|
{ |
|
"best_metric": 0.8227848101265823, |
|
"best_model_checkpoint": "beit-base-patch16-224-fold5/checkpoint-126", |
|
"epoch": 85.71428571428571, |
|
"eval_steps": 500, |
|
"global_step": 300, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 0.8571428571428571, |
|
"eval_accuracy": 0.4936708860759494, |
|
"eval_loss": 0.6963750720024109, |
|
"eval_runtime": 1.3397, |
|
"eval_samples_per_second": 58.971, |
|
"eval_steps_per_second": 2.239, |
|
"step": 3 |
|
}, |
|
{ |
|
"epoch": 2.0, |
|
"eval_accuracy": 0.6455696202531646, |
|
"eval_loss": 0.6335641145706177, |
|
"eval_runtime": 1.2784, |
|
"eval_samples_per_second": 61.795, |
|
"eval_steps_per_second": 2.347, |
|
"step": 7 |
|
}, |
|
{ |
|
"epoch": 2.857142857142857, |
|
"grad_norm": 10.374295234680176, |
|
"learning_rate": 1.6666666666666667e-05, |
|
"loss": 0.7161, |
|
"step": 10 |
|
}, |
|
{ |
|
"epoch": 2.857142857142857, |
|
"eval_accuracy": 0.5063291139240507, |
|
"eval_loss": 0.6711614727973938, |
|
"eval_runtime": 1.2465, |
|
"eval_samples_per_second": 63.375, |
|
"eval_steps_per_second": 2.407, |
|
"step": 10 |
|
}, |
|
{ |
|
"epoch": 4.0, |
|
"eval_accuracy": 0.6835443037974683, |
|
"eval_loss": 0.5884170532226562, |
|
"eval_runtime": 1.2485, |
|
"eval_samples_per_second": 63.276, |
|
"eval_steps_per_second": 2.403, |
|
"step": 14 |
|
}, |
|
{ |
|
"epoch": 4.857142857142857, |
|
"eval_accuracy": 0.5569620253164557, |
|
"eval_loss": 0.6325361728668213, |
|
"eval_runtime": 1.2645, |
|
"eval_samples_per_second": 62.477, |
|
"eval_steps_per_second": 2.373, |
|
"step": 17 |
|
}, |
|
{ |
|
"epoch": 5.714285714285714, |
|
"grad_norm": 6.489544868469238, |
|
"learning_rate": 3.3333333333333335e-05, |
|
"loss": 0.6466, |
|
"step": 20 |
|
}, |
|
{ |
|
"epoch": 6.0, |
|
"eval_accuracy": 0.6075949367088608, |
|
"eval_loss": 0.6306931376457214, |
|
"eval_runtime": 1.2522, |
|
"eval_samples_per_second": 63.087, |
|
"eval_steps_per_second": 2.396, |
|
"step": 21 |
|
}, |
|
{ |
|
"epoch": 6.857142857142857, |
|
"eval_accuracy": 0.6835443037974683, |
|
"eval_loss": 0.5518103837966919, |
|
"eval_runtime": 1.2518, |
|
"eval_samples_per_second": 63.11, |
|
"eval_steps_per_second": 2.397, |
|
"step": 24 |
|
}, |
|
{ |
|
"epoch": 8.0, |
|
"eval_accuracy": 0.6455696202531646, |
|
"eval_loss": 0.6855829358100891, |
|
"eval_runtime": 1.2601, |
|
"eval_samples_per_second": 62.693, |
|
"eval_steps_per_second": 2.381, |
|
"step": 28 |
|
}, |
|
{ |
|
"epoch": 8.571428571428571, |
|
"grad_norm": 15.54662799835205, |
|
"learning_rate": 5e-05, |
|
"loss": 0.5604, |
|
"step": 30 |
|
}, |
|
{ |
|
"epoch": 8.857142857142858, |
|
"eval_accuracy": 0.7088607594936709, |
|
"eval_loss": 0.5678081512451172, |
|
"eval_runtime": 1.2556, |
|
"eval_samples_per_second": 62.917, |
|
"eval_steps_per_second": 2.389, |
|
"step": 31 |
|
}, |
|
{ |
|
"epoch": 10.0, |
|
"eval_accuracy": 0.6329113924050633, |
|
"eval_loss": 0.5764104127883911, |
|
"eval_runtime": 1.2443, |
|
"eval_samples_per_second": 63.491, |
|
"eval_steps_per_second": 2.411, |
|
"step": 35 |
|
}, |
|
{ |
|
"epoch": 10.857142857142858, |
|
"eval_accuracy": 0.6455696202531646, |
|
"eval_loss": 0.6123582124710083, |
|
"eval_runtime": 1.2354, |
|
"eval_samples_per_second": 63.945, |
|
"eval_steps_per_second": 2.428, |
|
"step": 38 |
|
}, |
|
{ |
|
"epoch": 11.428571428571429, |
|
"grad_norm": 7.390775203704834, |
|
"learning_rate": 4.814814814814815e-05, |
|
"loss": 0.4941, |
|
"step": 40 |
|
}, |
|
{ |
|
"epoch": 12.0, |
|
"eval_accuracy": 0.5316455696202531, |
|
"eval_loss": 1.131524920463562, |
|
"eval_runtime": 1.2833, |
|
"eval_samples_per_second": 61.562, |
|
"eval_steps_per_second": 2.338, |
|
"step": 42 |
|
}, |
|
{ |
|
"epoch": 12.857142857142858, |
|
"eval_accuracy": 0.6835443037974683, |
|
"eval_loss": 0.5624261498451233, |
|
"eval_runtime": 1.2579, |
|
"eval_samples_per_second": 62.803, |
|
"eval_steps_per_second": 2.385, |
|
"step": 45 |
|
}, |
|
{ |
|
"epoch": 14.0, |
|
"eval_accuracy": 0.6962025316455697, |
|
"eval_loss": 0.6350968480110168, |
|
"eval_runtime": 1.2442, |
|
"eval_samples_per_second": 63.495, |
|
"eval_steps_per_second": 2.411, |
|
"step": 49 |
|
}, |
|
{ |
|
"epoch": 14.285714285714286, |
|
"grad_norm": 13.477153778076172, |
|
"learning_rate": 4.62962962962963e-05, |
|
"loss": 0.4516, |
|
"step": 50 |
|
}, |
|
{ |
|
"epoch": 14.857142857142858, |
|
"eval_accuracy": 0.6835443037974683, |
|
"eval_loss": 0.6810511350631714, |
|
"eval_runtime": 1.2437, |
|
"eval_samples_per_second": 63.519, |
|
"eval_steps_per_second": 2.412, |
|
"step": 52 |
|
}, |
|
{ |
|
"epoch": 16.0, |
|
"eval_accuracy": 0.6455696202531646, |
|
"eval_loss": 0.7385033965110779, |
|
"eval_runtime": 1.2445, |
|
"eval_samples_per_second": 63.478, |
|
"eval_steps_per_second": 2.411, |
|
"step": 56 |
|
}, |
|
{ |
|
"epoch": 16.857142857142858, |
|
"eval_accuracy": 0.7215189873417721, |
|
"eval_loss": 0.6122547388076782, |
|
"eval_runtime": 1.2613, |
|
"eval_samples_per_second": 62.632, |
|
"eval_steps_per_second": 2.378, |
|
"step": 59 |
|
}, |
|
{ |
|
"epoch": 17.142857142857142, |
|
"grad_norm": 5.328888893127441, |
|
"learning_rate": 4.4444444444444447e-05, |
|
"loss": 0.409, |
|
"step": 60 |
|
}, |
|
{ |
|
"epoch": 18.0, |
|
"eval_accuracy": 0.6835443037974683, |
|
"eval_loss": 0.587742805480957, |
|
"eval_runtime": 1.2428, |
|
"eval_samples_per_second": 63.564, |
|
"eval_steps_per_second": 2.414, |
|
"step": 63 |
|
}, |
|
{ |
|
"epoch": 18.857142857142858, |
|
"eval_accuracy": 0.6962025316455697, |
|
"eval_loss": 0.6250303387641907, |
|
"eval_runtime": 1.2444, |
|
"eval_samples_per_second": 63.486, |
|
"eval_steps_per_second": 2.411, |
|
"step": 66 |
|
}, |
|
{ |
|
"epoch": 20.0, |
|
"grad_norm": 7.51099157333374, |
|
"learning_rate": 4.259259259259259e-05, |
|
"loss": 0.3439, |
|
"step": 70 |
|
}, |
|
{ |
|
"epoch": 20.0, |
|
"eval_accuracy": 0.7341772151898734, |
|
"eval_loss": 0.6220384240150452, |
|
"eval_runtime": 1.2611, |
|
"eval_samples_per_second": 62.645, |
|
"eval_steps_per_second": 2.379, |
|
"step": 70 |
|
}, |
|
{ |
|
"epoch": 20.857142857142858, |
|
"eval_accuracy": 0.7088607594936709, |
|
"eval_loss": 0.724984347820282, |
|
"eval_runtime": 1.237, |
|
"eval_samples_per_second": 63.863, |
|
"eval_steps_per_second": 2.425, |
|
"step": 73 |
|
}, |
|
{ |
|
"epoch": 22.0, |
|
"eval_accuracy": 0.7341772151898734, |
|
"eval_loss": 0.6107321381568909, |
|
"eval_runtime": 1.2567, |
|
"eval_samples_per_second": 62.863, |
|
"eval_steps_per_second": 2.387, |
|
"step": 77 |
|
}, |
|
{ |
|
"epoch": 22.857142857142858, |
|
"grad_norm": 6.396337032318115, |
|
"learning_rate": 4.074074074074074e-05, |
|
"loss": 0.3268, |
|
"step": 80 |
|
}, |
|
{ |
|
"epoch": 22.857142857142858, |
|
"eval_accuracy": 0.7974683544303798, |
|
"eval_loss": 0.4999098479747772, |
|
"eval_runtime": 1.2544, |
|
"eval_samples_per_second": 62.978, |
|
"eval_steps_per_second": 2.392, |
|
"step": 80 |
|
}, |
|
{ |
|
"epoch": 24.0, |
|
"eval_accuracy": 0.7468354430379747, |
|
"eval_loss": 0.5325254201889038, |
|
"eval_runtime": 1.2825, |
|
"eval_samples_per_second": 61.597, |
|
"eval_steps_per_second": 2.339, |
|
"step": 84 |
|
}, |
|
{ |
|
"epoch": 24.857142857142858, |
|
"eval_accuracy": 0.7341772151898734, |
|
"eval_loss": 0.7209397554397583, |
|
"eval_runtime": 1.2439, |
|
"eval_samples_per_second": 63.508, |
|
"eval_steps_per_second": 2.412, |
|
"step": 87 |
|
}, |
|
{ |
|
"epoch": 25.714285714285715, |
|
"grad_norm": 4.736355781555176, |
|
"learning_rate": 3.888888888888889e-05, |
|
"loss": 0.2941, |
|
"step": 90 |
|
}, |
|
{ |
|
"epoch": 26.0, |
|
"eval_accuracy": 0.7721518987341772, |
|
"eval_loss": 0.5556657314300537, |
|
"eval_runtime": 1.2702, |
|
"eval_samples_per_second": 62.195, |
|
"eval_steps_per_second": 2.362, |
|
"step": 91 |
|
}, |
|
{ |
|
"epoch": 26.857142857142858, |
|
"eval_accuracy": 0.759493670886076, |
|
"eval_loss": 0.6655121445655823, |
|
"eval_runtime": 1.2506, |
|
"eval_samples_per_second": 63.172, |
|
"eval_steps_per_second": 2.399, |
|
"step": 94 |
|
}, |
|
{ |
|
"epoch": 28.0, |
|
"eval_accuracy": 0.6962025316455697, |
|
"eval_loss": 1.0775113105773926, |
|
"eval_runtime": 1.2537, |
|
"eval_samples_per_second": 63.013, |
|
"eval_steps_per_second": 2.393, |
|
"step": 98 |
|
}, |
|
{ |
|
"epoch": 28.571428571428573, |
|
"grad_norm": 19.96893310546875, |
|
"learning_rate": 3.7037037037037037e-05, |
|
"loss": 0.284, |
|
"step": 100 |
|
}, |
|
{ |
|
"epoch": 28.857142857142858, |
|
"eval_accuracy": 0.759493670886076, |
|
"eval_loss": 0.6816892027854919, |
|
"eval_runtime": 1.2538, |
|
"eval_samples_per_second": 63.006, |
|
"eval_steps_per_second": 2.393, |
|
"step": 101 |
|
}, |
|
{ |
|
"epoch": 30.0, |
|
"eval_accuracy": 0.6835443037974683, |
|
"eval_loss": 0.9235275983810425, |
|
"eval_runtime": 1.3158, |
|
"eval_samples_per_second": 60.038, |
|
"eval_steps_per_second": 2.28, |
|
"step": 105 |
|
}, |
|
{ |
|
"epoch": 30.857142857142858, |
|
"eval_accuracy": 0.759493670886076, |
|
"eval_loss": 0.6586753129959106, |
|
"eval_runtime": 1.2811, |
|
"eval_samples_per_second": 61.664, |
|
"eval_steps_per_second": 2.342, |
|
"step": 108 |
|
}, |
|
{ |
|
"epoch": 31.428571428571427, |
|
"grad_norm": 17.663036346435547, |
|
"learning_rate": 3.518518518518519e-05, |
|
"loss": 0.3134, |
|
"step": 110 |
|
}, |
|
{ |
|
"epoch": 32.0, |
|
"eval_accuracy": 0.7468354430379747, |
|
"eval_loss": 0.7085594534873962, |
|
"eval_runtime": 1.2261, |
|
"eval_samples_per_second": 64.432, |
|
"eval_steps_per_second": 2.447, |
|
"step": 112 |
|
}, |
|
{ |
|
"epoch": 32.857142857142854, |
|
"eval_accuracy": 0.7468354430379747, |
|
"eval_loss": 0.6894867420196533, |
|
"eval_runtime": 1.2687, |
|
"eval_samples_per_second": 62.267, |
|
"eval_steps_per_second": 2.365, |
|
"step": 115 |
|
}, |
|
{ |
|
"epoch": 34.0, |
|
"eval_accuracy": 0.7721518987341772, |
|
"eval_loss": 0.6417700052261353, |
|
"eval_runtime": 1.2606, |
|
"eval_samples_per_second": 62.669, |
|
"eval_steps_per_second": 2.38, |
|
"step": 119 |
|
}, |
|
{ |
|
"epoch": 34.285714285714285, |
|
"grad_norm": 5.122420310974121, |
|
"learning_rate": 3.3333333333333335e-05, |
|
"loss": 0.2266, |
|
"step": 120 |
|
}, |
|
{ |
|
"epoch": 34.857142857142854, |
|
"eval_accuracy": 0.7848101265822784, |
|
"eval_loss": 0.7007168531417847, |
|
"eval_runtime": 1.2576, |
|
"eval_samples_per_second": 62.819, |
|
"eval_steps_per_second": 2.386, |
|
"step": 122 |
|
}, |
|
{ |
|
"epoch": 36.0, |
|
"eval_accuracy": 0.8227848101265823, |
|
"eval_loss": 0.6919474005699158, |
|
"eval_runtime": 1.2393, |
|
"eval_samples_per_second": 63.744, |
|
"eval_steps_per_second": 2.421, |
|
"step": 126 |
|
}, |
|
{ |
|
"epoch": 36.857142857142854, |
|
"eval_accuracy": 0.7341772151898734, |
|
"eval_loss": 0.7562373876571655, |
|
"eval_runtime": 1.249, |
|
"eval_samples_per_second": 63.253, |
|
"eval_steps_per_second": 2.402, |
|
"step": 129 |
|
}, |
|
{ |
|
"epoch": 37.142857142857146, |
|
"grad_norm": 6.356712341308594, |
|
"learning_rate": 3.148148148148148e-05, |
|
"loss": 0.2249, |
|
"step": 130 |
|
}, |
|
{ |
|
"epoch": 38.0, |
|
"eval_accuracy": 0.7721518987341772, |
|
"eval_loss": 0.6774502992630005, |
|
"eval_runtime": 1.2642, |
|
"eval_samples_per_second": 62.491, |
|
"eval_steps_per_second": 2.373, |
|
"step": 133 |
|
}, |
|
{ |
|
"epoch": 38.857142857142854, |
|
"eval_accuracy": 0.759493670886076, |
|
"eval_loss": 0.7787100672721863, |
|
"eval_runtime": 1.2547, |
|
"eval_samples_per_second": 62.966, |
|
"eval_steps_per_second": 2.391, |
|
"step": 136 |
|
}, |
|
{ |
|
"epoch": 40.0, |
|
"grad_norm": 5.0371599197387695, |
|
"learning_rate": 2.962962962962963e-05, |
|
"loss": 0.2181, |
|
"step": 140 |
|
}, |
|
{ |
|
"epoch": 40.0, |
|
"eval_accuracy": 0.7721518987341772, |
|
"eval_loss": 0.7932192087173462, |
|
"eval_runtime": 1.2659, |
|
"eval_samples_per_second": 62.404, |
|
"eval_steps_per_second": 2.37, |
|
"step": 140 |
|
}, |
|
{ |
|
"epoch": 40.857142857142854, |
|
"eval_accuracy": 0.759493670886076, |
|
"eval_loss": 0.933390200138092, |
|
"eval_runtime": 1.2464, |
|
"eval_samples_per_second": 63.381, |
|
"eval_steps_per_second": 2.407, |
|
"step": 143 |
|
}, |
|
{ |
|
"epoch": 42.0, |
|
"eval_accuracy": 0.7468354430379747, |
|
"eval_loss": 0.822421133518219, |
|
"eval_runtime": 1.2469, |
|
"eval_samples_per_second": 63.358, |
|
"eval_steps_per_second": 2.406, |
|
"step": 147 |
|
}, |
|
{ |
|
"epoch": 42.857142857142854, |
|
"grad_norm": 5.825836181640625, |
|
"learning_rate": 2.777777777777778e-05, |
|
"loss": 0.186, |
|
"step": 150 |
|
}, |
|
{ |
|
"epoch": 42.857142857142854, |
|
"eval_accuracy": 0.7721518987341772, |
|
"eval_loss": 0.8443942666053772, |
|
"eval_runtime": 1.2708, |
|
"eval_samples_per_second": 62.166, |
|
"eval_steps_per_second": 2.361, |
|
"step": 150 |
|
}, |
|
{ |
|
"epoch": 44.0, |
|
"eval_accuracy": 0.7721518987341772, |
|
"eval_loss": 1.0349551439285278, |
|
"eval_runtime": 1.2484, |
|
"eval_samples_per_second": 63.281, |
|
"eval_steps_per_second": 2.403, |
|
"step": 154 |
|
}, |
|
{ |
|
"epoch": 44.857142857142854, |
|
"eval_accuracy": 0.7721518987341772, |
|
"eval_loss": 0.8385674953460693, |
|
"eval_runtime": 1.253, |
|
"eval_samples_per_second": 63.049, |
|
"eval_steps_per_second": 2.394, |
|
"step": 157 |
|
}, |
|
{ |
|
"epoch": 45.714285714285715, |
|
"grad_norm": 8.203125, |
|
"learning_rate": 2.5925925925925925e-05, |
|
"loss": 0.1882, |
|
"step": 160 |
|
}, |
|
{ |
|
"epoch": 46.0, |
|
"eval_accuracy": 0.7721518987341772, |
|
"eval_loss": 0.8384150266647339, |
|
"eval_runtime": 1.2347, |
|
"eval_samples_per_second": 63.983, |
|
"eval_steps_per_second": 2.43, |
|
"step": 161 |
|
}, |
|
{ |
|
"epoch": 46.857142857142854, |
|
"eval_accuracy": 0.759493670886076, |
|
"eval_loss": 0.7905216217041016, |
|
"eval_runtime": 1.2388, |
|
"eval_samples_per_second": 63.773, |
|
"eval_steps_per_second": 2.422, |
|
"step": 164 |
|
}, |
|
{ |
|
"epoch": 48.0, |
|
"eval_accuracy": 0.7721518987341772, |
|
"eval_loss": 0.7184091210365295, |
|
"eval_runtime": 1.2382, |
|
"eval_samples_per_second": 63.805, |
|
"eval_steps_per_second": 2.423, |
|
"step": 168 |
|
}, |
|
{ |
|
"epoch": 48.57142857142857, |
|
"grad_norm": 6.264881134033203, |
|
"learning_rate": 2.4074074074074074e-05, |
|
"loss": 0.1649, |
|
"step": 170 |
|
}, |
|
{ |
|
"epoch": 48.857142857142854, |
|
"eval_accuracy": 0.759493670886076, |
|
"eval_loss": 0.8118771910667419, |
|
"eval_runtime": 1.2516, |
|
"eval_samples_per_second": 63.119, |
|
"eval_steps_per_second": 2.397, |
|
"step": 171 |
|
}, |
|
{ |
|
"epoch": 50.0, |
|
"eval_accuracy": 0.7721518987341772, |
|
"eval_loss": 0.7171959280967712, |
|
"eval_runtime": 1.2501, |
|
"eval_samples_per_second": 63.194, |
|
"eval_steps_per_second": 2.4, |
|
"step": 175 |
|
}, |
|
{ |
|
"epoch": 50.857142857142854, |
|
"eval_accuracy": 0.7721518987341772, |
|
"eval_loss": 0.9248787760734558, |
|
"eval_runtime": 1.2392, |
|
"eval_samples_per_second": 63.752, |
|
"eval_steps_per_second": 2.421, |
|
"step": 178 |
|
}, |
|
{ |
|
"epoch": 51.42857142857143, |
|
"grad_norm": 7.78027868270874, |
|
"learning_rate": 2.2222222222222223e-05, |
|
"loss": 0.1847, |
|
"step": 180 |
|
}, |
|
{ |
|
"epoch": 52.0, |
|
"eval_accuracy": 0.7721518987341772, |
|
"eval_loss": 0.8068503737449646, |
|
"eval_runtime": 1.2411, |
|
"eval_samples_per_second": 63.654, |
|
"eval_steps_per_second": 2.417, |
|
"step": 182 |
|
}, |
|
{ |
|
"epoch": 52.857142857142854, |
|
"eval_accuracy": 0.7721518987341772, |
|
"eval_loss": 0.8752752542495728, |
|
"eval_runtime": 1.2551, |
|
"eval_samples_per_second": 62.941, |
|
"eval_steps_per_second": 2.39, |
|
"step": 185 |
|
}, |
|
{ |
|
"epoch": 54.0, |
|
"eval_accuracy": 0.7721518987341772, |
|
"eval_loss": 0.8822396397590637, |
|
"eval_runtime": 1.2453, |
|
"eval_samples_per_second": 63.441, |
|
"eval_steps_per_second": 2.409, |
|
"step": 189 |
|
}, |
|
{ |
|
"epoch": 54.285714285714285, |
|
"grad_norm": 6.302079200744629, |
|
"learning_rate": 2.037037037037037e-05, |
|
"loss": 0.1606, |
|
"step": 190 |
|
}, |
|
{ |
|
"epoch": 54.857142857142854, |
|
"eval_accuracy": 0.7974683544303798, |
|
"eval_loss": 0.7480602264404297, |
|
"eval_runtime": 1.2665, |
|
"eval_samples_per_second": 62.379, |
|
"eval_steps_per_second": 2.369, |
|
"step": 192 |
|
}, |
|
{ |
|
"epoch": 56.0, |
|
"eval_accuracy": 0.7721518987341772, |
|
"eval_loss": 0.7299729585647583, |
|
"eval_runtime": 1.2614, |
|
"eval_samples_per_second": 62.631, |
|
"eval_steps_per_second": 2.378, |
|
"step": 196 |
|
}, |
|
{ |
|
"epoch": 56.857142857142854, |
|
"eval_accuracy": 0.810126582278481, |
|
"eval_loss": 0.732498824596405, |
|
"eval_runtime": 1.2795, |
|
"eval_samples_per_second": 61.741, |
|
"eval_steps_per_second": 2.345, |
|
"step": 199 |
|
}, |
|
{ |
|
"epoch": 57.142857142857146, |
|
"grad_norm": 6.085663795471191, |
|
"learning_rate": 1.8518518518518518e-05, |
|
"loss": 0.142, |
|
"step": 200 |
|
}, |
|
{ |
|
"epoch": 58.0, |
|
"eval_accuracy": 0.8227848101265823, |
|
"eval_loss": 0.7087646126747131, |
|
"eval_runtime": 1.2611, |
|
"eval_samples_per_second": 62.643, |
|
"eval_steps_per_second": 2.379, |
|
"step": 203 |
|
}, |
|
{ |
|
"epoch": 58.857142857142854, |
|
"eval_accuracy": 0.8227848101265823, |
|
"eval_loss": 0.689956784248352, |
|
"eval_runtime": 1.2387, |
|
"eval_samples_per_second": 63.778, |
|
"eval_steps_per_second": 2.422, |
|
"step": 206 |
|
}, |
|
{ |
|
"epoch": 60.0, |
|
"grad_norm": 6.427424430847168, |
|
"learning_rate": 1.6666666666666667e-05, |
|
"loss": 0.1546, |
|
"step": 210 |
|
}, |
|
{ |
|
"epoch": 60.0, |
|
"eval_accuracy": 0.7721518987341772, |
|
"eval_loss": 0.8303462266921997, |
|
"eval_runtime": 1.2444, |
|
"eval_samples_per_second": 63.486, |
|
"eval_steps_per_second": 2.411, |
|
"step": 210 |
|
}, |
|
{ |
|
"epoch": 60.857142857142854, |
|
"eval_accuracy": 0.7974683544303798, |
|
"eval_loss": 0.8218382596969604, |
|
"eval_runtime": 1.2664, |
|
"eval_samples_per_second": 62.38, |
|
"eval_steps_per_second": 2.369, |
|
"step": 213 |
|
}, |
|
{ |
|
"epoch": 62.0, |
|
"eval_accuracy": 0.7848101265822784, |
|
"eval_loss": 0.9628246426582336, |
|
"eval_runtime": 1.2478, |
|
"eval_samples_per_second": 63.31, |
|
"eval_steps_per_second": 2.404, |
|
"step": 217 |
|
}, |
|
{ |
|
"epoch": 62.857142857142854, |
|
"grad_norm": 7.047874927520752, |
|
"learning_rate": 1.4814814814814815e-05, |
|
"loss": 0.1608, |
|
"step": 220 |
|
}, |
|
{ |
|
"epoch": 62.857142857142854, |
|
"eval_accuracy": 0.7848101265822784, |
|
"eval_loss": 0.9826211929321289, |
|
"eval_runtime": 1.2439, |
|
"eval_samples_per_second": 63.51, |
|
"eval_steps_per_second": 2.412, |
|
"step": 220 |
|
}, |
|
{ |
|
"epoch": 64.0, |
|
"eval_accuracy": 0.759493670886076, |
|
"eval_loss": 0.7603800892829895, |
|
"eval_runtime": 1.2499, |
|
"eval_samples_per_second": 63.204, |
|
"eval_steps_per_second": 2.4, |
|
"step": 224 |
|
}, |
|
{ |
|
"epoch": 64.85714285714286, |
|
"eval_accuracy": 0.7721518987341772, |
|
"eval_loss": 0.816529393196106, |
|
"eval_runtime": 1.2553, |
|
"eval_samples_per_second": 62.935, |
|
"eval_steps_per_second": 2.39, |
|
"step": 227 |
|
}, |
|
{ |
|
"epoch": 65.71428571428571, |
|
"grad_norm": 5.481523513793945, |
|
"learning_rate": 1.2962962962962962e-05, |
|
"loss": 0.1418, |
|
"step": 230 |
|
}, |
|
{ |
|
"epoch": 66.0, |
|
"eval_accuracy": 0.7721518987341772, |
|
"eval_loss": 0.8280141353607178, |
|
"eval_runtime": 1.2471, |
|
"eval_samples_per_second": 63.346, |
|
"eval_steps_per_second": 2.406, |
|
"step": 231 |
|
}, |
|
{ |
|
"epoch": 66.85714285714286, |
|
"eval_accuracy": 0.7848101265822784, |
|
"eval_loss": 0.9253725409507751, |
|
"eval_runtime": 1.2519, |
|
"eval_samples_per_second": 63.106, |
|
"eval_steps_per_second": 2.396, |
|
"step": 234 |
|
}, |
|
{ |
|
"epoch": 68.0, |
|
"eval_accuracy": 0.7848101265822784, |
|
"eval_loss": 0.843929648399353, |
|
"eval_runtime": 1.2503, |
|
"eval_samples_per_second": 63.185, |
|
"eval_steps_per_second": 2.399, |
|
"step": 238 |
|
}, |
|
{ |
|
"epoch": 68.57142857142857, |
|
"grad_norm": 7.343584060668945, |
|
"learning_rate": 1.1111111111111112e-05, |
|
"loss": 0.1505, |
|
"step": 240 |
|
}, |
|
{ |
|
"epoch": 68.85714285714286, |
|
"eval_accuracy": 0.7974683544303798, |
|
"eval_loss": 0.8095594644546509, |
|
"eval_runtime": 1.2513, |
|
"eval_samples_per_second": 63.132, |
|
"eval_steps_per_second": 2.397, |
|
"step": 241 |
|
}, |
|
{ |
|
"epoch": 70.0, |
|
"eval_accuracy": 0.759493670886076, |
|
"eval_loss": 0.947131872177124, |
|
"eval_runtime": 1.2591, |
|
"eval_samples_per_second": 62.745, |
|
"eval_steps_per_second": 2.383, |
|
"step": 245 |
|
}, |
|
{ |
|
"epoch": 70.85714285714286, |
|
"eval_accuracy": 0.759493670886076, |
|
"eval_loss": 0.902752161026001, |
|
"eval_runtime": 1.2601, |
|
"eval_samples_per_second": 62.695, |
|
"eval_steps_per_second": 2.381, |
|
"step": 248 |
|
}, |
|
{ |
|
"epoch": 71.42857142857143, |
|
"grad_norm": 2.4548463821411133, |
|
"learning_rate": 9.259259259259259e-06, |
|
"loss": 0.1211, |
|
"step": 250 |
|
}, |
|
{ |
|
"epoch": 72.0, |
|
"eval_accuracy": 0.7974683544303798, |
|
"eval_loss": 0.7743518352508545, |
|
"eval_runtime": 1.263, |
|
"eval_samples_per_second": 62.548, |
|
"eval_steps_per_second": 2.375, |
|
"step": 252 |
|
}, |
|
{ |
|
"epoch": 72.85714285714286, |
|
"eval_accuracy": 0.810126582278481, |
|
"eval_loss": 0.7755638360977173, |
|
"eval_runtime": 1.2392, |
|
"eval_samples_per_second": 63.75, |
|
"eval_steps_per_second": 2.421, |
|
"step": 255 |
|
}, |
|
{ |
|
"epoch": 74.0, |
|
"eval_accuracy": 0.810126582278481, |
|
"eval_loss": 0.7903711795806885, |
|
"eval_runtime": 1.2802, |
|
"eval_samples_per_second": 61.711, |
|
"eval_steps_per_second": 2.343, |
|
"step": 259 |
|
}, |
|
{ |
|
"epoch": 74.28571428571429, |
|
"grad_norm": 4.974315166473389, |
|
"learning_rate": 7.4074074074074075e-06, |
|
"loss": 0.1248, |
|
"step": 260 |
|
}, |
|
{ |
|
"epoch": 74.85714285714286, |
|
"eval_accuracy": 0.810126582278481, |
|
"eval_loss": 0.8036133646965027, |
|
"eval_runtime": 1.2526, |
|
"eval_samples_per_second": 63.067, |
|
"eval_steps_per_second": 2.395, |
|
"step": 262 |
|
}, |
|
{ |
|
"epoch": 76.0, |
|
"eval_accuracy": 0.7848101265822784, |
|
"eval_loss": 0.8591884970664978, |
|
"eval_runtime": 1.2518, |
|
"eval_samples_per_second": 63.108, |
|
"eval_steps_per_second": 2.397, |
|
"step": 266 |
|
}, |
|
{ |
|
"epoch": 76.85714285714286, |
|
"eval_accuracy": 0.7848101265822784, |
|
"eval_loss": 0.8945639133453369, |
|
"eval_runtime": 1.2541, |
|
"eval_samples_per_second": 62.996, |
|
"eval_steps_per_second": 2.392, |
|
"step": 269 |
|
}, |
|
{ |
|
"epoch": 77.14285714285714, |
|
"grad_norm": 5.69161319732666, |
|
"learning_rate": 5.555555555555556e-06, |
|
"loss": 0.122, |
|
"step": 270 |
|
}, |
|
{ |
|
"epoch": 78.0, |
|
"eval_accuracy": 0.7848101265822784, |
|
"eval_loss": 0.8575763702392578, |
|
"eval_runtime": 1.2492, |
|
"eval_samples_per_second": 63.241, |
|
"eval_steps_per_second": 2.402, |
|
"step": 273 |
|
}, |
|
{ |
|
"epoch": 78.85714285714286, |
|
"eval_accuracy": 0.7974683544303798, |
|
"eval_loss": 0.903687596321106, |
|
"eval_runtime": 1.2474, |
|
"eval_samples_per_second": 63.333, |
|
"eval_steps_per_second": 2.405, |
|
"step": 276 |
|
}, |
|
{ |
|
"epoch": 80.0, |
|
"grad_norm": 4.918107509613037, |
|
"learning_rate": 3.7037037037037037e-06, |
|
"loss": 0.1129, |
|
"step": 280 |
|
}, |
|
{ |
|
"epoch": 80.0, |
|
"eval_accuracy": 0.7848101265822784, |
|
"eval_loss": 0.998208224773407, |
|
"eval_runtime": 1.2524, |
|
"eval_samples_per_second": 63.079, |
|
"eval_steps_per_second": 2.395, |
|
"step": 280 |
|
}, |
|
{ |
|
"epoch": 80.85714285714286, |
|
"eval_accuracy": 0.7721518987341772, |
|
"eval_loss": 1.040386438369751, |
|
"eval_runtime": 1.2523, |
|
"eval_samples_per_second": 63.086, |
|
"eval_steps_per_second": 2.396, |
|
"step": 283 |
|
}, |
|
{ |
|
"epoch": 82.0, |
|
"eval_accuracy": 0.7974683544303798, |
|
"eval_loss": 0.9969496130943298, |
|
"eval_runtime": 1.2562, |
|
"eval_samples_per_second": 62.889, |
|
"eval_steps_per_second": 2.388, |
|
"step": 287 |
|
}, |
|
{ |
|
"epoch": 82.85714285714286, |
|
"grad_norm": 3.4930834770202637, |
|
"learning_rate": 1.8518518518518519e-06, |
|
"loss": 0.1136, |
|
"step": 290 |
|
}, |
|
{ |
|
"epoch": 82.85714285714286, |
|
"eval_accuracy": 0.7974683544303798, |
|
"eval_loss": 0.9585855007171631, |
|
"eval_runtime": 1.2777, |
|
"eval_samples_per_second": 61.829, |
|
"eval_steps_per_second": 2.348, |
|
"step": 290 |
|
}, |
|
{ |
|
"epoch": 84.0, |
|
"eval_accuracy": 0.7848101265822784, |
|
"eval_loss": 0.919994592666626, |
|
"eval_runtime": 1.2517, |
|
"eval_samples_per_second": 63.115, |
|
"eval_steps_per_second": 2.397, |
|
"step": 294 |
|
}, |
|
{ |
|
"epoch": 84.85714285714286, |
|
"eval_accuracy": 0.7721518987341772, |
|
"eval_loss": 0.9062978029251099, |
|
"eval_runtime": 1.2533, |
|
"eval_samples_per_second": 63.032, |
|
"eval_steps_per_second": 2.394, |
|
"step": 297 |
|
}, |
|
{ |
|
"epoch": 85.71428571428571, |
|
"grad_norm": 5.335267066955566, |
|
"learning_rate": 0.0, |
|
"loss": 0.1233, |
|
"step": 300 |
|
}, |
|
{ |
|
"epoch": 85.71428571428571, |
|
"eval_accuracy": 0.7721518987341772, |
|
"eval_loss": 0.9050403833389282, |
|
"eval_runtime": 1.2508, |
|
"eval_samples_per_second": 63.158, |
|
"eval_steps_per_second": 2.398, |
|
"step": 300 |
|
}, |
|
{ |
|
"epoch": 85.71428571428571, |
|
"step": 300, |
|
"total_flos": 2.9349165326823014e+18, |
|
"train_loss": 0.2620420479774475, |
|
"train_runtime": 1806.3529, |
|
"train_samples_per_second": 24.469, |
|
"train_steps_per_second": 0.166 |
|
}, |
|
{ |
|
"epoch": 85.71428571428571, |
|
"eval_accuracy": 0.8227848101265823, |
|
"eval_loss": 0.6919474005699158, |
|
"eval_runtime": 1.2565, |
|
"eval_samples_per_second": 62.873, |
|
"eval_steps_per_second": 2.388, |
|
"step": 300 |
|
} |
|
], |
|
"logging_steps": 10, |
|
"max_steps": 300, |
|
"num_input_tokens_seen": 0, |
|
"num_train_epochs": 100, |
|
"save_steps": 500, |
|
"total_flos": 2.9349165326823014e+18, |
|
"train_batch_size": 32, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|