| { | |
| "best_metric": 0.6337994337081909, | |
| "best_model_checkpoint": "finetuned-bangladeshi-traditional-food/checkpoint-320", | |
| "epoch": 10.0, | |
| "eval_steps": 500, | |
| "global_step": 640, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.15625, | |
| "grad_norm": 260322.40625, | |
| "learning_rate": 0.000196875, | |
| "loss": 0.1664, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.3125, | |
| "grad_norm": 9524.54296875, | |
| "learning_rate": 0.00019375000000000002, | |
| "loss": 0.132, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.46875, | |
| "grad_norm": 232655.25, | |
| "learning_rate": 0.000190625, | |
| "loss": 0.2052, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.625, | |
| "grad_norm": 2797.121337890625, | |
| "learning_rate": 0.0001875, | |
| "loss": 0.1608, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.78125, | |
| "grad_norm": 94306.1328125, | |
| "learning_rate": 0.000184375, | |
| "loss": 0.0879, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.9375, | |
| "grad_norm": 94746.8828125, | |
| "learning_rate": 0.00018125000000000001, | |
| "loss": 0.2159, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.840782122905028, | |
| "eval_loss": 0.6425307393074036, | |
| "eval_runtime": 6.1943, | |
| "eval_samples_per_second": 57.795, | |
| "eval_steps_per_second": 3.713, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.09375, | |
| "grad_norm": 224044.515625, | |
| "learning_rate": 0.000178125, | |
| "loss": 0.1128, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 179013.890625, | |
| "learning_rate": 0.000175, | |
| "loss": 0.1238, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.40625, | |
| "grad_norm": 322362.1875, | |
| "learning_rate": 0.00017187500000000002, | |
| "loss": 0.0767, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.5625, | |
| "grad_norm": 486709.90625, | |
| "learning_rate": 0.00016875, | |
| "loss": 0.1934, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.71875, | |
| "grad_norm": 339304.5, | |
| "learning_rate": 0.000165625, | |
| "loss": 0.1545, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.875, | |
| "grad_norm": 400540.96875, | |
| "learning_rate": 0.00016250000000000002, | |
| "loss": 0.1968, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8547486033519553, | |
| "eval_loss": 0.6727585792541504, | |
| "eval_runtime": 6.1785, | |
| "eval_samples_per_second": 57.943, | |
| "eval_steps_per_second": 3.723, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 2.03125, | |
| "grad_norm": 402626.6875, | |
| "learning_rate": 0.000159375, | |
| "loss": 0.1303, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.1875, | |
| "grad_norm": 84834.3515625, | |
| "learning_rate": 0.00015625, | |
| "loss": 0.1479, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.34375, | |
| "grad_norm": 120292.5703125, | |
| "learning_rate": 0.000153125, | |
| "loss": 0.1937, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 20186.486328125, | |
| "learning_rate": 0.00015000000000000001, | |
| "loss": 0.0804, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.65625, | |
| "grad_norm": 173780.65625, | |
| "learning_rate": 0.000146875, | |
| "loss": 0.0843, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.8125, | |
| "grad_norm": 46256.0546875, | |
| "learning_rate": 0.00014375, | |
| "loss": 0.0831, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.96875, | |
| "grad_norm": 165973.46875, | |
| "learning_rate": 0.00014062500000000002, | |
| "loss": 0.0686, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8631284916201117, | |
| "eval_loss": 0.71826171875, | |
| "eval_runtime": 6.475, | |
| "eval_samples_per_second": 55.29, | |
| "eval_steps_per_second": 3.552, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 3.125, | |
| "grad_norm": 811.771240234375, | |
| "learning_rate": 0.0001375, | |
| "loss": 0.0824, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.28125, | |
| "grad_norm": 421833.125, | |
| "learning_rate": 0.000134375, | |
| "loss": 0.1451, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 3.4375, | |
| "grad_norm": 347591.78125, | |
| "learning_rate": 0.00013125000000000002, | |
| "loss": 0.1193, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 3.59375, | |
| "grad_norm": 150345.671875, | |
| "learning_rate": 0.000128125, | |
| "loss": 0.1245, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 3.75, | |
| "grad_norm": 370.7461853027344, | |
| "learning_rate": 0.000125, | |
| "loss": 0.0934, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 3.90625, | |
| "grad_norm": 583907.0, | |
| "learning_rate": 0.00012187500000000001, | |
| "loss": 0.0975, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8463687150837989, | |
| "eval_loss": 0.8238765597343445, | |
| "eval_runtime": 6.2123, | |
| "eval_samples_per_second": 57.628, | |
| "eval_steps_per_second": 3.702, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 4.0625, | |
| "grad_norm": 75910.390625, | |
| "learning_rate": 0.00011875, | |
| "loss": 0.0952, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 4.21875, | |
| "grad_norm": 483136.6875, | |
| "learning_rate": 0.000115625, | |
| "loss": 0.0867, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 4.375, | |
| "grad_norm": 4663.18310546875, | |
| "learning_rate": 0.00011250000000000001, | |
| "loss": 0.0599, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 4.53125, | |
| "grad_norm": 1170.08251953125, | |
| "learning_rate": 0.000109375, | |
| "loss": 0.0671, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 4.6875, | |
| "grad_norm": 286.7016296386719, | |
| "learning_rate": 0.00010625000000000001, | |
| "loss": 0.1158, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 4.84375, | |
| "grad_norm": 237.0006561279297, | |
| "learning_rate": 0.000103125, | |
| "loss": 0.0821, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 137.74867248535156, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0399, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.888268156424581, | |
| "eval_loss": 0.6337994337081909, | |
| "eval_runtime": 6.2754, | |
| "eval_samples_per_second": 57.048, | |
| "eval_steps_per_second": 3.665, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 5.15625, | |
| "grad_norm": 1214.9869384765625, | |
| "learning_rate": 9.687500000000001e-05, | |
| "loss": 0.1064, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 5.3125, | |
| "grad_norm": 534309.6875, | |
| "learning_rate": 9.375e-05, | |
| "loss": 0.0826, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 5.46875, | |
| "grad_norm": 2779.36669921875, | |
| "learning_rate": 9.062500000000001e-05, | |
| "loss": 0.0524, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 5.625, | |
| "grad_norm": 2379.431640625, | |
| "learning_rate": 8.75e-05, | |
| "loss": 0.059, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 5.78125, | |
| "grad_norm": 567636.125, | |
| "learning_rate": 8.4375e-05, | |
| "loss": 0.1303, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 5.9375, | |
| "grad_norm": 582.1480712890625, | |
| "learning_rate": 8.125000000000001e-05, | |
| "loss": 0.0221, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.888268156424581, | |
| "eval_loss": 0.7707696557044983, | |
| "eval_runtime": 6.201, | |
| "eval_samples_per_second": 57.732, | |
| "eval_steps_per_second": 3.709, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 6.09375, | |
| "grad_norm": 60.556556701660156, | |
| "learning_rate": 7.8125e-05, | |
| "loss": 0.0341, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 6.25, | |
| "grad_norm": 59.626529693603516, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 0.047, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 6.40625, | |
| "grad_norm": 303419.21875, | |
| "learning_rate": 7.1875e-05, | |
| "loss": 0.065, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 6.5625, | |
| "grad_norm": 90.21700286865234, | |
| "learning_rate": 6.875e-05, | |
| "loss": 0.0541, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 6.71875, | |
| "grad_norm": 801652.9375, | |
| "learning_rate": 6.562500000000001e-05, | |
| "loss": 0.1373, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 6.875, | |
| "grad_norm": 281023.09375, | |
| "learning_rate": 6.25e-05, | |
| "loss": 0.0202, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.8910614525139665, | |
| "eval_loss": 0.7648816704750061, | |
| "eval_runtime": 6.2429, | |
| "eval_samples_per_second": 57.345, | |
| "eval_steps_per_second": 3.684, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 7.03125, | |
| "grad_norm": 43.21717834472656, | |
| "learning_rate": 5.9375e-05, | |
| "loss": 0.0798, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 7.1875, | |
| "grad_norm": 622758.6875, | |
| "learning_rate": 5.6250000000000005e-05, | |
| "loss": 0.0585, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 7.34375, | |
| "grad_norm": 4335.7919921875, | |
| "learning_rate": 5.3125000000000004e-05, | |
| "loss": 0.0967, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 7.5, | |
| "grad_norm": 56.46971130371094, | |
| "learning_rate": 5e-05, | |
| "loss": 0.0412, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 7.65625, | |
| "grad_norm": 26.425151824951172, | |
| "learning_rate": 4.6875e-05, | |
| "loss": 0.0834, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 7.8125, | |
| "grad_norm": 2260.0234375, | |
| "learning_rate": 4.375e-05, | |
| "loss": 0.0237, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 7.96875, | |
| "grad_norm": 55191.5859375, | |
| "learning_rate": 4.0625000000000005e-05, | |
| "loss": 0.0274, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.9162011173184358, | |
| "eval_loss": 0.7099100351333618, | |
| "eval_runtime": 6.3314, | |
| "eval_samples_per_second": 56.543, | |
| "eval_steps_per_second": 3.633, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 8.125, | |
| "grad_norm": 34213.54296875, | |
| "learning_rate": 3.7500000000000003e-05, | |
| "loss": 0.0295, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 8.28125, | |
| "grad_norm": 35.443115234375, | |
| "learning_rate": 3.4375e-05, | |
| "loss": 0.0814, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 8.4375, | |
| "grad_norm": 141960.859375, | |
| "learning_rate": 3.125e-05, | |
| "loss": 0.0377, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 8.59375, | |
| "grad_norm": 40429.1171875, | |
| "learning_rate": 2.8125000000000003e-05, | |
| "loss": 0.0481, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 8.75, | |
| "grad_norm": 11.297977447509766, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.0573, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 8.90625, | |
| "grad_norm": 12.592033386230469, | |
| "learning_rate": 2.1875e-05, | |
| "loss": 0.0507, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.9162011173184358, | |
| "eval_loss": 0.7499300837516785, | |
| "eval_runtime": 6.5756, | |
| "eval_samples_per_second": 54.444, | |
| "eval_steps_per_second": 3.498, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 9.0625, | |
| "grad_norm": 38.957550048828125, | |
| "learning_rate": 1.8750000000000002e-05, | |
| "loss": 0.0137, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 9.21875, | |
| "grad_norm": 37808.00390625, | |
| "learning_rate": 1.5625e-05, | |
| "loss": 0.1018, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 9.375, | |
| "grad_norm": 35257.2890625, | |
| "learning_rate": 1.25e-05, | |
| "loss": 0.0305, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 9.53125, | |
| "grad_norm": 14.713729858398438, | |
| "learning_rate": 9.375000000000001e-06, | |
| "loss": 0.0538, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 9.6875, | |
| "grad_norm": 123944.28125, | |
| "learning_rate": 6.25e-06, | |
| "loss": 0.0499, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 9.84375, | |
| "grad_norm": 199116.203125, | |
| "learning_rate": 3.125e-06, | |
| "loss": 0.1264, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 156690.1875, | |
| "learning_rate": 0.0, | |
| "loss": 0.0653, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.9162011173184358, | |
| "eval_loss": 0.7513992786407471, | |
| "eval_runtime": 6.3086, | |
| "eval_samples_per_second": 56.748, | |
| "eval_steps_per_second": 3.646, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 640, | |
| "total_flos": 1.571762912487506e+18, | |
| "train_loss": 0.0904881817754358, | |
| "train_runtime": 704.9996, | |
| "train_samples_per_second": 28.766, | |
| "train_steps_per_second": 0.908 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 640, | |
| "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": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.571762912487506e+18, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |