| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 10.0, | |
| "global_step": 50000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 1.98e-05, | |
| "loss": 1.4142, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 1.9600000000000002e-05, | |
| "loss": 1.2006, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 1.94e-05, | |
| "loss": 1.1483, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 1.9200000000000003e-05, | |
| "loss": 1.1203, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 1.9e-05, | |
| "loss": 1.1042, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 1.88e-05, | |
| "loss": 1.0782, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 1.86e-05, | |
| "loss": 1.0579, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 1.8400000000000003e-05, | |
| "loss": 1.0638, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 1.8200000000000002e-05, | |
| "loss": 1.0532, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 1.8e-05, | |
| "loss": 1.0355, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_gen_len": 9.6717, | |
| "eval_loss": 0.9534130692481995, | |
| "eval_rouge1": 11.1402, | |
| "eval_rouge2": 2.3949, | |
| "eval_rougeL": 11.0659, | |
| "eval_rougeLsum": 11.1028, | |
| "eval_runtime": 604.1787, | |
| "eval_samples_per_second": 16.551, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "learning_rate": 1.7800000000000002e-05, | |
| "loss": 1.0246, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "learning_rate": 1.76e-05, | |
| "loss": 1.0355, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "learning_rate": 1.7400000000000003e-05, | |
| "loss": 1.0076, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "learning_rate": 1.72e-05, | |
| "loss": 0.9938, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "learning_rate": 1.7e-05, | |
| "loss": 0.9928, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "learning_rate": 1.6800000000000002e-05, | |
| "loss": 1.0041, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "learning_rate": 1.66e-05, | |
| "loss": 0.9988, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "learning_rate": 1.64e-05, | |
| "loss": 0.9949, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "learning_rate": 1.62e-05, | |
| "loss": 0.9879, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 1.0008, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_gen_len": 10.509, | |
| "eval_loss": 0.9204540252685547, | |
| "eval_rouge1": 11.224, | |
| "eval_rouge2": 2.672, | |
| "eval_rougeL": 11.1021, | |
| "eval_rougeLsum": 11.1328, | |
| "eval_runtime": 596.5212, | |
| "eval_samples_per_second": 16.764, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "learning_rate": 1.58e-05, | |
| "loss": 0.9723, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "learning_rate": 1.5600000000000003e-05, | |
| "loss": 0.9784, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "learning_rate": 1.54e-05, | |
| "loss": 0.9629, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "learning_rate": 1.5200000000000002e-05, | |
| "loss": 0.9719, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "learning_rate": 1.5000000000000002e-05, | |
| "loss": 0.9644, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "learning_rate": 1.48e-05, | |
| "loss": 0.9847, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "learning_rate": 1.46e-05, | |
| "loss": 0.9717, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "learning_rate": 1.4400000000000001e-05, | |
| "loss": 0.9644, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "learning_rate": 1.4200000000000001e-05, | |
| "loss": 0.9618, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 1.4e-05, | |
| "loss": 0.9602, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_gen_len": 9.474, | |
| "eval_loss": 0.9030117392539978, | |
| "eval_rouge1": 12.4503, | |
| "eval_rouge2": 2.8721, | |
| "eval_rougeL": 12.354, | |
| "eval_rougeLsum": 12.3602, | |
| "eval_runtime": 585.0965, | |
| "eval_samples_per_second": 17.091, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 3.1, | |
| "learning_rate": 1.38e-05, | |
| "loss": 0.9477, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "learning_rate": 1.3600000000000002e-05, | |
| "loss": 0.9582, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 3.3, | |
| "learning_rate": 1.3400000000000002e-05, | |
| "loss": 0.9391, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 3.4, | |
| "learning_rate": 1.3200000000000002e-05, | |
| "loss": 0.9626, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 3.5, | |
| "learning_rate": 1.3000000000000001e-05, | |
| "loss": 0.951, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "learning_rate": 1.2800000000000001e-05, | |
| "loss": 0.9472, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 3.7, | |
| "learning_rate": 1.2600000000000001e-05, | |
| "loss": 0.9519, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 3.8, | |
| "learning_rate": 1.2400000000000002e-05, | |
| "loss": 0.952, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 3.9, | |
| "learning_rate": 1.22e-05, | |
| "loss": 0.945, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "learning_rate": 1.2e-05, | |
| "loss": 0.9381, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_gen_len": 9.4554, | |
| "eval_loss": 0.8888181447982788, | |
| "eval_rouge1": 12.4117, | |
| "eval_rouge2": 2.9965, | |
| "eval_rougeL": 12.3285, | |
| "eval_rougeLsum": 12.3327, | |
| "eval_runtime": 592.8034, | |
| "eval_samples_per_second": 16.869, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 4.1, | |
| "learning_rate": 1.18e-05, | |
| "loss": 0.9535, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 4.2, | |
| "learning_rate": 1.16e-05, | |
| "loss": 0.9458, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 4.3, | |
| "learning_rate": 1.14e-05, | |
| "loss": 0.9321, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 4.4, | |
| "learning_rate": 1.1200000000000001e-05, | |
| "loss": 0.9446, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 4.5, | |
| "learning_rate": 1.1000000000000001e-05, | |
| "loss": 0.9231, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 4.6, | |
| "learning_rate": 1.0800000000000002e-05, | |
| "loss": 0.9285, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 4.7, | |
| "learning_rate": 1.0600000000000002e-05, | |
| "loss": 0.9367, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "learning_rate": 1.04e-05, | |
| "loss": 0.9361, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 4.9, | |
| "learning_rate": 1.02e-05, | |
| "loss": 0.9194, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "learning_rate": 1e-05, | |
| "loss": 0.9288, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_gen_len": 9.6762, | |
| "eval_loss": 0.8804787993431091, | |
| "eval_rouge1": 12.6469, | |
| "eval_rouge2": 2.9175, | |
| "eval_rougeL": 12.5557, | |
| "eval_rougeLsum": 12.5805, | |
| "eval_runtime": 595.0767, | |
| "eval_samples_per_second": 16.805, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 5.1, | |
| "learning_rate": 9.800000000000001e-06, | |
| "loss": 0.9299, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 5.2, | |
| "learning_rate": 9.600000000000001e-06, | |
| "loss": 0.9313, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 5.3, | |
| "learning_rate": 9.4e-06, | |
| "loss": 0.9239, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 5.4, | |
| "learning_rate": 9.200000000000002e-06, | |
| "loss": 0.9238, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 5.5, | |
| "learning_rate": 9e-06, | |
| "loss": 0.9291, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 5.6, | |
| "learning_rate": 8.8e-06, | |
| "loss": 0.9207, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 5.7, | |
| "learning_rate": 8.6e-06, | |
| "loss": 0.9256, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 5.8, | |
| "learning_rate": 8.400000000000001e-06, | |
| "loss": 0.9199, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 5.9, | |
| "learning_rate": 8.2e-06, | |
| "loss": 0.9113, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 0.9243, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_gen_len": 10.0823, | |
| "eval_loss": 0.8752232193946838, | |
| "eval_rouge1": 14.0898, | |
| "eval_rouge2": 3.4389, | |
| "eval_rougeL": 13.9627, | |
| "eval_rougeLsum": 13.9793, | |
| "eval_runtime": 599.6266, | |
| "eval_samples_per_second": 16.677, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 6.1, | |
| "learning_rate": 7.800000000000002e-06, | |
| "loss": 0.9268, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 6.2, | |
| "learning_rate": 7.600000000000001e-06, | |
| "loss": 0.9144, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 6.3, | |
| "learning_rate": 7.4e-06, | |
| "loss": 0.9152, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 6.4, | |
| "learning_rate": 7.2000000000000005e-06, | |
| "loss": 0.9016, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 6.5, | |
| "learning_rate": 7e-06, | |
| "loss": 0.9119, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 6.6, | |
| "learning_rate": 6.800000000000001e-06, | |
| "loss": 0.916, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 6.7, | |
| "learning_rate": 6.600000000000001e-06, | |
| "loss": 0.9253, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 6.8, | |
| "learning_rate": 6.4000000000000006e-06, | |
| "loss": 0.9199, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 6.9, | |
| "learning_rate": 6.200000000000001e-06, | |
| "loss": 0.9168, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "learning_rate": 6e-06, | |
| "loss": 0.9087, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_gen_len": 9.7588, | |
| "eval_loss": 0.8731149435043335, | |
| "eval_rouge1": 13.9046, | |
| "eval_rouge2": 3.2515, | |
| "eval_rougeL": 13.7955, | |
| "eval_rougeLsum": 13.8149, | |
| "eval_runtime": 600.5144, | |
| "eval_samples_per_second": 16.652, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 7.1, | |
| "learning_rate": 5.8e-06, | |
| "loss": 0.905, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 7.2, | |
| "learning_rate": 5.600000000000001e-06, | |
| "loss": 0.9005, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 7.3, | |
| "learning_rate": 5.400000000000001e-06, | |
| "loss": 0.9214, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 7.4, | |
| "learning_rate": 5.2e-06, | |
| "loss": 0.9152, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 7.5, | |
| "learning_rate": 5e-06, | |
| "loss": 0.9089, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 7.6, | |
| "learning_rate": 4.800000000000001e-06, | |
| "loss": 0.9061, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 7.7, | |
| "learning_rate": 4.600000000000001e-06, | |
| "loss": 0.918, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 7.8, | |
| "learning_rate": 4.4e-06, | |
| "loss": 0.9055, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 7.9, | |
| "learning_rate": 4.2000000000000004e-06, | |
| "loss": 0.9067, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 0.9146, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_gen_len": 9.8881, | |
| "eval_loss": 0.871009886264801, | |
| "eval_rouge1": 14.9652, | |
| "eval_rouge2": 3.5015, | |
| "eval_rougeL": 14.8443, | |
| "eval_rougeLsum": 14.8507, | |
| "eval_runtime": 606.6389, | |
| "eval_samples_per_second": 16.484, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 8.1, | |
| "learning_rate": 3.8000000000000005e-06, | |
| "loss": 0.9119, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 8.2, | |
| "learning_rate": 3.6000000000000003e-06, | |
| "loss": 0.8942, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 8.3, | |
| "learning_rate": 3.4000000000000005e-06, | |
| "loss": 0.9066, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 8.4, | |
| "learning_rate": 3.2000000000000003e-06, | |
| "loss": 0.92, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 8.5, | |
| "learning_rate": 3e-06, | |
| "loss": 0.9046, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 8.6, | |
| "learning_rate": 2.8000000000000003e-06, | |
| "loss": 0.9027, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 8.7, | |
| "learning_rate": 2.6e-06, | |
| "loss": 0.9019, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 8.8, | |
| "learning_rate": 2.4000000000000003e-06, | |
| "loss": 0.9056, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 8.9, | |
| "learning_rate": 2.2e-06, | |
| "loss": 0.899, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 0.9034, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_gen_len": 9.8936, | |
| "eval_loss": 0.8681530952453613, | |
| "eval_rouge1": 14.9995, | |
| "eval_rouge2": 3.5176, | |
| "eval_rougeL": 14.8747, | |
| "eval_rougeLsum": 14.8748, | |
| "eval_runtime": 601.5997, | |
| "eval_samples_per_second": 16.622, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 9.1, | |
| "learning_rate": 1.8000000000000001e-06, | |
| "loss": 0.9079, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 9.2, | |
| "learning_rate": 1.6000000000000001e-06, | |
| "loss": 0.9027, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 9.3, | |
| "learning_rate": 1.4000000000000001e-06, | |
| "loss": 0.8944, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 9.4, | |
| "learning_rate": 1.2000000000000002e-06, | |
| "loss": 0.8955, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 9.5, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 0.8949, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 9.6, | |
| "learning_rate": 8.000000000000001e-07, | |
| "loss": 0.9037, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 9.7, | |
| "learning_rate": 6.000000000000001e-07, | |
| "loss": 0.9074, | |
| "step": 48500 | |
| }, | |
| { | |
| "epoch": 9.8, | |
| "learning_rate": 4.0000000000000003e-07, | |
| "loss": 0.9145, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 9.9, | |
| "learning_rate": 2.0000000000000002e-07, | |
| "loss": 0.9035, | |
| "step": 49500 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "learning_rate": 0.0, | |
| "loss": 0.9147, | |
| "step": 50000 | |
| } | |
| ], | |
| "max_steps": 50000, | |
| "num_train_epochs": 10, | |
| "total_flos": 6.850590755050291e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |