| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.988190836088805, | |
| "eval_steps": 50, | |
| "global_step": 880, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "dpo_loss": 0.6931471824645996, | |
| "epoch": 0.005668398677373642, | |
| "grad_norm": 134.13611854511257, | |
| "learning_rate": 1.1363636363636363e-08, | |
| "logits": -1.3147305250167847, | |
| "logps": -88.0877456665039, | |
| "loss": 0.4113, | |
| "objective": 0.41588976979255676, | |
| "ranking_idealized": 0.5416666865348816, | |
| "ranking_idealized_expo": 0.5208333134651184, | |
| "ranking_simple": 0.5208333134651184, | |
| "regularize": 0.41588976979255676, | |
| "step": 1 | |
| }, | |
| { | |
| "dpo_loss": 0.6930932998657227, | |
| "epoch": 0.02834199338686821, | |
| "grad_norm": 133.21235814006465, | |
| "learning_rate": 5.6818181818181815e-08, | |
| "logits": -1.3678579330444336, | |
| "logps": -84.42462921142578, | |
| "loss": 0.4128, | |
| "objective": 0.3755575716495514, | |
| "ranking_idealized": 0.5677083134651184, | |
| "ranking_idealized_expo": 0.546875, | |
| "ranking_simple": 0.546875, | |
| "regularize": 0.3755575716495514, | |
| "step": 5 | |
| }, | |
| { | |
| "dpo_loss": 0.6867414116859436, | |
| "epoch": 0.05668398677373642, | |
| "grad_norm": 142.75690820723997, | |
| "learning_rate": 1.1363636363636363e-07, | |
| "logits": -1.4464198350906372, | |
| "logps": -83.337890625, | |
| "loss": 0.4131, | |
| "objective": 0.43237432837486267, | |
| "ranking_idealized": 0.5375000238418579, | |
| "ranking_idealized_expo": 0.5083333253860474, | |
| "ranking_simple": 0.512499988079071, | |
| "regularize": 0.43237432837486267, | |
| "step": 10 | |
| }, | |
| { | |
| "dpo_loss": 0.6844848990440369, | |
| "epoch": 0.08502598016060463, | |
| "grad_norm": 134.06873444315383, | |
| "learning_rate": 1.7045454545454543e-07, | |
| "logits": -1.4124974012374878, | |
| "logps": -83.50977325439453, | |
| "loss": 0.4134, | |
| "objective": 0.3983347713947296, | |
| "ranking_idealized": 0.5249999761581421, | |
| "ranking_idealized_expo": 0.5041666626930237, | |
| "ranking_simple": 0.5083333253860474, | |
| "regularize": 0.3983347713947296, | |
| "step": 15 | |
| }, | |
| { | |
| "dpo_loss": 0.6885910034179688, | |
| "epoch": 0.11336797354747284, | |
| "grad_norm": 128.26617155521478, | |
| "learning_rate": 2.2727272727272726e-07, | |
| "logits": -1.4028390645980835, | |
| "logps": -84.32319641113281, | |
| "loss": 0.396, | |
| "objective": 0.40446150302886963, | |
| "ranking_idealized": 0.550000011920929, | |
| "ranking_idealized_expo": 0.5375000238418579, | |
| "ranking_simple": 0.5375000238418579, | |
| "regularize": 0.40446150302886963, | |
| "step": 20 | |
| }, | |
| { | |
| "dpo_loss": 0.664176881313324, | |
| "epoch": 0.14170996693434104, | |
| "grad_norm": 129.31399976570776, | |
| "learning_rate": 2.840909090909091e-07, | |
| "logits": -1.4484522342681885, | |
| "logps": -84.15487670898438, | |
| "loss": 0.3876, | |
| "objective": 0.3946884274482727, | |
| "ranking_idealized": 0.5041666626930237, | |
| "ranking_idealized_expo": 0.4958333373069763, | |
| "ranking_simple": 0.5, | |
| "regularize": 0.3946884274482727, | |
| "step": 25 | |
| }, | |
| { | |
| "dpo_loss": 0.6644570827484131, | |
| "epoch": 0.17005196032120926, | |
| "grad_norm": 129.8647460663128, | |
| "learning_rate": 3.4090909090909085e-07, | |
| "logits": -1.4183248281478882, | |
| "logps": -84.92523956298828, | |
| "loss": 0.3768, | |
| "objective": 0.38247305154800415, | |
| "ranking_idealized": 0.5583333373069763, | |
| "ranking_idealized_expo": 0.5416666865348816, | |
| "ranking_simple": 0.550000011920929, | |
| "regularize": 0.38247305154800415, | |
| "step": 30 | |
| }, | |
| { | |
| "dpo_loss": 0.6450185775756836, | |
| "epoch": 0.19839395370807747, | |
| "grad_norm": 130.99553488282152, | |
| "learning_rate": 3.977272727272727e-07, | |
| "logits": -1.3883568048477173, | |
| "logps": -83.79637145996094, | |
| "loss": 0.3692, | |
| "objective": 0.3310621976852417, | |
| "ranking_idealized": 0.5249999761581421, | |
| "ranking_idealized_expo": 0.5041666626930237, | |
| "ranking_simple": 0.5041666626930237, | |
| "regularize": 0.3310621976852417, | |
| "step": 35 | |
| }, | |
| { | |
| "dpo_loss": 0.6378442049026489, | |
| "epoch": 0.22673594709494568, | |
| "grad_norm": 124.69353315131622, | |
| "learning_rate": 4.545454545454545e-07, | |
| "logits": -1.3773659467697144, | |
| "logps": -83.89529418945312, | |
| "loss": 0.3685, | |
| "objective": 0.3828122019767761, | |
| "ranking_idealized": 0.5375000238418579, | |
| "ranking_idealized_expo": 0.5208333134651184, | |
| "ranking_simple": 0.5249999761581421, | |
| "regularize": 0.3828122019767761, | |
| "step": 40 | |
| }, | |
| { | |
| "dpo_loss": 0.6174018383026123, | |
| "epoch": 0.25507794048181387, | |
| "grad_norm": 131.46158666021728, | |
| "learning_rate": 5.113636363636363e-07, | |
| "logits": -1.4973019361495972, | |
| "logps": -85.6716537475586, | |
| "loss": 0.3654, | |
| "objective": 0.4057728052139282, | |
| "ranking_idealized": 0.5625, | |
| "ranking_idealized_expo": 0.5375000238418579, | |
| "ranking_simple": 0.550000011920929, | |
| "regularize": 0.40577277541160583, | |
| "step": 45 | |
| }, | |
| { | |
| "dpo_loss": 0.6175411343574524, | |
| "epoch": 0.2834199338686821, | |
| "grad_norm": 120.92405538498733, | |
| "learning_rate": 5.681818181818182e-07, | |
| "logits": -1.3952524662017822, | |
| "logps": -83.97753143310547, | |
| "loss": 0.3563, | |
| "objective": 0.3284095227718353, | |
| "ranking_idealized": 0.4416666626930237, | |
| "ranking_idealized_expo": 0.42500001192092896, | |
| "ranking_simple": 0.4333333373069763, | |
| "regularize": 0.3284095227718353, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.2834199338686821, | |
| "eval_dpo_loss": 0.687321662902832, | |
| "eval_logits": -1.420853853225708, | |
| "eval_logps": -92.33389282226562, | |
| "eval_loss": 0.42110419273376465, | |
| "eval_objective": 0.42802950739860535, | |
| "eval_ranking_idealized": 0.5247933864593506, | |
| "eval_ranking_idealized_expo": 0.5092975497245789, | |
| "eval_ranking_simple": 0.5113636255264282, | |
| "eval_regularize": 0.42802950739860535, | |
| "eval_runtime": 269.5653, | |
| "eval_samples_per_second": 21.479, | |
| "eval_steps_per_second": 0.898, | |
| "step": 50 | |
| }, | |
| { | |
| "dpo_loss": 0.6299411058425903, | |
| "epoch": 0.3117619272555503, | |
| "grad_norm": 143.5714789398323, | |
| "learning_rate": 6.249999999999999e-07, | |
| "logits": -1.4425883293151855, | |
| "logps": -86.13849639892578, | |
| "loss": 0.3849, | |
| "objective": 0.3963293433189392, | |
| "ranking_idealized": 0.550000011920929, | |
| "ranking_idealized_expo": 0.5416666865348816, | |
| "ranking_simple": 0.5416666865348816, | |
| "regularize": 0.3963293433189392, | |
| "step": 55 | |
| }, | |
| { | |
| "dpo_loss": 0.6125720143318176, | |
| "epoch": 0.3401039206424185, | |
| "grad_norm": 136.6167401869809, | |
| "learning_rate": 6.818181818181817e-07, | |
| "logits": -1.4236023426055908, | |
| "logps": -84.95130157470703, | |
| "loss": 0.3788, | |
| "objective": 0.3742086589336395, | |
| "ranking_idealized": 0.5333333611488342, | |
| "ranking_idealized_expo": 0.512499988079071, | |
| "ranking_simple": 0.5208333134651184, | |
| "regularize": 0.3742086589336395, | |
| "step": 60 | |
| }, | |
| { | |
| "dpo_loss": 0.6023539900779724, | |
| "epoch": 0.3684459140292867, | |
| "grad_norm": 145.32268284333037, | |
| "learning_rate": 7.386363636363636e-07, | |
| "logits": -1.3705960512161255, | |
| "logps": -85.3038330078125, | |
| "loss": 0.3824, | |
| "objective": 0.3741196095943451, | |
| "ranking_idealized": 0.5083333253860474, | |
| "ranking_idealized_expo": 0.5083333253860474, | |
| "ranking_simple": 0.512499988079071, | |
| "regularize": 0.3741196095943451, | |
| "step": 65 | |
| }, | |
| { | |
| "dpo_loss": 0.6065691709518433, | |
| "epoch": 0.39678790741615494, | |
| "grad_norm": 139.19992394086944, | |
| "learning_rate": 7.954545454545454e-07, | |
| "logits": -1.448442816734314, | |
| "logps": -86.62315368652344, | |
| "loss": 0.3952, | |
| "objective": 0.4089803397655487, | |
| "ranking_idealized": 0.5208333134651184, | |
| "ranking_idealized_expo": 0.5166666507720947, | |
| "ranking_simple": 0.5208333134651184, | |
| "regularize": 0.4089803397655487, | |
| "step": 70 | |
| }, | |
| { | |
| "dpo_loss": 0.5920066833496094, | |
| "epoch": 0.42512990080302315, | |
| "grad_norm": 124.21133200378506, | |
| "learning_rate": 8.522727272727273e-07, | |
| "logits": -1.4492307901382446, | |
| "logps": -86.64204406738281, | |
| "loss": 0.4203, | |
| "objective": 0.4280329644680023, | |
| "ranking_idealized": 0.5583333373069763, | |
| "ranking_idealized_expo": 0.5458333492279053, | |
| "ranking_simple": 0.5541666746139526, | |
| "regularize": 0.4280329644680023, | |
| "step": 75 | |
| }, | |
| { | |
| "dpo_loss": 0.5960291028022766, | |
| "epoch": 0.45347189418989137, | |
| "grad_norm": 189.2274993381048, | |
| "learning_rate": 9.09090909090909e-07, | |
| "logits": -1.4888454675674438, | |
| "logps": -86.4117431640625, | |
| "loss": 0.4188, | |
| "objective": 0.41209739446640015, | |
| "ranking_idealized": 0.5291666388511658, | |
| "ranking_idealized_expo": 0.5166666507720947, | |
| "ranking_simple": 0.512499988079071, | |
| "regularize": 0.41209739446640015, | |
| "step": 80 | |
| }, | |
| { | |
| "dpo_loss": 0.5794143676757812, | |
| "epoch": 0.4818138875767596, | |
| "grad_norm": 135.1857661271161, | |
| "learning_rate": 9.65909090909091e-07, | |
| "logits": -1.449118971824646, | |
| "logps": -85.0038833618164, | |
| "loss": 0.4453, | |
| "objective": 0.4282515347003937, | |
| "ranking_idealized": 0.5416666865348816, | |
| "ranking_idealized_expo": 0.5249999761581421, | |
| "ranking_simple": 0.5333333611488342, | |
| "regularize": 0.4282515347003937, | |
| "step": 85 | |
| }, | |
| { | |
| "dpo_loss": 0.5998800992965698, | |
| "epoch": 0.5101558809636277, | |
| "grad_norm": 176.339935388579, | |
| "learning_rate": 9.999842657116664e-07, | |
| "logits": -1.262819766998291, | |
| "logps": -84.80656433105469, | |
| "loss": 0.4721, | |
| "objective": 0.4337444007396698, | |
| "ranking_idealized": 0.4958333373069763, | |
| "ranking_idealized_expo": 0.4791666567325592, | |
| "ranking_simple": 0.49166667461395264, | |
| "regularize": 0.4337444007396698, | |
| "step": 90 | |
| }, | |
| { | |
| "dpo_loss": 0.5696186423301697, | |
| "epoch": 0.538497874350496, | |
| "grad_norm": 126.57058201712371, | |
| "learning_rate": 9.998072663403656e-07, | |
| "logits": -1.356713056564331, | |
| "logps": -84.7969741821289, | |
| "loss": 0.4657, | |
| "objective": 0.49121516942977905, | |
| "ranking_idealized": 0.5625, | |
| "ranking_idealized_expo": 0.5416666865348816, | |
| "ranking_simple": 0.5625, | |
| "regularize": 0.49121516942977905, | |
| "step": 95 | |
| }, | |
| { | |
| "dpo_loss": 0.5877792835235596, | |
| "epoch": 0.5668398677373642, | |
| "grad_norm": 132.17672172143168, | |
| "learning_rate": 9.99433669591504e-07, | |
| "logits": -1.4267613887786865, | |
| "logps": -85.57508850097656, | |
| "loss": 0.4749, | |
| "objective": 0.49508702754974365, | |
| "ranking_idealized": 0.550000011920929, | |
| "ranking_idealized_expo": 0.5416666865348816, | |
| "ranking_simple": 0.5375000238418579, | |
| "regularize": 0.49508702754974365, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.5668398677373642, | |
| "eval_dpo_loss": 0.6957917213439941, | |
| "eval_logits": -1.3981915712356567, | |
| "eval_logps": -91.31111907958984, | |
| "eval_loss": 0.46728843450546265, | |
| "eval_objective": 0.4711840748786926, | |
| "eval_ranking_idealized": 0.5247933864593506, | |
| "eval_ranking_idealized_expo": 0.5092975497245789, | |
| "eval_ranking_simple": 0.5144628286361694, | |
| "eval_regularize": 0.4711840748786926, | |
| "eval_runtime": 258.6905, | |
| "eval_samples_per_second": 22.382, | |
| "eval_steps_per_second": 0.935, | |
| "step": 100 | |
| }, | |
| { | |
| "dpo_loss": 0.5984219908714294, | |
| "epoch": 0.5951818611242324, | |
| "grad_norm": 141.88949705555274, | |
| "learning_rate": 9.988636224180095e-07, | |
| "logits": -1.3065966367721558, | |
| "logps": -85.25997924804688, | |
| "loss": 0.4971, | |
| "objective": 0.5261873602867126, | |
| "ranking_idealized": 0.550000011920929, | |
| "ranking_idealized_expo": 0.5333333611488342, | |
| "ranking_simple": 0.550000011920929, | |
| "regularize": 0.5261873602867126, | |
| "step": 105 | |
| }, | |
| { | |
| "dpo_loss": 0.5468531847000122, | |
| "epoch": 0.6235238545111006, | |
| "grad_norm": 118.1204631787259, | |
| "learning_rate": 9.980973490458728e-07, | |
| "logits": -1.4602758884429932, | |
| "logps": -83.59262084960938, | |
| "loss": 0.4868, | |
| "objective": 0.4539487659931183, | |
| "ranking_idealized": 0.47083333134651184, | |
| "ranking_idealized_expo": 0.44999998807907104, | |
| "ranking_simple": 0.47083333134651184, | |
| "regularize": 0.4539487659931183, | |
| "step": 110 | |
| }, | |
| { | |
| "dpo_loss": 0.595252275466919, | |
| "epoch": 0.6518658478979689, | |
| "grad_norm": 125.27205037996261, | |
| "learning_rate": 9.971351508859486e-07, | |
| "logits": -1.4180139303207397, | |
| "logps": -83.7674560546875, | |
| "loss": 0.5199, | |
| "objective": 0.49929893016815186, | |
| "ranking_idealized": 0.4958333373069763, | |
| "ranking_idealized_expo": 0.4833333194255829, | |
| "ranking_simple": 0.48750001192092896, | |
| "regularize": 0.49929893016815186, | |
| "step": 115 | |
| }, | |
| { | |
| "dpo_loss": 0.6112563014030457, | |
| "epoch": 0.680207841284837, | |
| "grad_norm": 115.02330529438134, | |
| "learning_rate": 9.959774064153975e-07, | |
| "logits": -1.3522943258285522, | |
| "logps": -84.06304168701172, | |
| "loss": 0.5115, | |
| "objective": 0.5348313450813293, | |
| "ranking_idealized": 0.5375000238418579, | |
| "ranking_idealized_expo": 0.5333333611488342, | |
| "ranking_simple": 0.5375000238418579, | |
| "regularize": 0.5348313450813293, | |
| "step": 120 | |
| }, | |
| { | |
| "dpo_loss": 0.5981053709983826, | |
| "epoch": 0.7085498346717053, | |
| "grad_norm": 125.9955613878692, | |
| "learning_rate": 9.94624571028813e-07, | |
| "logits": -1.3196688890457153, | |
| "logps": -82.79936981201172, | |
| "loss": 0.5262, | |
| "objective": 0.5262423157691956, | |
| "ranking_idealized": 0.574999988079071, | |
| "ranking_idealized_expo": 0.550000011920929, | |
| "ranking_simple": 0.5666666626930237, | |
| "regularize": 0.5262423157691956, | |
| "step": 125 | |
| }, | |
| { | |
| "dpo_loss": 0.5788747668266296, | |
| "epoch": 0.7368918280585735, | |
| "grad_norm": 127.79997373259253, | |
| "learning_rate": 9.930771768590933e-07, | |
| "logits": -1.4446656703948975, | |
| "logps": -82.71467590332031, | |
| "loss": 0.5057, | |
| "objective": 0.5254269242286682, | |
| "ranking_idealized": 0.4833333194255829, | |
| "ranking_idealized_expo": 0.44999998807907104, | |
| "ranking_simple": 0.47083333134651184, | |
| "regularize": 0.5254269242286682, | |
| "step": 130 | |
| }, | |
| { | |
| "dpo_loss": 0.5741093754768372, | |
| "epoch": 0.7652338214454416, | |
| "grad_norm": 119.65644724778777, | |
| "learning_rate": 9.91335832568129e-07, | |
| "logits": -1.412257432937622, | |
| "logps": -85.09854125976562, | |
| "loss": 0.5219, | |
| "objective": 0.5582370758056641, | |
| "ranking_idealized": 0.5458333492279053, | |
| "ranking_idealized_expo": 0.5291666388511658, | |
| "ranking_simple": 0.5416666865348816, | |
| "regularize": 0.5582370758056641, | |
| "step": 135 | |
| }, | |
| { | |
| "dpo_loss": 0.5746143460273743, | |
| "epoch": 0.7935758148323099, | |
| "grad_norm": 117.38563727058376, | |
| "learning_rate": 9.894012231073895e-07, | |
| "logits": -1.307041049003601, | |
| "logps": -83.801025390625, | |
| "loss": 0.5804, | |
| "objective": 0.5208708643913269, | |
| "ranking_idealized": 0.5249999761581421, | |
| "ranking_idealized_expo": 0.5041666626930237, | |
| "ranking_simple": 0.5291666388511658, | |
| "regularize": 0.5208708643913269, | |
| "step": 140 | |
| }, | |
| { | |
| "dpo_loss": 0.5846189856529236, | |
| "epoch": 0.821917808219178, | |
| "grad_norm": 130.2534703806674, | |
| "learning_rate": 9.872741094484964e-07, | |
| "logits": -1.3578908443450928, | |
| "logps": -85.40940856933594, | |
| "loss": 0.5595, | |
| "objective": 0.5616883635520935, | |
| "ranking_idealized": 0.5041666626930237, | |
| "ranking_idealized_expo": 0.4833333194255829, | |
| "ranking_simple": 0.48750001192092896, | |
| "regularize": 0.5616883635520935, | |
| "step": 145 | |
| }, | |
| { | |
| "dpo_loss": 0.5902164578437805, | |
| "epoch": 0.8502598016060463, | |
| "grad_norm": 124.65999306889285, | |
| "learning_rate": 9.849553282839024e-07, | |
| "logits": -1.3182071447372437, | |
| "logps": -83.66584014892578, | |
| "loss": 0.5468, | |
| "objective": 0.5547689199447632, | |
| "ranking_idealized": 0.5541666746139526, | |
| "ranking_idealized_expo": 0.5375000238418579, | |
| "ranking_simple": 0.5333333611488342, | |
| "regularize": 0.5547689199447632, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.8502598016060463, | |
| "eval_dpo_loss": 0.7061101198196411, | |
| "eval_logits": -1.3204107284545898, | |
| "eval_logps": -91.07585144042969, | |
| "eval_loss": 0.5596091151237488, | |
| "eval_objective": 0.5683301687240601, | |
| "eval_ranking_idealized": 0.5247933864593506, | |
| "eval_ranking_idealized_expo": 0.5092975497245789, | |
| "eval_ranking_simple": 0.5144628286361694, | |
| "eval_regularize": 0.5683301687240601, | |
| "eval_runtime": 262.2295, | |
| "eval_samples_per_second": 22.08, | |
| "eval_steps_per_second": 0.923, | |
| "step": 150 | |
| }, | |
| { | |
| "dpo_loss": 0.6094241738319397, | |
| "epoch": 0.8786017949929145, | |
| "grad_norm": 120.42788649324545, | |
| "learning_rate": 9.824457916977784e-07, | |
| "logits": -1.3233551979064941, | |
| "logps": -83.14967346191406, | |
| "loss": 0.5545, | |
| "objective": 0.5938875675201416, | |
| "ranking_idealized": 0.5541666746139526, | |
| "ranking_idealized_expo": 0.512499988079071, | |
| "ranking_simple": 0.5416666865348816, | |
| "regularize": 0.5938875675201416, | |
| "step": 155 | |
| }, | |
| { | |
| "dpo_loss": 0.5945659875869751, | |
| "epoch": 0.9069437883797827, | |
| "grad_norm": 130.17538489613264, | |
| "learning_rate": 9.797464868072486e-07, | |
| "logits": -1.2508231401443481, | |
| "logps": -84.22561645507812, | |
| "loss": 0.5582, | |
| "objective": 0.5726324319839478, | |
| "ranking_idealized": 0.5041666626930237, | |
| "ranking_idealized_expo": 0.4958333373069763, | |
| "ranking_simple": 0.5, | |
| "regularize": 0.5726324319839478, | |
| "step": 160 | |
| }, | |
| { | |
| "dpo_loss": 0.5802300572395325, | |
| "epoch": 0.9352857817666509, | |
| "grad_norm": 108.46916479196223, | |
| "learning_rate": 9.768584753741134e-07, | |
| "logits": -1.2593294382095337, | |
| "logps": -84.2457046508789, | |
| "loss": 0.5428, | |
| "objective": 0.5054454207420349, | |
| "ranking_idealized": 0.5416666865348816, | |
| "ranking_idealized_expo": 0.5208333134651184, | |
| "ranking_simple": 0.550000011920929, | |
| "regularize": 0.5054454207420349, | |
| "step": 165 | |
| }, | |
| { | |
| "dpo_loss": 0.5963813066482544, | |
| "epoch": 0.9636277751535192, | |
| "grad_norm": 112.45672430142012, | |
| "learning_rate": 9.737828933872073e-07, | |
| "logits": -1.3053935766220093, | |
| "logps": -84.02145385742188, | |
| "loss": 0.5633, | |
| "objective": 0.5686451196670532, | |
| "ranking_idealized": 0.5416666865348816, | |
| "ranking_idealized_expo": 0.512499988079071, | |
| "ranking_simple": 0.5333333611488342, | |
| "regularize": 0.5686451196670532, | |
| "step": 170 | |
| }, | |
| { | |
| "dpo_loss": 0.6094807982444763, | |
| "epoch": 0.9919697685403873, | |
| "grad_norm": 112.35460354829323, | |
| "learning_rate": 9.705209506155634e-07, | |
| "logits": -1.2585358619689941, | |
| "logps": -84.55461883544922, | |
| "loss": 0.5613, | |
| "objective": 0.567017674446106, | |
| "ranking_idealized": 0.5958333611488342, | |
| "ranking_idealized_expo": 0.574999988079071, | |
| "ranking_simple": 0.5874999761581421, | |
| "regularize": 0.567017674446106, | |
| "step": 175 | |
| }, | |
| { | |
| "dpo_loss": 0.5619717836380005, | |
| "epoch": 1.0203117619272555, | |
| "grad_norm": 111.48678467404953, | |
| "learning_rate": 9.670739301325534e-07, | |
| "logits": -1.4066673517227173, | |
| "logps": -83.15278625488281, | |
| "loss": 0.567, | |
| "objective": 0.633197009563446, | |
| "ranking_idealized": 0.550000011920929, | |
| "ranking_idealized_expo": 0.5416666865348816, | |
| "ranking_simple": 0.5416666865348816, | |
| "regularize": 0.633197009563446, | |
| "step": 180 | |
| }, | |
| { | |
| "dpo_loss": 0.5666179656982422, | |
| "epoch": 1.0486537553141237, | |
| "grad_norm": 131.38985186167173, | |
| "learning_rate": 9.63443187811197e-07, | |
| "logits": -1.2431377172470093, | |
| "logps": -83.49596405029297, | |
| "loss": 0.5655, | |
| "objective": 0.5990104079246521, | |
| "ranking_idealized": 0.5208333134651184, | |
| "ranking_idealized_expo": 0.4833333194255829, | |
| "ranking_simple": 0.512499988079071, | |
| "regularize": 0.5990104079246521, | |
| "step": 185 | |
| }, | |
| { | |
| "dpo_loss": 0.5910226106643677, | |
| "epoch": 1.076995748700992, | |
| "grad_norm": 121.16034753219803, | |
| "learning_rate": 9.596301517908328e-07, | |
| "logits": -1.3014639616012573, | |
| "logps": -85.08562469482422, | |
| "loss": 0.5836, | |
| "objective": 0.6070137023925781, | |
| "ranking_idealized": 0.5833333134651184, | |
| "ranking_idealized_expo": 0.550000011920929, | |
| "ranking_simple": 0.5833333134651184, | |
| "regularize": 0.6070137023925781, | |
| "step": 190 | |
| }, | |
| { | |
| "dpo_loss": 0.6265178918838501, | |
| "epoch": 1.10533774208786, | |
| "grad_norm": 113.24543776185278, | |
| "learning_rate": 9.556363219153662e-07, | |
| "logits": -1.2846567630767822, | |
| "logps": -83.17445373535156, | |
| "loss": 0.5509, | |
| "objective": 0.5626471638679504, | |
| "ranking_idealized": 0.574999988079071, | |
| "ranking_idealized_expo": 0.5666666626930237, | |
| "ranking_simple": 0.5833333134651184, | |
| "regularize": 0.5626471638679504, | |
| "step": 195 | |
| }, | |
| { | |
| "dpo_loss": 0.5952022671699524, | |
| "epoch": 1.1336797354747283, | |
| "grad_norm": 109.3282186534433, | |
| "learning_rate": 9.514632691433106e-07, | |
| "logits": -1.3481907844543457, | |
| "logps": -82.64450073242188, | |
| "loss": 0.5501, | |
| "objective": 0.5247953534126282, | |
| "ranking_idealized": 0.5208333134651184, | |
| "ranking_idealized_expo": 0.5083333253860474, | |
| "ranking_simple": 0.5249999761581421, | |
| "regularize": 0.5247953534126282, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.1336797354747283, | |
| "eval_dpo_loss": 0.7155318856239319, | |
| "eval_logits": -1.3003575801849365, | |
| "eval_logps": -89.7054672241211, | |
| "eval_loss": 0.6048148274421692, | |
| "eval_objective": 0.5995473861694336, | |
| "eval_ranking_idealized": 0.5247933864593506, | |
| "eval_ranking_idealized_expo": 0.5092975497245789, | |
| "eval_ranking_simple": 0.5061983466148376, | |
| "eval_regularize": 0.5995473861694336, | |
| "eval_runtime": 259.4012, | |
| "eval_samples_per_second": 22.321, | |
| "eval_steps_per_second": 0.933, | |
| "step": 200 | |
| }, | |
| { | |
| "dpo_loss": 0.5928524136543274, | |
| "epoch": 1.1620217288615966, | |
| "grad_norm": 117.4507736272679, | |
| "learning_rate": 9.471126349298556e-07, | |
| "logits": -1.304435133934021, | |
| "logps": -83.2208480834961, | |
| "loss": 0.5617, | |
| "objective": 0.569757878780365, | |
| "ranking_idealized": 0.5, | |
| "ranking_idealized_expo": 0.48750001192092896, | |
| "ranking_simple": 0.5041666626930237, | |
| "regularize": 0.569757878780365, | |
| "step": 205 | |
| }, | |
| { | |
| "dpo_loss": 0.5430352687835693, | |
| "epoch": 1.1903637222484649, | |
| "grad_norm": 108.28365774149364, | |
| "learning_rate": 9.425861305812081e-07, | |
| "logits": -1.28606116771698, | |
| "logps": -83.92695617675781, | |
| "loss": 0.5256, | |
| "objective": 0.4972154200077057, | |
| "ranking_idealized": 0.5083333253860474, | |
| "ranking_idealized_expo": 0.4958333373069763, | |
| "ranking_simple": 0.5083333253860474, | |
| "regularize": 0.4972154200077057, | |
| "step": 210 | |
| }, | |
| { | |
| "dpo_loss": 0.6411459445953369, | |
| "epoch": 1.2187057156353331, | |
| "grad_norm": 117.64538520858981, | |
| "learning_rate": 9.378855365814557e-07, | |
| "logits": -1.2710984945297241, | |
| "logps": -83.3528823852539, | |
| "loss": 0.5386, | |
| "objective": 0.6103346943855286, | |
| "ranking_idealized": 0.49166667461395264, | |
| "ranking_idealized_expo": 0.4833333194255829, | |
| "ranking_simple": 0.5041666626930237, | |
| "regularize": 0.6103346943855286, | |
| "step": 215 | |
| }, | |
| { | |
| "dpo_loss": 0.6046126484870911, | |
| "epoch": 1.2470477090222012, | |
| "grad_norm": 112.2724627593423, | |
| "learning_rate": 9.330127018922193e-07, | |
| "logits": -1.22818124294281, | |
| "logps": -82.01517486572266, | |
| "loss": 0.5413, | |
| "objective": 0.5305136442184448, | |
| "ranking_idealized": 0.6083333492279053, | |
| "ranking_idealized_expo": 0.6000000238418579, | |
| "ranking_simple": 0.6083333492279053, | |
| "regularize": 0.5305136442184448, | |
| "step": 220 | |
| }, | |
| { | |
| "dpo_loss": 0.5910329222679138, | |
| "epoch": 1.2753897024090695, | |
| "grad_norm": 114.94988107695234, | |
| "learning_rate": 9.279695432253708e-07, | |
| "logits": -1.3015425205230713, | |
| "logps": -83.29090881347656, | |
| "loss": 0.5262, | |
| "objective": 0.528804361820221, | |
| "ranking_idealized": 0.5375000238418579, | |
| "ranking_idealized_expo": 0.5291666388511658, | |
| "ranking_simple": 0.5375000238418579, | |
| "regularize": 0.528804361820221, | |
| "step": 225 | |
| }, | |
| { | |
| "dpo_loss": 0.5783368945121765, | |
| "epoch": 1.3037316957959377, | |
| "grad_norm": 112.00444337751402, | |
| "learning_rate": 9.227580442891021e-07, | |
| "logits": -1.2694045305252075, | |
| "logps": -82.34114837646484, | |
| "loss": 0.5259, | |
| "objective": 0.5298904180526733, | |
| "ranking_idealized": 0.5291666388511658, | |
| "ranking_idealized_expo": 0.5083333253860474, | |
| "ranking_simple": 0.5208333134651184, | |
| "regularize": 0.5298903584480286, | |
| "step": 230 | |
| }, | |
| { | |
| "dpo_loss": 0.559749186038971, | |
| "epoch": 1.3320736891828058, | |
| "grad_norm": 117.1742133311728, | |
| "learning_rate": 9.173802550076401e-07, | |
| "logits": -1.3496894836425781, | |
| "logps": -80.74533081054688, | |
| "loss": 0.5275, | |
| "objective": 0.5420801639556885, | |
| "ranking_idealized": 0.512499988079071, | |
| "ranking_idealized_expo": 0.4833333194255829, | |
| "ranking_simple": 0.512499988079071, | |
| "regularize": 0.5420801639556885, | |
| "step": 235 | |
| }, | |
| { | |
| "dpo_loss": 0.5604056715965271, | |
| "epoch": 1.360415682569674, | |
| "grad_norm": 120.20610234512412, | |
| "learning_rate": 9.118382907149163e-07, | |
| "logits": -1.2856090068817139, | |
| "logps": -81.93352508544922, | |
| "loss": 0.5307, | |
| "objective": 0.4905354678630829, | |
| "ranking_idealized": 0.4958333373069763, | |
| "ranking_idealized_expo": 0.4791666567325592, | |
| "ranking_simple": 0.49166667461395264, | |
| "regularize": 0.4905354678630829, | |
| "step": 240 | |
| }, | |
| { | |
| "dpo_loss": 0.5622699856758118, | |
| "epoch": 1.3887576759565423, | |
| "grad_norm": 112.51743847797803, | |
| "learning_rate": 9.061343313225087e-07, | |
| "logits": -1.3067680597305298, | |
| "logps": -81.71417999267578, | |
| "loss": 0.4769, | |
| "objective": 0.4980693757534027, | |
| "ranking_idealized": 0.5249999761581421, | |
| "ranking_idealized_expo": 0.5166666507720947, | |
| "ranking_simple": 0.5416666865348816, | |
| "regularize": 0.4980693757534027, | |
| "step": 245 | |
| }, | |
| { | |
| "dpo_loss": 0.5571755766868591, | |
| "epoch": 1.4170996693434104, | |
| "grad_norm": 105.16402218584498, | |
| "learning_rate": 9.002706204621802e-07, | |
| "logits": -1.2600089311599731, | |
| "logps": -81.47142028808594, | |
| "loss": 0.5045, | |
| "objective": 0.48103129863739014, | |
| "ranking_idealized": 0.48750001192092896, | |
| "ranking_idealized_expo": 0.47083333134651184, | |
| "ranking_simple": 0.48750001192092896, | |
| "regularize": 0.48103129863739014, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.4170996693434104, | |
| "eval_dpo_loss": 0.7336850166320801, | |
| "eval_logits": -1.319351077079773, | |
| "eval_logps": -88.7256088256836, | |
| "eval_loss": 0.6191251873970032, | |
| "eval_objective": 0.6396905779838562, | |
| "eval_ranking_idealized": 0.5247933864593506, | |
| "eval_ranking_idealized_expo": 0.5092975497245789, | |
| "eval_ranking_simple": 0.5175619721412659, | |
| "eval_regularize": 0.6396905779838562, | |
| "eval_runtime": 261.8978, | |
| "eval_samples_per_second": 22.108, | |
| "eval_steps_per_second": 0.924, | |
| "step": 250 | |
| }, | |
| { | |
| "dpo_loss": 0.5806334614753723, | |
| "epoch": 1.4454416627302786, | |
| "grad_norm": 108.9060179401822, | |
| "learning_rate": 8.942494646033554e-07, | |
| "logits": -1.268131971359253, | |
| "logps": -82.72068786621094, | |
| "loss": 0.5226, | |
| "objective": 0.5074787735939026, | |
| "ranking_idealized": 0.5708333253860474, | |
| "ranking_idealized_expo": 0.5458333492279053, | |
| "ranking_simple": 0.5625, | |
| "regularize": 0.5074787735939026, | |
| "step": 255 | |
| }, | |
| { | |
| "dpo_loss": 0.6120874285697937, | |
| "epoch": 1.473783656117147, | |
| "grad_norm": 113.8812403973121, | |
| "learning_rate": 8.880732321458784e-07, | |
| "logits": -1.3066486120224, | |
| "logps": -84.5510482788086, | |
| "loss": 0.5203, | |
| "objective": 0.5250194072723389, | |
| "ranking_idealized": 0.5041666626930237, | |
| "ranking_idealized_expo": 0.48750001192092896, | |
| "ranking_simple": 0.5041666626930237, | |
| "regularize": 0.5250194072723389, | |
| "step": 260 | |
| }, | |
| { | |
| "dpo_loss": 0.5722587704658508, | |
| "epoch": 1.5021256495040152, | |
| "grad_norm": 107.47217195176817, | |
| "learning_rate": 8.817443524884117e-07, | |
| "logits": -1.302085280418396, | |
| "logps": -82.64102935791016, | |
| "loss": 0.5125, | |
| "objective": 0.5349072813987732, | |
| "ranking_idealized": 0.5249999761581421, | |
| "ranking_idealized_expo": 0.5041666626930237, | |
| "ranking_simple": 0.5375000238418579, | |
| "regularize": 0.5349072813987732, | |
| "step": 265 | |
| }, | |
| { | |
| "dpo_loss": 0.5912469625473022, | |
| "epoch": 1.5304676428908834, | |
| "grad_norm": 102.61360328866256, | |
| "learning_rate": 8.752653150728411e-07, | |
| "logits": -1.33375883102417, | |
| "logps": -83.93022918701172, | |
| "loss": 0.4989, | |
| "objective": 0.5215144753456116, | |
| "ranking_idealized": 0.5208333134651184, | |
| "ranking_idealized_expo": 0.5041666626930237, | |
| "ranking_simple": 0.5166666507720947, | |
| "regularize": 0.5215144753456116, | |
| "step": 270 | |
| }, | |
| { | |
| "dpo_loss": 0.5620048642158508, | |
| "epoch": 1.5588096362777515, | |
| "grad_norm": 120.75621294263999, | |
| "learning_rate": 8.68638668405062e-07, | |
| "logits": -1.346874713897705, | |
| "logps": -83.51717376708984, | |
| "loss": 0.508, | |
| "objective": 0.49223792552948, | |
| "ranking_idealized": 0.550000011920929, | |
| "ranking_idealized_expo": 0.5249999761581421, | |
| "ranking_simple": 0.5416666865348816, | |
| "regularize": 0.49223792552948, | |
| "step": 275 | |
| }, | |
| { | |
| "dpo_loss": 0.5491313934326172, | |
| "epoch": 1.5871516296646198, | |
| "grad_norm": 111.38123469249379, | |
| "learning_rate": 8.61867019052535e-07, | |
| "logits": -1.2457977533340454, | |
| "logps": -82.07854461669922, | |
| "loss": 0.4912, | |
| "objective": 0.4965251088142395, | |
| "ranking_idealized": 0.5291666388511658, | |
| "ranking_idealized_expo": 0.5041666626930237, | |
| "ranking_simple": 0.5249999761581421, | |
| "regularize": 0.4965251088142395, | |
| "step": 280 | |
| }, | |
| { | |
| "dpo_loss": 0.597549319267273, | |
| "epoch": 1.615493623051488, | |
| "grad_norm": 108.78031876782775, | |
| "learning_rate": 8.549530306190014e-07, | |
| "logits": -1.3423717021942139, | |
| "logps": -83.64887237548828, | |
| "loss": 0.4878, | |
| "objective": 0.5166993141174316, | |
| "ranking_idealized": 0.5375000238418579, | |
| "ranking_idealized_expo": 0.5166666507720947, | |
| "ranking_simple": 0.5416666865348816, | |
| "regularize": 0.5166993141174316, | |
| "step": 285 | |
| }, | |
| { | |
| "dpo_loss": 0.5523730516433716, | |
| "epoch": 1.643835616438356, | |
| "grad_norm": 101.86340958175018, | |
| "learning_rate": 8.478994226967638e-07, | |
| "logits": -1.3935832977294922, | |
| "logps": -81.38288116455078, | |
| "loss": 0.4753, | |
| "objective": 0.45616135001182556, | |
| "ranking_idealized": 0.5708333253860474, | |
| "ranking_idealized_expo": 0.5625, | |
| "ranking_simple": 0.5791666507720947, | |
| "regularize": 0.45616135001182556, | |
| "step": 290 | |
| }, | |
| { | |
| "dpo_loss": 0.5671105980873108, | |
| "epoch": 1.6721776098252243, | |
| "grad_norm": 97.05334097299942, | |
| "learning_rate": 8.407089697969456e-07, | |
| "logits": -1.3089579343795776, | |
| "logps": -79.79814910888672, | |
| "loss": 0.4806, | |
| "objective": 0.48036739230155945, | |
| "ranking_idealized": 0.4958333373069763, | |
| "ranking_idealized_expo": 0.4791666567325592, | |
| "ranking_simple": 0.5041666626930237, | |
| "regularize": 0.48036739230155945, | |
| "step": 295 | |
| }, | |
| { | |
| "dpo_loss": 0.5690053701400757, | |
| "epoch": 1.7005196032120926, | |
| "grad_norm": 103.94959921577468, | |
| "learning_rate": 8.333845002581458e-07, | |
| "logits": -1.3550407886505127, | |
| "logps": -81.06673431396484, | |
| "loss": 0.5191, | |
| "objective": 0.5282995700836182, | |
| "ranking_idealized": 0.5916666388511658, | |
| "ranking_idealized_expo": 0.574999988079071, | |
| "ranking_simple": 0.6041666865348816, | |
| "regularize": 0.5282995700836182, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.7005196032120926, | |
| "eval_dpo_loss": 0.7341631650924683, | |
| "eval_logits": -1.3167152404785156, | |
| "eval_logps": -87.61885833740234, | |
| "eval_loss": 0.6501524448394775, | |
| "eval_objective": 0.6544150114059448, | |
| "eval_ranking_idealized": 0.5247933864593506, | |
| "eval_ranking_idealized_expo": 0.5092975497245789, | |
| "eval_ranking_simple": 0.5165289044380188, | |
| "eval_regularize": 0.6544150114059448, | |
| "eval_runtime": 258.5096, | |
| "eval_samples_per_second": 22.398, | |
| "eval_steps_per_second": 0.936, | |
| "step": 300 | |
| }, | |
| { | |
| "dpo_loss": 0.5860777497291565, | |
| "epoch": 1.7288615965989607, | |
| "grad_norm": 101.19563914464877, | |
| "learning_rate": 8.259288951339232e-07, | |
| "logits": -1.290361762046814, | |
| "logps": -81.49486541748047, | |
| "loss": 0.4715, | |
| "objective": 0.5236480236053467, | |
| "ranking_idealized": 0.5208333134651184, | |
| "ranking_idealized_expo": 0.5041666626930237, | |
| "ranking_simple": 0.5333333611488342, | |
| "regularize": 0.5236480236053467, | |
| "step": 305 | |
| }, | |
| { | |
| "dpo_loss": 0.5874544382095337, | |
| "epoch": 1.7572035899858292, | |
| "grad_norm": 101.32381454333631, | |
| "learning_rate": 8.183450870595441e-07, | |
| "logits": -1.323080062866211, | |
| "logps": -79.28206634521484, | |
| "loss": 0.4743, | |
| "objective": 0.4995422959327698, | |
| "ranking_idealized": 0.5375000238418579, | |
| "ranking_idealized_expo": 0.5208333134651184, | |
| "ranking_simple": 0.5333333611488342, | |
| "regularize": 0.4995422959327698, | |
| "step": 310 | |
| }, | |
| { | |
| "dpo_loss": 0.5733292698860168, | |
| "epoch": 1.7855455833726972, | |
| "grad_norm": 101.22251777355339, | |
| "learning_rate": 8.106360590984404e-07, | |
| "logits": -1.2535948753356934, | |
| "logps": -80.41255950927734, | |
| "loss": 0.4496, | |
| "objective": 0.44809386134147644, | |
| "ranking_idealized": 0.5208333134651184, | |
| "ranking_idealized_expo": 0.5083333253860474, | |
| "ranking_simple": 0.5166666507720947, | |
| "regularize": 0.44809386134147644, | |
| "step": 315 | |
| }, | |
| { | |
| "dpo_loss": 0.5748218297958374, | |
| "epoch": 1.8138875767595655, | |
| "grad_norm": 102.63800844635435, | |
| "learning_rate": 8.028048435688333e-07, | |
| "logits": -1.3195849657058716, | |
| "logps": -82.20491027832031, | |
| "loss": 0.464, | |
| "objective": 0.4412696659564972, | |
| "ranking_idealized": 0.5291666388511658, | |
| "ranking_idealized_expo": 0.5041666626930237, | |
| "ranking_simple": 0.5249999761581421, | |
| "regularize": 0.4412696659564972, | |
| "step": 320 | |
| }, | |
| { | |
| "dpo_loss": 0.5583847761154175, | |
| "epoch": 1.8422295701464337, | |
| "grad_norm": 106.29080888058539, | |
| "learning_rate": 7.948545208509811e-07, | |
| "logits": -1.3646790981292725, | |
| "logps": -83.66368865966797, | |
| "loss": 0.4757, | |
| "objective": 0.43752506375312805, | |
| "ranking_idealized": 0.5583333373069763, | |
| "ranking_idealized_expo": 0.5166666507720947, | |
| "ranking_simple": 0.5416666865348816, | |
| "regularize": 0.43752506375312805, | |
| "step": 325 | |
| }, | |
| { | |
| "dpo_loss": 0.5669862627983093, | |
| "epoch": 1.8705715635333018, | |
| "grad_norm": 114.61344867316569, | |
| "learning_rate": 7.86788218175523e-07, | |
| "logits": -1.1525540351867676, | |
| "logps": -81.54851531982422, | |
| "loss": 0.4342, | |
| "objective": 0.4504217505455017, | |
| "ranking_idealized": 0.48750001192092896, | |
| "ranking_idealized_expo": 0.47083333134651184, | |
| "ranking_simple": 0.48750001192092896, | |
| "regularize": 0.4504217505455017, | |
| "step": 330 | |
| }, | |
| { | |
| "dpo_loss": 0.5577248334884644, | |
| "epoch": 1.89891355692017, | |
| "grad_norm": 107.68338896171758, | |
| "learning_rate": 7.786091083933949e-07, | |
| "logits": -1.3191189765930176, | |
| "logps": -82.01612091064453, | |
| "loss": 0.4554, | |
| "objective": 0.4512566030025482, | |
| "ranking_idealized": 0.5249999761581421, | |
| "ranking_idealized_expo": 0.4958333373069763, | |
| "ranking_simple": 0.5208333134651184, | |
| "regularize": 0.4512566030025482, | |
| "step": 335 | |
| }, | |
| { | |
| "dpo_loss": 0.5417340993881226, | |
| "epoch": 1.9272555503070383, | |
| "grad_norm": 113.63225902754026, | |
| "learning_rate": 7.703204087277988e-07, | |
| "logits": -1.311868667602539, | |
| "logps": -84.563720703125, | |
| "loss": 0.4472, | |
| "objective": 0.4667154550552368, | |
| "ranking_idealized": 0.5541666746139526, | |
| "ranking_idealized_expo": 0.5416666865348816, | |
| "ranking_simple": 0.5666666626930237, | |
| "regularize": 0.46671542525291443, | |
| "step": 340 | |
| }, | |
| { | |
| "dpo_loss": 0.577439546585083, | |
| "epoch": 1.9555975436939064, | |
| "grad_norm": 105.47346696011448, | |
| "learning_rate": 7.619253795087208e-07, | |
| "logits": -1.309786081314087, | |
| "logps": -82.33320617675781, | |
| "loss": 0.4317, | |
| "objective": 0.45227622985839844, | |
| "ranking_idealized": 0.5416666865348816, | |
| "ranking_idealized_expo": 0.5333333611488342, | |
| "ranking_simple": 0.5416666865348816, | |
| "regularize": 0.45227622985839844, | |
| "step": 345 | |
| }, | |
| { | |
| "dpo_loss": 0.5885747671127319, | |
| "epoch": 1.9839395370807746, | |
| "grad_norm": 109.05004528153341, | |
| "learning_rate": 7.534273228904915e-07, | |
| "logits": -1.22944974899292, | |
| "logps": -83.1298599243164, | |
| "loss": 0.4473, | |
| "objective": 0.4539943039417267, | |
| "ranking_idealized": 0.5041666626930237, | |
| "ranking_idealized_expo": 0.48750001192092896, | |
| "ranking_simple": 0.5166666507720947, | |
| "regularize": 0.4539943039417267, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.9839395370807746, | |
| "eval_dpo_loss": 0.7476473450660706, | |
| "eval_logits": -1.3002201318740845, | |
| "eval_logps": -88.6204833984375, | |
| "eval_loss": 0.6869083642959595, | |
| "eval_objective": 0.6877632737159729, | |
| "eval_ranking_idealized": 0.5247933864593506, | |
| "eval_ranking_idealized_expo": 0.5092975497245789, | |
| "eval_ranking_simple": 0.5092975497245789, | |
| "eval_regularize": 0.6877632737159729, | |
| "eval_runtime": 258.9539, | |
| "eval_samples_per_second": 22.359, | |
| "eval_steps_per_second": 0.935, | |
| "step": 350 | |
| }, | |
| { | |
| "dpo_loss": 0.5712209343910217, | |
| "epoch": 2.012281530467643, | |
| "grad_norm": 101.66242864172723, | |
| "learning_rate": 7.448295815528956e-07, | |
| "logits": -1.2166054248809814, | |
| "logps": -82.02034759521484, | |
| "loss": 0.4247, | |
| "objective": 0.4026139974594116, | |
| "ranking_idealized": 0.5208333134651184, | |
| "ranking_idealized_expo": 0.5083333253860474, | |
| "ranking_simple": 0.5166666507720947, | |
| "regularize": 0.4026139974594116, | |
| "step": 355 | |
| }, | |
| { | |
| "dpo_loss": 0.5501471161842346, | |
| "epoch": 2.040623523854511, | |
| "grad_norm": 111.09255218939096, | |
| "learning_rate": 7.361355373863413e-07, | |
| "logits": -1.2762349843978882, | |
| "logps": -80.2708969116211, | |
| "loss": 0.4188, | |
| "objective": 0.3990139663219452, | |
| "ranking_idealized": 0.5541666746139526, | |
| "ranking_idealized_expo": 0.5291666388511658, | |
| "ranking_simple": 0.5541666746139526, | |
| "regularize": 0.3990139663219452, | |
| "step": 360 | |
| }, | |
| { | |
| "dpo_loss": 0.5589497089385986, | |
| "epoch": 2.0689655172413794, | |
| "grad_norm": 102.40571901595573, | |
| "learning_rate": 7.273486101616056e-07, | |
| "logits": -1.3285800218582153, | |
| "logps": -82.2364730834961, | |
| "loss": 0.4468, | |
| "objective": 0.4508870840072632, | |
| "ranking_idealized": 0.5249999761581421, | |
| "ranking_idealized_expo": 0.5041666626930237, | |
| "ranking_simple": 0.5333333611488342, | |
| "regularize": 0.4508870840072632, | |
| "step": 365 | |
| }, | |
| { | |
| "dpo_loss": 0.5617626309394836, | |
| "epoch": 2.0973075106282475, | |
| "grad_norm": 97.76405475943756, | |
| "learning_rate": 7.184722561846797e-07, | |
| "logits": -1.2896106243133545, | |
| "logps": -79.0242691040039, | |
| "loss": 0.4108, | |
| "objective": 0.4170443117618561, | |
| "ranking_idealized": 0.5041666626930237, | |
| "ranking_idealized_expo": 0.4833333194255829, | |
| "ranking_simple": 0.5, | |
| "regularize": 0.4170443117618561, | |
| "step": 370 | |
| }, | |
| { | |
| "dpo_loss": 0.5481951236724854, | |
| "epoch": 2.1256495040151155, | |
| "grad_norm": 102.54464966495118, | |
| "learning_rate": 7.095099669372443e-07, | |
| "logits": -1.2680145502090454, | |
| "logps": -80.49474334716797, | |
| "loss": 0.4088, | |
| "objective": 0.3847118020057678, | |
| "ranking_idealized": 0.625, | |
| "ranking_idealized_expo": 0.5874999761581421, | |
| "ranking_simple": 0.612500011920929, | |
| "regularize": 0.3847118020057678, | |
| "step": 375 | |
| }, | |
| { | |
| "dpo_loss": 0.5886502861976624, | |
| "epoch": 2.153991497401984, | |
| "grad_norm": 99.36933926359579, | |
| "learning_rate": 7.004652677033068e-07, | |
| "logits": -1.2002967596054077, | |
| "logps": -79.81450653076172, | |
| "loss": 0.4196, | |
| "objective": 0.43431776762008667, | |
| "ranking_idealized": 0.5458333492279053, | |
| "ranking_idealized_expo": 0.5291666388511658, | |
| "ranking_simple": 0.5416666865348816, | |
| "regularize": 0.43431776762008667, | |
| "step": 380 | |
| }, | |
| { | |
| "dpo_loss": 0.566298246383667, | |
| "epoch": 2.182333490788852, | |
| "grad_norm": 100.41990113674308, | |
| "learning_rate": 6.913417161825449e-07, | |
| "logits": -1.307159423828125, | |
| "logps": -81.99685668945312, | |
| "loss": 0.4125, | |
| "objective": 0.4449020326137543, | |
| "ranking_idealized": 0.5249999761581421, | |
| "ranking_idealized_expo": 0.5, | |
| "ranking_simple": 0.5375000238418579, | |
| "regularize": 0.4449020326137543, | |
| "step": 385 | |
| }, | |
| { | |
| "dpo_loss": 0.5851306319236755, | |
| "epoch": 2.21067548417572, | |
| "grad_norm": 101.58148277220718, | |
| "learning_rate": 6.821429010908971e-07, | |
| "logits": -1.2254120111465454, | |
| "logps": -80.0687484741211, | |
| "loss": 0.4184, | |
| "objective": 0.41351255774497986, | |
| "ranking_idealized": 0.48750001192092896, | |
| "ranking_idealized_expo": 0.4833333194255829, | |
| "ranking_simple": 0.4791666567325592, | |
| "regularize": 0.41351255774497986, | |
| "step": 390 | |
| }, | |
| { | |
| "dpo_loss": 0.5543534159660339, | |
| "epoch": 2.2390174775625886, | |
| "grad_norm": 102.754307701718, | |
| "learning_rate": 6.728724407489553e-07, | |
| "logits": -1.2246791124343872, | |
| "logps": -80.76174926757812, | |
| "loss": 0.394, | |
| "objective": 0.3953525424003601, | |
| "ranking_idealized": 0.5416666865348816, | |
| "ranking_idealized_expo": 0.5166666507720947, | |
| "ranking_simple": 0.5291666388511658, | |
| "regularize": 0.3953525424003601, | |
| "step": 395 | |
| }, | |
| { | |
| "dpo_loss": 0.5204497575759888, | |
| "epoch": 2.2673594709494567, | |
| "grad_norm": 105.48313372380032, | |
| "learning_rate": 6.635339816587108e-07, | |
| "logits": -1.2704825401306152, | |
| "logps": -82.68233489990234, | |
| "loss": 0.3926, | |
| "objective": 0.4119986891746521, | |
| "ranking_idealized": 0.6000000238418579, | |
| "ranking_idealized_expo": 0.574999988079071, | |
| "ranking_simple": 0.6208333373069763, | |
| "regularize": 0.4119986891746521, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.2673594709494567, | |
| "eval_dpo_loss": 0.7518758773803711, | |
| "eval_logits": -1.273954153060913, | |
| "eval_logps": -87.79329681396484, | |
| "eval_loss": 0.708670973777771, | |
| "eval_objective": 0.7146722078323364, | |
| "eval_ranking_idealized": 0.5247933864593506, | |
| "eval_ranking_idealized_expo": 0.5092975497245789, | |
| "eval_ranking_simple": 0.5113636255264282, | |
| "eval_regularize": 0.7146722078323364, | |
| "eval_runtime": 259.1203, | |
| "eval_samples_per_second": 22.345, | |
| "eval_steps_per_second": 0.934, | |
| "step": 400 | |
| }, | |
| { | |
| "dpo_loss": 0.5729255676269531, | |
| "epoch": 2.295701464336325, | |
| "grad_norm": 97.83556162175466, | |
| "learning_rate": 6.541311970692162e-07, | |
| "logits": -1.3065953254699707, | |
| "logps": -81.27420043945312, | |
| "loss": 0.388, | |
| "objective": 0.3636757433414459, | |
| "ranking_idealized": 0.5833333134651184, | |
| "ranking_idealized_expo": 0.5583333373069763, | |
| "ranking_simple": 0.574999988079071, | |
| "regularize": 0.3636757433414459, | |
| "step": 405 | |
| }, | |
| { | |
| "dpo_loss": 0.5597470998764038, | |
| "epoch": 2.324043457723193, | |
| "grad_norm": 103.83650251762533, | |
| "learning_rate": 6.446677855317264e-07, | |
| "logits": -1.1975177526474, | |
| "logps": -80.72605895996094, | |
| "loss": 0.3899, | |
| "objective": 0.39117977023124695, | |
| "ranking_idealized": 0.48750001192092896, | |
| "ranking_idealized_expo": 0.4749999940395355, | |
| "ranking_simple": 0.4958333373069763, | |
| "regularize": 0.39117977023124695, | |
| "step": 410 | |
| }, | |
| { | |
| "dpo_loss": 0.5665189027786255, | |
| "epoch": 2.3523854511100613, | |
| "grad_norm": 98.79350448658165, | |
| "learning_rate": 6.351474694448864e-07, | |
| "logits": -1.227319598197937, | |
| "logps": -81.35487365722656, | |
| "loss": 0.3741, | |
| "objective": 0.38428404927253723, | |
| "ranking_idealized": 0.550000011920929, | |
| "ranking_idealized_expo": 0.5333333611488342, | |
| "ranking_simple": 0.5541666746139526, | |
| "regularize": 0.38428401947021484, | |
| "step": 415 | |
| }, | |
| { | |
| "dpo_loss": 0.5536468029022217, | |
| "epoch": 2.3807274444969297, | |
| "grad_norm": 107.78716307916815, | |
| "learning_rate": 6.255739935905395e-07, | |
| "logits": -1.1353336572647095, | |
| "logps": -83.33781433105469, | |
| "loss": 0.3686, | |
| "objective": 0.3524840176105499, | |
| "ranking_idealized": 0.4791666567325592, | |
| "ranking_idealized_expo": 0.47083333134651184, | |
| "ranking_simple": 0.4833333194255829, | |
| "regularize": 0.3524840176105499, | |
| "step": 420 | |
| }, | |
| { | |
| "dpo_loss": 0.5507148504257202, | |
| "epoch": 2.409069437883798, | |
| "grad_norm": 95.31420015325158, | |
| "learning_rate": 6.159511236607315e-07, | |
| "logits": -1.186567783355713, | |
| "logps": -81.95353698730469, | |
| "loss": 0.3829, | |
| "objective": 0.3914056718349457, | |
| "ranking_idealized": 0.5208333134651184, | |
| "ranking_idealized_expo": 0.4958333373069763, | |
| "ranking_simple": 0.5166666507720947, | |
| "regularize": 0.3914056718349457, | |
| "step": 425 | |
| }, | |
| { | |
| "dpo_loss": 0.5727216601371765, | |
| "epoch": 2.4374114312706663, | |
| "grad_norm": 113.51215777470752, | |
| "learning_rate": 6.062826447764883e-07, | |
| "logits": -1.2177358865737915, | |
| "logps": -82.20878601074219, | |
| "loss": 0.3667, | |
| "objective": 0.38485896587371826, | |
| "ranking_idealized": 0.5416666865348816, | |
| "ranking_idealized_expo": 0.5291666388511658, | |
| "ranking_simple": 0.550000011920929, | |
| "regularize": 0.38485896587371826, | |
| "step": 430 | |
| }, | |
| { | |
| "dpo_loss": 0.5795385837554932, | |
| "epoch": 2.4657534246575343, | |
| "grad_norm": 109.86296645766184, | |
| "learning_rate": 5.965723599989528e-07, | |
| "logits": -1.2774423360824585, | |
| "logps": -81.68889617919922, | |
| "loss": 0.3742, | |
| "objective": 0.373045951128006, | |
| "ranking_idealized": 0.5416666865348816, | |
| "ranking_idealized_expo": 0.5083333253860474, | |
| "ranking_simple": 0.5208333134651184, | |
| "regularize": 0.373045951128006, | |
| "step": 435 | |
| }, | |
| { | |
| "dpo_loss": 0.5365008115768433, | |
| "epoch": 2.4940954180444024, | |
| "grad_norm": 111.81192989308617, | |
| "learning_rate": 5.868240888334652e-07, | |
| "logits": -1.2080743312835693, | |
| "logps": -81.38243865966797, | |
| "loss": 0.3754, | |
| "objective": 0.3666243851184845, | |
| "ranking_idealized": 0.5458333492279053, | |
| "ranking_idealized_expo": 0.5375000238418579, | |
| "ranking_simple": 0.5416666865348816, | |
| "regularize": 0.3666243851184845, | |
| "step": 440 | |
| }, | |
| { | |
| "dpo_loss": 0.5457199811935425, | |
| "epoch": 2.5224374114312704, | |
| "grad_norm": 111.8881514776807, | |
| "learning_rate": 5.770416657271728e-07, | |
| "logits": -1.220347285270691, | |
| "logps": -79.05127716064453, | |
| "loss": 0.3504, | |
| "objective": 0.3269334137439728, | |
| "ranking_idealized": 0.5208333134651184, | |
| "ranking_idealized_expo": 0.5166666507720947, | |
| "ranking_simple": 0.5375000238418579, | |
| "regularize": 0.3269334137439728, | |
| "step": 445 | |
| }, | |
| { | |
| "dpo_loss": 0.5769312977790833, | |
| "epoch": 2.550779404818139, | |
| "grad_norm": 99.97534603124807, | |
| "learning_rate": 5.67228938560766e-07, | |
| "logits": -1.2545174360275269, | |
| "logps": -81.12376403808594, | |
| "loss": 0.3583, | |
| "objective": 0.3686024844646454, | |
| "ranking_idealized": 0.512499988079071, | |
| "ranking_idealized_expo": 0.5, | |
| "ranking_simple": 0.512499988079071, | |
| "regularize": 0.3686024844646454, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 2.550779404818139, | |
| "eval_dpo_loss": 0.7467784881591797, | |
| "eval_logits": -1.2637890577316284, | |
| "eval_logps": -87.71803283691406, | |
| "eval_loss": 0.6997210383415222, | |
| "eval_objective": 0.7073113918304443, | |
| "eval_ranking_idealized": 0.5247933864593506, | |
| "eval_ranking_idealized_expo": 0.5092975497245789, | |
| "eval_ranking_simple": 0.5092975497245789, | |
| "eval_regularize": 0.7073113918304443, | |
| "eval_runtime": 259.5146, | |
| "eval_samples_per_second": 22.311, | |
| "eval_steps_per_second": 0.933, | |
| "step": 450 | |
| }, | |
| { | |
| "dpo_loss": 0.5546112656593323, | |
| "epoch": 2.579121398205007, | |
| "grad_norm": 103.50827450574826, | |
| "learning_rate": 5.573897671349268e-07, | |
| "logits": -1.282201886177063, | |
| "logps": -82.07099151611328, | |
| "loss": 0.3709, | |
| "objective": 0.3527103662490845, | |
| "ranking_idealized": 0.5166666507720947, | |
| "ranking_idealized_expo": 0.5, | |
| "ranking_simple": 0.512499988079071, | |
| "regularize": 0.3527103662490845, | |
| "step": 455 | |
| }, | |
| { | |
| "dpo_loss": 0.5413371920585632, | |
| "epoch": 2.6074633915918755, | |
| "grad_norm": 103.11622894521913, | |
| "learning_rate": 5.475280216520912e-07, | |
| "logits": -1.3678227663040161, | |
| "logps": -82.53510284423828, | |
| "loss": 0.3493, | |
| "objective": 0.3583069443702698, | |
| "ranking_idealized": 0.5291666388511658, | |
| "ranking_idealized_expo": 0.512499988079071, | |
| "ranking_simple": 0.5249999761581421, | |
| "regularize": 0.3583069443702698, | |
| "step": 460 | |
| }, | |
| { | |
| "dpo_loss": 0.5440939664840698, | |
| "epoch": 2.6358053849787435, | |
| "grad_norm": 101.71475758586865, | |
| "learning_rate": 5.376475811941191e-07, | |
| "logits": -1.2582870721817017, | |
| "logps": -81.93384552001953, | |
| "loss": 0.3418, | |
| "objective": 0.3399708867073059, | |
| "ranking_idealized": 0.5083333253860474, | |
| "ranking_idealized_expo": 0.4958333373069763, | |
| "ranking_simple": 0.512499988079071, | |
| "regularize": 0.3399708867073059, | |
| "step": 465 | |
| }, | |
| { | |
| "dpo_loss": 0.5473323464393616, | |
| "epoch": 2.6641473783656116, | |
| "grad_norm": 106.63182084368584, | |
| "learning_rate": 5.277523321964701e-07, | |
| "logits": -1.266352891921997, | |
| "logps": -80.5593490600586, | |
| "loss": 0.3361, | |
| "objective": 0.3204086124897003, | |
| "ranking_idealized": 0.5791666507720947, | |
| "ranking_idealized_expo": 0.550000011920929, | |
| "ranking_simple": 0.5833333134651184, | |
| "regularize": 0.3204086124897003, | |
| "step": 470 | |
| }, | |
| { | |
| "dpo_loss": 0.5676419734954834, | |
| "epoch": 2.69248937175248, | |
| "grad_norm": 100.07291475508246, | |
| "learning_rate": 5.178461669194903e-07, | |
| "logits": -1.2326582670211792, | |
| "logps": -81.09722137451172, | |
| "loss": 0.3221, | |
| "objective": 0.3266924321651459, | |
| "ranking_idealized": 0.512499988079071, | |
| "ranking_idealized_expo": 0.4958333373069763, | |
| "ranking_simple": 0.5083333253860474, | |
| "regularize": 0.3266924321651459, | |
| "step": 475 | |
| }, | |
| { | |
| "dpo_loss": 0.5700205564498901, | |
| "epoch": 2.720831365139348, | |
| "grad_norm": 114.43771495362513, | |
| "learning_rate": 5.07932981917404e-07, | |
| "logits": -1.3194983005523682, | |
| "logps": -80.66010284423828, | |
| "loss": 0.3398, | |
| "objective": 0.3393530547618866, | |
| "ranking_idealized": 0.5249999761581421, | |
| "ranking_idealized_expo": 0.5, | |
| "ranking_simple": 0.5166666507720947, | |
| "regularize": 0.3393530547618866, | |
| "step": 480 | |
| }, | |
| { | |
| "dpo_loss": 0.540759801864624, | |
| "epoch": 2.7491733585262166, | |
| "grad_norm": 103.16997454398687, | |
| "learning_rate": 4.980166765056193e-07, | |
| "logits": -1.254343032836914, | |
| "logps": -82.84033203125, | |
| "loss": 0.3085, | |
| "objective": 0.3383033871650696, | |
| "ranking_idealized": 0.5333333611488342, | |
| "ranking_idealized_expo": 0.5208333134651184, | |
| "ranking_simple": 0.5333333611488342, | |
| "regularize": 0.3383033871650696, | |
| "step": 485 | |
| }, | |
| { | |
| "dpo_loss": 0.5252385139465332, | |
| "epoch": 2.7775153519130846, | |
| "grad_norm": 106.99635988259067, | |
| "learning_rate": 4.881011512269463e-07, | |
| "logits": -1.227257251739502, | |
| "logps": -80.78081512451172, | |
| "loss": 0.3134, | |
| "objective": 0.3047190010547638, | |
| "ranking_idealized": 0.5041666626930237, | |
| "ranking_idealized_expo": 0.47083333134651184, | |
| "ranking_simple": 0.4958333373069763, | |
| "regularize": 0.3047190010547638, | |
| "step": 490 | |
| }, | |
| { | |
| "dpo_loss": 0.5381739139556885, | |
| "epoch": 2.8058573452999527, | |
| "grad_norm": 96.94469590251771, | |
| "learning_rate": 4.78190306317332e-07, | |
| "logits": -1.2355995178222656, | |
| "logps": -80.06063842773438, | |
| "loss": 0.3118, | |
| "objective": 0.31756675243377686, | |
| "ranking_idealized": 0.48750001192092896, | |
| "ranking_idealized_expo": 0.4791666567325592, | |
| "ranking_simple": 0.4958333373069763, | |
| "regularize": 0.31756675243377686, | |
| "step": 495 | |
| }, | |
| { | |
| "dpo_loss": 0.537841796875, | |
| "epoch": 2.8341993386868207, | |
| "grad_norm": 101.86509864093333, | |
| "learning_rate": 4.682880401717177e-07, | |
| "logits": -1.2800052165985107, | |
| "logps": -79.0132064819336, | |
| "loss": 0.2969, | |
| "objective": 0.2936791479587555, | |
| "ranking_idealized": 0.5333333611488342, | |
| "ranking_idealized_expo": 0.5208333134651184, | |
| "ranking_simple": 0.5416666865348816, | |
| "regularize": 0.2936791479587555, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.8341993386868207, | |
| "eval_dpo_loss": 0.7569719552993774, | |
| "eval_logits": -1.2819868326187134, | |
| "eval_logps": -87.59934997558594, | |
| "eval_loss": 0.7206214070320129, | |
| "eval_objective": 0.7300238609313965, | |
| "eval_ranking_idealized": 0.5247933864593506, | |
| "eval_ranking_idealized_expo": 0.5092975497245789, | |
| "eval_ranking_simple": 0.5134297609329224, | |
| "eval_regularize": 0.7300238609313965, | |
| "eval_runtime": 259.8973, | |
| "eval_samples_per_second": 22.278, | |
| "eval_steps_per_second": 0.931, | |
| "step": 500 | |
| }, | |
| { | |
| "dpo_loss": 0.5372523069381714, | |
| "epoch": 2.862541332073689, | |
| "grad_norm": 99.30099425250114, | |
| "learning_rate": 4.5839824781061886e-07, | |
| "logits": -1.239542841911316, | |
| "logps": -80.23788452148438, | |
| "loss": 0.2969, | |
| "objective": 0.2953982949256897, | |
| "ranking_idealized": 0.5458333492279053, | |
| "ranking_idealized_expo": 0.5083333253860474, | |
| "ranking_simple": 0.5458333492279053, | |
| "regularize": 0.2953982949256897, | |
| "step": 505 | |
| }, | |
| { | |
| "dpo_loss": 0.551425576210022, | |
| "epoch": 2.8908833254605573, | |
| "grad_norm": 97.19792911916865, | |
| "learning_rate": 4.4852481934803277e-07, | |
| "logits": -1.1939424276351929, | |
| "logps": -80.47624206542969, | |
| "loss": 0.2884, | |
| "objective": 0.2630988359451294, | |
| "ranking_idealized": 0.5416666865348816, | |
| "ranking_idealized_expo": 0.5249999761581421, | |
| "ranking_simple": 0.550000011920929, | |
| "regularize": 0.2630988359451294, | |
| "step": 510 | |
| }, | |
| { | |
| "dpo_loss": 0.553286612033844, | |
| "epoch": 2.9192253188474258, | |
| "grad_norm": 105.92002880687674, | |
| "learning_rate": 4.3867163846127674e-07, | |
| "logits": -1.3426212072372437, | |
| "logps": -80.19091796875, | |
| "loss": 0.2858, | |
| "objective": 0.2995811402797699, | |
| "ranking_idealized": 0.49166667461395264, | |
| "ranking_idealized_expo": 0.4791666567325592, | |
| "ranking_simple": 0.5, | |
| "regularize": 0.2995811104774475, | |
| "step": 515 | |
| }, | |
| { | |
| "dpo_loss": 0.5569177269935608, | |
| "epoch": 2.947567312234294, | |
| "grad_norm": 101.90420898944761, | |
| "learning_rate": 4.2884258086335745e-07, | |
| "logits": -1.2548803091049194, | |
| "logps": -82.30675506591797, | |
| "loss": 0.2885, | |
| "objective": 0.2949267625808716, | |
| "ranking_idealized": 0.5375000238418579, | |
| "ranking_idealized_expo": 0.5083333253860474, | |
| "ranking_simple": 0.5291666388511658, | |
| "regularize": 0.2949267625808716, | |
| "step": 520 | |
| }, | |
| { | |
| "dpo_loss": 0.5335093140602112, | |
| "epoch": 2.975909305621162, | |
| "grad_norm": 118.61355678596335, | |
| "learning_rate": 4.1904151277847305e-07, | |
| "logits": -1.2422376871109009, | |
| "logps": -81.52430725097656, | |
| "loss": 0.2939, | |
| "objective": 0.277982234954834, | |
| "ranking_idealized": 0.5791666507720947, | |
| "ranking_idealized_expo": 0.5666666626930237, | |
| "ranking_simple": 0.5874999761581421, | |
| "regularize": 0.277982234954834, | |
| "step": 525 | |
| }, | |
| { | |
| "dpo_loss": 0.5339943766593933, | |
| "epoch": 3.0042512990080303, | |
| "grad_norm": 95.3080580242307, | |
| "learning_rate": 4.092722894212487e-07, | |
| "logits": -1.245071530342102, | |
| "logps": -80.53169250488281, | |
| "loss": 0.2807, | |
| "objective": 0.28686758875846863, | |
| "ranking_idealized": 0.5333333611488342, | |
| "ranking_idealized_expo": 0.512499988079071, | |
| "ranking_simple": 0.5249999761581421, | |
| "regularize": 0.28686758875846863, | |
| "step": 530 | |
| }, | |
| { | |
| "dpo_loss": 0.5502054691314697, | |
| "epoch": 3.0325932923948984, | |
| "grad_norm": 102.03710421357799, | |
| "learning_rate": 3.995387534803005e-07, | |
| "logits": -1.2770276069641113, | |
| "logps": -82.10601043701172, | |
| "loss": 0.2686, | |
| "objective": 0.2612551748752594, | |
| "ranking_idealized": 0.5541666746139526, | |
| "ranking_idealized_expo": 0.5416666865348816, | |
| "ranking_simple": 0.5541666746139526, | |
| "regularize": 0.2612551748752594, | |
| "step": 535 | |
| }, | |
| { | |
| "dpo_loss": 0.5417371392250061, | |
| "epoch": 3.0609352857817664, | |
| "grad_norm": 93.88545782890976, | |
| "learning_rate": 3.8984473360672967e-07, | |
| "logits": -1.3437141180038452, | |
| "logps": -80.88935089111328, | |
| "loss": 0.2546, | |
| "objective": 0.2543182373046875, | |
| "ranking_idealized": 0.5583333373069763, | |
| "ranking_idealized_expo": 0.5375000238418579, | |
| "ranking_simple": 0.5416666865348816, | |
| "regularize": 0.2543182373046875, | |
| "step": 540 | |
| }, | |
| { | |
| "dpo_loss": 0.5468733906745911, | |
| "epoch": 3.089277279168635, | |
| "grad_norm": 104.98103692622652, | |
| "learning_rate": 3.801940429081345e-07, | |
| "logits": -1.29421067237854, | |
| "logps": -81.87580108642578, | |
| "loss": 0.2483, | |
| "objective": 0.23621630668640137, | |
| "ranking_idealized": 0.5, | |
| "ranking_idealized_expo": 0.49166667461395264, | |
| "ranking_simple": 0.49166667461395264, | |
| "regularize": 0.23621630668640137, | |
| "step": 545 | |
| }, | |
| { | |
| "dpo_loss": 0.5142976641654968, | |
| "epoch": 3.117619272555503, | |
| "grad_norm": 98.44480374148442, | |
| "learning_rate": 3.7059047744873955e-07, | |
| "logits": -1.1407238245010376, | |
| "logps": -82.03406524658203, | |
| "loss": 0.2456, | |
| "objective": 0.23313947021961212, | |
| "ranking_idealized": 0.5583333373069763, | |
| "ranking_idealized_expo": 0.5375000238418579, | |
| "ranking_simple": 0.5666666626930237, | |
| "regularize": 0.23313947021961212, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 3.117619272555503, | |
| "eval_dpo_loss": 0.7502080202102661, | |
| "eval_logits": -1.2747371196746826, | |
| "eval_logps": -87.48565673828125, | |
| "eval_loss": 0.7081578969955444, | |
| "eval_objective": 0.7094937562942505, | |
| "eval_ranking_idealized": 0.5247933864593506, | |
| "eval_ranking_idealized_expo": 0.5092975497245789, | |
| "eval_ranking_simple": 0.5134297609329224, | |
| "eval_regularize": 0.7094937562942505, | |
| "eval_runtime": 259.3716, | |
| "eval_samples_per_second": 22.323, | |
| "eval_steps_per_second": 0.933, | |
| "step": 550 | |
| }, | |
| { | |
| "dpo_loss": 0.519844651222229, | |
| "epoch": 3.1459612659423715, | |
| "grad_norm": 110.59601884737206, | |
| "learning_rate": 3.6103781475622786e-07, | |
| "logits": -1.2129274606704712, | |
| "logps": -82.34542083740234, | |
| "loss": 0.2514, | |
| "objective": 0.24437649548053741, | |
| "ranking_idealized": 0.5583333373069763, | |
| "ranking_idealized_expo": 0.550000011920929, | |
| "ranking_simple": 0.5791666507720947, | |
| "regularize": 0.24437648057937622, | |
| "step": 555 | |
| }, | |
| { | |
| "dpo_loss": 0.5489280223846436, | |
| "epoch": 3.1743032593292395, | |
| "grad_norm": 96.14728671233563, | |
| "learning_rate": 3.5153981233586274e-07, | |
| "logits": -1.213472843170166, | |
| "logps": -79.25808715820312, | |
| "loss": 0.2372, | |
| "objective": 0.2192198634147644, | |
| "ranking_idealized": 0.5083333253860474, | |
| "ranking_idealized_expo": 0.5041666626930237, | |
| "ranking_simple": 0.5166666507720947, | |
| "regularize": 0.2192198634147644, | |
| "step": 560 | |
| }, | |
| { | |
| "dpo_loss": 0.537466287612915, | |
| "epoch": 3.2026452527161076, | |
| "grad_norm": 107.72259319859731, | |
| "learning_rate": 3.421002061924876e-07, | |
| "logits": -1.2699893712997437, | |
| "logps": -80.88977813720703, | |
| "loss": 0.2311, | |
| "objective": 0.22347129881381989, | |
| "ranking_idealized": 0.5625, | |
| "ranking_idealized_expo": 0.5416666865348816, | |
| "ranking_simple": 0.5625, | |
| "regularize": 0.22347129881381989, | |
| "step": 565 | |
| }, | |
| { | |
| "dpo_loss": 0.5202825665473938, | |
| "epoch": 3.230987246102976, | |
| "grad_norm": 96.26968527459513, | |
| "learning_rate": 3.327227093609824e-07, | |
| "logits": -1.1865532398223877, | |
| "logps": -80.67752075195312, | |
| "loss": 0.2252, | |
| "objective": 0.21609430015087128, | |
| "ranking_idealized": 0.4958333373069763, | |
| "ranking_idealized_expo": 0.4791666567325592, | |
| "ranking_simple": 0.5, | |
| "regularize": 0.21609430015087128, | |
| "step": 570 | |
| }, | |
| { | |
| "dpo_loss": 0.5354868769645691, | |
| "epoch": 3.259329239489844, | |
| "grad_norm": 107.37637317824311, | |
| "learning_rate": 3.234110104457536e-07, | |
| "logits": -1.2506965398788452, | |
| "logps": -81.32064056396484, | |
| "loss": 0.2293, | |
| "objective": 0.2312368005514145, | |
| "ranking_idealized": 0.5208333134651184, | |
| "ranking_idealized_expo": 0.5, | |
| "ranking_simple": 0.5375000238418579, | |
| "regularize": 0.2312368005514145, | |
| "step": 575 | |
| }, | |
| { | |
| "dpo_loss": 0.5328246355056763, | |
| "epoch": 3.287671232876712, | |
| "grad_norm": 100.4320056036311, | |
| "learning_rate": 3.141687721698363e-07, | |
| "logits": -1.2623833417892456, | |
| "logps": -82.67001342773438, | |
| "loss": 0.2288, | |
| "objective": 0.22114989161491394, | |
| "ranking_idealized": 0.5583333373069763, | |
| "ranking_idealized_expo": 0.5541666746139526, | |
| "ranking_simple": 0.574999988079071, | |
| "regularize": 0.22114989161491394, | |
| "step": 580 | |
| }, | |
| { | |
| "dpo_loss": 0.5769094228744507, | |
| "epoch": 3.3160132262635806, | |
| "grad_norm": 95.72812686493451, | |
| "learning_rate": 3.049996299341742e-07, | |
| "logits": -1.279525876045227, | |
| "logps": -83.05570220947266, | |
| "loss": 0.2225, | |
| "objective": 0.2485423982143402, | |
| "ranking_idealized": 0.550000011920929, | |
| "ranking_idealized_expo": 0.5291666388511658, | |
| "ranking_simple": 0.5458333492279053, | |
| "regularize": 0.2485423982143402, | |
| "step": 585 | |
| }, | |
| { | |
| "dpo_loss": 0.5602288842201233, | |
| "epoch": 3.3443552196504487, | |
| "grad_norm": 114.56713956727607, | |
| "learning_rate": 2.959071903876486e-07, | |
| "logits": -1.294264554977417, | |
| "logps": -83.3205795288086, | |
| "loss": 0.2291, | |
| "objective": 0.22891105711460114, | |
| "ranking_idealized": 0.5208333134651184, | |
| "ranking_idealized_expo": 0.48750001192092896, | |
| "ranking_simple": 0.5166666507720947, | |
| "regularize": 0.22891105711460114, | |
| "step": 590 | |
| }, | |
| { | |
| "dpo_loss": 0.5364168286323547, | |
| "epoch": 3.372697213037317, | |
| "grad_norm": 110.16666826508619, | |
| "learning_rate": 2.86895030008416e-07, | |
| "logits": -1.2113687992095947, | |
| "logps": -82.52010345458984, | |
| "loss": 0.2129, | |
| "objective": 0.22527165710926056, | |
| "ranking_idealized": 0.5375000238418579, | |
| "ranking_idealized_expo": 0.5249999761581421, | |
| "ranking_simple": 0.5458333492279053, | |
| "regularize": 0.22527165710926056, | |
| "step": 595 | |
| }, | |
| { | |
| "dpo_loss": 0.5123292803764343, | |
| "epoch": 3.4010392064241852, | |
| "grad_norm": 102.02441515454663, | |
| "learning_rate": 2.779666936971129e-07, | |
| "logits": -1.2210050821304321, | |
| "logps": -81.7782974243164, | |
| "loss": 0.2121, | |
| "objective": 0.22540892660617828, | |
| "ranking_idealized": 0.512499988079071, | |
| "ranking_idealized_expo": 0.5041666626930237, | |
| "ranking_simple": 0.512499988079071, | |
| "regularize": 0.22540892660617828, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 3.4010392064241852, | |
| "eval_dpo_loss": 0.7512553930282593, | |
| "eval_logits": -1.2610689401626587, | |
| "eval_logps": -87.82508087158203, | |
| "eval_loss": 0.7149726152420044, | |
| "eval_objective": 0.7195205688476562, | |
| "eval_ranking_idealized": 0.5247933864593506, | |
| "eval_ranking_idealized_expo": 0.5092975497245789, | |
| "eval_ranking_simple": 0.5123966932296753, | |
| "eval_regularize": 0.7195205688476562, | |
| "eval_runtime": 260.141, | |
| "eval_samples_per_second": 22.257, | |
| "eval_steps_per_second": 0.93, | |
| "step": 600 | |
| }, | |
| { | |
| "dpo_loss": 0.5458318591117859, | |
| "epoch": 3.4293811998110533, | |
| "grad_norm": 109.5352185195456, | |
| "learning_rate": 2.6912569338248315e-07, | |
| "logits": -1.2689701318740845, | |
| "logps": -83.29906463623047, | |
| "loss": 0.2091, | |
| "objective": 0.18265347182750702, | |
| "ranking_idealized": 0.5458333492279053, | |
| "ranking_idealized_expo": 0.5333333611488342, | |
| "ranking_simple": 0.550000011920929, | |
| "regularize": 0.18265347182750702, | |
| "step": 605 | |
| }, | |
| { | |
| "dpo_loss": 0.5613773465156555, | |
| "epoch": 3.4577231931979218, | |
| "grad_norm": 98.96988548944374, | |
| "learning_rate": 2.603755066399718e-07, | |
| "logits": -1.240617036819458, | |
| "logps": -81.78998565673828, | |
| "loss": 0.208, | |
| "objective": 0.20637358725070953, | |
| "ranking_idealized": 0.5291666388511658, | |
| "ranking_idealized_expo": 0.512499988079071, | |
| "ranking_simple": 0.5291666388511658, | |
| "regularize": 0.20637358725070953, | |
| "step": 610 | |
| }, | |
| { | |
| "dpo_loss": 0.5179942846298218, | |
| "epoch": 3.48606518658479, | |
| "grad_norm": 101.26514646147335, | |
| "learning_rate": 2.517195753238345e-07, | |
| "logits": -1.2508882284164429, | |
| "logps": -81.7972183227539, | |
| "loss": 0.1973, | |
| "objective": 0.19295792281627655, | |
| "ranking_idealized": 0.4749999940395355, | |
| "ranking_idealized_expo": 0.46666666865348816, | |
| "ranking_simple": 0.47083333134651184, | |
| "regularize": 0.19295792281627655, | |
| "step": 615 | |
| }, | |
| { | |
| "dpo_loss": 0.49570992588996887, | |
| "epoch": 3.514407179971658, | |
| "grad_norm": 94.4850227020022, | |
| "learning_rate": 2.4316130421329696e-07, | |
| "logits": -1.1392742395401, | |
| "logps": -81.18209838867188, | |
| "loss": 0.1959, | |
| "objective": 0.19464272260665894, | |
| "ranking_idealized": 0.512499988079071, | |
| "ranking_idealized_expo": 0.4958333373069763, | |
| "ranking_simple": 0.512499988079071, | |
| "regularize": 0.19464272260665894, | |
| "step": 620 | |
| }, | |
| { | |
| "dpo_loss": 0.5331357717514038, | |
| "epoch": 3.5427491733585263, | |
| "grad_norm": 102.01306480785293, | |
| "learning_rate": 2.3470405967329604e-07, | |
| "logits": -1.2014575004577637, | |
| "logps": -80.8895492553711, | |
| "loss": 0.189, | |
| "objective": 0.20246191322803497, | |
| "ranking_idealized": 0.5333333611488342, | |
| "ranking_idealized_expo": 0.5208333134651184, | |
| "ranking_simple": 0.5291666388511658, | |
| "regularize": 0.20246191322803497, | |
| "step": 625 | |
| }, | |
| { | |
| "dpo_loss": 0.5293785929679871, | |
| "epoch": 3.5710911667453944, | |
| "grad_norm": 104.25371252839783, | |
| "learning_rate": 2.2635116833033392e-07, | |
| "logits": -1.28461754322052, | |
| "logps": -80.65286254882812, | |
| "loss": 0.1914, | |
| "objective": 0.1838250458240509, | |
| "ranking_idealized": 0.5291666388511658, | |
| "ranking_idealized_expo": 0.512499988079071, | |
| "ranking_simple": 0.5333333611488342, | |
| "regularize": 0.1838250458240509, | |
| "step": 630 | |
| }, | |
| { | |
| "dpo_loss": 0.5447375178337097, | |
| "epoch": 3.5994331601322624, | |
| "grad_norm": 98.62302663391017, | |
| "learning_rate": 2.181059157639598e-07, | |
| "logits": -1.2204653024673462, | |
| "logps": -81.04381561279297, | |
| "loss": 0.1798, | |
| "objective": 0.1847100853919983, | |
| "ranking_idealized": 0.4958333373069763, | |
| "ranking_idealized_expo": 0.48750001192092896, | |
| "ranking_simple": 0.49166667461395264, | |
| "regularize": 0.1847100704908371, | |
| "step": 635 | |
| }, | |
| { | |
| "dpo_loss": 0.5025536417961121, | |
| "epoch": 3.627775153519131, | |
| "grad_norm": 98.21161327414355, | |
| "learning_rate": 2.0997154521440097e-07, | |
| "logits": -1.178464412689209, | |
| "logps": -81.83779907226562, | |
| "loss": 0.1761, | |
| "objective": 0.18075956404209137, | |
| "ranking_idealized": 0.4958333373069763, | |
| "ranking_idealized_expo": 0.4625000059604645, | |
| "ranking_simple": 0.5, | |
| "regularize": 0.18075956404209137, | |
| "step": 640 | |
| }, | |
| { | |
| "dpo_loss": 0.5238084197044373, | |
| "epoch": 3.656117146905999, | |
| "grad_norm": 96.02562335204865, | |
| "learning_rate": 2.0195125630684428e-07, | |
| "logits": -1.2101988792419434, | |
| "logps": -80.41759490966797, | |
| "loss": 0.1754, | |
| "objective": 0.17556659877300262, | |
| "ranking_idealized": 0.5333333611488342, | |
| "ranking_idealized_expo": 0.5249999761581421, | |
| "ranking_simple": 0.5458333492279053, | |
| "regularize": 0.17556659877300262, | |
| "step": 645 | |
| }, | |
| { | |
| "dpo_loss": 0.5436745882034302, | |
| "epoch": 3.6844591402928675, | |
| "grad_norm": 99.15450590850473, | |
| "learning_rate": 1.9404820379287672e-07, | |
| "logits": -1.2024494409561157, | |
| "logps": -81.63809967041016, | |
| "loss": 0.1721, | |
| "objective": 0.15629148483276367, | |
| "ranking_idealized": 0.5166666507720947, | |
| "ranking_idealized_expo": 0.5041666626930237, | |
| "ranking_simple": 0.512499988079071, | |
| "regularize": 0.15629148483276367, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 3.6844591402928675, | |
| "eval_dpo_loss": 0.7529332041740417, | |
| "eval_logits": -1.2667409181594849, | |
| "eval_logps": -87.55421447753906, | |
| "eval_loss": 0.7180734276771545, | |
| "eval_objective": 0.7210157513618469, | |
| "eval_ranking_idealized": 0.5247933864593506, | |
| "eval_ranking_idealized_expo": 0.5092975497245789, | |
| "eval_ranking_simple": 0.5123966932296753, | |
| "eval_regularize": 0.7210157513618469, | |
| "eval_runtime": 258.8884, | |
| "eval_samples_per_second": 22.365, | |
| "eval_steps_per_second": 0.935, | |
| "step": 650 | |
| }, | |
| { | |
| "dpo_loss": 0.5238640904426575, | |
| "epoch": 3.7128011336797355, | |
| "grad_norm": 97.25223896173355, | |
| "learning_rate": 1.8626549630957395e-07, | |
| "logits": -1.240193486213684, | |
| "logps": -80.53789520263672, | |
| "loss": 0.1708, | |
| "objective": 0.1773752123117447, | |
| "ranking_idealized": 0.5375000238418579, | |
| "ranking_idealized_expo": 0.5208333134651184, | |
| "ranking_simple": 0.5416666865348816, | |
| "regularize": 0.1773752123117447, | |
| "step": 655 | |
| }, | |
| { | |
| "dpo_loss": 0.5414367914199829, | |
| "epoch": 3.7411431270666036, | |
| "grad_norm": 98.36333988268215, | |
| "learning_rate": 1.7860619515673032e-07, | |
| "logits": -1.318145513534546, | |
| "logps": -81.55570983886719, | |
| "loss": 0.1704, | |
| "objective": 0.1702161431312561, | |
| "ranking_idealized": 0.5666666626930237, | |
| "ranking_idealized_expo": 0.5541666746139526, | |
| "ranking_simple": 0.5625, | |
| "regularize": 0.1702161431312561, | |
| "step": 660 | |
| }, | |
| { | |
| "dpo_loss": 0.5241764187812805, | |
| "epoch": 3.769485120453472, | |
| "grad_norm": 104.11429458509453, | |
| "learning_rate": 1.7107331309270684e-07, | |
| "logits": -1.2156672477722168, | |
| "logps": -81.38536071777344, | |
| "loss": 0.1559, | |
| "objective": 0.16640320420265198, | |
| "ranking_idealized": 0.5208333134651184, | |
| "ranking_idealized_expo": 0.5166666507720947, | |
| "ranking_simple": 0.5208333134651184, | |
| "regularize": 0.16640320420265198, | |
| "step": 665 | |
| }, | |
| { | |
| "dpo_loss": 0.5278924703598022, | |
| "epoch": 3.79782711384034, | |
| "grad_norm": 101.56797379943947, | |
| "learning_rate": 1.6366981314937372e-07, | |
| "logits": -1.321338415145874, | |
| "logps": -81.34001922607422, | |
| "loss": 0.1551, | |
| "objective": 0.14367233216762543, | |
| "ranking_idealized": 0.6041666865348816, | |
| "ranking_idealized_expo": 0.5916666388511658, | |
| "ranking_simple": 0.6041666865348816, | |
| "regularize": 0.14367233216762543, | |
| "step": 670 | |
| }, | |
| { | |
| "dpo_loss": 0.529498279094696, | |
| "epoch": 3.826169107227208, | |
| "grad_norm": 93.64293062728967, | |
| "learning_rate": 1.5639860746661338e-07, | |
| "logits": -1.2852925062179565, | |
| "logps": -80.37777709960938, | |
| "loss": 0.1531, | |
| "objective": 0.1431640088558197, | |
| "ranking_idealized": 0.5666666626930237, | |
| "ranking_idealized_expo": 0.5375000238418579, | |
| "ranking_simple": 0.574999988079071, | |
| "regularize": 0.1431640088558197, | |
| "step": 675 | |
| }, | |
| { | |
| "dpo_loss": 0.5504888892173767, | |
| "epoch": 3.8545111006140766, | |
| "grad_norm": 101.91053202750965, | |
| "learning_rate": 1.492625561468393e-07, | |
| "logits": -1.2134761810302734, | |
| "logps": -80.77507781982422, | |
| "loss": 0.1495, | |
| "objective": 0.15138006210327148, | |
| "ranking_idealized": 0.5083333253860474, | |
| "ranking_idealized_expo": 0.5083333253860474, | |
| "ranking_simple": 0.5166666507720947, | |
| "regularize": 0.15138006210327148, | |
| "step": 680 | |
| }, | |
| { | |
| "dpo_loss": 0.5266403555870056, | |
| "epoch": 3.8828530940009447, | |
| "grad_norm": 105.20220630755058, | |
| "learning_rate": 1.4226446612998671e-07, | |
| "logits": -1.2806271314620972, | |
| "logps": -80.39690399169922, | |
| "loss": 0.1478, | |
| "objective": 0.1402387022972107, | |
| "ranking_idealized": 0.5458333492279053, | |
| "ranking_idealized_expo": 0.5166666507720947, | |
| "ranking_simple": 0.5416666865348816, | |
| "regularize": 0.1402387022972107, | |
| "step": 685 | |
| }, | |
| { | |
| "dpo_loss": 0.5246676802635193, | |
| "epoch": 3.9111950873878127, | |
| "grad_norm": 93.05384336107711, | |
| "learning_rate": 1.3540709008941147e-07, | |
| "logits": -1.2790921926498413, | |
| "logps": -80.16386413574219, | |
| "loss": 0.1448, | |
| "objective": 0.1356712281703949, | |
| "ranking_idealized": 0.5625, | |
| "ranking_idealized_expo": 0.5458333492279053, | |
| "ranking_simple": 0.574999988079071, | |
| "regularize": 0.1356712281703949, | |
| "step": 690 | |
| }, | |
| { | |
| "dpo_loss": 0.5320358276367188, | |
| "epoch": 3.9395370807746812, | |
| "grad_norm": 92.75785969700081, | |
| "learning_rate": 1.2869312534913685e-07, | |
| "logits": -1.1912401914596558, | |
| "logps": -81.54454803466797, | |
| "loss": 0.1365, | |
| "objective": 0.15173907577991486, | |
| "ranking_idealized": 0.4791666567325592, | |
| "ranking_idealized_expo": 0.46666666865348816, | |
| "ranking_simple": 0.4791666567325592, | |
| "regularize": 0.15173907577991486, | |
| "step": 695 | |
| }, | |
| { | |
| "dpo_loss": 0.5335038304328918, | |
| "epoch": 3.9678790741615493, | |
| "grad_norm": 98.04635697056264, | |
| "learning_rate": 1.2212521282287093e-07, | |
| "logits": -1.2281270027160645, | |
| "logps": -81.2812728881836, | |
| "loss": 0.1386, | |
| "objective": 0.15212775766849518, | |
| "ranking_idealized": 0.49166667461395264, | |
| "ranking_idealized_expo": 0.4833333194255829, | |
| "ranking_simple": 0.49166667461395264, | |
| "regularize": 0.15212775766849518, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 3.9678790741615493, | |
| "eval_dpo_loss": 0.7514122128486633, | |
| "eval_logits": -1.26543128490448, | |
| "eval_logps": -87.5438003540039, | |
| "eval_loss": 0.7064864635467529, | |
| "eval_objective": 0.7093826532363892, | |
| "eval_ranking_idealized": 0.5247933864593506, | |
| "eval_ranking_idealized_expo": 0.5092975497245789, | |
| "eval_ranking_simple": 0.5113636255264282, | |
| "eval_regularize": 0.7093826532363892, | |
| "eval_runtime": 259.4271, | |
| "eval_samples_per_second": 22.318, | |
| "eval_steps_per_second": 0.933, | |
| "step": 700 | |
| }, | |
| { | |
| "dpo_loss": 0.5151931047439575, | |
| "epoch": 3.9962210675484178, | |
| "grad_norm": 98.98779841672567, | |
| "learning_rate": 1.15705935975212e-07, | |
| "logits": -1.1545416116714478, | |
| "logps": -79.99464416503906, | |
| "loss": 0.133, | |
| "objective": 0.12396544963121414, | |
| "ranking_idealized": 0.5583333373069763, | |
| "ranking_idealized_expo": 0.5291666388511658, | |
| "ranking_simple": 0.5541666746139526, | |
| "regularize": 0.12396544963121414, | |
| "step": 705 | |
| }, | |
| { | |
| "dpo_loss": 0.5161293745040894, | |
| "epoch": 4.024563060935286, | |
| "grad_norm": 94.60148232757011, | |
| "learning_rate": 1.094378198054533e-07, | |
| "logits": -1.2483078241348267, | |
| "logps": -81.60172271728516, | |
| "loss": 0.1163, | |
| "objective": 0.09800054132938385, | |
| "ranking_idealized": 0.48750001192092896, | |
| "ranking_idealized_expo": 0.4791666567325592, | |
| "ranking_simple": 0.49166667461395264, | |
| "regularize": 0.09800054132938385, | |
| "step": 710 | |
| }, | |
| { | |
| "dpo_loss": 0.5236179232597351, | |
| "epoch": 4.052905054322154, | |
| "grad_norm": 95.22257327163594, | |
| "learning_rate": 1.0332332985438247e-07, | |
| "logits": -1.2045139074325562, | |
| "logps": -81.0995864868164, | |
| "loss": 0.1109, | |
| "objective": 0.10611793398857117, | |
| "ranking_idealized": 0.5458333492279053, | |
| "ranking_idealized_expo": 0.5333333611488342, | |
| "ranking_simple": 0.5458333492279053, | |
| "regularize": 0.10611793398857117, | |
| "step": 715 | |
| }, | |
| { | |
| "dpo_loss": 0.5168012976646423, | |
| "epoch": 4.081247047709022, | |
| "grad_norm": 101.20095806117061, | |
| "learning_rate": 9.736487123447068e-08, | |
| "logits": -1.1531497240066528, | |
| "logps": -83.11766052246094, | |
| "loss": 0.1094, | |
| "objective": 0.11501701176166534, | |
| "ranking_idealized": 0.5333333611488342, | |
| "ranking_idealized_expo": 0.5208333134651184, | |
| "ranking_simple": 0.5291666388511658, | |
| "regularize": 0.11501700431108475, | |
| "step": 720 | |
| }, | |
| { | |
| "dpo_loss": 0.5381873250007629, | |
| "epoch": 4.109589041095891, | |
| "grad_norm": 95.89781184589388, | |
| "learning_rate": 9.156478768383058e-08, | |
| "logits": -1.226915955543518, | |
| "logps": -81.01741790771484, | |
| "loss": 0.1131, | |
| "objective": 0.10321195423603058, | |
| "ranking_idealized": 0.5375000238418579, | |
| "ranking_idealized_expo": 0.5291666388511658, | |
| "ranking_simple": 0.5375000238418579, | |
| "regularize": 0.10321194678544998, | |
| "step": 725 | |
| }, | |
| { | |
| "dpo_loss": 0.5230408310890198, | |
| "epoch": 4.137931034482759, | |
| "grad_norm": 100.59516155263596, | |
| "learning_rate": 8.592536064431466e-08, | |
| "logits": -1.2960838079452515, | |
| "logps": -81.26818084716797, | |
| "loss": 0.1063, | |
| "objective": 0.09859620779752731, | |
| "ranking_idealized": 0.550000011920929, | |
| "ranking_idealized_expo": 0.5375000238418579, | |
| "ranking_simple": 0.5541666746139526, | |
| "regularize": 0.09859620034694672, | |
| "step": 730 | |
| }, | |
| { | |
| "dpo_loss": 0.5300396084785461, | |
| "epoch": 4.166273027869627, | |
| "grad_norm": 99.36395143789497, | |
| "learning_rate": 8.044880836411888e-08, | |
| "logits": -1.1851133108139038, | |
| "logps": -82.21375274658203, | |
| "loss": 0.1031, | |
| "objective": 0.10836609452962875, | |
| "ranking_idealized": 0.5541666746139526, | |
| "ranking_idealized_expo": 0.5249999761581421, | |
| "ranking_simple": 0.5458333492279053, | |
| "regularize": 0.10836609452962875, | |
| "step": 735 | |
| }, | |
| { | |
| "dpo_loss": 0.5081815719604492, | |
| "epoch": 4.194615021256495, | |
| "grad_norm": 97.85835254690582, | |
| "learning_rate": 7.513728502524286e-08, | |
| "logits": -1.214992880821228, | |
| "logps": -81.85246276855469, | |
| "loss": 0.1108, | |
| "objective": 0.09990967810153961, | |
| "ranking_idealized": 0.5375000238418579, | |
| "ranking_idealized_expo": 0.5208333134651184, | |
| "ranking_simple": 0.5375000238418579, | |
| "regularize": 0.09990967065095901, | |
| "step": 740 | |
| }, | |
| { | |
| "dpo_loss": 0.5218914747238159, | |
| "epoch": 4.222957014643363, | |
| "grad_norm": 102.73912212411751, | |
| "learning_rate": 6.999287989614971e-08, | |
| "logits": -1.2808277606964111, | |
| "logps": -79.33137512207031, | |
| "loss": 0.1003, | |
| "objective": 0.0994986817240715, | |
| "ranking_idealized": 0.5083333253860474, | |
| "ranking_idealized_expo": 0.48750001192092896, | |
| "ranking_simple": 0.5041666626930237, | |
| "regularize": 0.0994986817240715, | |
| "step": 745 | |
| }, | |
| { | |
| "dpo_loss": 0.5142425298690796, | |
| "epoch": 4.251299008030231, | |
| "grad_norm": 98.36808351404612, | |
| "learning_rate": 6.501761650996052e-08, | |
| "logits": -1.3703263998031616, | |
| "logps": -81.33523559570312, | |
| "loss": 0.0985, | |
| "objective": 0.09714233130216599, | |
| "ranking_idealized": 0.5416666865348816, | |
| "ranking_idealized_expo": 0.5249999761581421, | |
| "ranking_simple": 0.5416666865348816, | |
| "regularize": 0.09714233130216599, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 4.251299008030231, | |
| "eval_dpo_loss": 0.7508513331413269, | |
| "eval_logits": -1.2698993682861328, | |
| "eval_logps": -87.64305877685547, | |
| "eval_loss": 0.7095766663551331, | |
| "eval_objective": 0.7117514610290527, | |
| "eval_ranking_idealized": 0.5247933864593506, | |
| "eval_ranking_idealized_expo": 0.5092975497245789, | |
| "eval_ranking_simple": 0.5144628286361694, | |
| "eval_regularize": 0.7117514610290527, | |
| "eval_runtime": 259.3031, | |
| "eval_samples_per_second": 22.329, | |
| "eval_steps_per_second": 0.933, | |
| "step": 750 | |
| }, | |
| { | |
| "dpo_loss": 0.5111901760101318, | |
| "epoch": 4.2796410014171, | |
| "grad_norm": 95.47883788378027, | |
| "learning_rate": 6.021345186850418e-08, | |
| "logits": -1.2960675954818726, | |
| "logps": -81.29663848876953, | |
| "loss": 0.0959, | |
| "objective": 0.0885348990559578, | |
| "ranking_idealized": 0.5458333492279053, | |
| "ranking_idealized_expo": 0.5333333611488342, | |
| "ranking_simple": 0.5458333492279053, | |
| "regularize": 0.0885348990559578, | |
| "step": 755 | |
| }, | |
| { | |
| "dpo_loss": 0.5198584794998169, | |
| "epoch": 4.307982994803968, | |
| "grad_norm": 107.35986940095462, | |
| "learning_rate": 5.5582275672538316e-08, | |
| "logits": -1.2736433744430542, | |
| "logps": -81.2705078125, | |
| "loss": 0.0936, | |
| "objective": 0.08593542128801346, | |
| "ranking_idealized": 0.5791666507720947, | |
| "ranking_idealized_expo": 0.5625, | |
| "ranking_simple": 0.5791666507720947, | |
| "regularize": 0.08593542128801346, | |
| "step": 760 | |
| }, | |
| { | |
| "dpo_loss": 0.508173406124115, | |
| "epoch": 4.336324988190836, | |
| "grad_norm": 99.70173708140004, | |
| "learning_rate": 5.112590957844232e-08, | |
| "logits": -1.2959843873977661, | |
| "logps": -82.02350616455078, | |
| "loss": 0.0954, | |
| "objective": 0.09023122489452362, | |
| "ranking_idealized": 0.5458333492279053, | |
| "ranking_idealized_expo": 0.5249999761581421, | |
| "ranking_simple": 0.5458333492279053, | |
| "regularize": 0.09023122489452362, | |
| "step": 765 | |
| }, | |
| { | |
| "dpo_loss": 0.5115682482719421, | |
| "epoch": 4.364666981577704, | |
| "grad_norm": 94.66084425031231, | |
| "learning_rate": 4.684610648167503e-08, | |
| "logits": -1.1798807382583618, | |
| "logps": -80.28400421142578, | |
| "loss": 0.1006, | |
| "objective": 0.10149048268795013, | |
| "ranking_idealized": 0.5416666865348816, | |
| "ranking_idealized_expo": 0.5041666626930237, | |
| "ranking_simple": 0.5416666865348816, | |
| "regularize": 0.10149047523736954, | |
| "step": 770 | |
| }, | |
| { | |
| "dpo_loss": 0.5190815329551697, | |
| "epoch": 4.393008974964572, | |
| "grad_norm": 97.786866794993, | |
| "learning_rate": 4.274454982728032e-08, | |
| "logits": -1.2468054294586182, | |
| "logps": -81.20288848876953, | |
| "loss": 0.0877, | |
| "objective": 0.08333703130483627, | |
| "ranking_idealized": 0.5333333611488342, | |
| "ranking_idealized_expo": 0.5083333253860474, | |
| "ranking_simple": 0.5249999761581421, | |
| "regularize": 0.08333703130483627, | |
| "step": 775 | |
| }, | |
| { | |
| "dpo_loss": 0.5364963412284851, | |
| "epoch": 4.42135096835144, | |
| "grad_norm": 93.67850493598061, | |
| "learning_rate": 3.882285294770937e-08, | |
| "logits": -1.2755876779556274, | |
| "logps": -80.01787567138672, | |
| "loss": 0.0938, | |
| "objective": 0.09289808571338654, | |
| "ranking_idealized": 0.5083333253860474, | |
| "ranking_idealized_expo": 0.5, | |
| "ranking_simple": 0.5, | |
| "regularize": 0.09289808571338654, | |
| "step": 780 | |
| }, | |
| { | |
| "dpo_loss": 0.5412746667861938, | |
| "epoch": 4.449692961738309, | |
| "grad_norm": 94.9731716353141, | |
| "learning_rate": 3.508255842822255e-08, | |
| "logits": -1.292734980583191, | |
| "logps": -80.0306625366211, | |
| "loss": 0.0913, | |
| "objective": 0.09275906533002853, | |
| "ranking_idealized": 0.5458333492279053, | |
| "ranking_idealized_expo": 0.5291666388511658, | |
| "ranking_simple": 0.5416666865348816, | |
| "regularize": 0.09275905787944794, | |
| "step": 785 | |
| }, | |
| { | |
| "dpo_loss": 0.5126563906669617, | |
| "epoch": 4.478034955125177, | |
| "grad_norm": 96.74045611069916, | |
| "learning_rate": 3.15251375001192e-08, | |
| "logits": -1.253103494644165, | |
| "logps": -82.71080017089844, | |
| "loss": 0.0897, | |
| "objective": 0.08310778439044952, | |
| "ranking_idealized": 0.5833333134651184, | |
| "ranking_idealized_expo": 0.5583333373069763, | |
| "ranking_simple": 0.5791666507720947, | |
| "regularize": 0.08310777693986893, | |
| "step": 790 | |
| }, | |
| { | |
| "dpo_loss": 0.5277770161628723, | |
| "epoch": 4.506376948512045, | |
| "grad_norm": 102.60968536429716, | |
| "learning_rate": 2.8151989462033787e-08, | |
| "logits": -1.1557574272155762, | |
| "logps": -82.18769073486328, | |
| "loss": 0.0898, | |
| "objective": 0.08717220276594162, | |
| "ranking_idealized": 0.5583333373069763, | |
| "ranking_idealized_expo": 0.550000011920929, | |
| "ranking_simple": 0.5583333373069763, | |
| "regularize": 0.08717220276594162, | |
| "step": 795 | |
| }, | |
| { | |
| "dpo_loss": 0.5270230770111084, | |
| "epoch": 4.534718941898913, | |
| "grad_norm": 96.30549999556874, | |
| "learning_rate": 2.4964441129527335e-08, | |
| "logits": -1.1533504724502563, | |
| "logps": -81.5091323852539, | |
| "loss": 0.0882, | |
| "objective": 0.09708014130592346, | |
| "ranking_idealized": 0.5375000238418579, | |
| "ranking_idealized_expo": 0.5083333253860474, | |
| "ranking_simple": 0.5333333611488342, | |
| "regularize": 0.09708014130592346, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 4.534718941898913, | |
| "eval_dpo_loss": 0.7520135641098022, | |
| "eval_logits": -1.2693291902542114, | |
| "eval_logps": -87.7427749633789, | |
| "eval_loss": 0.7119117975234985, | |
| "eval_objective": 0.7144764065742493, | |
| "eval_ranking_idealized": 0.5247933864593506, | |
| "eval_ranking_idealized_expo": 0.5092975497245789, | |
| "eval_ranking_simple": 0.5113636255264282, | |
| "eval_regularize": 0.7144764065742493, | |
| "eval_runtime": 258.9953, | |
| "eval_samples_per_second": 22.356, | |
| "eval_steps_per_second": 0.934, | |
| "step": 800 | |
| }, | |
| { | |
| "dpo_loss": 0.5197975039482117, | |
| "epoch": 4.563060935285781, | |
| "grad_norm": 94.79722521361904, | |
| "learning_rate": 2.1963746313188757e-08, | |
| "logits": -1.226369023323059, | |
| "logps": -80.86296844482422, | |
| "loss": 0.0813, | |
| "objective": 0.07861018925905228, | |
| "ranking_idealized": 0.4833333194255829, | |
| "ranking_idealized_expo": 0.4625000059604645, | |
| "ranking_simple": 0.48750001192092896, | |
| "regularize": 0.07861018925905228, | |
| "step": 805 | |
| }, | |
| { | |
| "dpo_loss": 0.5268690586090088, | |
| "epoch": 4.59140292867265, | |
| "grad_norm": 95.24668751727883, | |
| "learning_rate": 1.915108532545351e-08, | |
| "logits": -1.290175199508667, | |
| "logps": -80.24264526367188, | |
| "loss": 0.0875, | |
| "objective": 0.08437123894691467, | |
| "ranking_idealized": 0.46666666865348816, | |
| "ranking_idealized_expo": 0.4583333432674408, | |
| "ranking_simple": 0.46666666865348816, | |
| "regularize": 0.08437123149633408, | |
| "step": 810 | |
| }, | |
| { | |
| "dpo_loss": 0.5314128398895264, | |
| "epoch": 4.619744922059518, | |
| "grad_norm": 94.93833650674884, | |
| "learning_rate": 1.6527564516331638e-08, | |
| "logits": -1.1623790264129639, | |
| "logps": -81.81945037841797, | |
| "loss": 0.0826, | |
| "objective": 0.09123408049345016, | |
| "ranking_idealized": 0.6166666746139526, | |
| "ranking_idealized_expo": 0.5874999761581421, | |
| "ranking_simple": 0.6166666746139526, | |
| "regularize": 0.09123408049345016, | |
| "step": 815 | |
| }, | |
| { | |
| "dpo_loss": 0.5448810458183289, | |
| "epoch": 4.648086915446386, | |
| "grad_norm": 99.02648077337858, | |
| "learning_rate": 1.4094215838229172e-08, | |
| "logits": -1.2590090036392212, | |
| "logps": -81.42196655273438, | |
| "loss": 0.0864, | |
| "objective": 0.09086828678846359, | |
| "ranking_idealized": 0.5625, | |
| "ranking_idealized_expo": 0.5625, | |
| "ranking_simple": 0.5583333373069763, | |
| "regularize": 0.09086828678846359, | |
| "step": 820 | |
| }, | |
| { | |
| "dpo_loss": 0.5343818664550781, | |
| "epoch": 4.6764289088332545, | |
| "grad_norm": 93.21086746148787, | |
| "learning_rate": 1.1851996440033318e-08, | |
| "logits": -1.1644405126571655, | |
| "logps": -79.48759460449219, | |
| "loss": 0.0796, | |
| "objective": 0.07582148164510727, | |
| "ranking_idealized": 0.5083333253860474, | |
| "ranking_idealized_expo": 0.5, | |
| "ranking_simple": 0.5083333253860474, | |
| "regularize": 0.07582148164510727, | |
| "step": 825 | |
| }, | |
| { | |
| "dpo_loss": 0.5317553281784058, | |
| "epoch": 4.7047709022201225, | |
| "grad_norm": 97.38269826526711, | |
| "learning_rate": 9.801788290621505e-09, | |
| "logits": -1.3180792331695557, | |
| "logps": -81.53900909423828, | |
| "loss": 0.0901, | |
| "objective": 0.10412049293518066, | |
| "ranking_idealized": 0.5375000238418579, | |
| "ranking_idealized_expo": 0.512499988079071, | |
| "ranking_simple": 0.5375000238418579, | |
| "regularize": 0.10412048548460007, | |
| "step": 830 | |
| }, | |
| { | |
| "dpo_loss": 0.5185703635215759, | |
| "epoch": 4.733112895606991, | |
| "grad_norm": 92.96268594737676, | |
| "learning_rate": 7.944397831941951e-09, | |
| "logits": -1.2221894264221191, | |
| "logps": -81.21542358398438, | |
| "loss": 0.0893, | |
| "objective": 0.0936354324221611, | |
| "ranking_idealized": 0.512499988079071, | |
| "ranking_idealized_expo": 0.48750001192092896, | |
| "ranking_simple": 0.5083333253860474, | |
| "regularize": 0.09363541752099991, | |
| "step": 835 | |
| }, | |
| { | |
| "dpo_loss": 0.5039588212966919, | |
| "epoch": 4.7614548889938595, | |
| "grad_norm": 93.6593830751853, | |
| "learning_rate": 6.280555661802856e-09, | |
| "logits": -1.2342692613601685, | |
| "logps": -81.57914733886719, | |
| "loss": 0.0827, | |
| "objective": 0.09424074739217758, | |
| "ranking_idealized": 0.5583333373069763, | |
| "ranking_idealized_expo": 0.5541666746139526, | |
| "ranking_simple": 0.5583333373069763, | |
| "regularize": 0.09424074739217758, | |
| "step": 840 | |
| }, | |
| { | |
| "dpo_loss": 0.5341091752052307, | |
| "epoch": 4.7897968823807275, | |
| "grad_norm": 94.7722019315539, | |
| "learning_rate": 4.810916246494157e-09, | |
| "logits": -1.2643479108810425, | |
| "logps": -81.62488555908203, | |
| "loss": 0.0815, | |
| "objective": 0.07745879143476486, | |
| "ranking_idealized": 0.4791666567325592, | |
| "ranking_idealized_expo": 0.4541666805744171, | |
| "ranking_simple": 0.47083333134651184, | |
| "regularize": 0.07745879143476486, | |
| "step": 845 | |
| }, | |
| { | |
| "dpo_loss": 0.5221010446548462, | |
| "epoch": 4.818138875767596, | |
| "grad_norm": 94.27228247906854, | |
| "learning_rate": 3.5360576633558513e-09, | |
| "logits": -1.2364792823791504, | |
| "logps": -80.28359985351562, | |
| "loss": 0.0796, | |
| "objective": 0.0774383395910263, | |
| "ranking_idealized": 0.5333333611488342, | |
| "ranking_idealized_expo": 0.5291666388511658, | |
| "ranking_simple": 0.5291666388511658, | |
| "regularize": 0.07743828743696213, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 4.818138875767596, | |
| "eval_dpo_loss": 0.7514771223068237, | |
| "eval_logits": -1.2688941955566406, | |
| "eval_logps": -87.71552276611328, | |
| "eval_loss": 0.7094753384590149, | |
| "eval_objective": 0.711796224117279, | |
| "eval_ranking_idealized": 0.5247933864593506, | |
| "eval_ranking_idealized_expo": 0.5092975497245789, | |
| "eval_ranking_simple": 0.5123966932296753, | |
| "eval_regularize": 0.711796224117279, | |
| "eval_runtime": 259.5779, | |
| "eval_samples_per_second": 22.305, | |
| "eval_steps_per_second": 0.932, | |
| "step": 850 | |
| }, | |
| { | |
| "dpo_loss": 0.5401983857154846, | |
| "epoch": 4.846480869154464, | |
| "grad_norm": 92.34602743429262, | |
| "learning_rate": 2.4564813733932155e-09, | |
| "logits": -1.205161690711975, | |
| "logps": -80.62550354003906, | |
| "loss": 0.08, | |
| "objective": 0.08250019699335098, | |
| "ranking_idealized": 0.5583333373069763, | |
| "ranking_idealized_expo": 0.5375000238418579, | |
| "ranking_simple": 0.5625, | |
| "regularize": 0.08250019699335098, | |
| "step": 855 | |
| }, | |
| { | |
| "dpo_loss": 0.5175269246101379, | |
| "epoch": 4.874822862541333, | |
| "grad_norm": 99.70852124493375, | |
| "learning_rate": 1.5726120240288631e-09, | |
| "logits": -1.1921393871307373, | |
| "logps": -80.51878356933594, | |
| "loss": 0.0814, | |
| "objective": 0.07799576222896576, | |
| "ranking_idealized": 0.5666666626930237, | |
| "ranking_idealized_expo": 0.550000011920929, | |
| "ranking_simple": 0.5666666626930237, | |
| "regularize": 0.07799576222896576, | |
| "step": 860 | |
| }, | |
| { | |
| "dpo_loss": 0.5134026408195496, | |
| "epoch": 4.903164855928201, | |
| "grad_norm": 93.71216642807723, | |
| "learning_rate": 8.847972820693051e-10, | |
| "logits": -1.2603986263275146, | |
| "logps": -79.67554473876953, | |
| "loss": 0.0745, | |
| "objective": 0.07266159355640411, | |
| "ranking_idealized": 0.5083333253860474, | |
| "ranking_idealized_expo": 0.4958333373069763, | |
| "ranking_simple": 0.5166666507720947, | |
| "regularize": 0.07266159355640411, | |
| "step": 865 | |
| }, | |
| { | |
| "dpo_loss": 0.5160818099975586, | |
| "epoch": 4.931506849315069, | |
| "grad_norm": 95.34876322242407, | |
| "learning_rate": 3.933076969516724e-10, | |
| "logits": -1.29552161693573, | |
| "logps": -81.27151489257812, | |
| "loss": 0.0794, | |
| "objective": 0.08014672249555588, | |
| "ranking_idealized": 0.5375000238418579, | |
| "ranking_idealized_expo": 0.5083333253860474, | |
| "ranking_simple": 0.5333333611488342, | |
| "regularize": 0.08014672249555588, | |
| "step": 870 | |
| }, | |
| { | |
| "dpo_loss": 0.522140622138977, | |
| "epoch": 4.959848842701937, | |
| "grad_norm": 95.71079860788076, | |
| "learning_rate": 9.833659432367803e-11, | |
| "logits": -1.2183899879455566, | |
| "logps": -81.16183471679688, | |
| "loss": 0.081, | |
| "objective": 0.08778008073568344, | |
| "ranking_idealized": 0.574999988079071, | |
| "ranking_idealized_expo": 0.5458333492279053, | |
| "ranking_simple": 0.574999988079071, | |
| "regularize": 0.08778008073568344, | |
| "step": 875 | |
| }, | |
| { | |
| "dpo_loss": 0.5225833058357239, | |
| "epoch": 4.988190836088805, | |
| "grad_norm": 104.51115430406888, | |
| "learning_rate": 0.0, | |
| "logits": -1.2887017726898193, | |
| "logps": -81.99433898925781, | |
| "loss": 0.0809, | |
| "objective": 0.0927945151925087, | |
| "ranking_idealized": 0.5375000238418579, | |
| "ranking_idealized_expo": 0.5291666388511658, | |
| "ranking_simple": 0.5375000238418579, | |
| "regularize": 0.0927945151925087, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 4.988190836088805, | |
| "step": 880, | |
| "total_flos": 0.0, | |
| "train_loss": 0.3223599465055899, | |
| "train_runtime": 35128.3246, | |
| "train_samples_per_second": 7.231, | |
| "train_steps_per_second": 0.025 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 880, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "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": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |