| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 46.15384615384615, | |
| "eval_steps": 50, | |
| "global_step": 600, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.07692307692307693, | |
| "eval_loss": 3.9168343544006348, | |
| "eval_runtime": 5.6922, | |
| "eval_samples_per_second": 263.695, | |
| "eval_steps_per_second": 4.216, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "grad_norm": 1.3671875, | |
| "learning_rate": 6.666666666666667e-05, | |
| "loss": 3.3633, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "grad_norm": 1.09375, | |
| "learning_rate": 0.00013333333333333334, | |
| "loss": 3.1311, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.3076923076923075, | |
| "grad_norm": 0.8203125, | |
| "learning_rate": 0.0002, | |
| "loss": 2.9153, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 3.076923076923077, | |
| "grad_norm": 0.7265625, | |
| "learning_rate": 0.00019984815164333163, | |
| "loss": 2.7539, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 3.8461538461538463, | |
| "grad_norm": 0.71875, | |
| "learning_rate": 0.00019939306773179497, | |
| "loss": 2.5978, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 3.8461538461538463, | |
| "eval_loss": 2.814912796020508, | |
| "eval_runtime": 5.762, | |
| "eval_samples_per_second": 260.499, | |
| "eval_steps_per_second": 4.165, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 4.615384615384615, | |
| "grad_norm": 0.71484375, | |
| "learning_rate": 0.00019863613034027224, | |
| "loss": 2.4698, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 5.384615384615385, | |
| "grad_norm": 0.7734375, | |
| "learning_rate": 0.00019757963826274357, | |
| "loss": 2.3909, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 6.153846153846154, | |
| "grad_norm": 0.765625, | |
| "learning_rate": 0.00019622680003092503, | |
| "loss": 2.2723, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 6.923076923076923, | |
| "grad_norm": 0.796875, | |
| "learning_rate": 0.00019458172417006347, | |
| "loss": 2.1816, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 7.6923076923076925, | |
| "grad_norm": 0.8203125, | |
| "learning_rate": 0.00019264940672148018, | |
| "loss": 2.0808, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 7.6923076923076925, | |
| "eval_loss": 2.966378927230835, | |
| "eval_runtime": 5.6204, | |
| "eval_samples_per_second": 267.064, | |
| "eval_steps_per_second": 4.27, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 8.461538461538462, | |
| "grad_norm": 0.83984375, | |
| "learning_rate": 0.00019043571606975777, | |
| "loss": 1.9902, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 9.23076923076923, | |
| "grad_norm": 0.99609375, | |
| "learning_rate": 0.0001879473751206489, | |
| "loss": 1.9033, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.83984375, | |
| "learning_rate": 0.00018519194088383273, | |
| "loss": 1.8173, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 10.76923076923077, | |
| "grad_norm": 0.87890625, | |
| "learning_rate": 0.0001821777815225245, | |
| "loss": 1.7058, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 11.538461538461538, | |
| "grad_norm": 0.9296875, | |
| "learning_rate": 0.00017891405093963938, | |
| "loss": 1.6294, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 11.538461538461538, | |
| "eval_loss": 3.233675003051758, | |
| "eval_runtime": 5.5692, | |
| "eval_samples_per_second": 269.517, | |
| "eval_steps_per_second": 4.309, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 12.307692307692308, | |
| "grad_norm": 0.9921875, | |
| "learning_rate": 0.00017541066097768963, | |
| "loss": 1.5563, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 13.076923076923077, | |
| "grad_norm": 1.0546875, | |
| "learning_rate": 0.00017167825131684513, | |
| "loss": 1.4865, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 13.846153846153847, | |
| "grad_norm": 0.94921875, | |
| "learning_rate": 0.00016772815716257412, | |
| "loss": 1.3938, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 14.615384615384615, | |
| "grad_norm": 0.90625, | |
| "learning_rate": 0.00016357237482099684, | |
| "loss": 1.3247, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 15.384615384615385, | |
| "grad_norm": 0.99609375, | |
| "learning_rate": 0.00015922352526649803, | |
| "loss": 1.2699, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 15.384615384615385, | |
| "eval_loss": 3.521737813949585, | |
| "eval_runtime": 5.6468, | |
| "eval_samples_per_second": 265.815, | |
| "eval_steps_per_second": 4.25, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 16.153846153846153, | |
| "grad_norm": 0.95703125, | |
| "learning_rate": 0.00015469481581224272, | |
| "loss": 1.2117, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 16.923076923076923, | |
| "grad_norm": 1.0390625, | |
| "learning_rate": 0.00015000000000000001, | |
| "loss": 1.1498, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 17.692307692307693, | |
| "grad_norm": 1.0078125, | |
| "learning_rate": 0.00014515333583108896, | |
| "loss": 1.0864, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 18.46153846153846, | |
| "grad_norm": 0.81640625, | |
| "learning_rate": 0.00014016954246529696, | |
| "loss": 1.0441, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 19.23076923076923, | |
| "grad_norm": 0.98828125, | |
| "learning_rate": 0.00013506375551927547, | |
| "loss": 1.0092, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 19.23076923076923, | |
| "eval_loss": 3.726165533065796, | |
| "eval_runtime": 5.8207, | |
| "eval_samples_per_second": 257.872, | |
| "eval_steps_per_second": 4.123, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 0.8359375, | |
| "learning_rate": 0.00012985148110016947, | |
| "loss": 0.9671, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 20.76923076923077, | |
| "grad_norm": 0.8203125, | |
| "learning_rate": 0.00012454854871407994, | |
| "loss": 0.9182, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 21.53846153846154, | |
| "grad_norm": 0.796875, | |
| "learning_rate": 0.00011917106319237386, | |
| "loss": 0.8857, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 22.307692307692307, | |
| "grad_norm": 0.79296875, | |
| "learning_rate": 0.00011373535578184082, | |
| "loss": 0.8623, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 23.076923076923077, | |
| "grad_norm": 0.7421875, | |
| "learning_rate": 0.00010825793454723325, | |
| "loss": 0.8392, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 23.076923076923077, | |
| "eval_loss": 3.868284225463867, | |
| "eval_runtime": 5.7191, | |
| "eval_samples_per_second": 262.455, | |
| "eval_steps_per_second": 4.196, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 23.846153846153847, | |
| "grad_norm": 0.75, | |
| "learning_rate": 0.00010275543423681621, | |
| "loss": 0.8103, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 24.615384615384617, | |
| "grad_norm": 0.75390625, | |
| "learning_rate": 9.724456576318381e-05, | |
| "loss": 0.7828, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 25.384615384615383, | |
| "grad_norm": 0.7265625, | |
| "learning_rate": 9.174206545276677e-05, | |
| "loss": 0.7749, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 26.153846153846153, | |
| "grad_norm": 0.74609375, | |
| "learning_rate": 8.626464421815919e-05, | |
| "loss": 0.7574, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 26.923076923076923, | |
| "grad_norm": 0.6875, | |
| "learning_rate": 8.082893680762619e-05, | |
| "loss": 0.7428, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 26.923076923076923, | |
| "eval_loss": 3.9434773921966553, | |
| "eval_runtime": 5.6325, | |
| "eval_samples_per_second": 266.49, | |
| "eval_steps_per_second": 4.261, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 27.692307692307693, | |
| "grad_norm": 0.6953125, | |
| "learning_rate": 7.54514512859201e-05, | |
| "loss": 0.7298, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 28.46153846153846, | |
| "grad_norm": 0.66015625, | |
| "learning_rate": 7.014851889983057e-05, | |
| "loss": 0.7167, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 29.23076923076923, | |
| "grad_norm": 0.65625, | |
| "learning_rate": 6.493624448072457e-05, | |
| "loss": 0.7147, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "grad_norm": 0.69140625, | |
| "learning_rate": 5.983045753470308e-05, | |
| "loss": 0.7019, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 30.76923076923077, | |
| "grad_norm": 0.66796875, | |
| "learning_rate": 5.484666416891109e-05, | |
| "loss": 0.6952, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 30.76923076923077, | |
| "eval_loss": 3.985978841781616, | |
| "eval_runtime": 5.6438, | |
| "eval_samples_per_second": 265.955, | |
| "eval_steps_per_second": 4.252, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 31.53846153846154, | |
| "grad_norm": 0.6484375, | |
| "learning_rate": 5.000000000000002e-05, | |
| "loss": 0.6901, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 32.30769230769231, | |
| "grad_norm": 0.640625, | |
| "learning_rate": 4.530518418775733e-05, | |
| "loss": 0.685, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 33.07692307692308, | |
| "grad_norm": 0.62890625, | |
| "learning_rate": 4.077647473350201e-05, | |
| "loss": 0.6851, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 33.84615384615385, | |
| "grad_norm": 0.62109375, | |
| "learning_rate": 3.642762517900322e-05, | |
| "loss": 0.6782, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 34.61538461538461, | |
| "grad_norm": 0.6171875, | |
| "learning_rate": 3.227184283742591e-05, | |
| "loss": 0.6762, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 34.61538461538461, | |
| "eval_loss": 3.998966932296753, | |
| "eval_runtime": 5.7156, | |
| "eval_samples_per_second": 262.613, | |
| "eval_steps_per_second": 4.199, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 35.38461538461539, | |
| "grad_norm": 0.62109375, | |
| "learning_rate": 2.8321748683154893e-05, | |
| "loss": 0.6742, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 36.15384615384615, | |
| "grad_norm": 0.6171875, | |
| "learning_rate": 2.4589339022310386e-05, | |
| "loss": 0.674, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 36.92307692307692, | |
| "grad_norm": 0.62109375, | |
| "learning_rate": 2.1085949060360654e-05, | |
| "loss": 0.6728, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 37.69230769230769, | |
| "grad_norm": 0.6171875, | |
| "learning_rate": 1.7822218477475494e-05, | |
| "loss": 0.6681, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 38.46153846153846, | |
| "grad_norm": 0.62109375, | |
| "learning_rate": 1.4808059116167305e-05, | |
| "loss": 0.6739, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 38.46153846153846, | |
| "eval_loss": 4.016704559326172, | |
| "eval_runtime": 5.9697, | |
| "eval_samples_per_second": 251.437, | |
| "eval_steps_per_second": 4.02, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 39.23076923076923, | |
| "grad_norm": 0.61328125, | |
| "learning_rate": 1.2052624879351104e-05, | |
| "loss": 0.6685, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "grad_norm": 0.625, | |
| "learning_rate": 9.564283930242257e-06, | |
| "loss": 0.6697, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 40.76923076923077, | |
| "grad_norm": 0.60546875, | |
| "learning_rate": 7.350593278519824e-06, | |
| "loss": 0.6691, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 41.53846153846154, | |
| "grad_norm": 0.609375, | |
| "learning_rate": 5.418275829936537e-06, | |
| "loss": 0.6709, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 42.30769230769231, | |
| "grad_norm": 0.62109375, | |
| "learning_rate": 3.7731999690749585e-06, | |
| "loss": 0.6691, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 42.30769230769231, | |
| "eval_loss": 4.020811080932617, | |
| "eval_runtime": 5.6558, | |
| "eval_samples_per_second": 265.392, | |
| "eval_steps_per_second": 4.243, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 43.07692307692308, | |
| "grad_norm": 0.61328125, | |
| "learning_rate": 2.420361737256438e-06, | |
| "loss": 0.6671, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 43.84615384615385, | |
| "grad_norm": 0.62890625, | |
| "learning_rate": 1.3638696597277679e-06, | |
| "loss": 0.6683, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 44.61538461538461, | |
| "grad_norm": 0.59375, | |
| "learning_rate": 6.069322682050516e-07, | |
| "loss": 0.6714, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 45.38461538461539, | |
| "grad_norm": 0.6015625, | |
| "learning_rate": 1.518483566683826e-07, | |
| "loss": 0.6695, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 46.15384615384615, | |
| "grad_norm": 0.62109375, | |
| "learning_rate": 0.0, | |
| "loss": 0.6667, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 46.15384615384615, | |
| "eval_loss": 4.010259628295898, | |
| "eval_runtime": 5.6787, | |
| "eval_samples_per_second": 264.32, | |
| "eval_steps_per_second": 4.226, | |
| "step": 600 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 600, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 47, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.05042678235136e+16, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |