adapters-llama2-bf16-QLORA-super_glue-axg
/
trainer_state-llama2-bf16-QLORA-super_glue-axg-sequence_classification.json
{ | |
"best_metric": null, | |
"best_model_checkpoint": null, | |
"epoch": 10.0, | |
"eval_steps": 1, | |
"global_step": 90, | |
"is_hyper_param_search": false, | |
"is_local_process_zero": true, | |
"is_world_process_zero": true, | |
"log_history": [ | |
{ | |
"epoch": 0.1111111111111111, | |
"grad_norm": 25.875, | |
"learning_rate": 2.5e-05, | |
"loss": 1.1308, | |
"step": 1 | |
}, | |
{ | |
"epoch": 0.1111111111111111, | |
"eval_accuracy": 0.5277777777777778, | |
"eval_loss": 0.8154025673866272, | |
"eval_runtime": 0.6646, | |
"eval_samples_per_second": 108.328, | |
"eval_steps_per_second": 7.523, | |
"step": 1 | |
}, | |
{ | |
"epoch": 0.2222222222222222, | |
"grad_norm": 31.75, | |
"learning_rate": 5e-05, | |
"loss": 0.9888, | |
"step": 2 | |
}, | |
{ | |
"epoch": 0.2222222222222222, | |
"eval_accuracy": 0.5416666666666666, | |
"eval_loss": 0.8097670078277588, | |
"eval_runtime": 0.7137, | |
"eval_samples_per_second": 100.883, | |
"eval_steps_per_second": 7.006, | |
"step": 2 | |
}, | |
{ | |
"epoch": 0.3333333333333333, | |
"grad_norm": 15.6875, | |
"learning_rate": 4.943181818181818e-05, | |
"loss": 0.8615, | |
"step": 3 | |
}, | |
{ | |
"epoch": 0.3333333333333333, | |
"eval_accuracy": 0.5833333333333334, | |
"eval_loss": 0.8370497226715088, | |
"eval_runtime": 0.6614, | |
"eval_samples_per_second": 108.857, | |
"eval_steps_per_second": 7.56, | |
"step": 3 | |
}, | |
{ | |
"epoch": 0.4444444444444444, | |
"grad_norm": 13.9375, | |
"learning_rate": 4.886363636363637e-05, | |
"loss": 0.901, | |
"step": 4 | |
}, | |
{ | |
"epoch": 0.4444444444444444, | |
"eval_accuracy": 0.5833333333333334, | |
"eval_loss": 0.8638712763786316, | |
"eval_runtime": 0.7123, | |
"eval_samples_per_second": 101.088, | |
"eval_steps_per_second": 7.02, | |
"step": 4 | |
}, | |
{ | |
"epoch": 0.5555555555555556, | |
"grad_norm": 13.375, | |
"learning_rate": 4.829545454545455e-05, | |
"loss": 1.1248, | |
"step": 5 | |
}, | |
{ | |
"epoch": 0.5555555555555556, | |
"eval_accuracy": 0.5972222222222222, | |
"eval_loss": 0.8621860146522522, | |
"eval_runtime": 0.7128, | |
"eval_samples_per_second": 101.003, | |
"eval_steps_per_second": 7.014, | |
"step": 5 | |
}, | |
{ | |
"epoch": 0.6666666666666666, | |
"grad_norm": 28.875, | |
"learning_rate": 4.772727272727273e-05, | |
"loss": 0.9482, | |
"step": 6 | |
}, | |
{ | |
"epoch": 0.6666666666666666, | |
"eval_accuracy": 0.5416666666666666, | |
"eval_loss": 0.8247477412223816, | |
"eval_runtime": 0.6622, | |
"eval_samples_per_second": 108.728, | |
"eval_steps_per_second": 7.551, | |
"step": 6 | |
}, | |
{ | |
"epoch": 0.7777777777777778, | |
"grad_norm": 33.25, | |
"learning_rate": 4.715909090909091e-05, | |
"loss": 0.8705, | |
"step": 7 | |
}, | |
{ | |
"epoch": 0.7777777777777778, | |
"eval_accuracy": 0.5833333333333334, | |
"eval_loss": 0.7855699062347412, | |
"eval_runtime": 0.7138, | |
"eval_samples_per_second": 100.863, | |
"eval_steps_per_second": 7.004, | |
"step": 7 | |
}, | |
{ | |
"epoch": 0.8888888888888888, | |
"grad_norm": 20.75, | |
"learning_rate": 4.659090909090909e-05, | |
"loss": 0.9702, | |
"step": 8 | |
}, | |
{ | |
"epoch": 0.8888888888888888, | |
"eval_accuracy": 0.5694444444444444, | |
"eval_loss": 0.7553032636642456, | |
"eval_runtime": 0.6125, | |
"eval_samples_per_second": 117.549, | |
"eval_steps_per_second": 8.163, | |
"step": 8 | |
}, | |
{ | |
"epoch": 1.0, | |
"grad_norm": 22.25, | |
"learning_rate": 4.602272727272727e-05, | |
"loss": 0.7868, | |
"step": 9 | |
}, | |
{ | |
"epoch": 1.0, | |
"eval_accuracy": 0.5555555555555556, | |
"eval_loss": 0.7463514804840088, | |
"eval_runtime": 0.6616, | |
"eval_samples_per_second": 108.828, | |
"eval_steps_per_second": 7.557, | |
"step": 9 | |
}, | |
{ | |
"epoch": 1.1111111111111112, | |
"grad_norm": 6.4375, | |
"learning_rate": 4.545454545454546e-05, | |
"loss": 0.7814, | |
"step": 10 | |
}, | |
{ | |
"epoch": 1.1111111111111112, | |
"eval_accuracy": 0.5416666666666666, | |
"eval_loss": 0.7470024824142456, | |
"eval_runtime": 0.6612, | |
"eval_samples_per_second": 108.9, | |
"eval_steps_per_second": 7.562, | |
"step": 10 | |
}, | |
{ | |
"epoch": 1.2222222222222223, | |
"grad_norm": 19.875, | |
"learning_rate": 4.488636363636364e-05, | |
"loss": 0.8527, | |
"step": 11 | |
}, | |
{ | |
"epoch": 1.2222222222222223, | |
"eval_accuracy": 0.5277777777777778, | |
"eval_loss": 0.7425130009651184, | |
"eval_runtime": 0.6143, | |
"eval_samples_per_second": 117.207, | |
"eval_steps_per_second": 8.139, | |
"step": 11 | |
}, | |
{ | |
"epoch": 1.3333333333333333, | |
"grad_norm": 15.1875, | |
"learning_rate": 4.431818181818182e-05, | |
"loss": 0.7484, | |
"step": 12 | |
}, | |
{ | |
"epoch": 1.3333333333333333, | |
"eval_accuracy": 0.5555555555555556, | |
"eval_loss": 0.7314046025276184, | |
"eval_runtime": 0.6182, | |
"eval_samples_per_second": 116.47, | |
"eval_steps_per_second": 8.088, | |
"step": 12 | |
}, | |
{ | |
"epoch": 1.4444444444444444, | |
"grad_norm": 21.375, | |
"learning_rate": 4.375e-05, | |
"loss": 0.6436, | |
"step": 13 | |
}, | |
{ | |
"epoch": 1.4444444444444444, | |
"eval_accuracy": 0.5833333333333334, | |
"eval_loss": 0.7158339023590088, | |
"eval_runtime": 0.7149, | |
"eval_samples_per_second": 100.712, | |
"eval_steps_per_second": 6.994, | |
"step": 13 | |
}, | |
{ | |
"epoch": 1.5555555555555556, | |
"grad_norm": 21.875, | |
"learning_rate": 4.318181818181819e-05, | |
"loss": 0.7159, | |
"step": 14 | |
}, | |
{ | |
"epoch": 1.5555555555555556, | |
"eval_accuracy": 0.6111111111111112, | |
"eval_loss": 0.7068142294883728, | |
"eval_runtime": 0.7129, | |
"eval_samples_per_second": 100.991, | |
"eval_steps_per_second": 7.013, | |
"step": 14 | |
}, | |
{ | |
"epoch": 1.6666666666666665, | |
"grad_norm": 5.6875, | |
"learning_rate": 4.261363636363637e-05, | |
"loss": 0.6247, | |
"step": 15 | |
}, | |
{ | |
"epoch": 1.6666666666666665, | |
"eval_accuracy": 0.5972222222222222, | |
"eval_loss": 0.7105034589767456, | |
"eval_runtime": 0.6625, | |
"eval_samples_per_second": 108.678, | |
"eval_steps_per_second": 7.547, | |
"step": 15 | |
}, | |
{ | |
"epoch": 1.7777777777777777, | |
"grad_norm": 12.9375, | |
"learning_rate": 4.204545454545455e-05, | |
"loss": 0.6662, | |
"step": 16 | |
}, | |
{ | |
"epoch": 1.7777777777777777, | |
"eval_accuracy": 0.5972222222222222, | |
"eval_loss": 0.7118598222732544, | |
"eval_runtime": 0.5619, | |
"eval_samples_per_second": 128.13, | |
"eval_steps_per_second": 8.898, | |
"step": 16 | |
}, | |
{ | |
"epoch": 1.8888888888888888, | |
"grad_norm": 9.0, | |
"learning_rate": 4.1477272727272734e-05, | |
"loss": 0.4505, | |
"step": 17 | |
}, | |
{ | |
"epoch": 1.8888888888888888, | |
"eval_accuracy": 0.5833333333333334, | |
"eval_loss": 0.7344834804534912, | |
"eval_runtime": 0.566, | |
"eval_samples_per_second": 127.198, | |
"eval_steps_per_second": 8.833, | |
"step": 17 | |
}, | |
{ | |
"epoch": 2.0, | |
"grad_norm": 9.75, | |
"learning_rate": 4.0909090909090915e-05, | |
"loss": 0.5626, | |
"step": 18 | |
}, | |
{ | |
"epoch": 2.0, | |
"eval_accuracy": 0.5277777777777778, | |
"eval_loss": 0.7482435703277588, | |
"eval_runtime": 0.6143, | |
"eval_samples_per_second": 117.216, | |
"eval_steps_per_second": 8.14, | |
"step": 18 | |
}, | |
{ | |
"epoch": 2.111111111111111, | |
"grad_norm": 16.125, | |
"learning_rate": 4.034090909090909e-05, | |
"loss": 0.4823, | |
"step": 19 | |
}, | |
{ | |
"epoch": 2.111111111111111, | |
"eval_accuracy": 0.5277777777777778, | |
"eval_loss": 0.7449544072151184, | |
"eval_runtime": 0.7126, | |
"eval_samples_per_second": 101.038, | |
"eval_steps_per_second": 7.016, | |
"step": 19 | |
}, | |
{ | |
"epoch": 2.2222222222222223, | |
"grad_norm": 15.4375, | |
"learning_rate": 3.9772727272727275e-05, | |
"loss": 0.4237, | |
"step": 20 | |
}, | |
{ | |
"epoch": 2.2222222222222223, | |
"eval_accuracy": 0.5416666666666666, | |
"eval_loss": 0.7253146767616272, | |
"eval_runtime": 0.6618, | |
"eval_samples_per_second": 108.786, | |
"eval_steps_per_second": 7.555, | |
"step": 20 | |
}, | |
{ | |
"epoch": 2.3333333333333335, | |
"grad_norm": 22.625, | |
"learning_rate": 3.9204545454545456e-05, | |
"loss": 0.5236, | |
"step": 21 | |
}, | |
{ | |
"epoch": 2.3333333333333335, | |
"eval_accuracy": 0.5833333333333334, | |
"eval_loss": 0.6920505166053772, | |
"eval_runtime": 0.613, | |
"eval_samples_per_second": 117.447, | |
"eval_steps_per_second": 8.156, | |
"step": 21 | |
}, | |
{ | |
"epoch": 2.4444444444444446, | |
"grad_norm": 11.75, | |
"learning_rate": 3.8636363636363636e-05, | |
"loss": 0.4028, | |
"step": 22 | |
}, | |
{ | |
"epoch": 2.4444444444444446, | |
"eval_accuracy": 0.6666666666666666, | |
"eval_loss": 0.6593967080116272, | |
"eval_runtime": 0.6119, | |
"eval_samples_per_second": 117.661, | |
"eval_steps_per_second": 8.171, | |
"step": 22 | |
}, | |
{ | |
"epoch": 2.5555555555555554, | |
"grad_norm": 18.875, | |
"learning_rate": 3.8068181818181816e-05, | |
"loss": 0.3952, | |
"step": 23 | |
}, | |
{ | |
"epoch": 2.5555555555555554, | |
"eval_accuracy": 0.6944444444444444, | |
"eval_loss": 0.6502448320388794, | |
"eval_runtime": 0.6118, | |
"eval_samples_per_second": 117.679, | |
"eval_steps_per_second": 8.172, | |
"step": 23 | |
}, | |
{ | |
"epoch": 2.6666666666666665, | |
"grad_norm": 8.3125, | |
"learning_rate": 3.7500000000000003e-05, | |
"loss": 0.3701, | |
"step": 24 | |
}, | |
{ | |
"epoch": 2.6666666666666665, | |
"eval_accuracy": 0.6944444444444444, | |
"eval_loss": 0.6597442626953125, | |
"eval_runtime": 0.7121, | |
"eval_samples_per_second": 101.112, | |
"eval_steps_per_second": 7.022, | |
"step": 24 | |
}, | |
{ | |
"epoch": 2.7777777777777777, | |
"grad_norm": 25.0, | |
"learning_rate": 3.6931818181818184e-05, | |
"loss": 0.4536, | |
"step": 25 | |
}, | |
{ | |
"epoch": 2.7777777777777777, | |
"eval_accuracy": 0.6805555555555556, | |
"eval_loss": 0.6615126132965088, | |
"eval_runtime": 0.6608, | |
"eval_samples_per_second": 108.959, | |
"eval_steps_per_second": 7.567, | |
"step": 25 | |
}, | |
{ | |
"epoch": 2.888888888888889, | |
"grad_norm": 25.125, | |
"learning_rate": 3.6363636363636364e-05, | |
"loss": 0.4004, | |
"step": 26 | |
}, | |
{ | |
"epoch": 2.888888888888889, | |
"eval_accuracy": 0.6666666666666666, | |
"eval_loss": 0.6555294394493103, | |
"eval_runtime": 0.6106, | |
"eval_samples_per_second": 117.911, | |
"eval_steps_per_second": 8.188, | |
"step": 26 | |
}, | |
{ | |
"epoch": 3.0, | |
"grad_norm": 16.875, | |
"learning_rate": 3.579545454545455e-05, | |
"loss": 0.2805, | |
"step": 27 | |
}, | |
{ | |
"epoch": 3.0, | |
"eval_accuracy": 0.6388888888888888, | |
"eval_loss": 0.650442361831665, | |
"eval_runtime": 0.6114, | |
"eval_samples_per_second": 117.754, | |
"eval_steps_per_second": 8.177, | |
"step": 27 | |
}, | |
{ | |
"epoch": 3.111111111111111, | |
"grad_norm": 5.6875, | |
"learning_rate": 3.522727272727273e-05, | |
"loss": 0.3309, | |
"step": 28 | |
}, | |
{ | |
"epoch": 3.111111111111111, | |
"eval_accuracy": 0.6111111111111112, | |
"eval_loss": 0.6491343379020691, | |
"eval_runtime": 0.7116, | |
"eval_samples_per_second": 101.178, | |
"eval_steps_per_second": 7.026, | |
"step": 28 | |
}, | |
{ | |
"epoch": 3.2222222222222223, | |
"grad_norm": 20.75, | |
"learning_rate": 3.465909090909091e-05, | |
"loss": 0.4176, | |
"step": 29 | |
}, | |
{ | |
"epoch": 3.2222222222222223, | |
"eval_accuracy": 0.6388888888888888, | |
"eval_loss": 0.6927388310432434, | |
"eval_runtime": 0.7116, | |
"eval_samples_per_second": 101.173, | |
"eval_steps_per_second": 7.026, | |
"step": 29 | |
}, | |
{ | |
"epoch": 3.3333333333333335, | |
"grad_norm": 11.375, | |
"learning_rate": 3.409090909090909e-05, | |
"loss": 0.1967, | |
"step": 30 | |
}, | |
{ | |
"epoch": 3.3333333333333335, | |
"eval_accuracy": 0.6666666666666666, | |
"eval_loss": 0.7233132123947144, | |
"eval_runtime": 0.6182, | |
"eval_samples_per_second": 116.474, | |
"eval_steps_per_second": 8.088, | |
"step": 30 | |
}, | |
{ | |
"epoch": 3.4444444444444446, | |
"grad_norm": 24.875, | |
"learning_rate": 3.352272727272727e-05, | |
"loss": 0.2568, | |
"step": 31 | |
}, | |
{ | |
"epoch": 3.4444444444444446, | |
"eval_accuracy": 0.6805555555555556, | |
"eval_loss": 0.7308578491210938, | |
"eval_runtime": 0.6112, | |
"eval_samples_per_second": 117.804, | |
"eval_steps_per_second": 8.181, | |
"step": 31 | |
}, | |
{ | |
"epoch": 3.5555555555555554, | |
"grad_norm": 11.25, | |
"learning_rate": 3.295454545454545e-05, | |
"loss": 0.2841, | |
"step": 32 | |
}, | |
{ | |
"epoch": 3.5555555555555554, | |
"eval_accuracy": 0.6666666666666666, | |
"eval_loss": 0.7040252685546875, | |
"eval_runtime": 0.6107, | |
"eval_samples_per_second": 117.902, | |
"eval_steps_per_second": 8.188, | |
"step": 32 | |
}, | |
{ | |
"epoch": 3.6666666666666665, | |
"grad_norm": 14.875, | |
"learning_rate": 3.238636363636364e-05, | |
"loss": 0.2797, | |
"step": 33 | |
}, | |
{ | |
"epoch": 3.6666666666666665, | |
"eval_accuracy": 0.625, | |
"eval_loss": 0.6633122563362122, | |
"eval_runtime": 0.6111, | |
"eval_samples_per_second": 117.821, | |
"eval_steps_per_second": 8.182, | |
"step": 33 | |
}, | |
{ | |
"epoch": 3.7777777777777777, | |
"grad_norm": 20.375, | |
"learning_rate": 3.181818181818182e-05, | |
"loss": 0.2031, | |
"step": 34 | |
}, | |
{ | |
"epoch": 3.7777777777777777, | |
"eval_accuracy": 0.6666666666666666, | |
"eval_loss": 0.6236428022384644, | |
"eval_runtime": 0.6112, | |
"eval_samples_per_second": 117.809, | |
"eval_steps_per_second": 8.181, | |
"step": 34 | |
}, | |
{ | |
"epoch": 3.888888888888889, | |
"grad_norm": 2.890625, | |
"learning_rate": 3.125e-05, | |
"loss": 0.1318, | |
"step": 35 | |
}, | |
{ | |
"epoch": 3.888888888888889, | |
"eval_accuracy": 0.6944444444444444, | |
"eval_loss": 0.6151860356330872, | |
"eval_runtime": 0.6134, | |
"eval_samples_per_second": 117.386, | |
"eval_steps_per_second": 8.152, | |
"step": 35 | |
}, | |
{ | |
"epoch": 4.0, | |
"grad_norm": 5.25, | |
"learning_rate": 3.068181818181818e-05, | |
"loss": 0.2346, | |
"step": 36 | |
}, | |
{ | |
"epoch": 4.0, | |
"eval_accuracy": 0.6944444444444444, | |
"eval_loss": 0.6115561723709106, | |
"eval_runtime": 0.7128, | |
"eval_samples_per_second": 101.01, | |
"eval_steps_per_second": 7.015, | |
"step": 36 | |
}, | |
{ | |
"epoch": 4.111111111111111, | |
"grad_norm": 13.125, | |
"learning_rate": 3.0113636363636365e-05, | |
"loss": 0.1741, | |
"step": 37 | |
}, | |
{ | |
"epoch": 4.111111111111111, | |
"eval_accuracy": 0.7083333333333334, | |
"eval_loss": 0.6088710427284241, | |
"eval_runtime": 0.662, | |
"eval_samples_per_second": 108.762, | |
"eval_steps_per_second": 7.553, | |
"step": 37 | |
}, | |
{ | |
"epoch": 4.222222222222222, | |
"grad_norm": 5.84375, | |
"learning_rate": 2.954545454545455e-05, | |
"loss": 0.186, | |
"step": 38 | |
}, | |
{ | |
"epoch": 4.222222222222222, | |
"eval_accuracy": 0.7222222222222222, | |
"eval_loss": 0.6091925501823425, | |
"eval_runtime": 0.7117, | |
"eval_samples_per_second": 101.172, | |
"eval_steps_per_second": 7.026, | |
"step": 38 | |
}, | |
{ | |
"epoch": 4.333333333333333, | |
"grad_norm": 3.46875, | |
"learning_rate": 2.8977272727272732e-05, | |
"loss": 0.1125, | |
"step": 39 | |
}, | |
{ | |
"epoch": 4.333333333333333, | |
"eval_accuracy": 0.7222222222222222, | |
"eval_loss": 0.6227471828460693, | |
"eval_runtime": 0.7115, | |
"eval_samples_per_second": 101.191, | |
"eval_steps_per_second": 7.027, | |
"step": 39 | |
}, | |
{ | |
"epoch": 4.444444444444445, | |
"grad_norm": 9.0625, | |
"learning_rate": 2.8409090909090912e-05, | |
"loss": 0.1299, | |
"step": 40 | |
}, | |
{ | |
"epoch": 4.444444444444445, | |
"eval_accuracy": 0.6805555555555556, | |
"eval_loss": 0.634276270866394, | |
"eval_runtime": 0.6124, | |
"eval_samples_per_second": 117.561, | |
"eval_steps_per_second": 8.164, | |
"step": 40 | |
}, | |
{ | |
"epoch": 4.555555555555555, | |
"grad_norm": 7.78125, | |
"learning_rate": 2.784090909090909e-05, | |
"loss": 0.1387, | |
"step": 41 | |
}, | |
{ | |
"epoch": 4.555555555555555, | |
"eval_accuracy": 0.6666666666666666, | |
"eval_loss": 0.6676434278488159, | |
"eval_runtime": 0.7113, | |
"eval_samples_per_second": 101.226, | |
"eval_steps_per_second": 7.03, | |
"step": 41 | |
}, | |
{ | |
"epoch": 4.666666666666667, | |
"grad_norm": 5.90625, | |
"learning_rate": 2.7272727272727273e-05, | |
"loss": 0.1129, | |
"step": 42 | |
}, | |
{ | |
"epoch": 4.666666666666667, | |
"eval_accuracy": 0.6805555555555556, | |
"eval_loss": 0.7227713465690613, | |
"eval_runtime": 0.7115, | |
"eval_samples_per_second": 101.191, | |
"eval_steps_per_second": 7.027, | |
"step": 42 | |
}, | |
{ | |
"epoch": 4.777777777777778, | |
"grad_norm": 7.125, | |
"learning_rate": 2.6704545454545453e-05, | |
"loss": 0.0691, | |
"step": 43 | |
}, | |
{ | |
"epoch": 4.777777777777778, | |
"eval_accuracy": 0.6944444444444444, | |
"eval_loss": 0.7673901319503784, | |
"eval_runtime": 0.6613, | |
"eval_samples_per_second": 108.871, | |
"eval_steps_per_second": 7.561, | |
"step": 43 | |
}, | |
{ | |
"epoch": 4.888888888888889, | |
"grad_norm": 4.03125, | |
"learning_rate": 2.6136363636363637e-05, | |
"loss": 0.0547, | |
"step": 44 | |
}, | |
{ | |
"epoch": 4.888888888888889, | |
"eval_accuracy": 0.6944444444444444, | |
"eval_loss": 0.7912532687187195, | |
"eval_runtime": 0.6118, | |
"eval_samples_per_second": 117.677, | |
"eval_steps_per_second": 8.172, | |
"step": 44 | |
}, | |
{ | |
"epoch": 5.0, | |
"grad_norm": 13.75, | |
"learning_rate": 2.5568181818181817e-05, | |
"loss": 0.0985, | |
"step": 45 | |
}, | |
{ | |
"epoch": 5.0, | |
"eval_accuracy": 0.6944444444444444, | |
"eval_loss": 0.7817304134368896, | |
"eval_runtime": 0.6108, | |
"eval_samples_per_second": 117.879, | |
"eval_steps_per_second": 8.186, | |
"step": 45 | |
}, | |
{ | |
"epoch": 5.111111111111111, | |
"grad_norm": 3.53125, | |
"learning_rate": 2.5e-05, | |
"loss": 0.0375, | |
"step": 46 | |
}, | |
{ | |
"epoch": 5.111111111111111, | |
"eval_accuracy": 0.7083333333333334, | |
"eval_loss": 0.7576977014541626, | |
"eval_runtime": 0.6112, | |
"eval_samples_per_second": 117.794, | |
"eval_steps_per_second": 8.18, | |
"step": 46 | |
}, | |
{ | |
"epoch": 5.222222222222222, | |
"grad_norm": 2.046875, | |
"learning_rate": 2.4431818181818185e-05, | |
"loss": 0.0279, | |
"step": 47 | |
}, | |
{ | |
"epoch": 5.222222222222222, | |
"eval_accuracy": 0.7083333333333334, | |
"eval_loss": 0.7299577593803406, | |
"eval_runtime": 0.6111, | |
"eval_samples_per_second": 117.826, | |
"eval_steps_per_second": 8.182, | |
"step": 47 | |
}, | |
{ | |
"epoch": 5.333333333333333, | |
"grad_norm": 1.640625, | |
"learning_rate": 2.3863636363636365e-05, | |
"loss": 0.0396, | |
"step": 48 | |
}, | |
{ | |
"epoch": 5.333333333333333, | |
"eval_accuracy": 0.75, | |
"eval_loss": 0.7089157104492188, | |
"eval_runtime": 0.7116, | |
"eval_samples_per_second": 101.18, | |
"eval_steps_per_second": 7.026, | |
"step": 48 | |
}, | |
{ | |
"epoch": 5.444444444444445, | |
"grad_norm": 3.640625, | |
"learning_rate": 2.3295454545454546e-05, | |
"loss": 0.049, | |
"step": 49 | |
}, | |
{ | |
"epoch": 5.444444444444445, | |
"eval_accuracy": 0.75, | |
"eval_loss": 0.7139127254486084, | |
"eval_runtime": 0.6114, | |
"eval_samples_per_second": 117.764, | |
"eval_steps_per_second": 8.178, | |
"step": 49 | |
}, | |
{ | |
"epoch": 5.555555555555555, | |
"grad_norm": 2.671875, | |
"learning_rate": 2.272727272727273e-05, | |
"loss": 0.0306, | |
"step": 50 | |
}, | |
{ | |
"epoch": 5.555555555555555, | |
"eval_accuracy": 0.75, | |
"eval_loss": 0.7333288192749023, | |
"eval_runtime": 0.6125, | |
"eval_samples_per_second": 117.554, | |
"eval_steps_per_second": 8.163, | |
"step": 50 | |
}, | |
{ | |
"epoch": 5.666666666666667, | |
"grad_norm": 0.76171875, | |
"learning_rate": 2.215909090909091e-05, | |
"loss": 0.0138, | |
"step": 51 | |
}, | |
{ | |
"epoch": 5.666666666666667, | |
"eval_accuracy": 0.7777777777777778, | |
"eval_loss": 0.7559387683868408, | |
"eval_runtime": 0.611, | |
"eval_samples_per_second": 117.848, | |
"eval_steps_per_second": 8.184, | |
"step": 51 | |
}, | |
{ | |
"epoch": 5.777777777777778, | |
"grad_norm": 0.50390625, | |
"learning_rate": 2.1590909090909093e-05, | |
"loss": 0.0076, | |
"step": 52 | |
}, | |
{ | |
"epoch": 5.777777777777778, | |
"eval_accuracy": 0.7777777777777778, | |
"eval_loss": 0.7911980152130127, | |
"eval_runtime": 0.6618, | |
"eval_samples_per_second": 108.8, | |
"eval_steps_per_second": 7.556, | |
"step": 52 | |
}, | |
{ | |
"epoch": 5.888888888888889, | |
"grad_norm": 1.421875, | |
"learning_rate": 2.1022727272727274e-05, | |
"loss": 0.02, | |
"step": 53 | |
}, | |
{ | |
"epoch": 5.888888888888889, | |
"eval_accuracy": 0.7777777777777778, | |
"eval_loss": 0.810262143611908, | |
"eval_runtime": 0.6623, | |
"eval_samples_per_second": 108.718, | |
"eval_steps_per_second": 7.55, | |
"step": 53 | |
}, | |
{ | |
"epoch": 6.0, | |
"grad_norm": 5.03125, | |
"learning_rate": 2.0454545454545457e-05, | |
"loss": 0.0509, | |
"step": 54 | |
}, | |
{ | |
"epoch": 6.0, | |
"eval_accuracy": 0.7777777777777778, | |
"eval_loss": 0.8537428975105286, | |
"eval_runtime": 0.712, | |
"eval_samples_per_second": 101.122, | |
"eval_steps_per_second": 7.022, | |
"step": 54 | |
}, | |
{ | |
"epoch": 6.111111111111111, | |
"grad_norm": 0.3203125, | |
"learning_rate": 1.9886363636363638e-05, | |
"loss": 0.0033, | |
"step": 55 | |
}, | |
{ | |
"epoch": 6.111111111111111, | |
"eval_accuracy": 0.7777777777777778, | |
"eval_loss": 0.9030678868293762, | |
"eval_runtime": 0.6622, | |
"eval_samples_per_second": 108.728, | |
"eval_steps_per_second": 7.551, | |
"step": 55 | |
}, | |
{ | |
"epoch": 6.222222222222222, | |
"grad_norm": 0.21484375, | |
"learning_rate": 1.9318181818181818e-05, | |
"loss": 0.0033, | |
"step": 56 | |
}, | |
{ | |
"epoch": 6.222222222222222, | |
"eval_accuracy": 0.75, | |
"eval_loss": 0.9585488438606262, | |
"eval_runtime": 0.611, | |
"eval_samples_per_second": 117.842, | |
"eval_steps_per_second": 8.183, | |
"step": 56 | |
}, | |
{ | |
"epoch": 6.333333333333333, | |
"grad_norm": 0.9296875, | |
"learning_rate": 1.8750000000000002e-05, | |
"loss": 0.0057, | |
"step": 57 | |
}, | |
{ | |
"epoch": 6.333333333333333, | |
"eval_accuracy": 0.7638888888888888, | |
"eval_loss": 0.9753527045249939, | |
"eval_runtime": 0.5617, | |
"eval_samples_per_second": 128.182, | |
"eval_steps_per_second": 8.902, | |
"step": 57 | |
}, | |
{ | |
"epoch": 6.444444444444445, | |
"grad_norm": 1.171875, | |
"learning_rate": 1.8181818181818182e-05, | |
"loss": 0.0091, | |
"step": 58 | |
}, | |
{ | |
"epoch": 6.444444444444445, | |
"eval_accuracy": 0.7638888888888888, | |
"eval_loss": 1.0243321657180786, | |
"eval_runtime": 0.6134, | |
"eval_samples_per_second": 117.377, | |
"eval_steps_per_second": 8.151, | |
"step": 58 | |
}, | |
{ | |
"epoch": 6.555555555555555, | |
"grad_norm": 0.25390625, | |
"learning_rate": 1.7613636363636366e-05, | |
"loss": 0.0024, | |
"step": 59 | |
}, | |
{ | |
"epoch": 6.555555555555555, | |
"eval_accuracy": 0.75, | |
"eval_loss": 1.0561052560806274, | |
"eval_runtime": 0.7119, | |
"eval_samples_per_second": 101.131, | |
"eval_steps_per_second": 7.023, | |
"step": 59 | |
}, | |
{ | |
"epoch": 6.666666666666667, | |
"grad_norm": 0.2080078125, | |
"learning_rate": 1.7045454545454546e-05, | |
"loss": 0.0011, | |
"step": 60 | |
}, | |
{ | |
"epoch": 6.666666666666667, | |
"eval_accuracy": 0.75, | |
"eval_loss": 1.0907957553863525, | |
"eval_runtime": 0.7135, | |
"eval_samples_per_second": 100.915, | |
"eval_steps_per_second": 7.008, | |
"step": 60 | |
}, | |
{ | |
"epoch": 6.777777777777778, | |
"grad_norm": 0.2001953125, | |
"learning_rate": 1.6477272727272726e-05, | |
"loss": 0.0017, | |
"step": 61 | |
}, | |
{ | |
"epoch": 6.777777777777778, | |
"eval_accuracy": 0.75, | |
"eval_loss": 1.1045655012130737, | |
"eval_runtime": 0.6619, | |
"eval_samples_per_second": 108.781, | |
"eval_steps_per_second": 7.554, | |
"step": 61 | |
}, | |
{ | |
"epoch": 6.888888888888889, | |
"grad_norm": 0.265625, | |
"learning_rate": 1.590909090909091e-05, | |
"loss": 0.0017, | |
"step": 62 | |
}, | |
{ | |
"epoch": 6.888888888888889, | |
"eval_accuracy": 0.7638888888888888, | |
"eval_loss": 1.1206629276275635, | |
"eval_runtime": 0.6117, | |
"eval_samples_per_second": 117.708, | |
"eval_steps_per_second": 8.174, | |
"step": 62 | |
}, | |
{ | |
"epoch": 7.0, | |
"grad_norm": 0.035400390625, | |
"learning_rate": 1.534090909090909e-05, | |
"loss": 0.0006, | |
"step": 63 | |
}, | |
{ | |
"epoch": 7.0, | |
"eval_accuracy": 0.7777777777777778, | |
"eval_loss": 1.1336255073547363, | |
"eval_runtime": 0.6634, | |
"eval_samples_per_second": 108.527, | |
"eval_steps_per_second": 7.537, | |
"step": 63 | |
}, | |
{ | |
"epoch": 7.111111111111111, | |
"grad_norm": 0.040283203125, | |
"learning_rate": 1.4772727272727274e-05, | |
"loss": 0.0004, | |
"step": 64 | |
}, | |
{ | |
"epoch": 7.111111111111111, | |
"eval_accuracy": 0.7777777777777778, | |
"eval_loss": 1.1413822174072266, | |
"eval_runtime": 0.6642, | |
"eval_samples_per_second": 108.409, | |
"eval_steps_per_second": 7.528, | |
"step": 64 | |
}, | |
{ | |
"epoch": 7.222222222222222, | |
"grad_norm": 0.076171875, | |
"learning_rate": 1.4204545454545456e-05, | |
"loss": 0.0005, | |
"step": 65 | |
}, | |
{ | |
"epoch": 7.222222222222222, | |
"eval_accuracy": 0.7777777777777778, | |
"eval_loss": 1.143465280532837, | |
"eval_runtime": 0.5614, | |
"eval_samples_per_second": 128.248, | |
"eval_steps_per_second": 8.906, | |
"step": 65 | |
}, | |
{ | |
"epoch": 7.333333333333333, | |
"grad_norm": 0.06201171875, | |
"learning_rate": 1.3636363636363637e-05, | |
"loss": 0.0005, | |
"step": 66 | |
}, | |
{ | |
"epoch": 7.333333333333333, | |
"eval_accuracy": 0.7777777777777778, | |
"eval_loss": 1.155677318572998, | |
"eval_runtime": 0.6131, | |
"eval_samples_per_second": 117.44, | |
"eval_steps_per_second": 8.156, | |
"step": 66 | |
}, | |
{ | |
"epoch": 7.444444444444445, | |
"grad_norm": 0.0888671875, | |
"learning_rate": 1.3068181818181819e-05, | |
"loss": 0.0008, | |
"step": 67 | |
}, | |
{ | |
"epoch": 7.444444444444445, | |
"eval_accuracy": 0.7777777777777778, | |
"eval_loss": 1.1653226613998413, | |
"eval_runtime": 0.6121, | |
"eval_samples_per_second": 117.637, | |
"eval_steps_per_second": 8.169, | |
"step": 67 | |
}, | |
{ | |
"epoch": 7.555555555555555, | |
"grad_norm": 0.515625, | |
"learning_rate": 1.25e-05, | |
"loss": 0.0021, | |
"step": 68 | |
}, | |
{ | |
"epoch": 7.555555555555555, | |
"eval_accuracy": 0.7777777777777778, | |
"eval_loss": 1.1580508947372437, | |
"eval_runtime": 0.7136, | |
"eval_samples_per_second": 100.903, | |
"eval_steps_per_second": 7.007, | |
"step": 68 | |
}, | |
{ | |
"epoch": 7.666666666666667, | |
"grad_norm": 0.0206298828125, | |
"learning_rate": 1.1931818181818183e-05, | |
"loss": 0.0002, | |
"step": 69 | |
}, | |
{ | |
"epoch": 7.666666666666667, | |
"eval_accuracy": 0.7777777777777778, | |
"eval_loss": 1.1536939144134521, | |
"eval_runtime": 0.6119, | |
"eval_samples_per_second": 117.657, | |
"eval_steps_per_second": 8.171, | |
"step": 69 | |
}, | |
{ | |
"epoch": 7.777777777777778, | |
"grad_norm": 0.03466796875, | |
"learning_rate": 1.1363636363636365e-05, | |
"loss": 0.0002, | |
"step": 70 | |
}, | |
{ | |
"epoch": 7.777777777777778, | |
"eval_accuracy": 0.7916666666666666, | |
"eval_loss": 1.1577788591384888, | |
"eval_runtime": 0.6119, | |
"eval_samples_per_second": 117.668, | |
"eval_steps_per_second": 8.171, | |
"step": 70 | |
}, | |
{ | |
"epoch": 7.888888888888889, | |
"grad_norm": 0.08349609375, | |
"learning_rate": 1.0795454545454547e-05, | |
"loss": 0.0006, | |
"step": 71 | |
}, | |
{ | |
"epoch": 7.888888888888889, | |
"eval_accuracy": 0.8055555555555556, | |
"eval_loss": 1.1397895812988281, | |
"eval_runtime": 0.7129, | |
"eval_samples_per_second": 100.999, | |
"eval_steps_per_second": 7.014, | |
"step": 71 | |
}, | |
{ | |
"epoch": 8.0, | |
"grad_norm": 0.04443359375, | |
"learning_rate": 1.0227272727272729e-05, | |
"loss": 0.0004, | |
"step": 72 | |
}, | |
{ | |
"epoch": 8.0, | |
"eval_accuracy": 0.8055555555555556, | |
"eval_loss": 1.1536896228790283, | |
"eval_runtime": 0.7131, | |
"eval_samples_per_second": 100.962, | |
"eval_steps_per_second": 7.011, | |
"step": 72 | |
}, | |
{ | |
"epoch": 8.11111111111111, | |
"grad_norm": 0.02392578125, | |
"learning_rate": 9.659090909090909e-06, | |
"loss": 0.0002, | |
"step": 73 | |
}, | |
{ | |
"epoch": 8.11111111111111, | |
"eval_accuracy": 0.8055555555555556, | |
"eval_loss": 1.1682804822921753, | |
"eval_runtime": 0.6627, | |
"eval_samples_per_second": 108.648, | |
"eval_steps_per_second": 7.545, | |
"step": 73 | |
}, | |
{ | |
"epoch": 8.222222222222221, | |
"grad_norm": 0.036865234375, | |
"learning_rate": 9.090909090909091e-06, | |
"loss": 0.0004, | |
"step": 74 | |
}, | |
{ | |
"epoch": 8.222222222222221, | |
"eval_accuracy": 0.8055555555555556, | |
"eval_loss": 1.1590551137924194, | |
"eval_runtime": 0.6635, | |
"eval_samples_per_second": 108.51, | |
"eval_steps_per_second": 7.535, | |
"step": 74 | |
}, | |
{ | |
"epoch": 8.333333333333334, | |
"grad_norm": 0.0213623046875, | |
"learning_rate": 8.522727272727273e-06, | |
"loss": 0.0003, | |
"step": 75 | |
}, | |
{ | |
"epoch": 8.333333333333334, | |
"eval_accuracy": 0.8055555555555556, | |
"eval_loss": 1.1632962226867676, | |
"eval_runtime": 0.6116, | |
"eval_samples_per_second": 117.718, | |
"eval_steps_per_second": 8.175, | |
"step": 75 | |
}, | |
{ | |
"epoch": 8.444444444444445, | |
"grad_norm": 0.08740234375, | |
"learning_rate": 7.954545454545455e-06, | |
"loss": 0.0004, | |
"step": 76 | |
}, | |
{ | |
"epoch": 8.444444444444445, | |
"eval_accuracy": 0.8055555555555556, | |
"eval_loss": 1.1668634414672852, | |
"eval_runtime": 0.611, | |
"eval_samples_per_second": 117.842, | |
"eval_steps_per_second": 8.183, | |
"step": 76 | |
}, | |
{ | |
"epoch": 8.555555555555555, | |
"grad_norm": 0.011474609375, | |
"learning_rate": 7.386363636363637e-06, | |
"loss": 0.0002, | |
"step": 77 | |
}, | |
{ | |
"epoch": 8.555555555555555, | |
"eval_accuracy": 0.8055555555555556, | |
"eval_loss": 1.1664338111877441, | |
"eval_runtime": 0.5614, | |
"eval_samples_per_second": 128.259, | |
"eval_steps_per_second": 8.907, | |
"step": 77 | |
}, | |
{ | |
"epoch": 8.666666666666666, | |
"grad_norm": 0.0308837890625, | |
"learning_rate": 6.818181818181818e-06, | |
"loss": 0.0002, | |
"step": 78 | |
}, | |
{ | |
"epoch": 8.666666666666666, | |
"eval_accuracy": 0.8055555555555556, | |
"eval_loss": 1.167876958847046, | |
"eval_runtime": 0.6112, | |
"eval_samples_per_second": 117.795, | |
"eval_steps_per_second": 8.18, | |
"step": 78 | |
}, | |
{ | |
"epoch": 8.777777777777779, | |
"grad_norm": 0.033935546875, | |
"learning_rate": 6.25e-06, | |
"loss": 0.0003, | |
"step": 79 | |
}, | |
{ | |
"epoch": 8.777777777777779, | |
"eval_accuracy": 0.8055555555555556, | |
"eval_loss": 1.153360366821289, | |
"eval_runtime": 0.712, | |
"eval_samples_per_second": 101.127, | |
"eval_steps_per_second": 7.023, | |
"step": 79 | |
}, | |
{ | |
"epoch": 8.88888888888889, | |
"grad_norm": 0.048828125, | |
"learning_rate": 5.681818181818182e-06, | |
"loss": 0.0003, | |
"step": 80 | |
}, | |
{ | |
"epoch": 8.88888888888889, | |
"eval_accuracy": 0.8055555555555556, | |
"eval_loss": 1.1648998260498047, | |
"eval_runtime": 0.6123, | |
"eval_samples_per_second": 117.581, | |
"eval_steps_per_second": 8.165, | |
"step": 80 | |
}, | |
{ | |
"epoch": 9.0, | |
"grad_norm": 0.0133056640625, | |
"learning_rate": 5.113636363636364e-06, | |
"loss": 0.0001, | |
"step": 81 | |
}, | |
{ | |
"epoch": 9.0, | |
"eval_accuracy": 0.8055555555555556, | |
"eval_loss": 1.1657260656356812, | |
"eval_runtime": 0.5613, | |
"eval_samples_per_second": 128.281, | |
"eval_steps_per_second": 8.908, | |
"step": 81 | |
}, | |
{ | |
"epoch": 9.11111111111111, | |
"grad_norm": 0.023193359375, | |
"learning_rate": 4.5454545454545455e-06, | |
"loss": 0.0001, | |
"step": 82 | |
}, | |
{ | |
"epoch": 9.11111111111111, | |
"eval_accuracy": 0.8055555555555556, | |
"eval_loss": 1.1588553190231323, | |
"eval_runtime": 0.6632, | |
"eval_samples_per_second": 108.563, | |
"eval_steps_per_second": 7.539, | |
"step": 82 | |
}, | |
{ | |
"epoch": 9.222222222222221, | |
"grad_norm": 0.028076171875, | |
"learning_rate": 3.9772727272727275e-06, | |
"loss": 0.0003, | |
"step": 83 | |
}, | |
{ | |
"epoch": 9.222222222222221, | |
"eval_accuracy": 0.8055555555555556, | |
"eval_loss": 1.1638314723968506, | |
"eval_runtime": 0.6615, | |
"eval_samples_per_second": 108.842, | |
"eval_steps_per_second": 7.558, | |
"step": 83 | |
}, | |
{ | |
"epoch": 9.333333333333334, | |
"grad_norm": 0.041259765625, | |
"learning_rate": 3.409090909090909e-06, | |
"loss": 0.0002, | |
"step": 84 | |
}, | |
{ | |
"epoch": 9.333333333333334, | |
"eval_accuracy": 0.8055555555555556, | |
"eval_loss": 1.1614993810653687, | |
"eval_runtime": 0.563, | |
"eval_samples_per_second": 127.875, | |
"eval_steps_per_second": 8.88, | |
"step": 84 | |
}, | |
{ | |
"epoch": 9.444444444444445, | |
"grad_norm": 0.033447265625, | |
"learning_rate": 2.840909090909091e-06, | |
"loss": 0.0002, | |
"step": 85 | |
}, | |
{ | |
"epoch": 9.444444444444445, | |
"eval_accuracy": 0.8055555555555556, | |
"eval_loss": 1.1663289070129395, | |
"eval_runtime": 0.6624, | |
"eval_samples_per_second": 108.689, | |
"eval_steps_per_second": 7.548, | |
"step": 85 | |
}, | |
{ | |
"epoch": 9.555555555555555, | |
"grad_norm": 0.060791015625, | |
"learning_rate": 2.2727272727272728e-06, | |
"loss": 0.0004, | |
"step": 86 | |
}, | |
{ | |
"epoch": 9.555555555555555, | |
"eval_accuracy": 0.8055555555555556, | |
"eval_loss": 1.1687304973602295, | |
"eval_runtime": 0.6624, | |
"eval_samples_per_second": 108.698, | |
"eval_steps_per_second": 7.548, | |
"step": 86 | |
}, | |
{ | |
"epoch": 9.666666666666666, | |
"grad_norm": 0.01275634765625, | |
"learning_rate": 1.7045454545454546e-06, | |
"loss": 0.0001, | |
"step": 87 | |
}, | |
{ | |
"epoch": 9.666666666666666, | |
"eval_accuracy": 0.8055555555555556, | |
"eval_loss": 1.170520544052124, | |
"eval_runtime": 0.6618, | |
"eval_samples_per_second": 108.798, | |
"eval_steps_per_second": 7.555, | |
"step": 87 | |
}, | |
{ | |
"epoch": 9.777777777777779, | |
"grad_norm": 0.0166015625, | |
"learning_rate": 1.1363636363636364e-06, | |
"loss": 0.0001, | |
"step": 88 | |
}, | |
{ | |
"epoch": 9.777777777777779, | |
"eval_accuracy": 0.8055555555555556, | |
"eval_loss": 1.1710082292556763, | |
"eval_runtime": 0.612, | |
"eval_samples_per_second": 117.65, | |
"eval_steps_per_second": 8.17, | |
"step": 88 | |
}, | |
{ | |
"epoch": 9.88888888888889, | |
"grad_norm": 0.01708984375, | |
"learning_rate": 5.681818181818182e-07, | |
"loss": 0.0002, | |
"step": 89 | |
}, | |
{ | |
"epoch": 9.88888888888889, | |
"eval_accuracy": 0.8055555555555556, | |
"eval_loss": 1.160547137260437, | |
"eval_runtime": 0.5621, | |
"eval_samples_per_second": 128.102, | |
"eval_steps_per_second": 8.896, | |
"step": 89 | |
}, | |
{ | |
"epoch": 10.0, | |
"grad_norm": 0.0625, | |
"learning_rate": 0.0, | |
"loss": 0.0003, | |
"step": 90 | |
}, | |
{ | |
"epoch": 10.0, | |
"eval_accuracy": 0.8055555555555556, | |
"eval_loss": 1.156036138534546, | |
"eval_runtime": 0.6629, | |
"eval_samples_per_second": 108.621, | |
"eval_steps_per_second": 7.543, | |
"step": 90 | |
}, | |
{ | |
"epoch": 10.0, | |
"step": 90, | |
"total_flos": 5016736558481408.0, | |
"train_loss": 0.24542952523463302, | |
"train_runtime": 139.7817, | |
"train_samples_per_second": 20.317, | |
"train_steps_per_second": 0.644 | |
} | |
], | |
"logging_steps": 1, | |
"max_steps": 90, | |
"num_input_tokens_seen": 0, | |
"num_train_epochs": 10, | |
"save_steps": 500, | |
"stateful_callbacks": { | |
"TrainerControl": { | |
"args": { | |
"should_epoch_stop": false, | |
"should_evaluate": false, | |
"should_log": false, | |
"should_save": false, | |
"should_training_stop": false | |
}, | |
"attributes": {} | |
} | |
}, | |
"total_flos": 5016736558481408.0, | |
"train_batch_size": 4, | |
"trial_name": null, | |
"trial_params": null | |
} | |