| { | |
| "best_metric": 0.49101120233535767, | |
| "best_model_checkpoint": "/home/jcanete/ft-data/all_results/xnli/bertin_roberta_base/epochs_3_bs_16_lr_1e-5/checkpoint-22000", | |
| "epoch": 3.0, | |
| "global_step": 73632, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 9.932366362451109e-06, | |
| "loss": 1.0444, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 9.864461103867885e-06, | |
| "loss": 0.8226, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 9.79655584528466e-06, | |
| "loss": 0.7505, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 9.728650586701435e-06, | |
| "loss": 0.7012, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "eval_accuracy": 0.7076305220883534, | |
| "eval_loss": 0.715941309928894, | |
| "eval_runtime": 1.744, | |
| "eval_samples_per_second": 1427.79, | |
| "eval_steps_per_second": 89.452, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 9.660881138635376e-06, | |
| "loss": 0.6765, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 9.592975880052153e-06, | |
| "loss": 0.6657, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 9.525070621468926e-06, | |
| "loss": 0.6506, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 9.457165362885704e-06, | |
| "loss": 0.6259, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "eval_accuracy": 0.7538152610441767, | |
| "eval_loss": 0.6095600724220276, | |
| "eval_runtime": 1.7428, | |
| "eval_samples_per_second": 1428.771, | |
| "eval_steps_per_second": 89.513, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 9.389260104302477e-06, | |
| "loss": 0.6367, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 9.321354845719254e-06, | |
| "loss": 0.6355, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 9.253449587136028e-06, | |
| "loss": 0.6289, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 9.185544328552804e-06, | |
| "loss": 0.6059, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "eval_accuracy": 0.7526104417670683, | |
| "eval_loss": 0.6048593521118164, | |
| "eval_runtime": 1.7395, | |
| "eval_samples_per_second": 1431.463, | |
| "eval_steps_per_second": 89.682, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 9.117639069969579e-06, | |
| "loss": 0.608, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 9.049869621903521e-06, | |
| "loss": 0.6132, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 8.981964363320296e-06, | |
| "loss": 0.6056, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 8.914059104737072e-06, | |
| "loss": 0.5869, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "eval_accuracy": 0.765863453815261, | |
| "eval_loss": 0.5669301152229309, | |
| "eval_runtime": 1.7452, | |
| "eval_samples_per_second": 1426.778, | |
| "eval_steps_per_second": 89.388, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 8.846153846153847e-06, | |
| "loss": 0.5859, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 8.778384398087787e-06, | |
| "loss": 0.5757, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 8.710479139504565e-06, | |
| "loss": 0.5837, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 8.642573880921338e-06, | |
| "loss": 0.5783, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "eval_accuracy": 0.7746987951807229, | |
| "eval_loss": 0.5589046478271484, | |
| "eval_runtime": 1.7403, | |
| "eval_samples_per_second": 1430.765, | |
| "eval_steps_per_second": 89.638, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 8.574668622338115e-06, | |
| "loss": 0.5792, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 8.506763363754889e-06, | |
| "loss": 0.5693, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 8.438858105171666e-06, | |
| "loss": 0.5649, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 8.371088657105606e-06, | |
| "loss": 0.5557, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "eval_accuracy": 0.7819277108433735, | |
| "eval_loss": 0.5610175132751465, | |
| "eval_runtime": 1.7411, | |
| "eval_samples_per_second": 1430.157, | |
| "eval_steps_per_second": 89.6, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 8.303183398522382e-06, | |
| "loss": 0.5548, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 8.235278139939157e-06, | |
| "loss": 0.5766, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 8.167372881355933e-06, | |
| "loss": 0.5551, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 8.099467622772708e-06, | |
| "loss": 0.5614, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "eval_accuracy": 0.7819277108433735, | |
| "eval_loss": 0.5594050884246826, | |
| "eval_runtime": 1.7412, | |
| "eval_samples_per_second": 1430.056, | |
| "eval_steps_per_second": 89.594, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 8.031562364189484e-06, | |
| "loss": 0.5486, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 7.963657105606259e-06, | |
| "loss": 0.5678, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 7.895751847023034e-06, | |
| "loss": 0.5458, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 7.827982398956976e-06, | |
| "loss": 0.5492, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "eval_accuracy": 0.7927710843373494, | |
| "eval_loss": 0.5257371068000793, | |
| "eval_runtime": 1.7251, | |
| "eval_samples_per_second": 1443.436, | |
| "eval_steps_per_second": 90.432, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 7.76007714037375e-06, | |
| "loss": 0.5466, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 7.692171881790527e-06, | |
| "loss": 0.5364, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 7.624266623207302e-06, | |
| "loss": 0.5526, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 7.556497175141243e-06, | |
| "loss": 0.526, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "eval_accuracy": 0.7730923694779116, | |
| "eval_loss": 0.5546149611473083, | |
| "eval_runtime": 1.7312, | |
| "eval_samples_per_second": 1438.268, | |
| "eval_steps_per_second": 90.108, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 7.488591916558019e-06, | |
| "loss": 0.5278, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 7.420686657974794e-06, | |
| "loss": 0.5469, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 7.35278139939157e-06, | |
| "loss": 0.5406, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 7.2848761408083445e-06, | |
| "loss": 0.5241, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "eval_accuracy": 0.7975903614457831, | |
| "eval_loss": 0.5097489953041077, | |
| "eval_runtime": 1.7046, | |
| "eval_samples_per_second": 1460.741, | |
| "eval_steps_per_second": 91.516, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 7.216970882225121e-06, | |
| "loss": 0.5384, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 7.149065623641895e-06, | |
| "loss": 0.5353, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 7.08116036505867e-06, | |
| "loss": 0.5194, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 7.013255106475446e-06, | |
| "loss": 0.5358, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "eval_accuracy": 0.8064257028112449, | |
| "eval_loss": 0.49101120233535767, | |
| "eval_runtime": 1.7316, | |
| "eval_samples_per_second": 1437.957, | |
| "eval_steps_per_second": 90.089, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 6.945485658409388e-06, | |
| "loss": 0.525, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 6.877580399826163e-06, | |
| "loss": 0.5274, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 6.809675141242938e-06, | |
| "loss": 0.5208, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 6.7417698826597135e-06, | |
| "loss": 0.5327, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "eval_accuracy": 0.7899598393574297, | |
| "eval_loss": 0.5255107283592224, | |
| "eval_runtime": 1.7081, | |
| "eval_samples_per_second": 1457.767, | |
| "eval_steps_per_second": 91.33, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 6.674000434593656e-06, | |
| "loss": 0.5227, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "learning_rate": 6.606095176010431e-06, | |
| "loss": 0.3802, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "learning_rate": 6.538189917427205e-06, | |
| "loss": 0.3898, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "learning_rate": 6.470420469361148e-06, | |
| "loss": 0.3598, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "eval_accuracy": 0.797991967871486, | |
| "eval_loss": 0.5192009210586548, | |
| "eval_runtime": 1.7342, | |
| "eval_samples_per_second": 1435.797, | |
| "eval_steps_per_second": 89.954, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "learning_rate": 6.4025152107779235e-06, | |
| "loss": 0.3638, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "learning_rate": 6.334609952194698e-06, | |
| "loss": 0.3636, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "learning_rate": 6.2667046936114736e-06, | |
| "loss": 0.3698, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "learning_rate": 6.198799435028249e-06, | |
| "loss": 0.3635, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "eval_accuracy": 0.8076305220883534, | |
| "eval_loss": 0.5339061617851257, | |
| "eval_runtime": 1.707, | |
| "eval_samples_per_second": 1458.725, | |
| "eval_steps_per_second": 91.39, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "learning_rate": 6.13102998696219e-06, | |
| "loss": 0.3472, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "learning_rate": 6.063124728378966e-06, | |
| "loss": 0.3599, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "learning_rate": 5.995219469795741e-06, | |
| "loss": 0.3521, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "learning_rate": 5.927314211212517e-06, | |
| "loss": 0.3633, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "eval_accuracy": 0.8040160642570281, | |
| "eval_loss": 0.5154483914375305, | |
| "eval_runtime": 1.7227, | |
| "eval_samples_per_second": 1445.384, | |
| "eval_steps_per_second": 90.554, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "learning_rate": 5.859408952629292e-06, | |
| "loss": 0.3621, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "learning_rate": 5.791503694046068e-06, | |
| "loss": 0.3604, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "learning_rate": 5.723734245980009e-06, | |
| "loss": 0.355, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "learning_rate": 5.6558289873967845e-06, | |
| "loss": 0.3586, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "eval_accuracy": 0.7975903614457831, | |
| "eval_loss": 0.5583391785621643, | |
| "eval_runtime": 1.7322, | |
| "eval_samples_per_second": 1437.489, | |
| "eval_steps_per_second": 90.06, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "learning_rate": 5.58792372881356e-06, | |
| "loss": 0.3527, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "learning_rate": 5.520018470230335e-06, | |
| "loss": 0.3644, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "learning_rate": 5.452249022164276e-06, | |
| "loss": 0.3769, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "learning_rate": 5.384343763581053e-06, | |
| "loss": 0.3614, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "eval_accuracy": 0.8056224899598393, | |
| "eval_loss": 0.5339176058769226, | |
| "eval_runtime": 1.706, | |
| "eval_samples_per_second": 1459.585, | |
| "eval_steps_per_second": 91.444, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "learning_rate": 5.316438504997827e-06, | |
| "loss": 0.3577, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "learning_rate": 5.2485332464146035e-06, | |
| "loss": 0.3593, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "learning_rate": 5.1807637983485445e-06, | |
| "loss": 0.3475, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "learning_rate": 5.11285853976532e-06, | |
| "loss": 0.3631, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "eval_accuracy": 0.8076305220883534, | |
| "eval_loss": 0.5045333504676819, | |
| "eval_runtime": 1.7273, | |
| "eval_samples_per_second": 1441.576, | |
| "eval_steps_per_second": 90.316, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "learning_rate": 5.044953281182095e-06, | |
| "loss": 0.3523, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "learning_rate": 4.977048022598871e-06, | |
| "loss": 0.3507, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "learning_rate": 4.909142764015646e-06, | |
| "loss": 0.3602, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "learning_rate": 4.841373315949587e-06, | |
| "loss": 0.3542, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "eval_accuracy": 0.8, | |
| "eval_loss": 0.5346028804779053, | |
| "eval_runtime": 1.7369, | |
| "eval_samples_per_second": 1433.604, | |
| "eval_steps_per_second": 89.816, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "learning_rate": 4.773468057366363e-06, | |
| "loss": 0.3527, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "learning_rate": 4.705562798783138e-06, | |
| "loss": 0.3753, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "learning_rate": 4.6376575401999135e-06, | |
| "loss": 0.3655, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "learning_rate": 4.569752281616689e-06, | |
| "loss": 0.3564, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "eval_accuracy": 0.8008032128514057, | |
| "eval_loss": 0.5383270382881165, | |
| "eval_runtime": 1.7313, | |
| "eval_samples_per_second": 1438.187, | |
| "eval_steps_per_second": 90.103, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "learning_rate": 4.501982833550631e-06, | |
| "loss": 0.3574, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "learning_rate": 4.434077574967406e-06, | |
| "loss": 0.3566, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 1.69, | |
| "learning_rate": 4.366172316384181e-06, | |
| "loss": 0.3629, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "learning_rate": 4.298267057800956e-06, | |
| "loss": 0.3508, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "eval_accuracy": 0.7987951807228916, | |
| "eval_loss": 0.5298879742622375, | |
| "eval_runtime": 1.7287, | |
| "eval_samples_per_second": 1440.391, | |
| "eval_steps_per_second": 90.241, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "learning_rate": 4.230497609734898e-06, | |
| "loss": 0.3587, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "learning_rate": 4.162592351151674e-06, | |
| "loss": 0.3622, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "learning_rate": 4.094687092568449e-06, | |
| "loss": 0.3599, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "learning_rate": 4.0267818339852245e-06, | |
| "loss": 0.354, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "eval_accuracy": 0.7927710843373494, | |
| "eval_loss": 0.5385558605194092, | |
| "eval_runtime": 1.733, | |
| "eval_samples_per_second": 1436.852, | |
| "eval_steps_per_second": 90.02, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "learning_rate": 3.958876575401999e-06, | |
| "loss": 0.3661, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "learning_rate": 3.891107127335941e-06, | |
| "loss": 0.3614, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "learning_rate": 3.823201868752716e-06, | |
| "loss": 0.3561, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "learning_rate": 3.755296610169492e-06, | |
| "loss": 0.3622, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "eval_accuracy": 0.8100401606425702, | |
| "eval_loss": 0.5162017941474915, | |
| "eval_runtime": 1.7327, | |
| "eval_samples_per_second": 1437.039, | |
| "eval_steps_per_second": 90.031, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "learning_rate": 3.6873913515862668e-06, | |
| "loss": 0.3472, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "learning_rate": 3.619621903520209e-06, | |
| "loss": 0.3381, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "learning_rate": 3.5517166449369845e-06, | |
| "loss": 0.3443, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "learning_rate": 3.483811386353759e-06, | |
| "loss": 0.3559, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "eval_accuracy": 0.8060240963855422, | |
| "eval_loss": 0.513914942741394, | |
| "eval_runtime": 1.7383, | |
| "eval_samples_per_second": 1432.408, | |
| "eval_steps_per_second": 89.741, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "learning_rate": 3.4159061277705345e-06, | |
| "loss": 0.3518, | |
| "step": 48500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 3.34800086918731e-06, | |
| "loss": 0.3554, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "learning_rate": 3.2803672316384183e-06, | |
| "loss": 0.2133, | |
| "step": 49500 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "learning_rate": 3.2124619730551937e-06, | |
| "loss": 0.1896, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "eval_accuracy": 0.7947791164658634, | |
| "eval_loss": 0.7987900972366333, | |
| "eval_runtime": 1.7334, | |
| "eval_samples_per_second": 1436.451, | |
| "eval_steps_per_second": 89.995, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 2.06, | |
| "learning_rate": 3.144556714471969e-06, | |
| "loss": 0.1934, | |
| "step": 50500 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "learning_rate": 3.0766514558887446e-06, | |
| "loss": 0.1986, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "learning_rate": 3.008882007822686e-06, | |
| "loss": 0.1973, | |
| "step": 51500 | |
| }, | |
| { | |
| "epoch": 2.12, | |
| "learning_rate": 2.9409767492394614e-06, | |
| "loss": 0.1867, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 2.12, | |
| "eval_accuracy": 0.7987951807228916, | |
| "eval_loss": 0.8654219508171082, | |
| "eval_runtime": 1.7101, | |
| "eval_samples_per_second": 1456.025, | |
| "eval_steps_per_second": 91.221, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 2.14, | |
| "learning_rate": 2.873071490656237e-06, | |
| "loss": 0.2036, | |
| "step": 52500 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "learning_rate": 2.8051662320730123e-06, | |
| "loss": 0.1956, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 2.18, | |
| "learning_rate": 2.737260973489787e-06, | |
| "loss": 0.1934, | |
| "step": 53500 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "learning_rate": 2.669491525423729e-06, | |
| "loss": 0.1945, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "eval_accuracy": 0.804417670682731, | |
| "eval_loss": 0.8309808373451233, | |
| "eval_runtime": 1.7106, | |
| "eval_samples_per_second": 1455.613, | |
| "eval_steps_per_second": 91.195, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "learning_rate": 2.6015862668405046e-06, | |
| "loss": 0.1885, | |
| "step": 54500 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "learning_rate": 2.53368100825728e-06, | |
| "loss": 0.1949, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 2.26, | |
| "learning_rate": 2.465775749674055e-06, | |
| "loss": 0.2041, | |
| "step": 55500 | |
| }, | |
| { | |
| "epoch": 2.28, | |
| "learning_rate": 2.39787049109083e-06, | |
| "loss": 0.2005, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 2.28, | |
| "eval_accuracy": 0.7899598393574297, | |
| "eval_loss": 0.9045226573944092, | |
| "eval_runtime": 1.7241, | |
| "eval_samples_per_second": 1444.229, | |
| "eval_steps_per_second": 90.482, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "learning_rate": 2.3299652325076055e-06, | |
| "loss": 0.1869, | |
| "step": 56500 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "learning_rate": 2.262059973924381e-06, | |
| "loss": 0.1704, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 2.34, | |
| "learning_rate": 2.1942905258583228e-06, | |
| "loss": 0.197, | |
| "step": 57500 | |
| }, | |
| { | |
| "epoch": 2.36, | |
| "learning_rate": 2.1263852672750978e-06, | |
| "loss": 0.1879, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 2.36, | |
| "eval_accuracy": 0.8036144578313253, | |
| "eval_loss": 0.8717219233512878, | |
| "eval_runtime": 1.7314, | |
| "eval_samples_per_second": 1438.131, | |
| "eval_steps_per_second": 90.1, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 2.38, | |
| "learning_rate": 2.058480008691873e-06, | |
| "loss": 0.1924, | |
| "step": 58500 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "learning_rate": 1.9905747501086486e-06, | |
| "loss": 0.2094, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 2.42, | |
| "learning_rate": 1.9226694915254236e-06, | |
| "loss": 0.1904, | |
| "step": 59500 | |
| }, | |
| { | |
| "epoch": 2.44, | |
| "learning_rate": 1.8549000434593655e-06, | |
| "loss": 0.2007, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 2.44, | |
| "eval_accuracy": 0.7995983935742972, | |
| "eval_loss": 0.8504888415336609, | |
| "eval_runtime": 1.7179, | |
| "eval_samples_per_second": 1449.471, | |
| "eval_steps_per_second": 90.81, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 2.46, | |
| "learning_rate": 1.786994784876141e-06, | |
| "loss": 0.1967, | |
| "step": 60500 | |
| }, | |
| { | |
| "epoch": 2.49, | |
| "learning_rate": 1.7190895262929164e-06, | |
| "loss": 0.1954, | |
| "step": 61000 | |
| }, | |
| { | |
| "epoch": 2.51, | |
| "learning_rate": 1.6511842677096914e-06, | |
| "loss": 0.1831, | |
| "step": 61500 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "learning_rate": 1.5832790091264668e-06, | |
| "loss": 0.1861, | |
| "step": 62000 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "eval_accuracy": 0.7983935742971887, | |
| "eval_loss": 0.8534940481185913, | |
| "eval_runtime": 1.7626, | |
| "eval_samples_per_second": 1412.7, | |
| "eval_steps_per_second": 88.507, | |
| "step": 62000 | |
| }, | |
| { | |
| "epoch": 2.55, | |
| "learning_rate": 1.5153737505432422e-06, | |
| "loss": 0.1869, | |
| "step": 62500 | |
| }, | |
| { | |
| "epoch": 2.57, | |
| "learning_rate": 1.4476043024771841e-06, | |
| "loss": 0.1767, | |
| "step": 63000 | |
| }, | |
| { | |
| "epoch": 2.59, | |
| "learning_rate": 1.3796990438939591e-06, | |
| "loss": 0.1902, | |
| "step": 63500 | |
| }, | |
| { | |
| "epoch": 2.61, | |
| "learning_rate": 1.3117937853107346e-06, | |
| "loss": 0.1836, | |
| "step": 64000 | |
| }, | |
| { | |
| "epoch": 2.61, | |
| "eval_accuracy": 0.8, | |
| "eval_loss": 0.8850274085998535, | |
| "eval_runtime": 1.7307, | |
| "eval_samples_per_second": 1438.73, | |
| "eval_steps_per_second": 90.137, | |
| "step": 64000 | |
| }, | |
| { | |
| "epoch": 2.63, | |
| "learning_rate": 1.2438885267275098e-06, | |
| "loss": 0.1765, | |
| "step": 64500 | |
| }, | |
| { | |
| "epoch": 2.65, | |
| "learning_rate": 1.1761190786614516e-06, | |
| "loss": 0.1877, | |
| "step": 65000 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "learning_rate": 1.1082138200782269e-06, | |
| "loss": 0.189, | |
| "step": 65500 | |
| }, | |
| { | |
| "epoch": 2.69, | |
| "learning_rate": 1.0403085614950023e-06, | |
| "loss": 0.2009, | |
| "step": 66000 | |
| }, | |
| { | |
| "epoch": 2.69, | |
| "eval_accuracy": 0.8, | |
| "eval_loss": 0.8615464568138123, | |
| "eval_runtime": 1.7274, | |
| "eval_samples_per_second": 1441.493, | |
| "eval_steps_per_second": 90.31, | |
| "step": 66000 | |
| }, | |
| { | |
| "epoch": 2.71, | |
| "learning_rate": 9.724033029117775e-07, | |
| "loss": 0.1853, | |
| "step": 66500 | |
| }, | |
| { | |
| "epoch": 2.73, | |
| "learning_rate": 9.044980443285528e-07, | |
| "loss": 0.1793, | |
| "step": 67000 | |
| }, | |
| { | |
| "epoch": 2.75, | |
| "learning_rate": 8.365927857453282e-07, | |
| "loss": 0.1938, | |
| "step": 67500 | |
| }, | |
| { | |
| "epoch": 2.77, | |
| "learning_rate": 7.686875271621035e-07, | |
| "loss": 0.1809, | |
| "step": 68000 | |
| }, | |
| { | |
| "epoch": 2.77, | |
| "eval_accuracy": 0.7991967871485943, | |
| "eval_loss": 0.9009087085723877, | |
| "eval_runtime": 1.7248, | |
| "eval_samples_per_second": 1443.619, | |
| "eval_steps_per_second": 90.444, | |
| "step": 68000 | |
| }, | |
| { | |
| "epoch": 2.79, | |
| "learning_rate": 7.009180790960452e-07, | |
| "loss": 0.1958, | |
| "step": 68500 | |
| }, | |
| { | |
| "epoch": 2.81, | |
| "learning_rate": 6.330128205128206e-07, | |
| "loss": 0.1866, | |
| "step": 69000 | |
| }, | |
| { | |
| "epoch": 2.83, | |
| "learning_rate": 5.651075619295959e-07, | |
| "loss": 0.189, | |
| "step": 69500 | |
| }, | |
| { | |
| "epoch": 2.85, | |
| "learning_rate": 4.972023033463712e-07, | |
| "loss": 0.1885, | |
| "step": 70000 | |
| }, | |
| { | |
| "epoch": 2.85, | |
| "eval_accuracy": 0.8064257028112449, | |
| "eval_loss": 0.89076167345047, | |
| "eval_runtime": 1.7155, | |
| "eval_samples_per_second": 1451.454, | |
| "eval_steps_per_second": 90.934, | |
| "step": 70000 | |
| }, | |
| { | |
| "epoch": 2.87, | |
| "learning_rate": 4.292970447631465e-07, | |
| "loss": 0.1845, | |
| "step": 70500 | |
| }, | |
| { | |
| "epoch": 2.89, | |
| "learning_rate": 3.613917861799218e-07, | |
| "loss": 0.1844, | |
| "step": 71000 | |
| }, | |
| { | |
| "epoch": 2.91, | |
| "learning_rate": 2.936223381138636e-07, | |
| "loss": 0.1919, | |
| "step": 71500 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "learning_rate": 2.2571707953063886e-07, | |
| "loss": 0.1941, | |
| "step": 72000 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "eval_accuracy": 0.8036144578313253, | |
| "eval_loss": 0.8937243819236755, | |
| "eval_runtime": 1.7336, | |
| "eval_samples_per_second": 1436.277, | |
| "eval_steps_per_second": 89.984, | |
| "step": 72000 | |
| }, | |
| { | |
| "epoch": 2.95, | |
| "learning_rate": 1.5781182094741418e-07, | |
| "loss": 0.1873, | |
| "step": 72500 | |
| }, | |
| { | |
| "epoch": 2.97, | |
| "learning_rate": 8.990656236418949e-08, | |
| "loss": 0.1923, | |
| "step": 73000 | |
| }, | |
| { | |
| "epoch": 2.99, | |
| "learning_rate": 2.2001303780964798e-08, | |
| "loss": 0.1831, | |
| "step": 73500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 73632, | |
| "total_flos": 5.651008562499955e+16, | |
| "train_loss": 0.38008860063159117, | |
| "train_runtime": 3735.6095, | |
| "train_samples_per_second": 315.372, | |
| "train_steps_per_second": 19.711 | |
| } | |
| ], | |
| "max_steps": 73632, | |
| "num_train_epochs": 3, | |
| "total_flos": 5.651008562499955e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |