| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9885148514851485, | |
| "eval_steps": 500, | |
| "global_step": 78, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.06336633663366337, | |
| "grad_norm": 6.15625, | |
| "learning_rate": 3.1249999999999997e-07, | |
| "log_odds_chosen": 0.9902670979499817, | |
| "log_odds_ratio": -0.4136257767677307, | |
| "logps/chosen": -0.6657758951187134, | |
| "logps/rejected": -1.2273353338241577, | |
| "loss": 2.0491, | |
| "nll_loss": 0.9891467094421387, | |
| "rewards/accuracies": 0.831250011920929, | |
| "rewards/chosen": -0.03328879922628403, | |
| "rewards/margins": 0.028077969327569008, | |
| "rewards/rejected": -0.061366766691207886, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.12673267326732673, | |
| "grad_norm": 8.5625, | |
| "learning_rate": 4.989935734988097e-07, | |
| "log_odds_chosen": 1.1518529653549194, | |
| "log_odds_ratio": -0.37196290493011475, | |
| "logps/chosen": -0.5914251208305359, | |
| "logps/rejected": -1.2302882671356201, | |
| "loss": 2.4679, | |
| "nll_loss": 0.8968847393989563, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": -0.029571253806352615, | |
| "rewards/margins": 0.03194316104054451, | |
| "rewards/rejected": -0.061514418572187424, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1900990099009901, | |
| "grad_norm": 7.15625, | |
| "learning_rate": 4.877641290737883e-07, | |
| "log_odds_chosen": 1.1437830924987793, | |
| "log_odds_ratio": -0.37077194452285767, | |
| "logps/chosen": -0.5667107701301575, | |
| "logps/rejected": -1.1911952495574951, | |
| "loss": 2.4763, | |
| "nll_loss": 0.8755158185958862, | |
| "rewards/accuracies": 0.8843749761581421, | |
| "rewards/chosen": -0.028335541486740112, | |
| "rewards/margins": 0.031224220991134644, | |
| "rewards/rejected": -0.05955975502729416, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.25346534653465347, | |
| "grad_norm": 7.53125, | |
| "learning_rate": 4.646121984004665e-07, | |
| "log_odds_chosen": 1.0641790628433228, | |
| "log_odds_ratio": -0.38870134949684143, | |
| "logps/chosen": -0.6166585683822632, | |
| "logps/rejected": -1.206767201423645, | |
| "loss": 2.3836, | |
| "nll_loss": 0.9371929168701172, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": -0.030832935124635696, | |
| "rewards/margins": 0.029505427926778793, | |
| "rewards/rejected": -0.06033835932612419, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.31683168316831684, | |
| "grad_norm": 7.875, | |
| "learning_rate": 4.3069871595684787e-07, | |
| "log_odds_chosen": 1.2286351919174194, | |
| "log_odds_ratio": -0.37730690836906433, | |
| "logps/chosen": -0.6100655198097229, | |
| "logps/rejected": -1.3068325519561768, | |
| "loss": 2.3784, | |
| "nll_loss": 0.8896926641464233, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": -0.030503276735544205, | |
| "rewards/margins": 0.034838344901800156, | |
| "rewards/rejected": -0.06534162163734436, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.3801980198019802, | |
| "grad_norm": 7.9375, | |
| "learning_rate": 3.877242453630256e-07, | |
| "log_odds_chosen": 1.3759397268295288, | |
| "log_odds_ratio": -0.3282025456428528, | |
| "logps/chosen": -0.5752947926521301, | |
| "logps/rejected": -1.338653326034546, | |
| "loss": 2.2376, | |
| "nll_loss": 0.8668821454048157, | |
| "rewards/accuracies": 0.893750011920929, | |
| "rewards/chosen": -0.028764739632606506, | |
| "rewards/margins": 0.038167934864759445, | |
| "rewards/rejected": -0.06693266332149506, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.44356435643564357, | |
| "grad_norm": 8.125, | |
| "learning_rate": 3.378437060203357e-07, | |
| "log_odds_chosen": 1.5571784973144531, | |
| "log_odds_ratio": -0.28482693433761597, | |
| "logps/chosen": -0.5232511758804321, | |
| "logps/rejected": -1.388687014579773, | |
| "loss": 2.8435, | |
| "nll_loss": 0.7976736426353455, | |
| "rewards/accuracies": 0.934374988079071, | |
| "rewards/chosen": -0.026162561029195786, | |
| "rewards/margins": 0.04327179118990898, | |
| "rewards/rejected": -0.06943435966968536, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.5069306930693069, | |
| "grad_norm": 9.1875, | |
| "learning_rate": 2.8355831645441387e-07, | |
| "log_odds_chosen": 1.7383034229278564, | |
| "log_odds_ratio": -0.24174487590789795, | |
| "logps/chosen": -0.48146629333496094, | |
| "logps/rejected": -1.4316900968551636, | |
| "loss": 2.8827, | |
| "nll_loss": 0.7479132413864136, | |
| "rewards/accuracies": 0.934374988079071, | |
| "rewards/chosen": -0.024073313921689987, | |
| "rewards/margins": 0.04751119762659073, | |
| "rewards/rejected": -0.07158450782299042, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.5702970297029702, | |
| "grad_norm": 9.0, | |
| "learning_rate": 2.2759017277414164e-07, | |
| "log_odds_chosen": 1.639399766921997, | |
| "log_odds_ratio": -0.26323410868644714, | |
| "logps/chosen": -0.5090625882148743, | |
| "logps/rejected": -1.4268980026245117, | |
| "loss": 2.6135, | |
| "nll_loss": 0.7730764746665955, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": -0.025453131645917892, | |
| "rewards/margins": 0.04589176923036575, | |
| "rewards/rejected": -0.07134490460157394, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.6336633663366337, | |
| "grad_norm": 12.5625, | |
| "learning_rate": 1.7274575140626315e-07, | |
| "log_odds_chosen": 1.7857434749603271, | |
| "log_odds_ratio": -0.23216819763183594, | |
| "logps/chosen": -0.4630921483039856, | |
| "logps/rejected": -1.3960620164871216, | |
| "loss": 3.6278, | |
| "nll_loss": 0.7098131775856018, | |
| "rewards/accuracies": 0.9593750238418579, | |
| "rewards/chosen": -0.0231546089053154, | |
| "rewards/margins": 0.04664849489927292, | |
| "rewards/rejected": -0.06980310380458832, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.697029702970297, | |
| "grad_norm": 10.125, | |
| "learning_rate": 1.2177518064852348e-07, | |
| "log_odds_chosen": 1.4938031435012817, | |
| "log_odds_ratio": -0.3022598326206207, | |
| "logps/chosen": -0.5556260347366333, | |
| "logps/rejected": -1.3820416927337646, | |
| "loss": 2.8267, | |
| "nll_loss": 0.7848575115203857, | |
| "rewards/accuracies": 0.9125000238418579, | |
| "rewards/chosen": -0.027781302109360695, | |
| "rewards/margins": 0.041320785880088806, | |
| "rewards/rejected": -0.06910209357738495, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.7603960396039604, | |
| "grad_norm": 11.4375, | |
| "learning_rate": 7.723433775328384e-08, | |
| "log_odds_chosen": 1.6789665222167969, | |
| "log_odds_ratio": -0.26475173234939575, | |
| "logps/chosen": -0.5475651621818542, | |
| "logps/rejected": -1.5100398063659668, | |
| "loss": 2.7577, | |
| "nll_loss": 0.7874841690063477, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/chosen": -0.02737826108932495, | |
| "rewards/margins": 0.048123735934495926, | |
| "rewards/rejected": -0.07550199329853058, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.8237623762376237, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 4.1356686569674335e-08, | |
| "log_odds_chosen": 1.652230978012085, | |
| "log_odds_ratio": -0.2656940817832947, | |
| "logps/chosen": -0.54820317029953, | |
| "logps/rejected": -1.4693254232406616, | |
| "loss": 2.5008, | |
| "nll_loss": 0.8066957592964172, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": -0.02741015888750553, | |
| "rewards/margins": 0.0460561141371727, | |
| "rewards/rejected": -0.07346627861261368, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.8871287128712871, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 1.5941282340065697e-08, | |
| "log_odds_chosen": 1.8457355499267578, | |
| "log_odds_ratio": -0.23030798137187958, | |
| "logps/chosen": -0.5012167692184448, | |
| "logps/rejected": -1.5387108325958252, | |
| "loss": 2.9364, | |
| "nll_loss": 0.7640018463134766, | |
| "rewards/accuracies": 0.953125, | |
| "rewards/chosen": -0.02506083808839321, | |
| "rewards/margins": 0.05187469720840454, | |
| "rewards/rejected": -0.0769355446100235, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.9504950495049505, | |
| "grad_norm": 13.5, | |
| "learning_rate": 2.2625595580163247e-09, | |
| "log_odds_chosen": 1.5747287273406982, | |
| "log_odds_ratio": -0.2784648537635803, | |
| "logps/chosen": -0.5585237741470337, | |
| "logps/rejected": -1.4540793895721436, | |
| "loss": 2.5767, | |
| "nll_loss": 0.823401153087616, | |
| "rewards/accuracies": 0.9312499761581421, | |
| "rewards/chosen": -0.027926195412874222, | |
| "rewards/margins": 0.044777773320674896, | |
| "rewards/rejected": -0.07270397245883942, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.9885148514851485, | |
| "step": 78, | |
| "total_flos": 0.0, | |
| "train_loss": 2.6355448135962853, | |
| "train_runtime": 1528.1787, | |
| "train_samples_per_second": 3.305, | |
| "train_steps_per_second": 0.051 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 78, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 300, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |