| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 10.0, | |
| "eval_steps": 500, | |
| "global_step": 100, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 204.0, | |
| "learning_rate": 2e-05, | |
| "loss": 35.5858, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 83.0, | |
| "learning_rate": 0.0001, | |
| "loss": 33.1551, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 19.0, | |
| "learning_rate": 0.0002, | |
| "loss": 21.1609, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 12.363748550415039, | |
| "eval_runtime": 0.2492, | |
| "eval_samples_per_second": 40.136, | |
| "eval_steps_per_second": 4.014, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 6.40625, | |
| "learning_rate": 0.00019848077530122083, | |
| "loss": 16.2359, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 3.234375, | |
| "learning_rate": 0.00019396926207859084, | |
| "loss": 14.2403, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 7.762616157531738, | |
| "eval_runtime": 0.234, | |
| "eval_samples_per_second": 42.735, | |
| "eval_steps_per_second": 4.274, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 4.5625, | |
| "learning_rate": 0.00018660254037844388, | |
| "loss": 13.444, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 8.3125, | |
| "learning_rate": 0.0001766044443118978, | |
| "loss": 12.4548, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 6.874077796936035, | |
| "eval_runtime": 0.2344, | |
| "eval_samples_per_second": 42.661, | |
| "eval_steps_per_second": 4.266, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 3.5, | |
| "grad_norm": 11.875, | |
| "learning_rate": 0.00016427876096865394, | |
| "loss": 10.8198, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 17.0, | |
| "learning_rate": 0.00015000000000000001, | |
| "loss": 8.6478, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 6.141203880310059, | |
| "eval_runtime": 0.2347, | |
| "eval_samples_per_second": 42.609, | |
| "eval_steps_per_second": 4.261, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 4.5, | |
| "grad_norm": 18.0, | |
| "learning_rate": 0.00013420201433256689, | |
| "loss": 5.648, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 8.5625, | |
| "learning_rate": 0.00011736481776669306, | |
| "loss": 3.1923, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 4.4401326179504395, | |
| "eval_runtime": 0.236, | |
| "eval_samples_per_second": 42.379, | |
| "eval_steps_per_second": 4.238, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 5.5, | |
| "grad_norm": 4.03125, | |
| "learning_rate": 0.0001, | |
| "loss": 2.2412, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 3.109375, | |
| "learning_rate": 8.263518223330697e-05, | |
| "loss": 1.9614, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 3.3291573524475098, | |
| "eval_runtime": 0.2367, | |
| "eval_samples_per_second": 42.255, | |
| "eval_steps_per_second": 4.225, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 6.5, | |
| "grad_norm": 1.9609375, | |
| "learning_rate": 6.579798566743314e-05, | |
| "loss": 1.7794, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 0.8046875, | |
| "learning_rate": 5.000000000000002e-05, | |
| "loss": 1.692, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_loss": 3.127150297164917, | |
| "eval_runtime": 0.2355, | |
| "eval_samples_per_second": 42.466, | |
| "eval_steps_per_second": 4.247, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 7.5, | |
| "grad_norm": 0.625, | |
| "learning_rate": 3.5721239031346066e-05, | |
| "loss": 1.6184, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.6328125, | |
| "learning_rate": 2.339555568810221e-05, | |
| "loss": 1.5661, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 3.0726068019866943, | |
| "eval_runtime": 0.2359, | |
| "eval_samples_per_second": 42.395, | |
| "eval_steps_per_second": 4.24, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 8.5, | |
| "grad_norm": 0.5625, | |
| "learning_rate": 1.339745962155613e-05, | |
| "loss": 1.5358, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 0.5546875, | |
| "learning_rate": 6.030737921409169e-06, | |
| "loss": 1.5417, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_loss": 3.0536386966705322, | |
| "eval_runtime": 0.2358, | |
| "eval_samples_per_second": 42.414, | |
| "eval_steps_per_second": 4.241, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 9.5, | |
| "grad_norm": 0.57421875, | |
| "learning_rate": 1.5192246987791981e-06, | |
| "loss": 1.5267, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.56640625, | |
| "learning_rate": 0.0, | |
| "loss": 1.5287, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 3.052680015563965, | |
| "eval_runtime": 0.234, | |
| "eval_samples_per_second": 42.735, | |
| "eval_steps_per_second": 4.274, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 100, | |
| "total_flos": 3.0491778760677786e+17, | |
| "train_loss": 7.823821220397949, | |
| "train_runtime": 245.6292, | |
| "train_samples_per_second": 25.038, | |
| "train_steps_per_second": 0.407 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 100, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 100, | |
| "total_flos": 3.0491778760677786e+17, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |