| { | |
| "best_metric": 2.0710394382476807, | |
| "best_model_checkpoint": "./checkpoints/barthez-deft-chimie/checkpoint-1062", | |
| "epoch": 20.0, | |
| "global_step": 2360, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 2.8805084745762714e-05, | |
| "loss": 3.8022, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_gen_len": 19.1538, | |
| "eval_loss": 2.5490684509277344, | |
| "eval_rouge1": 16.8208, | |
| "eval_rouge2": 7.0027, | |
| "eval_rougeL": 13.957, | |
| "eval_rougeLsum": 14.0479, | |
| "eval_runtime": 3.9642, | |
| "eval_samples_per_second": 19.676, | |
| "eval_steps_per_second": 5.045, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.69, | |
| "learning_rate": 2.7533898305084746e-05, | |
| "loss": 2.9286, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_gen_len": 19.9487, | |
| "eval_loss": 2.3073835372924805, | |
| "eval_rouge1": 17.5356, | |
| "eval_rouge2": 7.8717, | |
| "eval_rougeL": 14.4874, | |
| "eval_rougeLsum": 14.5044, | |
| "eval_runtime": 3.9919, | |
| "eval_samples_per_second": 19.54, | |
| "eval_steps_per_second": 5.01, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 2.54, | |
| "learning_rate": 2.627542372881356e-05, | |
| "loss": 2.5422, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_gen_len": 19.7051, | |
| "eval_loss": 2.232180118560791, | |
| "eval_rouge1": 19.6491, | |
| "eval_rouge2": 9.4156, | |
| "eval_rougeL": 15.9467, | |
| "eval_rougeLsum": 15.9433, | |
| "eval_runtime": 3.9423, | |
| "eval_samples_per_second": 19.785, | |
| "eval_steps_per_second": 5.073, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 3.39, | |
| "learning_rate": 2.5004237288135596e-05, | |
| "loss": 2.398, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_gen_len": 19.9231, | |
| "eval_loss": 2.150047540664673, | |
| "eval_rouge1": 18.7166, | |
| "eval_rouge2": 9.859, | |
| "eval_rougeL": 15.7535, | |
| "eval_rougeLsum": 15.8036, | |
| "eval_runtime": 3.904, | |
| "eval_samples_per_second": 19.98, | |
| "eval_steps_per_second": 5.123, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 4.24, | |
| "learning_rate": 2.373305084745763e-05, | |
| "loss": 2.2044, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_gen_len": 19.6154, | |
| "eval_loss": 2.1371994018554688, | |
| "eval_rouge1": 19.978, | |
| "eval_rouge2": 10.6235, | |
| "eval_rougeL": 16.1348, | |
| "eval_rougeLsum": 16.1274, | |
| "eval_runtime": 3.9032, | |
| "eval_samples_per_second": 19.984, | |
| "eval_steps_per_second": 5.124, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 5.08, | |
| "learning_rate": 2.2461864406779662e-05, | |
| "loss": 2.0419, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 5.93, | |
| "learning_rate": 2.1203389830508474e-05, | |
| "loss": 1.9405, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_gen_len": 19.9744, | |
| "eval_loss": 2.099212884902954, | |
| "eval_rouge1": 20.226, | |
| "eval_rouge2": 10.551, | |
| "eval_rougeL": 16.6928, | |
| "eval_rougeLsum": 16.7211, | |
| "eval_runtime": 3.9012, | |
| "eval_samples_per_second": 19.994, | |
| "eval_steps_per_second": 5.127, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 6.78, | |
| "learning_rate": 1.993220338983051e-05, | |
| "loss": 1.8544, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_gen_len": 19.8846, | |
| "eval_loss": 2.0841410160064697, | |
| "eval_rouge1": 19.8869, | |
| "eval_rouge2": 10.8456, | |
| "eval_rougeL": 16.1072, | |
| "eval_rougeLsum": 16.097, | |
| "eval_runtime": 3.8835, | |
| "eval_samples_per_second": 20.085, | |
| "eval_steps_per_second": 5.15, | |
| "step": 826 | |
| }, | |
| { | |
| "epoch": 7.63, | |
| "learning_rate": 1.866101694915254e-05, | |
| "loss": 1.7536, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_gen_len": 19.859, | |
| "eval_loss": 2.0790863037109375, | |
| "eval_rouge1": 19.3017, | |
| "eval_rouge2": 9.4921, | |
| "eval_rougeL": 16.1541, | |
| "eval_rougeLsum": 16.2167, | |
| "eval_runtime": 3.9024, | |
| "eval_samples_per_second": 19.988, | |
| "eval_steps_per_second": 5.125, | |
| "step": 944 | |
| }, | |
| { | |
| "epoch": 8.47, | |
| "learning_rate": 1.7389830508474576e-05, | |
| "loss": 1.6914, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_gen_len": 19.8846, | |
| "eval_loss": 2.0710394382476807, | |
| "eval_rouge1": 21.3848, | |
| "eval_rouge2": 10.4088, | |
| "eval_rougeL": 17.1963, | |
| "eval_rougeLsum": 17.2254, | |
| "eval_runtime": 3.8762, | |
| "eval_samples_per_second": 20.123, | |
| "eval_steps_per_second": 5.16, | |
| "step": 1062 | |
| }, | |
| { | |
| "epoch": 9.32, | |
| "learning_rate": 1.611864406779661e-05, | |
| "loss": 1.654, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_gen_len": 19.9231, | |
| "eval_loss": 2.106861114501953, | |
| "eval_rouge1": 22.3811, | |
| "eval_rouge2": 10.7987, | |
| "eval_rougeL": 18.7595, | |
| "eval_rougeLsum": 18.761, | |
| "eval_runtime": 3.9276, | |
| "eval_samples_per_second": 19.859, | |
| "eval_steps_per_second": 5.092, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 10.17, | |
| "learning_rate": 1.4847457627118646e-05, | |
| "loss": 1.5899, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_gen_len": 19.8077, | |
| "eval_loss": 2.09187650680542, | |
| "eval_rouge1": 20.8546, | |
| "eval_rouge2": 10.6958, | |
| "eval_rougeL": 16.8637, | |
| "eval_rougeLsum": 16.9499, | |
| "eval_runtime": 3.8954, | |
| "eval_samples_per_second": 20.024, | |
| "eval_steps_per_second": 5.134, | |
| "step": 1298 | |
| }, | |
| { | |
| "epoch": 11.02, | |
| "learning_rate": 1.357627118644068e-05, | |
| "loss": 1.534, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 11.86, | |
| "learning_rate": 1.2305084745762711e-05, | |
| "loss": 1.4661, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_gen_len": 19.9744, | |
| "eval_loss": 2.106508255004883, | |
| "eval_rouge1": 22.3677, | |
| "eval_rouge2": 11.7472, | |
| "eval_rougeL": 18.262, | |
| "eval_rougeLsum": 18.3, | |
| "eval_runtime": 3.8544, | |
| "eval_samples_per_second": 20.237, | |
| "eval_steps_per_second": 5.189, | |
| "step": 1416 | |
| }, | |
| { | |
| "epoch": 12.71, | |
| "learning_rate": 1.1033898305084746e-05, | |
| "loss": 1.4205, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_gen_len": 19.9359, | |
| "eval_loss": 2.116386651992798, | |
| "eval_rouge1": 20.5845, | |
| "eval_rouge2": 10.7825, | |
| "eval_rougeL": 16.9972, | |
| "eval_rougeLsum": 17.0216, | |
| "eval_runtime": 3.9246, | |
| "eval_samples_per_second": 19.875, | |
| "eval_steps_per_second": 5.096, | |
| "step": 1534 | |
| }, | |
| { | |
| "epoch": 13.56, | |
| "learning_rate": 9.76271186440678e-06, | |
| "loss": 1.3797, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_gen_len": 19.9744, | |
| "eval_loss": 2.124025344848633, | |
| "eval_rouge1": 22.2561, | |
| "eval_rouge2": 11.303, | |
| "eval_rougeL": 17.5064, | |
| "eval_rougeLsum": 17.5815, | |
| "eval_runtime": 3.8828, | |
| "eval_samples_per_second": 20.089, | |
| "eval_steps_per_second": 5.151, | |
| "step": 1652 | |
| }, | |
| { | |
| "epoch": 14.41, | |
| "learning_rate": 8.491525423728813e-06, | |
| "loss": 1.3724, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_gen_len": 19.9359, | |
| "eval_loss": 2.118665933609009, | |
| "eval_rouge1": 23.2825, | |
| "eval_rouge2": 11.912, | |
| "eval_rougeL": 18.5208, | |
| "eval_rougeLsum": 18.5499, | |
| "eval_runtime": 3.9164, | |
| "eval_samples_per_second": 19.916, | |
| "eval_steps_per_second": 5.107, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 15.25, | |
| "learning_rate": 7.220338983050848e-06, | |
| "loss": 1.3404, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_gen_len": 19.9744, | |
| "eval_loss": 2.139371395111084, | |
| "eval_rouge1": 22.1305, | |
| "eval_rouge2": 10.5258, | |
| "eval_rougeL": 17.772, | |
| "eval_rougeLsum": 17.8202, | |
| "eval_runtime": 3.9181, | |
| "eval_samples_per_second": 19.908, | |
| "eval_steps_per_second": 5.105, | |
| "step": 1888 | |
| }, | |
| { | |
| "epoch": 16.1, | |
| "learning_rate": 5.949152542372881e-06, | |
| "loss": 1.2967, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 16.95, | |
| "learning_rate": 4.677966101694916e-06, | |
| "loss": 1.2846, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_gen_len": 20.0, | |
| "eval_loss": 2.1501927375793457, | |
| "eval_rouge1": 21.567, | |
| "eval_rouge2": 11.0557, | |
| "eval_rougeL": 17.2562, | |
| "eval_rougeLsum": 17.2974, | |
| "eval_runtime": 3.8698, | |
| "eval_samples_per_second": 20.156, | |
| "eval_steps_per_second": 5.168, | |
| "step": 2006 | |
| }, | |
| { | |
| "epoch": 17.8, | |
| "learning_rate": 3.406779661016949e-06, | |
| "loss": 1.2871, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_gen_len": 19.9744, | |
| "eval_loss": 2.1571571826934814, | |
| "eval_rouge1": 22.5871, | |
| "eval_rouge2": 11.702, | |
| "eval_rougeL": 18.2906, | |
| "eval_rougeLsum": 18.3826, | |
| "eval_runtime": 3.8649, | |
| "eval_samples_per_second": 20.182, | |
| "eval_steps_per_second": 5.175, | |
| "step": 2124 | |
| }, | |
| { | |
| "epoch": 18.64, | |
| "learning_rate": 2.1355932203389833e-06, | |
| "loss": 1.2422, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_gen_len": 19.9744, | |
| "eval_loss": 2.161255359649658, | |
| "eval_rouge1": 23.0935, | |
| "eval_rouge2": 11.6824, | |
| "eval_rougeL": 18.6087, | |
| "eval_rougeLsum": 18.6777, | |
| "eval_runtime": 3.8614, | |
| "eval_samples_per_second": 20.2, | |
| "eval_steps_per_second": 5.18, | |
| "step": 2242 | |
| }, | |
| { | |
| "epoch": 19.49, | |
| "learning_rate": 8.64406779661017e-07, | |
| "loss": 1.2336, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_gen_len": 19.9487, | |
| "eval_loss": 2.158088445663452, | |
| "eval_rouge1": 22.6789, | |
| "eval_rouge2": 11.4363, | |
| "eval_rougeL": 18.1661, | |
| "eval_rougeLsum": 18.2346, | |
| "eval_runtime": 3.8624, | |
| "eval_samples_per_second": 20.195, | |
| "eval_steps_per_second": 5.178, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "step": 2360, | |
| "total_flos": 1837511390490624.0, | |
| "train_loss": 1.7793180174746757, | |
| "train_runtime": 373.3963, | |
| "train_samples_per_second": 25.121, | |
| "train_steps_per_second": 6.32 | |
| } | |
| ], | |
| "max_steps": 2360, | |
| "num_train_epochs": 20, | |
| "total_flos": 1837511390490624.0, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |