| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 500, | |
| "global_step": 175, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.005714285714285714, | |
| "grad_norm": 1251.908438964567, | |
| "learning_rate": 2.7777777777777774e-08, | |
| "logits/chosen": -4.099947929382324, | |
| "logits/rejected": -4.528928756713867, | |
| "logps/chosen": -297.4884033203125, | |
| "logps/rejected": -227.07449340820312, | |
| "loss": 0.6931, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.05714285714285714, | |
| "grad_norm": 1007.5595895273253, | |
| "learning_rate": 2.7777777777777776e-07, | |
| "logits/chosen": -4.256350040435791, | |
| "logits/rejected": -4.503963947296143, | |
| "logps/chosen": -316.07769775390625, | |
| "logps/rejected": -254.57467651367188, | |
| "loss": 0.6613, | |
| "rewards/accuracies": 0.5034722089767456, | |
| "rewards/chosen": 0.2096220850944519, | |
| "rewards/margins": 0.15642070770263672, | |
| "rewards/rejected": 0.0532013401389122, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.11428571428571428, | |
| "grad_norm": 675.3841086149566, | |
| "learning_rate": 4.997998237821233e-07, | |
| "logits/chosen": -4.360010623931885, | |
| "logits/rejected": -4.628513813018799, | |
| "logps/chosen": -298.9122009277344, | |
| "logps/rejected": -249.00918579101562, | |
| "loss": 0.4212, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": 3.4100475311279297, | |
| "rewards/margins": 2.592763900756836, | |
| "rewards/rejected": 0.8172838091850281, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.17142857142857143, | |
| "grad_norm": 675.1555265980924, | |
| "learning_rate": 4.928272579403969e-07, | |
| "logits/chosen": -4.373316287994385, | |
| "logits/rejected": -4.6160383224487305, | |
| "logps/chosen": -303.8053894042969, | |
| "logps/rejected": -266.44818115234375, | |
| "loss": 0.4762, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": 4.110724449157715, | |
| "rewards/margins": 6.506677150726318, | |
| "rewards/rejected": -2.3959527015686035, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.22857142857142856, | |
| "grad_norm": 600.7834437052495, | |
| "learning_rate": 4.7616414547743854e-07, | |
| "logits/chosen": -4.35813045501709, | |
| "logits/rejected": -4.55276346206665, | |
| "logps/chosen": -289.32257080078125, | |
| "logps/rejected": -250.25341796875, | |
| "loss": 0.5303, | |
| "rewards/accuracies": 0.8343750238418579, | |
| "rewards/chosen": 3.9142494201660156, | |
| "rewards/margins": 7.811418056488037, | |
| "rewards/rejected": -3.897169589996338, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.2857142857142857, | |
| "grad_norm": 788.2365830395779, | |
| "learning_rate": 4.5047546391491e-07, | |
| "logits/chosen": -4.276906967163086, | |
| "logits/rejected": -4.5039567947387695, | |
| "logps/chosen": -297.9548034667969, | |
| "logps/rejected": -260.8029479980469, | |
| "loss": 0.4673, | |
| "rewards/accuracies": 0.8531249761581421, | |
| "rewards/chosen": 4.0600905418396, | |
| "rewards/margins": 7.4909186363220215, | |
| "rewards/rejected": -3.430828094482422, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.34285714285714286, | |
| "grad_norm": 715.3302112367288, | |
| "learning_rate": 4.167863756189767e-07, | |
| "logits/chosen": -4.322784900665283, | |
| "logits/rejected": -4.564073085784912, | |
| "logps/chosen": -293.1005554199219, | |
| "logps/rejected": -254.21835327148438, | |
| "loss": 0.4621, | |
| "rewards/accuracies": 0.893750011920929, | |
| "rewards/chosen": 5.134177207946777, | |
| "rewards/margins": 7.875572204589844, | |
| "rewards/rejected": -2.741394519805908, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 633.4373267135044, | |
| "learning_rate": 3.764413164801049e-07, | |
| "logits/chosen": -4.282719612121582, | |
| "logits/rejected": -4.559357643127441, | |
| "logps/chosen": -287.670166015625, | |
| "logps/rejected": -240.59359741210938, | |
| "loss": 0.3978, | |
| "rewards/accuracies": 0.8812500238418579, | |
| "rewards/chosen": 4.107884407043457, | |
| "rewards/margins": 7.409787654876709, | |
| "rewards/rejected": -3.301903247833252, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.45714285714285713, | |
| "grad_norm": 585.2050073161457, | |
| "learning_rate": 3.3105034329273217e-07, | |
| "logits/chosen": -4.1827239990234375, | |
| "logits/rejected": -4.454409599304199, | |
| "logps/chosen": -294.9931640625, | |
| "logps/rejected": -253.04092407226562, | |
| "loss": 0.4435, | |
| "rewards/accuracies": 0.828125, | |
| "rewards/chosen": 4.6706342697143555, | |
| "rewards/margins": 7.3439764976501465, | |
| "rewards/rejected": -2.67334246635437, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.5142857142857142, | |
| "grad_norm": 744.3685065355755, | |
| "learning_rate": 2.8242488095860204e-07, | |
| "logits/chosen": -4.200292110443115, | |
| "logits/rejected": -4.432915210723877, | |
| "logps/chosen": -289.46466064453125, | |
| "logps/rejected": -249.84048461914062, | |
| "loss": 0.4059, | |
| "rewards/accuracies": 0.8656250238418579, | |
| "rewards/chosen": 5.276065349578857, | |
| "rewards/margins": 7.3948163986206055, | |
| "rewards/rejected": -2.118751287460327, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.5714285714285714, | |
| "grad_norm": 671.6213883992457, | |
| "learning_rate": 2.3250543366050071e-07, | |
| "logits/chosen": -4.27265739440918, | |
| "logits/rejected": -4.471877098083496, | |
| "logps/chosen": -299.2139892578125, | |
| "logps/rejected": -262.4172668457031, | |
| "loss": 0.3587, | |
| "rewards/accuracies": 0.8812500238418579, | |
| "rewards/chosen": 5.20701265335083, | |
| "rewards/margins": 7.200909614562988, | |
| "rewards/rejected": -1.9938958883285522, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.6285714285714286, | |
| "grad_norm": 731.7704645558294, | |
| "learning_rate": 1.8328414484826743e-07, | |
| "logits/chosen": -4.198658466339111, | |
| "logits/rejected": -4.5151848793029785, | |
| "logps/chosen": -299.356689453125, | |
| "logps/rejected": -248.8483428955078, | |
| "loss": 0.3809, | |
| "rewards/accuracies": 0.893750011920929, | |
| "rewards/chosen": 5.6889519691467285, | |
| "rewards/margins": 7.738437652587891, | |
| "rewards/rejected": -2.049485683441162, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.6857142857142857, | |
| "grad_norm": 763.5659705732334, | |
| "learning_rate": 1.3672529644823003e-07, | |
| "logits/chosen": -4.348945140838623, | |
| "logits/rejected": -4.604073524475098, | |
| "logps/chosen": -269.67547607421875, | |
| "logps/rejected": -229.8912811279297, | |
| "loss": 0.3889, | |
| "rewards/accuracies": 0.8656250238418579, | |
| "rewards/chosen": 4.305537700653076, | |
| "rewards/margins": 6.874751091003418, | |
| "rewards/rejected": -2.569213390350342, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.7428571428571429, | |
| "grad_norm": 609.1110882142142, | |
| "learning_rate": 9.468691994696146e-08, | |
| "logits/chosen": -4.341274261474609, | |
| "logits/rejected": -4.5989298820495605, | |
| "logps/chosen": -275.35833740234375, | |
| "logps/rejected": -245.8115692138672, | |
| "loss": 0.3699, | |
| "rewards/accuracies": 0.871874988079071, | |
| "rewards/chosen": 4.084762096405029, | |
| "rewards/margins": 6.567566871643066, | |
| "rewards/rejected": -2.4828040599823, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 833.6810153426302, | |
| "learning_rate": 5.884664762850466e-08, | |
| "logits/chosen": -4.367494106292725, | |
| "logits/rejected": -4.598031044006348, | |
| "logps/chosen": -273.83099365234375, | |
| "logps/rejected": -237.6991729736328, | |
| "loss": 0.381, | |
| "rewards/accuracies": 0.8656250238418579, | |
| "rewards/chosen": 4.479451656341553, | |
| "rewards/margins": 6.597804069519043, | |
| "rewards/rejected": -2.118351697921753, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.8571428571428571, | |
| "grad_norm": 677.0337306190108, | |
| "learning_rate": 3.063476303172388e-08, | |
| "logits/chosen": -4.267864227294922, | |
| "logits/rejected": -4.488691806793213, | |
| "logps/chosen": -288.9342041015625, | |
| "logps/rejected": -247.3463592529297, | |
| "loss": 0.392, | |
| "rewards/accuracies": 0.8843749761581421, | |
| "rewards/chosen": 5.401379585266113, | |
| "rewards/margins": 7.644896030426025, | |
| "rewards/rejected": -2.2435173988342285, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.9142857142857143, | |
| "grad_norm": 601.5321276048043, | |
| "learning_rate": 1.1177122393998372e-08, | |
| "logits/chosen": -4.331192970275879, | |
| "logits/rejected": -4.530573844909668, | |
| "logps/chosen": -277.52020263671875, | |
| "logps/rejected": -243.84323120117188, | |
| "loss": 0.363, | |
| "rewards/accuracies": 0.8687499761581421, | |
| "rewards/chosen": 4.819538593292236, | |
| "rewards/margins": 6.703360080718994, | |
| "rewards/rejected": -1.8838220834732056, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.9714285714285714, | |
| "grad_norm": 771.9761619985129, | |
| "learning_rate": 1.2502249244298879e-09, | |
| "logits/chosen": -4.293517112731934, | |
| "logits/rejected": -4.548079013824463, | |
| "logps/chosen": -295.1907653808594, | |
| "logps/rejected": -248.191650390625, | |
| "loss": 0.3693, | |
| "rewards/accuracies": 0.8843749761581421, | |
| "rewards/chosen": 4.456015586853027, | |
| "rewards/margins": 6.605706214904785, | |
| "rewards/rejected": -2.1496901512145996, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 175, | |
| "total_flos": 0.0, | |
| "train_loss": 0.4268451908656529, | |
| "train_runtime": 5571.7779, | |
| "train_samples_per_second": 8.032, | |
| "train_steps_per_second": 0.031 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 175, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |