| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.961038961038961, | |
| "eval_steps": 3, | |
| "global_step": 57, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.05194805194805195, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.0, | |
| "loss": 2.1051, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.15584415584415584, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.0, | |
| "loss": 2.0376, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.15584415584415584, | |
| "eval_loss": 2.0735840797424316, | |
| "eval_runtime": 188.468, | |
| "eval_samples_per_second": 1.443, | |
| "eval_steps_per_second": 0.027, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.3116883116883117, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.0, | |
| "loss": 2.0549, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.3116883116883117, | |
| "eval_loss": 2.0735840797424316, | |
| "eval_runtime": 183.8886, | |
| "eval_samples_per_second": 1.479, | |
| "eval_steps_per_second": 0.027, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.4675324675324675, | |
| "grad_norm": 1.9069625261641296, | |
| "learning_rate": 1.8e-06, | |
| "loss": 2.1192, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.4675324675324675, | |
| "eval_loss": 2.066406726837158, | |
| "eval_runtime": 188.4211, | |
| "eval_samples_per_second": 1.444, | |
| "eval_steps_per_second": 0.027, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.6233766233766234, | |
| "grad_norm": 2.195508371268519, | |
| "learning_rate": 3e-06, | |
| "loss": 2.1346, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.6233766233766234, | |
| "eval_loss": 2.055842638015747, | |
| "eval_runtime": 194.004, | |
| "eval_samples_per_second": 1.402, | |
| "eval_steps_per_second": 0.026, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.7792207792207793, | |
| "grad_norm": 1.6392960084592598, | |
| "learning_rate": 1.5844124594183009e-06, | |
| "loss": 2.08, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.7792207792207793, | |
| "eval_loss": 2.0401382446289062, | |
| "eval_runtime": 191.3975, | |
| "eval_samples_per_second": 1.421, | |
| "eval_steps_per_second": 0.026, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.935064935064935, | |
| "grad_norm": 1.4536448011292864, | |
| "learning_rate": 8.15810619632839e-07, | |
| "loss": 2.0576, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.935064935064935, | |
| "eval_loss": 2.024768829345703, | |
| "eval_runtime": 190.2805, | |
| "eval_samples_per_second": 1.429, | |
| "eval_steps_per_second": 0.026, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 1.0909090909090908, | |
| "grad_norm": 1.5244570928071939, | |
| "learning_rate": 4.147020859997813e-07, | |
| "loss": 2.0351, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 1.0909090909090908, | |
| "eval_loss": 2.0168356895446777, | |
| "eval_runtime": 183.6958, | |
| "eval_samples_per_second": 1.481, | |
| "eval_steps_per_second": 0.027, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 1.2467532467532467, | |
| "grad_norm": 0.9156345951361289, | |
| "learning_rate": 2.1460580257299744e-07, | |
| "loss": 1.9633, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 1.2467532467532467, | |
| "eval_loss": 2.0134994983673096, | |
| "eval_runtime": 183.9488, | |
| "eval_samples_per_second": 1.479, | |
| "eval_steps_per_second": 0.027, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 1.4025974025974026, | |
| "grad_norm": 2.2398676747041684, | |
| "learning_rate": 1.198239157793215e-07, | |
| "loss": 2.0389, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 1.4025974025974026, | |
| "eval_loss": 2.0117082595825195, | |
| "eval_runtime": 179.0815, | |
| "eval_samples_per_second": 1.519, | |
| "eval_steps_per_second": 0.028, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 1.5584415584415585, | |
| "grad_norm": 0.8504588484649336, | |
| "learning_rate": 7.754566893269768e-08, | |
| "loss": 1.9637, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.5584415584415585, | |
| "eval_loss": 2.0106735229492188, | |
| "eval_runtime": 184.0813, | |
| "eval_samples_per_second": 1.478, | |
| "eval_steps_per_second": 0.027, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.7142857142857144, | |
| "grad_norm": 2.8901234292003775, | |
| "learning_rate": 5.997157748148484e-08, | |
| "loss": 1.9749, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 1.7142857142857144, | |
| "eval_loss": 2.0099711418151855, | |
| "eval_runtime": 186.4048, | |
| "eval_samples_per_second": 1.459, | |
| "eval_steps_per_second": 0.027, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 1.87012987012987, | |
| "grad_norm": 1.81025589751347, | |
| "learning_rate": 5.3254794554805664e-08, | |
| "loss": 2.0231, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.87012987012987, | |
| "eval_loss": 2.0093705654144287, | |
| "eval_runtime": 181.0103, | |
| "eval_samples_per_second": 1.503, | |
| "eval_steps_per_second": 0.028, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 2.0259740259740258, | |
| "grad_norm": 0.6589563923312365, | |
| "learning_rate": 5.093567565825763e-08, | |
| "loss": 1.9785, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 2.0259740259740258, | |
| "eval_loss": 2.0088231563568115, | |
| "eval_runtime": 180.9094, | |
| "eval_samples_per_second": 1.504, | |
| "eval_steps_per_second": 0.028, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 2.1818181818181817, | |
| "grad_norm": 0.6589712270301414, | |
| "learning_rate": 5.022931619396663e-08, | |
| "loss": 1.9619, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 2.1818181818181817, | |
| "eval_loss": 2.008312463760376, | |
| "eval_runtime": 190.7131, | |
| "eval_samples_per_second": 1.426, | |
| "eval_steps_per_second": 0.026, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 2.3376623376623376, | |
| "grad_norm": 0.9435038636441478, | |
| "learning_rate": 5.0045716697115606e-08, | |
| "loss": 1.9971, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 2.3376623376623376, | |
| "eval_loss": 2.007845640182495, | |
| "eval_runtime": 193.4537, | |
| "eval_samples_per_second": 1.406, | |
| "eval_steps_per_second": 0.026, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 2.4935064935064934, | |
| "grad_norm": 2.734201702872017, | |
| "learning_rate": 5.0013450033658545e-08, | |
| "loss": 1.9992, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 2.4935064935064934, | |
| "eval_loss": 2.0075366497039795, | |
| "eval_runtime": 191.0125, | |
| "eval_samples_per_second": 1.424, | |
| "eval_steps_per_second": 0.026, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 2.6493506493506493, | |
| "grad_norm": 0.6652092963795725, | |
| "learning_rate": 5.0001589255153705e-08, | |
| "loss": 1.9912, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 2.6493506493506493, | |
| "eval_loss": 2.007113218307495, | |
| "eval_runtime": 183.1059, | |
| "eval_samples_per_second": 1.485, | |
| "eval_steps_per_second": 0.027, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 2.8051948051948052, | |
| "grad_norm": 0.8157759570772763, | |
| "learning_rate": 5.000011197079783e-08, | |
| "loss": 1.9894, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 2.8051948051948052, | |
| "eval_loss": 2.0066730976104736, | |
| "eval_runtime": 188.7087, | |
| "eval_samples_per_second": 1.441, | |
| "eval_steps_per_second": 0.026, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 2.961038961038961, | |
| "grad_norm": 2.3142513279299544, | |
| "learning_rate": 5.000000337112684e-08, | |
| "loss": 2.0143, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 2.961038961038961, | |
| "eval_loss": 2.006239414215088, | |
| "eval_runtime": 185.8251, | |
| "eval_samples_per_second": 1.464, | |
| "eval_steps_per_second": 0.027, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 2.961038961038961, | |
| "step": 57, | |
| "total_flos": 227154377834496.0, | |
| "train_loss": 2.023003205918429, | |
| "train_runtime": 21371.877, | |
| "train_samples_per_second": 0.343, | |
| "train_steps_per_second": 0.003 | |
| } | |
| ], | |
| "logging_steps": 3, | |
| "max_steps": 57, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 12, | |
| "total_flos": 227154377834496.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |