525522943801697, |
| "eval_loss": 1.6552181243896484, |
| "eval_runtime": 123.3249, |
| "eval_samples_per_second": 978.765, |
| "eval_steps_per_second": 6.122, |
| "step": 100000 |
| }, |
| { |
| "epoch": 7.0605592243923, |
| "grad_norm": 0.8851434588432312, |
| "learning_rate": 7.383258638137921e-05, |
| "loss": 1.9448, |
| "step": 100500 |
| }, |
| { |
| "epoch": 7.09568638471266, |
| "grad_norm": 0.9002060294151306, |
| "learning_rate": 7.368801503541999e-05, |
| "loss": 1.9446, |
| "step": 101000 |
| }, |
| { |
| "epoch": 7.1308135450330195, |
| "grad_norm": 0.9126923680305481, |
| "learning_rate": 7.354344368946075e-05, |
| "loss": 1.9427, |
| "step": 101500 |
| }, |
| { |
| "epoch": 7.165940705353379, |
| "grad_norm": 0.9131616353988647, |
| "learning_rate": 7.339887234350153e-05, |
| "loss": 1.9397, |
| "step": 102000 |
| }, |
| { |
| "epoch": 7.201067865673739, |
| "grad_norm": 0.8775026798248291, |
| "learning_rate": 7.32543009975423e-05, |
| "loss": 1.9386, |
| "step": 102500 |
| }, |
| { |
| "epoch": 7.236195025994099, |
| "grad_norm": 0.8769180774688721, |
| "learning_rate": 7.310972965158306e-05, |
| "loss": 1.9359, |
| "step": 103000 |
| }, |
| { |
| "epoch": 7.271322186314459, |
| "grad_norm": 0.885008692741394, |
| "learning_rate": 7.296515830562382e-05, |
| "loss": 1.9365, |
| "step": 103500 |
| }, |
| { |
| "epoch": 7.306449346634818, |
| "grad_norm": 0.9002963304519653, |
| "learning_rate": 7.28205869596646e-05, |
| "loss": 1.9316, |
| "step": 104000 |
| }, |
| { |
| "epoch": 7.341576506955178, |
| "grad_norm": 0.8847761750221252, |
| "learning_rate": 7.267601561370536e-05, |
| "loss": 1.9269, |
| "step": 104500 |
| }, |
| { |
| "epoch": 7.376703667275537, |
| "grad_norm": 0.8742501139640808, |
| "learning_rate": 7.253144426774614e-05, |
| "loss": 1.93, |
| "step": 105000 |
| }, |
| { |
| "epoch": 7.411830827595897, |
| "grad_norm": 0.88679039478302, |
| "learning_rate": 7.23868729217869e-05, |
| "loss": 1.9328, |
| "step": 105500 |
| }, |
| { |
| "epoch": 7.446957987916257, |
| "grad_norm": 0.880789041519165, |
| "learning_rate": 7.224230157582768e-05, |
| "loss": 1.9313, |
| "step": 106000 |
| }, |
| { |
| "epoch": 7.482085148236616, |
| "grad_norm": 0.8551275134086609, |
| "learning_rate": 7.209773022986845e-05, |
| "loss": 1.9268, |
| "step": 106500 |
| }, |
| { |
| "epoch": 7.517212308556976, |
| "grad_norm": 0.8933087587356567, |
| "learning_rate": 7.195315888390921e-05, |
| "loss": 1.9224, |
| "step": 107000 |
| }, |
| { |
| "epoch": 7.552339468877336, |
| "grad_norm": 0.888533890247345, |
| "learning_rate": 7.180858753794997e-05, |
| "loss": 1.9216, |
| "step": 107500 |
| }, |
| { |
| "epoch": 7.587466629197696, |
| "grad_norm": 0.8733634352684021, |
| "learning_rate": 7.166401619199075e-05, |
| "loss": 1.9213, |
| "step": 108000 |
| }, |
| { |
| "epoch": 7.6225937895180556, |
| "grad_norm": 0.9028536081314087, |
| "learning_rate": 7.151944484603153e-05, |
| "loss": 1.9236, |
| "step": 108500 |
| }, |
| { |
| "epoch": 7.657720949838415, |
| "grad_norm": 0.8724736571311951, |
| "learning_rate": 7.13748735000723e-05, |
| "loss": 1.9183, |
| "step": 109000 |
| }, |
| { |
| "epoch": 7.692848110158775, |
| "grad_norm": 0.8967872858047485, |
| "learning_rate": 7.123030215411306e-05, |
| "loss": 1.9182, |
| "step": 109500 |
| }, |
| { |
| "epoch": 7.727975270479135, |
| "grad_norm": 0.9060919284820557, |
| "learning_rate": 7.108573080815382e-05, |
| "loss": 1.919, |
| "step": 110000 |
| }, |
| { |
| "epoch": 7.727975270479135, |
| "eval_accuracy": 0.6572336073191586, |
| "eval_loss": 1.6279349327087402, |
| "eval_runtime": 122.761, |
| "eval_samples_per_second": 983.26, |
| "eval_steps_per_second": 6.15, |
| "step": 110000 |
| }, |
| { |
| "epoch": 7.763102430799494, |
| "grad_norm": 0.8861146569252014, |
| "learning_rate": 7.09411594621946e-05, |
| "loss": 1.9155, |
| "step": 110500 |
| }, |
| { |
| "epoch": 7.798229591119854, |
| "grad_norm": 0.8727577924728394, |
| "learning_rate": 7.079658811623536e-05, |
| "loss": 1.9128, |
| "step": 111000 |
| }, |
| { |
| "epoch": 7.833356751440213, |
| "grad_norm": 0.8794012665748596, |
| "learning_rate": 7.065201677027613e-05, |
| "loss": 1.9115, |
| "step": 111500 |
| }, |
| { |
| "epoch": 7.868483911760573, |
| "grad_norm": 0.9028168320655823, |
| "learning_rate": 7.05074454243169e-05, |
| "loss": 1.9126, |
| "step": 112000 |
| }, |
| { |
| "epoch": 7.903611072080933, |
| "grad_norm": 0.8709144592285156, |
| "learning_rate": 7.036287407835768e-05, |
| "loss": 1.911, |
| "step": 112500 |
| }, |
| { |
| "epoch": 7.938738232401293, |
| "grad_norm": 0.9144461154937744, |
| "learning_rate": 7.021830273239845e-05, |
| "loss": 1.9042, |
| "step": 113000 |
| }, |
| { |
| "epoch": 7.9738653927216525, |
| "grad_norm": 0.8901553153991699, |
| "learning_rate": 7.007373138643921e-05, |
| "loss": 1.9046, |
| "step": 113500 |
| }, |
| { |
| "epoch": 8.008992553042011, |
| "grad_norm": 0.8806023001670837, |
| "learning_rate": 6.992916004047997e-05, |
| "loss": 1.9069, |
| "step": 114000 |
| }, |
| { |
| "epoch": 8.044119713362372, |
| "grad_norm": 0.920773983001709, |
| "learning_rate": 6.978458869452075e-05, |
| "loss": 1.9015, |
| "step": 114500 |
| }, |
| { |
| "epoch": 8.079246873682731, |
| "grad_norm": 0.9078338146209717, |
| "learning_rate": 6.964001734856152e-05, |
| "loss": 1.9, |
| "step": 115000 |
| }, |
| { |
| "epoch": 8.114374034003092, |
| "grad_norm": 0.8935852646827698, |
| "learning_rate": 6.949544600260228e-05, |
| "loss": 1.8989, |
| "step": 115500 |
| }, |
| { |
| "epoch": 8.14950119432345, |
| "grad_norm": 0.888476550579071, |
| "learning_rate": 6.935087465664306e-05, |
| "loss": 1.8996, |
| "step": 116000 |
| }, |
| { |
| "epoch": 8.184628354643811, |
| "grad_norm": 0.8867985606193542, |
| "learning_rate": 6.920630331068383e-05, |
| "loss": 1.8995, |
| "step": 116500 |
| }, |
| { |
| "epoch": 8.21975551496417, |
| "grad_norm": 0.8914681673049927, |
| "learning_rate": 6.90617319647246e-05, |
| "loss": 1.8953, |
| "step": 117000 |
| }, |
| { |
| "epoch": 8.25488267528453, |
| "grad_norm": 0.9003250598907471, |
| "learning_rate": 6.891716061876536e-05, |
| "loss": 1.8952, |
| "step": 117500 |
| }, |
| { |
| "epoch": 8.29000983560489, |
| "grad_norm": 0.8978532552719116, |
| "learning_rate": 6.877258927280613e-05, |
| "loss": 1.8908, |
| "step": 118000 |
| }, |
| { |
| "epoch": 8.325136995925249, |
| "grad_norm": 0.9211486577987671, |
| "learning_rate": 6.86280179268469e-05, |
| "loss": 1.89, |
| "step": 118500 |
| }, |
| { |
| "epoch": 8.36026415624561, |
| "grad_norm": 0.9234055876731873, |
| "learning_rate": 6.848344658088767e-05, |
| "loss": 1.8906, |
| "step": 119000 |
| }, |
| { |
| "epoch": 8.395391316565968, |
| "grad_norm": 0.8973139524459839, |
| "learning_rate": 6.833887523492845e-05, |
| "loss": 1.8902, |
| "step": 119500 |
| }, |
| { |
| "epoch": 8.430518476886329, |
| "grad_norm": 0.8916562795639038, |
| "learning_rate": 6.819430388896921e-05, |
| "loss": 1.8899, |
| "step": 120000 |
| }, |
| { |
| "epoch": 8.430518476886329, |
| "eval_accuracy": 0.6613473610030635, |
| "eval_loss": 1.601199746131897, |
| "eval_runtime": 122.9016, |
| "eval_samples_per_second": 982.135, |
| "eval_steps_per_second": 6.143, |
| "step": 120000 |
| }, |
| { |
| "epoch": 8.465645637206688, |
| "grad_norm": 0.8888087868690491, |
| "learning_rate": 6.804973254300999e-05, |
| "loss": 1.8891, |
| "step": 120500 |
| }, |
| { |
| "epoch": 8.500772797527048, |
| "grad_norm": 0.9059433937072754, |
| "learning_rate": 6.790516119705075e-05, |
| "loss": 1.8897, |
| "step": 121000 |
| }, |
| { |
| "epoch": 8.535899957847407, |
| "grad_norm": 0.8812214732170105, |
| "learning_rate": 6.776058985109152e-05, |
| "loss": 1.8894, |
| "step": 121500 |
| }, |
| { |
| "epoch": 8.571027118167768, |
| "grad_norm": 0.8938203454017639, |
| "learning_rate": 6.761601850513228e-05, |
| "loss": 1.885, |
| "step": 122000 |
| }, |
| { |
| "epoch": 8.606154278488127, |
| "grad_norm": 0.9012354612350464, |
| "learning_rate": 6.747144715917306e-05, |
| "loss": 1.8836, |
| "step": 122500 |
| }, |
| { |
| "epoch": 8.641281438808488, |
| "grad_norm": 0.8827559351921082, |
| "learning_rate": 6.732687581321382e-05, |
| "loss": 1.8828, |
| "step": 123000 |
| }, |
| { |
| "epoch": 8.676408599128846, |
| "grad_norm": 0.9143077731132507, |
| "learning_rate": 6.71823044672546e-05, |
| "loss": 1.8775, |
| "step": 123500 |
| }, |
| { |
| "epoch": 8.711535759449205, |
| "grad_norm": 0.8918534517288208, |
| "learning_rate": 6.703773312129536e-05, |
| "loss": 1.8766, |
| "step": 124000 |
| }, |
| { |
| "epoch": 8.746662919769566, |
| "grad_norm": 0.8915277719497681, |
| "learning_rate": 6.689316177533613e-05, |
| "loss": 1.8719, |
| "step": 124500 |
| }, |
| { |
| "epoch": 8.781790080089925, |
| "grad_norm": 0.8825547099113464, |
| "learning_rate": 6.67485904293769e-05, |
| "loss": 1.8773, |
| "step": 125000 |
| }, |
| { |
| "epoch": 8.816917240410286, |
| "grad_norm": 0.9037731885910034, |
| "learning_rate": 6.660401908341767e-05, |
| "loss": 1.8752, |
| "step": 125500 |
| }, |
| { |
| "epoch": 8.852044400730644, |
| "grad_norm": 0.9040917754173279, |
| "learning_rate": 6.645944773745843e-05, |
| "loss": 1.8747, |
| "step": 126000 |
| }, |
| { |
| "epoch": 8.887171561051005, |
| "grad_norm": 0.920987069606781, |
| "learning_rate": 6.631487639149921e-05, |
| "loss": 1.8741, |
| "step": 126500 |
| }, |
| { |
| "epoch": 8.922298721371364, |
| "grad_norm": 0.9317889213562012, |
| "learning_rate": 6.617030504553999e-05, |
| "loss": 1.8728, |
| "step": 127000 |
| }, |
| { |
| "epoch": 8.957425881691725, |
| "grad_norm": 0.9033764004707336, |
| "learning_rate": 6.602573369958075e-05, |
| "loss": 1.8769, |
| "step": 127500 |
| }, |
| { |
| "epoch": 8.992553042012084, |
| "grad_norm": 0.9001402854919434, |
| "learning_rate": 6.588116235362152e-05, |
| "loss": 1.8756, |
| "step": 128000 |
| }, |
| { |
| "epoch": 9.027680202332444, |
| "grad_norm": 0.9200482964515686, |
| "learning_rate": 6.573659100766228e-05, |
| "loss": 1.8686, |
| "step": 128500 |
| }, |
| { |
| "epoch": 9.062807362652803, |
| "grad_norm": 0.8894959688186646, |
| "learning_rate": 6.559201966170306e-05, |
| "loss": 1.8672, |
| "step": 129000 |
| }, |
| { |
| "epoch": 9.097934522973162, |
| "grad_norm": 0.9352574348449707, |
| "learning_rate": 6.544744831574382e-05, |
| "loss": 1.8681, |
| "step": 129500 |
| }, |
| { |
| "epoch": 9.133061683293523, |
| "grad_norm": 0.8903130888938904, |
| "learning_rate": 6.530287696978458e-05, |
| "loss": 1.8634, |
| "step": 130000 |
| }, |
| { |
| "epoch": 9.133061683293523, |
| "eval_accuracy": 0.6640824195085319, |
| "eval_loss": 1.584887981414795, |
| "eval_runtime": 122.8969, |
| "eval_samples_per_second": 982.173, |
| "eval_steps_per_second": 6.143, |
| "step": 130000 |
| }, |
| { |
| "epoch": 9.168188843613882, |
| "grad_norm": 0.8989660143852234, |
| "learning_rate": 6.515830562382536e-05, |
| "loss": 1.8655, |
| "step": 130500 |
| }, |
| { |
| "epoch": 9.203316003934242, |
| "grad_norm": 0.9208776354789734, |
| "learning_rate": 6.501373427786614e-05, |
| "loss": 1.8666, |
| "step": 131000 |
| }, |
| { |
| "epoch": 9.238443164254601, |
| "grad_norm": 0.9106900691986084, |
| "learning_rate": 6.48691629319069e-05, |
| "loss": 1.8648, |
| "step": 131500 |
| }, |
| { |
| "epoch": 9.273570324574962, |
| "grad_norm": 0.8650423884391785, |
| "learning_rate": 6.472459158594767e-05, |
| "loss": 1.8663, |
| "step": 132000 |
| }, |
| { |
| "epoch": 9.30869748489532, |
| "grad_norm": 0.9090404510498047, |
| "learning_rate": 6.458002023998843e-05, |
| "loss": 1.8636, |
| "step": 132500 |
| }, |
| { |
| "epoch": 9.343824645215681, |
| "grad_norm": 0.9086259007453918, |
| "learning_rate": 6.443544889402921e-05, |
| "loss": 1.8607, |
| "step": 133000 |
| }, |
| { |
| "epoch": 9.37895180553604, |
| "grad_norm": 0.9031954407691956, |
| "learning_rate": 6.429087754806997e-05, |
| "loss": 1.8605, |
| "step": 133500 |
| }, |
| { |
| "epoch": 9.414078965856401, |
| "grad_norm": 0.9025021195411682, |
| "learning_rate": 6.414630620211074e-05, |
| "loss": 1.859, |
| "step": 134000 |
| }, |
| { |
| "epoch": 9.44920612617676, |
| "grad_norm": 0.9351354241371155, |
| "learning_rate": 6.400173485615151e-05, |
| "loss": 1.8559, |
| "step": 134500 |
| }, |
| { |
| "epoch": 9.484333286497119, |
| "grad_norm": 0.9082109332084656, |
| "learning_rate": 6.385716351019229e-05, |
| "loss": 1.8552, |
| "step": 135000 |
| }, |
| { |
| "epoch": 9.51946044681748, |
| "grad_norm": 0.8957135677337646, |
| "learning_rate": 6.371259216423306e-05, |
| "loss": 1.8559, |
| "step": 135500 |
| }, |
| { |
| "epoch": 9.554587607137838, |
| "grad_norm": 0.9105542898178101, |
| "learning_rate": 6.356802081827382e-05, |
| "loss": 1.8555, |
| "step": 136000 |
| }, |
| { |
| "epoch": 9.589714767458199, |
| "grad_norm": 0.9041818380355835, |
| "learning_rate": 6.342344947231458e-05, |
| "loss": 1.8562, |
| "step": 136500 |
| }, |
| { |
| "epoch": 9.624841927778558, |
| "grad_norm": 0.90251225233078, |
| "learning_rate": 6.327887812635536e-05, |
| "loss": 1.8522, |
| "step": 137000 |
| }, |
| { |
| "epoch": 9.659969088098919, |
| "grad_norm": 0.9044151306152344, |
| "learning_rate": 6.313430678039613e-05, |
| "loss": 1.8533, |
| "step": 137500 |
| }, |
| { |
| "epoch": 9.695096248419278, |
| "grad_norm": 0.8829325437545776, |
| "learning_rate": 6.29897354344369e-05, |
| "loss": 1.848, |
| "step": 138000 |
| }, |
| { |
| "epoch": 9.730223408739638, |
| "grad_norm": 0.8989315629005432, |
| "learning_rate": 6.284516408847767e-05, |
| "loss": 1.8548, |
| "step": 138500 |
| }, |
| { |
| "epoch": 9.765350569059997, |
| "grad_norm": 0.890407383441925, |
| "learning_rate": 6.270059274251843e-05, |
| "loss": 1.8493, |
| "step": 139000 |
| }, |
| { |
| "epoch": 9.800477729380358, |
| "grad_norm": 0.8821042776107788, |
| "learning_rate": 6.255602139655921e-05, |
| "loss": 1.8493, |
| "step": 139500 |
| }, |
| { |
| "epoch": 9.835604889700717, |
| "grad_norm": 0.8859909772872925, |
| "learning_rate": 6.241145005059997e-05, |
| "loss": 1.8481, |
| "step": 140000 |
| }, |
| { |
| "epoch": 9.835604889700717, |
| "eval_accuracy": 0.6679107399542193, |
| "eval_loss": 1.5630303621292114, |
| "eval_runtime": 122.8436, |
| "eval_samples_per_second": 982.599, |
| "eval_steps_per_second": 6.146, |
| "step": 140000 |
| }, |
| { |
| "epoch": 9.870732050021076, |
| "grad_norm": 0.9338480234146118, |
| "learning_rate": 6.226687870464074e-05, |
| "loss": 1.8479, |
| "step": 140500 |
| }, |
| { |
| "epoch": 9.905859210341436, |
| "grad_norm": 0.9007723927497864, |
| "learning_rate": 6.212230735868151e-05, |
| "loss": 1.8458, |
| "step": 141000 |
| }, |
| { |
| "epoch": 9.940986370661795, |
| "grad_norm": 0.9068960547447205, |
| "learning_rate": 6.197773601272228e-05, |
| "loss": 1.846, |
| "step": 141500 |
| }, |
| { |
| "epoch": 9.976113530982156, |
| "grad_norm": 0.8967083692550659, |
| "learning_rate": 6.183316466676306e-05, |
| "loss": 1.8429, |
| "step": 142000 |
| }, |
| { |
| "epoch": 10.011240691302515, |
| "grad_norm": 0.9090276956558228, |
| "learning_rate": 6.168859332080382e-05, |
| "loss": 1.8448, |
| "step": 142500 |
| }, |
| { |
| "epoch": 10.046367851622875, |
| "grad_norm": 0.8974506855010986, |
| "learning_rate": 6.154402197484458e-05, |
| "loss": 1.8444, |
| "step": 143000 |
| }, |
| { |
| "epoch": 10.081495011943234, |
| "grad_norm": 0.897613525390625, |
| "learning_rate": 6.139945062888536e-05, |
| "loss": 1.8399, |
| "step": 143500 |
| }, |
| { |
| "epoch": 10.116622172263595, |
| "grad_norm": 0.920937716960907, |
| "learning_rate": 6.125487928292613e-05, |
| "loss": 1.8383, |
| "step": 144000 |
| }, |
| { |
| "epoch": 10.151749332583954, |
| "grad_norm": 0.9360824823379517, |
| "learning_rate": 6.111030793696689e-05, |
| "loss": 1.8393, |
| "step": 144500 |
| }, |
| { |
| "epoch": 10.186876492904315, |
| "grad_norm": 0.9206079244613647, |
| "learning_rate": 6.096573659100766e-05, |
| "loss": 1.8391, |
| "step": 145000 |
| }, |
| { |
| "epoch": 10.222003653224673, |
| "grad_norm": 0.9022341966629028, |
| "learning_rate": 6.082116524504844e-05, |
| "loss": 1.8408, |
| "step": 145500 |
| }, |
| { |
| "epoch": 10.257130813545032, |
| "grad_norm": 0.8904668688774109, |
| "learning_rate": 6.067659389908921e-05, |
| "loss": 1.8405, |
| "step": 146000 |
| }, |
| { |
| "epoch": 10.292257973865393, |
| "grad_norm": 0.8963555097579956, |
| "learning_rate": 6.053202255312997e-05, |
| "loss": 1.8337, |
| "step": 146500 |
| }, |
| { |
| "epoch": 10.327385134185752, |
| "grad_norm": 0.9340492486953735, |
| "learning_rate": 6.0387451207170744e-05, |
| "loss": 1.836, |
| "step": 147000 |
| }, |
| { |
| "epoch": 10.362512294506113, |
| "grad_norm": 0.9032008647918701, |
| "learning_rate": 6.024287986121151e-05, |
| "loss": 1.8349, |
| "step": 147500 |
| }, |
| { |
| "epoch": 10.397639454826471, |
| "grad_norm": 0.9033811092376709, |
| "learning_rate": 6.009830851525228e-05, |
| "loss": 1.8325, |
| "step": 148000 |
| }, |
| { |
| "epoch": 10.432766615146832, |
| "grad_norm": 0.9075065851211548, |
| "learning_rate": 5.995373716929304e-05, |
| "loss": 1.8361, |
| "step": 148500 |
| }, |
| { |
| "epoch": 10.467893775467191, |
| "grad_norm": 0.9036703705787659, |
| "learning_rate": 5.980916582333381e-05, |
| "loss": 1.8312, |
| "step": 149000 |
| }, |
| { |
| "epoch": 10.503020935787552, |
| "grad_norm": 0.8876055479049683, |
| "learning_rate": 5.966459447737459e-05, |
| "loss": 1.8295, |
| "step": 149500 |
| }, |
| { |
| "epoch": 10.53814809610791, |
| "grad_norm": 0.9093754291534424, |
| "learning_rate": 5.952002313141536e-05, |
| "loss": 1.8315, |
| "step": 150000 |
| }, |
| { |
| "epoch": 10.53814809610791, |
| "eval_accuracy": 0.6703074092205885, |
| "eval_loss": 1.5475610494613647, |
| "eval_runtime": 122.5506, |
| "eval_samples_per_second": 984.949, |
| "eval_steps_per_second": 6.161, |
| "step": 150000 |
| }, |
| { |
| "epoch": 10.573275256428271, |
| "grad_norm": 0.9191955924034119, |
| "learning_rate": 5.9375451785456126e-05, |
| "loss": 1.8343, |
| "step": 150500 |
| }, |
| { |
| "epoch": 10.60840241674863, |
| "grad_norm": 0.9078773856163025, |
| "learning_rate": 5.9230880439496896e-05, |
| "loss": 1.828, |
| "step": 151000 |
| }, |
| { |
| "epoch": 10.643529577068989, |
| "grad_norm": 0.918960452079773, |
| "learning_rate": 5.908630909353766e-05, |
| "loss": 1.8291, |
| "step": 151500 |
| }, |
| { |
| "epoch": 10.67865673738935, |
| "grad_norm": 0.8893048167228699, |
| "learning_rate": 5.894173774757843e-05, |
| "loss": 1.8289, |
| "step": 152000 |
| }, |
| { |
| "epoch": 10.713783897709709, |
| "grad_norm": 0.8991037011146545, |
| "learning_rate": 5.8797166401619195e-05, |
| "loss": 1.8263, |
| "step": 152500 |
| }, |
| { |
| "epoch": 10.74891105803007, |
| "grad_norm": 0.9058040380477905, |
| "learning_rate": 5.865259505565997e-05, |
| "loss": 1.8264, |
| "step": 153000 |
| }, |
| { |
| "epoch": 10.784038218350428, |
| "grad_norm": 0.9202247262001038, |
| "learning_rate": 5.8508023709700743e-05, |
| "loss": 1.8289, |
| "step": 153500 |
| }, |
| { |
| "epoch": 10.819165378670789, |
| "grad_norm": 0.8929277658462524, |
| "learning_rate": 5.8363452363741514e-05, |
| "loss": 1.8266, |
| "step": 154000 |
| }, |
| { |
| "epoch": 10.854292538991148, |
| "grad_norm": 0.9024432301521301, |
| "learning_rate": 5.821888101778228e-05, |
| "loss": 1.8263, |
| "step": 154500 |
| }, |
| { |
| "epoch": 10.889419699311508, |
| "grad_norm": 0.8983497023582458, |
| "learning_rate": 5.807430967182305e-05, |
| "loss": 1.8256, |
| "step": 155000 |
| }, |
| { |
| "epoch": 10.924546859631867, |
| "grad_norm": 0.9260616898536682, |
| "learning_rate": 5.792973832586381e-05, |
| "loss": 1.8262, |
| "step": 155500 |
| }, |
| { |
| "epoch": 10.959674019952226, |
| "grad_norm": 0.9316391944885254, |
| "learning_rate": 5.7785166979904584e-05, |
| "loss": 1.8239, |
| "step": 156000 |
| }, |
| { |
| "epoch": 10.994801180272587, |
| "grad_norm": 0.9033026695251465, |
| "learning_rate": 5.764059563394536e-05, |
| "loss": 1.8232, |
| "step": 156500 |
| }, |
| { |
| "epoch": 11.029928340592946, |
| "grad_norm": 0.933590292930603, |
| "learning_rate": 5.7496024287986125e-05, |
| "loss": 1.8167, |
| "step": 157000 |
| }, |
| { |
| "epoch": 11.065055500913306, |
| "grad_norm": 0.8982483148574829, |
| "learning_rate": 5.7351452942026896e-05, |
| "loss": 1.8205, |
| "step": 157500 |
| }, |
| { |
| "epoch": 11.100182661233665, |
| "grad_norm": 0.9130824208259583, |
| "learning_rate": 5.720688159606766e-05, |
| "loss": 1.8198, |
| "step": 158000 |
| }, |
| { |
| "epoch": 11.135309821554026, |
| "grad_norm": 0.9298592805862427, |
| "learning_rate": 5.706231025010843e-05, |
| "loss": 1.8166, |
| "step": 158500 |
| }, |
| { |
| "epoch": 11.170436981874385, |
| "grad_norm": 0.910667896270752, |
| "learning_rate": 5.6917738904149195e-05, |
| "loss": 1.8184, |
| "step": 159000 |
| }, |
| { |
| "epoch": 11.205564142194746, |
| "grad_norm": 0.9171704053878784, |
| "learning_rate": 5.6773167558189966e-05, |
| "loss": 1.8191, |
| "step": 159500 |
| }, |
| { |
| "epoch": 11.240691302515105, |
| "grad_norm": 0.9253237247467041, |
| "learning_rate": 5.6628596212230736e-05, |
| "loss": 1.8209, |
| "step": 160000 |
| }, |
| { |
| "epoch": 11.240691302515105, |
| "eval_accuracy": 0.6716161989137798, |
| "eval_loss": 1.5394113063812256, |
| "eval_runtime": 122.8945, |
| "eval_samples_per_second": 982.192, |
| "eval_steps_per_second": 6.143, |
| "step": 160000 |
| }, |
| { |
| "epoch": 11.275818462835465, |
| "grad_norm": 0.9159039855003357, |
| "learning_rate": 5.6484024866271514e-05, |
| "loss": 1.8151, |
| "step": 160500 |
| }, |
| { |
| "epoch": 11.310945623155824, |
| "grad_norm": 0.9057639837265015, |
| "learning_rate": 5.633945352031228e-05, |
| "loss": 1.8194, |
| "step": 161000 |
| }, |
| { |
| "epoch": 11.346072783476183, |
| "grad_norm": 0.91380375623703, |
| "learning_rate": 5.619488217435305e-05, |
| "loss": 1.8154, |
| "step": 161500 |
| }, |
| { |
| "epoch": 11.381199943796544, |
| "grad_norm": 0.9542326927185059, |
| "learning_rate": 5.605031082839381e-05, |
| "loss": 1.8151, |
| "step": 162000 |
| }, |
| { |
| "epoch": 11.416327104116903, |
| "grad_norm": 0.9179979562759399, |
| "learning_rate": 5.5905739482434584e-05, |
| "loss": 1.8128, |
| "step": 162500 |
| }, |
| { |
| "epoch": 11.451454264437263, |
| "grad_norm": 0.8947669863700867, |
| "learning_rate": 5.576116813647535e-05, |
| "loss": 1.8084, |
| "step": 163000 |
| }, |
| { |
| "epoch": 11.486581424757622, |
| "grad_norm": 0.9551398754119873, |
| "learning_rate": 5.561659679051612e-05, |
| "loss": 1.8121, |
| "step": 163500 |
| }, |
| { |
| "epoch": 11.521708585077983, |
| "grad_norm": 0.9089252352714539, |
| "learning_rate": 5.5472025444556896e-05, |
| "loss": 1.8126, |
| "step": 164000 |
| }, |
| { |
| "epoch": 11.556835745398342, |
| "grad_norm": 0.9001522064208984, |
| "learning_rate": 5.532745409859767e-05, |
| "loss": 1.81, |
| "step": 164500 |
| }, |
| { |
| "epoch": 11.591962905718702, |
| "grad_norm": 0.9125813841819763, |
| "learning_rate": 5.518288275263843e-05, |
| "loss": 1.8109, |
| "step": 165000 |
| }, |
| { |
| "epoch": 11.627090066039061, |
| "grad_norm": 0.9127496480941772, |
| "learning_rate": 5.50383114066792e-05, |
| "loss": 1.8107, |
| "step": 165500 |
| }, |
| { |
| "epoch": 11.662217226359422, |
| "grad_norm": 0.9150344133377075, |
| "learning_rate": 5.4893740060719965e-05, |
| "loss": 1.8074, |
| "step": 166000 |
| }, |
| { |
| "epoch": 11.69734438667978, |
| "grad_norm": 0.9088866710662842, |
| "learning_rate": 5.4749168714760736e-05, |
| "loss": 1.8094, |
| "step": 166500 |
| }, |
| { |
| "epoch": 11.73247154700014, |
| "grad_norm": 0.8988353610038757, |
| "learning_rate": 5.46045973688015e-05, |
| "loss": 1.8042, |
| "step": 167000 |
| }, |
| { |
| "epoch": 11.7675987073205, |
| "grad_norm": 0.9148843288421631, |
| "learning_rate": 5.446002602284227e-05, |
| "loss": 1.8045, |
| "step": 167500 |
| }, |
| { |
| "epoch": 11.80272586764086, |
| "grad_norm": 0.9276425838470459, |
| "learning_rate": 5.431545467688305e-05, |
| "loss": 1.8077, |
| "step": 168000 |
| }, |
| { |
| "epoch": 11.83785302796122, |
| "grad_norm": 0.9034392237663269, |
| "learning_rate": 5.417088333092382e-05, |
| "loss": 1.8033, |
| "step": 168500 |
| }, |
| { |
| "epoch": 11.872980188281579, |
| "grad_norm": 0.9140313863754272, |
| "learning_rate": 5.402631198496458e-05, |
| "loss": 1.8037, |
| "step": 169000 |
| }, |
| { |
| "epoch": 11.90810734860194, |
| "grad_norm": 0.9139729142189026, |
| "learning_rate": 5.3881740639005354e-05, |
| "loss": 1.8036, |
| "step": 169500 |
| }, |
| { |
| "epoch": 11.943234508922298, |
| "grad_norm": 0.9192969799041748, |
| "learning_rate": 5.373716929304612e-05, |
| "loss": 1.8038, |
| "step": 170000 |
| }, |
| { |
| "epoch": 11.943234508922298, |
| "eval_accuracy": 0.6740451513112237, |
| "eval_loss": 1.5250400304794312, |
| "eval_runtime": 122.3163, |
| "eval_samples_per_second": 986.835, |
| "eval_steps_per_second": 6.173, |
| "step": 170000 |
| }, |
| { |
| "epoch": 11.978361669242659, |
| "grad_norm": 0.9270504117012024, |
| "learning_rate": 5.359259794708689e-05, |
| "loss": 1.8029, |
| "step": 170500 |
| }, |
| { |
| "epoch": 12.013488829563018, |
| "grad_norm": 0.9690572619438171, |
| "learning_rate": 5.344802660112765e-05, |
| "loss": 1.8042, |
| "step": 171000 |
| }, |
| { |
| "epoch": 12.048615989883379, |
| "grad_norm": 0.9404934644699097, |
| "learning_rate": 5.330345525516843e-05, |
| "loss": 1.8012, |
| "step": 171500 |
| }, |
| { |
| "epoch": 12.083743150203738, |
| "grad_norm": 0.9095417857170105, |
| "learning_rate": 5.31588839092092e-05, |
| "loss": 1.7982, |
| "step": 172000 |
| }, |
| { |
| "epoch": 12.118870310524096, |
| "grad_norm": 0.9265354871749878, |
| "learning_rate": 5.3014312563249965e-05, |
| "loss": 1.8005, |
| "step": 172500 |
| }, |
| { |
| "epoch": 12.153997470844457, |
| "grad_norm": 0.9355581402778625, |
| "learning_rate": 5.2869741217290736e-05, |
| "loss": 1.8036, |
| "step": 173000 |
| }, |
| { |
| "epoch": 12.189124631164816, |
| "grad_norm": 0.9052658081054688, |
| "learning_rate": 5.272516987133151e-05, |
| "loss": 1.7996, |
| "step": 173500 |
| }, |
| { |
| "epoch": 12.224251791485177, |
| "grad_norm": 0.9294010996818542, |
| "learning_rate": 5.258059852537227e-05, |
| "loss": 1.7945, |
| "step": 174000 |
| }, |
| { |
| "epoch": 12.259378951805536, |
| "grad_norm": 0.9216769933700562, |
| "learning_rate": 5.243602717941304e-05, |
| "loss": 1.7995, |
| "step": 174500 |
| }, |
| { |
| "epoch": 12.294506112125896, |
| "grad_norm": 0.9287693500518799, |
| "learning_rate": 5.229145583345382e-05, |
| "loss": 1.7946, |
| "step": 175000 |
| }, |
| { |
| "epoch": 12.329633272446255, |
| "grad_norm": 0.9121254682540894, |
| "learning_rate": 5.214688448749458e-05, |
| "loss": 1.8, |
| "step": 175500 |
| }, |
| { |
| "epoch": 12.364760432766616, |
| "grad_norm": 0.9176110625267029, |
| "learning_rate": 5.2002313141535354e-05, |
| "loss": 1.7959, |
| "step": 176000 |
| }, |
| { |
| "epoch": 12.399887593086975, |
| "grad_norm": 0.9151242971420288, |
| "learning_rate": 5.185774179557612e-05, |
| "loss": 1.7952, |
| "step": 176500 |
| }, |
| { |
| "epoch": 12.435014753407334, |
| "grad_norm": 0.9072933197021484, |
| "learning_rate": 5.171317044961689e-05, |
| "loss": 1.7956, |
| "step": 177000 |
| }, |
| { |
| "epoch": 12.470141913727694, |
| "grad_norm": 0.890991747379303, |
| "learning_rate": 5.156859910365765e-05, |
| "loss": 1.794, |
| "step": 177500 |
| }, |
| { |
| "epoch": 12.505269074048053, |
| "grad_norm": 0.9174100756645203, |
| "learning_rate": 5.1424027757698423e-05, |
| "loss": 1.7959, |
| "step": 178000 |
| }, |
| { |
| "epoch": 12.540396234368414, |
| "grad_norm": 0.9365373253822327, |
| "learning_rate": 5.127945641173919e-05, |
| "loss": 1.7971, |
| "step": 178500 |
| }, |
| { |
| "epoch": 12.575523394688773, |
| "grad_norm": 0.92861407995224, |
| "learning_rate": 5.113488506577997e-05, |
| "loss": 1.7924, |
| "step": 179000 |
| }, |
| { |
| "epoch": 12.610650555009133, |
| "grad_norm": 0.9171894788742065, |
| "learning_rate": 5.0990313719820736e-05, |
| "loss": 1.7921, |
| "step": 179500 |
| }, |
| { |
| "epoch": 12.645777715329492, |
| "grad_norm": 0.9318227171897888, |
| "learning_rate": 5.0845742373861507e-05, |
| "loss": 1.7932, |
| "step": 180000 |
| }, |
| { |
| "epoch": 12.645777715329492, |
| "eval_accuracy": 0.6760542921805517, |
| "eval_loss": 1.5106685161590576, |
| "eval_runtime": 122.7601, |
| "eval_samples_per_second": 983.268, |
| "eval_steps_per_second": 6.15, |
| "step": 180000 |
| }, |
| { |
| "epoch": 12.680904875649853, |
| "grad_norm": 0.9716969132423401, |
| "learning_rate": 5.070117102790227e-05, |
| "loss": 1.791, |
| "step": 180500 |
| }, |
| { |
| "epoch": 12.716032035970212, |
| "grad_norm": 0.9394809603691101, |
| "learning_rate": 5.055659968194304e-05, |
| "loss": 1.7917, |
| "step": 181000 |
| }, |
| { |
| "epoch": 12.751159196290573, |
| "grad_norm": 0.925958514213562, |
| "learning_rate": 5.0412028335983805e-05, |
| "loss": 1.7924, |
| "step": 181500 |
| }, |
| { |
| "epoch": 12.786286356610931, |
| "grad_norm": 0.9119183421134949, |
| "learning_rate": 5.0267456990024576e-05, |
| "loss": 1.7918, |
| "step": 182000 |
| }, |
| { |
| "epoch": 12.82141351693129, |
| "grad_norm": 0.9053396582603455, |
| "learning_rate": 5.0122885644065354e-05, |
| "loss": 1.7929, |
| "step": 182500 |
| }, |
| { |
| "epoch": 12.856540677251651, |
| "grad_norm": 0.9542437195777893, |
| "learning_rate": 4.997831429810612e-05, |
| "loss": 1.7896, |
| "step": 183000 |
| }, |
| { |
| "epoch": 12.89166783757201, |
| "grad_norm": 0.9358499050140381, |
| "learning_rate": 4.983374295214689e-05, |
| "loss": 1.789, |
| "step": 183500 |
| }, |
| { |
| "epoch": 12.92679499789237, |
| "grad_norm": 0.917269229888916, |
| "learning_rate": 4.968917160618766e-05, |
| "loss": 1.7864, |
| "step": 184000 |
| }, |
| { |
| "epoch": 12.96192215821273, |
| "grad_norm": 0.9288418292999268, |
| "learning_rate": 4.954460026022842e-05, |
| "loss": 1.7902, |
| "step": 184500 |
| }, |
| { |
| "epoch": 12.99704931853309, |
| "grad_norm": 0.9365912079811096, |
| "learning_rate": 4.9400028914269194e-05, |
| "loss": 1.7886, |
| "step": 185000 |
| }, |
| { |
| "epoch": 13.032176478853449, |
| "grad_norm": 0.9191272854804993, |
| "learning_rate": 4.9255457568309965e-05, |
| "loss": 1.784, |
| "step": 185500 |
| }, |
| { |
| "epoch": 13.06730363917381, |
| "grad_norm": 0.9435843825340271, |
| "learning_rate": 4.9110886222350736e-05, |
| "loss": 1.7851, |
| "step": 186000 |
| }, |
| { |
| "epoch": 13.102430799494169, |
| "grad_norm": 0.9292587637901306, |
| "learning_rate": 4.89663148763915e-05, |
| "loss": 1.7856, |
| "step": 186500 |
| }, |
| { |
| "epoch": 13.13755795981453, |
| "grad_norm": 0.9227132797241211, |
| "learning_rate": 4.882174353043227e-05, |
| "loss": 1.7816, |
| "step": 187000 |
| }, |
| { |
| "epoch": 13.172685120134888, |
| "grad_norm": 0.9432901740074158, |
| "learning_rate": 4.867717218447304e-05, |
| "loss": 1.7822, |
| "step": 187500 |
| }, |
| { |
| "epoch": 13.207812280455247, |
| "grad_norm": 0.9219586849212646, |
| "learning_rate": 4.853260083851381e-05, |
| "loss": 1.784, |
| "step": 188000 |
| }, |
| { |
| "epoch": 13.242939440775608, |
| "grad_norm": 0.9259042143821716, |
| "learning_rate": 4.8388029492554576e-05, |
| "loss": 1.7835, |
| "step": 188500 |
| }, |
| { |
| "epoch": 13.278066601095967, |
| "grad_norm": 0.907507061958313, |
| "learning_rate": 4.824345814659535e-05, |
| "loss": 1.7821, |
| "step": 189000 |
| }, |
| { |
| "epoch": 13.313193761416327, |
| "grad_norm": 0.921046793460846, |
| "learning_rate": 4.809888680063612e-05, |
| "loss": 1.7856, |
| "step": 189500 |
| }, |
| { |
| "epoch": 13.348320921736686, |
| "grad_norm": 0.9301853775978088, |
| "learning_rate": 4.795431545467688e-05, |
| "loss": 1.7826, |
| "step": 190000 |
| }, |
| { |
| "epoch": 13.348320921736686, |
| "eval_accuracy": 0.677024651915173, |
| "eval_loss": 1.5057569742202759, |
| "eval_runtime": 123.203, |
| "eval_samples_per_second": 979.733, |
| "eval_steps_per_second": 6.128, |
| "step": 190000 |
| }, |
| { |
| "epoch": 13.383448082057047, |
| "grad_norm": 0.9536148905754089, |
| "learning_rate": 4.780974410871765e-05, |
| "loss": 1.778, |
| "step": 190500 |
| }, |
| { |
| "epoch": 13.418575242377406, |
| "grad_norm": 0.9380630850791931, |
| "learning_rate": 4.766517276275842e-05, |
| "loss": 1.7816, |
| "step": 191000 |
| }, |
| { |
| "epoch": 13.453702402697767, |
| "grad_norm": 0.9119521975517273, |
| "learning_rate": 4.7520601416799194e-05, |
| "loss": 1.7782, |
| "step": 191500 |
| }, |
| { |
| "epoch": 13.488829563018125, |
| "grad_norm": 0.9392079710960388, |
| "learning_rate": 4.737603007083996e-05, |
| "loss": 1.7802, |
| "step": 192000 |
| }, |
| { |
| "epoch": 13.523956723338486, |
| "grad_norm": 0.9062619209289551, |
| "learning_rate": 4.723145872488073e-05, |
| "loss": 1.7822, |
| "step": 192500 |
| }, |
| { |
| "epoch": 13.559083883658845, |
| "grad_norm": 0.9596025943756104, |
| "learning_rate": 4.70868873789215e-05, |
| "loss": 1.7799, |
| "step": 193000 |
| }, |
| { |
| "epoch": 13.594211043979204, |
| "grad_norm": 0.9480317234992981, |
| "learning_rate": 4.694231603296227e-05, |
| "loss": 1.78, |
| "step": 193500 |
| }, |
| { |
| "epoch": 13.629338204299565, |
| "grad_norm": 0.9420462250709534, |
| "learning_rate": 4.6797744687003034e-05, |
| "loss": 1.7765, |
| "step": 194000 |
| }, |
| { |
| "epoch": 13.664465364619923, |
| "grad_norm": 0.944510817527771, |
| "learning_rate": 4.665317334104381e-05, |
| "loss": 1.7756, |
| "step": 194500 |
| }, |
| { |
| "epoch": 13.699592524940284, |
| "grad_norm": 0.939622700214386, |
| "learning_rate": 4.6508601995084576e-05, |
| "loss": 1.7776, |
| "step": 195000 |
| }, |
| { |
| "epoch": 13.734719685260643, |
| "grad_norm": 0.9391052722930908, |
| "learning_rate": 4.6364030649125346e-05, |
| "loss": 1.7769, |
| "step": 195500 |
| }, |
| { |
| "epoch": 13.769846845581004, |
| "grad_norm": 0.9237781167030334, |
| "learning_rate": 4.621945930316611e-05, |
| "loss": 1.7771, |
| "step": 196000 |
| }, |
| { |
| "epoch": 13.804974005901363, |
| "grad_norm": 0.9500161409378052, |
| "learning_rate": 4.607488795720689e-05, |
| "loss": 1.7741, |
| "step": 196500 |
| }, |
| { |
| "epoch": 13.840101166221723, |
| "grad_norm": 0.933706521987915, |
| "learning_rate": 4.593031661124765e-05, |
| "loss": 1.7756, |
| "step": 197000 |
| }, |
| { |
| "epoch": 13.875228326542082, |
| "grad_norm": 0.9328217506408691, |
| "learning_rate": 4.578574526528842e-05, |
| "loss": 1.7728, |
| "step": 197500 |
| }, |
| { |
| "epoch": 13.910355486862443, |
| "grad_norm": 0.922516942024231, |
| "learning_rate": 4.564117391932919e-05, |
| "loss": 1.7744, |
| "step": 198000 |
| }, |
| { |
| "epoch": 13.945482647182802, |
| "grad_norm": 0.9467635750770569, |
| "learning_rate": 4.5496602573369964e-05, |
| "loss": 1.7731, |
| "step": 198500 |
| }, |
| { |
| "epoch": 13.98060980750316, |
| "grad_norm": 0.9323879480361938, |
| "learning_rate": 4.535203122741073e-05, |
| "loss": 1.7779, |
| "step": 199000 |
| }, |
| { |
| "epoch": 14.015736967823521, |
| "grad_norm": 0.923172116279602, |
| "learning_rate": 4.52074598814515e-05, |
| "loss": 1.7726, |
| "step": 199500 |
| }, |
| { |
| "epoch": 14.05086412814388, |
| "grad_norm": 0.9327266216278076, |
| "learning_rate": 4.506288853549227e-05, |
| "loss": 1.7736, |
| "step": 200000 |
| }, |
| { |
| "epoch": 14.05086412814388, |
| "eval_accuracy": 0.6785191532727831, |
| "eval_loss": 1.4952304363250732, |
| "eval_runtime": 123.3293, |
| "eval_samples_per_second": 978.729, |
| "eval_steps_per_second": 6.122, |
| "step": 200000 |
| }, |
| { |
| "epoch": 14.08599128846424, |
| "grad_norm": 0.9325737953186035, |
| "learning_rate": 4.491831718953304e-05, |
| "loss": 1.7701, |
| "step": 200500 |
| }, |
| { |
| "epoch": 14.1211184487846, |
| "grad_norm": 0.9446019530296326, |
| "learning_rate": 4.4773745843573805e-05, |
| "loss": 1.772, |
| "step": 201000 |
| }, |
| { |
| "epoch": 14.15624560910496, |
| "grad_norm": 0.9441095590591431, |
| "learning_rate": 4.4629174497614575e-05, |
| "loss": 1.7698, |
| "step": 201500 |
| }, |
| { |
| "epoch": 14.19137276942532, |
| "grad_norm": 0.9392231702804565, |
| "learning_rate": 4.4484603151655346e-05, |
| "loss": 1.7679, |
| "step": 202000 |
| }, |
| { |
| "epoch": 14.22649992974568, |
| "grad_norm": 0.9567651748657227, |
| "learning_rate": 4.434003180569612e-05, |
| "loss": 1.7695, |
| "step": 202500 |
| }, |
| { |
| "epoch": 14.261627090066039, |
| "grad_norm": 0.9683824181556702, |
| "learning_rate": 4.419546045973688e-05, |
| "loss": 1.7704, |
| "step": 203000 |
| }, |
| { |
| "epoch": 14.2967542503864, |
| "grad_norm": 0.9330751299858093, |
| "learning_rate": 4.405088911377765e-05, |
| "loss": 1.7687, |
| "step": 203500 |
| }, |
| { |
| "epoch": 14.331881410706758, |
| "grad_norm": 0.942535936832428, |
| "learning_rate": 4.390631776781842e-05, |
| "loss": 1.7715, |
| "step": 204000 |
| }, |
| { |
| "epoch": 14.367008571027117, |
| "grad_norm": 0.9362878203392029, |
| "learning_rate": 4.3761746421859187e-05, |
| "loss": 1.7689, |
| "step": 204500 |
| }, |
| { |
| "epoch": 14.402135731347478, |
| "grad_norm": 0.9386815428733826, |
| "learning_rate": 4.361717507589996e-05, |
| "loss": 1.7674, |
| "step": 205000 |
| }, |
| { |
| "epoch": 14.437262891667837, |
| "grad_norm": 0.9423466920852661, |
| "learning_rate": 4.347260372994073e-05, |
| "loss": 1.7696, |
| "step": 205500 |
| }, |
| { |
| "epoch": 14.472390051988198, |
| "grad_norm": 0.9570279717445374, |
| "learning_rate": 4.33280323839815e-05, |
| "loss": 1.7666, |
| "step": 206000 |
| }, |
| { |
| "epoch": 14.507517212308557, |
| "grad_norm": 0.9636929631233215, |
| "learning_rate": 4.318346103802226e-05, |
| "loss": 1.7656, |
| "step": 206500 |
| }, |
| { |
| "epoch": 14.542644372628917, |
| "grad_norm": 0.933325469493866, |
| "learning_rate": 4.3038889692063034e-05, |
| "loss": 1.7644, |
| "step": 207000 |
| }, |
| { |
| "epoch": 14.577771532949276, |
| "grad_norm": 0.9618740081787109, |
| "learning_rate": 4.2894318346103804e-05, |
| "loss": 1.7657, |
| "step": 207500 |
| }, |
| { |
| "epoch": 14.612898693269637, |
| "grad_norm": 0.959227442741394, |
| "learning_rate": 4.2749747000144575e-05, |
| "loss": 1.7663, |
| "step": 208000 |
| }, |
| { |
| "epoch": 14.648025853589996, |
| "grad_norm": 0.9232593178749084, |
| "learning_rate": 4.260517565418534e-05, |
| "loss": 1.7646, |
| "step": 208500 |
| }, |
| { |
| "epoch": 14.683153013910356, |
| "grad_norm": 0.9647856950759888, |
| "learning_rate": 4.246060430822611e-05, |
| "loss": 1.7645, |
| "step": 209000 |
| }, |
| { |
| "epoch": 14.718280174230715, |
| "grad_norm": 0.929384171962738, |
| "learning_rate": 4.231603296226688e-05, |
| "loss": 1.7642, |
| "step": 209500 |
| }, |
| { |
| "epoch": 14.753407334551074, |
| "grad_norm": 0.9312683343887329, |
| "learning_rate": 4.217146161630765e-05, |
| "loss": 1.7635, |
| "step": 210000 |
| }, |
| { |
| "epoch": 14.753407334551074, |
| "eval_accuracy": 0.6799087869761088, |
| "eval_loss": 1.4861806631088257, |
| "eval_runtime": 123.8053, |
| "eval_samples_per_second": 974.967, |
| "eval_steps_per_second": 6.098, |
| "step": 210000 |
| }, |
| { |
| "epoch": 14.788534494871435, |
| "grad_norm": 0.964197039604187, |
| "learning_rate": 4.2026890270348416e-05, |
| "loss": 1.7633, |
| "step": 210500 |
| }, |
| { |
| "epoch": 14.823661655191794, |
| "grad_norm": 0.9745231866836548, |
| "learning_rate": 4.1882318924389186e-05, |
| "loss": 1.7623, |
| "step": 211000 |
| }, |
| { |
| "epoch": 14.858788815512154, |
| "grad_norm": 0.9433575868606567, |
| "learning_rate": 4.173774757842996e-05, |
| "loss": 1.7635, |
| "step": 211500 |
| }, |
| { |
| "epoch": 14.893915975832513, |
| "grad_norm": 0.9631170630455017, |
| "learning_rate": 4.159317623247073e-05, |
| "loss": 1.7609, |
| "step": 212000 |
| }, |
| { |
| "epoch": 14.929043136152874, |
| "grad_norm": 0.9696474075317383, |
| "learning_rate": 4.144860488651149e-05, |
| "loss": 1.7639, |
| "step": 212500 |
| }, |
| { |
| "epoch": 14.964170296473233, |
| "grad_norm": 0.9274208545684814, |
| "learning_rate": 4.130403354055227e-05, |
| "loss": 1.7633, |
| "step": 213000 |
| }, |
| { |
| "epoch": 14.999297456793593, |
| "grad_norm": 0.9089575409889221, |
| "learning_rate": 4.1159462194593033e-05, |
| "loss": 1.7634, |
| "step": 213500 |
| }, |
| { |
| "epoch": 15.034424617113952, |
| "grad_norm": 0.9492926597595215, |
| "learning_rate": 4.1014890848633804e-05, |
| "loss": 1.758, |
| "step": 214000 |
| }, |
| { |
| "epoch": 15.069551777434313, |
| "grad_norm": 0.939822793006897, |
| "learning_rate": 4.087031950267457e-05, |
| "loss": 1.7586, |
| "step": 214500 |
| }, |
| { |
| "epoch": 15.104678937754672, |
| "grad_norm": 0.9360830783843994, |
| "learning_rate": 4.0725748156715346e-05, |
| "loss": 1.76, |
| "step": 215000 |
| }, |
| { |
| "epoch": 15.13980609807503, |
| "grad_norm": 0.944041907787323, |
| "learning_rate": 4.058117681075611e-05, |
| "loss": 1.7554, |
| "step": 215500 |
| }, |
| { |
| "epoch": 15.174933258395392, |
| "grad_norm": 0.9470905661582947, |
| "learning_rate": 4.043660546479688e-05, |
| "loss": 1.7602, |
| "step": 216000 |
| }, |
| { |
| "epoch": 15.21006041871575, |
| "grad_norm": 0.9378679990768433, |
| "learning_rate": 4.0292034118837645e-05, |
| "loss": 1.7594, |
| "step": 216500 |
| }, |
| { |
| "epoch": 15.245187579036111, |
| "grad_norm": 0.9557765126228333, |
| "learning_rate": 4.014746277287842e-05, |
| "loss": 1.7581, |
| "step": 217000 |
| }, |
| { |
| "epoch": 15.28031473935647, |
| "grad_norm": 0.9435907602310181, |
| "learning_rate": 4.0002891426919186e-05, |
| "loss": 1.7569, |
| "step": 217500 |
| }, |
| { |
| "epoch": 15.31544189967683, |
| "grad_norm": 0.956589937210083, |
| "learning_rate": 3.985832008095996e-05, |
| "loss": 1.7579, |
| "step": 218000 |
| }, |
| { |
| "epoch": 15.35056905999719, |
| "grad_norm": 0.9466123580932617, |
| "learning_rate": 3.971374873500072e-05, |
| "loss": 1.758, |
| "step": 218500 |
| }, |
| { |
| "epoch": 15.38569622031755, |
| "grad_norm": 0.9562757015228271, |
| "learning_rate": 3.95691773890415e-05, |
| "loss": 1.7564, |
| "step": 219000 |
| }, |
| { |
| "epoch": 15.42082338063791, |
| "grad_norm": 0.9425830245018005, |
| "learning_rate": 3.942460604308226e-05, |
| "loss": 1.7569, |
| "step": 219500 |
| }, |
| { |
| "epoch": 15.45595054095827, |
| "grad_norm": 0.9630948901176453, |
| "learning_rate": 3.928003469712303e-05, |
| "loss": 1.7549, |
| "step": 220000 |
| }, |
| { |
| "epoch": 15.45595054095827, |
| "eval_accuracy": 0.6815102684170856, |
| "eval_loss": 1.478232979774475, |
| "eval_runtime": 123.9931, |
| "eval_samples_per_second": 973.49, |
| "eval_steps_per_second": 6.089, |
| "step": 220000 |
| }, |
| { |
| "epoch": 15.491077701278629, |
| "grad_norm": 0.9481454491615295, |
| "learning_rate": 3.9135463351163804e-05, |
| "loss": 1.7566, |
| "step": 220500 |
| }, |
| { |
| "epoch": 15.526204861598988, |
| "grad_norm": 0.9637031555175781, |
| "learning_rate": 3.899089200520457e-05, |
| "loss": 1.7572, |
| "step": 221000 |
| }, |
| { |
| "epoch": 15.561332021919348, |
| "grad_norm": 0.9362086653709412, |
| "learning_rate": 3.884632065924534e-05, |
| "loss": 1.7555, |
| "step": 221500 |
| }, |
| { |
| "epoch": 15.596459182239707, |
| "grad_norm": 0.9242130517959595, |
| "learning_rate": 3.870174931328611e-05, |
| "loss": 1.7564, |
| "step": 222000 |
| }, |
| { |
| "epoch": 15.631586342560068, |
| "grad_norm": 0.9315471053123474, |
| "learning_rate": 3.855717796732688e-05, |
| "loss": 1.7524, |
| "step": 222500 |
| }, |
| { |
| "epoch": 15.666713502880427, |
| "grad_norm": 0.9530342817306519, |
| "learning_rate": 3.8412606621367644e-05, |
| "loss": 1.7542, |
| "step": 223000 |
| }, |
| { |
| "epoch": 15.701840663200787, |
| "grad_norm": 0.9682760834693909, |
| "learning_rate": 3.8268035275408415e-05, |
| "loss": 1.7547, |
| "step": 223500 |
| }, |
| { |
| "epoch": 15.736967823521146, |
| "grad_norm": 0.9661684036254883, |
| "learning_rate": 3.812346392944918e-05, |
| "loss": 1.7523, |
| "step": 224000 |
| }, |
| { |
| "epoch": 15.772094983841507, |
| "grad_norm": 0.9484301209449768, |
| "learning_rate": 3.797889258348996e-05, |
| "loss": 1.7503, |
| "step": 224500 |
| }, |
| { |
| "epoch": 15.807222144161866, |
| "grad_norm": 0.9527466297149658, |
| "learning_rate": 3.783432123753072e-05, |
| "loss": 1.7488, |
| "step": 225000 |
| }, |
| { |
| "epoch": 15.842349304482227, |
| "grad_norm": 0.9570134878158569, |
| "learning_rate": 3.768974989157149e-05, |
| "loss": 1.7497, |
| "step": 225500 |
| }, |
| { |
| "epoch": 15.877476464802585, |
| "grad_norm": 0.9423008561134338, |
| "learning_rate": 3.754517854561226e-05, |
| "loss": 1.7534, |
| "step": 226000 |
| }, |
| { |
| "epoch": 15.912603625122944, |
| "grad_norm": 0.9657214879989624, |
| "learning_rate": 3.740060719965303e-05, |
| "loss": 1.7539, |
| "step": 226500 |
| }, |
| { |
| "epoch": 15.947730785443305, |
| "grad_norm": 0.9357506632804871, |
| "learning_rate": 3.72560358536938e-05, |
| "loss": 1.7501, |
| "step": 227000 |
| }, |
| { |
| "epoch": 15.982857945763664, |
| "grad_norm": 0.9499670267105103, |
| "learning_rate": 3.711146450773457e-05, |
| "loss": 1.7519, |
| "step": 227500 |
| }, |
| { |
| "epoch": 16.017985106084023, |
| "grad_norm": 0.9760450124740601, |
| "learning_rate": 3.696689316177534e-05, |
| "loss": 1.7481, |
| "step": 228000 |
| }, |
| { |
| "epoch": 16.053112266404383, |
| "grad_norm": 0.9631822109222412, |
| "learning_rate": 3.682232181581611e-05, |
| "loss": 1.7465, |
| "step": 228500 |
| }, |
| { |
| "epoch": 16.088239426724744, |
| "grad_norm": 0.987704336643219, |
| "learning_rate": 3.667775046985687e-05, |
| "loss": 1.7478, |
| "step": 229000 |
| }, |
| { |
| "epoch": 16.123366587045105, |
| "grad_norm": 0.9503914713859558, |
| "learning_rate": 3.6533179123897644e-05, |
| "loss": 1.7481, |
| "step": 229500 |
| }, |
| { |
| "epoch": 16.158493747365462, |
| "grad_norm": 0.9707350730895996, |
| "learning_rate": 3.6388607777938415e-05, |
| "loss": 1.7497, |
| "step": 230000 |
| }, |
| { |
| "epoch": 16.158493747365462, |
| "eval_accuracy": 0.6821131056264182, |
| "eval_loss": 1.4737012386322021, |
| "eval_runtime": 123.9198, |
| "eval_samples_per_second": 974.066, |
| "eval_steps_per_second": 6.093, |
| "step": 230000 |
| }, |
| { |
| "epoch": 16.193620907685823, |
| "grad_norm": 0.9594478011131287, |
| "learning_rate": 3.6244036431979186e-05, |
| "loss": 1.7447, |
| "step": 230500 |
| }, |
| { |
| "epoch": 16.228748068006183, |
| "grad_norm": 0.9405125975608826, |
| "learning_rate": 3.609946508601995e-05, |
| "loss": 1.7469, |
| "step": 231000 |
| }, |
| { |
| "epoch": 16.26387522832654, |
| "grad_norm": 0.9684107899665833, |
| "learning_rate": 3.595489374006073e-05, |
| "loss": 1.7437, |
| "step": 231500 |
| }, |
| { |
| "epoch": 16.2990023886469, |
| "grad_norm": 0.935355007648468, |
| "learning_rate": 3.581032239410149e-05, |
| "loss": 1.7489, |
| "step": 232000 |
| }, |
| { |
| "epoch": 16.33412954896726, |
| "grad_norm": 0.9581801891326904, |
| "learning_rate": 3.566575104814226e-05, |
| "loss": 1.7445, |
| "step": 232500 |
| }, |
| { |
| "epoch": 16.369256709287622, |
| "grad_norm": 0.9640411734580994, |
| "learning_rate": 3.5521179702183026e-05, |
| "loss": 1.7452, |
| "step": 233000 |
| }, |
| { |
| "epoch": 16.40438386960798, |
| "grad_norm": 0.957968533039093, |
| "learning_rate": 3.5376608356223804e-05, |
| "loss": 1.745, |
| "step": 233500 |
| }, |
| { |
| "epoch": 16.43951102992834, |
| "grad_norm": 0.9438958764076233, |
| "learning_rate": 3.523203701026457e-05, |
| "loss": 1.7445, |
| "step": 234000 |
| }, |
| { |
| "epoch": 16.4746381902487, |
| "grad_norm": 0.9451521635055542, |
| "learning_rate": 3.508746566430534e-05, |
| "loss": 1.7463, |
| "step": 234500 |
| }, |
| { |
| "epoch": 16.50976535056906, |
| "grad_norm": 0.9519336819648743, |
| "learning_rate": 3.49428943183461e-05, |
| "loss": 1.7446, |
| "step": 235000 |
| }, |
| { |
| "epoch": 16.54489251088942, |
| "grad_norm": 0.954872190952301, |
| "learning_rate": 3.479832297238687e-05, |
| "loss": 1.7421, |
| "step": 235500 |
| }, |
| { |
| "epoch": 16.58001967120978, |
| "grad_norm": 0.9438602328300476, |
| "learning_rate": 3.4653751626427644e-05, |
| "loss": 1.7464, |
| "step": 236000 |
| }, |
| { |
| "epoch": 16.61514683153014, |
| "grad_norm": 0.9813643097877502, |
| "learning_rate": 3.4509180280468415e-05, |
| "loss": 1.7419, |
| "step": 236500 |
| }, |
| { |
| "epoch": 16.650273991850497, |
| "grad_norm": 0.9836772680282593, |
| "learning_rate": 3.436460893450918e-05, |
| "loss": 1.749, |
| "step": 237000 |
| }, |
| { |
| "epoch": 16.685401152170858, |
| "grad_norm": 0.9869303107261658, |
| "learning_rate": 3.422003758854995e-05, |
| "loss": 1.7441, |
| "step": 237500 |
| }, |
| { |
| "epoch": 16.72052831249122, |
| "grad_norm": 0.960365891456604, |
| "learning_rate": 3.407546624259072e-05, |
| "loss": 1.7428, |
| "step": 238000 |
| }, |
| { |
| "epoch": 16.75565547281158, |
| "grad_norm": 0.9576560258865356, |
| "learning_rate": 3.393089489663149e-05, |
| "loss": 1.7451, |
| "step": 238500 |
| }, |
| { |
| "epoch": 16.790782633131936, |
| "grad_norm": 0.9759810566902161, |
| "learning_rate": 3.378632355067226e-05, |
| "loss": 1.7422, |
| "step": 239000 |
| }, |
| { |
| "epoch": 16.825909793452297, |
| "grad_norm": 0.9629964828491211, |
| "learning_rate": 3.3641752204713026e-05, |
| "loss": 1.7413, |
| "step": 239500 |
| }, |
| { |
| "epoch": 16.861036953772658, |
| "grad_norm": 0.9406253695487976, |
| "learning_rate": 3.3497180858753797e-05, |
| "loss": 1.7407, |
| "step": 240000 |
| }, |
| { |
| "epoch": 16.861036953772658, |
| "eval_accuracy": 0.683232423256991, |
| "eval_loss": 1.467530369758606, |
| "eval_runtime": 123.8716, |
| "eval_samples_per_second": 974.445, |
| "eval_steps_per_second": 6.095, |
| "step": 240000 |
| }, |
| { |
| "epoch": 16.89616411409302, |
| "grad_norm": 0.970605731010437, |
| "learning_rate": 3.335260951279456e-05, |
| "loss": 1.7406, |
| "step": 240500 |
| }, |
| { |
| "epoch": 16.931291274413375, |
| "grad_norm": 0.9687949419021606, |
| "learning_rate": 3.320803816683534e-05, |
| "loss": 1.7445, |
| "step": 241000 |
| }, |
| { |
| "epoch": 16.966418434733736, |
| "grad_norm": 0.9652205109596252, |
| "learning_rate": 3.30634668208761e-05, |
| "loss": 1.7392, |
| "step": 241500 |
| }, |
| { |
| "epoch": 17.001545595054097, |
| "grad_norm": 0.9640699028968811, |
| "learning_rate": 3.291889547491687e-05, |
| "loss": 1.7441, |
| "step": 242000 |
| }, |
| { |
| "epoch": 17.036672755374454, |
| "grad_norm": 0.9808828830718994, |
| "learning_rate": 3.277432412895764e-05, |
| "loss": 1.7422, |
| "step": 242500 |
| }, |
| { |
| "epoch": 17.071799915694815, |
| "grad_norm": 0.9614331126213074, |
| "learning_rate": 3.2629752782998414e-05, |
| "loss": 1.7384, |
| "step": 243000 |
| }, |
| { |
| "epoch": 17.106927076015175, |
| "grad_norm": 0.9977091550827026, |
| "learning_rate": 3.248518143703918e-05, |
| "loss": 1.7393, |
| "step": 243500 |
| }, |
| { |
| "epoch": 17.142054236335536, |
| "grad_norm": 0.9659290313720703, |
| "learning_rate": 3.234061009107995e-05, |
| "loss": 1.7366, |
| "step": 244000 |
| }, |
| { |
| "epoch": 17.177181396655893, |
| "grad_norm": 0.9564438462257385, |
| "learning_rate": 3.219603874512072e-05, |
| "loss": 1.7364, |
| "step": 244500 |
| }, |
| { |
| "epoch": 17.212308556976254, |
| "grad_norm": 0.9877740144729614, |
| "learning_rate": 3.205146739916149e-05, |
| "loss": 1.7395, |
| "step": 245000 |
| }, |
| { |
| "epoch": 17.247435717296614, |
| "grad_norm": 0.9651238322257996, |
| "learning_rate": 3.1906896053202255e-05, |
| "loss": 1.7366, |
| "step": 245500 |
| }, |
| { |
| "epoch": 17.282562877616975, |
| "grad_norm": 0.9977210760116577, |
| "learning_rate": 3.1762324707243026e-05, |
| "loss": 1.7408, |
| "step": 246000 |
| }, |
| { |
| "epoch": 17.317690037937332, |
| "grad_norm": 0.9484860897064209, |
| "learning_rate": 3.1617753361283796e-05, |
| "loss": 1.7369, |
| "step": 246500 |
| }, |
| { |
| "epoch": 17.352817198257693, |
| "grad_norm": 0.9521619081497192, |
| "learning_rate": 3.147318201532457e-05, |
| "loss": 1.7357, |
| "step": 247000 |
| }, |
| { |
| "epoch": 17.387944358578054, |
| "grad_norm": 0.9652268290519714, |
| "learning_rate": 3.132861066936533e-05, |
| "loss": 1.7383, |
| "step": 247500 |
| }, |
| { |
| "epoch": 17.42307151889841, |
| "grad_norm": 0.9888260364532471, |
| "learning_rate": 3.11840393234061e-05, |
| "loss": 1.7316, |
| "step": 248000 |
| }, |
| { |
| "epoch": 17.45819867921877, |
| "grad_norm": 0.9842555522918701, |
| "learning_rate": 3.103946797744687e-05, |
| "loss": 1.7343, |
| "step": 248500 |
| }, |
| { |
| "epoch": 17.493325839539132, |
| "grad_norm": 0.9811616539955139, |
| "learning_rate": 3.0894896631487643e-05, |
| "loss": 1.7327, |
| "step": 249000 |
| }, |
| { |
| "epoch": 17.528452999859493, |
| "grad_norm": 1.0258958339691162, |
| "learning_rate": 3.075032528552841e-05, |
| "loss": 1.7322, |
| "step": 249500 |
| }, |
| { |
| "epoch": 17.56358016017985, |
| "grad_norm": 0.9356208443641663, |
| "learning_rate": 3.060575393956918e-05, |
| "loss": 1.7334, |
| "step": 250000 |
| }, |
| { |
| "epoch": 17.56358016017985, |
| "eval_accuracy": 0.6843066982594397, |
| "eval_loss": 1.4612157344818115, |
| "eval_runtime": 125.2144, |
| "eval_samples_per_second": 963.995, |
| "eval_steps_per_second": 6.03, |
| "step": 250000 |
| }, |
| { |
| "epoch": 17.59870732050021, |
| "grad_norm": 0.9861596822738647, |
| "learning_rate": 3.046118259360995e-05, |
| "loss": 1.7385, |
| "step": 250500 |
| }, |
| { |
| "epoch": 17.63383448082057, |
| "grad_norm": 0.9695125222206116, |
| "learning_rate": 3.0316611247650716e-05, |
| "loss": 1.732, |
| "step": 251000 |
| }, |
| { |
| "epoch": 17.668961641140932, |
| "grad_norm": 0.9782454371452332, |
| "learning_rate": 3.0172039901691484e-05, |
| "loss": 1.7333, |
| "step": 251500 |
| }, |
| { |
| "epoch": 17.70408880146129, |
| "grad_norm": 0.9732998013496399, |
| "learning_rate": 3.0027468555732258e-05, |
| "loss": 1.7343, |
| "step": 252000 |
| }, |
| { |
| "epoch": 17.73921596178165, |
| "grad_norm": 0.9806649684906006, |
| "learning_rate": 2.9882897209773025e-05, |
| "loss": 1.7343, |
| "step": 252500 |
| }, |
| { |
| "epoch": 17.77434312210201, |
| "grad_norm": 0.9701500535011292, |
| "learning_rate": 2.9738325863813793e-05, |
| "loss": 1.732, |
| "step": 253000 |
| }, |
| { |
| "epoch": 17.809470282422367, |
| "grad_norm": 0.9571565985679626, |
| "learning_rate": 2.959375451785456e-05, |
| "loss": 1.7343, |
| "step": 253500 |
| }, |
| { |
| "epoch": 17.844597442742728, |
| "grad_norm": 0.9742622375488281, |
| "learning_rate": 2.9449183171895334e-05, |
| "loss": 1.7311, |
| "step": 254000 |
| }, |
| { |
| "epoch": 17.87972460306309, |
| "grad_norm": 0.9732539653778076, |
| "learning_rate": 2.93046118259361e-05, |
| "loss": 1.7361, |
| "step": 254500 |
| }, |
| { |
| "epoch": 17.91485176338345, |
| "grad_norm": 0.9911869764328003, |
| "learning_rate": 2.916004047997687e-05, |
| "loss": 1.734, |
| "step": 255000 |
| }, |
| { |
| "epoch": 17.949978923703807, |
| "grad_norm": 0.9931451678276062, |
| "learning_rate": 2.9015469134017636e-05, |
| "loss": 1.7341, |
| "step": 255500 |
| }, |
| { |
| "epoch": 17.985106084024167, |
| "grad_norm": 1.0015878677368164, |
| "learning_rate": 2.887089778805841e-05, |
| "loss": 1.7316, |
| "step": 256000 |
| }, |
| { |
| "epoch": 18.020233244344528, |
| "grad_norm": 0.9583291411399841, |
| "learning_rate": 2.8726326442099178e-05, |
| "loss": 1.733, |
| "step": 256500 |
| }, |
| { |
| "epoch": 18.05536040466489, |
| "grad_norm": 0.9565033316612244, |
| "learning_rate": 2.8581755096139945e-05, |
| "loss": 1.7319, |
| "step": 257000 |
| }, |
| { |
| "epoch": 18.090487564985246, |
| "grad_norm": 0.9722208380699158, |
| "learning_rate": 2.843718375018072e-05, |
| "loss": 1.7287, |
| "step": 257500 |
| }, |
| { |
| "epoch": 18.125614725305606, |
| "grad_norm": 0.9902445673942566, |
| "learning_rate": 2.8292612404221487e-05, |
| "loss": 1.7312, |
| "step": 258000 |
| }, |
| { |
| "epoch": 18.160741885625967, |
| "grad_norm": 0.9808877110481262, |
| "learning_rate": 2.8148041058262254e-05, |
| "loss": 1.7295, |
| "step": 258500 |
| }, |
| { |
| "epoch": 18.195869045946324, |
| "grad_norm": 0.9743230938911438, |
| "learning_rate": 2.800346971230302e-05, |
| "loss": 1.7308, |
| "step": 259000 |
| }, |
| { |
| "epoch": 18.230996206266685, |
| "grad_norm": 0.9624738097190857, |
| "learning_rate": 2.7858898366343792e-05, |
| "loss": 1.7307, |
| "step": 259500 |
| }, |
| { |
| "epoch": 18.266123366587045, |
| "grad_norm": 1.0031484365463257, |
| "learning_rate": 2.7714327020384563e-05, |
| "loss": 1.7288, |
| "step": 260000 |
| }, |
| { |
| "epoch": 18.266123366587045, |
| "eval_accuracy": 0.6849457057156627, |
| "eval_loss": 1.4567586183547974, |
| "eval_runtime": 126.6444, |
| "eval_samples_per_second": 953.11, |
| "eval_steps_per_second": 5.962, |
| "step": 260000 |
| }, |
| { |
| "epoch": 18.301250526907406, |
| "grad_norm": 0.9802526831626892, |
| "learning_rate": 2.756975567442533e-05, |
| "loss": 1.7281, |
| "step": 260500 |
| }, |
| { |
| "epoch": 18.336377687227763, |
| "grad_norm": 1.008023977279663, |
| "learning_rate": 2.7425184328466098e-05, |
| "loss": 1.7264, |
| "step": 261000 |
| }, |
| { |
| "epoch": 18.371504847548124, |
| "grad_norm": 0.9780415892601013, |
| "learning_rate": 2.728061298250687e-05, |
| "loss": 1.7313, |
| "step": 261500 |
| }, |
| { |
| "epoch": 18.406632007868485, |
| "grad_norm": 0.9491579532623291, |
| "learning_rate": 2.7136041636547636e-05, |
| "loss": 1.728, |
| "step": 262000 |
| }, |
| { |
| "epoch": 18.441759168188845, |
| "grad_norm": 0.9651474356651306, |
| "learning_rate": 2.6991470290588407e-05, |
| "loss": 1.7251, |
| "step": 262500 |
| }, |
| { |
| "epoch": 18.476886328509202, |
| "grad_norm": 0.9988080859184265, |
| "learning_rate": 2.6846898944629178e-05, |
| "loss": 1.7251, |
| "step": 263000 |
| }, |
| { |
| "epoch": 18.512013488829563, |
| "grad_norm": 0.9819346070289612, |
| "learning_rate": 2.6702327598669945e-05, |
| "loss": 1.7244, |
| "step": 263500 |
| }, |
| { |
| "epoch": 18.547140649149924, |
| "grad_norm": 0.9724238514900208, |
| "learning_rate": 2.6557756252710713e-05, |
| "loss": 1.7247, |
| "step": 264000 |
| }, |
| { |
| "epoch": 18.58226780947028, |
| "grad_norm": 0.9832892417907715, |
| "learning_rate": 2.641318490675148e-05, |
| "loss": 1.7251, |
| "step": 264500 |
| }, |
| { |
| "epoch": 18.61739496979064, |
| "grad_norm": 1.020037293434143, |
| "learning_rate": 2.6268613560792254e-05, |
| "loss": 1.7289, |
| "step": 265000 |
| }, |
| { |
| "epoch": 18.652522130111002, |
| "grad_norm": 0.9707494974136353, |
| "learning_rate": 2.612404221483302e-05, |
| "loss": 1.7259, |
| "step": 265500 |
| }, |
| { |
| "epoch": 18.687649290431363, |
| "grad_norm": 0.9776583909988403, |
| "learning_rate": 2.597947086887379e-05, |
| "loss": 1.7238, |
| "step": 266000 |
| }, |
| { |
| "epoch": 18.72277645075172, |
| "grad_norm": 0.9959068298339844, |
| "learning_rate": 2.5834899522914556e-05, |
| "loss": 1.7271, |
| "step": 266500 |
| }, |
| { |
| "epoch": 18.75790361107208, |
| "grad_norm": 0.9585506319999695, |
| "learning_rate": 2.569032817695533e-05, |
| "loss": 1.7262, |
| "step": 267000 |
| }, |
| { |
| "epoch": 18.79303077139244, |
| "grad_norm": 0.9782285690307617, |
| "learning_rate": 2.5545756830996098e-05, |
| "loss": 1.728, |
| "step": 267500 |
| }, |
| { |
| "epoch": 18.828157931712802, |
| "grad_norm": 0.976788341999054, |
| "learning_rate": 2.5401185485036865e-05, |
| "loss": 1.726, |
| "step": 268000 |
| }, |
| { |
| "epoch": 18.86328509203316, |
| "grad_norm": 1.0110868215560913, |
| "learning_rate": 2.5256614139077633e-05, |
| "loss": 1.7247, |
| "step": 268500 |
| }, |
| { |
| "epoch": 18.89841225235352, |
| "grad_norm": 0.9687120318412781, |
| "learning_rate": 2.5112042793118407e-05, |
| "loss": 1.7251, |
| "step": 269000 |
| }, |
| { |
| "epoch": 18.93353941267388, |
| "grad_norm": 1.0036640167236328, |
| "learning_rate": 2.4967471447159174e-05, |
| "loss": 1.7242, |
| "step": 269500 |
| }, |
| { |
| "epoch": 18.968666572994238, |
| "grad_norm": 0.9706979393959045, |
| "learning_rate": 2.4822900101199945e-05, |
| "loss": 1.7265, |
| "step": 270000 |
| }, |
| { |
| "epoch": 18.968666572994238, |
| "eval_accuracy": 0.6858128224484996, |
| "eval_loss": 1.4519171714782715, |
| "eval_runtime": 125.5425, |
| "eval_samples_per_second": 961.475, |
| "eval_steps_per_second": 6.014, |
| "step": 270000 |
| }, |
| { |
| "epoch": 19.0037937333146, |
| "grad_norm": 1.0062251091003418, |
| "learning_rate": 2.4678328755240712e-05, |
| "loss": 1.7243, |
| "step": 270500 |
| }, |
| { |
| "epoch": 19.03892089363496, |
| "grad_norm": 0.9886362552642822, |
| "learning_rate": 2.4533757409281483e-05, |
| "loss": 1.7218, |
| "step": 271000 |
| }, |
| { |
| "epoch": 19.07404805395532, |
| "grad_norm": 0.9718700051307678, |
| "learning_rate": 2.438918606332225e-05, |
| "loss": 1.723, |
| "step": 271500 |
| }, |
| { |
| "epoch": 19.109175214275677, |
| "grad_norm": 0.9607687592506409, |
| "learning_rate": 2.424461471736302e-05, |
| "loss": 1.7217, |
| "step": 272000 |
| }, |
| { |
| "epoch": 19.144302374596037, |
| "grad_norm": 0.9833087921142578, |
| "learning_rate": 2.410004337140379e-05, |
| "loss": 1.7228, |
| "step": 272500 |
| }, |
| { |
| "epoch": 19.179429534916398, |
| "grad_norm": 0.9877829551696777, |
| "learning_rate": 2.395547202544456e-05, |
| "loss": 1.7227, |
| "step": 273000 |
| }, |
| { |
| "epoch": 19.21455669523676, |
| "grad_norm": 0.9727568030357361, |
| "learning_rate": 2.3810900679485327e-05, |
| "loss": 1.7189, |
| "step": 273500 |
| }, |
| { |
| "epoch": 19.249683855557116, |
| "grad_norm": 0.9535881876945496, |
| "learning_rate": 2.3666329333526098e-05, |
| "loss": 1.7197, |
| "step": 274000 |
| }, |
| { |
| "epoch": 19.284811015877477, |
| "grad_norm": 0.9929323792457581, |
| "learning_rate": 2.3521757987566865e-05, |
| "loss": 1.7207, |
| "step": 274500 |
| }, |
| { |
| "epoch": 19.319938176197837, |
| "grad_norm": 1.0060275793075562, |
| "learning_rate": 2.3377186641607636e-05, |
| "loss": 1.7227, |
| "step": 275000 |
| }, |
| { |
| "epoch": 19.355065336518194, |
| "grad_norm": 0.9844222664833069, |
| "learning_rate": 2.3232615295648403e-05, |
| "loss": 1.7238, |
| "step": 275500 |
| }, |
| { |
| "epoch": 19.390192496838555, |
| "grad_norm": 0.9736326336860657, |
| "learning_rate": 2.3088043949689174e-05, |
| "loss": 1.7187, |
| "step": 276000 |
| }, |
| { |
| "epoch": 19.425319657158916, |
| "grad_norm": 0.9709787964820862, |
| "learning_rate": 2.2943472603729945e-05, |
| "loss": 1.7223, |
| "step": 276500 |
| }, |
| { |
| "epoch": 19.460446817479276, |
| "grad_norm": 0.982831597328186, |
| "learning_rate": 2.2798901257770712e-05, |
| "loss": 1.7182, |
| "step": 277000 |
| }, |
| { |
| "epoch": 19.495573977799634, |
| "grad_norm": 0.9745506644248962, |
| "learning_rate": 2.265432991181148e-05, |
| "loss": 1.721, |
| "step": 277500 |
| }, |
| { |
| "epoch": 19.530701138119994, |
| "grad_norm": 0.99278324842453, |
| "learning_rate": 2.250975856585225e-05, |
| "loss": 1.7182, |
| "step": 278000 |
| }, |
| { |
| "epoch": 19.565828298440355, |
| "grad_norm": 0.9980936646461487, |
| "learning_rate": 2.2365187219893018e-05, |
| "loss": 1.7207, |
| "step": 278500 |
| }, |
| { |
| "epoch": 19.600955458760716, |
| "grad_norm": 0.9948651790618896, |
| "learning_rate": 2.2220615873933785e-05, |
| "loss": 1.7181, |
| "step": 279000 |
| }, |
| { |
| "epoch": 19.636082619081073, |
| "grad_norm": 1.0024000406265259, |
| "learning_rate": 2.2076044527974556e-05, |
| "loss": 1.7204, |
| "step": 279500 |
| }, |
| { |
| "epoch": 19.671209779401433, |
| "grad_norm": 0.9898077845573425, |
| "learning_rate": 2.1931473182015323e-05, |
| "loss": 1.7179, |
| "step": 280000 |
| }, |
| { |
| "epoch": 19.671209779401433, |
| "eval_accuracy": 0.6868170847468235, |
| "eval_loss": 1.4468756914138794, |
| "eval_runtime": 123.3991, |
| "eval_samples_per_second": 978.176, |
| "eval_steps_per_second": 6.118, |
| "step": 280000 |
| }, |
| { |
| "epoch": 19.706336939721794, |
| "grad_norm": 0.9522126913070679, |
| "learning_rate": 2.1786901836056094e-05, |
| "loss": 1.7204, |
| "step": 280500 |
| }, |
| { |
| "epoch": 19.74146410004215, |
| "grad_norm": 0.9856494069099426, |
| "learning_rate": 2.164233049009686e-05, |
| "loss": 1.7194, |
| "step": 281000 |
| }, |
| { |
| "epoch": 19.776591260362512, |
| "grad_norm": 1.0011855363845825, |
| "learning_rate": 2.1497759144137632e-05, |
| "loss": 1.7193, |
| "step": 281500 |
| }, |
| { |
| "epoch": 19.811718420682872, |
| "grad_norm": 0.9867976307868958, |
| "learning_rate": 2.1353187798178403e-05, |
| "loss": 1.7176, |
| "step": 282000 |
| }, |
| { |
| "epoch": 19.846845581003233, |
| "grad_norm": 0.9874268770217896, |
| "learning_rate": 2.120861645221917e-05, |
| "loss": 1.7179, |
| "step": 282500 |
| }, |
| { |
| "epoch": 19.88197274132359, |
| "grad_norm": 0.9992992281913757, |
| "learning_rate": 2.106404510625994e-05, |
| "loss": 1.7161, |
| "step": 283000 |
| }, |
| { |
| "epoch": 19.91709990164395, |
| "grad_norm": 1.0240249633789062, |
| "learning_rate": 2.091947376030071e-05, |
| "loss": 1.7202, |
| "step": 283500 |
| }, |
| { |
| "epoch": 19.95222706196431, |
| "grad_norm": 0.9993736743927002, |
| "learning_rate": 2.077490241434148e-05, |
| "loss": 1.7168, |
| "step": 284000 |
| }, |
| { |
| "epoch": 19.987354222284672, |
| "grad_norm": 0.9884563684463501, |
| "learning_rate": 2.0630331068382247e-05, |
| "loss": 1.7168, |
| "step": 284500 |
| }, |
| { |
| "epoch": 20.02248138260503, |
| "grad_norm": 0.99095219373703, |
| "learning_rate": 2.0485759722423017e-05, |
| "loss": 1.7194, |
| "step": 285000 |
| }, |
| { |
| "epoch": 20.05760854292539, |
| "grad_norm": 0.9920971393585205, |
| "learning_rate": 2.0341188376463785e-05, |
| "loss": 1.7153, |
| "step": 285500 |
| }, |
| { |
| "epoch": 20.09273570324575, |
| "grad_norm": 1.0193287134170532, |
| "learning_rate": 2.0196617030504556e-05, |
| "loss": 1.715, |
| "step": 286000 |
| }, |
| { |
| "epoch": 20.127862863566108, |
| "grad_norm": 1.0010589361190796, |
| "learning_rate": 2.0052045684545323e-05, |
| "loss": 1.7151, |
| "step": 286500 |
| }, |
| { |
| "epoch": 20.16299002388647, |
| "grad_norm": 0.9678018689155579, |
| "learning_rate": 1.9907474338586094e-05, |
| "loss": 1.7143, |
| "step": 287000 |
| }, |
| { |
| "epoch": 20.19811718420683, |
| "grad_norm": 0.9888837337493896, |
| "learning_rate": 1.976290299262686e-05, |
| "loss": 1.7132, |
| "step": 287500 |
| }, |
| { |
| "epoch": 20.23324434452719, |
| "grad_norm": 0.9658753871917725, |
| "learning_rate": 1.9618331646667632e-05, |
| "loss": 1.7146, |
| "step": 288000 |
| }, |
| { |
| "epoch": 20.268371504847547, |
| "grad_norm": 0.9823668599128723, |
| "learning_rate": 1.9473760300708403e-05, |
| "loss": 1.7123, |
| "step": 288500 |
| }, |
| { |
| "epoch": 20.303498665167908, |
| "grad_norm": 0.9993078112602234, |
| "learning_rate": 1.932918895474917e-05, |
| "loss": 1.7149, |
| "step": 289000 |
| }, |
| { |
| "epoch": 20.33862582548827, |
| "grad_norm": 1.0103904008865356, |
| "learning_rate": 1.918461760878994e-05, |
| "loss": 1.7138, |
| "step": 289500 |
| }, |
| { |
| "epoch": 20.37375298580863, |
| "grad_norm": 0.988292396068573, |
| "learning_rate": 1.9040046262830708e-05, |
| "loss": 1.7156, |
| "step": 290000 |
| }, |
| { |
| "epoch": 20.37375298580863, |
| "eval_accuracy": 0.6872867106014598, |
| "eval_loss": 1.442797303199768, |
| "eval_runtime": 125.0974, |
| "eval_samples_per_second": 964.896, |
| "eval_steps_per_second": 6.035, |
| "step": 290000 |
| }, |
| { |
| "epoch": 20.408880146128986, |
| "grad_norm": 0.9929233193397522, |
| "learning_rate": 1.889547491687148e-05, |
| "loss": 1.7138, |
| "step": 290500 |
| }, |
| { |
| "epoch": 20.444007306449347, |
| "grad_norm": 1.0042837858200073, |
| "learning_rate": 1.8750903570912246e-05, |
| "loss": 1.7158, |
| "step": 291000 |
| }, |
| { |
| "epoch": 20.479134466769708, |
| "grad_norm": 1.028939127922058, |
| "learning_rate": 1.8606332224953017e-05, |
| "loss": 1.7111, |
| "step": 291500 |
| }, |
| { |
| "epoch": 20.514261627090065, |
| "grad_norm": 1.00380277633667, |
| "learning_rate": 1.8461760878993785e-05, |
| "loss": 1.714, |
| "step": 292000 |
| }, |
| { |
| "epoch": 20.549388787410425, |
| "grad_norm": 0.9856330156326294, |
| "learning_rate": 1.8317189533034555e-05, |
| "loss": 1.7114, |
| "step": 292500 |
| }, |
| { |
| "epoch": 20.584515947730786, |
| "grad_norm": 0.9770887494087219, |
| "learning_rate": 1.8172618187075323e-05, |
| "loss": 1.7134, |
| "step": 293000 |
| }, |
| { |
| "epoch": 20.619643108051147, |
| "grad_norm": 1.0203800201416016, |
| "learning_rate": 1.8028046841116093e-05, |
| "loss": 1.7141, |
| "step": 293500 |
| }, |
| { |
| "epoch": 20.654770268371504, |
| "grad_norm": 0.9915537238121033, |
| "learning_rate": 1.788347549515686e-05, |
| "loss": 1.7105, |
| "step": 294000 |
| }, |
| { |
| "epoch": 20.689897428691864, |
| "grad_norm": 0.9719441533088684, |
| "learning_rate": 1.7738904149197628e-05, |
| "loss": 1.7135, |
| "step": 294500 |
| }, |
| { |
| "epoch": 20.725024589012225, |
| "grad_norm": 0.9837605953216553, |
| "learning_rate": 1.75943328032384e-05, |
| "loss": 1.7128, |
| "step": 295000 |
| }, |
| { |
| "epoch": 20.760151749332586, |
| "grad_norm": 1.002923607826233, |
| "learning_rate": 1.7449761457279166e-05, |
| "loss": 1.7144, |
| "step": 295500 |
| }, |
| { |
| "epoch": 20.795278909652943, |
| "grad_norm": 1.0188437700271606, |
| "learning_rate": 1.7305190111319937e-05, |
| "loss": 1.7099, |
| "step": 296000 |
| }, |
| { |
| "epoch": 20.830406069973304, |
| "grad_norm": 0.9990536570549011, |
| "learning_rate": 1.7160618765360705e-05, |
| "loss": 1.7113, |
| "step": 296500 |
| }, |
| { |
| "epoch": 20.865533230293664, |
| "grad_norm": 0.9832262396812439, |
| "learning_rate": 1.7016047419401475e-05, |
| "loss": 1.7086, |
| "step": 297000 |
| }, |
| { |
| "epoch": 20.90066039061402, |
| "grad_norm": 1.011824131011963, |
| "learning_rate": 1.6871476073442243e-05, |
| "loss": 1.7125, |
| "step": 297500 |
| }, |
| { |
| "epoch": 20.935787550934382, |
| "grad_norm": 0.9975372552871704, |
| "learning_rate": 1.6726904727483014e-05, |
| "loss": 1.7117, |
| "step": 298000 |
| }, |
| { |
| "epoch": 20.970914711254743, |
| "grad_norm": 1.0138835906982422, |
| "learning_rate": 1.658233338152378e-05, |
| "loss": 1.7121, |
| "step": 298500 |
| }, |
| { |
| "epoch": 21.006041871575103, |
| "grad_norm": 0.9824172854423523, |
| "learning_rate": 1.6437762035564552e-05, |
| "loss": 1.7113, |
| "step": 299000 |
| }, |
| { |
| "epoch": 21.04116903189546, |
| "grad_norm": 1.004136085510254, |
| "learning_rate": 1.629319068960532e-05, |
| "loss": 1.7116, |
| "step": 299500 |
| }, |
| { |
| "epoch": 21.07629619221582, |
| "grad_norm": 0.9585360884666443, |
| "learning_rate": 1.614861934364609e-05, |
| "loss": 1.7086, |
| "step": 300000 |
| }, |
| { |
| "epoch": 21.07629619221582, |
| "eval_accuracy": 0.6881649414522338, |
| "eval_loss": 1.438929557800293, |
| "eval_runtime": 122.0316, |
| "eval_samples_per_second": 989.137, |
| "eval_steps_per_second": 6.187, |
| "step": 300000 |
| }, |
| { |
| "epoch": 21.111423352536182, |
| "grad_norm": 0.9871618747711182, |
| "learning_rate": 1.6004047997686857e-05, |
| "loss": 1.7081, |
| "step": 300500 |
| }, |
| { |
| "epoch": 21.14655051285654, |
| "grad_norm": 0.9984627366065979, |
| "learning_rate": 1.5859476651727628e-05, |
| "loss": 1.7105, |
| "step": 301000 |
| }, |
| { |
| "epoch": 21.1816776731769, |
| "grad_norm": 1.026432991027832, |
| "learning_rate": 1.57149053057684e-05, |
| "loss": 1.7092, |
| "step": 301500 |
| }, |
| { |
| "epoch": 21.21680483349726, |
| "grad_norm": 0.9899799227714539, |
| "learning_rate": 1.5570333959809166e-05, |
| "loss": 1.7131, |
| "step": 302000 |
| }, |
| { |
| "epoch": 21.25193199381762, |
| "grad_norm": 1.0005983114242554, |
| "learning_rate": 1.5425762613849937e-05, |
| "loss": 1.7064, |
| "step": 302500 |
| }, |
| { |
| "epoch": 21.287059154137978, |
| "grad_norm": 0.9878481030464172, |
| "learning_rate": 1.5281191267890704e-05, |
| "loss": 1.7087, |
| "step": 303000 |
| }, |
| { |
| "epoch": 21.32218631445834, |
| "grad_norm": 1.015194296836853, |
| "learning_rate": 1.5136619921931475e-05, |
| "loss": 1.7084, |
| "step": 303500 |
| }, |
| { |
| "epoch": 21.3573134747787, |
| "grad_norm": 0.9895281791687012, |
| "learning_rate": 1.4992048575972243e-05, |
| "loss": 1.7052, |
| "step": 304000 |
| }, |
| { |
| "epoch": 21.39244063509906, |
| "grad_norm": 1.0375460386276245, |
| "learning_rate": 1.4847477230013013e-05, |
| "loss": 1.7096, |
| "step": 304500 |
| }, |
| { |
| "epoch": 21.427567795419417, |
| "grad_norm": 0.996401309967041, |
| "learning_rate": 1.470290588405378e-05, |
| "loss": 1.707, |
| "step": 305000 |
| }, |
| { |
| "epoch": 21.462694955739778, |
| "grad_norm": 0.9940415620803833, |
| "learning_rate": 1.4558334538094551e-05, |
| "loss": 1.7077, |
| "step": 305500 |
| }, |
| { |
| "epoch": 21.49782211606014, |
| "grad_norm": 1.0166724920272827, |
| "learning_rate": 1.4413763192135319e-05, |
| "loss": 1.7072, |
| "step": 306000 |
| }, |
| { |
| "epoch": 21.532949276380496, |
| "grad_norm": 1.0302542448043823, |
| "learning_rate": 1.4269191846176088e-05, |
| "loss": 1.707, |
| "step": 306500 |
| }, |
| { |
| "epoch": 21.568076436700856, |
| "grad_norm": 1.0291839838027954, |
| "learning_rate": 1.4124620500216857e-05, |
| "loss": 1.706, |
| "step": 307000 |
| }, |
| { |
| "epoch": 21.603203597021217, |
| "grad_norm": 0.9947880506515503, |
| "learning_rate": 1.3980049154257626e-05, |
| "loss": 1.7071, |
| "step": 307500 |
| }, |
| { |
| "epoch": 21.638330757341578, |
| "grad_norm": 1.0455776453018188, |
| "learning_rate": 1.3835477808298397e-05, |
| "loss": 1.712, |
| "step": 308000 |
| }, |
| { |
| "epoch": 21.673457917661935, |
| "grad_norm": 1.0297991037368774, |
| "learning_rate": 1.3690906462339164e-05, |
| "loss": 1.7078, |
| "step": 308500 |
| }, |
| { |
| "epoch": 21.708585077982296, |
| "grad_norm": 0.9970694184303284, |
| "learning_rate": 1.3546335116379935e-05, |
| "loss": 1.706, |
| "step": 309000 |
| }, |
| { |
| "epoch": 21.743712238302656, |
| "grad_norm": 0.9997472763061523, |
| "learning_rate": 1.3401763770420702e-05, |
| "loss": 1.7053, |
| "step": 309500 |
| }, |
| { |
| "epoch": 21.778839398623017, |
| "grad_norm": 0.9959579706192017, |
| "learning_rate": 1.3257192424461473e-05, |
| "loss": 1.7064, |
| "step": 310000 |
| }, |
| { |
| "epoch": 21.778839398623017, |
| "eval_accuracy": 0.6891847713020477, |
| "eval_loss": 1.4333512783050537, |
| "eval_runtime": 122.696, |
| "eval_samples_per_second": 983.781, |
| "eval_steps_per_second": 6.153, |
| "step": 310000 |
| }, |
| { |
| "epoch": 21.813966558943374, |
| "grad_norm": 0.9923667907714844, |
| "learning_rate": 1.311262107850224e-05, |
| "loss": 1.7042, |
| "step": 310500 |
| }, |
| { |
| "epoch": 21.849093719263735, |
| "grad_norm": 1.0111578702926636, |
| "learning_rate": 1.2968049732543011e-05, |
| "loss": 1.7024, |
| "step": 311000 |
| }, |
| { |
| "epoch": 21.884220879584095, |
| "grad_norm": 1.0152249336242676, |
| "learning_rate": 1.2823478386583779e-05, |
| "loss": 1.7081, |
| "step": 311500 |
| }, |
| { |
| "epoch": 21.919348039904452, |
| "grad_norm": 0.9608742594718933, |
| "learning_rate": 1.267890704062455e-05, |
| "loss": 1.7038, |
| "step": 312000 |
| }, |
| { |
| "epoch": 21.954475200224813, |
| "grad_norm": 0.9814004898071289, |
| "learning_rate": 1.2534335694665317e-05, |
| "loss": 1.7057, |
| "step": 312500 |
| }, |
| { |
| "epoch": 21.989602360545174, |
| "grad_norm": 1.0021171569824219, |
| "learning_rate": 1.2389764348706088e-05, |
| "loss": 1.7067, |
| "step": 313000 |
| }, |
| { |
| "epoch": 22.024729520865534, |
| "grad_norm": 1.024749994277954, |
| "learning_rate": 1.2245193002746857e-05, |
| "loss": 1.704, |
| "step": 313500 |
| }, |
| { |
| "epoch": 22.05985668118589, |
| "grad_norm": 1.017061471939087, |
| "learning_rate": 1.2100621656787626e-05, |
| "loss": 1.7013, |
| "step": 314000 |
| }, |
| { |
| "epoch": 22.094983841506252, |
| "grad_norm": 1.004145622253418, |
| "learning_rate": 1.1956050310828395e-05, |
| "loss": 1.706, |
| "step": 314500 |
| }, |
| { |
| "epoch": 22.130111001826613, |
| "grad_norm": 1.0147069692611694, |
| "learning_rate": 1.1811478964869162e-05, |
| "loss": 1.702, |
| "step": 315000 |
| }, |
| { |
| "epoch": 22.165238162146974, |
| "grad_norm": 1.0169391632080078, |
| "learning_rate": 1.1666907618909931e-05, |
| "loss": 1.7008, |
| "step": 315500 |
| }, |
| { |
| "epoch": 22.20036532246733, |
| "grad_norm": 1.001263976097107, |
| "learning_rate": 1.15223362729507e-05, |
| "loss": 1.7016, |
| "step": 316000 |
| }, |
| { |
| "epoch": 22.23549248278769, |
| "grad_norm": 1.0239739418029785, |
| "learning_rate": 1.137776492699147e-05, |
| "loss": 1.7035, |
| "step": 316500 |
| }, |
| { |
| "epoch": 22.270619643108052, |
| "grad_norm": 1.0169011354446411, |
| "learning_rate": 1.123319358103224e-05, |
| "loss": 1.7044, |
| "step": 317000 |
| }, |
| { |
| "epoch": 22.30574680342841, |
| "grad_norm": 0.9854140877723694, |
| "learning_rate": 1.108862223507301e-05, |
| "loss": 1.7035, |
| "step": 317500 |
| }, |
| { |
| "epoch": 22.34087396374877, |
| "grad_norm": 0.9744647741317749, |
| "learning_rate": 1.0944050889113779e-05, |
| "loss": 1.7017, |
| "step": 318000 |
| }, |
| { |
| "epoch": 22.37600112406913, |
| "grad_norm": 1.0284457206726074, |
| "learning_rate": 1.0799479543154548e-05, |
| "loss": 1.7053, |
| "step": 318500 |
| }, |
| { |
| "epoch": 22.41112828438949, |
| "grad_norm": 0.998509407043457, |
| "learning_rate": 1.0654908197195317e-05, |
| "loss": 1.7013, |
| "step": 319000 |
| }, |
| { |
| "epoch": 22.44625544470985, |
| "grad_norm": 1.0026978254318237, |
| "learning_rate": 1.0510336851236086e-05, |
| "loss": 1.7002, |
| "step": 319500 |
| }, |
| { |
| "epoch": 22.48138260503021, |
| "grad_norm": 1.0223945379257202, |
| "learning_rate": 1.0365765505276855e-05, |
| "loss": 1.7009, |
| "step": 320000 |
| }, |
| { |
| "epoch": 22.48138260503021, |
| "eval_accuracy": 0.6894801420018741, |
| "eval_loss": 1.4309388399124146, |
| "eval_runtime": 123.5652, |
| "eval_samples_per_second": 976.86, |
| "eval_steps_per_second": 6.11, |
| "step": 320000 |
| }, |
| { |
| "epoch": 22.51650976535057, |
| "grad_norm": 1.0213385820388794, |
| "learning_rate": 1.0221194159317624e-05, |
| "loss": 1.7023, |
| "step": 320500 |
| }, |
| { |
| "epoch": 22.55163692567093, |
| "grad_norm": 1.0591439008712769, |
| "learning_rate": 1.0076622813358393e-05, |
| "loss": 1.7057, |
| "step": 321000 |
| }, |
| { |
| "epoch": 22.586764085991287, |
| "grad_norm": 1.0163780450820923, |
| "learning_rate": 9.932051467399162e-06, |
| "loss": 1.7041, |
| "step": 321500 |
| }, |
| { |
| "epoch": 22.621891246311648, |
| "grad_norm": 0.9861245155334473, |
| "learning_rate": 9.787480121439931e-06, |
| "loss": 1.7007, |
| "step": 322000 |
| }, |
| { |
| "epoch": 22.65701840663201, |
| "grad_norm": 1.0122628211975098, |
| "learning_rate": 9.6429087754807e-06, |
| "loss": 1.7025, |
| "step": 322500 |
| }, |
| { |
| "epoch": 22.692145566952366, |
| "grad_norm": 1.0044549703598022, |
| "learning_rate": 9.49833742952147e-06, |
| "loss": 1.7023, |
| "step": 323000 |
| }, |
| { |
| "epoch": 22.727272727272727, |
| "grad_norm": 0.9898348450660706, |
| "learning_rate": 9.353766083562238e-06, |
| "loss": 1.7001, |
| "step": 323500 |
| }, |
| { |
| "epoch": 22.762399887593087, |
| "grad_norm": 1.0124576091766357, |
| "learning_rate": 9.209194737603008e-06, |
| "loss": 1.6989, |
| "step": 324000 |
| }, |
| { |
| "epoch": 22.797527047913448, |
| "grad_norm": 1.0146456956863403, |
| "learning_rate": 9.064623391643777e-06, |
| "loss": 1.7026, |
| "step": 324500 |
| }, |
| { |
| "epoch": 22.832654208233805, |
| "grad_norm": 1.0264073610305786, |
| "learning_rate": 8.920052045684546e-06, |
| "loss": 1.7013, |
| "step": 325000 |
| }, |
| { |
| "epoch": 22.867781368554166, |
| "grad_norm": 1.0368303060531616, |
| "learning_rate": 8.775480699725315e-06, |
| "loss": 1.7015, |
| "step": 325500 |
| }, |
| { |
| "epoch": 22.902908528874526, |
| "grad_norm": 1.0334751605987549, |
| "learning_rate": 8.630909353766084e-06, |
| "loss": 1.7035, |
| "step": 326000 |
| }, |
| { |
| "epoch": 22.938035689194887, |
| "grad_norm": 1.0076967477798462, |
| "learning_rate": 8.486338007806853e-06, |
| "loss": 1.7003, |
| "step": 326500 |
| }, |
| { |
| "epoch": 22.973162849515244, |
| "grad_norm": 1.0371466875076294, |
| "learning_rate": 8.341766661847622e-06, |
| "loss": 1.696, |
| "step": 327000 |
| }, |
| { |
| "epoch": 23.008290009835605, |
| "grad_norm": 1.0344361066818237, |
| "learning_rate": 8.197195315888391e-06, |
| "loss": 1.7008, |
| "step": 327500 |
| }, |
| { |
| "epoch": 23.043417170155966, |
| "grad_norm": 1.0290839672088623, |
| "learning_rate": 8.05262396992916e-06, |
| "loss": 1.6999, |
| "step": 328000 |
| }, |
| { |
| "epoch": 23.078544330476323, |
| "grad_norm": 1.0254687070846558, |
| "learning_rate": 7.90805262396993e-06, |
| "loss": 1.6964, |
| "step": 328500 |
| }, |
| { |
| "epoch": 23.113671490796683, |
| "grad_norm": 1.0352903604507446, |
| "learning_rate": 7.763481278010698e-06, |
| "loss": 1.6997, |
| "step": 329000 |
| }, |
| { |
| "epoch": 23.148798651117044, |
| "grad_norm": 1.0133601427078247, |
| "learning_rate": 7.618909932051468e-06, |
| "loss": 1.7013, |
| "step": 329500 |
| }, |
| { |
| "epoch": 23.183925811437405, |
| "grad_norm": 1.023796796798706, |
| "learning_rate": 7.474338586092237e-06, |
| "loss": 1.6989, |
| "step": 330000 |
| }, |
| { |
| "epoch": 23.183925811437405, |
| "eval_accuracy": 0.6896140558328754, |
| "eval_loss": 1.429188847541809, |
| "eval_runtime": 123.0195, |
| "eval_samples_per_second": 981.194, |
| "eval_steps_per_second": 6.137, |
| "step": 330000 |
| }, |
| { |
| "epoch": 23.219052971757762, |
| "grad_norm": 1.032958745956421, |
| "learning_rate": 7.3297672401330065e-06, |
| "loss": 1.6951, |
| "step": 330500 |
| }, |
| { |
| "epoch": 23.254180132078123, |
| "grad_norm": 1.0342133045196533, |
| "learning_rate": 7.1851958941737756e-06, |
| "loss": 1.6988, |
| "step": 331000 |
| }, |
| { |
| "epoch": 23.289307292398483, |
| "grad_norm": 0.9886177778244019, |
| "learning_rate": 7.040624548214545e-06, |
| "loss": 1.699, |
| "step": 331500 |
| }, |
| { |
| "epoch": 23.324434452718844, |
| "grad_norm": 1.0002549886703491, |
| "learning_rate": 6.896053202255314e-06, |
| "loss": 1.7021, |
| "step": 332000 |
| }, |
| { |
| "epoch": 23.3595616130392, |
| "grad_norm": 1.0241076946258545, |
| "learning_rate": 6.751481856296083e-06, |
| "loss": 1.6992, |
| "step": 332500 |
| }, |
| { |
| "epoch": 23.39468877335956, |
| "grad_norm": 1.022056221961975, |
| "learning_rate": 6.606910510336851e-06, |
| "loss": 1.696, |
| "step": 333000 |
| }, |
| { |
| "epoch": 23.429815933679922, |
| "grad_norm": 1.0093097686767578, |
| "learning_rate": 6.46233916437762e-06, |
| "loss": 1.6973, |
| "step": 333500 |
| }, |
| { |
| "epoch": 23.46494309400028, |
| "grad_norm": 1.0172024965286255, |
| "learning_rate": 6.317767818418389e-06, |
| "loss": 1.6981, |
| "step": 334000 |
| }, |
| { |
| "epoch": 23.50007025432064, |
| "grad_norm": 1.0267430543899536, |
| "learning_rate": 6.173196472459159e-06, |
| "loss": 1.7, |
| "step": 334500 |
| }, |
| { |
| "epoch": 23.535197414641, |
| "grad_norm": 1.022218942642212, |
| "learning_rate": 6.028625126499928e-06, |
| "loss": 1.7001, |
| "step": 335000 |
| }, |
| { |
| "epoch": 23.57032457496136, |
| "grad_norm": 1.0257142782211304, |
| "learning_rate": 5.884053780540697e-06, |
| "loss": 1.6985, |
| "step": 335500 |
| }, |
| { |
| "epoch": 23.60545173528172, |
| "grad_norm": 1.0375627279281616, |
| "learning_rate": 5.739482434581466e-06, |
| "loss": 1.6962, |
| "step": 336000 |
| }, |
| { |
| "epoch": 23.64057889560208, |
| "grad_norm": 1.0243185758590698, |
| "learning_rate": 5.5949110886222355e-06, |
| "loss": 1.6937, |
| "step": 336500 |
| }, |
| { |
| "epoch": 23.67570605592244, |
| "grad_norm": 1.0458335876464844, |
| "learning_rate": 5.4503397426630046e-06, |
| "loss": 1.6979, |
| "step": 337000 |
| }, |
| { |
| "epoch": 23.7108332162428, |
| "grad_norm": 1.0181182622909546, |
| "learning_rate": 5.305768396703774e-06, |
| "loss": 1.6919, |
| "step": 337500 |
| }, |
| { |
| "epoch": 23.745960376563158, |
| "grad_norm": 1.0168306827545166, |
| "learning_rate": 5.161197050744543e-06, |
| "loss": 1.6949, |
| "step": 338000 |
| }, |
| { |
| "epoch": 23.78108753688352, |
| "grad_norm": 1.0054911375045776, |
| "learning_rate": 5.016625704785312e-06, |
| "loss": 1.6944, |
| "step": 338500 |
| }, |
| { |
| "epoch": 23.81621469720388, |
| "grad_norm": 1.0396686792373657, |
| "learning_rate": 4.872054358826081e-06, |
| "loss": 1.6968, |
| "step": 339000 |
| }, |
| { |
| "epoch": 23.851341857524236, |
| "grad_norm": 1.0173486471176147, |
| "learning_rate": 4.72748301286685e-06, |
| "loss": 1.6961, |
| "step": 339500 |
| }, |
| { |
| "epoch": 23.886469017844597, |
| "grad_norm": 1.0065337419509888, |
| "learning_rate": 4.582911666907619e-06, |
| "loss": 1.6971, |
| "step": 340000 |
| }, |
| { |
| "epoch": 23.886469017844597, |
| "eval_accuracy": 0.6902412566331204, |
| "eval_loss": 1.426208734512329, |
| "eval_runtime": 123.8287, |
| "eval_samples_per_second": 974.782, |
| "eval_steps_per_second": 6.097, |
| "step": 340000 |
| }, |
| { |
| "epoch": 23.921596178164958, |
| "grad_norm": 1.0310503244400024, |
| "learning_rate": 4.438340320948388e-06, |
| "loss": 1.6948, |
| "step": 340500 |
| }, |
| { |
| "epoch": 23.956723338485318, |
| "grad_norm": 0.98621666431427, |
| "learning_rate": 4.293768974989157e-06, |
| "loss": 1.6965, |
| "step": 341000 |
| }, |
| { |
| "epoch": 23.991850498805675, |
| "grad_norm": 0.9949444532394409, |
| "learning_rate": 4.149197629029926e-06, |
| "loss": 1.6963, |
| "step": 341500 |
| }, |
| { |
| "epoch": 24.026977659126036, |
| "grad_norm": 1.0038100481033325, |
| "learning_rate": 4.004626283070695e-06, |
| "loss": 1.6977, |
| "step": 342000 |
| }, |
| { |
| "epoch": 24.062104819446397, |
| "grad_norm": 1.0704007148742676, |
| "learning_rate": 3.8600549371114645e-06, |
| "loss": 1.6954, |
| "step": 342500 |
| }, |
| { |
| "epoch": 24.097231979766757, |
| "grad_norm": 1.0111685991287231, |
| "learning_rate": 3.7154835911522336e-06, |
| "loss": 1.6945, |
| "step": 343000 |
| }, |
| { |
| "epoch": 24.132359140087114, |
| "grad_norm": 1.0191147327423096, |
| "learning_rate": 3.5709122451930026e-06, |
| "loss": 1.6963, |
| "step": 343500 |
| }, |
| { |
| "epoch": 24.167486300407475, |
| "grad_norm": 1.0488834381103516, |
| "learning_rate": 3.426340899233772e-06, |
| "loss": 1.6986, |
| "step": 344000 |
| }, |
| { |
| "epoch": 24.202613460727836, |
| "grad_norm": 1.0078057050704956, |
| "learning_rate": 3.2817695532745412e-06, |
| "loss": 1.6962, |
| "step": 344500 |
| }, |
| { |
| "epoch": 24.237740621048193, |
| "grad_norm": 1.020544409751892, |
| "learning_rate": 3.1371982073153103e-06, |
| "loss": 1.6982, |
| "step": 345000 |
| }, |
| { |
| "epoch": 24.272867781368554, |
| "grad_norm": 1.0428537130355835, |
| "learning_rate": 2.9926268613560794e-06, |
| "loss": 1.6956, |
| "step": 345500 |
| }, |
| { |
| "epoch": 24.307994941688914, |
| "grad_norm": 1.0283273458480835, |
| "learning_rate": 2.8480555153968485e-06, |
| "loss": 1.6941, |
| "step": 346000 |
| }, |
| { |
| "epoch": 24.343122102009275, |
| "grad_norm": 1.0288498401641846, |
| "learning_rate": 2.7034841694376176e-06, |
| "loss": 1.6931, |
| "step": 346500 |
| }, |
| { |
| "epoch": 24.378249262329632, |
| "grad_norm": 1.0071648359298706, |
| "learning_rate": 2.5589128234783866e-06, |
| "loss": 1.695, |
| "step": 347000 |
| }, |
| { |
| "epoch": 24.413376422649993, |
| "grad_norm": 1.0438432693481445, |
| "learning_rate": 2.414341477519156e-06, |
| "loss": 1.6938, |
| "step": 347500 |
| }, |
| { |
| "epoch": 24.448503582970353, |
| "grad_norm": 1.0068647861480713, |
| "learning_rate": 2.269770131559925e-06, |
| "loss": 1.6952, |
| "step": 348000 |
| }, |
| { |
| "epoch": 24.483630743290714, |
| "grad_norm": 1.0173580646514893, |
| "learning_rate": 2.125198785600694e-06, |
| "loss": 1.6929, |
| "step": 348500 |
| }, |
| { |
| "epoch": 24.51875790361107, |
| "grad_norm": 1.0046675205230713, |
| "learning_rate": 1.980627439641463e-06, |
| "loss": 1.6917, |
| "step": 349000 |
| }, |
| { |
| "epoch": 24.553885063931432, |
| "grad_norm": 1.0160244703292847, |
| "learning_rate": 1.8360560936822323e-06, |
| "loss": 1.6916, |
| "step": 349500 |
| }, |
| { |
| "epoch": 24.589012224251793, |
| "grad_norm": 1.0335299968719482, |
| "learning_rate": 1.6914847477230013e-06, |
| "loss": 1.6958, |
| "step": 350000 |
| }, |
| { |
| "epoch": 24.589012224251793, |
| "eval_accuracy": 0.6906203824669922, |
| "eval_loss": 1.4238797426223755, |
| "eval_runtime": 122.2634, |
| "eval_samples_per_second": 987.262, |
| "eval_steps_per_second": 6.175, |
| "step": 350000 |
| }, |
| { |
| "epoch": 24.62413938457215, |
| "grad_norm": 1.0259653329849243, |
| "learning_rate": 1.5469134017637704e-06, |
| "loss": 1.6938, |
| "step": 350500 |
| }, |
| { |
| "epoch": 24.65926654489251, |
| "grad_norm": 1.0362039804458618, |
| "learning_rate": 1.4023420558045395e-06, |
| "loss": 1.6923, |
| "step": 351000 |
| }, |
| { |
| "epoch": 24.69439370521287, |
| "grad_norm": 1.0353460311889648, |
| "learning_rate": 1.2577707098453088e-06, |
| "loss": 1.6934, |
| "step": 351500 |
| }, |
| { |
| "epoch": 24.72952086553323, |
| "grad_norm": 1.044894814491272, |
| "learning_rate": 1.1131993638860779e-06, |
| "loss": 1.695, |
| "step": 352000 |
| }, |
| { |
| "epoch": 24.76464802585359, |
| "grad_norm": 1.0238316059112549, |
| "learning_rate": 9.68628017926847e-07, |
| "loss": 1.6936, |
| "step": 352500 |
| }, |
| { |
| "epoch": 24.79977518617395, |
| "grad_norm": 1.008371114730835, |
| "learning_rate": 8.240566719676161e-07, |
| "loss": 1.6938, |
| "step": 353000 |
| }, |
| { |
| "epoch": 24.83490234649431, |
| "grad_norm": 1.0096451044082642, |
| "learning_rate": 6.794853260083851e-07, |
| "loss": 1.693, |
| "step": 353500 |
| }, |
| { |
| "epoch": 24.870029506814667, |
| "grad_norm": 1.0062962770462036, |
| "learning_rate": 5.349139800491543e-07, |
| "loss": 1.6943, |
| "step": 354000 |
| }, |
| { |
| "epoch": 24.905156667135028, |
| "grad_norm": 1.0379866361618042, |
| "learning_rate": 3.903426340899234e-07, |
| "loss": 1.6945, |
| "step": 354500 |
| }, |
| { |
| "epoch": 24.94028382745539, |
| "grad_norm": 1.0395712852478027, |
| "learning_rate": 2.457712881306925e-07, |
| "loss": 1.6903, |
| "step": 355000 |
| }, |
| { |
| "epoch": 24.97541098777575, |
| "grad_norm": 0.9931861162185669, |
| "learning_rate": 1.0119994217146162e-07, |
| "loss": 1.6939, |
| "step": 355500 |
| }, |
| { |
| "epoch": 25.0, |
| "step": 355850, |
| "total_flos": 2.991340735135027e+18, |
| "train_loss": 2.147127732371654, |
| "train_runtime": 88863.3111, |
| "train_samples_per_second": 640.687, |
| "train_steps_per_second": 4.004 |
| } |
| ], |
| "logging_steps": 500, |
| "max_steps": 355850, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 25, |
| "save_steps": 10000, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": true |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 2.991340735135027e+18, |
| "train_batch_size": 160, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
