{ "best_metric": 1.4811300039291382, "best_model_checkpoint": "smilemikan/nllb-finetuned-jpn-to-ain-2/checkpoint-21000", "epoch": 4.626569729015202, "eval_steps": 500, "global_step": 21000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.11, "learning_rate": 1.9833333333333335e-05, "loss": 4.6196, "step": 500 }, { "epoch": 0.11, "eval_loss": 3.7452237606048584, "eval_runtime": 43.586, "eval_samples_per_second": 208.278, "eval_steps_per_second": 13.032, "step": 500 }, { "epoch": 0.22, "learning_rate": 1.9666666666666666e-05, "loss": 3.5915, "step": 1000 }, { "epoch": 0.22, "eval_loss": 3.2064383029937744, "eval_runtime": 43.0726, "eval_samples_per_second": 210.76, "eval_steps_per_second": 13.187, "step": 1000 }, { "epoch": 0.33, "learning_rate": 1.95e-05, "loss": 3.1484, "step": 1500 }, { "epoch": 0.33, "eval_loss": 2.8919074535369873, "eval_runtime": 42.8255, "eval_samples_per_second": 211.977, "eval_steps_per_second": 13.263, "step": 1500 }, { "epoch": 0.44, "learning_rate": 1.9333333333333333e-05, "loss": 2.902, "step": 2000 }, { "epoch": 0.44, "eval_loss": 2.679006338119507, "eval_runtime": 43.6162, "eval_samples_per_second": 208.134, "eval_steps_per_second": 13.023, "step": 2000 }, { "epoch": 0.55, "learning_rate": 1.916666666666667e-05, "loss": 2.7296, "step": 2500 }, { "epoch": 0.55, "eval_loss": 2.5206823348999023, "eval_runtime": 42.8291, "eval_samples_per_second": 211.959, "eval_steps_per_second": 13.262, "step": 2500 }, { "epoch": 0.66, "learning_rate": 1.9000333333333335e-05, "loss": 2.6105, "step": 3000 }, { "epoch": 0.66, "eval_loss": 2.394014835357666, "eval_runtime": 42.8022, "eval_samples_per_second": 212.092, "eval_steps_per_second": 13.27, "step": 3000 }, { "epoch": 0.77, "learning_rate": 1.883366666666667e-05, "loss": 2.5068, "step": 3500 }, { "epoch": 0.77, "eval_loss": 2.30385422706604, "eval_runtime": 44.6981, "eval_samples_per_second": 203.096, "eval_steps_per_second": 12.707, "step": 3500 }, { "epoch": 0.88, "learning_rate": 1.8667000000000003e-05, "loss": 2.3848, "step": 4000 }, { "epoch": 0.88, "eval_loss": 2.2251899242401123, "eval_runtime": 42.8042, "eval_samples_per_second": 212.082, "eval_steps_per_second": 13.27, "step": 4000 }, { "epoch": 0.99, "learning_rate": 1.8500333333333337e-05, "loss": 2.3391, "step": 4500 }, { "epoch": 0.99, "eval_loss": 2.148716449737549, "eval_runtime": 43.0552, "eval_samples_per_second": 210.845, "eval_steps_per_second": 13.192, "step": 4500 }, { "epoch": 1.1, "learning_rate": 1.8334e-05, "loss": 2.1994, "step": 5000 }, { "epoch": 1.1, "eval_loss": 2.091869354248047, "eval_runtime": 43.5857, "eval_samples_per_second": 208.279, "eval_steps_per_second": 13.032, "step": 5000 }, { "epoch": 1.21, "learning_rate": 1.8167333333333335e-05, "loss": 2.1446, "step": 5500 }, { "epoch": 1.21, "eval_loss": 2.0449471473693848, "eval_runtime": 43.2545, "eval_samples_per_second": 209.874, "eval_steps_per_second": 13.132, "step": 5500 }, { "epoch": 1.32, "learning_rate": 1.800066666666667e-05, "loss": 2.1012, "step": 6000 }, { "epoch": 1.32, "eval_loss": 2.0015575885772705, "eval_runtime": 43.2116, "eval_samples_per_second": 210.083, "eval_steps_per_second": 13.145, "step": 6000 }, { "epoch": 1.43, "learning_rate": 1.7834000000000002e-05, "loss": 2.0542, "step": 6500 }, { "epoch": 1.43, "eval_loss": 1.951575517654419, "eval_runtime": 43.3919, "eval_samples_per_second": 209.209, "eval_steps_per_second": 13.09, "step": 6500 }, { "epoch": 1.54, "learning_rate": 1.7667333333333333e-05, "loss": 2.0216, "step": 7000 }, { "epoch": 1.54, "eval_loss": 1.9161829948425293, "eval_runtime": 43.6948, "eval_samples_per_second": 207.759, "eval_steps_per_second": 12.999, "step": 7000 }, { "epoch": 1.65, "learning_rate": 1.7500666666666667e-05, "loss": 1.9926, "step": 7500 }, { "epoch": 1.65, "eval_loss": 1.8790709972381592, "eval_runtime": 44.1328, "eval_samples_per_second": 205.698, "eval_steps_per_second": 12.87, "step": 7500 }, { "epoch": 1.76, "learning_rate": 1.7334e-05, "loss": 1.9278, "step": 8000 }, { "epoch": 1.76, "eval_loss": 1.845486044883728, "eval_runtime": 44.2649, "eval_samples_per_second": 205.084, "eval_steps_per_second": 12.832, "step": 8000 }, { "epoch": 1.87, "learning_rate": 1.7167333333333334e-05, "loss": 1.9305, "step": 8500 }, { "epoch": 1.87, "eval_loss": 1.8124595880508423, "eval_runtime": 43.561, "eval_samples_per_second": 208.398, "eval_steps_per_second": 13.039, "step": 8500 }, { "epoch": 1.98, "learning_rate": 1.7000666666666668e-05, "loss": 1.905, "step": 9000 }, { "epoch": 1.98, "eval_loss": 1.7888203859329224, "eval_runtime": 43.4924, "eval_samples_per_second": 208.726, "eval_steps_per_second": 13.06, "step": 9000 }, { "epoch": 2.09, "learning_rate": 1.6834666666666667e-05, "loss": 1.828, "step": 9500 }, { "epoch": 2.09, "eval_loss": 1.7763384580612183, "eval_runtime": 43.5326, "eval_samples_per_second": 208.534, "eval_steps_per_second": 13.048, "step": 9500 }, { "epoch": 2.2, "learning_rate": 1.6668e-05, "loss": 1.808, "step": 10000 }, { "epoch": 2.2, "eval_loss": 1.7431403398513794, "eval_runtime": 43.6327, "eval_samples_per_second": 208.055, "eval_steps_per_second": 13.018, "step": 10000 }, { "epoch": 2.31, "learning_rate": 1.6501333333333334e-05, "loss": 1.7434, "step": 10500 }, { "epoch": 2.31, "eval_loss": 1.7204526662826538, "eval_runtime": 44.0612, "eval_samples_per_second": 206.032, "eval_steps_per_second": 12.891, "step": 10500 }, { "epoch": 2.42, "learning_rate": 1.6334666666666668e-05, "loss": 1.7527, "step": 11000 }, { "epoch": 2.42, "eval_loss": 1.7068791389465332, "eval_runtime": 43.4892, "eval_samples_per_second": 208.741, "eval_steps_per_second": 13.061, "step": 11000 }, { "epoch": 2.53, "learning_rate": 1.6168333333333336e-05, "loss": 1.6977, "step": 11500 }, { "epoch": 2.53, "eval_loss": 1.6881217956542969, "eval_runtime": 43.1769, "eval_samples_per_second": 210.252, "eval_steps_per_second": 13.155, "step": 11500 }, { "epoch": 2.64, "learning_rate": 1.600166666666667e-05, "loss": 1.6854, "step": 12000 }, { "epoch": 2.64, "eval_loss": 1.6784749031066895, "eval_runtime": 42.7581, "eval_samples_per_second": 212.311, "eval_steps_per_second": 13.284, "step": 12000 }, { "epoch": 2.75, "learning_rate": 1.5835e-05, "loss": 1.6603, "step": 12500 }, { "epoch": 2.75, "eval_loss": 1.6570212841033936, "eval_runtime": 42.6669, "eval_samples_per_second": 212.765, "eval_steps_per_second": 13.312, "step": 12500 }, { "epoch": 2.86, "learning_rate": 1.5668333333333334e-05, "loss": 1.6725, "step": 13000 }, { "epoch": 2.86, "eval_loss": 1.6376469135284424, "eval_runtime": 43.1686, "eval_samples_per_second": 210.292, "eval_steps_per_second": 13.158, "step": 13000 }, { "epoch": 2.97, "learning_rate": 1.5501666666666668e-05, "loss": 1.6384, "step": 13500 }, { "epoch": 2.97, "eval_loss": 1.6240772008895874, "eval_runtime": 43.2327, "eval_samples_per_second": 209.98, "eval_steps_per_second": 13.138, "step": 13500 }, { "epoch": 3.08, "learning_rate": 1.5335e-05, "loss": 1.5917, "step": 14000 }, { "epoch": 3.08, "eval_loss": 1.6128900051116943, "eval_runtime": 42.9712, "eval_samples_per_second": 211.258, "eval_steps_per_second": 13.218, "step": 14000 }, { "epoch": 3.19, "learning_rate": 1.5168333333333334e-05, "loss": 1.591, "step": 14500 }, { "epoch": 3.19, "eval_loss": 1.6054280996322632, "eval_runtime": 43.0919, "eval_samples_per_second": 210.666, "eval_steps_per_second": 13.181, "step": 14500 }, { "epoch": 3.3, "learning_rate": 1.5001666666666667e-05, "loss": 1.5936, "step": 15000 }, { "epoch": 3.3, "eval_loss": 1.5842323303222656, "eval_runtime": 43.5874, "eval_samples_per_second": 208.271, "eval_steps_per_second": 13.031, "step": 15000 }, { "epoch": 3.41, "learning_rate": 1.4835000000000001e-05, "loss": 1.6031, "step": 15500 }, { "epoch": 3.41, "eval_loss": 1.5732085704803467, "eval_runtime": 42.6594, "eval_samples_per_second": 212.802, "eval_steps_per_second": 13.315, "step": 15500 }, { "epoch": 3.53, "learning_rate": 1.4668666666666669e-05, "loss": 1.554, "step": 16000 }, { "epoch": 3.53, "eval_loss": 1.5651994943618774, "eval_runtime": 42.8927, "eval_samples_per_second": 211.645, "eval_steps_per_second": 13.242, "step": 16000 }, { "epoch": 3.64, "learning_rate": 1.4502000000000001e-05, "loss": 1.5284, "step": 16500 }, { "epoch": 3.64, "eval_loss": 1.5577419996261597, "eval_runtime": 43.6635, "eval_samples_per_second": 207.908, "eval_steps_per_second": 13.009, "step": 16500 }, { "epoch": 3.75, "learning_rate": 1.4335666666666667e-05, "loss": 1.5219, "step": 17000 }, { "epoch": 3.75, "eval_loss": 1.544019103050232, "eval_runtime": 43.4005, "eval_samples_per_second": 209.168, "eval_steps_per_second": 13.087, "step": 17000 }, { "epoch": 3.86, "learning_rate": 1.4169000000000001e-05, "loss": 1.5173, "step": 17500 }, { "epoch": 3.86, "eval_loss": 1.5335613489151, "eval_runtime": 43.512, "eval_samples_per_second": 208.632, "eval_steps_per_second": 13.054, "step": 17500 }, { "epoch": 3.97, "learning_rate": 1.4002333333333335e-05, "loss": 1.4781, "step": 18000 }, { "epoch": 3.97, "eval_loss": 1.5262504816055298, "eval_runtime": 43.3365, "eval_samples_per_second": 209.477, "eval_steps_per_second": 13.107, "step": 18000 }, { "epoch": 4.08, "learning_rate": 1.3835666666666667e-05, "loss": 1.4771, "step": 18500 }, { "epoch": 4.08, "eval_loss": 1.519882082939148, "eval_runtime": 43.8272, "eval_samples_per_second": 207.132, "eval_steps_per_second": 12.96, "step": 18500 }, { "epoch": 4.19, "learning_rate": 1.3669e-05, "loss": 1.4613, "step": 19000 }, { "epoch": 4.19, "eval_loss": 1.5142260789871216, "eval_runtime": 43.3978, "eval_samples_per_second": 209.181, "eval_steps_per_second": 13.088, "step": 19000 }, { "epoch": 4.3, "learning_rate": 1.3502333333333335e-05, "loss": 1.452, "step": 19500 }, { "epoch": 4.3, "eval_loss": 1.503504753112793, "eval_runtime": 43.2384, "eval_samples_per_second": 209.952, "eval_steps_per_second": 13.136, "step": 19500 }, { "epoch": 4.41, "learning_rate": 1.3335666666666667e-05, "loss": 1.4563, "step": 20000 }, { "epoch": 4.41, "eval_loss": 1.495379090309143, "eval_runtime": 43.6248, "eval_samples_per_second": 208.093, "eval_steps_per_second": 13.02, "step": 20000 }, { "epoch": 4.52, "learning_rate": 1.3169e-05, "loss": 1.46, "step": 20500 }, { "epoch": 4.52, "eval_loss": 1.4834085702896118, "eval_runtime": 43.9437, "eval_samples_per_second": 206.583, "eval_steps_per_second": 12.926, "step": 20500 }, { "epoch": 4.63, "learning_rate": 1.3002333333333334e-05, "loss": 1.4284, "step": 21000 }, { "epoch": 4.63, "eval_loss": 1.4811300039291382, "eval_runtime": 43.2207, "eval_samples_per_second": 210.038, "eval_steps_per_second": 13.142, "step": 21000 } ], "logging_steps": 500, "max_steps": 60000, "num_train_epochs": 14, "save_steps": 500, "total_flos": 2.130825916632269e+16, "trial_name": null, "trial_params": null }