| { | |
| "best_metric": 2.6168134212493896, | |
| "best_model_checkpoint": "bert_12_layer_model_v1_complete_training_new_wt_init_48/checkpoint-180000", | |
| "epoch": 1.5484114109685558, | |
| "global_step": 188996, | |
| "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": 9.4266, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 7.6939, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 1.5e-06, | |
| "loss": 7.4576, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 7.297, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 2.5e-06, | |
| "loss": 7.213, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 3e-06, | |
| "loss": 7.1534, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 3.5e-06, | |
| "loss": 7.0715, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 6.9897, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 4.5e-06, | |
| "loss": 6.9045, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 5e-06, | |
| "loss": 6.8414, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 5.500000000000001e-06, | |
| "loss": 6.7709, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 6e-06, | |
| "loss": 6.7065, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 6.5000000000000004e-06, | |
| "loss": 6.6573, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 7e-06, | |
| "loss": 6.6071, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 6.5686, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 6.5246, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 8.5e-06, | |
| "loss": 6.4901, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 9e-06, | |
| "loss": 6.4663, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 9.5e-06, | |
| "loss": 6.4372, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 1e-05, | |
| "loss": 6.4172, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "eval_accuracy": 0.13172429743617925, | |
| "eval_loss": 6.366359710693359, | |
| "eval_runtime": 1883.8145, | |
| "eval_samples_per_second": 163.687, | |
| "eval_steps_per_second": 3.411, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 9.991670692498626e-06, | |
| "loss": 6.3939, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 9.983341384997253e-06, | |
| "loss": 6.3788, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 9.975012077495878e-06, | |
| "loss": 6.3558, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 9.966682769994504e-06, | |
| "loss": 6.3399, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 9.95835346249313e-06, | |
| "loss": 6.3157, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 9.950024154991756e-06, | |
| "loss": 6.3052, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 9.941694847490381e-06, | |
| "loss": 6.2934, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 9.933365539989006e-06, | |
| "loss": 6.2813, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 9.925036232487631e-06, | |
| "loss": 6.2647, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 9.916706924986258e-06, | |
| "loss": 6.2555, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 9.908377617484883e-06, | |
| "loss": 6.238, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 9.90004830998351e-06, | |
| "loss": 6.2365, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 9.891719002482134e-06, | |
| "loss": 6.2343, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 9.883389694980761e-06, | |
| "loss": 6.2157, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 9.875060387479386e-06, | |
| "loss": 6.2164, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 9.866731079978011e-06, | |
| "loss": 6.2077, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 9.858401772476636e-06, | |
| "loss": 6.2059, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 9.850072464975263e-06, | |
| "loss": 6.188, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 9.84174315747389e-06, | |
| "loss": 6.1352, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 9.833413849972514e-06, | |
| "loss": 6.1014, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "eval_accuracy": 0.14966543588927317, | |
| "eval_loss": 6.053226470947266, | |
| "eval_runtime": 1883.5651, | |
| "eval_samples_per_second": 163.709, | |
| "eval_steps_per_second": 3.411, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 9.82508454247114e-06, | |
| "loss": 6.0778, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 9.816755234969766e-06, | |
| "loss": 6.058, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 9.808425927468391e-06, | |
| "loss": 6.045, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 9.800096619967016e-06, | |
| "loss": 6.0169, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 9.791767312465643e-06, | |
| "loss": 6.0074, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 9.783438004964268e-06, | |
| "loss": 5.9823, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 9.775108697462895e-06, | |
| "loss": 5.9702, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 9.76677938996152e-06, | |
| "loss": 5.9545, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 9.758450082460145e-06, | |
| "loss": 5.9453, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 9.750120774958771e-06, | |
| "loss": 5.9323, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 9.741791467457396e-06, | |
| "loss": 5.9177, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 9.733462159956021e-06, | |
| "loss": 5.9071, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 9.725132852454648e-06, | |
| "loss": 5.9029, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 9.716803544953275e-06, | |
| "loss": 5.8895, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 9.7084742374519e-06, | |
| "loss": 5.8761, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 9.700144929950525e-06, | |
| "loss": 5.8666, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 9.69181562244915e-06, | |
| "loss": 5.8604, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 9.683486314947776e-06, | |
| "loss": 5.8484, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 9.675157007446401e-06, | |
| "loss": 5.8343, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 9.666827699945028e-06, | |
| "loss": 5.8255, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "eval_accuracy": 0.16555430235844348, | |
| "eval_loss": 5.746612071990967, | |
| "eval_runtime": 1885.237, | |
| "eval_samples_per_second": 163.564, | |
| "eval_steps_per_second": 3.408, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 9.658498392443653e-06, | |
| "loss": 5.8017, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 9.65016908494228e-06, | |
| "loss": 5.7818, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 9.641839777440905e-06, | |
| "loss": 5.7582, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 9.63351046993953e-06, | |
| "loss": 5.7172, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 9.625181162438155e-06, | |
| "loss": 5.6734, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 9.616851854936781e-06, | |
| "loss": 5.6323, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 9.608522547435406e-06, | |
| "loss": 5.5832, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 9.600193239934033e-06, | |
| "loss": 5.5368, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 9.591863932432658e-06, | |
| "loss": 5.4847, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 9.583534624931285e-06, | |
| "loss": 5.4441, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 9.57520531742991e-06, | |
| "loss": 5.4073, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 9.566876009928535e-06, | |
| "loss": 5.3619, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 9.55854670242716e-06, | |
| "loss": 5.3196, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 9.550217394925787e-06, | |
| "loss": 5.2839, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 9.541888087424413e-06, | |
| "loss": 5.2458, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 9.533558779923038e-06, | |
| "loss": 5.2056, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 9.525229472421663e-06, | |
| "loss": 5.1656, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 9.51690016492029e-06, | |
| "loss": 5.1229, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 9.508570857418915e-06, | |
| "loss": 5.0891, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 9.50024154991754e-06, | |
| "loss": 5.0543, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "eval_accuracy": 0.27521187544371295, | |
| "eval_loss": 4.753915786743164, | |
| "eval_runtime": 1883.0128, | |
| "eval_samples_per_second": 163.757, | |
| "eval_steps_per_second": 3.412, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 9.491912242416167e-06, | |
| "loss": 4.9967, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 9.483582934914792e-06, | |
| "loss": 4.9497, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 9.475253627413418e-06, | |
| "loss": 4.9019, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 9.466924319912043e-06, | |
| "loss": 4.8657, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 9.458595012410668e-06, | |
| "loss": 4.8201, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 9.450265704909295e-06, | |
| "loss": 4.7853, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 9.44193639740792e-06, | |
| "loss": 4.7487, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 9.433607089906545e-06, | |
| "loss": 4.7208, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 9.425277782405172e-06, | |
| "loss": 4.6774, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 9.416948474903798e-06, | |
| "loss": 4.6161, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 9.408619167402423e-06, | |
| "loss": 4.5571, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 9.400289859901048e-06, | |
| "loss": 4.4937, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 9.391960552399673e-06, | |
| "loss": 4.4455, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 9.3836312448983e-06, | |
| "loss": 4.3969, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 9.375301937396925e-06, | |
| "loss": 4.3589, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 9.366972629895552e-06, | |
| "loss": 4.3168, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 9.358643322394177e-06, | |
| "loss": 4.2717, | |
| "step": 48500 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 9.350314014892804e-06, | |
| "loss": 4.2346, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 9.341984707391429e-06, | |
| "loss": 4.1988, | |
| "step": 49500 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 9.333655399890054e-06, | |
| "loss": 4.1669, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "eval_accuracy": 0.3738831835398605, | |
| "eval_loss": 3.8567991256713867, | |
| "eval_runtime": 1885.2234, | |
| "eval_samples_per_second": 163.565, | |
| "eval_steps_per_second": 3.408, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 9.325326092388679e-06, | |
| "loss": 4.1292, | |
| "step": 50500 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 9.316996784887305e-06, | |
| "loss": 4.1036, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 9.30866747738593e-06, | |
| "loss": 4.075, | |
| "step": 51500 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 9.300338169884557e-06, | |
| "loss": 4.0441, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 9.292008862383182e-06, | |
| "loss": 4.0186, | |
| "step": 52500 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 9.283679554881809e-06, | |
| "loss": 3.9923, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 9.275350247380434e-06, | |
| "loss": 3.9693, | |
| "step": 53500 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 9.267020939879059e-06, | |
| "loss": 3.9433, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 9.258691632377685e-06, | |
| "loss": 3.9272, | |
| "step": 54500 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 9.25036232487631e-06, | |
| "loss": 3.898, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 9.242033017374937e-06, | |
| "loss": 3.8845, | |
| "step": 55500 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 9.233703709873562e-06, | |
| "loss": 3.8562, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 9.225374402372189e-06, | |
| "loss": 3.8395, | |
| "step": 56500 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 9.217045094870814e-06, | |
| "loss": 3.8162, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 9.208715787369439e-06, | |
| "loss": 3.8019, | |
| "step": 57500 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 9.200386479868064e-06, | |
| "loss": 3.7828, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 9.19205717236669e-06, | |
| "loss": 3.766, | |
| "step": 58500 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 9.183727864865315e-06, | |
| "loss": 3.752, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 9.175398557363942e-06, | |
| "loss": 3.7387, | |
| "step": 59500 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 9.167069249862567e-06, | |
| "loss": 3.7259, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "eval_accuracy": 0.42145711116396833, | |
| "eval_loss": 3.475050687789917, | |
| "eval_runtime": 1883.4825, | |
| "eval_samples_per_second": 163.716, | |
| "eval_steps_per_second": 3.411, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 9.158739942361194e-06, | |
| "loss": 3.7011, | |
| "step": 60500 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 9.150410634859819e-06, | |
| "loss": 3.6825, | |
| "step": 61000 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 9.142081327358444e-06, | |
| "loss": 3.6735, | |
| "step": 61500 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 9.133752019857069e-06, | |
| "loss": 3.6501, | |
| "step": 62000 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 9.125422712355696e-06, | |
| "loss": 3.6448, | |
| "step": 62500 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 9.11709340485432e-06, | |
| "loss": 3.6283, | |
| "step": 63000 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 9.108764097352947e-06, | |
| "loss": 3.6028, | |
| "step": 63500 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 9.100434789851572e-06, | |
| "loss": 3.593, | |
| "step": 64000 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 9.092105482350199e-06, | |
| "loss": 3.5824, | |
| "step": 64500 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 9.083776174848824e-06, | |
| "loss": 3.5685, | |
| "step": 65000 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 9.075446867347449e-06, | |
| "loss": 3.5641, | |
| "step": 65500 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 9.067117559846076e-06, | |
| "loss": 3.5463, | |
| "step": 66000 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 9.0587882523447e-06, | |
| "loss": 3.5313, | |
| "step": 66500 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 9.050458944843327e-06, | |
| "loss": 3.5104, | |
| "step": 67000 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 9.042129637341952e-06, | |
| "loss": 3.5072, | |
| "step": 67500 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 9.033800329840577e-06, | |
| "loss": 3.4979, | |
| "step": 68000 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 9.025471022339204e-06, | |
| "loss": 3.4915, | |
| "step": 68500 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 9.017141714837829e-06, | |
| "loss": 3.4737, | |
| "step": 69000 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 9.008812407336454e-06, | |
| "loss": 3.4733, | |
| "step": 69500 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 9.00048309983508e-06, | |
| "loss": 3.4547, | |
| "step": 70000 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "eval_accuracy": 0.4468989613536479, | |
| "eval_loss": 3.2615108489990234, | |
| "eval_runtime": 1883.5752, | |
| "eval_samples_per_second": 163.708, | |
| "eval_steps_per_second": 3.411, | |
| "step": 70000 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 8.992153792333706e-06, | |
| "loss": 3.4417, | |
| "step": 70500 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 8.983824484832332e-06, | |
| "loss": 3.4319, | |
| "step": 71000 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 8.975495177330957e-06, | |
| "loss": 3.4328, | |
| "step": 71500 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 8.967165869829582e-06, | |
| "loss": 3.4131, | |
| "step": 72000 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 8.958836562328209e-06, | |
| "loss": 3.4097, | |
| "step": 72500 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 8.950507254826834e-06, | |
| "loss": 3.3952, | |
| "step": 73000 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 8.942177947325459e-06, | |
| "loss": 3.3942, | |
| "step": 73500 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 8.933848639824086e-06, | |
| "loss": 3.3815, | |
| "step": 74000 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 8.925519332322713e-06, | |
| "loss": 3.3762, | |
| "step": 74500 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 8.917190024821338e-06, | |
| "loss": 3.3701, | |
| "step": 75000 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 8.908860717319963e-06, | |
| "loss": 3.3556, | |
| "step": 75500 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 8.900531409818588e-06, | |
| "loss": 3.3481, | |
| "step": 76000 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 8.892202102317214e-06, | |
| "loss": 3.3401, | |
| "step": 76500 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 8.88387279481584e-06, | |
| "loss": 3.3316, | |
| "step": 77000 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 8.875543487314466e-06, | |
| "loss": 3.327, | |
| "step": 77500 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 8.867214179813091e-06, | |
| "loss": 3.3161, | |
| "step": 78000 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 8.858884872311718e-06, | |
| "loss": 3.3151, | |
| "step": 78500 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 8.850555564810343e-06, | |
| "loss": 3.3038, | |
| "step": 79000 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 8.842226257308968e-06, | |
| "loss": 3.3062, | |
| "step": 79500 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 8.833896949807593e-06, | |
| "loss": 3.2968, | |
| "step": 80000 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "eval_accuracy": 0.4637908800686817, | |
| "eval_loss": 3.124985694885254, | |
| "eval_runtime": 1885.2786, | |
| "eval_samples_per_second": 163.56, | |
| "eval_steps_per_second": 3.408, | |
| "step": 80000 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 8.82556764230622e-06, | |
| "loss": 3.2889, | |
| "step": 80500 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 8.817238334804844e-06, | |
| "loss": 3.2831, | |
| "step": 81000 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 8.808909027303471e-06, | |
| "loss": 3.2741, | |
| "step": 81500 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 8.800579719802096e-06, | |
| "loss": 3.2666, | |
| "step": 82000 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 8.792250412300723e-06, | |
| "loss": 3.253, | |
| "step": 82500 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 8.783921104799348e-06, | |
| "loss": 3.2582, | |
| "step": 83000 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 8.775591797297973e-06, | |
| "loss": 3.2537, | |
| "step": 83500 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 8.767262489796598e-06, | |
| "loss": 3.2429, | |
| "step": 84000 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 8.758933182295224e-06, | |
| "loss": 3.2352, | |
| "step": 84500 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 8.750603874793851e-06, | |
| "loss": 3.233, | |
| "step": 85000 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 8.742274567292476e-06, | |
| "loss": 3.2309, | |
| "step": 85500 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 8.733945259791101e-06, | |
| "loss": 3.2169, | |
| "step": 86000 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 8.725615952289728e-06, | |
| "loss": 3.2167, | |
| "step": 86500 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 8.717286644788353e-06, | |
| "loss": 3.2139, | |
| "step": 87000 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 8.708957337286978e-06, | |
| "loss": 3.2059, | |
| "step": 87500 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 8.700628029785605e-06, | |
| "loss": 3.1974, | |
| "step": 88000 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 8.69229872228423e-06, | |
| "loss": 3.1979, | |
| "step": 88500 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 8.683969414782856e-06, | |
| "loss": 3.1898, | |
| "step": 89000 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 8.675640107281481e-06, | |
| "loss": 3.1824, | |
| "step": 89500 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 8.667310799780106e-06, | |
| "loss": 3.1805, | |
| "step": 90000 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "eval_accuracy": 0.47595959409087013, | |
| "eval_loss": 3.02461576461792, | |
| "eval_runtime": 1882.556, | |
| "eval_samples_per_second": 163.796, | |
| "eval_steps_per_second": 3.413, | |
| "step": 90000 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 8.658981492278733e-06, | |
| "loss": 3.1702, | |
| "step": 90500 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 8.650652184777358e-06, | |
| "loss": 3.1706, | |
| "step": 91000 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 8.642322877275983e-06, | |
| "loss": 3.1672, | |
| "step": 91500 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 8.63399356977461e-06, | |
| "loss": 3.1693, | |
| "step": 92000 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 8.625664262273236e-06, | |
| "loss": 3.1572, | |
| "step": 92500 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 8.617334954771861e-06, | |
| "loss": 3.1571, | |
| "step": 93000 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 8.609005647270486e-06, | |
| "loss": 3.1444, | |
| "step": 93500 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 8.600676339769111e-06, | |
| "loss": 3.1387, | |
| "step": 94000 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 8.592347032267738e-06, | |
| "loss": 3.1432, | |
| "step": 94500 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 8.584017724766363e-06, | |
| "loss": 3.1357, | |
| "step": 95000 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 8.57568841726499e-06, | |
| "loss": 3.1341, | |
| "step": 95500 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 8.567359109763615e-06, | |
| "loss": 3.129, | |
| "step": 96000 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 8.559029802262241e-06, | |
| "loss": 3.1224, | |
| "step": 96500 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 8.550700494760866e-06, | |
| "loss": 3.1203, | |
| "step": 97000 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 8.542371187259491e-06, | |
| "loss": 3.1172, | |
| "step": 97500 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 8.534041879758118e-06, | |
| "loss": 3.1132, | |
| "step": 98000 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 8.525712572256743e-06, | |
| "loss": 3.1145, | |
| "step": 98500 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 8.517383264755368e-06, | |
| "loss": 3.1093, | |
| "step": 99000 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 8.509053957253995e-06, | |
| "loss": 3.0963, | |
| "step": 99500 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 8.500724649752622e-06, | |
| "loss": 3.0963, | |
| "step": 100000 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "eval_accuracy": 0.48577514930940024, | |
| "eval_loss": 2.949064254760742, | |
| "eval_runtime": 1882.7548, | |
| "eval_samples_per_second": 163.779, | |
| "eval_steps_per_second": 3.413, | |
| "step": 100000 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 8.492395342251247e-06, | |
| "loss": 3.0919, | |
| "step": 100500 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 8.484066034749872e-06, | |
| "loss": 3.0798, | |
| "step": 101000 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 8.475736727248497e-06, | |
| "loss": 3.0876, | |
| "step": 101500 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 8.467407419747123e-06, | |
| "loss": 3.0796, | |
| "step": 102000 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 8.459078112245748e-06, | |
| "loss": 3.075, | |
| "step": 102500 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 8.450748804744375e-06, | |
| "loss": 3.0751, | |
| "step": 103000 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 8.442419497243e-06, | |
| "loss": 3.0693, | |
| "step": 103500 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 8.434090189741627e-06, | |
| "loss": 3.0604, | |
| "step": 104000 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 8.425760882240252e-06, | |
| "loss": 3.0722, | |
| "step": 104500 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 8.417431574738877e-06, | |
| "loss": 3.0586, | |
| "step": 105000 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 8.409102267237502e-06, | |
| "loss": 3.0644, | |
| "step": 105500 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 8.400772959736128e-06, | |
| "loss": 3.0537, | |
| "step": 106000 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 8.392443652234753e-06, | |
| "loss": 3.0509, | |
| "step": 106500 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 8.38411434473338e-06, | |
| "loss": 3.0388, | |
| "step": 107000 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 8.375785037232005e-06, | |
| "loss": 3.036, | |
| "step": 107500 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 8.367455729730632e-06, | |
| "loss": 3.0418, | |
| "step": 108000 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "learning_rate": 8.359126422229257e-06, | |
| "loss": 3.0331, | |
| "step": 108500 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "learning_rate": 8.350797114727882e-06, | |
| "loss": 3.0299, | |
| "step": 109000 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 8.342467807226507e-06, | |
| "loss": 3.0293, | |
| "step": 109500 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 8.334138499725133e-06, | |
| "loss": 3.0238, | |
| "step": 110000 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "eval_accuracy": 0.4932534853709394, | |
| "eval_loss": 2.8874318599700928, | |
| "eval_runtime": 1884.192, | |
| "eval_samples_per_second": 163.654, | |
| "eval_steps_per_second": 3.41, | |
| "step": 110000 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 8.32580919222376e-06, | |
| "loss": 3.0237, | |
| "step": 110500 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 8.317479884722385e-06, | |
| "loss": 3.0207, | |
| "step": 111000 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 8.30915057722101e-06, | |
| "loss": 3.0169, | |
| "step": 111500 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 8.300821269719637e-06, | |
| "loss": 3.0095, | |
| "step": 112000 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 8.292491962218262e-06, | |
| "loss": 3.0155, | |
| "step": 112500 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 8.284162654716887e-06, | |
| "loss": 3.0076, | |
| "step": 113000 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 8.275833347215514e-06, | |
| "loss": 3.0019, | |
| "step": 113500 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 8.267504039714139e-06, | |
| "loss": 3.0073, | |
| "step": 114000 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 8.259174732212765e-06, | |
| "loss": 2.9991, | |
| "step": 114500 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 8.25084542471139e-06, | |
| "loss": 2.9988, | |
| "step": 115000 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 8.242516117210015e-06, | |
| "loss": 2.9945, | |
| "step": 115500 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 8.234186809708642e-06, | |
| "loss": 2.9888, | |
| "step": 116000 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 8.225857502207267e-06, | |
| "loss": 2.9852, | |
| "step": 116500 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 8.217528194705892e-06, | |
| "loss": 2.9842, | |
| "step": 117000 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 8.209198887204519e-06, | |
| "loss": 2.9841, | |
| "step": 117500 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "learning_rate": 8.200869579703145e-06, | |
| "loss": 2.9797, | |
| "step": 118000 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "learning_rate": 8.19254027220177e-06, | |
| "loss": 2.984, | |
| "step": 118500 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "learning_rate": 8.184210964700395e-06, | |
| "loss": 2.977, | |
| "step": 119000 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 8.17588165719902e-06, | |
| "loss": 2.9709, | |
| "step": 119500 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 8.167552349697647e-06, | |
| "loss": 2.9729, | |
| "step": 120000 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "eval_accuracy": 0.5002575105769624, | |
| "eval_loss": 2.8317418098449707, | |
| "eval_runtime": 1883.3782, | |
| "eval_samples_per_second": 163.725, | |
| "eval_steps_per_second": 3.411, | |
| "step": 120000 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 8.159223042196272e-06, | |
| "loss": 2.966, | |
| "step": 120500 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 8.150893734694899e-06, | |
| "loss": 2.9624, | |
| "step": 121000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 8.142564427193524e-06, | |
| "loss": 2.9632, | |
| "step": 121500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 8.13423511969215e-06, | |
| "loss": 2.9551, | |
| "step": 122000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 8.125905812190775e-06, | |
| "loss": 2.9525, | |
| "step": 122500 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "learning_rate": 8.1175765046894e-06, | |
| "loss": 2.9535, | |
| "step": 123000 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "learning_rate": 8.109247197188025e-06, | |
| "loss": 2.9468, | |
| "step": 123500 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "learning_rate": 8.100917889686652e-06, | |
| "loss": 2.9421, | |
| "step": 124000 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "learning_rate": 8.092588582185277e-06, | |
| "loss": 2.9443, | |
| "step": 124500 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "learning_rate": 8.084259274683904e-06, | |
| "loss": 2.9399, | |
| "step": 125000 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "learning_rate": 8.075929967182529e-06, | |
| "loss": 2.9429, | |
| "step": 125500 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "learning_rate": 8.067600659681156e-06, | |
| "loss": 2.9341, | |
| "step": 126000 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "learning_rate": 8.05927135217978e-06, | |
| "loss": 2.9356, | |
| "step": 126500 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "learning_rate": 8.050942044678406e-06, | |
| "loss": 2.9308, | |
| "step": 127000 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "learning_rate": 8.04261273717703e-06, | |
| "loss": 2.932, | |
| "step": 127500 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "learning_rate": 8.034283429675657e-06, | |
| "loss": 2.9285, | |
| "step": 128000 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "learning_rate": 8.025954122174284e-06, | |
| "loss": 2.9218, | |
| "step": 128500 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "learning_rate": 8.017624814672909e-06, | |
| "loss": 2.9283, | |
| "step": 129000 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "learning_rate": 8.009295507171534e-06, | |
| "loss": 2.9209, | |
| "step": 129500 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "learning_rate": 8.00096619967016e-06, | |
| "loss": 2.9156, | |
| "step": 130000 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "eval_accuracy": 0.5062766691352024, | |
| "eval_loss": 2.782405376434326, | |
| "eval_runtime": 1885.6493, | |
| "eval_samples_per_second": 163.528, | |
| "eval_steps_per_second": 3.407, | |
| "step": 130000 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "learning_rate": 7.992636892168786e-06, | |
| "loss": 2.9139, | |
| "step": 130500 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "learning_rate": 7.98430758466741e-06, | |
| "loss": 2.9215, | |
| "step": 131000 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "learning_rate": 7.975978277166037e-06, | |
| "loss": 2.9168, | |
| "step": 131500 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "learning_rate": 7.967648969664662e-06, | |
| "loss": 2.9121, | |
| "step": 132000 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "learning_rate": 7.959319662163289e-06, | |
| "loss": 2.9136, | |
| "step": 132500 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "learning_rate": 7.950990354661914e-06, | |
| "loss": 2.9067, | |
| "step": 133000 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "learning_rate": 7.942661047160539e-06, | |
| "loss": 2.9035, | |
| "step": 133500 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "learning_rate": 7.934331739659166e-06, | |
| "loss": 2.8999, | |
| "step": 134000 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "learning_rate": 7.92600243215779e-06, | |
| "loss": 2.9039, | |
| "step": 134500 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "learning_rate": 7.917673124656416e-06, | |
| "loss": 2.892, | |
| "step": 135000 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "learning_rate": 7.909343817155042e-06, | |
| "loss": 2.8987, | |
| "step": 135500 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "learning_rate": 7.901014509653669e-06, | |
| "loss": 2.8955, | |
| "step": 136000 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "learning_rate": 7.892685202152294e-06, | |
| "loss": 2.8871, | |
| "step": 136500 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "learning_rate": 7.884355894650919e-06, | |
| "loss": 2.8949, | |
| "step": 137000 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "learning_rate": 7.876026587149544e-06, | |
| "loss": 2.8844, | |
| "step": 137500 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "learning_rate": 7.86769727964817e-06, | |
| "loss": 2.8853, | |
| "step": 138000 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "learning_rate": 7.859367972146796e-06, | |
| "loss": 2.8813, | |
| "step": 138500 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "learning_rate": 7.851038664645423e-06, | |
| "loss": 2.8821, | |
| "step": 139000 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "learning_rate": 7.842709357144048e-06, | |
| "loss": 2.8774, | |
| "step": 139500 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "learning_rate": 7.834380049642674e-06, | |
| "loss": 2.8784, | |
| "step": 140000 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "eval_accuracy": 0.5110604926517122, | |
| "eval_loss": 2.74336314201355, | |
| "eval_runtime": 1883.0968, | |
| "eval_samples_per_second": 163.749, | |
| "eval_steps_per_second": 3.412, | |
| "step": 140000 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "learning_rate": 7.8260507421413e-06, | |
| "loss": 2.8763, | |
| "step": 140500 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "learning_rate": 7.817721434639924e-06, | |
| "loss": 2.8725, | |
| "step": 141000 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "learning_rate": 7.809392127138551e-06, | |
| "loss": 2.8635, | |
| "step": 141500 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "learning_rate": 7.801062819637176e-06, | |
| "loss": 2.8643, | |
| "step": 142000 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "learning_rate": 7.792733512135801e-06, | |
| "loss": 2.8715, | |
| "step": 142500 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "learning_rate": 7.784404204634428e-06, | |
| "loss": 2.8642, | |
| "step": 143000 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "learning_rate": 7.776074897133054e-06, | |
| "loss": 2.8635, | |
| "step": 143500 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "learning_rate": 7.76774558963168e-06, | |
| "loss": 2.8657, | |
| "step": 144000 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "learning_rate": 7.759416282130304e-06, | |
| "loss": 2.8611, | |
| "step": 144500 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "learning_rate": 7.75108697462893e-06, | |
| "loss": 2.8554, | |
| "step": 145000 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "learning_rate": 7.742757667127556e-06, | |
| "loss": 2.8482, | |
| "step": 145500 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "learning_rate": 7.734428359626181e-06, | |
| "loss": 2.8502, | |
| "step": 146000 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "learning_rate": 7.726099052124808e-06, | |
| "loss": 2.853, | |
| "step": 146500 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "learning_rate": 7.717769744623433e-06, | |
| "loss": 2.8496, | |
| "step": 147000 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "learning_rate": 7.70944043712206e-06, | |
| "loss": 2.8406, | |
| "step": 147500 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "learning_rate": 7.701111129620684e-06, | |
| "loss": 2.8435, | |
| "step": 148000 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "learning_rate": 7.69278182211931e-06, | |
| "loss": 2.8476, | |
| "step": 148500 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "learning_rate": 7.684452514617934e-06, | |
| "loss": 2.8424, | |
| "step": 149000 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "learning_rate": 7.676123207116561e-06, | |
| "loss": 2.8426, | |
| "step": 149500 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "learning_rate": 7.667793899615186e-06, | |
| "loss": 2.8333, | |
| "step": 150000 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "eval_accuracy": 0.5164025371480699, | |
| "eval_loss": 2.7081446647644043, | |
| "eval_runtime": 1885.6747, | |
| "eval_samples_per_second": 163.526, | |
| "eval_steps_per_second": 3.407, | |
| "step": 150000 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "learning_rate": 7.659464592113813e-06, | |
| "loss": 2.832, | |
| "step": 150500 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "learning_rate": 7.651135284612438e-06, | |
| "loss": 2.8347, | |
| "step": 151000 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "learning_rate": 7.642805977111065e-06, | |
| "loss": 2.8235, | |
| "step": 151500 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "learning_rate": 7.63447666960969e-06, | |
| "loss": 2.8266, | |
| "step": 152000 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "learning_rate": 7.6261473621083145e-06, | |
| "loss": 2.8249, | |
| "step": 152500 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "learning_rate": 7.61781805460694e-06, | |
| "loss": 2.821, | |
| "step": 153000 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "learning_rate": 7.609488747105567e-06, | |
| "loss": 2.8267, | |
| "step": 153500 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "learning_rate": 7.601159439604192e-06, | |
| "loss": 2.8174, | |
| "step": 154000 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "learning_rate": 7.592830132102817e-06, | |
| "loss": 2.8148, | |
| "step": 154500 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "learning_rate": 7.584500824601443e-06, | |
| "loss": 2.8161, | |
| "step": 155000 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "learning_rate": 7.57617151710007e-06, | |
| "loss": 2.8089, | |
| "step": 155500 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "learning_rate": 7.567842209598695e-06, | |
| "loss": 2.8119, | |
| "step": 156000 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "learning_rate": 7.5595129020973205e-06, | |
| "loss": 2.8074, | |
| "step": 156500 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "learning_rate": 7.5511835945959455e-06, | |
| "loss": 2.8137, | |
| "step": 157000 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "learning_rate": 7.542854287094572e-06, | |
| "loss": 2.8003, | |
| "step": 157500 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "learning_rate": 7.534524979593197e-06, | |
| "loss": 2.807, | |
| "step": 158000 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "learning_rate": 7.526195672091823e-06, | |
| "loss": 2.7972, | |
| "step": 158500 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "learning_rate": 7.517866364590448e-06, | |
| "loss": 2.8055, | |
| "step": 159000 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "learning_rate": 7.509537057089075e-06, | |
| "loss": 2.7986, | |
| "step": 159500 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "learning_rate": 7.5012077495877e-06, | |
| "loss": 2.8043, | |
| "step": 160000 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "eval_accuracy": 0.5205336690252295, | |
| "eval_loss": 2.6730058193206787, | |
| "eval_runtime": 1883.7924, | |
| "eval_samples_per_second": 163.689, | |
| "eval_steps_per_second": 3.411, | |
| "step": 160000 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "learning_rate": 7.492878442086326e-06, | |
| "loss": 2.8048, | |
| "step": 160500 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "learning_rate": 7.484549134584951e-06, | |
| "loss": 2.7957, | |
| "step": 161000 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "learning_rate": 7.476219827083577e-06, | |
| "loss": 2.7908, | |
| "step": 161500 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "learning_rate": 7.467890519582202e-06, | |
| "loss": 2.7931, | |
| "step": 162000 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "learning_rate": 7.459561212080828e-06, | |
| "loss": 2.7849, | |
| "step": 162500 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "learning_rate": 7.451231904579453e-06, | |
| "loss": 2.7917, | |
| "step": 163000 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "learning_rate": 7.44290259707808e-06, | |
| "loss": 2.7877, | |
| "step": 163500 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "learning_rate": 7.434573289576706e-06, | |
| "loss": 2.7834, | |
| "step": 164000 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "learning_rate": 7.426243982075331e-06, | |
| "loss": 2.7821, | |
| "step": 164500 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "learning_rate": 7.417914674573956e-06, | |
| "loss": 2.7804, | |
| "step": 165000 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "learning_rate": 7.409585367072582e-06, | |
| "loss": 2.7802, | |
| "step": 165500 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "learning_rate": 7.401256059571208e-06, | |
| "loss": 2.7816, | |
| "step": 166000 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "learning_rate": 7.392926752069833e-06, | |
| "loss": 2.7721, | |
| "step": 166500 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "learning_rate": 7.384597444568459e-06, | |
| "loss": 2.7703, | |
| "step": 167000 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "learning_rate": 7.376268137067085e-06, | |
| "loss": 2.7757, | |
| "step": 167500 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "learning_rate": 7.367938829565711e-06, | |
| "loss": 2.7725, | |
| "step": 168000 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "learning_rate": 7.359609522064336e-06, | |
| "loss": 2.7716, | |
| "step": 168500 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "learning_rate": 7.351280214562962e-06, | |
| "loss": 2.7708, | |
| "step": 169000 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "learning_rate": 7.3429509070615875e-06, | |
| "loss": 2.7708, | |
| "step": 169500 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "learning_rate": 7.334621599560213e-06, | |
| "loss": 2.7646, | |
| "step": 170000 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "eval_accuracy": 0.5247538851191895, | |
| "eval_loss": 2.6436333656311035, | |
| "eval_runtime": 1885.7036, | |
| "eval_samples_per_second": 163.523, | |
| "eval_steps_per_second": 3.407, | |
| "step": 170000 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "learning_rate": 7.326292292058838e-06, | |
| "loss": 2.7601, | |
| "step": 170500 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "learning_rate": 7.317962984557464e-06, | |
| "loss": 2.7683, | |
| "step": 171000 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "learning_rate": 7.309633677056091e-06, | |
| "loss": 2.7655, | |
| "step": 171500 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "learning_rate": 7.301304369554716e-06, | |
| "loss": 2.7614, | |
| "step": 172000 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "learning_rate": 7.292975062053341e-06, | |
| "loss": 2.7566, | |
| "step": 172500 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "learning_rate": 7.284645754551967e-06, | |
| "loss": 2.7601, | |
| "step": 173000 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "learning_rate": 7.2763164470505934e-06, | |
| "loss": 2.7608, | |
| "step": 173500 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "learning_rate": 7.2679871395492184e-06, | |
| "loss": 2.7563, | |
| "step": 174000 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "learning_rate": 7.259657832047844e-06, | |
| "loss": 2.7535, | |
| "step": 174500 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "learning_rate": 7.251328524546469e-06, | |
| "loss": 2.7507, | |
| "step": 175000 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "learning_rate": 7.242999217045096e-06, | |
| "loss": 2.7577, | |
| "step": 175500 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "learning_rate": 7.234669909543721e-06, | |
| "loss": 2.7485, | |
| "step": 176000 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "learning_rate": 7.226340602042347e-06, | |
| "loss": 2.7501, | |
| "step": 176500 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "learning_rate": 7.218011294540972e-06, | |
| "loss": 2.7445, | |
| "step": 177000 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "learning_rate": 7.2096819870395985e-06, | |
| "loss": 2.7482, | |
| "step": 177500 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "learning_rate": 7.2013526795382235e-06, | |
| "loss": 2.7488, | |
| "step": 178000 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "learning_rate": 7.193023372036849e-06, | |
| "loss": 2.7501, | |
| "step": 178500 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "learning_rate": 7.184694064535474e-06, | |
| "loss": 2.741, | |
| "step": 179000 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "learning_rate": 7.176364757034101e-06, | |
| "loss": 2.7371, | |
| "step": 179500 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "learning_rate": 7.168035449532726e-06, | |
| "loss": 2.7403, | |
| "step": 180000 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "eval_accuracy": 0.5280822410313738, | |
| "eval_loss": 2.6168134212493896, | |
| "eval_runtime": 1882.9339, | |
| "eval_samples_per_second": 163.764, | |
| "eval_steps_per_second": 3.412, | |
| "step": 180000 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "learning_rate": 7.159706142031352e-06, | |
| "loss": 2.7343, | |
| "step": 180500 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "learning_rate": 7.151376834529977e-06, | |
| "loss": 2.7362, | |
| "step": 181000 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "learning_rate": 7.143047527028604e-06, | |
| "loss": 2.735, | |
| "step": 181500 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "learning_rate": 7.1347182195272295e-06, | |
| "loss": 2.7257, | |
| "step": 182000 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "learning_rate": 7.1263889120258545e-06, | |
| "loss": 2.7271, | |
| "step": 182500 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "learning_rate": 7.118059604524481e-06, | |
| "loss": 2.7285, | |
| "step": 183000 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "learning_rate": 7.109730297023106e-06, | |
| "loss": 2.7347, | |
| "step": 183500 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "learning_rate": 7.101400989521732e-06, | |
| "loss": 2.7286, | |
| "step": 184000 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "learning_rate": 7.093071682020357e-06, | |
| "loss": 2.733, | |
| "step": 184500 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "learning_rate": 7.084742374518984e-06, | |
| "loss": 2.7304, | |
| "step": 185000 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "learning_rate": 7.076413067017609e-06, | |
| "loss": 2.7234, | |
| "step": 185500 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "learning_rate": 7.068083759516235e-06, | |
| "loss": 2.7288, | |
| "step": 186000 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "learning_rate": 7.05975445201486e-06, | |
| "loss": 2.725, | |
| "step": 186500 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "learning_rate": 7.051425144513486e-06, | |
| "loss": 2.7207, | |
| "step": 187000 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "learning_rate": 7.043095837012111e-06, | |
| "loss": 2.7142, | |
| "step": 187500 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "learning_rate": 7.034766529510737e-06, | |
| "loss": 2.7263, | |
| "step": 188000 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "learning_rate": 7.026437222009362e-06, | |
| "loss": 2.7128, | |
| "step": 188500 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "step": 188996, | |
| "total_flos": 2.6878539060792525e+18, | |
| "train_loss": 3.806450256516879, | |
| "train_runtime": 197979.3754, | |
| "train_samples_per_second": 147.964, | |
| "train_steps_per_second": 3.083 | |
| } | |
| ], | |
| "max_steps": 610290, | |
| "num_train_epochs": 5, | |
| "total_flos": 2.6878539060792525e+18, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |