| { | |
| "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": 12.625, | |
| "learning_rate": 3.1249999999999997e-07, | |
| "log_odds_chosen": 0.9938775897026062, | |
| "log_odds_ratio": -0.41239112615585327, | |
| "logps/chosen": -0.6662770509719849, | |
| "logps/rejected": -1.2300238609313965, | |
| "loss": 9.4217, | |
| "nll_loss": 0.9876836538314819, | |
| "rewards/accuracies": 0.831250011920929, | |
| "rewards/chosen": -0.03331385552883148, | |
| "rewards/margins": 0.02818734012544155, | |
| "rewards/rejected": -0.06150120496749878, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.12673267326732673, | |
| "grad_norm": 16.875, | |
| "learning_rate": 4.989935734988097e-07, | |
| "log_odds_chosen": 1.151381492614746, | |
| "log_odds_ratio": -0.37275490164756775, | |
| "logps/chosen": -0.5901239514350891, | |
| "logps/rejected": -1.2277742624282837, | |
| "loss": 11.2705, | |
| "nll_loss": 0.896571159362793, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": -0.029506197199225426, | |
| "rewards/margins": 0.03188251703977585, | |
| "rewards/rejected": -0.061388712376356125, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1900990099009901, | |
| "grad_norm": 15.5, | |
| "learning_rate": 4.877641290737883e-07, | |
| "log_odds_chosen": 1.1365224123001099, | |
| "log_odds_ratio": -0.3716532588005066, | |
| "logps/chosen": -0.5673869252204895, | |
| "logps/rejected": -1.1881535053253174, | |
| "loss": 12.0359, | |
| "nll_loss": 0.8758338093757629, | |
| "rewards/accuracies": 0.8812500238418579, | |
| "rewards/chosen": -0.028369342908263206, | |
| "rewards/margins": 0.031038332730531693, | |
| "rewards/rejected": -0.05940768122673035, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.25346534653465347, | |
| "grad_norm": 15.0625, | |
| "learning_rate": 4.646121984004665e-07, | |
| "log_odds_chosen": 1.0660278797149658, | |
| "log_odds_ratio": -0.3872835040092468, | |
| "logps/chosen": -0.6164752840995789, | |
| "logps/rejected": -1.207406997680664, | |
| "loss": 10.6234, | |
| "nll_loss": 0.9384651184082031, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": -0.030823761597275734, | |
| "rewards/margins": 0.0295465886592865, | |
| "rewards/rejected": -0.06037035584449768, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.31683168316831684, | |
| "grad_norm": 16.875, | |
| "learning_rate": 4.3069871595684787e-07, | |
| "log_odds_chosen": 1.2324841022491455, | |
| "log_odds_ratio": -0.37691444158554077, | |
| "logps/chosen": -0.6098319888114929, | |
| "logps/rejected": -1.3082152605056763, | |
| "loss": 11.208, | |
| "nll_loss": 0.8900537490844727, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": -0.030491601675748825, | |
| "rewards/margins": 0.03491916134953499, | |
| "rewards/rejected": -0.06541076302528381, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.3801980198019802, | |
| "grad_norm": 16.375, | |
| "learning_rate": 3.877242453630256e-07, | |
| "log_odds_chosen": 1.3739392757415771, | |
| "log_odds_ratio": -0.3297683596611023, | |
| "logps/chosen": -0.5759536027908325, | |
| "logps/rejected": -1.337491750717163, | |
| "loss": 10.5079, | |
| "nll_loss": 0.8683460354804993, | |
| "rewards/accuracies": 0.8968750238418579, | |
| "rewards/chosen": -0.028797682374715805, | |
| "rewards/margins": 0.038076914846897125, | |
| "rewards/rejected": -0.06687458604574203, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.44356435643564357, | |
| "grad_norm": 18.5, | |
| "learning_rate": 3.378437060203357e-07, | |
| "log_odds_chosen": 1.5558397769927979, | |
| "log_odds_ratio": -0.28507766127586365, | |
| "logps/chosen": -0.5231963992118835, | |
| "logps/rejected": -1.3864152431488037, | |
| "loss": 14.1291, | |
| "nll_loss": 0.7970401048660278, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": -0.026159819215536118, | |
| "rewards/margins": 0.04316094145178795, | |
| "rewards/rejected": -0.06932076811790466, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.5069306930693069, | |
| "grad_norm": 21.625, | |
| "learning_rate": 2.8355831645441387e-07, | |
| "log_odds_chosen": 1.7381893396377563, | |
| "log_odds_ratio": -0.2419164478778839, | |
| "logps/chosen": -0.4817093014717102, | |
| "logps/rejected": -1.4321532249450684, | |
| "loss": 17.0352, | |
| "nll_loss": 0.7467607259750366, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": -0.02408546581864357, | |
| "rewards/margins": 0.047522205859422684, | |
| "rewards/rejected": -0.07160767167806625, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.5702970297029702, | |
| "grad_norm": 21.375, | |
| "learning_rate": 2.2759017277414164e-07, | |
| "log_odds_chosen": 1.6346168518066406, | |
| "log_odds_ratio": -0.2641943395137787, | |
| "logps/chosen": -0.5079860091209412, | |
| "logps/rejected": -1.423356056213379, | |
| "loss": 14.0588, | |
| "nll_loss": 0.7725532650947571, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/chosen": -0.025399303063750267, | |
| "rewards/margins": 0.04576849564909935, | |
| "rewards/rejected": -0.07116780430078506, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.6336633663366337, | |
| "grad_norm": 26.375, | |
| "learning_rate": 1.7274575140626315e-07, | |
| "log_odds_chosen": 1.783243179321289, | |
| "log_odds_ratio": -0.2324325144290924, | |
| "logps/chosen": -0.46209725737571716, | |
| "logps/rejected": -1.3920438289642334, | |
| "loss": 20.5353, | |
| "nll_loss": 0.7096225023269653, | |
| "rewards/accuracies": 0.953125, | |
| "rewards/chosen": -0.023104865103960037, | |
| "rewards/margins": 0.04649733379483223, | |
| "rewards/rejected": -0.06960219889879227, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.697029702970297, | |
| "grad_norm": 24.125, | |
| "learning_rate": 1.2177518064852348e-07, | |
| "log_odds_chosen": 1.4938585758209229, | |
| "log_odds_ratio": -0.30352550745010376, | |
| "logps/chosen": -0.5551624298095703, | |
| "logps/rejected": -1.3799673318862915, | |
| "loss": 15.6133, | |
| "nll_loss": 0.7854876518249512, | |
| "rewards/accuracies": 0.9156249761581421, | |
| "rewards/chosen": -0.027758121490478516, | |
| "rewards/margins": 0.041240252554416656, | |
| "rewards/rejected": -0.06899837404489517, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.7603960396039604, | |
| "grad_norm": 28.125, | |
| "learning_rate": 7.723433775328384e-08, | |
| "log_odds_chosen": 1.6778091192245483, | |
| "log_odds_ratio": -0.26565200090408325, | |
| "logps/chosen": -0.5478382706642151, | |
| "logps/rejected": -1.5084162950515747, | |
| "loss": 17.3363, | |
| "nll_loss": 0.7867327928543091, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": -0.027391914278268814, | |
| "rewards/margins": 0.04802890121936798, | |
| "rewards/rejected": -0.0754208117723465, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.8237623762376237, | |
| "grad_norm": 26.5, | |
| "learning_rate": 4.1356686569674335e-08, | |
| "log_odds_chosen": 1.653925895690918, | |
| "log_odds_ratio": -0.26407569646835327, | |
| "logps/chosen": -0.5457277297973633, | |
| "logps/rejected": -1.4650758504867554, | |
| "loss": 18.0275, | |
| "nll_loss": 0.8072822690010071, | |
| "rewards/accuracies": 0.940625011920929, | |
| "rewards/chosen": -0.027286384254693985, | |
| "rewards/margins": 0.04596741124987602, | |
| "rewards/rejected": -0.07325379550457001, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.8871287128712871, | |
| "grad_norm": 22.75, | |
| "learning_rate": 1.5941282340065697e-08, | |
| "log_odds_chosen": 1.8514635562896729, | |
| "log_odds_ratio": -0.22955870628356934, | |
| "logps/chosen": -0.4996102452278137, | |
| "logps/rejected": -1.5373380184173584, | |
| "loss": 14.8205, | |
| "nll_loss": 0.7640124559402466, | |
| "rewards/accuracies": 0.953125, | |
| "rewards/chosen": -0.024980511516332626, | |
| "rewards/margins": 0.05188639834523201, | |
| "rewards/rejected": -0.07686690986156464, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.9504950495049505, | |
| "grad_norm": 30.875, | |
| "learning_rate": 2.2625595580163247e-09, | |
| "log_odds_chosen": 1.5729515552520752, | |
| "log_odds_ratio": -0.2771000564098358, | |
| "logps/chosen": -0.5589004755020142, | |
| "logps/rejected": -1.4541728496551514, | |
| "loss": 14.6472, | |
| "nll_loss": 0.8230066299438477, | |
| "rewards/accuracies": 0.940625011920929, | |
| "rewards/chosen": -0.02794502303004265, | |
| "rewards/margins": 0.04476362094283104, | |
| "rewards/rejected": -0.07270864397287369, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.9885148514851485, | |
| "step": 78, | |
| "total_flos": 0.0, | |
| "train_loss": 14.124817579220503, | |
| "train_runtime": 4601.8147, | |
| "train_samples_per_second": 1.097, | |
| "train_steps_per_second": 0.017 | |
| } | |
| ], | |
| "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 | |
| } | |