| { | |
| "best_metric": 0.9052917957305908, | |
| "best_model_checkpoint": "./tst-translation-output/checkpoint-12000", | |
| "epoch": 12.126224156692057, | |
| "eval_steps": 3000, | |
| "global_step": 39000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 9.712509712509713e-06, | |
| "loss": 1.9995, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 1.9425019425019425e-05, | |
| "loss": 1.4173, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 2.913752913752914e-05, | |
| "loss": 1.3137, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 3.885003885003885e-05, | |
| "loss": 1.2381, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 4.856254856254856e-05, | |
| "loss": 1.2023, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 4.991638730343166e-05, | |
| "loss": 1.1751, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "eval_bleu": 7.6766, | |
| "eval_gen_len": 17.4644, | |
| "eval_loss": 1.1288801431655884, | |
| "eval_runtime": 851.3289, | |
| "eval_samples_per_second": 15.112, | |
| "eval_steps_per_second": 0.946, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 4.981825033562843e-05, | |
| "loss": 1.1267, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 4.972011336782521e-05, | |
| "loss": 1.1013, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 4.962197640002198e-05, | |
| "loss": 1.0821, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 4.952383943221876e-05, | |
| "loss": 1.0599, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 4.9425702464415536e-05, | |
| "loss": 1.0336, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 4.9327565496612316e-05, | |
| "loss": 1.0268, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "eval_bleu": 9.2277, | |
| "eval_gen_len": 17.7668, | |
| "eval_loss": 0.9894970655441284, | |
| "eval_runtime": 891.3112, | |
| "eval_samples_per_second": 14.434, | |
| "eval_steps_per_second": 0.903, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "learning_rate": 4.922942852880909e-05, | |
| "loss": 0.9631, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "learning_rate": 4.913129156100586e-05, | |
| "loss": 0.8021, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "learning_rate": 4.903315459320265e-05, | |
| "loss": 0.8189, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "learning_rate": 4.893501762539942e-05, | |
| "loss": 0.8034, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "learning_rate": 4.8836880657596196e-05, | |
| "loss": 0.8156, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "learning_rate": 4.873874368979297e-05, | |
| "loss": 0.8075, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "eval_bleu": 9.1197, | |
| "eval_gen_len": 17.6811, | |
| "eval_loss": 0.945686399936676, | |
| "eval_runtime": 839.7466, | |
| "eval_samples_per_second": 15.32, | |
| "eval_steps_per_second": 0.959, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "learning_rate": 4.864060672198975e-05, | |
| "loss": 0.81, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "learning_rate": 4.854246975418653e-05, | |
| "loss": 0.8101, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "learning_rate": 4.84443327863833e-05, | |
| "loss": 0.7964, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "learning_rate": 4.8346195818580075e-05, | |
| "loss": 0.806, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "learning_rate": 4.8248058850776855e-05, | |
| "loss": 0.8042, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "learning_rate": 4.814992188297363e-05, | |
| "loss": 0.8082, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "eval_bleu": 8.4837, | |
| "eval_gen_len": 17.4826, | |
| "eval_loss": 0.9052917957305908, | |
| "eval_runtime": 837.7614, | |
| "eval_samples_per_second": 15.356, | |
| "eval_steps_per_second": 0.961, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "learning_rate": 4.805178491517041e-05, | |
| "loss": 0.8032, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "learning_rate": 4.795364794736719e-05, | |
| "loss": 0.7252, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "learning_rate": 4.785551097956396e-05, | |
| "loss": 0.5682, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 2.18, | |
| "learning_rate": 4.7757374011760734e-05, | |
| "loss": 0.5741, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 2.25, | |
| "learning_rate": 4.7659237043957514e-05, | |
| "loss": 0.5818, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 2.33, | |
| "learning_rate": 4.7561100076154294e-05, | |
| "loss": 0.5841, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 2.33, | |
| "eval_bleu": 9.8887, | |
| "eval_gen_len": 17.5166, | |
| "eval_loss": 0.9302938580513, | |
| "eval_runtime": 835.8268, | |
| "eval_samples_per_second": 15.392, | |
| "eval_steps_per_second": 0.963, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 2.41, | |
| "learning_rate": 4.746296310835107e-05, | |
| "loss": 0.5893, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 2.49, | |
| "learning_rate": 4.736482614054784e-05, | |
| "loss": 0.5978, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "learning_rate": 4.726668917274461e-05, | |
| "loss": 0.6057, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "learning_rate": 4.716855220494139e-05, | |
| "loss": 0.6048, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 2.72, | |
| "learning_rate": 4.707041523713817e-05, | |
| "loss": 0.6113, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "learning_rate": 4.6972278269334946e-05, | |
| "loss": 0.6142, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "eval_bleu": 9.547, | |
| "eval_gen_len": 17.426, | |
| "eval_loss": 0.9142343401908875, | |
| "eval_runtime": 836.5663, | |
| "eval_samples_per_second": 15.378, | |
| "eval_steps_per_second": 0.962, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "learning_rate": 4.6874141301531726e-05, | |
| "loss": 0.6165, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 2.95, | |
| "learning_rate": 4.67760043337285e-05, | |
| "loss": 0.6087, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 3.03, | |
| "learning_rate": 4.667786736592528e-05, | |
| "loss": 0.5237, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 3.11, | |
| "learning_rate": 4.657973039812205e-05, | |
| "loss": 0.3981, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 3.19, | |
| "learning_rate": 4.648159343031883e-05, | |
| "loss": 0.4018, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 3.26, | |
| "learning_rate": 4.6383456462515606e-05, | |
| "loss": 0.4119, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 3.26, | |
| "eval_bleu": 9.5055, | |
| "eval_gen_len": 17.3378, | |
| "eval_loss": 0.9879267811775208, | |
| "eval_runtime": 820.5652, | |
| "eval_samples_per_second": 15.678, | |
| "eval_steps_per_second": 0.981, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 6.69, | |
| "learning_rate": 3.947146767740577e-05, | |
| "loss": 0.3417, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 6.84, | |
| "learning_rate": 3.920825436934091e-05, | |
| "loss": 0.3358, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "learning_rate": 3.8945041061276064e-05, | |
| "loss": 0.3331, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 7.15, | |
| "learning_rate": 3.86818277532112e-05, | |
| "loss": 0.2743, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 7.31, | |
| "learning_rate": 3.841861444514635e-05, | |
| "loss": 0.2776, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 7.46, | |
| "learning_rate": 3.8155401137081495e-05, | |
| "loss": 0.2837, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 7.46, | |
| "eval_bleu": 11.0549, | |
| "eval_gen_len": 17.2982, | |
| "eval_loss": 1.0062528848648071, | |
| "eval_runtime": 964.1851, | |
| "eval_samples_per_second": 13.343, | |
| "eval_steps_per_second": 1.669, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 7.62, | |
| "learning_rate": 3.789218782901663e-05, | |
| "loss": 0.2896, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 7.77, | |
| "learning_rate": 3.762897452095178e-05, | |
| "loss": 0.2927, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 7.93, | |
| "learning_rate": 3.7365761212886926e-05, | |
| "loss": 0.298, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 8.08, | |
| "learning_rate": 3.710254790482207e-05, | |
| "loss": 0.2277, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 8.24, | |
| "learning_rate": 3.683933459675721e-05, | |
| "loss": 0.1723, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 8.4, | |
| "learning_rate": 3.657612128869236e-05, | |
| "loss": 0.1792, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 8.4, | |
| "eval_bleu": 8.9031, | |
| "eval_gen_len": 17.2801, | |
| "eval_loss": 1.085568904876709, | |
| "eval_runtime": 954.4683, | |
| "eval_samples_per_second": 13.479, | |
| "eval_steps_per_second": 1.686, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 8.55, | |
| "learning_rate": 3.63129079806275e-05, | |
| "loss": 0.1867, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 8.71, | |
| "learning_rate": 3.6049694672562644e-05, | |
| "loss": 0.1921, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 8.86, | |
| "learning_rate": 3.5786481364497795e-05, | |
| "loss": 0.1974, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 9.02, | |
| "learning_rate": 3.552326805643293e-05, | |
| "loss": 0.1903, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 9.17, | |
| "learning_rate": 3.5260054748368075e-05, | |
| "loss": 0.1132, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 9.33, | |
| "learning_rate": 3.4996841440303226e-05, | |
| "loss": 0.1204, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 9.33, | |
| "eval_bleu": 11.3498, | |
| "eval_gen_len": 17.2986, | |
| "eval_loss": 1.1643482446670532, | |
| "eval_runtime": 964.1632, | |
| "eval_samples_per_second": 13.343, | |
| "eval_steps_per_second": 1.669, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 9.48, | |
| "learning_rate": 3.473362813223837e-05, | |
| "loss": 0.125, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 9.64, | |
| "learning_rate": 3.447041482417351e-05, | |
| "loss": 0.1299, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 9.79, | |
| "learning_rate": 3.4207201516108657e-05, | |
| "loss": 0.1327, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 9.95, | |
| "learning_rate": 3.39439882080438e-05, | |
| "loss": 0.1355, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 10.11, | |
| "learning_rate": 3.3680774899978944e-05, | |
| "loss": 0.0975, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 10.26, | |
| "learning_rate": 3.341756159191409e-05, | |
| "loss": 0.0826, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 10.26, | |
| "eval_bleu": 10.796, | |
| "eval_gen_len": 17.3627, | |
| "eval_loss": 1.231860876083374, | |
| "eval_runtime": 963.3695, | |
| "eval_samples_per_second": 13.354, | |
| "eval_steps_per_second": 1.67, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 10.42, | |
| "learning_rate": 3.315434828384924e-05, | |
| "loss": 0.0874, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 10.57, | |
| "learning_rate": 3.2891134975784375e-05, | |
| "loss": 0.0894, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 10.73, | |
| "learning_rate": 3.262792166771952e-05, | |
| "loss": 0.0952, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 10.88, | |
| "learning_rate": 3.236470835965467e-05, | |
| "loss": 0.097, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 11.04, | |
| "learning_rate": 3.2101495051589806e-05, | |
| "loss": 0.0888, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 11.19, | |
| "learning_rate": 3.1838281743524956e-05, | |
| "loss": 0.0617, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 11.19, | |
| "eval_bleu": 10.6211, | |
| "eval_gen_len": 17.3748, | |
| "eval_loss": 1.2784796953201294, | |
| "eval_runtime": 963.6124, | |
| "eval_samples_per_second": 13.351, | |
| "eval_steps_per_second": 1.67, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 11.35, | |
| "learning_rate": 3.15750684354601e-05, | |
| "loss": 0.0635, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 11.5, | |
| "learning_rate": 3.131185512739524e-05, | |
| "loss": 0.0682, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 11.66, | |
| "learning_rate": 3.104864181933039e-05, | |
| "loss": 0.0703, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 11.82, | |
| "learning_rate": 3.078542851126553e-05, | |
| "loss": 0.0717, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 11.97, | |
| "learning_rate": 3.0522215203200675e-05, | |
| "loss": 0.0744, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 12.13, | |
| "learning_rate": 3.025900189513582e-05, | |
| "loss": 0.0523, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 12.13, | |
| "eval_bleu": 9.8848, | |
| "eval_gen_len": 17.3358, | |
| "eval_loss": 1.3216667175292969, | |
| "eval_runtime": 962.5757, | |
| "eval_samples_per_second": 13.365, | |
| "eval_steps_per_second": 1.672, | |
| "step": 39000 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 96480, | |
| "num_train_epochs": 30, | |
| "save_steps": 3000, | |
| "total_flos": 1.976422348357632e+18, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |