| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.827521206409048, | |
| "global_step": 12000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 5e-06, | |
| "loss": 0.8357, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 1e-05, | |
| "loss": 0.6021, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 1.5e-05, | |
| "loss": 0.216, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1557, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.1325, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "eval_bleu": 8.163188643432274, | |
| "eval_chrf": 9.012540691377044, | |
| "eval_loss": 0.21054820716381073, | |
| "eval_runtime": 641.8265, | |
| "eval_samples_per_second": 2.493, | |
| "eval_steps_per_second": 0.078, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 3e-05, | |
| "loss": 0.1126, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 3.5e-05, | |
| "loss": 0.1059, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 4e-05, | |
| "loss": 0.096, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 4.5e-05, | |
| "loss": 0.0972, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 5e-05, | |
| "loss": 0.0895, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "eval_bleu": 12.747283674419554, | |
| "eval_chrf": 14.975219232972814, | |
| "eval_loss": 0.2011401355266571, | |
| "eval_runtime": 899.369, | |
| "eval_samples_per_second": 1.779, | |
| "eval_steps_per_second": 0.056, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 4.906820723071189e-05, | |
| "loss": 0.0782, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 4.813641446142379e-05, | |
| "loss": 0.0775, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 4.720462169213567e-05, | |
| "loss": 0.0754, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 4.627282892284756e-05, | |
| "loss": 0.0698, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 4.5341036153559454e-05, | |
| "loss": 0.0682, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "eval_bleu": 6.876683844974673, | |
| "eval_chrf": 14.013775496967607, | |
| "eval_loss": 0.2018859088420868, | |
| "eval_runtime": 978.3085, | |
| "eval_samples_per_second": 1.635, | |
| "eval_steps_per_second": 0.051, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 4.440924338427134e-05, | |
| "loss": 0.0633, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 4.347745061498323e-05, | |
| "loss": 0.0625, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 4.254565784569512e-05, | |
| "loss": 0.0627, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 4.161386507640701e-05, | |
| "loss": 0.0595, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 4.06820723071189e-05, | |
| "loss": 0.0591, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "eval_bleu": 28.04277575880573, | |
| "eval_chrf": 24.223488772329105, | |
| "eval_loss": 0.19976669549942017, | |
| "eval_runtime": 250.605, | |
| "eval_samples_per_second": 6.385, | |
| "eval_steps_per_second": 0.2, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 3.975027953783079e-05, | |
| "loss": 0.0583, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "learning_rate": 3.8818486768542676e-05, | |
| "loss": 0.0528, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "learning_rate": 3.788669399925457e-05, | |
| "loss": 0.0511, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "learning_rate": 3.695490122996646e-05, | |
| "loss": 0.0512, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "learning_rate": 3.602310846067834e-05, | |
| "loss": 0.0482, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "eval_bleu": 28.9415072135438, | |
| "eval_chrf": 24.984505458577555, | |
| "eval_loss": 0.19697046279907227, | |
| "eval_runtime": 264.9755, | |
| "eval_samples_per_second": 6.038, | |
| "eval_steps_per_second": 0.189, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "learning_rate": 3.509131569139024e-05, | |
| "loss": 0.0477, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "learning_rate": 3.415952292210213e-05, | |
| "loss": 0.0475, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "learning_rate": 3.3227730152814016e-05, | |
| "loss": 0.0477, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "learning_rate": 3.2295937383525905e-05, | |
| "loss": 0.0471, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "learning_rate": 3.1364144614237794e-05, | |
| "loss": 0.0461, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "eval_bleu": 29.631223435519598, | |
| "eval_chrf": 25.61588898393464, | |
| "eval_loss": 0.19584685564041138, | |
| "eval_runtime": 306.0261, | |
| "eval_samples_per_second": 5.228, | |
| "eval_steps_per_second": 0.163, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "learning_rate": 3.0432351844949686e-05, | |
| "loss": 0.0465, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "learning_rate": 2.950055907566158e-05, | |
| "loss": 0.0458, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "learning_rate": 2.8568766306373464e-05, | |
| "loss": 0.0433, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "learning_rate": 2.7636973537085353e-05, | |
| "loss": 0.045, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "learning_rate": 2.6705180767797245e-05, | |
| "loss": 0.044, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "eval_bleu": 29.440869914401336, | |
| "eval_chrf": 25.503480748268714, | |
| "eval_loss": 0.201187402009964, | |
| "eval_runtime": 327.7594, | |
| "eval_samples_per_second": 4.882, | |
| "eval_steps_per_second": 0.153, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "learning_rate": 2.5773387998509134e-05, | |
| "loss": 0.0434, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "learning_rate": 2.4841595229221023e-05, | |
| "loss": 0.0431, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "learning_rate": 2.3909802459932912e-05, | |
| "loss": 0.0429, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "learning_rate": 2.2978009690644804e-05, | |
| "loss": 0.0411, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "learning_rate": 2.204621692135669e-05, | |
| "loss": 0.0397, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "eval_bleu": 29.715590270005077, | |
| "eval_chrf": 25.651817457402736, | |
| "eval_loss": 0.19735974073410034, | |
| "eval_runtime": 382.7369, | |
| "eval_samples_per_second": 4.18, | |
| "eval_steps_per_second": 0.131, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "learning_rate": 2.1114424152068582e-05, | |
| "loss": 0.0416, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "learning_rate": 2.018263138278047e-05, | |
| "loss": 0.0404, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "learning_rate": 1.925083861349236e-05, | |
| "loss": 0.0398, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 2.07, | |
| "learning_rate": 1.831904584420425e-05, | |
| "loss": 0.0371, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 2.12, | |
| "learning_rate": 1.738725307491614e-05, | |
| "loss": 0.0367, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 2.12, | |
| "eval_bleu": 29.539694856534723, | |
| "eval_chrf": 25.49563217007088, | |
| "eval_loss": 0.20109611749649048, | |
| "eval_runtime": 343.2907, | |
| "eval_samples_per_second": 4.661, | |
| "eval_steps_per_second": 0.146, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "learning_rate": 1.645546030562803e-05, | |
| "loss": 0.0357, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 2.21, | |
| "learning_rate": 1.552366753633992e-05, | |
| "loss": 0.0348, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 2.26, | |
| "learning_rate": 1.4591874767051808e-05, | |
| "loss": 0.0365, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 2.31, | |
| "learning_rate": 1.36600819977637e-05, | |
| "loss": 0.0356, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 2.36, | |
| "learning_rate": 1.2728289228475587e-05, | |
| "loss": 0.0361, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 2.36, | |
| "eval_bleu": 29.780602519540743, | |
| "eval_chrf": 25.721793630268674, | |
| "eval_loss": 0.20224666595458984, | |
| "eval_runtime": 349.4231, | |
| "eval_samples_per_second": 4.579, | |
| "eval_steps_per_second": 0.143, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "learning_rate": 1.1796496459187478e-05, | |
| "loss": 0.0354, | |
| "step": 10200 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "learning_rate": 1.0864703689899367e-05, | |
| "loss": 0.0349, | |
| "step": 10400 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "learning_rate": 9.932910920611255e-06, | |
| "loss": 0.0352, | |
| "step": 10600 | |
| }, | |
| { | |
| "epoch": 2.54, | |
| "learning_rate": 9.001118151323146e-06, | |
| "loss": 0.0353, | |
| "step": 10800 | |
| }, | |
| { | |
| "epoch": 2.59, | |
| "learning_rate": 8.069325382035035e-06, | |
| "loss": 0.0357, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 2.59, | |
| "eval_bleu": 30.453580686642933, | |
| "eval_chrf": 26.259452504202525, | |
| "eval_loss": 0.19840951263904572, | |
| "eval_runtime": 395.5462, | |
| "eval_samples_per_second": 4.045, | |
| "eval_steps_per_second": 0.126, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "learning_rate": 7.137532612746925e-06, | |
| "loss": 0.0352, | |
| "step": 11200 | |
| }, | |
| { | |
| "epoch": 2.69, | |
| "learning_rate": 6.205739843458815e-06, | |
| "loss": 0.0344, | |
| "step": 11400 | |
| }, | |
| { | |
| "epoch": 2.73, | |
| "learning_rate": 5.273947074170705e-06, | |
| "loss": 0.0335, | |
| "step": 11600 | |
| }, | |
| { | |
| "epoch": 2.78, | |
| "learning_rate": 4.342154304882595e-06, | |
| "loss": 0.0346, | |
| "step": 11800 | |
| }, | |
| { | |
| "epoch": 2.83, | |
| "learning_rate": 3.410361535594484e-06, | |
| "loss": 0.0352, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 2.83, | |
| "eval_bleu": 30.412569602104455, | |
| "eval_chrf": 26.262864851975078, | |
| "eval_loss": 0.1988079845905304, | |
| "eval_runtime": 424.6307, | |
| "eval_samples_per_second": 3.768, | |
| "eval_steps_per_second": 0.118, | |
| "step": 12000 | |
| } | |
| ], | |
| "max_steps": 12732, | |
| "num_train_epochs": 3, | |
| "total_flos": 2.926491104968704e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |