| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9191458071632097, | |
| "global_step": 30000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 5.999999999999999e-06, | |
| "loss": 0.9345, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 1.1999999999999999e-05, | |
| "loss": 0.7514, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "eval_loss": 0.9052248597145081, | |
| "eval_runtime": 0.5073, | |
| "eval_samples_per_second": 1971.284, | |
| "eval_steps_per_second": 31.541, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 1.7999999999999997e-05, | |
| "loss": 0.7408, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 2.3999999999999997e-05, | |
| "loss": 0.74, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "eval_loss": 0.9058064818382263, | |
| "eval_runtime": 0.5123, | |
| "eval_samples_per_second": 1952.096, | |
| "eval_steps_per_second": 31.234, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 2.9999999999999997e-05, | |
| "loss": 0.7398, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 3.5999999999999994e-05, | |
| "loss": 0.7395, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "eval_loss": 0.9068936705589294, | |
| "eval_runtime": 0.499, | |
| "eval_samples_per_second": 2003.855, | |
| "eval_steps_per_second": 32.062, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 4.2e-05, | |
| "loss": 0.7394, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 4.7999999999999994e-05, | |
| "loss": 0.7392, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "eval_loss": 0.9032273888587952, | |
| "eval_runtime": 0.5086, | |
| "eval_samples_per_second": 1966.372, | |
| "eval_steps_per_second": 31.462, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 5.399999999999999e-05, | |
| "loss": 0.7389, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 5.9999999999999995e-05, | |
| "loss": 0.7386, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "eval_loss": 0.9034351706504822, | |
| "eval_runtime": 0.5101, | |
| "eval_samples_per_second": 1960.24, | |
| "eval_steps_per_second": 31.364, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 6.599999999999999e-05, | |
| "loss": 0.7382, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 7.199999999999999e-05, | |
| "loss": 0.7377, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "eval_loss": 0.8661078810691833, | |
| "eval_runtime": 0.5119, | |
| "eval_samples_per_second": 1953.649, | |
| "eval_steps_per_second": 31.258, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 7.8e-05, | |
| "loss": 0.7375, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 8.4e-05, | |
| "loss": 0.7373, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "eval_loss": 0.8658460974693298, | |
| "eval_runtime": 0.5198, | |
| "eval_samples_per_second": 1923.698, | |
| "eval_steps_per_second": 30.779, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 8.999999999999999e-05, | |
| "loss": 0.7367, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 9.599999999999999e-05, | |
| "loss": 0.7207, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "eval_loss": 0.8675529956817627, | |
| "eval_runtime": 0.5261, | |
| "eval_samples_per_second": 1900.917, | |
| "eval_steps_per_second": 30.415, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 0.000102, | |
| "loss": 0.6905, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 0.00010799999999999998, | |
| "loss": 0.6746, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "eval_loss": 0.8773286938667297, | |
| "eval_runtime": 0.5159, | |
| "eval_samples_per_second": 1938.548, | |
| "eval_steps_per_second": 31.017, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 0.00011399999999999999, | |
| "loss": 0.6617, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 0.00011999999999999999, | |
| "loss": 0.6406, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "eval_loss": 0.8803548812866211, | |
| "eval_runtime": 0.5084, | |
| "eval_samples_per_second": 1966.828, | |
| "eval_steps_per_second": 31.469, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 0.00012599999999999997, | |
| "loss": 0.6275, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 0.00013199999999999998, | |
| "loss": 0.6079, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "eval_loss": 0.8707549571990967, | |
| "eval_runtime": 0.5111, | |
| "eval_samples_per_second": 1956.647, | |
| "eval_steps_per_second": 31.306, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 0.000138, | |
| "loss": 0.5977, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 0.00014399999999999998, | |
| "loss": 0.5907, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "eval_loss": 0.8676372766494751, | |
| "eval_runtime": 0.5179, | |
| "eval_samples_per_second": 1930.992, | |
| "eval_steps_per_second": 30.896, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 0.00015, | |
| "loss": 0.5834, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 0.000156, | |
| "loss": 0.5764, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "eval_loss": 0.8676320910453796, | |
| "eval_runtime": 0.5238, | |
| "eval_samples_per_second": 1909.011, | |
| "eval_steps_per_second": 30.544, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 0.000162, | |
| "loss": 0.5696, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 0.000168, | |
| "loss": 0.5648, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "eval_loss": 0.860458254814148, | |
| "eval_runtime": 0.5396, | |
| "eval_samples_per_second": 1853.225, | |
| "eval_steps_per_second": 29.652, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 0.00017399999999999997, | |
| "loss": 0.5602, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 0.00017999999999999998, | |
| "loss": 0.556, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "eval_loss": 0.8584501147270203, | |
| "eval_runtime": 0.5323, | |
| "eval_samples_per_second": 1878.593, | |
| "eval_steps_per_second": 30.057, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 0.000186, | |
| "loss": 0.5519, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 0.00019199999999999998, | |
| "loss": 0.5459, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "eval_loss": 0.8554501533508301, | |
| "eval_runtime": 0.5107, | |
| "eval_samples_per_second": 1957.988, | |
| "eval_steps_per_second": 31.328, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 0.000198, | |
| "loss": 0.5412, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 0.000204, | |
| "loss": 0.5374, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "eval_loss": 0.8554509282112122, | |
| "eval_runtime": 0.5145, | |
| "eval_samples_per_second": 1943.819, | |
| "eval_steps_per_second": 31.101, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 0.00020999999999999998, | |
| "loss": 0.5338, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 0.00021599999999999996, | |
| "loss": 0.5305, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "eval_loss": 0.8557196259498596, | |
| "eval_runtime": 0.5163, | |
| "eval_samples_per_second": 1936.818, | |
| "eval_steps_per_second": 30.989, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 0.00022199999999999998, | |
| "loss": 0.5273, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 0.00022799999999999999, | |
| "loss": 0.5243, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "eval_loss": 0.8606428503990173, | |
| "eval_runtime": 0.5087, | |
| "eval_samples_per_second": 1965.911, | |
| "eval_steps_per_second": 31.455, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 0.000234, | |
| "loss": 0.5215, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 0.00023999999999999998, | |
| "loss": 0.5188, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "eval_loss": 0.8626545071601868, | |
| "eval_runtime": 0.4996, | |
| "eval_samples_per_second": 2001.549, | |
| "eval_steps_per_second": 32.025, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 0.00024599999999999996, | |
| "loss": 0.516, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 0.00025199999999999995, | |
| "loss": 0.5136, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "eval_loss": 0.8466567993164062, | |
| "eval_runtime": 0.521, | |
| "eval_samples_per_second": 1919.499, | |
| "eval_steps_per_second": 30.712, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 0.000258, | |
| "loss": 0.5112, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 0.00026399999999999997, | |
| "loss": 0.5089, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "eval_loss": 0.8465701937675476, | |
| "eval_runtime": 0.5015, | |
| "eval_samples_per_second": 1994.167, | |
| "eval_steps_per_second": 31.907, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 0.00027, | |
| "loss": 0.5067, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 0.000276, | |
| "loss": 0.5048, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "eval_loss": 0.8367106914520264, | |
| "eval_runtime": 0.5157, | |
| "eval_samples_per_second": 1939.06, | |
| "eval_steps_per_second": 31.025, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 0.00028199999999999997, | |
| "loss": 0.5029, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 0.00028799999999999995, | |
| "loss": 0.5009, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "eval_loss": 0.8282895088195801, | |
| "eval_runtime": 0.5191, | |
| "eval_samples_per_second": 1926.314, | |
| "eval_steps_per_second": 30.821, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 0.000294, | |
| "loss": 0.4986, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 0.0003, | |
| "loss": 0.4968, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "eval_loss": 0.8393005132675171, | |
| "eval_runtime": 0.5106, | |
| "eval_samples_per_second": 1958.371, | |
| "eval_steps_per_second": 31.334, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 0.00029999920715161553, | |
| "loss": 0.495, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 0.0002999968286151326, | |
| "loss": 0.4932, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "eval_loss": 0.8276752829551697, | |
| "eval_runtime": 0.5157, | |
| "eval_samples_per_second": 1939.179, | |
| "eval_steps_per_second": 31.027, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 0.0002999928644165624, | |
| "loss": 0.4916, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 0.0002999873145992569, | |
| "loss": 0.4899, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "eval_loss": 0.8151644468307495, | |
| "eval_runtime": 0.5093, | |
| "eval_samples_per_second": 1963.529, | |
| "eval_steps_per_second": 31.416, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 0.000299980179223908, | |
| "loss": 0.4883, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 0.0002999714583685469, | |
| "loss": 0.4868, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "eval_loss": 0.8103233575820923, | |
| "eval_runtime": 0.5126, | |
| "eval_samples_per_second": 1950.786, | |
| "eval_steps_per_second": 31.213, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 0.00029996115212854366, | |
| "loss": 0.4855, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "learning_rate": 0.00029994926061660554, | |
| "loss": 0.4841, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "eval_loss": 0.8084909319877625, | |
| "eval_runtime": 0.5298, | |
| "eval_samples_per_second": 1887.571, | |
| "eval_steps_per_second": 30.201, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 0.0002999357839627762, | |
| "loss": 0.4823, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 0.00029992072231443425, | |
| "loss": 0.4806, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "eval_loss": 0.8131279945373535, | |
| "eval_runtime": 0.5066, | |
| "eval_samples_per_second": 1974.102, | |
| "eval_steps_per_second": 31.586, | |
| "step": 30000 | |
| } | |
| ], | |
| "max_steps": 500000, | |
| "num_train_epochs": 16, | |
| "total_flos": 9.584614163235e+20, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |