| { |
| "best_metric": 2.991220235824585, |
| "best_model_checkpoint": "bert_12_layer_model_v1_complete_training_new_48/checkpoint-180000", |
| "epoch": 1.5441593340870734, |
| "global_step": 188477, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.0, |
| "learning_rate": 5.000000000000001e-07, |
| "loss": 9.9299, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 1.0000000000000002e-06, |
| "loss": 9.1977, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.01, |
| "learning_rate": 1.5e-06, |
| "loss": 8.8067, |
| "step": 1500 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 2.0000000000000003e-06, |
| "loss": 8.3304, |
| "step": 2000 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 2.5e-06, |
| "loss": 7.828, |
| "step": 2500 |
| }, |
| { |
| "epoch": 0.02, |
| "learning_rate": 3e-06, |
| "loss": 7.4293, |
| "step": 3000 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 3.5e-06, |
| "loss": 7.1807, |
| "step": 3500 |
| }, |
| { |
| "epoch": 0.03, |
| "learning_rate": 4.000000000000001e-06, |
| "loss": 7.053, |
| "step": 4000 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 4.5e-06, |
| "loss": 6.9587, |
| "step": 4500 |
| }, |
| { |
| "epoch": 0.04, |
| "learning_rate": 5e-06, |
| "loss": 6.8993, |
| "step": 5000 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 5.500000000000001e-06, |
| "loss": 6.8415, |
| "step": 5500 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 6e-06, |
| "loss": 6.7959, |
| "step": 6000 |
| }, |
| { |
| "epoch": 0.05, |
| "learning_rate": 6.5000000000000004e-06, |
| "loss": 6.7647, |
| "step": 6500 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 7e-06, |
| "loss": 6.7286, |
| "step": 7000 |
| }, |
| { |
| "epoch": 0.06, |
| "learning_rate": 7.500000000000001e-06, |
| "loss": 6.7002, |
| "step": 7500 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 8.000000000000001e-06, |
| "loss": 6.6642, |
| "step": 8000 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 8.5e-06, |
| "loss": 6.636, |
| "step": 8500 |
| }, |
| { |
| "epoch": 0.07, |
| "learning_rate": 9e-06, |
| "loss": 6.616, |
| "step": 9000 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 9.5e-06, |
| "loss": 6.5912, |
| "step": 9500 |
| }, |
| { |
| "epoch": 0.08, |
| "learning_rate": 1e-05, |
| "loss": 6.5749, |
| "step": 10000 |
| }, |
| { |
| "epoch": 0.08, |
| "eval_accuracy": 0.1270349366338523, |
| "eval_loss": 6.538110256195068, |
| "eval_runtime": 1887.8625, |
| "eval_samples_per_second": 163.336, |
| "eval_steps_per_second": 3.403, |
| "step": 10000 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 9.991670692498626e-06, |
| "loss": 6.5529, |
| "step": 10500 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 9.983341384997253e-06, |
| "loss": 6.5409, |
| "step": 11000 |
| }, |
| { |
| "epoch": 0.09, |
| "learning_rate": 9.975012077495878e-06, |
| "loss": 6.5186, |
| "step": 11500 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 9.966682769994504e-06, |
| "loss": 6.504, |
| "step": 12000 |
| }, |
| { |
| "epoch": 0.1, |
| "learning_rate": 9.95835346249313e-06, |
| "loss": 6.4798, |
| "step": 12500 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 9.950024154991756e-06, |
| "loss": 6.4697, |
| "step": 13000 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 9.941694847490381e-06, |
| "loss": 6.4584, |
| "step": 13500 |
| }, |
| { |
| "epoch": 0.11, |
| "learning_rate": 9.933365539989006e-06, |
| "loss": 6.4478, |
| "step": 14000 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 9.925036232487631e-06, |
| "loss": 6.4286, |
| "step": 14500 |
| }, |
| { |
| "epoch": 0.12, |
| "learning_rate": 9.916706924986258e-06, |
| "loss": 6.4178, |
| "step": 15000 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 9.908377617484883e-06, |
| "loss": 6.4003, |
| "step": 15500 |
| }, |
| { |
| "epoch": 0.13, |
| "learning_rate": 9.90004830998351e-06, |
| "loss": 6.3985, |
| "step": 16000 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 9.891719002482134e-06, |
| "loss": 6.3955, |
| "step": 16500 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 9.883389694980761e-06, |
| "loss": 6.3752, |
| "step": 17000 |
| }, |
| { |
| "epoch": 0.14, |
| "learning_rate": 9.875060387479386e-06, |
| "loss": 6.3744, |
| "step": 17500 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 9.866731079978011e-06, |
| "loss": 6.3656, |
| "step": 18000 |
| }, |
| { |
| "epoch": 0.15, |
| "learning_rate": 9.858401772476636e-06, |
| "loss": 6.3628, |
| "step": 18500 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 9.850072464975263e-06, |
| "loss": 6.3458, |
| "step": 19000 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 9.84174315747389e-06, |
| "loss": 6.3384, |
| "step": 19500 |
| }, |
| { |
| "epoch": 0.16, |
| "learning_rate": 9.833413849972514e-06, |
| "loss": 6.333, |
| "step": 20000 |
| }, |
| { |
| "epoch": 0.16, |
| "eval_accuracy": 0.14102632562372713, |
| "eval_loss": 6.309741020202637, |
| "eval_runtime": 1889.6396, |
| "eval_samples_per_second": 163.182, |
| "eval_steps_per_second": 3.4, |
| "step": 20000 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 9.82508454247114e-06, |
| "loss": 6.3248, |
| "step": 20500 |
| }, |
| { |
| "epoch": 0.17, |
| "learning_rate": 9.816755234969766e-06, |
| "loss": 6.3151, |
| "step": 21000 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 9.808425927468391e-06, |
| "loss": 6.3198, |
| "step": 21500 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 9.800096619967016e-06, |
| "loss": 6.3031, |
| "step": 22000 |
| }, |
| { |
| "epoch": 0.18, |
| "learning_rate": 9.791767312465643e-06, |
| "loss": 6.3055, |
| "step": 22500 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 9.783438004964268e-06, |
| "loss": 6.2871, |
| "step": 23000 |
| }, |
| { |
| "epoch": 0.19, |
| "learning_rate": 9.775108697462895e-06, |
| "loss": 6.2835, |
| "step": 23500 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 9.76677938996152e-06, |
| "loss": 6.2762, |
| "step": 24000 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 9.758450082460145e-06, |
| "loss": 6.279, |
| "step": 24500 |
| }, |
| { |
| "epoch": 0.2, |
| "learning_rate": 9.750120774958771e-06, |
| "loss": 6.2732, |
| "step": 25000 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 9.741791467457396e-06, |
| "loss": 6.264, |
| "step": 25500 |
| }, |
| { |
| "epoch": 0.21, |
| "learning_rate": 9.733462159956021e-06, |
| "loss": 6.2618, |
| "step": 26000 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 9.725132852454648e-06, |
| "loss": 6.2658, |
| "step": 26500 |
| }, |
| { |
| "epoch": 0.22, |
| "learning_rate": 9.716803544953275e-06, |
| "loss": 6.2597, |
| "step": 27000 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 9.7084742374519e-06, |
| "loss": 6.2504, |
| "step": 27500 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 9.700144929950525e-06, |
| "loss": 6.2482, |
| "step": 28000 |
| }, |
| { |
| "epoch": 0.23, |
| "learning_rate": 9.69181562244915e-06, |
| "loss": 6.247, |
| "step": 28500 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 9.683486314947776e-06, |
| "loss": 6.2413, |
| "step": 29000 |
| }, |
| { |
| "epoch": 0.24, |
| "learning_rate": 9.675157007446401e-06, |
| "loss": 6.2348, |
| "step": 29500 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 9.666827699945028e-06, |
| "loss": 6.2341, |
| "step": 30000 |
| }, |
| { |
| "epoch": 0.25, |
| "eval_accuracy": 0.14499767659211227, |
| "eval_loss": 6.217945098876953, |
| "eval_runtime": 1890.2968, |
| "eval_samples_per_second": 163.126, |
| "eval_steps_per_second": 3.399, |
| "step": 30000 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 9.658498392443653e-06, |
| "loss": 6.2268, |
| "step": 30500 |
| }, |
| { |
| "epoch": 0.25, |
| "learning_rate": 9.65016908494228e-06, |
| "loss": 6.2268, |
| "step": 31000 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 9.641839777440905e-06, |
| "loss": 6.2275, |
| "step": 31500 |
| }, |
| { |
| "epoch": 0.26, |
| "learning_rate": 9.63351046993953e-06, |
| "loss": 6.2192, |
| "step": 32000 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 9.625181162438155e-06, |
| "loss": 6.2144, |
| "step": 32500 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 9.616851854936781e-06, |
| "loss": 6.2118, |
| "step": 33000 |
| }, |
| { |
| "epoch": 0.27, |
| "learning_rate": 9.608522547435406e-06, |
| "loss": 6.2062, |
| "step": 33500 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 9.600193239934033e-06, |
| "loss": 6.206, |
| "step": 34000 |
| }, |
| { |
| "epoch": 0.28, |
| "learning_rate": 9.591863932432658e-06, |
| "loss": 6.199, |
| "step": 34500 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 9.583534624931285e-06, |
| "loss": 6.2003, |
| "step": 35000 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 9.57520531742991e-06, |
| "loss": 6.2049, |
| "step": 35500 |
| }, |
| { |
| "epoch": 0.29, |
| "learning_rate": 9.566876009928535e-06, |
| "loss": 6.1984, |
| "step": 36000 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 9.55854670242716e-06, |
| "loss": 6.1945, |
| "step": 36500 |
| }, |
| { |
| "epoch": 0.3, |
| "learning_rate": 9.550217394925787e-06, |
| "loss": 6.1913, |
| "step": 37000 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 9.541888087424413e-06, |
| "loss": 6.1878, |
| "step": 37500 |
| }, |
| { |
| "epoch": 0.31, |
| "learning_rate": 9.533558779923038e-06, |
| "loss": 6.1778, |
| "step": 38000 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 9.525229472421663e-06, |
| "loss": 6.1807, |
| "step": 38500 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 9.51690016492029e-06, |
| "loss": 6.1739, |
| "step": 39000 |
| }, |
| { |
| "epoch": 0.32, |
| "learning_rate": 9.508570857418915e-06, |
| "loss": 6.1769, |
| "step": 39500 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 9.50024154991754e-06, |
| "loss": 6.1803, |
| "step": 40000 |
| }, |
| { |
| "epoch": 0.33, |
| "eval_accuracy": 0.14784209481060512, |
| "eval_loss": 6.158620357513428, |
| "eval_runtime": 1891.6028, |
| "eval_samples_per_second": 163.013, |
| "eval_steps_per_second": 3.397, |
| "step": 40000 |
| }, |
| { |
| "epoch": 0.33, |
| "learning_rate": 9.491912242416167e-06, |
| "loss": 6.1717, |
| "step": 40500 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 9.483582934914792e-06, |
| "loss": 6.1696, |
| "step": 41000 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 9.475253627413418e-06, |
| "loss": 6.1667, |
| "step": 41500 |
| }, |
| { |
| "epoch": 0.34, |
| "learning_rate": 9.466924319912043e-06, |
| "loss": 6.1708, |
| "step": 42000 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 9.458595012410668e-06, |
| "loss": 6.1576, |
| "step": 42500 |
| }, |
| { |
| "epoch": 0.35, |
| "learning_rate": 9.450265704909295e-06, |
| "loss": 6.1572, |
| "step": 43000 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 9.44193639740792e-06, |
| "loss": 6.1538, |
| "step": 43500 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 9.433607089906545e-06, |
| "loss": 6.1569, |
| "step": 44000 |
| }, |
| { |
| "epoch": 0.36, |
| "learning_rate": 9.425277782405172e-06, |
| "loss": 6.1522, |
| "step": 44500 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 9.416948474903798e-06, |
| "loss": 6.1502, |
| "step": 45000 |
| }, |
| { |
| "epoch": 0.37, |
| "learning_rate": 9.408619167402423e-06, |
| "loss": 6.1513, |
| "step": 45500 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 9.400289859901048e-06, |
| "loss": 6.1375, |
| "step": 46000 |
| }, |
| { |
| "epoch": 0.38, |
| "learning_rate": 9.391960552399673e-06, |
| "loss": 6.1341, |
| "step": 46500 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 9.3836312448983e-06, |
| "loss": 6.1253, |
| "step": 47000 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 9.375301937396925e-06, |
| "loss": 6.1237, |
| "step": 47500 |
| }, |
| { |
| "epoch": 0.39, |
| "learning_rate": 9.366972629895552e-06, |
| "loss": 6.1197, |
| "step": 48000 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 9.358643322394177e-06, |
| "loss": 6.1111, |
| "step": 48500 |
| }, |
| { |
| "epoch": 0.4, |
| "learning_rate": 9.350314014892804e-06, |
| "loss": 6.1033, |
| "step": 49000 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 9.341984707391429e-06, |
| "loss": 6.0921, |
| "step": 49500 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 9.333655399890054e-06, |
| "loss": 6.0775, |
| "step": 50000 |
| }, |
| { |
| "epoch": 0.41, |
| "eval_accuracy": 0.15201612962376762, |
| "eval_loss": 6.047074794769287, |
| "eval_runtime": 1892.7085, |
| "eval_samples_per_second": 162.918, |
| "eval_steps_per_second": 3.395, |
| "step": 50000 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 9.325326092388679e-06, |
| "loss": 6.0559, |
| "step": 50500 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 9.316996784887305e-06, |
| "loss": 6.0454, |
| "step": 51000 |
| }, |
| { |
| "epoch": 0.42, |
| "learning_rate": 9.30866747738593e-06, |
| "loss": 6.0361, |
| "step": 51500 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 9.300338169884557e-06, |
| "loss": 6.0208, |
| "step": 52000 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 9.292008862383182e-06, |
| "loss": 6.004, |
| "step": 52500 |
| }, |
| { |
| "epoch": 0.43, |
| "learning_rate": 9.283679554881809e-06, |
| "loss": 5.9976, |
| "step": 53000 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 9.275350247380434e-06, |
| "loss": 5.9822, |
| "step": 53500 |
| }, |
| { |
| "epoch": 0.44, |
| "learning_rate": 9.267020939879059e-06, |
| "loss": 5.9751, |
| "step": 54000 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 9.258691632377685e-06, |
| "loss": 5.9676, |
| "step": 54500 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 9.25036232487631e-06, |
| "loss": 5.954, |
| "step": 55000 |
| }, |
| { |
| "epoch": 0.45, |
| "learning_rate": 9.242033017374937e-06, |
| "loss": 5.955, |
| "step": 55500 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 9.233703709873562e-06, |
| "loss": 5.9421, |
| "step": 56000 |
| }, |
| { |
| "epoch": 0.46, |
| "learning_rate": 9.225374402372189e-06, |
| "loss": 5.9321, |
| "step": 56500 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 9.217045094870814e-06, |
| "loss": 5.9178, |
| "step": 57000 |
| }, |
| { |
| "epoch": 0.47, |
| "learning_rate": 9.208715787369439e-06, |
| "loss": 5.9229, |
| "step": 57500 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 9.200386479868064e-06, |
| "loss": 5.9179, |
| "step": 58000 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 9.19205717236669e-06, |
| "loss": 5.9098, |
| "step": 58500 |
| }, |
| { |
| "epoch": 0.48, |
| "learning_rate": 9.183727864865315e-06, |
| "loss": 5.9018, |
| "step": 59000 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 9.175398557363942e-06, |
| "loss": 5.9023, |
| "step": 59500 |
| }, |
| { |
| "epoch": 0.49, |
| "learning_rate": 9.167069249862567e-06, |
| "loss": 5.8957, |
| "step": 60000 |
| }, |
| { |
| "epoch": 0.49, |
| "eval_accuracy": 0.165545505269546, |
| "eval_loss": 5.845799446105957, |
| "eval_runtime": 1892.1568, |
| "eval_samples_per_second": 162.965, |
| "eval_steps_per_second": 3.396, |
| "step": 60000 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 9.158739942361194e-06, |
| "loss": 5.882, |
| "step": 60500 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 9.150410634859819e-06, |
| "loss": 5.8724, |
| "step": 61000 |
| }, |
| { |
| "epoch": 0.5, |
| "learning_rate": 9.142081327358444e-06, |
| "loss": 5.8718, |
| "step": 61500 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 9.133752019857069e-06, |
| "loss": 5.8614, |
| "step": 62000 |
| }, |
| { |
| "epoch": 0.51, |
| "learning_rate": 9.125422712355696e-06, |
| "loss": 5.8606, |
| "step": 62500 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 9.11709340485432e-06, |
| "loss": 5.8554, |
| "step": 63000 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 9.108764097352947e-06, |
| "loss": 5.8398, |
| "step": 63500 |
| }, |
| { |
| "epoch": 0.52, |
| "learning_rate": 9.100434789851572e-06, |
| "loss": 5.8349, |
| "step": 64000 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 9.092105482350199e-06, |
| "loss": 5.8345, |
| "step": 64500 |
| }, |
| { |
| "epoch": 0.53, |
| "learning_rate": 9.083776174848824e-06, |
| "loss": 5.8301, |
| "step": 65000 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 9.075446867347449e-06, |
| "loss": 5.8252, |
| "step": 65500 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 9.067117559846076e-06, |
| "loss": 5.8213, |
| "step": 66000 |
| }, |
| { |
| "epoch": 0.54, |
| "learning_rate": 9.0587882523447e-06, |
| "loss": 5.8079, |
| "step": 66500 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 9.050458944843327e-06, |
| "loss": 5.8021, |
| "step": 67000 |
| }, |
| { |
| "epoch": 0.55, |
| "learning_rate": 9.042129637341952e-06, |
| "loss": 5.7962, |
| "step": 67500 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 9.033800329840577e-06, |
| "loss": 5.7945, |
| "step": 68000 |
| }, |
| { |
| "epoch": 0.56, |
| "learning_rate": 9.025471022339204e-06, |
| "loss": 5.7872, |
| "step": 68500 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 9.017141714837829e-06, |
| "loss": 5.7794, |
| "step": 69000 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 9.008812407336454e-06, |
| "loss": 5.7742, |
| "step": 69500 |
| }, |
| { |
| "epoch": 0.57, |
| "learning_rate": 9.00048309983508e-06, |
| "loss": 5.7655, |
| "step": 70000 |
| }, |
| { |
| "epoch": 0.57, |
| "eval_accuracy": 0.18459943409109789, |
| "eval_loss": 5.704047203063965, |
| "eval_runtime": 1889.6221, |
| "eval_samples_per_second": 163.184, |
| "eval_steps_per_second": 3.4, |
| "step": 70000 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 8.992153792333706e-06, |
| "loss": 5.7536, |
| "step": 70500 |
| }, |
| { |
| "epoch": 0.58, |
| "learning_rate": 8.983824484832332e-06, |
| "loss": 5.7448, |
| "step": 71000 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 8.975495177330957e-06, |
| "loss": 5.7462, |
| "step": 71500 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 8.967165869829582e-06, |
| "loss": 5.7384, |
| "step": 72000 |
| }, |
| { |
| "epoch": 0.59, |
| "learning_rate": 8.958836562328209e-06, |
| "loss": 5.7339, |
| "step": 72500 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 8.950507254826834e-06, |
| "loss": 5.722, |
| "step": 73000 |
| }, |
| { |
| "epoch": 0.6, |
| "learning_rate": 8.942177947325459e-06, |
| "loss": 5.7217, |
| "step": 73500 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 8.933848639824086e-06, |
| "loss": 5.7092, |
| "step": 74000 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 8.925519332322713e-06, |
| "loss": 5.7035, |
| "step": 74500 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 8.917190024821338e-06, |
| "loss": 5.7032, |
| "step": 75000 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 8.908860717319963e-06, |
| "loss": 5.6864, |
| "step": 75500 |
| }, |
| { |
| "epoch": 0.62, |
| "learning_rate": 8.900531409818588e-06, |
| "loss": 5.6855, |
| "step": 76000 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 8.892202102317214e-06, |
| "loss": 5.6752, |
| "step": 76500 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 8.88387279481584e-06, |
| "loss": 5.6704, |
| "step": 77000 |
| }, |
| { |
| "epoch": 0.63, |
| "learning_rate": 8.875543487314466e-06, |
| "loss": 5.6612, |
| "step": 77500 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 8.867214179813091e-06, |
| "loss": 5.6556, |
| "step": 78000 |
| }, |
| { |
| "epoch": 0.64, |
| "learning_rate": 8.858884872311718e-06, |
| "loss": 5.652, |
| "step": 78500 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 8.850555564810343e-06, |
| "loss": 5.6424, |
| "step": 79000 |
| }, |
| { |
| "epoch": 0.65, |
| "learning_rate": 8.842226257308968e-06, |
| "loss": 5.6391, |
| "step": 79500 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 8.833896949807593e-06, |
| "loss": 5.6281, |
| "step": 80000 |
| }, |
| { |
| "epoch": 0.66, |
| "eval_accuracy": 0.20261157607549118, |
| "eval_loss": 5.548015117645264, |
| "eval_runtime": 1889.2251, |
| "eval_samples_per_second": 163.218, |
| "eval_steps_per_second": 3.401, |
| "step": 80000 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 8.82556764230622e-06, |
| "loss": 5.6183, |
| "step": 80500 |
| }, |
| { |
| "epoch": 0.66, |
| "learning_rate": 8.817238334804844e-06, |
| "loss": 5.6107, |
| "step": 81000 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 8.808909027303471e-06, |
| "loss": 5.5977, |
| "step": 81500 |
| }, |
| { |
| "epoch": 0.67, |
| "learning_rate": 8.800579719802096e-06, |
| "loss": 5.5938, |
| "step": 82000 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 8.792250412300723e-06, |
| "loss": 5.5727, |
| "step": 82500 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 8.783921104799348e-06, |
| "loss": 5.5687, |
| "step": 83000 |
| }, |
| { |
| "epoch": 0.68, |
| "learning_rate": 8.775591797297973e-06, |
| "loss": 5.5463, |
| "step": 83500 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 8.767262489796598e-06, |
| "loss": 5.5223, |
| "step": 84000 |
| }, |
| { |
| "epoch": 0.69, |
| "learning_rate": 8.758933182295224e-06, |
| "loss": 5.4914, |
| "step": 84500 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 8.750603874793851e-06, |
| "loss": 5.4664, |
| "step": 85000 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 8.742274567292476e-06, |
| "loss": 5.4408, |
| "step": 85500 |
| }, |
| { |
| "epoch": 0.7, |
| "learning_rate": 8.733945259791101e-06, |
| "loss": 5.4078, |
| "step": 86000 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 8.725615952289728e-06, |
| "loss": 5.3864, |
| "step": 86500 |
| }, |
| { |
| "epoch": 0.71, |
| "learning_rate": 8.717286644788353e-06, |
| "loss": 5.3537, |
| "step": 87000 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 8.708957337286978e-06, |
| "loss": 5.3277, |
| "step": 87500 |
| }, |
| { |
| "epoch": 0.72, |
| "learning_rate": 8.700628029785605e-06, |
| "loss": 5.2929, |
| "step": 88000 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 8.69229872228423e-06, |
| "loss": 5.2634, |
| "step": 88500 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 8.683969414782856e-06, |
| "loss": 5.2354, |
| "step": 89000 |
| }, |
| { |
| "epoch": 0.73, |
| "learning_rate": 8.675640107281481e-06, |
| "loss": 5.2045, |
| "step": 89500 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 8.667310799780106e-06, |
| "loss": 5.1797, |
| "step": 90000 |
| }, |
| { |
| "epoch": 0.74, |
| "eval_accuracy": 0.26609113272078677, |
| "eval_loss": 5.00044584274292, |
| "eval_runtime": 1887.6694, |
| "eval_samples_per_second": 163.353, |
| "eval_steps_per_second": 3.404, |
| "step": 90000 |
| }, |
| { |
| "epoch": 0.74, |
| "learning_rate": 8.658981492278733e-06, |
| "loss": 5.1458, |
| "step": 90500 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 8.650652184777358e-06, |
| "loss": 5.123, |
| "step": 91000 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 8.642322877275983e-06, |
| "loss": 5.0956, |
| "step": 91500 |
| }, |
| { |
| "epoch": 0.75, |
| "learning_rate": 8.63399356977461e-06, |
| "loss": 5.0819, |
| "step": 92000 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 8.625664262273236e-06, |
| "loss": 5.0458, |
| "step": 92500 |
| }, |
| { |
| "epoch": 0.76, |
| "learning_rate": 8.617334954771861e-06, |
| "loss": 5.0314, |
| "step": 93000 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 8.609005647270486e-06, |
| "loss": 4.997, |
| "step": 93500 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 8.600676339769111e-06, |
| "loss": 4.9804, |
| "step": 94000 |
| }, |
| { |
| "epoch": 0.77, |
| "learning_rate": 8.592347032267738e-06, |
| "loss": 4.9661, |
| "step": 94500 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 8.584017724766363e-06, |
| "loss": 4.9355, |
| "step": 95000 |
| }, |
| { |
| "epoch": 0.78, |
| "learning_rate": 8.57568841726499e-06, |
| "loss": 4.9253, |
| "step": 95500 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 8.567359109763615e-06, |
| "loss": 4.9062, |
| "step": 96000 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 8.559029802262241e-06, |
| "loss": 4.8857, |
| "step": 96500 |
| }, |
| { |
| "epoch": 0.79, |
| "learning_rate": 8.550700494760866e-06, |
| "loss": 4.8644, |
| "step": 97000 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 8.542371187259491e-06, |
| "loss": 4.8439, |
| "step": 97500 |
| }, |
| { |
| "epoch": 0.8, |
| "learning_rate": 8.534041879758118e-06, |
| "loss": 4.8313, |
| "step": 98000 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 8.525712572256743e-06, |
| "loss": 4.815, |
| "step": 98500 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 8.517383264755368e-06, |
| "loss": 4.7952, |
| "step": 99000 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 8.509053957253995e-06, |
| "loss": 4.7711, |
| "step": 99500 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 8.500724649752622e-06, |
| "loss": 4.7518, |
| "step": 100000 |
| }, |
| { |
| "epoch": 0.82, |
| "eval_accuracy": 0.3096753842739182, |
| "eval_loss": 4.575112342834473, |
| "eval_runtime": 1890.8301, |
| "eval_samples_per_second": 163.08, |
| "eval_steps_per_second": 3.398, |
| "step": 100000 |
| }, |
| { |
| "epoch": 0.82, |
| "learning_rate": 8.492395342251247e-06, |
| "loss": 4.7381, |
| "step": 100500 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 8.484066034749872e-06, |
| "loss": 4.7172, |
| "step": 101000 |
| }, |
| { |
| "epoch": 0.83, |
| "learning_rate": 8.475736727248497e-06, |
| "loss": 4.714, |
| "step": 101500 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 8.467407419747123e-06, |
| "loss": 4.6895, |
| "step": 102000 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 8.459078112245748e-06, |
| "loss": 4.6775, |
| "step": 102500 |
| }, |
| { |
| "epoch": 0.84, |
| "learning_rate": 8.450748804744375e-06, |
| "loss": 4.663, |
| "step": 103000 |
| }, |
| { |
| "epoch": 0.85, |
| "learning_rate": 8.442419497243e-06, |
| "loss": 4.6431, |
| "step": 103500 |
| }, |
| { |
| "epoch": 0.85, |
| "learning_rate": 8.434090189741627e-06, |
| "loss": 4.6214, |
| "step": 104000 |
| }, |
| { |
| "epoch": 0.86, |
| "learning_rate": 8.425760882240252e-06, |
| "loss": 4.6219, |
| "step": 104500 |
| }, |
| { |
| "epoch": 0.86, |
| "learning_rate": 8.417431574738877e-06, |
| "loss": 4.5977, |
| "step": 105000 |
| }, |
| { |
| "epoch": 0.86, |
| "learning_rate": 8.409102267237502e-06, |
| "loss": 4.5904, |
| "step": 105500 |
| }, |
| { |
| "epoch": 0.87, |
| "learning_rate": 8.400772959736128e-06, |
| "loss": 4.5611, |
| "step": 106000 |
| }, |
| { |
| "epoch": 0.87, |
| "learning_rate": 8.392443652234753e-06, |
| "loss": 4.5417, |
| "step": 106500 |
| }, |
| { |
| "epoch": 0.88, |
| "learning_rate": 8.38411434473338e-06, |
| "loss": 4.519, |
| "step": 107000 |
| }, |
| { |
| "epoch": 0.88, |
| "learning_rate": 8.375785037232005e-06, |
| "loss": 4.4832, |
| "step": 107500 |
| }, |
| { |
| "epoch": 0.88, |
| "learning_rate": 8.367455729730632e-06, |
| "loss": 4.4519, |
| "step": 108000 |
| }, |
| { |
| "epoch": 0.89, |
| "learning_rate": 8.359126422229257e-06, |
| "loss": 4.4154, |
| "step": 108500 |
| }, |
| { |
| "epoch": 0.89, |
| "learning_rate": 8.350797114727882e-06, |
| "loss": 4.3833, |
| "step": 109000 |
| }, |
| { |
| "epoch": 0.9, |
| "learning_rate": 8.342467807226507e-06, |
| "loss": 4.3618, |
| "step": 109500 |
| }, |
| { |
| "epoch": 0.9, |
| "learning_rate": 8.334138499725133e-06, |
| "loss": 4.3368, |
| "step": 110000 |
| }, |
| { |
| "epoch": 0.9, |
| "eval_accuracy": 0.3518140220152654, |
| "eval_loss": 4.145502090454102, |
| "eval_runtime": 1889.0177, |
| "eval_samples_per_second": 163.236, |
| "eval_steps_per_second": 3.401, |
| "step": 110000 |
| }, |
| { |
| "epoch": 0.91, |
| "learning_rate": 8.32580919222376e-06, |
| "loss": 4.3177, |
| "step": 110500 |
| }, |
| { |
| "epoch": 0.91, |
| "learning_rate": 8.317479884722385e-06, |
| "loss": 4.2898, |
| "step": 111000 |
| }, |
| { |
| "epoch": 0.91, |
| "learning_rate": 8.30915057722101e-06, |
| "loss": 4.2661, |
| "step": 111500 |
| }, |
| { |
| "epoch": 0.92, |
| "learning_rate": 8.300821269719637e-06, |
| "loss": 4.2442, |
| "step": 112000 |
| }, |
| { |
| "epoch": 0.92, |
| "learning_rate": 8.292491962218262e-06, |
| "loss": 4.2316, |
| "step": 112500 |
| }, |
| { |
| "epoch": 0.93, |
| "learning_rate": 8.284162654716887e-06, |
| "loss": 4.2075, |
| "step": 113000 |
| }, |
| { |
| "epoch": 0.93, |
| "learning_rate": 8.275833347215514e-06, |
| "loss": 4.1821, |
| "step": 113500 |
| }, |
| { |
| "epoch": 0.93, |
| "learning_rate": 8.267504039714139e-06, |
| "loss": 4.1699, |
| "step": 114000 |
| }, |
| { |
| "epoch": 0.94, |
| "learning_rate": 8.259174732212765e-06, |
| "loss": 4.15, |
| "step": 114500 |
| }, |
| { |
| "epoch": 0.94, |
| "learning_rate": 8.25084542471139e-06, |
| "loss": 4.1313, |
| "step": 115000 |
| }, |
| { |
| "epoch": 0.95, |
| "learning_rate": 8.242516117210015e-06, |
| "loss": 4.1072, |
| "step": 115500 |
| }, |
| { |
| "epoch": 0.95, |
| "learning_rate": 8.234186809708642e-06, |
| "loss": 4.0882, |
| "step": 116000 |
| }, |
| { |
| "epoch": 0.95, |
| "learning_rate": 8.225857502207267e-06, |
| "loss": 4.0645, |
| "step": 116500 |
| }, |
| { |
| "epoch": 0.96, |
| "learning_rate": 8.217528194705892e-06, |
| "loss": 4.051, |
| "step": 117000 |
| }, |
| { |
| "epoch": 0.96, |
| "learning_rate": 8.209198887204519e-06, |
| "loss": 4.0385, |
| "step": 117500 |
| }, |
| { |
| "epoch": 0.97, |
| "learning_rate": 8.200869579703145e-06, |
| "loss": 4.0142, |
| "step": 118000 |
| }, |
| { |
| "epoch": 0.97, |
| "learning_rate": 8.19254027220177e-06, |
| "loss": 4.0046, |
| "step": 118500 |
| }, |
| { |
| "epoch": 0.97, |
| "learning_rate": 8.184210964700395e-06, |
| "loss": 3.9869, |
| "step": 119000 |
| }, |
| { |
| "epoch": 0.98, |
| "learning_rate": 8.17588165719902e-06, |
| "loss": 3.9595, |
| "step": 119500 |
| }, |
| { |
| "epoch": 0.98, |
| "learning_rate": 8.167552349697647e-06, |
| "loss": 3.9513, |
| "step": 120000 |
| }, |
| { |
| "epoch": 0.98, |
| "eval_accuracy": 0.3964395339200828, |
| "eval_loss": 3.765943765640259, |
| "eval_runtime": 1895.0159, |
| "eval_samples_per_second": 162.719, |
| "eval_steps_per_second": 3.39, |
| "step": 120000 |
| }, |
| { |
| "epoch": 0.99, |
| "learning_rate": 8.159223042196272e-06, |
| "loss": 3.9299, |
| "step": 120500 |
| }, |
| { |
| "epoch": 0.99, |
| "learning_rate": 8.150893734694899e-06, |
| "loss": 3.9074, |
| "step": 121000 |
| }, |
| { |
| "epoch": 1.0, |
| "learning_rate": 8.142564427193524e-06, |
| "loss": 3.9006, |
| "step": 121500 |
| }, |
| { |
| "epoch": 1.0, |
| "learning_rate": 8.13423511969215e-06, |
| "loss": 3.8772, |
| "step": 122000 |
| }, |
| { |
| "epoch": 1.0, |
| "learning_rate": 8.125905812190775e-06, |
| "loss": 3.863, |
| "step": 122500 |
| }, |
| { |
| "epoch": 1.01, |
| "learning_rate": 8.1175765046894e-06, |
| "loss": 3.8501, |
| "step": 123000 |
| }, |
| { |
| "epoch": 1.01, |
| "learning_rate": 8.109247197188025e-06, |
| "loss": 3.8387, |
| "step": 123500 |
| }, |
| { |
| "epoch": 1.02, |
| "learning_rate": 8.100917889686652e-06, |
| "loss": 3.8185, |
| "step": 124000 |
| }, |
| { |
| "epoch": 1.02, |
| "learning_rate": 8.092588582185277e-06, |
| "loss": 3.8097, |
| "step": 124500 |
| }, |
| { |
| "epoch": 1.02, |
| "learning_rate": 8.084259274683904e-06, |
| "loss": 3.7949, |
| "step": 125000 |
| }, |
| { |
| "epoch": 1.03, |
| "learning_rate": 8.075929967182529e-06, |
| "loss": 3.7928, |
| "step": 125500 |
| }, |
| { |
| "epoch": 1.03, |
| "learning_rate": 8.067600659681156e-06, |
| "loss": 3.7713, |
| "step": 126000 |
| }, |
| { |
| "epoch": 1.04, |
| "learning_rate": 8.05927135217978e-06, |
| "loss": 3.7639, |
| "step": 126500 |
| }, |
| { |
| "epoch": 1.04, |
| "learning_rate": 8.050942044678406e-06, |
| "loss": 3.7498, |
| "step": 127000 |
| }, |
| { |
| "epoch": 1.04, |
| "learning_rate": 8.04261273717703e-06, |
| "loss": 3.7394, |
| "step": 127500 |
| }, |
| { |
| "epoch": 1.05, |
| "learning_rate": 8.034283429675657e-06, |
| "loss": 3.7273, |
| "step": 128000 |
| }, |
| { |
| "epoch": 1.05, |
| "learning_rate": 8.025954122174284e-06, |
| "loss": 3.7125, |
| "step": 128500 |
| }, |
| { |
| "epoch": 1.06, |
| "learning_rate": 8.017624814672909e-06, |
| "loss": 3.7099, |
| "step": 129000 |
| }, |
| { |
| "epoch": 1.06, |
| "learning_rate": 8.009295507171534e-06, |
| "loss": 3.6948, |
| "step": 129500 |
| }, |
| { |
| "epoch": 1.07, |
| "learning_rate": 8.00096619967016e-06, |
| "loss": 3.682, |
| "step": 130000 |
| }, |
| { |
| "epoch": 1.07, |
| "eval_accuracy": 0.42475269250855335, |
| "eval_loss": 3.5328402519226074, |
| "eval_runtime": 1894.6078, |
| "eval_samples_per_second": 162.755, |
| "eval_steps_per_second": 3.391, |
| "step": 130000 |
| }, |
| { |
| "epoch": 1.07, |
| "learning_rate": 7.992636892168786e-06, |
| "loss": 3.67, |
| "step": 130500 |
| }, |
| { |
| "epoch": 1.07, |
| "learning_rate": 7.98430758466741e-06, |
| "loss": 3.6717, |
| "step": 131000 |
| }, |
| { |
| "epoch": 1.08, |
| "learning_rate": 7.975978277166037e-06, |
| "loss": 3.6574, |
| "step": 131500 |
| }, |
| { |
| "epoch": 1.08, |
| "learning_rate": 7.967648969664662e-06, |
| "loss": 3.6461, |
| "step": 132000 |
| }, |
| { |
| "epoch": 1.09, |
| "learning_rate": 7.959319662163289e-06, |
| "loss": 3.6403, |
| "step": 132500 |
| }, |
| { |
| "epoch": 1.09, |
| "learning_rate": 7.950990354661914e-06, |
| "loss": 3.6264, |
| "step": 133000 |
| }, |
| { |
| "epoch": 1.09, |
| "learning_rate": 7.942661047160539e-06, |
| "loss": 3.618, |
| "step": 133500 |
| }, |
| { |
| "epoch": 1.1, |
| "learning_rate": 7.934331739659166e-06, |
| "loss": 3.6069, |
| "step": 134000 |
| }, |
| { |
| "epoch": 1.1, |
| "learning_rate": 7.92600243215779e-06, |
| "loss": 3.6018, |
| "step": 134500 |
| }, |
| { |
| "epoch": 1.11, |
| "learning_rate": 7.917673124656416e-06, |
| "loss": 3.5847, |
| "step": 135000 |
| }, |
| { |
| "epoch": 1.11, |
| "learning_rate": 7.909343817155042e-06, |
| "loss": 3.5878, |
| "step": 135500 |
| }, |
| { |
| "epoch": 1.11, |
| "learning_rate": 7.901014509653669e-06, |
| "loss": 3.5743, |
| "step": 136000 |
| }, |
| { |
| "epoch": 1.12, |
| "learning_rate": 7.892685202152294e-06, |
| "loss": 3.5603, |
| "step": 136500 |
| }, |
| { |
| "epoch": 1.12, |
| "learning_rate": 7.884355894650919e-06, |
| "loss": 3.5653, |
| "step": 137000 |
| }, |
| { |
| "epoch": 1.13, |
| "learning_rate": 7.876026587149544e-06, |
| "loss": 3.5437, |
| "step": 137500 |
| }, |
| { |
| "epoch": 1.13, |
| "learning_rate": 7.86769727964817e-06, |
| "loss": 3.5388, |
| "step": 138000 |
| }, |
| { |
| "epoch": 1.13, |
| "learning_rate": 7.859367972146796e-06, |
| "loss": 3.5319, |
| "step": 138500 |
| }, |
| { |
| "epoch": 1.14, |
| "learning_rate": 7.851038664645423e-06, |
| "loss": 3.5249, |
| "step": 139000 |
| }, |
| { |
| "epoch": 1.14, |
| "learning_rate": 7.842709357144048e-06, |
| "loss": 3.5169, |
| "step": 139500 |
| }, |
| { |
| "epoch": 1.15, |
| "learning_rate": 7.834380049642674e-06, |
| "loss": 3.5114, |
| "step": 140000 |
| }, |
| { |
| "epoch": 1.15, |
| "eval_accuracy": 0.444097590282513, |
| "eval_loss": 3.3714852333068848, |
| "eval_runtime": 1893.5204, |
| "eval_samples_per_second": 162.848, |
| "eval_steps_per_second": 3.393, |
| "step": 140000 |
| }, |
| { |
| "epoch": 1.15, |
| "learning_rate": 7.8260507421413e-06, |
| "loss": 3.5045, |
| "step": 140500 |
| }, |
| { |
| "epoch": 1.16, |
| "learning_rate": 7.817721434639924e-06, |
| "loss": 3.4977, |
| "step": 141000 |
| }, |
| { |
| "epoch": 1.16, |
| "learning_rate": 7.809392127138551e-06, |
| "loss": 3.4785, |
| "step": 141500 |
| }, |
| { |
| "epoch": 1.16, |
| "learning_rate": 7.801062819637176e-06, |
| "loss": 3.4776, |
| "step": 142000 |
| }, |
| { |
| "epoch": 1.17, |
| "learning_rate": 7.792733512135801e-06, |
| "loss": 3.4772, |
| "step": 142500 |
| }, |
| { |
| "epoch": 1.17, |
| "learning_rate": 7.784404204634428e-06, |
| "loss": 3.4675, |
| "step": 143000 |
| }, |
| { |
| "epoch": 1.18, |
| "learning_rate": 7.776074897133054e-06, |
| "loss": 3.4608, |
| "step": 143500 |
| }, |
| { |
| "epoch": 1.18, |
| "learning_rate": 7.76774558963168e-06, |
| "loss": 3.4577, |
| "step": 144000 |
| }, |
| { |
| "epoch": 1.18, |
| "learning_rate": 7.759416282130304e-06, |
| "loss": 3.4489, |
| "step": 144500 |
| }, |
| { |
| "epoch": 1.19, |
| "learning_rate": 7.75108697462893e-06, |
| "loss": 3.4387, |
| "step": 145000 |
| }, |
| { |
| "epoch": 1.19, |
| "learning_rate": 7.742757667127556e-06, |
| "loss": 3.4275, |
| "step": 145500 |
| }, |
| { |
| "epoch": 1.2, |
| "learning_rate": 7.734428359626181e-06, |
| "loss": 3.42, |
| "step": 146000 |
| }, |
| { |
| "epoch": 1.2, |
| "learning_rate": 7.726099052124808e-06, |
| "loss": 3.4231, |
| "step": 146500 |
| }, |
| { |
| "epoch": 1.2, |
| "learning_rate": 7.717769744623433e-06, |
| "loss": 3.4142, |
| "step": 147000 |
| }, |
| { |
| "epoch": 1.21, |
| "learning_rate": 7.70944043712206e-06, |
| "loss": 3.3983, |
| "step": 147500 |
| }, |
| { |
| "epoch": 1.21, |
| "learning_rate": 7.701111129620684e-06, |
| "loss": 3.3993, |
| "step": 148000 |
| }, |
| { |
| "epoch": 1.22, |
| "learning_rate": 7.69278182211931e-06, |
| "loss": 3.4018, |
| "step": 148500 |
| }, |
| { |
| "epoch": 1.22, |
| "learning_rate": 7.684452514617934e-06, |
| "loss": 3.3899, |
| "step": 149000 |
| }, |
| { |
| "epoch": 1.22, |
| "learning_rate": 7.676123207116561e-06, |
| "loss": 3.3898, |
| "step": 149500 |
| }, |
| { |
| "epoch": 1.23, |
| "learning_rate": 7.667793899615186e-06, |
| "loss": 3.3789, |
| "step": 150000 |
| }, |
| { |
| "epoch": 1.23, |
| "eval_accuracy": 0.4591399132250148, |
| "eval_loss": 3.2500367164611816, |
| "eval_runtime": 1896.6147, |
| "eval_samples_per_second": 162.582, |
| "eval_steps_per_second": 3.388, |
| "step": 150000 |
| }, |
| { |
| "epoch": 1.23, |
| "learning_rate": 7.659464592113813e-06, |
| "loss": 3.3734, |
| "step": 150500 |
| }, |
| { |
| "epoch": 1.24, |
| "learning_rate": 7.651135284612438e-06, |
| "loss": 3.3718, |
| "step": 151000 |
| }, |
| { |
| "epoch": 1.24, |
| "learning_rate": 7.642805977111065e-06, |
| "loss": 3.3584, |
| "step": 151500 |
| }, |
| { |
| "epoch": 1.25, |
| "learning_rate": 7.63447666960969e-06, |
| "loss": 3.3562, |
| "step": 152000 |
| }, |
| { |
| "epoch": 1.25, |
| "learning_rate": 7.6261473621083145e-06, |
| "loss": 3.3518, |
| "step": 152500 |
| }, |
| { |
| "epoch": 1.25, |
| "learning_rate": 7.61781805460694e-06, |
| "loss": 3.3418, |
| "step": 153000 |
| }, |
| { |
| "epoch": 1.26, |
| "learning_rate": 7.609488747105567e-06, |
| "loss": 3.3428, |
| "step": 153500 |
| }, |
| { |
| "epoch": 1.26, |
| "learning_rate": 7.601159439604192e-06, |
| "loss": 3.3316, |
| "step": 154000 |
| }, |
| { |
| "epoch": 1.27, |
| "learning_rate": 7.592830132102817e-06, |
| "loss": 3.3251, |
| "step": 154500 |
| }, |
| { |
| "epoch": 1.27, |
| "learning_rate": 7.584500824601443e-06, |
| "loss": 3.3229, |
| "step": 155000 |
| }, |
| { |
| "epoch": 1.27, |
| "learning_rate": 7.57617151710007e-06, |
| "loss": 3.3077, |
| "step": 155500 |
| }, |
| { |
| "epoch": 1.28, |
| "learning_rate": 7.567842209598695e-06, |
| "loss": 3.3107, |
| "step": 156000 |
| }, |
| { |
| "epoch": 1.28, |
| "learning_rate": 7.5595129020973205e-06, |
| "loss": 3.3034, |
| "step": 156500 |
| }, |
| { |
| "epoch": 1.29, |
| "learning_rate": 7.5511835945959455e-06, |
| "loss": 3.3045, |
| "step": 157000 |
| }, |
| { |
| "epoch": 1.29, |
| "learning_rate": 7.542854287094572e-06, |
| "loss": 3.2894, |
| "step": 157500 |
| }, |
| { |
| "epoch": 1.29, |
| "learning_rate": 7.534524979593197e-06, |
| "loss": 3.2906, |
| "step": 158000 |
| }, |
| { |
| "epoch": 1.3, |
| "learning_rate": 7.526195672091823e-06, |
| "loss": 3.2767, |
| "step": 158500 |
| }, |
| { |
| "epoch": 1.3, |
| "learning_rate": 7.517866364590448e-06, |
| "loss": 3.2807, |
| "step": 159000 |
| }, |
| { |
| "epoch": 1.31, |
| "learning_rate": 7.509537057089075e-06, |
| "loss": 3.273, |
| "step": 159500 |
| }, |
| { |
| "epoch": 1.31, |
| "learning_rate": 7.5012077495877e-06, |
| "loss": 3.2776, |
| "step": 160000 |
| }, |
| { |
| "epoch": 1.31, |
| "eval_accuracy": 0.4708693238738307, |
| "eval_loss": 3.1468160152435303, |
| "eval_runtime": 1892.9525, |
| "eval_samples_per_second": 162.897, |
| "eval_steps_per_second": 3.394, |
| "step": 160000 |
| }, |
| { |
| "epoch": 1.31, |
| "learning_rate": 7.492878442086326e-06, |
| "loss": 3.2733, |
| "step": 160500 |
| }, |
| { |
| "epoch": 1.32, |
| "learning_rate": 7.484549134584951e-06, |
| "loss": 3.2581, |
| "step": 161000 |
| }, |
| { |
| "epoch": 1.32, |
| "learning_rate": 7.476219827083577e-06, |
| "loss": 3.2499, |
| "step": 161500 |
| }, |
| { |
| "epoch": 1.33, |
| "learning_rate": 7.467890519582202e-06, |
| "loss": 3.251, |
| "step": 162000 |
| }, |
| { |
| "epoch": 1.33, |
| "learning_rate": 7.459561212080828e-06, |
| "loss": 3.2393, |
| "step": 162500 |
| }, |
| { |
| "epoch": 1.34, |
| "learning_rate": 7.451231904579453e-06, |
| "loss": 3.2428, |
| "step": 163000 |
| }, |
| { |
| "epoch": 1.34, |
| "learning_rate": 7.44290259707808e-06, |
| "loss": 3.2365, |
| "step": 163500 |
| }, |
| { |
| "epoch": 1.34, |
| "learning_rate": 7.434573289576706e-06, |
| "loss": 3.2368, |
| "step": 164000 |
| }, |
| { |
| "epoch": 1.35, |
| "learning_rate": 7.426243982075331e-06, |
| "loss": 3.2384, |
| "step": 164500 |
| }, |
| { |
| "epoch": 1.35, |
| "learning_rate": 7.417914674573956e-06, |
| "loss": 3.2204, |
| "step": 165000 |
| }, |
| { |
| "epoch": 1.36, |
| "learning_rate": 7.409585367072582e-06, |
| "loss": 3.215, |
| "step": 165500 |
| }, |
| { |
| "epoch": 1.36, |
| "learning_rate": 7.401256059571208e-06, |
| "loss": 3.2115, |
| "step": 166000 |
| }, |
| { |
| "epoch": 1.36, |
| "learning_rate": 7.392926752069833e-06, |
| "loss": 3.2026, |
| "step": 166500 |
| }, |
| { |
| "epoch": 1.37, |
| "learning_rate": 7.384597444568459e-06, |
| "loss": 3.1968, |
| "step": 167000 |
| }, |
| { |
| "epoch": 1.37, |
| "learning_rate": 7.376268137067085e-06, |
| "loss": 3.1985, |
| "step": 167500 |
| }, |
| { |
| "epoch": 1.38, |
| "learning_rate": 7.367938829565711e-06, |
| "loss": 3.1937, |
| "step": 168000 |
| }, |
| { |
| "epoch": 1.38, |
| "learning_rate": 7.359609522064336e-06, |
| "loss": 3.1884, |
| "step": 168500 |
| }, |
| { |
| "epoch": 1.38, |
| "learning_rate": 7.351280214562962e-06, |
| "loss": 3.2047, |
| "step": 169000 |
| }, |
| { |
| "epoch": 1.39, |
| "learning_rate": 7.3429509070615875e-06, |
| "loss": 3.2154, |
| "step": 169500 |
| }, |
| { |
| "epoch": 1.39, |
| "learning_rate": 7.334621599560213e-06, |
| "loss": 3.204, |
| "step": 170000 |
| }, |
| { |
| "epoch": 1.39, |
| "eval_accuracy": 0.47839236055001166, |
| "eval_loss": 3.0898542404174805, |
| "eval_runtime": 1890.9247, |
| "eval_samples_per_second": 163.072, |
| "eval_steps_per_second": 3.398, |
| "step": 170000 |
| }, |
| { |
| "epoch": 1.4, |
| "learning_rate": 7.326292292058838e-06, |
| "loss": 3.1964, |
| "step": 170500 |
| }, |
| { |
| "epoch": 1.4, |
| "learning_rate": 7.317962984557464e-06, |
| "loss": 3.2063, |
| "step": 171000 |
| }, |
| { |
| "epoch": 1.41, |
| "learning_rate": 7.309633677056091e-06, |
| "loss": 3.19, |
| "step": 171500 |
| }, |
| { |
| "epoch": 1.41, |
| "learning_rate": 7.301304369554716e-06, |
| "loss": 3.1761, |
| "step": 172000 |
| }, |
| { |
| "epoch": 1.41, |
| "learning_rate": 7.292975062053341e-06, |
| "loss": 3.1667, |
| "step": 172500 |
| }, |
| { |
| "epoch": 1.42, |
| "learning_rate": 7.284645754551967e-06, |
| "loss": 3.1599, |
| "step": 173000 |
| }, |
| { |
| "epoch": 1.42, |
| "learning_rate": 7.2763164470505934e-06, |
| "loss": 3.1548, |
| "step": 173500 |
| }, |
| { |
| "epoch": 1.43, |
| "learning_rate": 7.2679871395492184e-06, |
| "loss": 3.1478, |
| "step": 174000 |
| }, |
| { |
| "epoch": 1.43, |
| "learning_rate": 7.259657832047844e-06, |
| "loss": 3.1411, |
| "step": 174500 |
| }, |
| { |
| "epoch": 1.43, |
| "learning_rate": 7.251328524546469e-06, |
| "loss": 3.1372, |
| "step": 175000 |
| }, |
| { |
| "epoch": 1.44, |
| "learning_rate": 7.242999217045096e-06, |
| "loss": 3.1412, |
| "step": 175500 |
| }, |
| { |
| "epoch": 1.44, |
| "learning_rate": 7.234669909543721e-06, |
| "loss": 3.131, |
| "step": 176000 |
| }, |
| { |
| "epoch": 1.45, |
| "learning_rate": 7.226340602042347e-06, |
| "loss": 3.1289, |
| "step": 176500 |
| }, |
| { |
| "epoch": 1.45, |
| "learning_rate": 7.218011294540972e-06, |
| "loss": 3.1213, |
| "step": 177000 |
| }, |
| { |
| "epoch": 1.45, |
| "learning_rate": 7.2096819870395985e-06, |
| "loss": 3.1237, |
| "step": 177500 |
| }, |
| { |
| "epoch": 1.46, |
| "learning_rate": 7.2013526795382235e-06, |
| "loss": 3.1241, |
| "step": 178000 |
| }, |
| { |
| "epoch": 1.46, |
| "learning_rate": 7.193023372036849e-06, |
| "loss": 3.1235, |
| "step": 178500 |
| }, |
| { |
| "epoch": 1.47, |
| "learning_rate": 7.184694064535474e-06, |
| "loss": 3.1086, |
| "step": 179000 |
| }, |
| { |
| "epoch": 1.47, |
| "learning_rate": 7.176364757034101e-06, |
| "loss": 3.1052, |
| "step": 179500 |
| }, |
| { |
| "epoch": 1.47, |
| "learning_rate": 7.168035449532726e-06, |
| "loss": 3.1051, |
| "step": 180000 |
| }, |
| { |
| "epoch": 1.47, |
| "eval_accuracy": 0.48981425578624954, |
| "eval_loss": 2.991220235824585, |
| "eval_runtime": 1884.2806, |
| "eval_samples_per_second": 163.647, |
| "eval_steps_per_second": 3.41, |
| "step": 180000 |
| }, |
| { |
| "epoch": 1.48, |
| "learning_rate": 7.159706142031352e-06, |
| "loss": 3.098, |
| "step": 180500 |
| }, |
| { |
| "epoch": 1.48, |
| "learning_rate": 7.151376834529977e-06, |
| "loss": 3.0953, |
| "step": 181000 |
| }, |
| { |
| "epoch": 1.49, |
| "learning_rate": 7.143047527028604e-06, |
| "loss": 3.0922, |
| "step": 181500 |
| }, |
| { |
| "epoch": 1.49, |
| "learning_rate": 7.1347182195272295e-06, |
| "loss": 3.0805, |
| "step": 182000 |
| }, |
| { |
| "epoch": 1.5, |
| "learning_rate": 7.1263889120258545e-06, |
| "loss": 3.083, |
| "step": 182500 |
| }, |
| { |
| "epoch": 1.5, |
| "learning_rate": 7.118059604524481e-06, |
| "loss": 3.0809, |
| "step": 183000 |
| }, |
| { |
| "epoch": 1.5, |
| "learning_rate": 7.109730297023106e-06, |
| "loss": 3.0931, |
| "step": 183500 |
| }, |
| { |
| "epoch": 1.51, |
| "learning_rate": 7.101400989521732e-06, |
| "loss": 3.0883, |
| "step": 184000 |
| }, |
| { |
| "epoch": 1.51, |
| "learning_rate": 7.093071682020357e-06, |
| "loss": 3.0829, |
| "step": 184500 |
| }, |
| { |
| "epoch": 1.52, |
| "learning_rate": 7.084742374518984e-06, |
| "loss": 3.0723, |
| "step": 185000 |
| }, |
| { |
| "epoch": 1.52, |
| "learning_rate": 7.076413067017609e-06, |
| "loss": 3.064, |
| "step": 185500 |
| }, |
| { |
| "epoch": 1.52, |
| "learning_rate": 7.068083759516235e-06, |
| "loss": 3.0667, |
| "step": 186000 |
| }, |
| { |
| "epoch": 1.53, |
| "learning_rate": 7.05975445201486e-06, |
| "loss": 3.064, |
| "step": 186500 |
| }, |
| { |
| "epoch": 1.53, |
| "learning_rate": 7.051425144513486e-06, |
| "loss": 3.0576, |
| "step": 187000 |
| }, |
| { |
| "epoch": 1.54, |
| "learning_rate": 7.043095837012111e-06, |
| "loss": 3.0503, |
| "step": 187500 |
| }, |
| { |
| "epoch": 1.54, |
| "learning_rate": 7.034766529510737e-06, |
| "loss": 3.0592, |
| "step": 188000 |
| }, |
| { |
| "epoch": 1.54, |
| "step": 188477, |
| "total_flos": 2.6804727964751626e+18, |
| "train_loss": 4.882590240800892, |
| "train_runtime": 197983.7555, |
| "train_samples_per_second": 147.961, |
| "train_steps_per_second": 3.083 |
| } |
| ], |
| "max_steps": 610290, |
| "num_train_epochs": 5, |
| "total_flos": 2.6804727964751626e+18, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
