| { |
| "best_metric": 1.202436089515686, |
| "best_model_checkpoint": "distilbert_base_train_book/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": 1.9654814004898071, |
| "learning_rate": 5e-06, |
| "loss": 8.8875, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.0702543206407194, |
| "grad_norm": 0.728134274482727, |
| "learning_rate": 1e-05, |
| "loss": 6.5828, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.1053814809610791, |
| "grad_norm": 0.6658562421798706, |
| "learning_rate": 1.5e-05, |
| "loss": 6.0119, |
| "step": 1500 |
| }, |
| { |
| "epoch": 0.1405086412814388, |
| "grad_norm": 0.691459596157074, |
| "learning_rate": 2e-05, |
| "loss": 5.8991, |
| "step": 2000 |
| }, |
| { |
| "epoch": 0.1756358016017985, |
| "grad_norm": 0.6991929411888123, |
| "learning_rate": 2.5e-05, |
| "loss": 5.8401, |
| "step": 2500 |
| }, |
| { |
| "epoch": 0.2107629619221582, |
| "grad_norm": 0.594689667224884, |
| "learning_rate": 3e-05, |
| "loss": 5.7955, |
| "step": 3000 |
| }, |
| { |
| "epoch": 0.2458901222425179, |
| "grad_norm": 0.49550116062164307, |
| "learning_rate": 3.5e-05, |
| "loss": 5.7665, |
| "step": 3500 |
| }, |
| { |
| "epoch": 0.2810172825628776, |
| "grad_norm": 0.48818740248680115, |
| "learning_rate": 4e-05, |
| "loss": 5.7418, |
| "step": 4000 |
| }, |
| { |
| "epoch": 0.31614444288323734, |
| "grad_norm": 0.5408225655555725, |
| "learning_rate": 4.5e-05, |
| "loss": 5.7212, |
| "step": 4500 |
| }, |
| { |
| "epoch": 0.351271603203597, |
| "grad_norm": 0.5848438739776611, |
| "learning_rate": 5e-05, |
| "loss": 5.7049, |
| "step": 5000 |
| }, |
| { |
| "epoch": 0.38639876352395675, |
| "grad_norm": 0.5231767296791077, |
| "learning_rate": 5.500000000000001e-05, |
| "loss": 5.6856, |
| "step": 5500 |
| }, |
| { |
| "epoch": 0.4215259238443164, |
| "grad_norm": 0.4542430639266968, |
| "learning_rate": 6e-05, |
| "loss": 5.6762, |
| "step": 6000 |
| }, |
| { |
| "epoch": 0.45665308416467615, |
| "grad_norm": 0.5210428833961487, |
| "learning_rate": 6.500000000000001e-05, |
| "loss": 5.6635, |
| "step": 6500 |
| }, |
| { |
| "epoch": 0.4917802444850358, |
| "grad_norm": 0.5147435665130615, |
| "learning_rate": 7e-05, |
| "loss": 5.6566, |
| "step": 7000 |
| }, |
| { |
| "epoch": 0.5269074048053956, |
| "grad_norm": 0.5612876415252686, |
| "learning_rate": 7.500000000000001e-05, |
| "loss": 5.6469, |
| "step": 7500 |
| }, |
| { |
| "epoch": 0.5620345651257552, |
| "grad_norm": 0.554747462272644, |
| "learning_rate": 8e-05, |
| "loss": 5.6365, |
| "step": 8000 |
| }, |
| { |
| "epoch": 0.5971617254461149, |
| "grad_norm": 0.6292888522148132, |
| "learning_rate": 8.5e-05, |
| "loss": 5.627, |
| "step": 8500 |
| }, |
| { |
| "epoch": 0.6322888857664747, |
| "grad_norm": 0.6702057719230652, |
| "learning_rate": 9e-05, |
| "loss": 5.6161, |
| "step": 9000 |
| }, |
| { |
| "epoch": 0.6674160460868344, |
| "grad_norm": 0.4405311048030853, |
| "learning_rate": 9.5e-05, |
| "loss": 5.6112, |
| "step": 9500 |
| }, |
| { |
| "epoch": 0.702543206407194, |
| "grad_norm": 0.4621039927005768, |
| "learning_rate": 0.0001, |
| "loss": 5.6039, |
| "step": 10000 |
| }, |
| { |
| "epoch": 0.702543206407194, |
| "eval_accuracy": 0.1652548912175523, |
| "eval_loss": 5.450643062591553, |
| "eval_runtime": 156.0549, |
| "eval_samples_per_second": 773.484, |
| "eval_steps_per_second": 4.838, |
| "step": 10000 |
| }, |
| { |
| "epoch": 0.7376703667275537, |
| "grad_norm": 0.517584502696991, |
| "learning_rate": 9.985542865404077e-05, |
| "loss": 5.5964, |
| "step": 10500 |
| }, |
| { |
| "epoch": 0.7727975270479135, |
| "grad_norm": 0.4528399705886841, |
| "learning_rate": 9.971085730808155e-05, |
| "loss": 5.5891, |
| "step": 11000 |
| }, |
| { |
| "epoch": 0.8079246873682732, |
| "grad_norm": 0.476567804813385, |
| "learning_rate": 9.956628596212231e-05, |
| "loss": 5.5855, |
| "step": 11500 |
| }, |
| { |
| "epoch": 0.8430518476886328, |
| "grad_norm": 0.4747626483440399, |
| "learning_rate": 9.942171461616309e-05, |
| "loss": 5.5763, |
| "step": 12000 |
| }, |
| { |
| "epoch": 0.8781790080089925, |
| "grad_norm": 0.4742708206176758, |
| "learning_rate": 9.927714327020385e-05, |
| "loss": 5.571, |
| "step": 12500 |
| }, |
| { |
| "epoch": 0.9133061683293523, |
| "grad_norm": 0.4602668285369873, |
| "learning_rate": 9.913257192424462e-05, |
| "loss": 5.5608, |
| "step": 13000 |
| }, |
| { |
| "epoch": 0.948433328649712, |
| "grad_norm": 0.5417143106460571, |
| "learning_rate": 9.89880005782854e-05, |
| "loss": 5.5541, |
| "step": 13500 |
| }, |
| { |
| "epoch": 0.9835604889700716, |
| "grad_norm": 0.4820058345794678, |
| "learning_rate": 9.884342923232616e-05, |
| "loss": 5.5499, |
| "step": 14000 |
| }, |
| { |
| "epoch": 1.0186876492904313, |
| "grad_norm": 0.44978609681129456, |
| "learning_rate": 9.869885788636692e-05, |
| "loss": 5.5414, |
| "step": 14500 |
| }, |
| { |
| "epoch": 1.053814809610791, |
| "grad_norm": 0.4410560429096222, |
| "learning_rate": 9.855428654040769e-05, |
| "loss": 5.5374, |
| "step": 15000 |
| }, |
| { |
| "epoch": 1.0889419699311507, |
| "grad_norm": 0.47571223974227905, |
| "learning_rate": 9.840971519444846e-05, |
| "loss": 5.5298, |
| "step": 15500 |
| }, |
| { |
| "epoch": 1.1240691302515105, |
| "grad_norm": 0.49150145053863525, |
| "learning_rate": 9.826514384848924e-05, |
| "loss": 5.5255, |
| "step": 16000 |
| }, |
| { |
| "epoch": 1.1591962905718702, |
| "grad_norm": 0.4894424080848694, |
| "learning_rate": 9.812057250253e-05, |
| "loss": 5.5201, |
| "step": 16500 |
| }, |
| { |
| "epoch": 1.1943234508922298, |
| "grad_norm": 0.4323597550392151, |
| "learning_rate": 9.797600115657077e-05, |
| "loss": 5.5167, |
| "step": 17000 |
| }, |
| { |
| "epoch": 1.2294506112125896, |
| "grad_norm": 0.5220290422439575, |
| "learning_rate": 9.783142981061155e-05, |
| "loss": 5.5086, |
| "step": 17500 |
| }, |
| { |
| "epoch": 1.2645777715329491, |
| "grad_norm": 0.46441909670829773, |
| "learning_rate": 9.768685846465231e-05, |
| "loss": 5.507, |
| "step": 18000 |
| }, |
| { |
| "epoch": 1.299704931853309, |
| "grad_norm": 0.6826435923576355, |
| "learning_rate": 9.754228711869307e-05, |
| "loss": 5.4763, |
| "step": 18500 |
| }, |
| { |
| "epoch": 1.3348320921736687, |
| "grad_norm": 1.1718729734420776, |
| "learning_rate": 9.739771577273384e-05, |
| "loss": 5.3617, |
| "step": 19000 |
| }, |
| { |
| "epoch": 1.3699592524940285, |
| "grad_norm": 1.7364224195480347, |
| "learning_rate": 9.725314442677463e-05, |
| "loss": 5.1371, |
| "step": 19500 |
| }, |
| { |
| "epoch": 1.405086412814388, |
| "grad_norm": 2.350893259048462, |
| "learning_rate": 9.71085730808154e-05, |
| "loss": 4.4684, |
| "step": 20000 |
| }, |
| { |
| "epoch": 1.405086412814388, |
| "eval_accuracy": 0.3849075452860559, |
| "eval_loss": 3.745048761367798, |
| "eval_runtime": 171.6019, |
| "eval_samples_per_second": 703.407, |
| "eval_steps_per_second": 4.4, |
| "step": 20000 |
| }, |
| { |
| "epoch": 1.4402135731347478, |
| "grad_norm": 1.7915072441101074, |
| "learning_rate": 9.696400173485616e-05, |
| "loss": 3.9932, |
| "step": 20500 |
| }, |
| { |
| "epoch": 1.4753407334551074, |
| "grad_norm": 1.3795692920684814, |
| "learning_rate": 9.681943038889692e-05, |
| "loss": 3.5874, |
| "step": 21000 |
| }, |
| { |
| "epoch": 1.5104678937754672, |
| "grad_norm": 1.3643194437026978, |
| "learning_rate": 9.66748590429377e-05, |
| "loss": 3.3011, |
| "step": 21500 |
| }, |
| { |
| "epoch": 1.545595054095827, |
| "grad_norm": 1.1432992219924927, |
| "learning_rate": 9.653028769697846e-05, |
| "loss": 3.1295, |
| "step": 22000 |
| }, |
| { |
| "epoch": 1.5807222144161865, |
| "grad_norm": 1.1360324621200562, |
| "learning_rate": 9.638571635101923e-05, |
| "loss": 3.0156, |
| "step": 22500 |
| }, |
| { |
| "epoch": 1.6158493747365463, |
| "grad_norm": 1.045045018196106, |
| "learning_rate": 9.624114500506e-05, |
| "loss": 2.9242, |
| "step": 23000 |
| }, |
| { |
| "epoch": 1.650976535056906, |
| "grad_norm": 1.0264136791229248, |
| "learning_rate": 9.609657365910077e-05, |
| "loss": 2.8484, |
| "step": 23500 |
| }, |
| { |
| "epoch": 1.6861036953772657, |
| "grad_norm": 1.0071278810501099, |
| "learning_rate": 9.595200231314155e-05, |
| "loss": 2.7804, |
| "step": 24000 |
| }, |
| { |
| "epoch": 1.7212308556976255, |
| "grad_norm": 1.0051913261413574, |
| "learning_rate": 9.580743096718231e-05, |
| "loss": 2.727, |
| "step": 24500 |
| }, |
| { |
| "epoch": 1.7563580160179852, |
| "grad_norm": 1.0014828443527222, |
| "learning_rate": 9.566285962122307e-05, |
| "loss": 2.6771, |
| "step": 25000 |
| }, |
| { |
| "epoch": 1.7914851763383448, |
| "grad_norm": 0.99198979139328, |
| "learning_rate": 9.551828827526385e-05, |
| "loss": 2.6332, |
| "step": 25500 |
| }, |
| { |
| "epoch": 1.8266123366587044, |
| "grad_norm": 0.9274836778640747, |
| "learning_rate": 9.537371692930462e-05, |
| "loss": 2.5896, |
| "step": 26000 |
| }, |
| { |
| "epoch": 1.8617394969790642, |
| "grad_norm": 1.0025986433029175, |
| "learning_rate": 9.522914558334538e-05, |
| "loss": 2.5482, |
| "step": 26500 |
| }, |
| { |
| "epoch": 1.896866657299424, |
| "grad_norm": 0.9490007162094116, |
| "learning_rate": 9.508457423738616e-05, |
| "loss": 2.5157, |
| "step": 27000 |
| }, |
| { |
| "epoch": 1.9319938176197837, |
| "grad_norm": 0.9618148803710938, |
| "learning_rate": 9.494000289142692e-05, |
| "loss": 2.4826, |
| "step": 27500 |
| }, |
| { |
| "epoch": 1.9671209779401433, |
| "grad_norm": 0.945733368396759, |
| "learning_rate": 9.47954315454677e-05, |
| "loss": 2.4494, |
| "step": 28000 |
| }, |
| { |
| "epoch": 2.002248138260503, |
| "grad_norm": 0.9726658463478088, |
| "learning_rate": 9.465086019950846e-05, |
| "loss": 2.4297, |
| "step": 28500 |
| }, |
| { |
| "epoch": 2.0373752985808626, |
| "grad_norm": 0.9335887432098389, |
| "learning_rate": 9.450628885354923e-05, |
| "loss": 2.3963, |
| "step": 29000 |
| }, |
| { |
| "epoch": 2.0725024589012224, |
| "grad_norm": 0.9659011960029602, |
| "learning_rate": 9.436171750758999e-05, |
| "loss": 2.3715, |
| "step": 29500 |
| }, |
| { |
| "epoch": 2.107629619221582, |
| "grad_norm": 0.9352829456329346, |
| "learning_rate": 9.421714616163077e-05, |
| "loss": 2.3547, |
| "step": 30000 |
| }, |
| { |
| "epoch": 2.107629619221582, |
| "eval_accuracy": 0.5926471710257318, |
| "eval_loss": 2.044137477874756, |
| "eval_runtime": 159.9246, |
| "eval_samples_per_second": 754.768, |
| "eval_steps_per_second": 4.721, |
| "step": 30000 |
| }, |
| { |
| "epoch": 2.142756779541942, |
| "grad_norm": 0.9466668963432312, |
| "learning_rate": 9.407257481567155e-05, |
| "loss": 2.3333, |
| "step": 30500 |
| }, |
| { |
| "epoch": 2.1778839398623013, |
| "grad_norm": 0.9470064640045166, |
| "learning_rate": 9.392800346971231e-05, |
| "loss": 2.3114, |
| "step": 31000 |
| }, |
| { |
| "epoch": 2.213011100182661, |
| "grad_norm": 0.9255052208900452, |
| "learning_rate": 9.378343212375307e-05, |
| "loss": 2.2956, |
| "step": 31500 |
| }, |
| { |
| "epoch": 2.248138260503021, |
| "grad_norm": 0.9332784414291382, |
| "learning_rate": 9.363886077779385e-05, |
| "loss": 2.278, |
| "step": 32000 |
| }, |
| { |
| "epoch": 2.2832654208233807, |
| "grad_norm": 0.8893001079559326, |
| "learning_rate": 9.349428943183462e-05, |
| "loss": 2.2614, |
| "step": 32500 |
| }, |
| { |
| "epoch": 2.3183925811437405, |
| "grad_norm": 0.9374568462371826, |
| "learning_rate": 9.334971808587538e-05, |
| "loss": 2.2429, |
| "step": 33000 |
| }, |
| { |
| "epoch": 2.3535197414641003, |
| "grad_norm": 0.922341525554657, |
| "learning_rate": 9.320514673991614e-05, |
| "loss": 2.2282, |
| "step": 33500 |
| }, |
| { |
| "epoch": 2.3886469017844596, |
| "grad_norm": 0.9099931716918945, |
| "learning_rate": 9.306057539395692e-05, |
| "loss": 2.2156, |
| "step": 34000 |
| }, |
| { |
| "epoch": 2.4237740621048194, |
| "grad_norm": 0.8796118497848511, |
| "learning_rate": 9.29160040479977e-05, |
| "loss": 2.1989, |
| "step": 34500 |
| }, |
| { |
| "epoch": 2.458901222425179, |
| "grad_norm": 0.900063157081604, |
| "learning_rate": 9.277143270203846e-05, |
| "loss": 2.1873, |
| "step": 35000 |
| }, |
| { |
| "epoch": 2.494028382745539, |
| "grad_norm": 0.9187572002410889, |
| "learning_rate": 9.262686135607923e-05, |
| "loss": 2.177, |
| "step": 35500 |
| }, |
| { |
| "epoch": 2.5291555430658983, |
| "grad_norm": 0.8697600364685059, |
| "learning_rate": 9.248229001012e-05, |
| "loss": 2.164, |
| "step": 36000 |
| }, |
| { |
| "epoch": 2.564282703386258, |
| "grad_norm": 0.891128420829773, |
| "learning_rate": 9.233771866416077e-05, |
| "loss": 2.151, |
| "step": 36500 |
| }, |
| { |
| "epoch": 2.599409863706618, |
| "grad_norm": 0.8724724650382996, |
| "learning_rate": 9.219314731820153e-05, |
| "loss": 2.139, |
| "step": 37000 |
| }, |
| { |
| "epoch": 2.6345370240269776, |
| "grad_norm": 0.90709388256073, |
| "learning_rate": 9.20485759722423e-05, |
| "loss": 2.1304, |
| "step": 37500 |
| }, |
| { |
| "epoch": 2.6696641843473374, |
| "grad_norm": 0.9295563697814941, |
| "learning_rate": 9.190400462628307e-05, |
| "loss": 2.1174, |
| "step": 38000 |
| }, |
| { |
| "epoch": 2.704791344667697, |
| "grad_norm": 0.8665570020675659, |
| "learning_rate": 9.175943328032385e-05, |
| "loss": 2.1087, |
| "step": 38500 |
| }, |
| { |
| "epoch": 2.739918504988057, |
| "grad_norm": 0.8724929690361023, |
| "learning_rate": 9.161486193436461e-05, |
| "loss": 2.0976, |
| "step": 39000 |
| }, |
| { |
| "epoch": 2.7750456653084163, |
| "grad_norm": 0.8766260147094727, |
| "learning_rate": 9.147029058840538e-05, |
| "loss": 2.0894, |
| "step": 39500 |
| }, |
| { |
| "epoch": 2.810172825628776, |
| "grad_norm": 0.9151031374931335, |
| "learning_rate": 9.132571924244616e-05, |
| "loss": 2.0785, |
| "step": 40000 |
| }, |
| { |
| "epoch": 2.810172825628776, |
| "eval_accuracy": 0.6308508733347783, |
| "eval_loss": 1.7986087799072266, |
| "eval_runtime": 167.2706, |
| "eval_samples_per_second": 721.621, |
| "eval_steps_per_second": 4.514, |
| "step": 40000 |
| }, |
| { |
| "epoch": 2.845299985949136, |
| "grad_norm": 0.896449863910675, |
| "learning_rate": 9.118114789648692e-05, |
| "loss": 2.0673, |
| "step": 40500 |
| }, |
| { |
| "epoch": 2.8804271462694957, |
| "grad_norm": 0.8919656872749329, |
| "learning_rate": 9.103657655052768e-05, |
| "loss": 2.0589, |
| "step": 41000 |
| }, |
| { |
| "epoch": 2.915554306589855, |
| "grad_norm": 0.9112071394920349, |
| "learning_rate": 9.089200520456845e-05, |
| "loss": 2.051, |
| "step": 41500 |
| }, |
| { |
| "epoch": 2.950681466910215, |
| "grad_norm": 0.857875645160675, |
| "learning_rate": 9.074743385860923e-05, |
| "loss": 2.0446, |
| "step": 42000 |
| }, |
| { |
| "epoch": 2.9858086272305746, |
| "grad_norm": 0.854753851890564, |
| "learning_rate": 9.060286251265e-05, |
| "loss": 2.0366, |
| "step": 42500 |
| }, |
| { |
| "epoch": 3.0209357875509344, |
| "grad_norm": 0.8534505367279053, |
| "learning_rate": 9.045829116669077e-05, |
| "loss": 2.0218, |
| "step": 43000 |
| }, |
| { |
| "epoch": 3.056062947871294, |
| "grad_norm": 0.8935545086860657, |
| "learning_rate": 9.031371982073153e-05, |
| "loss": 2.0175, |
| "step": 43500 |
| }, |
| { |
| "epoch": 3.091190108191654, |
| "grad_norm": 0.855864405632019, |
| "learning_rate": 9.016914847477231e-05, |
| "loss": 2.0144, |
| "step": 44000 |
| }, |
| { |
| "epoch": 3.1263172685120133, |
| "grad_norm": 0.8761529326438904, |
| "learning_rate": 9.002457712881307e-05, |
| "loss": 2.0036, |
| "step": 44500 |
| }, |
| { |
| "epoch": 3.161444428832373, |
| "grad_norm": 0.8668981790542603, |
| "learning_rate": 8.988000578285384e-05, |
| "loss": 1.9982, |
| "step": 45000 |
| }, |
| { |
| "epoch": 3.196571589152733, |
| "grad_norm": 0.8865150809288025, |
| "learning_rate": 8.973543443689461e-05, |
| "loss": 1.9885, |
| "step": 45500 |
| }, |
| { |
| "epoch": 3.2316987494730927, |
| "grad_norm": 0.8691010475158691, |
| "learning_rate": 8.959086309093538e-05, |
| "loss": 1.986, |
| "step": 46000 |
| }, |
| { |
| "epoch": 3.2668259097934524, |
| "grad_norm": 0.8647297620773315, |
| "learning_rate": 8.944629174497616e-05, |
| "loss": 1.9779, |
| "step": 46500 |
| }, |
| { |
| "epoch": 3.301953070113812, |
| "grad_norm": 0.8569030165672302, |
| "learning_rate": 8.930172039901692e-05, |
| "loss": 1.9715, |
| "step": 47000 |
| }, |
| { |
| "epoch": 3.3370802304341716, |
| "grad_norm": 0.8286603093147278, |
| "learning_rate": 8.915714905305768e-05, |
| "loss": 1.9653, |
| "step": 47500 |
| }, |
| { |
| "epoch": 3.3722073907545314, |
| "grad_norm": 0.9221807718276978, |
| "learning_rate": 8.901257770709845e-05, |
| "loss": 1.9612, |
| "step": 48000 |
| }, |
| { |
| "epoch": 3.407334551074891, |
| "grad_norm": 0.8589962124824524, |
| "learning_rate": 8.886800636113923e-05, |
| "loss": 1.9529, |
| "step": 48500 |
| }, |
| { |
| "epoch": 3.442461711395251, |
| "grad_norm": 0.8679049611091614, |
| "learning_rate": 8.872343501518e-05, |
| "loss": 1.9461, |
| "step": 49000 |
| }, |
| { |
| "epoch": 3.4775888717156107, |
| "grad_norm": 0.8534972071647644, |
| "learning_rate": 8.857886366922077e-05, |
| "loss": 1.9397, |
| "step": 49500 |
| }, |
| { |
| "epoch": 3.51271603203597, |
| "grad_norm": 0.8502809405326843, |
| "learning_rate": 8.843429232326153e-05, |
| "loss": 1.938, |
| "step": 50000 |
| }, |
| { |
| "epoch": 3.51271603203597, |
| "eval_accuracy": 0.6520104559976765, |
| "eval_loss": 1.6649836301803589, |
| "eval_runtime": 159.9891, |
| "eval_samples_per_second": 754.464, |
| "eval_steps_per_second": 4.719, |
| "step": 50000 |
| }, |
| { |
| "epoch": 3.54784319235633, |
| "grad_norm": 0.896166205406189, |
| "learning_rate": 8.828972097730231e-05, |
| "loss": 1.9343, |
| "step": 50500 |
| }, |
| { |
| "epoch": 3.5829703526766896, |
| "grad_norm": 0.8760735392570496, |
| "learning_rate": 8.814514963134307e-05, |
| "loss": 1.9265, |
| "step": 51000 |
| }, |
| { |
| "epoch": 3.6180975129970494, |
| "grad_norm": 0.8405304551124573, |
| "learning_rate": 8.800057828538384e-05, |
| "loss": 1.9241, |
| "step": 51500 |
| }, |
| { |
| "epoch": 3.6532246733174087, |
| "grad_norm": 0.8592715859413147, |
| "learning_rate": 8.78560069394246e-05, |
| "loss": 1.9182, |
| "step": 52000 |
| }, |
| { |
| "epoch": 3.6883518336377685, |
| "grad_norm": 0.8587563633918762, |
| "learning_rate": 8.771143559346538e-05, |
| "loss": 1.9113, |
| "step": 52500 |
| }, |
| { |
| "epoch": 3.7234789939581283, |
| "grad_norm": 0.835717499256134, |
| "learning_rate": 8.756686424750616e-05, |
| "loss": 1.9102, |
| "step": 53000 |
| }, |
| { |
| "epoch": 3.758606154278488, |
| "grad_norm": 0.8733358383178711, |
| "learning_rate": 8.742229290154692e-05, |
| "loss": 1.9044, |
| "step": 53500 |
| }, |
| { |
| "epoch": 3.793733314598848, |
| "grad_norm": 0.8711694478988647, |
| "learning_rate": 8.727772155558768e-05, |
| "loss": 1.8976, |
| "step": 54000 |
| }, |
| { |
| "epoch": 3.8288604749192077, |
| "grad_norm": 0.8459246158599854, |
| "learning_rate": 8.713315020962846e-05, |
| "loss": 1.8954, |
| "step": 54500 |
| }, |
| { |
| "epoch": 3.8639876352395675, |
| "grad_norm": 0.8664897680282593, |
| "learning_rate": 8.698857886366923e-05, |
| "loss": 1.8882, |
| "step": 55000 |
| }, |
| { |
| "epoch": 3.899114795559927, |
| "grad_norm": 0.845977783203125, |
| "learning_rate": 8.684400751770999e-05, |
| "loss": 1.8838, |
| "step": 55500 |
| }, |
| { |
| "epoch": 3.9342419558802866, |
| "grad_norm": 0.8510854244232178, |
| "learning_rate": 8.669943617175075e-05, |
| "loss": 1.8804, |
| "step": 56000 |
| }, |
| { |
| "epoch": 3.9693691162006464, |
| "grad_norm": 0.8640584945678711, |
| "learning_rate": 8.655486482579153e-05, |
| "loss": 1.8779, |
| "step": 56500 |
| }, |
| { |
| "epoch": 4.004496276521006, |
| "grad_norm": 0.8487034440040588, |
| "learning_rate": 8.641029347983231e-05, |
| "loss": 1.8714, |
| "step": 57000 |
| }, |
| { |
| "epoch": 4.0396234368413655, |
| "grad_norm": 0.8744542598724365, |
| "learning_rate": 8.626572213387307e-05, |
| "loss": 1.863, |
| "step": 57500 |
| }, |
| { |
| "epoch": 4.074750597161725, |
| "grad_norm": 0.8659520745277405, |
| "learning_rate": 8.612115078791384e-05, |
| "loss": 1.8612, |
| "step": 58000 |
| }, |
| { |
| "epoch": 4.109877757482085, |
| "grad_norm": 0.8496120572090149, |
| "learning_rate": 8.597657944195461e-05, |
| "loss": 1.8582, |
| "step": 58500 |
| }, |
| { |
| "epoch": 4.145004917802445, |
| "grad_norm": 0.8670371174812317, |
| "learning_rate": 8.583200809599538e-05, |
| "loss": 1.8534, |
| "step": 59000 |
| }, |
| { |
| "epoch": 4.180132078122805, |
| "grad_norm": 0.8423211574554443, |
| "learning_rate": 8.568743675003614e-05, |
| "loss": 1.8539, |
| "step": 59500 |
| }, |
| { |
| "epoch": 4.215259238443164, |
| "grad_norm": 0.8854081630706787, |
| "learning_rate": 8.55428654040769e-05, |
| "loss": 1.8476, |
| "step": 60000 |
| }, |
| { |
| "epoch": 4.215259238443164, |
| "eval_accuracy": 0.6647783402761954, |
| "eval_loss": 1.5861711502075195, |
| "eval_runtime": 159.2871, |
| "eval_samples_per_second": 757.789, |
| "eval_steps_per_second": 4.74, |
| "step": 60000 |
| }, |
| { |
| "epoch": 4.250386398763524, |
| "grad_norm": 0.867537260055542, |
| "learning_rate": 8.539829405811768e-05, |
| "loss": 1.847, |
| "step": 60500 |
| }, |
| { |
| "epoch": 4.285513559083884, |
| "grad_norm": 0.8403070569038391, |
| "learning_rate": 8.525372271215846e-05, |
| "loss": 1.8402, |
| "step": 61000 |
| }, |
| { |
| "epoch": 4.320640719404244, |
| "grad_norm": 0.8398751616477966, |
| "learning_rate": 8.510915136619923e-05, |
| "loss": 1.8351, |
| "step": 61500 |
| }, |
| { |
| "epoch": 4.355767879724603, |
| "grad_norm": 0.8706727623939514, |
| "learning_rate": 8.496458002023999e-05, |
| "loss": 1.8333, |
| "step": 62000 |
| }, |
| { |
| "epoch": 4.3908950400449624, |
| "grad_norm": 0.8499593138694763, |
| "learning_rate": 8.482000867428075e-05, |
| "loss": 1.8353, |
| "step": 62500 |
| }, |
| { |
| "epoch": 4.426022200365322, |
| "grad_norm": 0.8617079854011536, |
| "learning_rate": 8.467543732832153e-05, |
| "loss": 1.83, |
| "step": 63000 |
| }, |
| { |
| "epoch": 4.461149360685682, |
| "grad_norm": 0.8642334938049316, |
| "learning_rate": 8.45308659823623e-05, |
| "loss": 1.8244, |
| "step": 63500 |
| }, |
| { |
| "epoch": 4.496276521006042, |
| "grad_norm": 0.9034135937690735, |
| "learning_rate": 8.438629463640307e-05, |
| "loss": 1.8225, |
| "step": 64000 |
| }, |
| { |
| "epoch": 4.531403681326402, |
| "grad_norm": 0.8784350156784058, |
| "learning_rate": 8.424172329044384e-05, |
| "loss": 1.8195, |
| "step": 64500 |
| }, |
| { |
| "epoch": 4.566530841646761, |
| "grad_norm": 0.859846293926239, |
| "learning_rate": 8.409715194448461e-05, |
| "loss": 1.8163, |
| "step": 65000 |
| }, |
| { |
| "epoch": 4.601658001967121, |
| "grad_norm": 0.8535738587379456, |
| "learning_rate": 8.395258059852538e-05, |
| "loss": 1.814, |
| "step": 65500 |
| }, |
| { |
| "epoch": 4.636785162287481, |
| "grad_norm": 0.8490976691246033, |
| "learning_rate": 8.380800925256614e-05, |
| "loss": 1.8105, |
| "step": 66000 |
| }, |
| { |
| "epoch": 4.671912322607841, |
| "grad_norm": 0.8383164405822754, |
| "learning_rate": 8.36634379066069e-05, |
| "loss": 1.8098, |
| "step": 66500 |
| }, |
| { |
| "epoch": 4.7070394829282005, |
| "grad_norm": 0.8779833316802979, |
| "learning_rate": 8.351886656064768e-05, |
| "loss": 1.8032, |
| "step": 67000 |
| }, |
| { |
| "epoch": 4.742166643248559, |
| "grad_norm": 0.8833803534507751, |
| "learning_rate": 8.337429521468846e-05, |
| "loss": 1.8018, |
| "step": 67500 |
| }, |
| { |
| "epoch": 4.777293803568919, |
| "grad_norm": 0.8527798652648926, |
| "learning_rate": 8.322972386872923e-05, |
| "loss": 1.8009, |
| "step": 68000 |
| }, |
| { |
| "epoch": 4.812420963889279, |
| "grad_norm": 0.8733193278312683, |
| "learning_rate": 8.308515252276999e-05, |
| "loss": 1.7989, |
| "step": 68500 |
| }, |
| { |
| "epoch": 4.847548124209639, |
| "grad_norm": 0.8703482151031494, |
| "learning_rate": 8.294058117681077e-05, |
| "loss": 1.7924, |
| "step": 69000 |
| }, |
| { |
| "epoch": 4.8826752845299985, |
| "grad_norm": 0.8766271471977234, |
| "learning_rate": 8.279600983085153e-05, |
| "loss": 1.7922, |
| "step": 69500 |
| }, |
| { |
| "epoch": 4.917802444850358, |
| "grad_norm": 0.8780910968780518, |
| "learning_rate": 8.26514384848923e-05, |
| "loss": 1.7905, |
| "step": 70000 |
| }, |
| { |
| "epoch": 4.917802444850358, |
| "eval_accuracy": 0.6745820550741288, |
| "eval_loss": 1.5280698537826538, |
| "eval_runtime": 159.6215, |
| "eval_samples_per_second": 756.202, |
| "eval_steps_per_second": 4.73, |
| "step": 70000 |
| }, |
| { |
| "epoch": 4.952929605170718, |
| "grad_norm": 0.8586461544036865, |
| "learning_rate": 8.250686713893306e-05, |
| "loss": 1.7883, |
| "step": 70500 |
| }, |
| { |
| "epoch": 4.988056765491078, |
| "grad_norm": 0.8624871373176575, |
| "learning_rate": 8.236229579297384e-05, |
| "loss": 1.7828, |
| "step": 71000 |
| }, |
| { |
| "epoch": 5.023183925811438, |
| "grad_norm": 0.8912454843521118, |
| "learning_rate": 8.221772444701461e-05, |
| "loss": 1.7817, |
| "step": 71500 |
| }, |
| { |
| "epoch": 5.0583110861317975, |
| "grad_norm": 0.8608023524284363, |
| "learning_rate": 8.207315310105538e-05, |
| "loss": 1.7749, |
| "step": 72000 |
| }, |
| { |
| "epoch": 5.093438246452157, |
| "grad_norm": 0.8637052178382874, |
| "learning_rate": 8.192858175509614e-05, |
| "loss": 1.7718, |
| "step": 72500 |
| }, |
| { |
| "epoch": 5.128565406772516, |
| "grad_norm": 0.8690493702888489, |
| "learning_rate": 8.178401040913692e-05, |
| "loss": 1.7687, |
| "step": 73000 |
| }, |
| { |
| "epoch": 5.163692567092876, |
| "grad_norm": 0.857749342918396, |
| "learning_rate": 8.163943906317768e-05, |
| "loss": 1.7671, |
| "step": 73500 |
| }, |
| { |
| "epoch": 5.198819727413236, |
| "grad_norm": 0.8630342483520508, |
| "learning_rate": 8.149486771721845e-05, |
| "loss": 1.7641, |
| "step": 74000 |
| }, |
| { |
| "epoch": 5.2339468877335955, |
| "grad_norm": 0.8738946318626404, |
| "learning_rate": 8.135029637125921e-05, |
| "loss": 1.7656, |
| "step": 74500 |
| }, |
| { |
| "epoch": 5.269074048053955, |
| "grad_norm": 0.8837184309959412, |
| "learning_rate": 8.120572502529999e-05, |
| "loss": 1.762, |
| "step": 75000 |
| }, |
| { |
| "epoch": 5.304201208374315, |
| "grad_norm": 0.8617969155311584, |
| "learning_rate": 8.106115367934077e-05, |
| "loss": 1.7609, |
| "step": 75500 |
| }, |
| { |
| "epoch": 5.339328368694675, |
| "grad_norm": 0.8774200081825256, |
| "learning_rate": 8.091658233338153e-05, |
| "loss": 1.7576, |
| "step": 76000 |
| }, |
| { |
| "epoch": 5.374455529015035, |
| "grad_norm": 0.88800048828125, |
| "learning_rate": 8.07720109874223e-05, |
| "loss": 1.7542, |
| "step": 76500 |
| }, |
| { |
| "epoch": 5.409582689335394, |
| "grad_norm": 0.8689431548118591, |
| "learning_rate": 8.062743964146306e-05, |
| "loss": 1.7506, |
| "step": 77000 |
| }, |
| { |
| "epoch": 5.444709849655754, |
| "grad_norm": 0.9089791774749756, |
| "learning_rate": 8.048286829550384e-05, |
| "loss": 1.7474, |
| "step": 77500 |
| }, |
| { |
| "epoch": 5.479837009976113, |
| "grad_norm": 0.9031184911727905, |
| "learning_rate": 8.03382969495446e-05, |
| "loss": 1.7488, |
| "step": 78000 |
| }, |
| { |
| "epoch": 5.514964170296473, |
| "grad_norm": 0.8577529788017273, |
| "learning_rate": 8.019372560358536e-05, |
| "loss": 1.7464, |
| "step": 78500 |
| }, |
| { |
| "epoch": 5.550091330616833, |
| "grad_norm": 0.8868847489356995, |
| "learning_rate": 8.004915425762614e-05, |
| "loss": 1.7447, |
| "step": 79000 |
| }, |
| { |
| "epoch": 5.5852184909371925, |
| "grad_norm": 0.880105197429657, |
| "learning_rate": 7.990458291166692e-05, |
| "loss": 1.7427, |
| "step": 79500 |
| }, |
| { |
| "epoch": 5.620345651257552, |
| "grad_norm": 0.8873801231384277, |
| "learning_rate": 7.976001156570768e-05, |
| "loss": 1.74, |
| "step": 80000 |
| }, |
| { |
| "epoch": 5.620345651257552, |
| "eval_accuracy": 0.6814942626423945, |
| "eval_loss": 1.48452627658844, |
| "eval_runtime": 157.5372, |
| "eval_samples_per_second": 766.206, |
| "eval_steps_per_second": 4.793, |
| "step": 80000 |
| }, |
| { |
| "epoch": 5.655472811577912, |
| "grad_norm": 0.8973670601844788, |
| "learning_rate": 7.961544021974845e-05, |
| "loss": 1.7369, |
| "step": 80500 |
| }, |
| { |
| "epoch": 5.690599971898272, |
| "grad_norm": 0.8752869963645935, |
| "learning_rate": 7.947086887378921e-05, |
| "loss": 1.7423, |
| "step": 81000 |
| }, |
| { |
| "epoch": 5.725727132218632, |
| "grad_norm": 0.8732084035873413, |
| "learning_rate": 7.932629752782999e-05, |
| "loss": 1.7363, |
| "step": 81500 |
| }, |
| { |
| "epoch": 5.760854292538991, |
| "grad_norm": 0.9183601140975952, |
| "learning_rate": 7.918172618187075e-05, |
| "loss": 1.7317, |
| "step": 82000 |
| }, |
| { |
| "epoch": 5.795981452859351, |
| "grad_norm": 0.9126468896865845, |
| "learning_rate": 7.903715483591153e-05, |
| "loss": 1.7324, |
| "step": 82500 |
| }, |
| { |
| "epoch": 5.83110861317971, |
| "grad_norm": 0.8697426915168762, |
| "learning_rate": 7.88925834899523e-05, |
| "loss": 1.7315, |
| "step": 83000 |
| }, |
| { |
| "epoch": 5.866235773500071, |
| "grad_norm": 0.8750841021537781, |
| "learning_rate": 7.874801214399307e-05, |
| "loss": 1.7287, |
| "step": 83500 |
| }, |
| { |
| "epoch": 5.90136293382043, |
| "grad_norm": 0.884046733379364, |
| "learning_rate": 7.860344079803384e-05, |
| "loss": 1.726, |
| "step": 84000 |
| }, |
| { |
| "epoch": 5.936490094140789, |
| "grad_norm": 0.8718340396881104, |
| "learning_rate": 7.84588694520746e-05, |
| "loss": 1.7242, |
| "step": 84500 |
| }, |
| { |
| "epoch": 5.971617254461149, |
| "grad_norm": 0.902029812335968, |
| "learning_rate": 7.831429810611536e-05, |
| "loss": 1.7226, |
| "step": 85000 |
| }, |
| { |
| "epoch": 6.006744414781509, |
| "grad_norm": 0.9095391035079956, |
| "learning_rate": 7.816972676015614e-05, |
| "loss": 1.7209, |
| "step": 85500 |
| }, |
| { |
| "epoch": 6.041871575101869, |
| "grad_norm": 0.8775455951690674, |
| "learning_rate": 7.802515541419692e-05, |
| "loss": 1.715, |
| "step": 86000 |
| }, |
| { |
| "epoch": 6.076998735422229, |
| "grad_norm": 0.8738338947296143, |
| "learning_rate": 7.788058406823768e-05, |
| "loss": 1.717, |
| "step": 86500 |
| }, |
| { |
| "epoch": 6.112125895742588, |
| "grad_norm": 0.8942703008651733, |
| "learning_rate": 7.773601272227845e-05, |
| "loss": 1.7111, |
| "step": 87000 |
| }, |
| { |
| "epoch": 6.147253056062948, |
| "grad_norm": 0.910753607749939, |
| "learning_rate": 7.759144137631922e-05, |
| "loss": 1.7098, |
| "step": 87500 |
| }, |
| { |
| "epoch": 6.182380216383308, |
| "grad_norm": 0.8818731307983398, |
| "learning_rate": 7.744687003035999e-05, |
| "loss": 1.7095, |
| "step": 88000 |
| }, |
| { |
| "epoch": 6.217507376703667, |
| "grad_norm": 0.8830975890159607, |
| "learning_rate": 7.730229868440075e-05, |
| "loss": 1.7088, |
| "step": 88500 |
| }, |
| { |
| "epoch": 6.252634537024027, |
| "grad_norm": 0.8831595182418823, |
| "learning_rate": 7.715772733844152e-05, |
| "loss": 1.7042, |
| "step": 89000 |
| }, |
| { |
| "epoch": 6.287761697344386, |
| "grad_norm": 0.9083411693572998, |
| "learning_rate": 7.70131559924823e-05, |
| "loss": 1.7047, |
| "step": 89500 |
| }, |
| { |
| "epoch": 6.322888857664746, |
| "grad_norm": 0.8737950325012207, |
| "learning_rate": 7.686858464652307e-05, |
| "loss": 1.7042, |
| "step": 90000 |
| }, |
| { |
| "epoch": 6.322888857664746, |
| "eval_accuracy": 0.6868321219641907, |
| "eval_loss": 1.4543001651763916, |
| "eval_runtime": 167.7074, |
| "eval_samples_per_second": 719.742, |
| "eval_steps_per_second": 4.502, |
| "step": 90000 |
| }, |
| { |
| "epoch": 6.358016017985106, |
| "grad_norm": 0.8860738277435303, |
| "learning_rate": 7.672401330056384e-05, |
| "loss": 1.7036, |
| "step": 90500 |
| }, |
| { |
| "epoch": 6.393143178305466, |
| "grad_norm": 0.8951513767242432, |
| "learning_rate": 7.65794419546046e-05, |
| "loss": 1.7007, |
| "step": 91000 |
| }, |
| { |
| "epoch": 6.4282703386258255, |
| "grad_norm": 0.8835942149162292, |
| "learning_rate": 7.643487060864536e-05, |
| "loss": 1.6977, |
| "step": 91500 |
| }, |
| { |
| "epoch": 6.463397498946185, |
| "grad_norm": 0.923398494720459, |
| "learning_rate": 7.629029926268614e-05, |
| "loss": 1.6975, |
| "step": 92000 |
| }, |
| { |
| "epoch": 6.498524659266545, |
| "grad_norm": 0.8661450147628784, |
| "learning_rate": 7.61457279167269e-05, |
| "loss": 1.6967, |
| "step": 92500 |
| }, |
| { |
| "epoch": 6.533651819586905, |
| "grad_norm": 0.8681382536888123, |
| "learning_rate": 7.600115657076767e-05, |
| "loss": 1.6965, |
| "step": 93000 |
| }, |
| { |
| "epoch": 6.568778979907265, |
| "grad_norm": 0.9150139689445496, |
| "learning_rate": 7.585658522480845e-05, |
| "loss": 1.694, |
| "step": 93500 |
| }, |
| { |
| "epoch": 6.603906140227624, |
| "grad_norm": 0.8800594806671143, |
| "learning_rate": 7.571201387884922e-05, |
| "loss": 1.6898, |
| "step": 94000 |
| }, |
| { |
| "epoch": 6.639033300547983, |
| "grad_norm": 0.8568074703216553, |
| "learning_rate": 7.556744253288999e-05, |
| "loss": 1.6892, |
| "step": 94500 |
| }, |
| { |
| "epoch": 6.674160460868343, |
| "grad_norm": 0.8834248781204224, |
| "learning_rate": 7.542287118693075e-05, |
| "loss": 1.6869, |
| "step": 95000 |
| }, |
| { |
| "epoch": 6.709287621188703, |
| "grad_norm": 0.8736454248428345, |
| "learning_rate": 7.527829984097152e-05, |
| "loss": 1.6867, |
| "step": 95500 |
| }, |
| { |
| "epoch": 6.744414781509063, |
| "grad_norm": 0.9212759137153625, |
| "learning_rate": 7.51337284950123e-05, |
| "loss": 1.6912, |
| "step": 96000 |
| }, |
| { |
| "epoch": 6.7795419418294225, |
| "grad_norm": 0.8739559650421143, |
| "learning_rate": 7.498915714905306e-05, |
| "loss": 1.6832, |
| "step": 96500 |
| }, |
| { |
| "epoch": 6.814669102149782, |
| "grad_norm": 0.9126378893852234, |
| "learning_rate": 7.484458580309382e-05, |
| "loss": 1.6844, |
| "step": 97000 |
| }, |
| { |
| "epoch": 6.849796262470142, |
| "grad_norm": 0.8926857709884644, |
| "learning_rate": 7.47000144571346e-05, |
| "loss": 1.678, |
| "step": 97500 |
| }, |
| { |
| "epoch": 6.884923422790502, |
| "grad_norm": 0.895026683807373, |
| "learning_rate": 7.455544311117538e-05, |
| "loss": 1.6792, |
| "step": 98000 |
| }, |
| { |
| "epoch": 6.920050583110862, |
| "grad_norm": 0.8956107497215271, |
| "learning_rate": 7.441087176521614e-05, |
| "loss": 1.6762, |
| "step": 98500 |
| }, |
| { |
| "epoch": 6.955177743431221, |
| "grad_norm": 0.8957840800285339, |
| "learning_rate": 7.42663004192569e-05, |
| "loss": 1.6769, |
| "step": 99000 |
| }, |
| { |
| "epoch": 6.99030490375158, |
| "grad_norm": 0.8974307775497437, |
| "learning_rate": 7.412172907329767e-05, |
| "loss": 1.6724, |
| "step": 99500 |
| }, |
| { |
| "epoch": 7.02543206407194, |
| "grad_norm": 0.8978477120399475, |
| "learning_rate": 7.397715772733845e-05, |
| "loss": 1.6725, |
| "step": 100000 |
| }, |
| { |
| "epoch": 7.02543206407194, |
| "eval_accuracy": 0.6917472831960627, |
| "eval_loss": 1.4226056337356567, |
| "eval_runtime": 164.788, |
| "eval_samples_per_second": 732.492, |
| "eval_steps_per_second": 4.582, |
| "step": 100000 |
| }, |
| { |
| "epoch": 7.0605592243923, |
| "grad_norm": 0.8918427228927612, |
| "learning_rate": 7.383258638137921e-05, |
| "loss": 1.6711, |
| "step": 100500 |
| }, |
| { |
| "epoch": 7.09568638471266, |
| "grad_norm": 0.899376392364502, |
| "learning_rate": 7.368801503541999e-05, |
| "loss": 1.6712, |
| "step": 101000 |
| }, |
| { |
| "epoch": 7.1308135450330195, |
| "grad_norm": 0.9002095460891724, |
| "learning_rate": 7.354344368946075e-05, |
| "loss": 1.6693, |
| "step": 101500 |
| }, |
| { |
| "epoch": 7.165940705353379, |
| "grad_norm": 0.9232343435287476, |
| "learning_rate": 7.339887234350153e-05, |
| "loss": 1.667, |
| "step": 102000 |
| }, |
| { |
| "epoch": 7.201067865673739, |
| "grad_norm": 0.8874390721321106, |
| "learning_rate": 7.32543009975423e-05, |
| "loss": 1.666, |
| "step": 102500 |
| }, |
| { |
| "epoch": 7.236195025994099, |
| "grad_norm": 0.8941359519958496, |
| "learning_rate": 7.310972965158306e-05, |
| "loss": 1.6641, |
| "step": 103000 |
| }, |
| { |
| "epoch": 7.271322186314459, |
| "grad_norm": 0.8822116851806641, |
| "learning_rate": 7.296515830562382e-05, |
| "loss": 1.665, |
| "step": 103500 |
| }, |
| { |
| "epoch": 7.306449346634818, |
| "grad_norm": 0.911529541015625, |
| "learning_rate": 7.28205869596646e-05, |
| "loss": 1.6605, |
| "step": 104000 |
| }, |
| { |
| "epoch": 7.341576506955178, |
| "grad_norm": 0.8901451230049133, |
| "learning_rate": 7.267601561370536e-05, |
| "loss": 1.6565, |
| "step": 104500 |
| }, |
| { |
| "epoch": 7.376703667275537, |
| "grad_norm": 0.9035801291465759, |
| "learning_rate": 7.253144426774614e-05, |
| "loss": 1.6595, |
| "step": 105000 |
| }, |
| { |
| "epoch": 7.411830827595897, |
| "grad_norm": 0.9026615619659424, |
| "learning_rate": 7.23868729217869e-05, |
| "loss": 1.6629, |
| "step": 105500 |
| }, |
| { |
| "epoch": 7.446957987916257, |
| "grad_norm": 0.886187732219696, |
| "learning_rate": 7.224230157582768e-05, |
| "loss": 1.66, |
| "step": 106000 |
| }, |
| { |
| "epoch": 7.482085148236616, |
| "grad_norm": 0.8838157057762146, |
| "learning_rate": 7.209773022986845e-05, |
| "loss": 1.656, |
| "step": 106500 |
| }, |
| { |
| "epoch": 7.517212308556976, |
| "grad_norm": 0.9090375900268555, |
| "learning_rate": 7.195315888390921e-05, |
| "loss": 1.6533, |
| "step": 107000 |
| }, |
| { |
| "epoch": 7.552339468877336, |
| "grad_norm": 0.9034336805343628, |
| "learning_rate": 7.180858753794997e-05, |
| "loss": 1.6524, |
| "step": 107500 |
| }, |
| { |
| "epoch": 7.587466629197696, |
| "grad_norm": 0.8821242451667786, |
| "learning_rate": 7.166401619199075e-05, |
| "loss": 1.653, |
| "step": 108000 |
| }, |
| { |
| "epoch": 7.6225937895180556, |
| "grad_norm": 0.9131813645362854, |
| "learning_rate": 7.151944484603153e-05, |
| "loss": 1.6548, |
| "step": 108500 |
| }, |
| { |
| "epoch": 7.657720949838415, |
| "grad_norm": 0.9031689167022705, |
| "learning_rate": 7.13748735000723e-05, |
| "loss": 1.6506, |
| "step": 109000 |
| }, |
| { |
| "epoch": 7.692848110158775, |
| "grad_norm": 0.9220744967460632, |
| "learning_rate": 7.123030215411306e-05, |
| "loss": 1.6516, |
| "step": 109500 |
| }, |
| { |
| "epoch": 7.727975270479135, |
| "grad_norm": 0.9103692770004272, |
| "learning_rate": 7.108573080815382e-05, |
| "loss": 1.6516, |
| "step": 110000 |
| }, |
| { |
| "epoch": 7.727975270479135, |
| "eval_accuracy": 0.6957056341555516, |
| "eval_loss": 1.4015986919403076, |
| "eval_runtime": 162.006, |
| "eval_samples_per_second": 745.071, |
| "eval_steps_per_second": 4.66, |
| "step": 110000 |
| }, |
| { |
| "epoch": 7.763102430799494, |
| "grad_norm": 0.9088935256004333, |
| "learning_rate": 7.09411594621946e-05, |
| "loss": 1.6479, |
| "step": 110500 |
| }, |
| { |
| "epoch": 7.798229591119854, |
| "grad_norm": 0.9063370227813721, |
| "learning_rate": 7.079658811623536e-05, |
| "loss": 1.6448, |
| "step": 111000 |
| }, |
| { |
| "epoch": 7.833356751440213, |
| "grad_norm": 0.9396301507949829, |
| "learning_rate": 7.065201677027613e-05, |
| "loss": 1.645, |
| "step": 111500 |
| }, |
| { |
| "epoch": 7.868483911760573, |
| "grad_norm": 0.9160650372505188, |
| "learning_rate": 7.05074454243169e-05, |
| "loss": 1.6462, |
| "step": 112000 |
| }, |
| { |
| "epoch": 7.903611072080933, |
| "grad_norm": 0.8814093470573425, |
| "learning_rate": 7.036287407835768e-05, |
| "loss": 1.6447, |
| "step": 112500 |
| }, |
| { |
| "epoch": 7.938738232401293, |
| "grad_norm": 0.9333155155181885, |
| "learning_rate": 7.021830273239845e-05, |
| "loss": 1.6386, |
| "step": 113000 |
| }, |
| { |
| "epoch": 7.9738653927216525, |
| "grad_norm": 0.8837305307388306, |
| "learning_rate": 7.007373138643921e-05, |
| "loss": 1.6398, |
| "step": 113500 |
| }, |
| { |
| "epoch": 8.008992553042011, |
| "grad_norm": 0.9042782783508301, |
| "learning_rate": 6.992916004047997e-05, |
| "loss": 1.6407, |
| "step": 114000 |
| }, |
| { |
| "epoch": 8.044119713362372, |
| "grad_norm": 0.942909836769104, |
| "learning_rate": 6.978458869452075e-05, |
| "loss": 1.6352, |
| "step": 114500 |
| }, |
| { |
| "epoch": 8.079246873682731, |
| "grad_norm": 0.9293888807296753, |
| "learning_rate": 6.964001734856152e-05, |
| "loss": 1.6348, |
| "step": 115000 |
| }, |
| { |
| "epoch": 8.114374034003092, |
| "grad_norm": 0.9147842526435852, |
| "learning_rate": 6.949544600260228e-05, |
| "loss": 1.6334, |
| "step": 115500 |
| }, |
| { |
| "epoch": 8.14950119432345, |
| "grad_norm": 0.9225263595581055, |
| "learning_rate": 6.935087465664306e-05, |
| "loss": 1.635, |
| "step": 116000 |
| }, |
| { |
| "epoch": 8.184628354643811, |
| "grad_norm": 0.9055420756340027, |
| "learning_rate": 6.920630331068383e-05, |
| "loss": 1.6338, |
| "step": 116500 |
| }, |
| { |
| "epoch": 8.21975551496417, |
| "grad_norm": 0.9056685566902161, |
| "learning_rate": 6.90617319647246e-05, |
| "loss": 1.6307, |
| "step": 117000 |
| }, |
| { |
| "epoch": 8.25488267528453, |
| "grad_norm": 0.9413787722587585, |
| "learning_rate": 6.891716061876536e-05, |
| "loss": 1.6313, |
| "step": 117500 |
| }, |
| { |
| "epoch": 8.29000983560489, |
| "grad_norm": 0.8983238935470581, |
| "learning_rate": 6.877258927280613e-05, |
| "loss": 1.6271, |
| "step": 118000 |
| }, |
| { |
| "epoch": 8.325136995925249, |
| "grad_norm": 0.914631187915802, |
| "learning_rate": 6.86280179268469e-05, |
| "loss": 1.6265, |
| "step": 118500 |
| }, |
| { |
| "epoch": 8.36026415624561, |
| "grad_norm": 0.932934582233429, |
| "learning_rate": 6.848344658088767e-05, |
| "loss": 1.6272, |
| "step": 119000 |
| }, |
| { |
| "epoch": 8.395391316565968, |
| "grad_norm": 0.9087077975273132, |
| "learning_rate": 6.833887523492845e-05, |
| "loss": 1.627, |
| "step": 119500 |
| }, |
| { |
| "epoch": 8.430518476886329, |
| "grad_norm": 0.9194434881210327, |
| "learning_rate": 6.819430388896921e-05, |
| "loss": 1.6269, |
| "step": 120000 |
| }, |
| { |
| "epoch": 8.430518476886329, |
| "eval_accuracy": 0.6991192238224887, |
| "eval_loss": 1.3791452646255493, |
| "eval_runtime": 171.6084, |
| "eval_samples_per_second": 703.381, |
| "eval_steps_per_second": 4.4, |
| "step": 120000 |
| }, |
| { |
| "epoch": 8.465645637206688, |
| "grad_norm": 0.9238064885139465, |
| "learning_rate": 6.804973254300999e-05, |
| "loss": 1.6258, |
| "step": 120500 |
| }, |
| { |
| "epoch": 8.500772797527048, |
| "grad_norm": 0.9381468296051025, |
| "learning_rate": 6.790516119705075e-05, |
| "loss": 1.6261, |
| "step": 121000 |
| }, |
| { |
| "epoch": 8.535899957847407, |
| "grad_norm": 0.8942645192146301, |
| "learning_rate": 6.776058985109152e-05, |
| "loss": 1.6273, |
| "step": 121500 |
| }, |
| { |
| "epoch": 8.571027118167768, |
| "grad_norm": 0.9223001599311829, |
| "learning_rate": 6.761601850513228e-05, |
| "loss": 1.6225, |
| "step": 122000 |
| }, |
| { |
| "epoch": 8.606154278488127, |
| "grad_norm": 0.9025728702545166, |
| "learning_rate": 6.747144715917306e-05, |
| "loss": 1.6208, |
| "step": 122500 |
| }, |
| { |
| "epoch": 8.641281438808488, |
| "grad_norm": 0.8991536498069763, |
| "learning_rate": 6.732687581321382e-05, |
| "loss": 1.6209, |
| "step": 123000 |
| }, |
| { |
| "epoch": 8.676408599128846, |
| "grad_norm": 0.9359357357025146, |
| "learning_rate": 6.71823044672546e-05, |
| "loss": 1.6166, |
| "step": 123500 |
| }, |
| { |
| "epoch": 8.711535759449205, |
| "grad_norm": 0.9257062673568726, |
| "learning_rate": 6.703773312129536e-05, |
| "loss": 1.616, |
| "step": 124000 |
| }, |
| { |
| "epoch": 8.746662919769566, |
| "grad_norm": 0.9320250749588013, |
| "learning_rate": 6.689316177533613e-05, |
| "loss": 1.6115, |
| "step": 124500 |
| }, |
| { |
| "epoch": 8.781790080089925, |
| "grad_norm": 0.9052867889404297, |
| "learning_rate": 6.67485904293769e-05, |
| "loss": 1.6163, |
| "step": 125000 |
| }, |
| { |
| "epoch": 8.816917240410286, |
| "grad_norm": 0.9434939026832581, |
| "learning_rate": 6.660401908341767e-05, |
| "loss": 1.6147, |
| "step": 125500 |
| }, |
| { |
| "epoch": 8.852044400730644, |
| "grad_norm": 0.9227339625358582, |
| "learning_rate": 6.645944773745843e-05, |
| "loss": 1.6141, |
| "step": 126000 |
| }, |
| { |
| "epoch": 8.887171561051005, |
| "grad_norm": 0.9441553950309753, |
| "learning_rate": 6.631487639149921e-05, |
| "loss": 1.6142, |
| "step": 126500 |
| }, |
| { |
| "epoch": 8.922298721371364, |
| "grad_norm": 0.9345538020133972, |
| "learning_rate": 6.617030504553999e-05, |
| "loss": 1.6121, |
| "step": 127000 |
| }, |
| { |
| "epoch": 8.957425881691725, |
| "grad_norm": 0.9714908003807068, |
| "learning_rate": 6.602573369958075e-05, |
| "loss": 1.6159, |
| "step": 127500 |
| }, |
| { |
| "epoch": 8.992553042012084, |
| "grad_norm": 0.942217230796814, |
| "learning_rate": 6.588116235362152e-05, |
| "loss": 1.6149, |
| "step": 128000 |
| }, |
| { |
| "epoch": 9.027680202332444, |
| "grad_norm": 0.9442844390869141, |
| "learning_rate": 6.573659100766228e-05, |
| "loss": 1.6073, |
| "step": 128500 |
| }, |
| { |
| "epoch": 9.062807362652803, |
| "grad_norm": 0.9177855849266052, |
| "learning_rate": 6.559201966170306e-05, |
| "loss": 1.6068, |
| "step": 129000 |
| }, |
| { |
| "epoch": 9.097934522973162, |
| "grad_norm": 0.9610123038291931, |
| "learning_rate": 6.544744831574382e-05, |
| "loss": 1.6084, |
| "step": 129500 |
| }, |
| { |
| "epoch": 9.133061683293523, |
| "grad_norm": 0.8969529867172241, |
| "learning_rate": 6.530287696978458e-05, |
| "loss": 1.6032, |
| "step": 130000 |
| }, |
| { |
| "epoch": 9.133061683293523, |
| "eval_accuracy": 0.701581667800137, |
| "eval_loss": 1.3646811246871948, |
| "eval_runtime": 168.8385, |
| "eval_samples_per_second": 714.92, |
| "eval_steps_per_second": 4.472, |
| "step": 130000 |
| }, |
| { |
| "epoch": 9.168188843613882, |
| "grad_norm": 0.9467385411262512, |
| "learning_rate": 6.515830562382536e-05, |
| "loss": 1.6052, |
| "step": 130500 |
| }, |
| { |
| "epoch": 9.203316003934242, |
| "grad_norm": 0.9412127137184143, |
| "learning_rate": 6.501373427786614e-05, |
| "loss": 1.6062, |
| "step": 131000 |
| }, |
| { |
| "epoch": 9.238443164254601, |
| "grad_norm": 0.9351688623428345, |
| "learning_rate": 6.48691629319069e-05, |
| "loss": 1.605, |
| "step": 131500 |
| }, |
| { |
| "epoch": 9.273570324574962, |
| "grad_norm": 0.9037009477615356, |
| "learning_rate": 6.472459158594767e-05, |
| "loss": 1.6061, |
| "step": 132000 |
| }, |
| { |
| "epoch": 9.30869748489532, |
| "grad_norm": 0.9293714165687561, |
| "learning_rate": 6.458002023998843e-05, |
| "loss": 1.6037, |
| "step": 132500 |
| }, |
| { |
| "epoch": 9.343824645215681, |
| "grad_norm": 0.9317144751548767, |
| "learning_rate": 6.443544889402921e-05, |
| "loss": 1.6015, |
| "step": 133000 |
| }, |
| { |
| "epoch": 9.37895180553604, |
| "grad_norm": 0.937980055809021, |
| "learning_rate": 6.429087754806997e-05, |
| "loss": 1.6009, |
| "step": 133500 |
| }, |
| { |
| "epoch": 9.414078965856401, |
| "grad_norm": 0.9135161638259888, |
| "learning_rate": 6.414630620211074e-05, |
| "loss": 1.5999, |
| "step": 134000 |
| }, |
| { |
| "epoch": 9.44920612617676, |
| "grad_norm": 0.9643651247024536, |
| "learning_rate": 6.400173485615151e-05, |
| "loss": 1.5967, |
| "step": 134500 |
| }, |
| { |
| "epoch": 9.484333286497119, |
| "grad_norm": 0.918627142906189, |
| "learning_rate": 6.385716351019229e-05, |
| "loss": 1.5966, |
| "step": 135000 |
| }, |
| { |
| "epoch": 9.51946044681748, |
| "grad_norm": 0.9351798892021179, |
| "learning_rate": 6.371259216423306e-05, |
| "loss": 1.5984, |
| "step": 135500 |
| }, |
| { |
| "epoch": 9.554587607137838, |
| "grad_norm": 0.9409826397895813, |
| "learning_rate": 6.356802081827382e-05, |
| "loss": 1.5969, |
| "step": 136000 |
| }, |
| { |
| "epoch": 9.589714767458199, |
| "grad_norm": 0.9116629362106323, |
| "learning_rate": 6.342344947231458e-05, |
| "loss": 1.5983, |
| "step": 136500 |
| }, |
| { |
| "epoch": 9.624841927778558, |
| "grad_norm": 0.9535496830940247, |
| "learning_rate": 6.327887812635536e-05, |
| "loss": 1.5939, |
| "step": 137000 |
| }, |
| { |
| "epoch": 9.659969088098919, |
| "grad_norm": 0.9358441233634949, |
| "learning_rate": 6.313430678039613e-05, |
| "loss": 1.5951, |
| "step": 137500 |
| }, |
| { |
| "epoch": 9.695096248419278, |
| "grad_norm": 0.9049669504165649, |
| "learning_rate": 6.29897354344369e-05, |
| "loss": 1.5902, |
| "step": 138000 |
| }, |
| { |
| "epoch": 9.730223408739638, |
| "grad_norm": 0.9285699725151062, |
| "learning_rate": 6.284516408847767e-05, |
| "loss": 1.5964, |
| "step": 138500 |
| }, |
| { |
| "epoch": 9.765350569059997, |
| "grad_norm": 0.9280917644500732, |
| "learning_rate": 6.270059274251843e-05, |
| "loss": 1.5914, |
| "step": 139000 |
| }, |
| { |
| "epoch": 9.800477729380358, |
| "grad_norm": 0.9302689433097839, |
| "learning_rate": 6.255602139655921e-05, |
| "loss": 1.5923, |
| "step": 139500 |
| }, |
| { |
| "epoch": 9.835604889700717, |
| "grad_norm": 0.9023528099060059, |
| "learning_rate": 6.241145005059997e-05, |
| "loss": 1.5903, |
| "step": 140000 |
| }, |
| { |
| "epoch": 9.835604889700717, |
| "eval_accuracy": 0.7050626192663446, |
| "eval_loss": 1.3465116024017334, |
| "eval_runtime": 162.3772, |
| "eval_samples_per_second": 743.368, |
| "eval_steps_per_second": 4.65, |
| "step": 140000 |
| }, |
| { |
| "epoch": 9.870732050021076, |
| "grad_norm": 0.9710935950279236, |
| "learning_rate": 6.226687870464074e-05, |
| "loss": 1.5907, |
| "step": 140500 |
| }, |
| { |
| "epoch": 9.905859210341436, |
| "grad_norm": 0.9528540968894958, |
| "learning_rate": 6.212230735868151e-05, |
| "loss": 1.5888, |
| "step": 141000 |
| }, |
| { |
| "epoch": 9.940986370661795, |
| "grad_norm": 0.9252150058746338, |
| "learning_rate": 6.197773601272228e-05, |
| "loss": 1.5888, |
| "step": 141500 |
| }, |
| { |
| "epoch": 9.976113530982156, |
| "grad_norm": 0.9385340213775635, |
| "learning_rate": 6.183316466676306e-05, |
| "loss": 1.5853, |
| "step": 142000 |
| }, |
| { |
| "epoch": 10.011240691302515, |
| "grad_norm": 0.946033239364624, |
| "learning_rate": 6.168859332080382e-05, |
| "loss": 1.5873, |
| "step": 142500 |
| }, |
| { |
| "epoch": 10.046367851622875, |
| "grad_norm": 0.9161434769630432, |
| "learning_rate": 6.154402197484458e-05, |
| "loss": 1.5872, |
| "step": 143000 |
| }, |
| { |
| "epoch": 10.081495011943234, |
| "grad_norm": 0.9464936852455139, |
| "learning_rate": 6.139945062888536e-05, |
| "loss": 1.5825, |
| "step": 143500 |
| }, |
| { |
| "epoch": 10.116622172263595, |
| "grad_norm": 0.947498619556427, |
| "learning_rate": 6.125487928292613e-05, |
| "loss": 1.5813, |
| "step": 144000 |
| }, |
| { |
| "epoch": 10.151749332583954, |
| "grad_norm": 0.9718836545944214, |
| "learning_rate": 6.111030793696689e-05, |
| "loss": 1.5816, |
| "step": 144500 |
| }, |
| { |
| "epoch": 10.186876492904315, |
| "grad_norm": 0.9489443898200989, |
| "learning_rate": 6.096573659100766e-05, |
| "loss": 1.5816, |
| "step": 145000 |
| }, |
| { |
| "epoch": 10.222003653224673, |
| "grad_norm": 0.9455053210258484, |
| "learning_rate": 6.082116524504844e-05, |
| "loss": 1.5835, |
| "step": 145500 |
| }, |
| { |
| "epoch": 10.257130813545032, |
| "grad_norm": 0.9187600016593933, |
| "learning_rate": 6.067659389908921e-05, |
| "loss": 1.5827, |
| "step": 146000 |
| }, |
| { |
| "epoch": 10.292257973865393, |
| "grad_norm": 0.9401862025260925, |
| "learning_rate": 6.053202255312997e-05, |
| "loss": 1.5771, |
| "step": 146500 |
| }, |
| { |
| "epoch": 10.327385134185752, |
| "grad_norm": 0.9436215758323669, |
| "learning_rate": 6.0387451207170744e-05, |
| "loss": 1.5784, |
| "step": 147000 |
| }, |
| { |
| "epoch": 10.362512294506113, |
| "grad_norm": 0.9564245343208313, |
| "learning_rate": 6.024287986121151e-05, |
| "loss": 1.5786, |
| "step": 147500 |
| }, |
| { |
| "epoch": 10.397639454826471, |
| "grad_norm": 0.9166202545166016, |
| "learning_rate": 6.009830851525228e-05, |
| "loss": 1.5765, |
| "step": 148000 |
| }, |
| { |
| "epoch": 10.432766615146832, |
| "grad_norm": 0.9648714065551758, |
| "learning_rate": 5.995373716929304e-05, |
| "loss": 1.5796, |
| "step": 148500 |
| }, |
| { |
| "epoch": 10.467893775467191, |
| "grad_norm": 0.9763674736022949, |
| "learning_rate": 5.980916582333381e-05, |
| "loss": 1.5747, |
| "step": 149000 |
| }, |
| { |
| "epoch": 10.503020935787552, |
| "grad_norm": 0.9395629167556763, |
| "learning_rate": 5.966459447737459e-05, |
| "loss": 1.5727, |
| "step": 149500 |
| }, |
| { |
| "epoch": 10.53814809610791, |
| "grad_norm": 0.9242696166038513, |
| "learning_rate": 5.952002313141536e-05, |
| "loss": 1.5759, |
| "step": 150000 |
| }, |
| { |
| "epoch": 10.53814809610791, |
| "eval_accuracy": 0.7074019542721169, |
| "eval_loss": 1.3325526714324951, |
| "eval_runtime": 164.3033, |
| "eval_samples_per_second": 734.654, |
| "eval_steps_per_second": 4.595, |
| "step": 150000 |
| }, |
| { |
| "epoch": 10.573275256428271, |
| "grad_norm": 0.9491183757781982, |
| "learning_rate": 5.9375451785456126e-05, |
| "loss": 1.5778, |
| "step": 150500 |
| }, |
| { |
| "epoch": 10.60840241674863, |
| "grad_norm": 0.9535388350486755, |
| "learning_rate": 5.9230880439496896e-05, |
| "loss": 1.5726, |
| "step": 151000 |
| }, |
| { |
| "epoch": 10.643529577068989, |
| "grad_norm": 0.9607381820678711, |
| "learning_rate": 5.908630909353766e-05, |
| "loss": 1.573, |
| "step": 151500 |
| }, |
| { |
| "epoch": 10.67865673738935, |
| "grad_norm": 0.9314340949058533, |
| "learning_rate": 5.894173774757843e-05, |
| "loss": 1.5728, |
| "step": 152000 |
| }, |
| { |
| "epoch": 10.713783897709709, |
| "grad_norm": 0.9708797335624695, |
| "learning_rate": 5.8797166401619195e-05, |
| "loss": 1.5696, |
| "step": 152500 |
| }, |
| { |
| "epoch": 10.74891105803007, |
| "grad_norm": 0.94566810131073, |
| "learning_rate": 5.865259505565997e-05, |
| "loss": 1.5706, |
| "step": 153000 |
| }, |
| { |
| "epoch": 10.784038218350428, |
| "grad_norm": 0.9453552961349487, |
| "learning_rate": 5.8508023709700743e-05, |
| "loss": 1.5726, |
| "step": 153500 |
| }, |
| { |
| "epoch": 10.819165378670789, |
| "grad_norm": 0.9269335865974426, |
| "learning_rate": 5.8363452363741514e-05, |
| "loss": 1.5708, |
| "step": 154000 |
| }, |
| { |
| "epoch": 10.854292538991148, |
| "grad_norm": 0.9515119791030884, |
| "learning_rate": 5.821888101778228e-05, |
| "loss": 1.5706, |
| "step": 154500 |
| }, |
| { |
| "epoch": 10.889419699311508, |
| "grad_norm": 0.9325230717658997, |
| "learning_rate": 5.807430967182305e-05, |
| "loss": 1.5702, |
| "step": 155000 |
| }, |
| { |
| "epoch": 10.924546859631867, |
| "grad_norm": 0.9474920630455017, |
| "learning_rate": 5.792973832586381e-05, |
| "loss": 1.5701, |
| "step": 155500 |
| }, |
| { |
| "epoch": 10.959674019952226, |
| "grad_norm": 0.9623352885246277, |
| "learning_rate": 5.7785166979904584e-05, |
| "loss": 1.5682, |
| "step": 156000 |
| }, |
| { |
| "epoch": 10.994801180272587, |
| "grad_norm": 0.919079065322876, |
| "learning_rate": 5.764059563394536e-05, |
| "loss": 1.5676, |
| "step": 156500 |
| }, |
| { |
| "epoch": 11.029928340592946, |
| "grad_norm": 0.9759007096290588, |
| "learning_rate": 5.7496024287986125e-05, |
| "loss": 1.5609, |
| "step": 157000 |
| }, |
| { |
| "epoch": 11.065055500913306, |
| "grad_norm": 0.9327608942985535, |
| "learning_rate": 5.7351452942026896e-05, |
| "loss": 1.5638, |
| "step": 157500 |
| }, |
| { |
| "epoch": 11.100182661233665, |
| "grad_norm": 0.9498403668403625, |
| "learning_rate": 5.720688159606766e-05, |
| "loss": 1.5638, |
| "step": 158000 |
| }, |
| { |
| "epoch": 11.135309821554026, |
| "grad_norm": 0.9694502949714661, |
| "learning_rate": 5.706231025010843e-05, |
| "loss": 1.5611, |
| "step": 158500 |
| }, |
| { |
| "epoch": 11.170436981874385, |
| "grad_norm": 0.9644596576690674, |
| "learning_rate": 5.6917738904149195e-05, |
| "loss": 1.5622, |
| "step": 159000 |
| }, |
| { |
| "epoch": 11.205564142194746, |
| "grad_norm": 0.9816145300865173, |
| "learning_rate": 5.6773167558189966e-05, |
| "loss": 1.5629, |
| "step": 159500 |
| }, |
| { |
| "epoch": 11.240691302515105, |
| "grad_norm": 0.9645638465881348, |
| "learning_rate": 5.6628596212230736e-05, |
| "loss": 1.5641, |
| "step": 160000 |
| }, |
| { |
| "epoch": 11.240691302515105, |
| "eval_accuracy": 0.7089585050166161, |
| "eval_loss": 1.323495864868164, |
| "eval_runtime": 166.5008, |
| "eval_samples_per_second": 724.957, |
| "eval_steps_per_second": 4.535, |
| "step": 160000 |
| }, |
| { |
| "epoch": 11.275818462835465, |
| "grad_norm": 0.9568628668785095, |
| "learning_rate": 5.6484024866271514e-05, |
| "loss": 1.5595, |
| "step": 160500 |
| }, |
| { |
| "epoch": 11.310945623155824, |
| "grad_norm": 0.9562154412269592, |
| "learning_rate": 5.633945352031228e-05, |
| "loss": 1.5632, |
| "step": 161000 |
| }, |
| { |
| "epoch": 11.346072783476183, |
| "grad_norm": 0.9697450995445251, |
| "learning_rate": 5.619488217435305e-05, |
| "loss": 1.5591, |
| "step": 161500 |
| }, |
| { |
| "epoch": 11.381199943796544, |
| "grad_norm": 0.9945010542869568, |
| "learning_rate": 5.605031082839381e-05, |
| "loss": 1.56, |
| "step": 162000 |
| }, |
| { |
| "epoch": 11.416327104116903, |
| "grad_norm": 0.9685713052749634, |
| "learning_rate": 5.5905739482434584e-05, |
| "loss": 1.5574, |
| "step": 162500 |
| }, |
| { |
| "epoch": 11.451454264437263, |
| "grad_norm": 0.9100556969642639, |
| "learning_rate": 5.576116813647535e-05, |
| "loss": 1.5536, |
| "step": 163000 |
| }, |
| { |
| "epoch": 11.486581424757622, |
| "grad_norm": 0.9792532920837402, |
| "learning_rate": 5.561659679051612e-05, |
| "loss": 1.5573, |
| "step": 163500 |
| }, |
| { |
| "epoch": 11.521708585077983, |
| "grad_norm": 0.9711182713508606, |
| "learning_rate": 5.5472025444556896e-05, |
| "loss": 1.5573, |
| "step": 164000 |
| }, |
| { |
| "epoch": 11.556835745398342, |
| "grad_norm": 0.9407551884651184, |
| "learning_rate": 5.532745409859767e-05, |
| "loss": 1.555, |
| "step": 164500 |
| }, |
| { |
| "epoch": 11.591962905718702, |
| "grad_norm": 0.9609689116477966, |
| "learning_rate": 5.518288275263843e-05, |
| "loss": 1.556, |
| "step": 165000 |
| }, |
| { |
| "epoch": 11.627090066039061, |
| "grad_norm": 0.9470645785331726, |
| "learning_rate": 5.50383114066792e-05, |
| "loss": 1.5556, |
| "step": 165500 |
| }, |
| { |
| "epoch": 11.662217226359422, |
| "grad_norm": 0.9639591574668884, |
| "learning_rate": 5.4893740060719965e-05, |
| "loss": 1.5534, |
| "step": 166000 |
| }, |
| { |
| "epoch": 11.69734438667978, |
| "grad_norm": 0.9554125070571899, |
| "learning_rate": 5.4749168714760736e-05, |
| "loss": 1.5545, |
| "step": 166500 |
| }, |
| { |
| "epoch": 11.73247154700014, |
| "grad_norm": 0.9270333647727966, |
| "learning_rate": 5.46045973688015e-05, |
| "loss": 1.55, |
| "step": 167000 |
| }, |
| { |
| "epoch": 11.7675987073205, |
| "grad_norm": 0.967582643032074, |
| "learning_rate": 5.446002602284227e-05, |
| "loss": 1.5505, |
| "step": 167500 |
| }, |
| { |
| "epoch": 11.80272586764086, |
| "grad_norm": 0.9647204875946045, |
| "learning_rate": 5.431545467688305e-05, |
| "loss": 1.5533, |
| "step": 168000 |
| }, |
| { |
| "epoch": 11.83785302796122, |
| "grad_norm": 0.9632757902145386, |
| "learning_rate": 5.417088333092382e-05, |
| "loss": 1.5492, |
| "step": 168500 |
| }, |
| { |
| "epoch": 11.872980188281579, |
| "grad_norm": 0.9651910066604614, |
| "learning_rate": 5.402631198496458e-05, |
| "loss": 1.5486, |
| "step": 169000 |
| }, |
| { |
| "epoch": 11.90810734860194, |
| "grad_norm": 0.9492982029914856, |
| "learning_rate": 5.3881740639005354e-05, |
| "loss": 1.5479, |
| "step": 169500 |
| }, |
| { |
| "epoch": 11.943234508922298, |
| "grad_norm": 0.9542218446731567, |
| "learning_rate": 5.373716929304612e-05, |
| "loss": 1.5487, |
| "step": 170000 |
| }, |
| { |
| "epoch": 11.943234508922298, |
| "eval_accuracy": 0.7110132857979123, |
| "eval_loss": 1.3103113174438477, |
| "eval_runtime": 161.8637, |
| "eval_samples_per_second": 745.726, |
| "eval_steps_per_second": 4.664, |
| "step": 170000 |
| }, |
| { |
| "epoch": 11.978361669242659, |
| "grad_norm": 0.9556529521942139, |
| "learning_rate": 5.359259794708689e-05, |
| "loss": 1.5487, |
| "step": 170500 |
| }, |
| { |
| "epoch": 12.013488829563018, |
| "grad_norm": 0.9854975938796997, |
| "learning_rate": 5.344802660112765e-05, |
| "loss": 1.5485, |
| "step": 171000 |
| }, |
| { |
| "epoch": 12.048615989883379, |
| "grad_norm": 0.9763370752334595, |
| "learning_rate": 5.330345525516843e-05, |
| "loss": 1.5455, |
| "step": 171500 |
| }, |
| { |
| "epoch": 12.083743150203738, |
| "grad_norm": 0.9626860618591309, |
| "learning_rate": 5.31588839092092e-05, |
| "loss": 1.5433, |
| "step": 172000 |
| }, |
| { |
| "epoch": 12.118870310524096, |
| "grad_norm": 0.9775327444076538, |
| "learning_rate": 5.3014312563249965e-05, |
| "loss": 1.546, |
| "step": 172500 |
| }, |
| { |
| "epoch": 12.153997470844457, |
| "grad_norm": 0.9717371463775635, |
| "learning_rate": 5.2869741217290736e-05, |
| "loss": 1.5481, |
| "step": 173000 |
| }, |
| { |
| "epoch": 12.189124631164816, |
| "grad_norm": 0.9556875824928284, |
| "learning_rate": 5.272516987133151e-05, |
| "loss": 1.5445, |
| "step": 173500 |
| }, |
| { |
| "epoch": 12.224251791485177, |
| "grad_norm": 0.973354697227478, |
| "learning_rate": 5.258059852537227e-05, |
| "loss": 1.5399, |
| "step": 174000 |
| }, |
| { |
| "epoch": 12.259378951805536, |
| "grad_norm": 0.9697452187538147, |
| "learning_rate": 5.243602717941304e-05, |
| "loss": 1.5444, |
| "step": 174500 |
| }, |
| { |
| "epoch": 12.294506112125896, |
| "grad_norm": 0.955089807510376, |
| "learning_rate": 5.229145583345382e-05, |
| "loss": 1.5396, |
| "step": 175000 |
| }, |
| { |
| "epoch": 12.329633272446255, |
| "grad_norm": 0.9705688953399658, |
| "learning_rate": 5.214688448749458e-05, |
| "loss": 1.5442, |
| "step": 175500 |
| }, |
| { |
| "epoch": 12.364760432766616, |
| "grad_norm": 0.9606062173843384, |
| "learning_rate": 5.2002313141535354e-05, |
| "loss": 1.5416, |
| "step": 176000 |
| }, |
| { |
| "epoch": 12.399887593086975, |
| "grad_norm": 0.9730265140533447, |
| "learning_rate": 5.185774179557612e-05, |
| "loss": 1.5402, |
| "step": 176500 |
| }, |
| { |
| "epoch": 12.435014753407334, |
| "grad_norm": 0.9675940871238708, |
| "learning_rate": 5.171317044961689e-05, |
| "loss": 1.5403, |
| "step": 177000 |
| }, |
| { |
| "epoch": 12.470141913727694, |
| "grad_norm": 0.9416015148162842, |
| "learning_rate": 5.156859910365765e-05, |
| "loss": 1.539, |
| "step": 177500 |
| }, |
| { |
| "epoch": 12.505269074048053, |
| "grad_norm": 0.9610737562179565, |
| "learning_rate": 5.1424027757698423e-05, |
| "loss": 1.5417, |
| "step": 178000 |
| }, |
| { |
| "epoch": 12.540396234368414, |
| "grad_norm": 0.9966002106666565, |
| "learning_rate": 5.127945641173919e-05, |
| "loss": 1.5422, |
| "step": 178500 |
| }, |
| { |
| "epoch": 12.575523394688773, |
| "grad_norm": 0.9663391709327698, |
| "learning_rate": 5.113488506577997e-05, |
| "loss": 1.5386, |
| "step": 179000 |
| }, |
| { |
| "epoch": 12.610650555009133, |
| "grad_norm": 0.9937753081321716, |
| "learning_rate": 5.0990313719820736e-05, |
| "loss": 1.5378, |
| "step": 179500 |
| }, |
| { |
| "epoch": 12.645777715329492, |
| "grad_norm": 0.9821291565895081, |
| "learning_rate": 5.0845742373861507e-05, |
| "loss": 1.5384, |
| "step": 180000 |
| }, |
| { |
| "epoch": 12.645777715329492, |
| "eval_accuracy": 0.7132840020003155, |
| "eval_loss": 1.296420693397522, |
| "eval_runtime": 166.1807, |
| "eval_samples_per_second": 726.354, |
| "eval_steps_per_second": 4.543, |
| "step": 180000 |
| }, |
| { |
| "epoch": 12.680904875649853, |
| "grad_norm": 0.9923209547996521, |
| "learning_rate": 5.070117102790227e-05, |
| "loss": 1.5362, |
| "step": 180500 |
| }, |
| { |
| "epoch": 12.716032035970212, |
| "grad_norm": 0.9979475140571594, |
| "learning_rate": 5.055659968194304e-05, |
| "loss": 1.5373, |
| "step": 181000 |
| }, |
| { |
| "epoch": 12.751159196290573, |
| "grad_norm": 0.9900950789451599, |
| "learning_rate": 5.0412028335983805e-05, |
| "loss": 1.5367, |
| "step": 181500 |
| }, |
| { |
| "epoch": 12.786286356610931, |
| "grad_norm": 0.9692813754081726, |
| "learning_rate": 5.0267456990024576e-05, |
| "loss": 1.5375, |
| "step": 182000 |
| }, |
| { |
| "epoch": 12.82141351693129, |
| "grad_norm": 0.9559634327888489, |
| "learning_rate": 5.0122885644065354e-05, |
| "loss": 1.5375, |
| "step": 182500 |
| }, |
| { |
| "epoch": 12.856540677251651, |
| "grad_norm": 1.0166131258010864, |
| "learning_rate": 4.997831429810612e-05, |
| "loss": 1.5353, |
| "step": 183000 |
| }, |
| { |
| "epoch": 12.89166783757201, |
| "grad_norm": 0.9816883206367493, |
| "learning_rate": 4.983374295214689e-05, |
| "loss": 1.5347, |
| "step": 183500 |
| }, |
| { |
| "epoch": 12.92679499789237, |
| "grad_norm": 0.9761995673179626, |
| "learning_rate": 4.968917160618766e-05, |
| "loss": 1.5312, |
| "step": 184000 |
| }, |
| { |
| "epoch": 12.96192215821273, |
| "grad_norm": 0.9669660329818726, |
| "learning_rate": 4.954460026022842e-05, |
| "loss": 1.5364, |
| "step": 184500 |
| }, |
| { |
| "epoch": 12.99704931853309, |
| "grad_norm": 0.9894793033599854, |
| "learning_rate": 4.9400028914269194e-05, |
| "loss": 1.534, |
| "step": 185000 |
| }, |
| { |
| "epoch": 13.032176478853449, |
| "grad_norm": 0.9801907539367676, |
| "learning_rate": 4.9255457568309965e-05, |
| "loss": 1.5292, |
| "step": 185500 |
| }, |
| { |
| "epoch": 13.06730363917381, |
| "grad_norm": 0.9880996346473694, |
| "learning_rate": 4.9110886222350736e-05, |
| "loss": 1.5301, |
| "step": 186000 |
| }, |
| { |
| "epoch": 13.102430799494169, |
| "grad_norm": 0.9890075325965881, |
| "learning_rate": 4.89663148763915e-05, |
| "loss": 1.5297, |
| "step": 186500 |
| }, |
| { |
| "epoch": 13.13755795981453, |
| "grad_norm": 0.9811028242111206, |
| "learning_rate": 4.882174353043227e-05, |
| "loss": 1.5272, |
| "step": 187000 |
| }, |
| { |
| "epoch": 13.172685120134888, |
| "grad_norm": 0.992874801158905, |
| "learning_rate": 4.867717218447304e-05, |
| "loss": 1.5273, |
| "step": 187500 |
| }, |
| { |
| "epoch": 13.207812280455247, |
| "grad_norm": 0.9366714358329773, |
| "learning_rate": 4.853260083851381e-05, |
| "loss": 1.529, |
| "step": 188000 |
| }, |
| { |
| "epoch": 13.242939440775608, |
| "grad_norm": 0.951941967010498, |
| "learning_rate": 4.8388029492554576e-05, |
| "loss": 1.5276, |
| "step": 188500 |
| }, |
| { |
| "epoch": 13.278066601095967, |
| "grad_norm": 0.9553075432777405, |
| "learning_rate": 4.824345814659535e-05, |
| "loss": 1.5269, |
| "step": 189000 |
| }, |
| { |
| "epoch": 13.313193761416327, |
| "grad_norm": 0.970656156539917, |
| "learning_rate": 4.809888680063612e-05, |
| "loss": 1.5306, |
| "step": 189500 |
| }, |
| { |
| "epoch": 13.348320921736686, |
| "grad_norm": 0.9519357085227966, |
| "learning_rate": 4.795431545467688e-05, |
| "loss": 1.527, |
| "step": 190000 |
| }, |
| { |
| "epoch": 13.348320921736686, |
| "eval_accuracy": 0.7144312153355764, |
| "eval_loss": 1.2920178174972534, |
| "eval_runtime": 167.9005, |
| "eval_samples_per_second": 718.914, |
| "eval_steps_per_second": 4.497, |
| "step": 190000 |
| }, |
| { |
| "epoch": 13.383448082057047, |
| "grad_norm": 0.9864979386329651, |
| "learning_rate": 4.780974410871765e-05, |
| "loss": 1.5239, |
| "step": 190500 |
| }, |
| { |
| "epoch": 13.418575242377406, |
| "grad_norm": 0.9722352027893066, |
| "learning_rate": 4.766517276275842e-05, |
| "loss": 1.5266, |
| "step": 191000 |
| }, |
| { |
| "epoch": 13.453702402697767, |
| "grad_norm": 0.943938136100769, |
| "learning_rate": 4.7520601416799194e-05, |
| "loss": 1.5235, |
| "step": 191500 |
| }, |
| { |
| "epoch": 13.488829563018125, |
| "grad_norm": 0.9698721170425415, |
| "learning_rate": 4.737603007083996e-05, |
| "loss": 1.5253, |
| "step": 192000 |
| }, |
| { |
| "epoch": 13.523956723338486, |
| "grad_norm": 0.9482982754707336, |
| "learning_rate": 4.723145872488073e-05, |
| "loss": 1.5274, |
| "step": 192500 |
| }, |
| { |
| "epoch": 13.559083883658845, |
| "grad_norm": 0.9819760918617249, |
| "learning_rate": 4.70868873789215e-05, |
| "loss": 1.5246, |
| "step": 193000 |
| }, |
| { |
| "epoch": 13.594211043979204, |
| "grad_norm": 1.0020642280578613, |
| "learning_rate": 4.694231603296227e-05, |
| "loss": 1.5251, |
| "step": 193500 |
| }, |
| { |
| "epoch": 13.629338204299565, |
| "grad_norm": 1.013135313987732, |
| "learning_rate": 4.6797744687003034e-05, |
| "loss": 1.5219, |
| "step": 194000 |
| }, |
| { |
| "epoch": 13.664465364619923, |
| "grad_norm": 0.9866718649864197, |
| "learning_rate": 4.665317334104381e-05, |
| "loss": 1.5205, |
| "step": 194500 |
| }, |
| { |
| "epoch": 13.699592524940284, |
| "grad_norm": 1.0128811597824097, |
| "learning_rate": 4.6508601995084576e-05, |
| "loss": 1.5229, |
| "step": 195000 |
| }, |
| { |
| "epoch": 13.734719685260643, |
| "grad_norm": 0.9859340786933899, |
| "learning_rate": 4.6364030649125346e-05, |
| "loss": 1.522, |
| "step": 195500 |
| }, |
| { |
| "epoch": 13.769846845581004, |
| "grad_norm": 0.9608091711997986, |
| "learning_rate": 4.621945930316611e-05, |
| "loss": 1.5217, |
| "step": 196000 |
| }, |
| { |
| "epoch": 13.804974005901363, |
| "grad_norm": 1.0180490016937256, |
| "learning_rate": 4.607488795720689e-05, |
| "loss": 1.5199, |
| "step": 196500 |
| }, |
| { |
| "epoch": 13.840101166221723, |
| "grad_norm": 0.9972442388534546, |
| "learning_rate": 4.593031661124765e-05, |
| "loss": 1.5212, |
| "step": 197000 |
| }, |
| { |
| "epoch": 13.875228326542082, |
| "grad_norm": 0.9892627596855164, |
| "learning_rate": 4.578574526528842e-05, |
| "loss": 1.5188, |
| "step": 197500 |
| }, |
| { |
| "epoch": 13.910355486862443, |
| "grad_norm": 0.9838631749153137, |
| "learning_rate": 4.564117391932919e-05, |
| "loss": 1.5198, |
| "step": 198000 |
| }, |
| { |
| "epoch": 13.945482647182802, |
| "grad_norm": 1.0061650276184082, |
| "learning_rate": 4.5496602573369964e-05, |
| "loss": 1.5182, |
| "step": 198500 |
| }, |
| { |
| "epoch": 13.98060980750316, |
| "grad_norm": 0.9801396131515503, |
| "learning_rate": 4.535203122741073e-05, |
| "loss": 1.5229, |
| "step": 199000 |
| }, |
| { |
| "epoch": 14.015736967823521, |
| "grad_norm": 0.9728179574012756, |
| "learning_rate": 4.52074598814515e-05, |
| "loss": 1.5179, |
| "step": 199500 |
| }, |
| { |
| "epoch": 14.05086412814388, |
| "grad_norm": 0.9667819142341614, |
| "learning_rate": 4.506288853549227e-05, |
| "loss": 1.5186, |
| "step": 200000 |
| }, |
| { |
| "epoch": 14.05086412814388, |
| "eval_accuracy": 0.7159755077448605, |
| "eval_loss": 1.2808479070663452, |
| "eval_runtime": 178.7243, |
| "eval_samples_per_second": 675.375, |
| "eval_steps_per_second": 4.224, |
| "step": 200000 |
| }, |
| { |
| "epoch": 14.08599128846424, |
| "grad_norm": 0.9804860353469849, |
| "learning_rate": 4.491831718953304e-05, |
| "loss": 1.5146, |
| "step": 200500 |
| }, |
| { |
| "epoch": 14.1211184487846, |
| "grad_norm": 0.998810887336731, |
| "learning_rate": 4.4773745843573805e-05, |
| "loss": 1.5159, |
| "step": 201000 |
| }, |
| { |
| "epoch": 14.15624560910496, |
| "grad_norm": 0.9885842800140381, |
| "learning_rate": 4.4629174497614575e-05, |
| "loss": 1.514, |
| "step": 201500 |
| }, |
| { |
| "epoch": 14.19137276942532, |
| "grad_norm": 0.9932366609573364, |
| "learning_rate": 4.4484603151655346e-05, |
| "loss": 1.5121, |
| "step": 202000 |
| }, |
| { |
| "epoch": 14.22649992974568, |
| "grad_norm": 0.9998967051506042, |
| "learning_rate": 4.434003180569612e-05, |
| "loss": 1.5146, |
| "step": 202500 |
| }, |
| { |
| "epoch": 14.261627090066039, |
| "grad_norm": 1.0137783288955688, |
| "learning_rate": 4.419546045973688e-05, |
| "loss": 1.515, |
| "step": 203000 |
| }, |
| { |
| "epoch": 14.2967542503864, |
| "grad_norm": 1.0021193027496338, |
| "learning_rate": 4.405088911377765e-05, |
| "loss": 1.5133, |
| "step": 203500 |
| }, |
| { |
| "epoch": 14.331881410706758, |
| "grad_norm": 1.0028791427612305, |
| "learning_rate": 4.390631776781842e-05, |
| "loss": 1.5157, |
| "step": 204000 |
| }, |
| { |
| "epoch": 14.367008571027117, |
| "grad_norm": 0.9845914244651794, |
| "learning_rate": 4.3761746421859187e-05, |
| "loss": 1.5129, |
| "step": 204500 |
| }, |
| { |
| "epoch": 14.402135731347478, |
| "grad_norm": 0.981665849685669, |
| "learning_rate": 4.361717507589996e-05, |
| "loss": 1.5124, |
| "step": 205000 |
| }, |
| { |
| "epoch": 14.437262891667837, |
| "grad_norm": 0.9732480049133301, |
| "learning_rate": 4.347260372994073e-05, |
| "loss": 1.5135, |
| "step": 205500 |
| }, |
| { |
| "epoch": 14.472390051988198, |
| "grad_norm": 1.0320522785186768, |
| "learning_rate": 4.33280323839815e-05, |
| "loss": 1.5118, |
| "step": 206000 |
| }, |
| { |
| "epoch": 14.507517212308557, |
| "grad_norm": 1.0051497220993042, |
| "learning_rate": 4.318346103802226e-05, |
| "loss": 1.5101, |
| "step": 206500 |
| }, |
| { |
| "epoch": 14.542644372628917, |
| "grad_norm": 0.9936639666557312, |
| "learning_rate": 4.3038889692063034e-05, |
| "loss": 1.5088, |
| "step": 207000 |
| }, |
| { |
| "epoch": 14.577771532949276, |
| "grad_norm": 0.994345486164093, |
| "learning_rate": 4.2894318346103804e-05, |
| "loss": 1.5101, |
| "step": 207500 |
| }, |
| { |
| "epoch": 14.612898693269637, |
| "grad_norm": 1.0319503545761108, |
| "learning_rate": 4.2749747000144575e-05, |
| "loss": 1.5105, |
| "step": 208000 |
| }, |
| { |
| "epoch": 14.648025853589996, |
| "grad_norm": 1.0258818864822388, |
| "learning_rate": 4.260517565418534e-05, |
| "loss": 1.5092, |
| "step": 208500 |
| }, |
| { |
| "epoch": 14.683153013910356, |
| "grad_norm": 1.015943169593811, |
| "learning_rate": 4.246060430822611e-05, |
| "loss": 1.5096, |
| "step": 209000 |
| }, |
| { |
| "epoch": 14.718280174230715, |
| "grad_norm": 0.9799129366874695, |
| "learning_rate": 4.231603296226688e-05, |
| "loss": 1.508, |
| "step": 209500 |
| }, |
| { |
| "epoch": 14.753407334551074, |
| "grad_norm": 0.9862245917320251, |
| "learning_rate": 4.217146161630765e-05, |
| "loss": 1.5086, |
| "step": 210000 |
| }, |
| { |
| "epoch": 14.753407334551074, |
| "eval_accuracy": 0.7173609367641633, |
| "eval_loss": 1.2729206085205078, |
| "eval_runtime": 164.5565, |
| "eval_samples_per_second": 733.523, |
| "eval_steps_per_second": 4.588, |
| "step": 210000 |
| }, |
| { |
| "epoch": 14.788534494871435, |
| "grad_norm": 0.9937060475349426, |
| "learning_rate": 4.2026890270348416e-05, |
| "loss": 1.5084, |
| "step": 210500 |
| }, |
| { |
| "epoch": 14.823661655191794, |
| "grad_norm": 1.0387269258499146, |
| "learning_rate": 4.1882318924389186e-05, |
| "loss": 1.5076, |
| "step": 211000 |
| }, |
| { |
| "epoch": 14.858788815512154, |
| "grad_norm": 0.9882417917251587, |
| "learning_rate": 4.173774757842996e-05, |
| "loss": 1.5078, |
| "step": 211500 |
| }, |
| { |
| "epoch": 14.893915975832513, |
| "grad_norm": 1.020268440246582, |
| "learning_rate": 4.159317623247073e-05, |
| "loss": 1.5059, |
| "step": 212000 |
| }, |
| { |
| "epoch": 14.929043136152874, |
| "grad_norm": 1.0184102058410645, |
| "learning_rate": 4.144860488651149e-05, |
| "loss": 1.5084, |
| "step": 212500 |
| }, |
| { |
| "epoch": 14.964170296473233, |
| "grad_norm": 0.9803560972213745, |
| "learning_rate": 4.130403354055227e-05, |
| "loss": 1.5082, |
| "step": 213000 |
| }, |
| { |
| "epoch": 14.999297456793593, |
| "grad_norm": 0.9785324931144714, |
| "learning_rate": 4.1159462194593033e-05, |
| "loss": 1.508, |
| "step": 213500 |
| }, |
| { |
| "epoch": 15.034424617113952, |
| "grad_norm": 1.0045809745788574, |
| "learning_rate": 4.1014890848633804e-05, |
| "loss": 1.5018, |
| "step": 214000 |
| }, |
| { |
| "epoch": 15.069551777434313, |
| "grad_norm": 0.9606882333755493, |
| "learning_rate": 4.087031950267457e-05, |
| "loss": 1.5031, |
| "step": 214500 |
| }, |
| { |
| "epoch": 15.104678937754672, |
| "grad_norm": 0.968102753162384, |
| "learning_rate": 4.0725748156715346e-05, |
| "loss": 1.5032, |
| "step": 215000 |
| }, |
| { |
| "epoch": 15.13980609807503, |
| "grad_norm": 1.017193078994751, |
| "learning_rate": 4.058117681075611e-05, |
| "loss": 1.4991, |
| "step": 215500 |
| }, |
| { |
| "epoch": 15.174933258395392, |
| "grad_norm": 1.001583456993103, |
| "learning_rate": 4.043660546479688e-05, |
| "loss": 1.5041, |
| "step": 216000 |
| }, |
| { |
| "epoch": 15.21006041871575, |
| "grad_norm": 1.0024583339691162, |
| "learning_rate": 4.0292034118837645e-05, |
| "loss": 1.5032, |
| "step": 216500 |
| }, |
| { |
| "epoch": 15.245187579036111, |
| "grad_norm": 1.0062695741653442, |
| "learning_rate": 4.014746277287842e-05, |
| "loss": 1.5022, |
| "step": 217000 |
| }, |
| { |
| "epoch": 15.28031473935647, |
| "grad_norm": 0.9849976301193237, |
| "learning_rate": 4.0002891426919186e-05, |
| "loss": 1.5005, |
| "step": 217500 |
| }, |
| { |
| "epoch": 15.31544189967683, |
| "grad_norm": 1.0209317207336426, |
| "learning_rate": 3.985832008095996e-05, |
| "loss": 1.5011, |
| "step": 218000 |
| }, |
| { |
| "epoch": 15.35056905999719, |
| "grad_norm": 1.0125476121902466, |
| "learning_rate": 3.971374873500072e-05, |
| "loss": 1.5022, |
| "step": 218500 |
| }, |
| { |
| "epoch": 15.38569622031755, |
| "grad_norm": 1.0083107948303223, |
| "learning_rate": 3.95691773890415e-05, |
| "loss": 1.5005, |
| "step": 219000 |
| }, |
| { |
| "epoch": 15.42082338063791, |
| "grad_norm": 1.0500251054763794, |
| "learning_rate": 3.942460604308226e-05, |
| "loss": 1.5007, |
| "step": 219500 |
| }, |
| { |
| "epoch": 15.45595054095827, |
| "grad_norm": 1.0102968215942383, |
| "learning_rate": 3.928003469712303e-05, |
| "loss": 1.4991, |
| "step": 220000 |
| }, |
| { |
| "epoch": 15.45595054095827, |
| "eval_accuracy": 0.7190619701240156, |
| "eval_loss": 1.2636791467666626, |
| "eval_runtime": 174.3942, |
| "eval_samples_per_second": 692.144, |
| "eval_steps_per_second": 4.329, |
| "step": 220000 |
| }, |
| { |
| "epoch": 15.491077701278629, |
| "grad_norm": 0.9830746650695801, |
| "learning_rate": 3.9135463351163804e-05, |
| "loss": 1.5007, |
| "step": 220500 |
| }, |
| { |
| "epoch": 15.526204861598988, |
| "grad_norm": 1.0218993425369263, |
| "learning_rate": 3.899089200520457e-05, |
| "loss": 1.5015, |
| "step": 221000 |
| }, |
| { |
| "epoch": 15.561332021919348, |
| "grad_norm": 1.0001802444458008, |
| "learning_rate": 3.884632065924534e-05, |
| "loss": 1.4994, |
| "step": 221500 |
| }, |
| { |
| "epoch": 15.596459182239707, |
| "grad_norm": 0.9662476181983948, |
| "learning_rate": 3.870174931328611e-05, |
| "loss": 1.5, |
| "step": 222000 |
| }, |
| { |
| "epoch": 15.631586342560068, |
| "grad_norm": 0.9796843528747559, |
| "learning_rate": 3.855717796732688e-05, |
| "loss": 1.4967, |
| "step": 222500 |
| }, |
| { |
| "epoch": 15.666713502880427, |
| "grad_norm": 1.0108364820480347, |
| "learning_rate": 3.8412606621367644e-05, |
| "loss": 1.4982, |
| "step": 223000 |
| }, |
| { |
| "epoch": 15.701840663200787, |
| "grad_norm": 1.0097131729125977, |
| "learning_rate": 3.8268035275408415e-05, |
| "loss": 1.4978, |
| "step": 223500 |
| }, |
| { |
| "epoch": 15.736967823521146, |
| "grad_norm": 1.0322747230529785, |
| "learning_rate": 3.812346392944918e-05, |
| "loss": 1.4959, |
| "step": 224000 |
| }, |
| { |
| "epoch": 15.772094983841507, |
| "grad_norm": 0.9876171350479126, |
| "learning_rate": 3.797889258348996e-05, |
| "loss": 1.4947, |
| "step": 224500 |
| }, |
| { |
| "epoch": 15.807222144161866, |
| "grad_norm": 1.0190730094909668, |
| "learning_rate": 3.783432123753072e-05, |
| "loss": 1.4936, |
| "step": 225000 |
| }, |
| { |
| "epoch": 15.842349304482227, |
| "grad_norm": 1.0361992120742798, |
| "learning_rate": 3.768974989157149e-05, |
| "loss": 1.4942, |
| "step": 225500 |
| }, |
| { |
| "epoch": 15.877476464802585, |
| "grad_norm": 1.0055601596832275, |
| "learning_rate": 3.754517854561226e-05, |
| "loss": 1.498, |
| "step": 226000 |
| }, |
| { |
| "epoch": 15.912603625122944, |
| "grad_norm": 1.028681755065918, |
| "learning_rate": 3.740060719965303e-05, |
| "loss": 1.498, |
| "step": 226500 |
| }, |
| { |
| "epoch": 15.947730785443305, |
| "grad_norm": 1.008866310119629, |
| "learning_rate": 3.72560358536938e-05, |
| "loss": 1.4933, |
| "step": 227000 |
| }, |
| { |
| "epoch": 15.982857945763664, |
| "grad_norm": 1.0164419412612915, |
| "learning_rate": 3.711146450773457e-05, |
| "loss": 1.4956, |
| "step": 227500 |
| }, |
| { |
| "epoch": 16.017985106084023, |
| "grad_norm": 1.0534642934799194, |
| "learning_rate": 3.696689316177534e-05, |
| "loss": 1.492, |
| "step": 228000 |
| }, |
| { |
| "epoch": 16.053112266404383, |
| "grad_norm": 1.0368808507919312, |
| "learning_rate": 3.682232181581611e-05, |
| "loss": 1.4905, |
| "step": 228500 |
| }, |
| { |
| "epoch": 16.088239426724744, |
| "grad_norm": 1.0361706018447876, |
| "learning_rate": 3.667775046985687e-05, |
| "loss": 1.4909, |
| "step": 229000 |
| }, |
| { |
| "epoch": 16.123366587045105, |
| "grad_norm": 0.9920340776443481, |
| "learning_rate": 3.6533179123897644e-05, |
| "loss": 1.491, |
| "step": 229500 |
| }, |
| { |
| "epoch": 16.158493747365462, |
| "grad_norm": 1.0142643451690674, |
| "learning_rate": 3.6388607777938415e-05, |
| "loss": 1.4936, |
| "step": 230000 |
| }, |
| { |
| "epoch": 16.158493747365462, |
| "eval_accuracy": 0.7198381367948815, |
| "eval_loss": 1.25892174243927, |
| "eval_runtime": 161.3083, |
| "eval_samples_per_second": 748.294, |
| "eval_steps_per_second": 4.68, |
| "step": 230000 |
| }, |
| { |
| "epoch": 16.193620907685823, |
| "grad_norm": 1.016242265701294, |
| "learning_rate": 3.6244036431979186e-05, |
| "loss": 1.4886, |
| "step": 230500 |
| }, |
| { |
| "epoch": 16.228748068006183, |
| "grad_norm": 0.9944464564323425, |
| "learning_rate": 3.609946508601995e-05, |
| "loss": 1.4907, |
| "step": 231000 |
| }, |
| { |
| "epoch": 16.26387522832654, |
| "grad_norm": 1.0434259176254272, |
| "learning_rate": 3.595489374006073e-05, |
| "loss": 1.487, |
| "step": 231500 |
| }, |
| { |
| "epoch": 16.2990023886469, |
| "grad_norm": 0.9813429713249207, |
| "learning_rate": 3.581032239410149e-05, |
| "loss": 1.4914, |
| "step": 232000 |
| }, |
| { |
| "epoch": 16.33412954896726, |
| "grad_norm": 1.0342869758605957, |
| "learning_rate": 3.566575104814226e-05, |
| "loss": 1.4886, |
| "step": 232500 |
| }, |
| { |
| "epoch": 16.369256709287622, |
| "grad_norm": 1.0258421897888184, |
| "learning_rate": 3.5521179702183026e-05, |
| "loss": 1.4894, |
| "step": 233000 |
| }, |
| { |
| "epoch": 16.40438386960798, |
| "grad_norm": 1.0049115419387817, |
| "learning_rate": 3.5376608356223804e-05, |
| "loss": 1.4888, |
| "step": 233500 |
| }, |
| { |
| "epoch": 16.43951102992834, |
| "grad_norm": 0.9911952614784241, |
| "learning_rate": 3.523203701026457e-05, |
| "loss": 1.4877, |
| "step": 234000 |
| }, |
| { |
| "epoch": 16.4746381902487, |
| "grad_norm": 0.9992234706878662, |
| "learning_rate": 3.508746566430534e-05, |
| "loss": 1.4891, |
| "step": 234500 |
| }, |
| { |
| "epoch": 16.50976535056906, |
| "grad_norm": 1.027385950088501, |
| "learning_rate": 3.49428943183461e-05, |
| "loss": 1.4884, |
| "step": 235000 |
| }, |
| { |
| "epoch": 16.54489251088942, |
| "grad_norm": 1.0000590085983276, |
| "learning_rate": 3.479832297238687e-05, |
| "loss": 1.4855, |
| "step": 235500 |
| }, |
| { |
| "epoch": 16.58001967120978, |
| "grad_norm": 0.9984703660011292, |
| "learning_rate": 3.4653751626427644e-05, |
| "loss": 1.4886, |
| "step": 236000 |
| }, |
| { |
| "epoch": 16.61514683153014, |
| "grad_norm": 1.0558722019195557, |
| "learning_rate": 3.4509180280468415e-05, |
| "loss": 1.4853, |
| "step": 236500 |
| }, |
| { |
| "epoch": 16.650273991850497, |
| "grad_norm": 1.062028408050537, |
| "learning_rate": 3.436460893450918e-05, |
| "loss": 1.4918, |
| "step": 237000 |
| }, |
| { |
| "epoch": 16.685401152170858, |
| "grad_norm": 1.066724181175232, |
| "learning_rate": 3.422003758854995e-05, |
| "loss": 1.488, |
| "step": 237500 |
| }, |
| { |
| "epoch": 16.72052831249122, |
| "grad_norm": 1.0459178686141968, |
| "learning_rate": 3.407546624259072e-05, |
| "loss": 1.4862, |
| "step": 238000 |
| }, |
| { |
| "epoch": 16.75565547281158, |
| "grad_norm": 1.0179390907287598, |
| "learning_rate": 3.393089489663149e-05, |
| "loss": 1.488, |
| "step": 238500 |
| }, |
| { |
| "epoch": 16.790782633131936, |
| "grad_norm": 1.0292638540267944, |
| "learning_rate": 3.378632355067226e-05, |
| "loss": 1.4855, |
| "step": 239000 |
| }, |
| { |
| "epoch": 16.825909793452297, |
| "grad_norm": 1.0212992429733276, |
| "learning_rate": 3.3641752204713026e-05, |
| "loss": 1.4845, |
| "step": 239500 |
| }, |
| { |
| "epoch": 16.861036953772658, |
| "grad_norm": 1.0024462938308716, |
| "learning_rate": 3.3497180858753797e-05, |
| "loss": 1.4843, |
| "step": 240000 |
| }, |
| { |
| "epoch": 16.861036953772658, |
| "eval_accuracy": 0.7209039011781289, |
| "eval_loss": 1.2533527612686157, |
| "eval_runtime": 164.4861, |
| "eval_samples_per_second": 733.837, |
| "eval_steps_per_second": 4.59, |
| "step": 240000 |
| }, |
| { |
| "epoch": 16.89616411409302, |
| "grad_norm": 1.0215648412704468, |
| "learning_rate": 3.335260951279456e-05, |
| "loss": 1.4838, |
| "step": 240500 |
| }, |
| { |
| "epoch": 16.931291274413375, |
| "grad_norm": 1.0557113885879517, |
| "learning_rate": 3.320803816683534e-05, |
| "loss": 1.4867, |
| "step": 241000 |
| }, |
| { |
| "epoch": 16.966418434733736, |
| "grad_norm": 1.0456198453903198, |
| "learning_rate": 3.30634668208761e-05, |
| "loss": 1.483, |
| "step": 241500 |
| }, |
| { |
| "epoch": 17.001545595054097, |
| "grad_norm": 1.0394978523254395, |
| "learning_rate": 3.291889547491687e-05, |
| "loss": 1.4869, |
| "step": 242000 |
| }, |
| { |
| "epoch": 17.036672755374454, |
| "grad_norm": 1.0153169631958008, |
| "learning_rate": 3.277432412895764e-05, |
| "loss": 1.484, |
| "step": 242500 |
| }, |
| { |
| "epoch": 17.071799915694815, |
| "grad_norm": 1.0242444276809692, |
| "learning_rate": 3.2629752782998414e-05, |
| "loss": 1.4805, |
| "step": 243000 |
| }, |
| { |
| "epoch": 17.106927076015175, |
| "grad_norm": 1.0739620923995972, |
| "learning_rate": 3.248518143703918e-05, |
| "loss": 1.4819, |
| "step": 243500 |
| }, |
| { |
| "epoch": 17.142054236335536, |
| "grad_norm": 1.0303699970245361, |
| "learning_rate": 3.234061009107995e-05, |
| "loss": 1.4791, |
| "step": 244000 |
| }, |
| { |
| "epoch": 17.177181396655893, |
| "grad_norm": 1.0153394937515259, |
| "learning_rate": 3.219603874512072e-05, |
| "loss": 1.479, |
| "step": 244500 |
| }, |
| { |
| "epoch": 17.212308556976254, |
| "grad_norm": 1.0197999477386475, |
| "learning_rate": 3.205146739916149e-05, |
| "loss": 1.4816, |
| "step": 245000 |
| }, |
| { |
| "epoch": 17.247435717296614, |
| "grad_norm": 1.031987190246582, |
| "learning_rate": 3.1906896053202255e-05, |
| "loss": 1.4781, |
| "step": 245500 |
| }, |
| { |
| "epoch": 17.282562877616975, |
| "grad_norm": 1.036007046699524, |
| "learning_rate": 3.1762324707243026e-05, |
| "loss": 1.4829, |
| "step": 246000 |
| }, |
| { |
| "epoch": 17.317690037937332, |
| "grad_norm": 1.0153439044952393, |
| "learning_rate": 3.1617753361283796e-05, |
| "loss": 1.4793, |
| "step": 246500 |
| }, |
| { |
| "epoch": 17.352817198257693, |
| "grad_norm": 1.0097757577896118, |
| "learning_rate": 3.147318201532457e-05, |
| "loss": 1.4782, |
| "step": 247000 |
| }, |
| { |
| "epoch": 17.387944358578054, |
| "grad_norm": 1.0287526845932007, |
| "learning_rate": 3.132861066936533e-05, |
| "loss": 1.4808, |
| "step": 247500 |
| }, |
| { |
| "epoch": 17.42307151889841, |
| "grad_norm": 1.0353249311447144, |
| "learning_rate": 3.11840393234061e-05, |
| "loss": 1.4742, |
| "step": 248000 |
| }, |
| { |
| "epoch": 17.45819867921877, |
| "grad_norm": 1.0666767358779907, |
| "learning_rate": 3.103946797744687e-05, |
| "loss": 1.4768, |
| "step": 248500 |
| }, |
| { |
| "epoch": 17.493325839539132, |
| "grad_norm": 1.0369137525558472, |
| "learning_rate": 3.0894896631487643e-05, |
| "loss": 1.4757, |
| "step": 249000 |
| }, |
| { |
| "epoch": 17.528452999859493, |
| "grad_norm": 1.0696829557418823, |
| "learning_rate": 3.075032528552841e-05, |
| "loss": 1.4746, |
| "step": 249500 |
| }, |
| { |
| "epoch": 17.56358016017985, |
| "grad_norm": 1.0157139301300049, |
| "learning_rate": 3.060575393956918e-05, |
| "loss": 1.4763, |
| "step": 250000 |
| }, |
| { |
| "epoch": 17.56358016017985, |
| "eval_accuracy": 0.7219265186910462, |
| "eval_loss": 1.2467384338378906, |
| "eval_runtime": 162.8498, |
| "eval_samples_per_second": 741.21, |
| "eval_steps_per_second": 4.636, |
| "step": 250000 |
| }, |
| { |
| "epoch": 17.59870732050021, |
| "grad_norm": 1.040384292602539, |
| "learning_rate": 3.046118259360995e-05, |
| "loss": 1.4802, |
| "step": 250500 |
| }, |
| { |
| "epoch": 17.63383448082057, |
| "grad_norm": 1.018039584159851, |
| "learning_rate": 3.0316611247650716e-05, |
| "loss": 1.4741, |
| "step": 251000 |
| }, |
| { |
| "epoch": 17.668961641140932, |
| "grad_norm": 1.0444985628128052, |
| "learning_rate": 3.0172039901691484e-05, |
| "loss": 1.4759, |
| "step": 251500 |
| }, |
| { |
| "epoch": 17.70408880146129, |
| "grad_norm": 1.0292766094207764, |
| "learning_rate": 3.0027468555732258e-05, |
| "loss": 1.477, |
| "step": 252000 |
| }, |
| { |
| "epoch": 17.73921596178165, |
| "grad_norm": 1.0553712844848633, |
| "learning_rate": 2.9882897209773025e-05, |
| "loss": 1.4761, |
| "step": 252500 |
| }, |
| { |
| "epoch": 17.77434312210201, |
| "grad_norm": 1.0308884382247925, |
| "learning_rate": 2.9738325863813793e-05, |
| "loss": 1.4745, |
| "step": 253000 |
| }, |
| { |
| "epoch": 17.809470282422367, |
| "grad_norm": 1.0115193128585815, |
| "learning_rate": 2.959375451785456e-05, |
| "loss": 1.4764, |
| "step": 253500 |
| }, |
| { |
| "epoch": 17.844597442742728, |
| "grad_norm": 1.0114680528640747, |
| "learning_rate": 2.9449183171895334e-05, |
| "loss": 1.4739, |
| "step": 254000 |
| }, |
| { |
| "epoch": 17.87972460306309, |
| "grad_norm": 1.0322569608688354, |
| "learning_rate": 2.93046118259361e-05, |
| "loss": 1.4779, |
| "step": 254500 |
| }, |
| { |
| "epoch": 17.91485176338345, |
| "grad_norm": 1.0674282312393188, |
| "learning_rate": 2.916004047997687e-05, |
| "loss": 1.4757, |
| "step": 255000 |
| }, |
| { |
| "epoch": 17.949978923703807, |
| "grad_norm": 1.041752576828003, |
| "learning_rate": 2.9015469134017636e-05, |
| "loss": 1.4766, |
| "step": 255500 |
| }, |
| { |
| "epoch": 17.985106084024167, |
| "grad_norm": 1.0347331762313843, |
| "learning_rate": 2.887089778805841e-05, |
| "loss": 1.4735, |
| "step": 256000 |
| }, |
| { |
| "epoch": 18.020233244344528, |
| "grad_norm": 1.0649460554122925, |
| "learning_rate": 2.8726326442099178e-05, |
| "loss": 1.4747, |
| "step": 256500 |
| }, |
| { |
| "epoch": 18.05536040466489, |
| "grad_norm": 1.0295964479446411, |
| "learning_rate": 2.8581755096139945e-05, |
| "loss": 1.4731, |
| "step": 257000 |
| }, |
| { |
| "epoch": 18.090487564985246, |
| "grad_norm": 1.0168757438659668, |
| "learning_rate": 2.843718375018072e-05, |
| "loss": 1.4696, |
| "step": 257500 |
| }, |
| { |
| "epoch": 18.125614725305606, |
| "grad_norm": 1.060194969177246, |
| "learning_rate": 2.8292612404221487e-05, |
| "loss": 1.4732, |
| "step": 258000 |
| }, |
| { |
| "epoch": 18.160741885625967, |
| "grad_norm": 1.0267513990402222, |
| "learning_rate": 2.8148041058262254e-05, |
| "loss": 1.4708, |
| "step": 258500 |
| }, |
| { |
| "epoch": 18.195869045946324, |
| "grad_norm": 1.0169157981872559, |
| "learning_rate": 2.800346971230302e-05, |
| "loss": 1.4721, |
| "step": 259000 |
| }, |
| { |
| "epoch": 18.230996206266685, |
| "grad_norm": 1.0195244550704956, |
| "learning_rate": 2.7858898366343792e-05, |
| "loss": 1.4723, |
| "step": 259500 |
| }, |
| { |
| "epoch": 18.266123366587045, |
| "grad_norm": 1.0502545833587646, |
| "learning_rate": 2.7714327020384563e-05, |
| "loss": 1.4701, |
| "step": 260000 |
| }, |
| { |
| "epoch": 18.266123366587045, |
| "eval_accuracy": 0.7230056212538117, |
| "eval_loss": 1.240757942199707, |
| "eval_runtime": 149.6462, |
| "eval_samples_per_second": 806.609, |
| "eval_steps_per_second": 5.045, |
| "step": 260000 |
| }, |
| { |
| "epoch": 18.301250526907406, |
| "grad_norm": 1.0381680727005005, |
| "learning_rate": 2.756975567442533e-05, |
| "loss": 1.4698, |
| "step": 260500 |
| }, |
| { |
| "epoch": 18.336377687227763, |
| "grad_norm": 1.0591702461242676, |
| "learning_rate": 2.7425184328466098e-05, |
| "loss": 1.4679, |
| "step": 261000 |
| }, |
| { |
| "epoch": 18.371504847548124, |
| "grad_norm": 1.0286833047866821, |
| "learning_rate": 2.728061298250687e-05, |
| "loss": 1.4714, |
| "step": 261500 |
| }, |
| { |
| "epoch": 18.406632007868485, |
| "grad_norm": 1.0340920686721802, |
| "learning_rate": 2.7136041636547636e-05, |
| "loss": 1.4692, |
| "step": 262000 |
| }, |
| { |
| "epoch": 18.441759168188845, |
| "grad_norm": 1.0377320051193237, |
| "learning_rate": 2.6991470290588407e-05, |
| "loss": 1.4665, |
| "step": 262500 |
| }, |
| { |
| "epoch": 18.476886328509202, |
| "grad_norm": 1.0692384243011475, |
| "learning_rate": 2.6846898944629178e-05, |
| "loss": 1.4667, |
| "step": 263000 |
| }, |
| { |
| "epoch": 18.512013488829563, |
| "grad_norm": 1.0369582176208496, |
| "learning_rate": 2.6702327598669945e-05, |
| "loss": 1.466, |
| "step": 263500 |
| }, |
| { |
| "epoch": 18.547140649149924, |
| "grad_norm": 1.0349966287612915, |
| "learning_rate": 2.6557756252710713e-05, |
| "loss": 1.4656, |
| "step": 264000 |
| }, |
| { |
| "epoch": 18.58226780947028, |
| "grad_norm": 1.0353617668151855, |
| "learning_rate": 2.641318490675148e-05, |
| "loss": 1.4663, |
| "step": 264500 |
| }, |
| { |
| "epoch": 18.61739496979064, |
| "grad_norm": 1.0669108629226685, |
| "learning_rate": 2.6268613560792254e-05, |
| "loss": 1.4699, |
| "step": 265000 |
| }, |
| { |
| "epoch": 18.652522130111002, |
| "grad_norm": 1.0231959819793701, |
| "learning_rate": 2.612404221483302e-05, |
| "loss": 1.4672, |
| "step": 265500 |
| }, |
| { |
| "epoch": 18.687649290431363, |
| "grad_norm": 1.0390506982803345, |
| "learning_rate": 2.597947086887379e-05, |
| "loss": 1.4652, |
| "step": 266000 |
| }, |
| { |
| "epoch": 18.72277645075172, |
| "grad_norm": 1.053722620010376, |
| "learning_rate": 2.5834899522914556e-05, |
| "loss": 1.4678, |
| "step": 266500 |
| }, |
| { |
| "epoch": 18.75790361107208, |
| "grad_norm": 1.0169175863265991, |
| "learning_rate": 2.569032817695533e-05, |
| "loss": 1.4675, |
| "step": 267000 |
| }, |
| { |
| "epoch": 18.79303077139244, |
| "grad_norm": 1.0717405080795288, |
| "learning_rate": 2.5545756830996098e-05, |
| "loss": 1.4688, |
| "step": 267500 |
| }, |
| { |
| "epoch": 18.828157931712802, |
| "grad_norm": 1.0536720752716064, |
| "learning_rate": 2.5401185485036865e-05, |
| "loss": 1.4664, |
| "step": 268000 |
| }, |
| { |
| "epoch": 18.86328509203316, |
| "grad_norm": 1.1142454147338867, |
| "learning_rate": 2.5256614139077633e-05, |
| "loss": 1.4659, |
| "step": 268500 |
| }, |
| { |
| "epoch": 18.89841225235352, |
| "grad_norm": 1.03350031375885, |
| "learning_rate": 2.5112042793118407e-05, |
| "loss": 1.466, |
| "step": 269000 |
| }, |
| { |
| "epoch": 18.93353941267388, |
| "grad_norm": 1.074278473854065, |
| "learning_rate": 2.4967471447159174e-05, |
| "loss": 1.4653, |
| "step": 269500 |
| }, |
| { |
| "epoch": 18.968666572994238, |
| "grad_norm": 1.0311990976333618, |
| "learning_rate": 2.4822900101199945e-05, |
| "loss": 1.4668, |
| "step": 270000 |
| }, |
| { |
| "epoch": 18.968666572994238, |
| "eval_accuracy": 0.7240031008620313, |
| "eval_loss": 1.2352993488311768, |
| "eval_runtime": 147.7789, |
| "eval_samples_per_second": 816.801, |
| "eval_steps_per_second": 5.109, |
| "step": 270000 |
| }, |
| { |
| "epoch": 19.0037937333146, |
| "grad_norm": 1.0610495805740356, |
| "learning_rate": 2.4678328755240712e-05, |
| "loss": 1.466, |
| "step": 270500 |
| }, |
| { |
| "epoch": 19.03892089363496, |
| "grad_norm": 1.0686758756637573, |
| "learning_rate": 2.4533757409281483e-05, |
| "loss": 1.4623, |
| "step": 271000 |
| }, |
| { |
| "epoch": 19.07404805395532, |
| "grad_norm": 1.0690410137176514, |
| "learning_rate": 2.438918606332225e-05, |
| "loss": 1.4627, |
| "step": 271500 |
| }, |
| { |
| "epoch": 19.109175214275677, |
| "grad_norm": 1.056462287902832, |
| "learning_rate": 2.424461471736302e-05, |
| "loss": 1.4619, |
| "step": 272000 |
| }, |
| { |
| "epoch": 19.144302374596037, |
| "grad_norm": 1.0437265634536743, |
| "learning_rate": 2.410004337140379e-05, |
| "loss": 1.4626, |
| "step": 272500 |
| }, |
| { |
| "epoch": 19.179429534916398, |
| "grad_norm": 1.047309160232544, |
| "learning_rate": 2.395547202544456e-05, |
| "loss": 1.4627, |
| "step": 273000 |
| }, |
| { |
| "epoch": 19.21455669523676, |
| "grad_norm": 1.0340656042099, |
| "learning_rate": 2.3810900679485327e-05, |
| "loss": 1.4589, |
| "step": 273500 |
| }, |
| { |
| "epoch": 19.249683855557116, |
| "grad_norm": 1.023406982421875, |
| "learning_rate": 2.3666329333526098e-05, |
| "loss": 1.4605, |
| "step": 274000 |
| }, |
| { |
| "epoch": 19.284811015877477, |
| "grad_norm": 1.0535467863082886, |
| "learning_rate": 2.3521757987566865e-05, |
| "loss": 1.4605, |
| "step": 274500 |
| }, |
| { |
| "epoch": 19.319938176197837, |
| "grad_norm": 1.0728813409805298, |
| "learning_rate": 2.3377186641607636e-05, |
| "loss": 1.4626, |
| "step": 275000 |
| }, |
| { |
| "epoch": 19.355065336518194, |
| "grad_norm": 1.0349727869033813, |
| "learning_rate": 2.3232615295648403e-05, |
| "loss": 1.4629, |
| "step": 275500 |
| }, |
| { |
| "epoch": 19.390192496838555, |
| "grad_norm": 1.028206467628479, |
| "learning_rate": 2.3088043949689174e-05, |
| "loss": 1.4586, |
| "step": 276000 |
| }, |
| { |
| "epoch": 19.425319657158916, |
| "grad_norm": 1.040684700012207, |
| "learning_rate": 2.2943472603729945e-05, |
| "loss": 1.4627, |
| "step": 276500 |
| }, |
| { |
| "epoch": 19.460446817479276, |
| "grad_norm": 1.0300320386886597, |
| "learning_rate": 2.2798901257770712e-05, |
| "loss": 1.459, |
| "step": 277000 |
| }, |
| { |
| "epoch": 19.495573977799634, |
| "grad_norm": 1.0533299446105957, |
| "learning_rate": 2.265432991181148e-05, |
| "loss": 1.4612, |
| "step": 277500 |
| }, |
| { |
| "epoch": 19.530701138119994, |
| "grad_norm": 1.041367530822754, |
| "learning_rate": 2.250975856585225e-05, |
| "loss": 1.4587, |
| "step": 278000 |
| }, |
| { |
| "epoch": 19.565828298440355, |
| "grad_norm": 1.078899621963501, |
| "learning_rate": 2.2365187219893018e-05, |
| "loss": 1.4605, |
| "step": 278500 |
| }, |
| { |
| "epoch": 19.600955458760716, |
| "grad_norm": 1.065246343612671, |
| "learning_rate": 2.2220615873933785e-05, |
| "loss": 1.4575, |
| "step": 279000 |
| }, |
| { |
| "epoch": 19.636082619081073, |
| "grad_norm": 1.0640360116958618, |
| "learning_rate": 2.2076044527974556e-05, |
| "loss": 1.46, |
| "step": 279500 |
| }, |
| { |
| "epoch": 19.671209779401433, |
| "grad_norm": 1.0512970685958862, |
| "learning_rate": 2.1931473182015323e-05, |
| "loss": 1.458, |
| "step": 280000 |
| }, |
| { |
| "epoch": 19.671209779401433, |
| "eval_accuracy": 0.72493969774434, |
| "eval_loss": 1.2306617498397827, |
| "eval_runtime": 155.6234, |
| "eval_samples_per_second": 775.629, |
| "eval_steps_per_second": 4.851, |
| "step": 280000 |
| }, |
| { |
| "epoch": 19.706336939721794, |
| "grad_norm": 1.0301251411437988, |
| "learning_rate": 2.1786901836056094e-05, |
| "loss": 1.4597, |
| "step": 280500 |
| }, |
| { |
| "epoch": 19.74146410004215, |
| "grad_norm": 1.0216301679611206, |
| "learning_rate": 2.164233049009686e-05, |
| "loss": 1.4589, |
| "step": 281000 |
| }, |
| { |
| "epoch": 19.776591260362512, |
| "grad_norm": 1.0636018514633179, |
| "learning_rate": 2.1497759144137632e-05, |
| "loss": 1.4591, |
| "step": 281500 |
| }, |
| { |
| "epoch": 19.811718420682872, |
| "grad_norm": 1.052390456199646, |
| "learning_rate": 2.1353187798178403e-05, |
| "loss": 1.4583, |
| "step": 282000 |
| }, |
| { |
| "epoch": 19.846845581003233, |
| "grad_norm": 1.0745834112167358, |
| "learning_rate": 2.120861645221917e-05, |
| "loss": 1.4577, |
| "step": 282500 |
| }, |
| { |
| "epoch": 19.88197274132359, |
| "grad_norm": 1.0797315835952759, |
| "learning_rate": 2.106404510625994e-05, |
| "loss": 1.4567, |
| "step": 283000 |
| }, |
| { |
| "epoch": 19.91709990164395, |
| "grad_norm": 1.072534203529358, |
| "learning_rate": 2.091947376030071e-05, |
| "loss": 1.4595, |
| "step": 283500 |
| }, |
| { |
| "epoch": 19.95222706196431, |
| "grad_norm": 1.0777778625488281, |
| "learning_rate": 2.077490241434148e-05, |
| "loss": 1.4562, |
| "step": 284000 |
| }, |
| { |
| "epoch": 19.987354222284672, |
| "grad_norm": 1.0876667499542236, |
| "learning_rate": 2.0630331068382247e-05, |
| "loss": 1.4556, |
| "step": 284500 |
| }, |
| { |
| "epoch": 20.02248138260503, |
| "grad_norm": 1.0577116012573242, |
| "learning_rate": 2.0485759722423017e-05, |
| "loss": 1.4587, |
| "step": 285000 |
| }, |
| { |
| "epoch": 20.05760854292539, |
| "grad_norm": 1.0711121559143066, |
| "learning_rate": 2.0341188376463785e-05, |
| "loss": 1.4537, |
| "step": 285500 |
| }, |
| { |
| "epoch": 20.09273570324575, |
| "grad_norm": 1.0864572525024414, |
| "learning_rate": 2.0196617030504556e-05, |
| "loss": 1.4534, |
| "step": 286000 |
| }, |
| { |
| "epoch": 20.127862863566108, |
| "grad_norm": 1.068872332572937, |
| "learning_rate": 2.0052045684545323e-05, |
| "loss": 1.4537, |
| "step": 286500 |
| }, |
| { |
| "epoch": 20.16299002388647, |
| "grad_norm": 1.0317089557647705, |
| "learning_rate": 1.9907474338586094e-05, |
| "loss": 1.4536, |
| "step": 287000 |
| }, |
| { |
| "epoch": 20.19811718420683, |
| "grad_norm": 1.030041217803955, |
| "learning_rate": 1.976290299262686e-05, |
| "loss": 1.452, |
| "step": 287500 |
| }, |
| { |
| "epoch": 20.23324434452719, |
| "grad_norm": 1.0281263589859009, |
| "learning_rate": 1.9618331646667632e-05, |
| "loss": 1.4532, |
| "step": 288000 |
| }, |
| { |
| "epoch": 20.268371504847547, |
| "grad_norm": 1.0654585361480713, |
| "learning_rate": 1.9473760300708403e-05, |
| "loss": 1.4506, |
| "step": 288500 |
| }, |
| { |
| "epoch": 20.303498665167908, |
| "grad_norm": 1.0688068866729736, |
| "learning_rate": 1.932918895474917e-05, |
| "loss": 1.4538, |
| "step": 289000 |
| }, |
| { |
| "epoch": 20.33862582548827, |
| "grad_norm": 1.0967620611190796, |
| "learning_rate": 1.918461760878994e-05, |
| "loss": 1.4527, |
| "step": 289500 |
| }, |
| { |
| "epoch": 20.37375298580863, |
| "grad_norm": 1.0728892087936401, |
| "learning_rate": 1.9040046262830708e-05, |
| "loss": 1.4547, |
| "step": 290000 |
| }, |
| { |
| "epoch": 20.37375298580863, |
| "eval_accuracy": 0.7258094694195006, |
| "eval_loss": 1.2251224517822266, |
| "eval_runtime": 150.5035, |
| "eval_samples_per_second": 802.015, |
| "eval_steps_per_second": 5.016, |
| "step": 290000 |
| }, |
| { |
| "epoch": 20.408880146128986, |
| "grad_norm": 1.0635648965835571, |
| "learning_rate": 1.889547491687148e-05, |
| "loss": 1.4523, |
| "step": 290500 |
| }, |
| { |
| "epoch": 20.444007306449347, |
| "grad_norm": 1.0642104148864746, |
| "learning_rate": 1.8750903570912246e-05, |
| "loss": 1.4539, |
| "step": 291000 |
| }, |
| { |
| "epoch": 20.479134466769708, |
| "grad_norm": 1.0912190675735474, |
| "learning_rate": 1.8606332224953017e-05, |
| "loss": 1.4497, |
| "step": 291500 |
| }, |
| { |
| "epoch": 20.514261627090065, |
| "grad_norm": 1.063309907913208, |
| "learning_rate": 1.8461760878993785e-05, |
| "loss": 1.453, |
| "step": 292000 |
| }, |
| { |
| "epoch": 20.549388787410425, |
| "grad_norm": 1.0404052734375, |
| "learning_rate": 1.8317189533034555e-05, |
| "loss": 1.45, |
| "step": 292500 |
| }, |
| { |
| "epoch": 20.584515947730786, |
| "grad_norm": 1.071198582649231, |
| "learning_rate": 1.8172618187075323e-05, |
| "loss": 1.4522, |
| "step": 293000 |
| }, |
| { |
| "epoch": 20.619643108051147, |
| "grad_norm": 1.0776095390319824, |
| "learning_rate": 1.8028046841116093e-05, |
| "loss": 1.4531, |
| "step": 293500 |
| }, |
| { |
| "epoch": 20.654770268371504, |
| "grad_norm": 1.046939730644226, |
| "learning_rate": 1.788347549515686e-05, |
| "loss": 1.4496, |
| "step": 294000 |
| }, |
| { |
| "epoch": 20.689897428691864, |
| "grad_norm": 1.0666289329528809, |
| "learning_rate": 1.7738904149197628e-05, |
| "loss": 1.4519, |
| "step": 294500 |
| }, |
| { |
| "epoch": 20.725024589012225, |
| "grad_norm": 1.0487384796142578, |
| "learning_rate": 1.75943328032384e-05, |
| "loss": 1.4514, |
| "step": 295000 |
| }, |
| { |
| "epoch": 20.760151749332586, |
| "grad_norm": 1.084795355796814, |
| "learning_rate": 1.7449761457279166e-05, |
| "loss": 1.4536, |
| "step": 295500 |
| }, |
| { |
| "epoch": 20.795278909652943, |
| "grad_norm": 1.0882389545440674, |
| "learning_rate": 1.7305190111319937e-05, |
| "loss": 1.4481, |
| "step": 296000 |
| }, |
| { |
| "epoch": 20.830406069973304, |
| "grad_norm": 1.0670382976531982, |
| "learning_rate": 1.7160618765360705e-05, |
| "loss": 1.4496, |
| "step": 296500 |
| }, |
| { |
| "epoch": 20.865533230293664, |
| "grad_norm": 1.066261887550354, |
| "learning_rate": 1.7016047419401475e-05, |
| "loss": 1.4473, |
| "step": 297000 |
| }, |
| { |
| "epoch": 20.90066039061402, |
| "grad_norm": 1.0589436292648315, |
| "learning_rate": 1.6871476073442243e-05, |
| "loss": 1.4512, |
| "step": 297500 |
| }, |
| { |
| "epoch": 20.935787550934382, |
| "grad_norm": 1.0456101894378662, |
| "learning_rate": 1.6726904727483014e-05, |
| "loss": 1.4499, |
| "step": 298000 |
| }, |
| { |
| "epoch": 20.970914711254743, |
| "grad_norm": 1.0359961986541748, |
| "learning_rate": 1.658233338152378e-05, |
| "loss": 1.4509, |
| "step": 298500 |
| }, |
| { |
| "epoch": 21.006041871575103, |
| "grad_norm": 1.0852166414260864, |
| "learning_rate": 1.6437762035564552e-05, |
| "loss": 1.45, |
| "step": 299000 |
| }, |
| { |
| "epoch": 21.04116903189546, |
| "grad_norm": 1.0755125284194946, |
| "learning_rate": 1.629319068960532e-05, |
| "loss": 1.45, |
| "step": 299500 |
| }, |
| { |
| "epoch": 21.07629619221582, |
| "grad_norm": 1.0145972967147827, |
| "learning_rate": 1.614861934364609e-05, |
| "loss": 1.4466, |
| "step": 300000 |
| }, |
| { |
| "epoch": 21.07629619221582, |
| "eval_accuracy": 0.7266361341679182, |
| "eval_loss": 1.2207320928573608, |
| "eval_runtime": 148.3967, |
| "eval_samples_per_second": 813.401, |
| "eval_steps_per_second": 5.088, |
| "step": 300000 |
| }, |
| { |
| "epoch": 21.111423352536182, |
| "grad_norm": 1.0357249975204468, |
| "learning_rate": 1.6004047997686857e-05, |
| "loss": 1.4468, |
| "step": 300500 |
| }, |
| { |
| "epoch": 21.14655051285654, |
| "grad_norm": 1.0667498111724854, |
| "learning_rate": 1.5859476651727628e-05, |
| "loss": 1.4479, |
| "step": 301000 |
| }, |
| { |
| "epoch": 21.1816776731769, |
| "grad_norm": 1.070271611213684, |
| "learning_rate": 1.57149053057684e-05, |
| "loss": 1.447, |
| "step": 301500 |
| }, |
| { |
| "epoch": 21.21680483349726, |
| "grad_norm": 1.0554113388061523, |
| "learning_rate": 1.5570333959809166e-05, |
| "loss": 1.4505, |
| "step": 302000 |
| }, |
| { |
| "epoch": 21.25193199381762, |
| "grad_norm": 1.0505762100219727, |
| "learning_rate": 1.5425762613849937e-05, |
| "loss": 1.4443, |
| "step": 302500 |
| }, |
| { |
| "epoch": 21.287059154137978, |
| "grad_norm": 1.0589311122894287, |
| "learning_rate": 1.5281191267890704e-05, |
| "loss": 1.4463, |
| "step": 303000 |
| }, |
| { |
| "epoch": 21.32218631445834, |
| "grad_norm": 1.0565818548202515, |
| "learning_rate": 1.5136619921931475e-05, |
| "loss": 1.4464, |
| "step": 303500 |
| }, |
| { |
| "epoch": 21.3573134747787, |
| "grad_norm": 1.028573751449585, |
| "learning_rate": 1.4992048575972243e-05, |
| "loss": 1.443, |
| "step": 304000 |
| }, |
| { |
| "epoch": 21.39244063509906, |
| "grad_norm": 1.0877870321273804, |
| "learning_rate": 1.4847477230013013e-05, |
| "loss": 1.4466, |
| "step": 304500 |
| }, |
| { |
| "epoch": 21.427567795419417, |
| "grad_norm": 1.077754259109497, |
| "learning_rate": 1.470290588405378e-05, |
| "loss": 1.4452, |
| "step": 305000 |
| }, |
| { |
| "epoch": 21.462694955739778, |
| "grad_norm": 1.075392246246338, |
| "learning_rate": 1.4558334538094551e-05, |
| "loss": 1.446, |
| "step": 305500 |
| }, |
| { |
| "epoch": 21.49782211606014, |
| "grad_norm": 1.0782703161239624, |
| "learning_rate": 1.4413763192135319e-05, |
| "loss": 1.4445, |
| "step": 306000 |
| }, |
| { |
| "epoch": 21.532949276380496, |
| "grad_norm": 1.1199102401733398, |
| "learning_rate": 1.4269191846176088e-05, |
| "loss": 1.4444, |
| "step": 306500 |
| }, |
| { |
| "epoch": 21.568076436700856, |
| "grad_norm": 1.101174235343933, |
| "learning_rate": 1.4124620500216857e-05, |
| "loss": 1.4434, |
| "step": 307000 |
| }, |
| { |
| "epoch": 21.603203597021217, |
| "grad_norm": 1.0650967359542847, |
| "learning_rate": 1.3980049154257626e-05, |
| "loss": 1.4443, |
| "step": 307500 |
| }, |
| { |
| "epoch": 21.638330757341578, |
| "grad_norm": 1.1260513067245483, |
| "learning_rate": 1.3835477808298397e-05, |
| "loss": 1.449, |
| "step": 308000 |
| }, |
| { |
| "epoch": 21.673457917661935, |
| "grad_norm": 1.0960819721221924, |
| "learning_rate": 1.3690906462339164e-05, |
| "loss": 1.4455, |
| "step": 308500 |
| }, |
| { |
| "epoch": 21.708585077982296, |
| "grad_norm": 1.0622438192367554, |
| "learning_rate": 1.3546335116379935e-05, |
| "loss": 1.4442, |
| "step": 309000 |
| }, |
| { |
| "epoch": 21.743712238302656, |
| "grad_norm": 1.0769811868667603, |
| "learning_rate": 1.3401763770420702e-05, |
| "loss": 1.4428, |
| "step": 309500 |
| }, |
| { |
| "epoch": 21.778839398623017, |
| "grad_norm": 1.0648568868637085, |
| "learning_rate": 1.3257192424461473e-05, |
| "loss": 1.4446, |
| "step": 310000 |
| }, |
| { |
| "epoch": 21.778839398623017, |
| "eval_accuracy": 0.7275222868853471, |
| "eval_loss": 1.2153449058532715, |
| "eval_runtime": 151.7, |
| "eval_samples_per_second": 795.689, |
| "eval_steps_per_second": 4.977, |
| "step": 310000 |
| }, |
| { |
| "epoch": 21.813966558943374, |
| "grad_norm": 1.076948642730713, |
| "learning_rate": 1.311262107850224e-05, |
| "loss": 1.4408, |
| "step": 310500 |
| }, |
| { |
| "epoch": 21.849093719263735, |
| "grad_norm": 1.0741335153579712, |
| "learning_rate": 1.2968049732543011e-05, |
| "loss": 1.4405, |
| "step": 311000 |
| }, |
| { |
| "epoch": 21.884220879584095, |
| "grad_norm": 1.0808049440383911, |
| "learning_rate": 1.2823478386583779e-05, |
| "loss": 1.4452, |
| "step": 311500 |
| }, |
| { |
| "epoch": 21.919348039904452, |
| "grad_norm": 1.0341604948043823, |
| "learning_rate": 1.267890704062455e-05, |
| "loss": 1.4414, |
| "step": 312000 |
| }, |
| { |
| "epoch": 21.954475200224813, |
| "grad_norm": 1.074894905090332, |
| "learning_rate": 1.2534335694665317e-05, |
| "loss": 1.4432, |
| "step": 312500 |
| }, |
| { |
| "epoch": 21.989602360545174, |
| "grad_norm": 1.1056467294692993, |
| "learning_rate": 1.2389764348706088e-05, |
| "loss": 1.4441, |
| "step": 313000 |
| }, |
| { |
| "epoch": 22.024729520865534, |
| "grad_norm": 1.066142201423645, |
| "learning_rate": 1.2245193002746857e-05, |
| "loss": 1.4409, |
| "step": 313500 |
| }, |
| { |
| "epoch": 22.05985668118589, |
| "grad_norm": 1.0810632705688477, |
| "learning_rate": 1.2100621656787626e-05, |
| "loss": 1.4385, |
| "step": 314000 |
| }, |
| { |
| "epoch": 22.094983841506252, |
| "grad_norm": 1.0772794485092163, |
| "learning_rate": 1.1956050310828395e-05, |
| "loss": 1.4425, |
| "step": 314500 |
| }, |
| { |
| "epoch": 22.130111001826613, |
| "grad_norm": 1.0714941024780273, |
| "learning_rate": 1.1811478964869162e-05, |
| "loss": 1.439, |
| "step": 315000 |
| }, |
| { |
| "epoch": 22.165238162146974, |
| "grad_norm": 1.087117075920105, |
| "learning_rate": 1.1666907618909931e-05, |
| "loss": 1.4374, |
| "step": 315500 |
| }, |
| { |
| "epoch": 22.20036532246733, |
| "grad_norm": 1.068716049194336, |
| "learning_rate": 1.15223362729507e-05, |
| "loss": 1.4384, |
| "step": 316000 |
| }, |
| { |
| "epoch": 22.23549248278769, |
| "grad_norm": 1.0669277906417847, |
| "learning_rate": 1.137776492699147e-05, |
| "loss": 1.44, |
| "step": 316500 |
| }, |
| { |
| "epoch": 22.270619643108052, |
| "grad_norm": 1.0556156635284424, |
| "learning_rate": 1.123319358103224e-05, |
| "loss": 1.4411, |
| "step": 317000 |
| }, |
| { |
| "epoch": 22.30574680342841, |
| "grad_norm": 1.081476092338562, |
| "learning_rate": 1.108862223507301e-05, |
| "loss": 1.4396, |
| "step": 317500 |
| }, |
| { |
| "epoch": 22.34087396374877, |
| "grad_norm": 1.0456115007400513, |
| "learning_rate": 1.0944050889113779e-05, |
| "loss": 1.4386, |
| "step": 318000 |
| }, |
| { |
| "epoch": 22.37600112406913, |
| "grad_norm": 1.0953654050827026, |
| "learning_rate": 1.0799479543154548e-05, |
| "loss": 1.4416, |
| "step": 318500 |
| }, |
| { |
| "epoch": 22.41112828438949, |
| "grad_norm": 1.0733470916748047, |
| "learning_rate": 1.0654908197195317e-05, |
| "loss": 1.4378, |
| "step": 319000 |
| }, |
| { |
| "epoch": 22.44625544470985, |
| "grad_norm": 1.107221007347107, |
| "learning_rate": 1.0510336851236086e-05, |
| "loss": 1.4365, |
| "step": 319500 |
| }, |
| { |
| "epoch": 22.48138260503021, |
| "grad_norm": 1.114027500152588, |
| "learning_rate": 1.0365765505276855e-05, |
| "loss": 1.4375, |
| "step": 320000 |
| }, |
| { |
| "epoch": 22.48138260503021, |
| "eval_accuracy": 0.7281385863914055, |
| "eval_loss": 1.2119001150131226, |
| "eval_runtime": 146.4023, |
| "eval_samples_per_second": 824.481, |
| "eval_steps_per_second": 5.157, |
| "step": 320000 |
| }, |
| { |
| "epoch": 22.51650976535057, |
| "grad_norm": 1.0958184003829956, |
| "learning_rate": 1.0221194159317624e-05, |
| "loss": 1.4385, |
| "step": 320500 |
| }, |
| { |
| "epoch": 22.55163692567093, |
| "grad_norm": 1.1146174669265747, |
| "learning_rate": 1.0076622813358393e-05, |
| "loss": 1.4416, |
| "step": 321000 |
| }, |
| { |
| "epoch": 22.586764085991287, |
| "grad_norm": 1.076619029045105, |
| "learning_rate": 9.932051467399162e-06, |
| "loss": 1.4395, |
| "step": 321500 |
| }, |
| { |
| "epoch": 22.621891246311648, |
| "grad_norm": 1.049025297164917, |
| "learning_rate": 9.787480121439931e-06, |
| "loss": 1.4371, |
| "step": 322000 |
| }, |
| { |
| "epoch": 22.65701840663201, |
| "grad_norm": 1.1100108623504639, |
| "learning_rate": 9.6429087754807e-06, |
| "loss": 1.4392, |
| "step": 322500 |
| }, |
| { |
| "epoch": 22.692145566952366, |
| "grad_norm": 1.070065975189209, |
| "learning_rate": 9.49833742952147e-06, |
| "loss": 1.439, |
| "step": 323000 |
| }, |
| { |
| "epoch": 22.727272727272727, |
| "grad_norm": 1.0542230606079102, |
| "learning_rate": 9.353766083562238e-06, |
| "loss": 1.4369, |
| "step": 323500 |
| }, |
| { |
| "epoch": 22.762399887593087, |
| "grad_norm": 1.096826434135437, |
| "learning_rate": 9.209194737603008e-06, |
| "loss": 1.4349, |
| "step": 324000 |
| }, |
| { |
| "epoch": 22.797527047913448, |
| "grad_norm": 1.0940192937850952, |
| "learning_rate": 9.064623391643777e-06, |
| "loss": 1.4387, |
| "step": 324500 |
| }, |
| { |
| "epoch": 22.832654208233805, |
| "grad_norm": 1.095182180404663, |
| "learning_rate": 8.920052045684546e-06, |
| "loss": 1.4376, |
| "step": 325000 |
| }, |
| { |
| "epoch": 22.867781368554166, |
| "grad_norm": 1.1126294136047363, |
| "learning_rate": 8.775480699725315e-06, |
| "loss": 1.4371, |
| "step": 325500 |
| }, |
| { |
| "epoch": 22.902908528874526, |
| "grad_norm": 1.0970271825790405, |
| "learning_rate": 8.630909353766084e-06, |
| "loss": 1.4394, |
| "step": 326000 |
| }, |
| { |
| "epoch": 22.938035689194887, |
| "grad_norm": 1.0789997577667236, |
| "learning_rate": 8.486338007806853e-06, |
| "loss": 1.4365, |
| "step": 326500 |
| }, |
| { |
| "epoch": 22.973162849515244, |
| "grad_norm": 1.0986006259918213, |
| "learning_rate": 8.341766661847622e-06, |
| "loss": 1.4326, |
| "step": 327000 |
| }, |
| { |
| "epoch": 23.008290009835605, |
| "grad_norm": 1.0825368165969849, |
| "learning_rate": 8.197195315888391e-06, |
| "loss": 1.4371, |
| "step": 327500 |
| }, |
| { |
| "epoch": 23.043417170155966, |
| "grad_norm": 1.0966144800186157, |
| "learning_rate": 8.05262396992916e-06, |
| "loss": 1.4343, |
| "step": 328000 |
| }, |
| { |
| "epoch": 23.078544330476323, |
| "grad_norm": 1.1192373037338257, |
| "learning_rate": 7.90805262396993e-06, |
| "loss": 1.4317, |
| "step": 328500 |
| }, |
| { |
| "epoch": 23.113671490796683, |
| "grad_norm": 1.0746561288833618, |
| "learning_rate": 7.763481278010698e-06, |
| "loss": 1.4352, |
| "step": 329000 |
| }, |
| { |
| "epoch": 23.148798651117044, |
| "grad_norm": 1.1043540239334106, |
| "learning_rate": 7.618909932051468e-06, |
| "loss": 1.4363, |
| "step": 329500 |
| }, |
| { |
| "epoch": 23.183925811437405, |
| "grad_norm": 1.0738343000411987, |
| "learning_rate": 7.474338586092237e-06, |
| "loss": 1.4343, |
| "step": 330000 |
| }, |
| { |
| "epoch": 23.183925811437405, |
| "eval_accuracy": 0.7286091536052918, |
| "eval_loss": 1.2086355686187744, |
| "eval_runtime": 149.8495, |
| "eval_samples_per_second": 805.515, |
| "eval_steps_per_second": 5.038, |
| "step": 330000 |
| }, |
| { |
| "epoch": 23.219052971757762, |
| "grad_norm": 1.0852594375610352, |
| "learning_rate": 7.3297672401330065e-06, |
| "loss": 1.4311, |
| "step": 330500 |
| }, |
| { |
| "epoch": 23.254180132078123, |
| "grad_norm": 1.1096608638763428, |
| "learning_rate": 7.1851958941737756e-06, |
| "loss": 1.4347, |
| "step": 331000 |
| }, |
| { |
| "epoch": 23.289307292398483, |
| "grad_norm": 1.0801420211791992, |
| "learning_rate": 7.040624548214545e-06, |
| "loss": 1.4349, |
| "step": 331500 |
| }, |
| { |
| "epoch": 23.324434452718844, |
| "grad_norm": 1.0770845413208008, |
| "learning_rate": 6.896053202255314e-06, |
| "loss": 1.4369, |
| "step": 332000 |
| }, |
| { |
| "epoch": 23.3595616130392, |
| "grad_norm": 1.1009504795074463, |
| "learning_rate": 6.751481856296083e-06, |
| "loss": 1.4344, |
| "step": 332500 |
| }, |
| { |
| "epoch": 23.39468877335956, |
| "grad_norm": 1.1018662452697754, |
| "learning_rate": 6.606910510336851e-06, |
| "loss": 1.431, |
| "step": 333000 |
| }, |
| { |
| "epoch": 23.429815933679922, |
| "grad_norm": 1.081549048423767, |
| "learning_rate": 6.46233916437762e-06, |
| "loss": 1.4329, |
| "step": 333500 |
| }, |
| { |
| "epoch": 23.46494309400028, |
| "grad_norm": 1.0795695781707764, |
| "learning_rate": 6.317767818418389e-06, |
| "loss": 1.4329, |
| "step": 334000 |
| }, |
| { |
| "epoch": 23.50007025432064, |
| "grad_norm": 1.0918314456939697, |
| "learning_rate": 6.173196472459159e-06, |
| "loss": 1.4352, |
| "step": 334500 |
| }, |
| { |
| "epoch": 23.535197414641, |
| "grad_norm": 1.094010829925537, |
| "learning_rate": 6.028625126499928e-06, |
| "loss": 1.4351, |
| "step": 335000 |
| }, |
| { |
| "epoch": 23.57032457496136, |
| "grad_norm": 1.0907319784164429, |
| "learning_rate": 5.884053780540697e-06, |
| "loss": 1.4341, |
| "step": 335500 |
| }, |
| { |
| "epoch": 23.60545173528172, |
| "grad_norm": 1.047654390335083, |
| "learning_rate": 5.739482434581466e-06, |
| "loss": 1.4316, |
| "step": 336000 |
| }, |
| { |
| "epoch": 23.64057889560208, |
| "grad_norm": 1.1064766645431519, |
| "learning_rate": 5.5949110886222355e-06, |
| "loss": 1.429, |
| "step": 336500 |
| }, |
| { |
| "epoch": 23.67570605592244, |
| "grad_norm": 1.0993943214416504, |
| "learning_rate": 5.4503397426630046e-06, |
| "loss": 1.4337, |
| "step": 337000 |
| }, |
| { |
| "epoch": 23.7108332162428, |
| "grad_norm": 1.0862998962402344, |
| "learning_rate": 5.305768396703774e-06, |
| "loss": 1.4273, |
| "step": 337500 |
| }, |
| { |
| "epoch": 23.745960376563158, |
| "grad_norm": 1.0952229499816895, |
| "learning_rate": 5.161197050744543e-06, |
| "loss": 1.4303, |
| "step": 338000 |
| }, |
| { |
| "epoch": 23.78108753688352, |
| "grad_norm": 1.0840052366256714, |
| "learning_rate": 5.016625704785312e-06, |
| "loss": 1.4294, |
| "step": 338500 |
| }, |
| { |
| "epoch": 23.81621469720388, |
| "grad_norm": 1.0815508365631104, |
| "learning_rate": 4.872054358826081e-06, |
| "loss": 1.4319, |
| "step": 339000 |
| }, |
| { |
| "epoch": 23.851341857524236, |
| "grad_norm": 1.1285847425460815, |
| "learning_rate": 4.72748301286685e-06, |
| "loss": 1.4317, |
| "step": 339500 |
| }, |
| { |
| "epoch": 23.886469017844597, |
| "grad_norm": 1.0636698007583618, |
| "learning_rate": 4.582911666907619e-06, |
| "loss": 1.4325, |
| "step": 340000 |
| }, |
| { |
| "epoch": 23.886469017844597, |
| "eval_accuracy": 0.729324170791156, |
| "eval_loss": 1.2056583166122437, |
| "eval_runtime": 149.2803, |
| "eval_samples_per_second": 808.586, |
| "eval_steps_per_second": 5.058, |
| "step": 340000 |
| }, |
| { |
| "epoch": 23.921596178164958, |
| "grad_norm": 1.1081428527832031, |
| "learning_rate": 4.438340320948388e-06, |
| "loss": 1.4302, |
| "step": 340500 |
| }, |
| { |
| "epoch": 23.956723338485318, |
| "grad_norm": 1.0586096048355103, |
| "learning_rate": 4.293768974989157e-06, |
| "loss": 1.4315, |
| "step": 341000 |
| }, |
| { |
| "epoch": 23.991850498805675, |
| "grad_norm": 1.0705420970916748, |
| "learning_rate": 4.149197629029926e-06, |
| "loss": 1.4315, |
| "step": 341500 |
| }, |
| { |
| "epoch": 24.026977659126036, |
| "grad_norm": 1.0813066959381104, |
| "learning_rate": 4.004626283070695e-06, |
| "loss": 1.4328, |
| "step": 342000 |
| }, |
| { |
| "epoch": 24.062104819446397, |
| "grad_norm": 1.1351895332336426, |
| "learning_rate": 3.8600549371114645e-06, |
| "loss": 1.4302, |
| "step": 342500 |
| }, |
| { |
| "epoch": 24.097231979766757, |
| "grad_norm": 1.0965903997421265, |
| "learning_rate": 3.7154835911522336e-06, |
| "loss": 1.4289, |
| "step": 343000 |
| }, |
| { |
| "epoch": 24.132359140087114, |
| "grad_norm": 1.0875136852264404, |
| "learning_rate": 3.5709122451930026e-06, |
| "loss": 1.4314, |
| "step": 343500 |
| }, |
| { |
| "epoch": 24.167486300407475, |
| "grad_norm": 1.0911264419555664, |
| "learning_rate": 3.426340899233772e-06, |
| "loss": 1.4329, |
| "step": 344000 |
| }, |
| { |
| "epoch": 24.202613460727836, |
| "grad_norm": 1.0920251607894897, |
| "learning_rate": 3.2817695532745412e-06, |
| "loss": 1.4304, |
| "step": 344500 |
| }, |
| { |
| "epoch": 24.237740621048193, |
| "grad_norm": 1.1164170503616333, |
| "learning_rate": 3.1371982073153103e-06, |
| "loss": 1.4325, |
| "step": 345000 |
| }, |
| { |
| "epoch": 24.272867781368554, |
| "grad_norm": 1.1081171035766602, |
| "learning_rate": 2.9926268613560794e-06, |
| "loss": 1.4302, |
| "step": 345500 |
| }, |
| { |
| "epoch": 24.307994941688914, |
| "grad_norm": 1.1110938787460327, |
| "learning_rate": 2.8480555153968485e-06, |
| "loss": 1.4289, |
| "step": 346000 |
| }, |
| { |
| "epoch": 24.343122102009275, |
| "grad_norm": 1.0824356079101562, |
| "learning_rate": 2.7034841694376176e-06, |
| "loss": 1.4267, |
| "step": 346500 |
| }, |
| { |
| "epoch": 24.378249262329632, |
| "grad_norm": 1.09736967086792, |
| "learning_rate": 2.5589128234783866e-06, |
| "loss": 1.4296, |
| "step": 347000 |
| }, |
| { |
| "epoch": 24.413376422649993, |
| "grad_norm": 1.1021647453308105, |
| "learning_rate": 2.414341477519156e-06, |
| "loss": 1.4288, |
| "step": 347500 |
| }, |
| { |
| "epoch": 24.448503582970353, |
| "grad_norm": 1.0728425979614258, |
| "learning_rate": 2.269770131559925e-06, |
| "loss": 1.4303, |
| "step": 348000 |
| }, |
| { |
| "epoch": 24.483630743290714, |
| "grad_norm": 1.0710495710372925, |
| "learning_rate": 2.125198785600694e-06, |
| "loss": 1.4276, |
| "step": 348500 |
| }, |
| { |
| "epoch": 24.51875790361107, |
| "grad_norm": 1.0602104663848877, |
| "learning_rate": 1.980627439641463e-06, |
| "loss": 1.4259, |
| "step": 349000 |
| }, |
| { |
| "epoch": 24.553885063931432, |
| "grad_norm": 1.0960617065429688, |
| "learning_rate": 1.8360560936822323e-06, |
| "loss": 1.4269, |
| "step": 349500 |
| }, |
| { |
| "epoch": 24.589012224251793, |
| "grad_norm": 1.0834195613861084, |
| "learning_rate": 1.6914847477230013e-06, |
| "loss": 1.4294, |
| "step": 350000 |
| }, |
| { |
| "epoch": 24.589012224251793, |
| "eval_accuracy": 0.7296860337374499, |
| "eval_loss": 1.202436089515686, |
| "eval_runtime": 150.0537, |
| "eval_samples_per_second": 804.419, |
| "eval_steps_per_second": 5.032, |
| "step": 350000 |
| }, |
| { |
| "epoch": 24.62413938457215, |
| "grad_norm": 1.0872336626052856, |
| "learning_rate": 1.5469134017637704e-06, |
| "loss": 1.4277, |
| "step": 350500 |
| }, |
| { |
| "epoch": 24.65926654489251, |
| "grad_norm": 1.104445457458496, |
| "learning_rate": 1.4023420558045395e-06, |
| "loss": 1.4268, |
| "step": 351000 |
| }, |
| { |
| "epoch": 24.69439370521287, |
| "grad_norm": 1.076301097869873, |
| "learning_rate": 1.2577707098453088e-06, |
| "loss": 1.427, |
| "step": 351500 |
| }, |
| { |
| "epoch": 24.72952086553323, |
| "grad_norm": 1.1510794162750244, |
| "learning_rate": 1.1131993638860779e-06, |
| "loss": 1.429, |
| "step": 352000 |
| }, |
| { |
| "epoch": 24.76464802585359, |
| "grad_norm": 1.12981116771698, |
| "learning_rate": 9.68628017926847e-07, |
| "loss": 1.428, |
| "step": 352500 |
| }, |
| { |
| "epoch": 24.79977518617395, |
| "grad_norm": 1.0890146493911743, |
| "learning_rate": 8.240566719676161e-07, |
| "loss": 1.4283, |
| "step": 353000 |
| }, |
| { |
| "epoch": 24.83490234649431, |
| "grad_norm": 1.084420919418335, |
| "learning_rate": 6.794853260083851e-07, |
| "loss": 1.4274, |
| "step": 353500 |
| }, |
| { |
| "epoch": 24.870029506814667, |
| "grad_norm": 1.090317964553833, |
| "learning_rate": 5.349139800491543e-07, |
| "loss": 1.4283, |
| "step": 354000 |
| }, |
| { |
| "epoch": 24.905156667135028, |
| "grad_norm": 1.108689546585083, |
| "learning_rate": 3.903426340899234e-07, |
| "loss": 1.4277, |
| "step": 354500 |
| }, |
| { |
| "epoch": 24.94028382745539, |
| "grad_norm": 1.1043752431869507, |
| "learning_rate": 2.457712881306925e-07, |
| "loss": 1.4245, |
| "step": 355000 |
| }, |
| { |
| "epoch": 24.97541098777575, |
| "grad_norm": 1.0509768724441528, |
| "learning_rate": 1.0119994217146162e-07, |
| "loss": 1.4286, |
| "step": 355500 |
| }, |
| { |
| "epoch": 25.0, |
| "step": 355850, |
| "total_flos": 7.547177208649421e+18, |
| "train_loss": 1.8504363005138171, |
| "train_runtime": 132228.6011, |
| "train_samples_per_second": 430.569, |
| "train_steps_per_second": 2.691 |
| } |
| ], |
| "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": 7.547177208649421e+18, |
| "train_batch_size": 160, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
