| { | |
| "best_metric": 0.2546972334384918, | |
| "best_model_checkpoint": "./ryan_model314_3/checkpoint-250", | |
| "epoch": 2.4, | |
| "eval_steps": 25, | |
| "global_step": 300, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.8899219036102295, | |
| "learning_rate": 9.8e-05, | |
| "loss": 0.6027, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.9669970273971558, | |
| "learning_rate": 9.6e-05, | |
| "loss": 0.4505, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "eval_loss": 0.4262112081050873, | |
| "eval_na_accuracy": 0.9, | |
| "eval_ordinal_mae": 1.0091899644308249, | |
| "eval_runtime": 24.8283, | |
| "eval_samples_per_second": 8.055, | |
| "eval_steps_per_second": 1.007, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 1.0023255348205566, | |
| "learning_rate": 9.4e-05, | |
| "loss": 0.4166, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.7017713189125061, | |
| "learning_rate": 9.200000000000001e-05, | |
| "loss": 0.3632, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 1.028002142906189, | |
| "learning_rate": 9e-05, | |
| "loss": 0.3847, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "eval_loss": 0.367563396692276, | |
| "eval_na_accuracy": 0.935, | |
| "eval_ordinal_mae": 1.3719091470156184, | |
| "eval_runtime": 7.6715, | |
| "eval_samples_per_second": 26.071, | |
| "eval_steps_per_second": 3.259, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.9666626453399658, | |
| "learning_rate": 8.800000000000001e-05, | |
| "loss": 0.3503, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 1.6747545003890991, | |
| "learning_rate": 8.6e-05, | |
| "loss": 0.3061, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "eval_loss": 0.32615897059440613, | |
| "eval_na_accuracy": 0.945, | |
| "eval_ordinal_mae": 0.7485670199170652, | |
| "eval_runtime": 7.5813, | |
| "eval_samples_per_second": 26.381, | |
| "eval_steps_per_second": 3.298, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.9667518138885498, | |
| "learning_rate": 8.4e-05, | |
| "loss": 0.3304, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 1.2316919565200806, | |
| "learning_rate": 8.2e-05, | |
| "loss": 0.3149, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.8591766953468323, | |
| "learning_rate": 8e-05, | |
| "loss": 0.2744, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "eval_loss": 0.3524325489997864, | |
| "eval_na_accuracy": 0.905, | |
| "eval_ordinal_mae": 1.1407896330054461, | |
| "eval_runtime": 7.8965, | |
| "eval_samples_per_second": 25.328, | |
| "eval_steps_per_second": 3.166, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 0.9472360014915466, | |
| "learning_rate": 7.800000000000001e-05, | |
| "loss": 0.3389, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.5021968483924866, | |
| "learning_rate": 7.6e-05, | |
| "loss": 0.2384, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.36111611127853394, | |
| "eval_na_accuracy": 0.93, | |
| "eval_ordinal_mae": 0.6746639459123592, | |
| "eval_runtime": 7.827, | |
| "eval_samples_per_second": 25.553, | |
| "eval_steps_per_second": 3.194, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 0.4840359687805176, | |
| "learning_rate": 7.4e-05, | |
| "loss": 0.1954, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 1.4421360492706299, | |
| "learning_rate": 7.2e-05, | |
| "loss": 0.223, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 0.36317479610443115, | |
| "learning_rate": 7e-05, | |
| "loss": 0.2021, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "eval_loss": 0.31051769852638245, | |
| "eval_na_accuracy": 0.95, | |
| "eval_ordinal_mae": 1.0440793198453366, | |
| "eval_runtime": 8.093, | |
| "eval_samples_per_second": 24.713, | |
| "eval_steps_per_second": 3.089, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 0.8046426773071289, | |
| "learning_rate": 6.800000000000001e-05, | |
| "loss": 0.21, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "grad_norm": 0.8122718334197998, | |
| "learning_rate": 6.6e-05, | |
| "loss": 0.2234, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "eval_loss": 0.27382710576057434, | |
| "eval_na_accuracy": 0.955, | |
| "eval_ordinal_mae": 1.4168444083902865, | |
| "eval_runtime": 7.6337, | |
| "eval_samples_per_second": 26.2, | |
| "eval_steps_per_second": 3.275, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 1.5608426332473755, | |
| "learning_rate": 6.400000000000001e-05, | |
| "loss": 0.2306, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 0.9130323529243469, | |
| "learning_rate": 6.2e-05, | |
| "loss": 0.1506, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.6261627674102783, | |
| "learning_rate": 6e-05, | |
| "loss": 0.187, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "eval_loss": 0.26879259943962097, | |
| "eval_na_accuracy": 0.955, | |
| "eval_ordinal_mae": 1.365302862794827, | |
| "eval_runtime": 7.5957, | |
| "eval_samples_per_second": 26.331, | |
| "eval_steps_per_second": 3.291, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "grad_norm": 0.7222486138343811, | |
| "learning_rate": 5.8e-05, | |
| "loss": 0.2111, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 0.7348190546035767, | |
| "learning_rate": 5.6000000000000006e-05, | |
| "loss": 0.2008, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "eval_loss": 0.2668895423412323, | |
| "eval_na_accuracy": 0.96, | |
| "eval_ordinal_mae": 0.8935630971378721, | |
| "eval_runtime": 7.6573, | |
| "eval_samples_per_second": 26.119, | |
| "eval_steps_per_second": 3.265, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "grad_norm": 1.3947832584381104, | |
| "learning_rate": 5.4000000000000005e-05, | |
| "loss": 0.2176, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 1.1390098333358765, | |
| "learning_rate": 5.2000000000000004e-05, | |
| "loss": 0.2037, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.7233979105949402, | |
| "learning_rate": 5e-05, | |
| "loss": 0.1541, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.2546972334384918, | |
| "eval_na_accuracy": 0.95, | |
| "eval_ordinal_mae": 1.2089859222915764, | |
| "eval_runtime": 7.685, | |
| "eval_samples_per_second": 26.025, | |
| "eval_steps_per_second": 3.253, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 0.5704214572906494, | |
| "learning_rate": 4.8e-05, | |
| "loss": 0.1104, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "grad_norm": 0.648725688457489, | |
| "learning_rate": 4.600000000000001e-05, | |
| "loss": 0.1201, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "eval_loss": 0.2725123465061188, | |
| "eval_na_accuracy": 0.95, | |
| "eval_ordinal_mae": 0.7955228271403142, | |
| "eval_runtime": 7.4815, | |
| "eval_samples_per_second": 26.733, | |
| "eval_steps_per_second": 3.342, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 0.5427641272544861, | |
| "learning_rate": 4.4000000000000006e-05, | |
| "loss": 0.096, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "grad_norm": 1.687751293182373, | |
| "learning_rate": 4.2e-05, | |
| "loss": 0.155, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 0.6842709183692932, | |
| "learning_rate": 4e-05, | |
| "loss": 0.113, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "eval_loss": 0.2817830443382263, | |
| "eval_na_accuracy": 0.955, | |
| "eval_ordinal_mae": 1.2378182741668735, | |
| "eval_runtime": 8.0351, | |
| "eval_samples_per_second": 24.891, | |
| "eval_steps_per_second": 3.111, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "step": 300, | |
| "total_flos": 3.71974885244928e+17, | |
| "train_loss": 0.25249010701974234, | |
| "train_runtime": 688.8214, | |
| "train_samples_per_second": 11.614, | |
| "train_steps_per_second": 0.726 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 500, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 25, | |
| "total_flos": 3.71974885244928e+17, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |