| { | |
| "best_metric": 7.343978404998779, | |
| "best_model_checkpoint": "/data1/attanasiog/babylm/roberta-tiny-4l-10M/checkpoint-200", | |
| "epoch": 29.16441875401413, | |
| "global_step": 350, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 0.00014000000000000001, | |
| "loss": 10.2422, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "learning_rate": 0.00028000000000000003, | |
| "loss": 8.631, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.49, | |
| "learning_rate": 0.00041999999999999996, | |
| "loss": 7.5098, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 3.33, | |
| "learning_rate": 0.0005600000000000001, | |
| "loss": 9.1052, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 4.16, | |
| "learning_rate": 0.0007, | |
| "loss": 7.4785, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 4.16, | |
| "eval_accuracy": 0.051358309226408695, | |
| "eval_loss": 7.383351802825928, | |
| "eval_runtime": 145.6157, | |
| "eval_samples_per_second": 165.195, | |
| "eval_steps_per_second": 5.164, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 4.99, | |
| "learning_rate": 0.0006998694084416687, | |
| "loss": 7.3182, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 5.82, | |
| "learning_rate": 0.0006994777312189898, | |
| "loss": 7.4319, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 6.66, | |
| "learning_rate": 0.0006988252606161853, | |
| "loss": 7.4297, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 7.49, | |
| "learning_rate": 0.0006979124835312712, | |
| "loss": 7.426, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 8.33, | |
| "learning_rate": 0.0006967400811127157, | |
| "loss": 7.425, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 8.33, | |
| "eval_accuracy": 0.05140231728427503, | |
| "eval_loss": 7.355892181396484, | |
| "eval_runtime": 145.5988, | |
| "eval_samples_per_second": 165.214, | |
| "eval_steps_per_second": 5.165, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 9.16, | |
| "learning_rate": 0.0006953089282511407, | |
| "loss": 7.427, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 9.99, | |
| "learning_rate": 0.0006936200929264454, | |
| "loss": 7.2955, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 10.82, | |
| "learning_rate": 0.0006916748354108406, | |
| "loss": 7.4214, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 11.66, | |
| "learning_rate": 0.0006894746073283849, | |
| "loss": 7.4198, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 12.49, | |
| "learning_rate": 0.0006870210505717298, | |
| "loss": 7.4187, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 12.49, | |
| "eval_accuracy": 0.051204356500776864, | |
| "eval_loss": 7.351686477661133, | |
| "eval_runtime": 309.0562, | |
| "eval_samples_per_second": 77.834, | |
| "eval_steps_per_second": 2.433, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 13.33, | |
| "learning_rate": 0.0006843159960768768, | |
| "loss": 7.4248, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 14.16, | |
| "learning_rate": 0.0006813614624568652, | |
| "loss": 7.418, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 14.99, | |
| "learning_rate": 0.0006781596544954071, | |
| "loss": 7.295, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 15.82, | |
| "learning_rate": 0.0006747129615015945, | |
| "loss": 7.413, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 16.66, | |
| "learning_rate": 0.0006710239555269085, | |
| "loss": 7.4204, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 16.66, | |
| "eval_accuracy": 0.05138556381472711, | |
| "eval_loss": 7.343978404998779, | |
| "eval_runtime": 145.4825, | |
| "eval_samples_per_second": 165.346, | |
| "eval_steps_per_second": 5.169, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 17.49, | |
| "learning_rate": 0.0006670953894458573, | |
| "loss": 7.4064, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 18.33, | |
| "learning_rate": 0.0006629301949016779, | |
| "loss": 7.4197, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 19.16, | |
| "learning_rate": 0.000658531480118635, | |
| "loss": 7.4168, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 19.99, | |
| "learning_rate": 0.0006539025275825468, | |
| "loss": 7.2923, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 20.82, | |
| "learning_rate": 0.000649046791591271, | |
| "loss": 7.4099, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 20.82, | |
| "eval_accuracy": 0.051534592005405604, | |
| "eval_loss": 7.345396041870117, | |
| "eval_runtime": 145.5459, | |
| "eval_samples_per_second": 165.274, | |
| "eval_steps_per_second": 5.167, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 21.66, | |
| "learning_rate": 0.0006439678956769787, | |
| "loss": 7.4172, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 22.49, | |
| "learning_rate": 0.0006386696299021389, | |
| "loss": 7.4133, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 23.33, | |
| "learning_rate": 0.0006331559480312316, | |
| "loss": 7.4126, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 24.16, | |
| "learning_rate": 0.0006274309645803004, | |
| "loss": 7.4111, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 24.99, | |
| "learning_rate": 0.000621498951746547, | |
| "loss": 7.2916, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 24.99, | |
| "eval_accuracy": 0.05147383671825258, | |
| "eval_loss": 7.34423303604126, | |
| "eval_runtime": 145.4168, | |
| "eval_samples_per_second": 165.421, | |
| "eval_steps_per_second": 5.171, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 25.82, | |
| "learning_rate": 0.0006153643362202569, | |
| "loss": 7.408, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 26.66, | |
| "learning_rate": 0.0006090316958814382, | |
| "loss": 7.4066, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 27.49, | |
| "learning_rate": 0.0006025057563836345, | |
| "loss": 7.4151, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 28.33, | |
| "learning_rate": 0.0005957913876274665, | |
| "loss": 7.4088, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 29.16, | |
| "learning_rate": 0.0005888936001265289, | |
| "loss": 7.4117, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 29.16, | |
| "eval_accuracy": 0.05125750394432031, | |
| "eval_loss": 7.344011306762695, | |
| "eval_runtime": 145.4424, | |
| "eval_samples_per_second": 165.392, | |
| "eval_steps_per_second": 5.17, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 29.16, | |
| "step": 350, | |
| "total_flos": 9.618048555527578e+16, | |
| "train_loss": 7.568343920026507, | |
| "train_runtime": 10701.7058, | |
| "train_samples_per_second": 232.767, | |
| "train_steps_per_second": 0.112 | |
| } | |
| ], | |
| "max_steps": 1200, | |
| "num_train_epochs": 100, | |
| "total_flos": 9.618048555527578e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |