| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.9719890036406866, | |
| "eval_steps": 4000, | |
| "global_step": 40000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.2733132243156433, | |
| "learning_rate": 5e-05, | |
| "loss": 0.3088, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.011344216763973236, | |
| "learning_rate": 4.873022322675674e-05, | |
| "loss": 0.0012, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 0.002656959230080247, | |
| "learning_rate": 4.746044645351347e-05, | |
| "loss": 0.0001, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 0.0012492843670770526, | |
| "learning_rate": 4.619066968027021e-05, | |
| "loss": 0.0, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "eval_loss": 5.678263278241502e-06, | |
| "eval_rouge2_fmeasure": 0.0, | |
| "eval_rouge2_precision": 0.0, | |
| "eval_rouge2_recall": 0.0, | |
| "eval_runtime": 3660.3045, | |
| "eval_samples_per_second": 2.739, | |
| "eval_steps_per_second": 0.171, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 0.0005404581315815449, | |
| "learning_rate": 4.4920892907026946e-05, | |
| "loss": 0.0009, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 0.002274229424074292, | |
| "learning_rate": 4.365111613378368e-05, | |
| "loss": 0.0001, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 0.001682213624008, | |
| "learning_rate": 4.2381339360540415e-05, | |
| "loss": 0.0001, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 0.00026362488279119134, | |
| "learning_rate": 4.1111562587297156e-05, | |
| "loss": 0.0002, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "eval_loss": 2.755745526883402e-06, | |
| "eval_rouge2_fmeasure": 0.0, | |
| "eval_rouge2_precision": 0.0, | |
| "eval_rouge2_recall": 0.0, | |
| "eval_runtime": 3669.0705, | |
| "eval_samples_per_second": 2.733, | |
| "eval_steps_per_second": 0.171, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 0.010094952769577503, | |
| "learning_rate": 3.984178581405389e-05, | |
| "loss": 0.0014, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 0.0005482935812324286, | |
| "learning_rate": 3.8572009040810624e-05, | |
| "loss": 0.0001, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 0.00010689393093343824, | |
| "learning_rate": 3.730223226756736e-05, | |
| "loss": 0.0, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 0.0003704932169057429, | |
| "learning_rate": 3.60324554943241e-05, | |
| "loss": 0.0006, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "eval_loss": 3.983559054177022e-06, | |
| "eval_rouge2_fmeasure": 0.0, | |
| "eval_rouge2_precision": 0.0, | |
| "eval_rouge2_recall": 0.0, | |
| "eval_runtime": 3661.9564, | |
| "eval_samples_per_second": 2.738, | |
| "eval_steps_per_second": 0.171, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 0.00011691737017827109, | |
| "learning_rate": 3.476267872108084e-05, | |
| "loss": 0.0, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 8.349567360710353e-05, | |
| "learning_rate": 3.3492901947837574e-05, | |
| "loss": 0.0001, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "grad_norm": 0.0001232757931575179, | |
| "learning_rate": 3.222312517459431e-05, | |
| "loss": 0.0001, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "grad_norm": 0.00011773507139878348, | |
| "learning_rate": 3.095334840135104e-05, | |
| "loss": 0.0, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "eval_loss": 1.058370685314003e-06, | |
| "eval_rouge2_fmeasure": 0.0, | |
| "eval_rouge2_precision": 0.0, | |
| "eval_rouge2_recall": 0.0, | |
| "eval_runtime": 3660.4191, | |
| "eval_samples_per_second": 2.739, | |
| "eval_steps_per_second": 0.171, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "grad_norm": 0.0022140583023428917, | |
| "learning_rate": 2.968357162810778e-05, | |
| "loss": 0.0, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "grad_norm": 3.650138387456536e-05, | |
| "learning_rate": 2.8413794854864518e-05, | |
| "loss": 0.0, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "grad_norm": 0.010362447239458561, | |
| "learning_rate": 2.7144018081621252e-05, | |
| "loss": 0.0001, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "grad_norm": 0.005041338969022036, | |
| "learning_rate": 2.587424130837799e-05, | |
| "loss": 0.0001, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "eval_loss": 6.733321242791135e-06, | |
| "eval_rouge2_fmeasure": 0.0, | |
| "eval_rouge2_precision": 0.0, | |
| "eval_rouge2_recall": 0.0, | |
| "eval_runtime": 3656.0835, | |
| "eval_samples_per_second": 2.742, | |
| "eval_steps_per_second": 0.171, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": 6.371775089064613e-05, | |
| "learning_rate": 2.4604464535134724e-05, | |
| "loss": 0.0, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "grad_norm": 0.0006053569377399981, | |
| "learning_rate": 2.3334687761891462e-05, | |
| "loss": 0.0001, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "grad_norm": 0.0021953503601253033, | |
| "learning_rate": 2.2064910988648196e-05, | |
| "loss": 0.0, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "grad_norm": 2.853705154848285e-05, | |
| "learning_rate": 2.0795134215404934e-05, | |
| "loss": 0.0, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "eval_loss": 4.6661673991366115e-07, | |
| "eval_rouge2_fmeasure": 0.0, | |
| "eval_rouge2_precision": 0.0, | |
| "eval_rouge2_recall": 0.0, | |
| "eval_runtime": 3661.0043, | |
| "eval_samples_per_second": 2.739, | |
| "eval_steps_per_second": 0.171, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "grad_norm": 6.040197695256211e-05, | |
| "learning_rate": 1.9525357442161668e-05, | |
| "loss": 0.0, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "grad_norm": 0.00013890626723878086, | |
| "learning_rate": 1.8255580668918406e-05, | |
| "loss": 0.0, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "grad_norm": 2.3448434149031527e-05, | |
| "learning_rate": 1.698580389567514e-05, | |
| "loss": 0.0, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 0.00018542050383985043, | |
| "learning_rate": 1.5716027122431877e-05, | |
| "loss": 0.0, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "eval_loss": 1.1049346539948601e-06, | |
| "eval_rouge2_fmeasure": 0.0001, | |
| "eval_rouge2_precision": 0.0, | |
| "eval_rouge2_recall": 0.0001, | |
| "eval_runtime": 3659.7337, | |
| "eval_samples_per_second": 2.74, | |
| "eval_steps_per_second": 0.171, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 2.15, | |
| "grad_norm": 1.2732043614960276e-05, | |
| "learning_rate": 1.4446250349188615e-05, | |
| "loss": 0.0, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 2.23, | |
| "grad_norm": 1.618100213818252e-05, | |
| "learning_rate": 1.317647357594535e-05, | |
| "loss": 0.0, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "grad_norm": 5.8585923397913575e-05, | |
| "learning_rate": 1.1906696802702085e-05, | |
| "loss": 0.0, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 2.38, | |
| "grad_norm": 0.00015024282038211823, | |
| "learning_rate": 1.0636920029458821e-05, | |
| "loss": 0.0004, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 2.38, | |
| "eval_loss": 1.2435020835255273e-06, | |
| "eval_rouge2_fmeasure": 0.0, | |
| "eval_rouge2_precision": 0.0001, | |
| "eval_rouge2_recall": 0.0, | |
| "eval_runtime": 3675.9943, | |
| "eval_samples_per_second": 2.727, | |
| "eval_steps_per_second": 0.171, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "grad_norm": 4.083578460267745e-05, | |
| "learning_rate": 9.367143256215557e-06, | |
| "loss": 0.0, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "grad_norm": 8.640537998871878e-06, | |
| "learning_rate": 8.097366482972293e-06, | |
| "loss": 0.0, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 1.1439586160122417e-05, | |
| "learning_rate": 6.8275897097290306e-06, | |
| "loss": 0.0, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "grad_norm": 0.00013945900718681514, | |
| "learning_rate": 5.557812936485766e-06, | |
| "loss": 0.0, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "eval_loss": 2.3348137290213344e-07, | |
| "eval_rouge2_fmeasure": 0.0, | |
| "eval_rouge2_precision": 0.0, | |
| "eval_rouge2_recall": 0.0, | |
| "eval_runtime": 3659.0736, | |
| "eval_samples_per_second": 2.74, | |
| "eval_steps_per_second": 0.171, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 2.75, | |
| "grad_norm": 0.0003026222693733871, | |
| "learning_rate": 4.288036163242502e-06, | |
| "loss": 0.0, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 2.82, | |
| "grad_norm": 6.744077109033242e-05, | |
| "learning_rate": 3.018259389999238e-06, | |
| "loss": 0.0, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "grad_norm": 3.917148569598794e-05, | |
| "learning_rate": 1.7484826167559743e-06, | |
| "loss": 0.0, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 2.97, | |
| "grad_norm": 1.1611381523835007e-05, | |
| "learning_rate": 4.787058435127105e-07, | |
| "loss": 0.0, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 2.97, | |
| "eval_loss": 1.7171299759866088e-07, | |
| "eval_rouge2_fmeasure": 0.0, | |
| "eval_rouge2_precision": 0.0, | |
| "eval_rouge2_recall": 0.0001, | |
| "eval_runtime": 3666.3048, | |
| "eval_samples_per_second": 2.735, | |
| "eval_steps_per_second": 0.171, | |
| "step": 40000 | |
| } | |
| ], | |
| "logging_steps": 1000, | |
| "max_steps": 40377, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "total_flos": 3.9259958213339136e+17, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |