| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9221902017291066, | |
| "eval_steps": 500, | |
| "global_step": 10, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.09221902017291066, | |
| "grad_norm": 0.3019522428512573, | |
| "learning_rate": 5e-07, | |
| "logits/chosen": 0.06190446391701698, | |
| "logits/rejected": 0.24964340031147003, | |
| "logps/chosen": -133.9302215576172, | |
| "logps/ref_chosen": -134.30517578125, | |
| "logps/ref_rejected": -173.56590270996094, | |
| "logps/rejected": -173.0438232421875, | |
| "loss": 0.6939, | |
| "rewards/accuracies": 0.4453125, | |
| "rewards/chosen": 0.0037494890857487917, | |
| "rewards/margins": -0.0014713926939293742, | |
| "rewards/rejected": 0.005220881663262844, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.1844380403458213, | |
| "grad_norm": 0.41573917865753174, | |
| "learning_rate": 4.849231551964771e-07, | |
| "logits/chosen": 0.1268736720085144, | |
| "logits/rejected": 0.2718392312526703, | |
| "logps/chosen": -138.44927978515625, | |
| "logps/ref_chosen": -138.6737518310547, | |
| "logps/ref_rejected": -170.81944274902344, | |
| "logps/rejected": -170.447265625, | |
| "loss": 0.6939, | |
| "rewards/accuracies": 0.34375, | |
| "rewards/chosen": 0.0022446608636528254, | |
| "rewards/margins": -0.0014773242874071002, | |
| "rewards/rejected": 0.0037219852674752474, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.276657060518732, | |
| "grad_norm": 0.3277066946029663, | |
| "learning_rate": 4.415111107797445e-07, | |
| "logits/chosen": 0.1311892718076706, | |
| "logits/rejected": 0.26672443747520447, | |
| "logps/chosen": -132.68569946289062, | |
| "logps/ref_chosen": -132.93014526367188, | |
| "logps/ref_rejected": -165.84507751464844, | |
| "logps/rejected": -165.56008911132812, | |
| "loss": 0.6934, | |
| "rewards/accuracies": 0.4765625, | |
| "rewards/chosen": 0.0024443636648356915, | |
| "rewards/margins": -0.00040555946179665625, | |
| "rewards/rejected": 0.0028499234467744827, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.3688760806916426, | |
| "grad_norm": 0.41292324662208557, | |
| "learning_rate": 3.75e-07, | |
| "logits/chosen": 0.08853106200695038, | |
| "logits/rejected": 0.23616701364517212, | |
| "logps/chosen": -137.2488250732422, | |
| "logps/ref_chosen": -137.4427032470703, | |
| "logps/ref_rejected": -177.97886657714844, | |
| "logps/rejected": -177.62860107421875, | |
| "loss": 0.6939, | |
| "rewards/accuracies": 0.421875, | |
| "rewards/chosen": 0.001938714412972331, | |
| "rewards/margins": -0.0015639358898624778, | |
| "rewards/rejected": 0.003502650186419487, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.4610951008645533, | |
| "grad_norm": 0.31257063150405884, | |
| "learning_rate": 2.934120444167326e-07, | |
| "logits/chosen": 0.12685821950435638, | |
| "logits/rejected": 0.23144984245300293, | |
| "logps/chosen": -131.0224609375, | |
| "logps/ref_chosen": -131.1569061279297, | |
| "logps/ref_rejected": -164.11549377441406, | |
| "logps/rejected": -163.87142944335938, | |
| "loss": 0.6937, | |
| "rewards/accuracies": 0.4453125, | |
| "rewards/chosen": 0.0013444966170936823, | |
| "rewards/margins": -0.001095889019779861, | |
| "rewards/rejected": 0.0024403855204582214, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.553314121037464, | |
| "grad_norm": 0.36982661485671997, | |
| "learning_rate": 2.065879555832674e-07, | |
| "logits/chosen": 0.16632890701293945, | |
| "logits/rejected": 0.27399736642837524, | |
| "logps/chosen": -130.6382598876953, | |
| "logps/ref_chosen": -130.83815002441406, | |
| "logps/ref_rejected": -160.47244262695312, | |
| "logps/rejected": -160.15945434570312, | |
| "loss": 0.6937, | |
| "rewards/accuracies": 0.3671875, | |
| "rewards/chosen": 0.001998710911720991, | |
| "rewards/margins": -0.0011310731060802937, | |
| "rewards/rejected": 0.003129784483462572, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.6455331412103746, | |
| "grad_norm": 0.27344921231269836, | |
| "learning_rate": 1.2500000000000005e-07, | |
| "logits/chosen": 0.11610936373472214, | |
| "logits/rejected": 0.24762782454490662, | |
| "logps/chosen": -131.18833923339844, | |
| "logps/ref_chosen": -131.2764892578125, | |
| "logps/ref_rejected": -175.32669067382812, | |
| "logps/rejected": -175.1273193359375, | |
| "loss": 0.6937, | |
| "rewards/accuracies": 0.4296875, | |
| "rewards/chosen": 0.0008815132896415889, | |
| "rewards/margins": -0.0011121523566544056, | |
| "rewards/rejected": 0.0019936657045036554, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.7377521613832853, | |
| "grad_norm": 0.3876575529575348, | |
| "learning_rate": 5.848888922025552e-08, | |
| "logits/chosen": 0.0759858638048172, | |
| "logits/rejected": 0.2628093659877777, | |
| "logps/chosen": -131.0194549560547, | |
| "logps/ref_chosen": -131.0524139404297, | |
| "logps/ref_rejected": -162.96224975585938, | |
| "logps/rejected": -162.89967346191406, | |
| "loss": 0.6933, | |
| "rewards/accuracies": 0.484375, | |
| "rewards/chosen": 0.00032957797520793974, | |
| "rewards/margins": -0.00029635371174663305, | |
| "rewards/rejected": 0.0006259315996430814, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.829971181556196, | |
| "grad_norm": 0.3749904930591583, | |
| "learning_rate": 1.507684480352292e-08, | |
| "logits/chosen": 0.12389053404331207, | |
| "logits/rejected": 0.27916383743286133, | |
| "logps/chosen": -136.4560546875, | |
| "logps/ref_chosen": -136.45892333984375, | |
| "logps/ref_rejected": -167.01512145996094, | |
| "logps/rejected": -166.9549560546875, | |
| "loss": 0.6934, | |
| "rewards/accuracies": 0.4609375, | |
| "rewards/chosen": 2.849580778274685e-05, | |
| "rewards/margins": -0.0005732894060201943, | |
| "rewards/rejected": 0.0006017851992510259, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.9221902017291066, | |
| "grad_norm": 0.32978877425193787, | |
| "learning_rate": 0.0, | |
| "logits/chosen": 0.1598304808139801, | |
| "logits/rejected": 0.3594120740890503, | |
| "logps/chosen": -140.08253479003906, | |
| "logps/ref_chosen": -140.051513671875, | |
| "logps/ref_rejected": -173.16290283203125, | |
| "logps/rejected": -173.14581298828125, | |
| "loss": 0.6934, | |
| "rewards/accuracies": 0.4375, | |
| "rewards/chosen": -0.0003102564951404929, | |
| "rewards/margins": -0.0004811614053323865, | |
| "rewards/rejected": 0.00017090495384763926, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.9221902017291066, | |
| "step": 10, | |
| "total_flos": 0.0, | |
| "train_loss": 0.6936326384544372, | |
| "train_runtime": 616.2156, | |
| "train_samples_per_second": 2.251, | |
| "train_steps_per_second": 0.016 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 10, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": false, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |