| { | |
| "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": 134135.90281322593, | |
| "learning_rate": 5.681818181818182e-08, | |
| "logits": -1.3147305250167847, | |
| "logps": -88.0877456665039, | |
| "loss": 0.4113, | |
| "objective": 0.41588976979255676, | |
| "ranking_idealized": 0.5208333134651184, | |
| "ranking_idealized_expo": 0.5208333134651184, | |
| "ranking_simple": 0.5208333134651184, | |
| "regularize": 0.41588976979255676, | |
| "step": 1 | |
| }, | |
| { | |
| "dpo_loss": 13.11236572265625, | |
| "epoch": 0.02834199338686821, | |
| "grad_norm": 235831.3417258827, | |
| "learning_rate": 2.840909090909091e-07, | |
| "logits": -1.3680161237716675, | |
| "logps": -84.41597747802734, | |
| "loss": 26.5575, | |
| "objective": 27.22771453857422, | |
| "ranking_idealized": 0.546875, | |
| "ranking_idealized_expo": 0.546875, | |
| "ranking_simple": 0.546875, | |
| "regularize": 27.22771453857422, | |
| "step": 5 | |
| }, | |
| { | |
| "dpo_loss": 32.3316535949707, | |
| "epoch": 0.05668398677373642, | |
| "grad_norm": 273612.5381400241, | |
| "learning_rate": 5.681818181818182e-07, | |
| "logits": -1.44873046875, | |
| "logps": -83.52092742919922, | |
| "loss": 73.6592, | |
| "objective": 67.45610809326172, | |
| "ranking_idealized": 0.5083333253860474, | |
| "ranking_idealized_expo": 0.5083333253860474, | |
| "ranking_simple": 0.5083333253860474, | |
| "regularize": 67.45610809326172, | |
| "step": 10 | |
| }, | |
| { | |
| "dpo_loss": 62.25831604003906, | |
| "epoch": 0.08502598016060463, | |
| "grad_norm": 199991.8454482197, | |
| "learning_rate": 8.522727272727273e-07, | |
| "logits": -1.4123358726501465, | |
| "logps": -83.7873306274414, | |
| "loss": 136.6551, | |
| "objective": 129.5699005126953, | |
| "ranking_idealized": 0.5041666626930237, | |
| "ranking_idealized_expo": 0.5041666626930237, | |
| "ranking_simple": 0.5083333253860474, | |
| "regularize": 129.5699005126953, | |
| "step": 15 | |
| }, | |
| { | |
| "dpo_loss": 97.5913314819336, | |
| "epoch": 0.11336797354747284, | |
| "grad_norm": 137050.47037487474, | |
| "learning_rate": 1.1363636363636364e-06, | |
| "logits": -1.39812433719635, | |
| "logps": -84.45435333251953, | |
| "loss": 198.982, | |
| "objective": 196.08615112304688, | |
| "ranking_idealized": 0.5375000238418579, | |
| "ranking_idealized_expo": 0.5375000238418579, | |
| "ranking_simple": 0.5375000238418579, | |
| "regularize": 196.08615112304688, | |
| "step": 20 | |
| }, | |
| { | |
| "dpo_loss": 134.1880340576172, | |
| "epoch": 0.14170996693434104, | |
| "grad_norm": 128505.67292666303, | |
| "learning_rate": 1.4204545454545458e-06, | |
| "logits": -1.4363852739334106, | |
| "logps": -84.79105377197266, | |
| "loss": 276.7151, | |
| "objective": 282.50897216796875, | |
| "ranking_idealized": 0.4958333373069763, | |
| "ranking_idealized_expo": 0.4958333373069763, | |
| "ranking_simple": 0.5041666626930237, | |
| "regularize": 282.50897216796875, | |
| "step": 25 | |
| }, | |
| { | |
| "dpo_loss": 166.0544891357422, | |
| "epoch": 0.17005196032120926, | |
| "grad_norm": 181462.71968306202, | |
| "learning_rate": 1.7045454545454546e-06, | |
| "logits": -1.4044525623321533, | |
| "logps": -85.5672378540039, | |
| "loss": 371.4895, | |
| "objective": 362.3011169433594, | |
| "ranking_idealized": 0.5416666865348816, | |
| "ranking_idealized_expo": 0.5416666865348816, | |
| "ranking_simple": 0.550000011920929, | |
| "regularize": 362.3011169433594, | |
| "step": 30 | |
| }, | |
| { | |
| "dpo_loss": 249.85475158691406, | |
| "epoch": 0.19839395370807747, | |
| "grad_norm": 131607.94522082573, | |
| "learning_rate": 1.9886363636363638e-06, | |
| "logits": -1.3713446855545044, | |
| "logps": -82.86891174316406, | |
| "loss": 459.5799, | |
| "objective": 475.0990905761719, | |
| "ranking_idealized": 0.5041666626930237, | |
| "ranking_idealized_expo": 0.5041666626930237, | |
| "ranking_simple": 0.5, | |
| "regularize": 475.0990905761719, | |
| "step": 35 | |
| }, | |
| { | |
| "dpo_loss": 300.4722900390625, | |
| "epoch": 0.22673594709494568, | |
| "grad_norm": 134716.80909593205, | |
| "learning_rate": 2.2727272727272728e-06, | |
| "logits": -1.349226713180542, | |
| "logps": -83.15601348876953, | |
| "loss": 533.5176, | |
| "objective": 542.9343872070312, | |
| "ranking_idealized": 0.5208333134651184, | |
| "ranking_idealized_expo": 0.5208333134651184, | |
| "ranking_simple": 0.5208333134651184, | |
| "regularize": 542.9343872070312, | |
| "step": 40 | |
| }, | |
| { | |
| "dpo_loss": 362.15899658203125, | |
| "epoch": 0.25507794048181387, | |
| "grad_norm": 117630.82464120697, | |
| "learning_rate": 2.556818181818182e-06, | |
| "logits": -1.4267643690109253, | |
| "logps": -84.82117462158203, | |
| "loss": 670.75, | |
| "objective": 727.4680786132812, | |
| "ranking_idealized": 0.5375000238418579, | |
| "ranking_idealized_expo": 0.5375000238418579, | |
| "ranking_simple": 0.5291666388511658, | |
| "regularize": 727.4680786132812, | |
| "step": 45 | |
| }, | |
| { | |
| "dpo_loss": 410.1810302734375, | |
| "epoch": 0.2834199338686821, | |
| "grad_norm": 108815.79636709657, | |
| "learning_rate": 2.8409090909090916e-06, | |
| "logits": -1.2878683805465698, | |
| "logps": -84.17786407470703, | |
| "loss": 737.2396, | |
| "objective": 742.7522583007812, | |
| "ranking_idealized": 0.42500001192092896, | |
| "ranking_idealized_expo": 0.42500001192092896, | |
| "ranking_simple": 0.4375, | |
| "regularize": 742.7522583007812, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.2834199338686821, | |
| "eval_dpo_loss": 221.7373504638672, | |
| "eval_logits": -1.3124604225158691, | |
| "eval_logps": -92.02384185791016, | |
| "eval_loss": 421.9188232421875, | |
| "eval_objective": 428.8151550292969, | |
| "eval_ranking_idealized": 0.5092975497245789, | |
| "eval_ranking_idealized_expo": 0.5092975497245789, | |
| "eval_ranking_simple": 0.5092975497245789, | |
| "eval_regularize": 428.8151550292969, | |
| "eval_runtime": 260.7367, | |
| "eval_samples_per_second": 22.206, | |
| "eval_steps_per_second": 0.928, | |
| "step": 50 | |
| }, | |
| { | |
| "dpo_loss": 504.61846923828125, | |
| "epoch": 0.3117619272555503, | |
| "grad_norm": 119543.6777826292, | |
| "learning_rate": 3.125e-06, | |
| "logits": -1.3338755369186401, | |
| "logps": -85.74323272705078, | |
| "loss": 876.1629, | |
| "objective": 957.7632446289062, | |
| "ranking_idealized": 0.5416666865348816, | |
| "ranking_idealized_expo": 0.5416666865348816, | |
| "ranking_simple": 0.550000011920929, | |
| "regularize": 957.7632446289062, | |
| "step": 55 | |
| }, | |
| { | |
| "dpo_loss": 526.1536254882812, | |
| "epoch": 0.3401039206424185, | |
| "grad_norm": 107179.68101966614, | |
| "learning_rate": 3.409090909090909e-06, | |
| "logits": -1.3320441246032715, | |
| "logps": -82.8685531616211, | |
| "loss": 989.766, | |
| "objective": 928.43359375, | |
| "ranking_idealized": 0.512499988079071, | |
| "ranking_idealized_expo": 0.512499988079071, | |
| "ranking_simple": 0.5083333253860474, | |
| "regularize": 928.43359375, | |
| "step": 60 | |
| }, | |
| { | |
| "dpo_loss": 544.1323852539062, | |
| "epoch": 0.3684459140292867, | |
| "grad_norm": 108350.88761754551, | |
| "learning_rate": 3.6931818181818186e-06, | |
| "logits": -1.2518764734268188, | |
| "logps": -79.81038665771484, | |
| "loss": 1084.4579, | |
| "objective": 1179.59228515625, | |
| "ranking_idealized": 0.5083333253860474, | |
| "ranking_idealized_expo": 0.5083333253860474, | |
| "ranking_simple": 0.48750001192092896, | |
| "regularize": 1179.59228515625, | |
| "step": 65 | |
| }, | |
| { | |
| "dpo_loss": 636.8450927734375, | |
| "epoch": 0.39678790741615494, | |
| "grad_norm": 136538.92716952198, | |
| "learning_rate": 3.9772727272727275e-06, | |
| "logits": -1.2635269165039062, | |
| "logps": -79.40995788574219, | |
| "loss": 1189.2933, | |
| "objective": 1093.83642578125, | |
| "ranking_idealized": 0.5166666507720947, | |
| "ranking_idealized_expo": 0.5166666507720947, | |
| "ranking_simple": 0.4958333373069763, | |
| "regularize": 1093.83642578125, | |
| "step": 70 | |
| }, | |
| { | |
| "dpo_loss": 679.4451904296875, | |
| "epoch": 0.42512990080302315, | |
| "grad_norm": 97239.20388663356, | |
| "learning_rate": 4.2613636363636365e-06, | |
| "logits": -1.2477833032608032, | |
| "logps": -78.69037628173828, | |
| "loss": 1162.864, | |
| "objective": 1142.31640625, | |
| "ranking_idealized": 0.5458333492279053, | |
| "ranking_idealized_expo": 0.5458333492279053, | |
| "ranking_simple": 0.5375000238418579, | |
| "regularize": 1142.31640625, | |
| "step": 75 | |
| }, | |
| { | |
| "dpo_loss": 752.3662719726562, | |
| "epoch": 0.45347189418989137, | |
| "grad_norm": 87641.25402861548, | |
| "learning_rate": 4.5454545454545455e-06, | |
| "logits": -1.2982757091522217, | |
| "logps": -78.63927459716797, | |
| "loss": 1282.2214, | |
| "objective": 1388.2518310546875, | |
| "ranking_idealized": 0.5166666507720947, | |
| "ranking_idealized_expo": 0.5166666507720947, | |
| "ranking_simple": 0.512499988079071, | |
| "regularize": 1388.2518310546875, | |
| "step": 80 | |
| }, | |
| { | |
| "dpo_loss": 713.9522705078125, | |
| "epoch": 0.4818138875767596, | |
| "grad_norm": 88675.6017495157, | |
| "learning_rate": 4.829545454545455e-06, | |
| "logits": -1.225476622581482, | |
| "logps": -76.3520278930664, | |
| "loss": 1352.1976, | |
| "objective": 1389.2098388671875, | |
| "ranking_idealized": 0.5249999761581421, | |
| "ranking_idealized_expo": 0.5249999761581421, | |
| "ranking_simple": 0.5166666507720947, | |
| "regularize": 1389.2098388671875, | |
| "step": 85 | |
| }, | |
| { | |
| "dpo_loss": 791.7650756835938, | |
| "epoch": 0.5101558809636277, | |
| "grad_norm": 89967.15997141204, | |
| "learning_rate": 4.999921328558333e-06, | |
| "logits": -1.026414155960083, | |
| "logps": -77.92498016357422, | |
| "loss": 1440.1646, | |
| "objective": 1395.8138427734375, | |
| "ranking_idealized": 0.4791666567325592, | |
| "ranking_idealized_expo": 0.4791666567325592, | |
| "ranking_simple": 0.44583332538604736, | |
| "regularize": 1395.8138427734375, | |
| "step": 90 | |
| }, | |
| { | |
| "dpo_loss": 791.7685546875, | |
| "epoch": 0.538497874350496, | |
| "grad_norm": 91320.19274895087, | |
| "learning_rate": 4.999036331701828e-06, | |
| "logits": -1.0597717761993408, | |
| "logps": -76.71431732177734, | |
| "loss": 1517.9578, | |
| "objective": 1621.507080078125, | |
| "ranking_idealized": 0.5416666865348816, | |
| "ranking_idealized_expo": 0.5416666865348816, | |
| "ranking_simple": 0.550000011920929, | |
| "regularize": 1621.507080078125, | |
| "step": 95 | |
| }, | |
| { | |
| "dpo_loss": 673.1317749023438, | |
| "epoch": 0.5668398677373642, | |
| "grad_norm": 76415.34463234508, | |
| "learning_rate": 4.997168347957521e-06, | |
| "logits": -1.0299068689346313, | |
| "logps": -79.222412109375, | |
| "loss": 1559.0492, | |
| "objective": 1632.038330078125, | |
| "ranking_idealized": 0.5416666865348816, | |
| "ranking_idealized_expo": 0.5416666865348816, | |
| "ranking_simple": 0.5666666626930237, | |
| "regularize": 1632.038330078125, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.5668398677373642, | |
| "eval_dpo_loss": 733.09912109375, | |
| "eval_logits": -1.0226072072982788, | |
| "eval_logps": -82.79534149169922, | |
| "eval_loss": 1500.8583984375, | |
| "eval_objective": 1492.8040771484375, | |
| "eval_ranking_idealized": 0.5092975497245789, | |
| "eval_ranking_idealized_expo": 0.5092975497245789, | |
| "eval_ranking_simple": 0.5072314143180847, | |
| "eval_regularize": 1492.8040771484375, | |
| "eval_runtime": 260.8895, | |
| "eval_samples_per_second": 22.193, | |
| "eval_steps_per_second": 0.928, | |
| "step": 100 | |
| }, | |
| { | |
| "dpo_loss": 821.6388549804688, | |
| "epoch": 0.5951818611242324, | |
| "grad_norm": 78646.58325790247, | |
| "learning_rate": 4.994318112090048e-06, | |
| "logits": -0.8470949530601501, | |
| "logps": -78.14350891113281, | |
| "loss": 1588.0706, | |
| "objective": 1580.76416015625, | |
| "ranking_idealized": 0.5333333611488342, | |
| "ranking_idealized_expo": 0.5333333611488342, | |
| "ranking_simple": 0.5333333611488342, | |
| "regularize": 1580.76416015625, | |
| "step": 105 | |
| }, | |
| { | |
| "dpo_loss": 813.548583984375, | |
| "epoch": 0.6235238545111006, | |
| "grad_norm": 80078.74787055497, | |
| "learning_rate": 4.990486745229364e-06, | |
| "logits": -0.9708471298217773, | |
| "logps": -76.51264953613281, | |
| "loss": 1625.5101, | |
| "objective": 1670.44873046875, | |
| "ranking_idealized": 0.44999998807907104, | |
| "ranking_idealized_expo": 0.44999998807907104, | |
| "ranking_simple": 0.4416666626930237, | |
| "regularize": 1670.44873046875, | |
| "step": 110 | |
| }, | |
| { | |
| "dpo_loss": 967.8834838867188, | |
| "epoch": 0.6518658478979689, | |
| "grad_norm": 81186.63407260107, | |
| "learning_rate": 4.985675754429744e-06, | |
| "logits": -0.9930670261383057, | |
| "logps": -75.4658432006836, | |
| "loss": 1605.4639, | |
| "objective": 1732.50439453125, | |
| "ranking_idealized": 0.4833333194255829, | |
| "ranking_idealized_expo": 0.4833333194255829, | |
| "ranking_simple": 0.4749999940395355, | |
| "regularize": 1732.50439453125, | |
| "step": 115 | |
| }, | |
| { | |
| "dpo_loss": 781.5300903320312, | |
| "epoch": 0.680207841284837, | |
| "grad_norm": 73919.01988932796, | |
| "learning_rate": 4.9798870320769884e-06, | |
| "logits": -0.9557650685310364, | |
| "logps": -76.91903686523438, | |
| "loss": 1498.1906, | |
| "objective": 1478.94482421875, | |
| "ranking_idealized": 0.5333333611488342, | |
| "ranking_idealized_expo": 0.5333333611488342, | |
| "ranking_simple": 0.5249999761581421, | |
| "regularize": 1478.94482421875, | |
| "step": 120 | |
| }, | |
| { | |
| "dpo_loss": 951.0021362304688, | |
| "epoch": 0.7085498346717053, | |
| "grad_norm": 72729.38636733989, | |
| "learning_rate": 4.973122855144066e-06, | |
| "logits": -0.9417923092842102, | |
| "logps": -75.4057846069336, | |
| "loss": 1555.9701, | |
| "objective": 1580.7845458984375, | |
| "ranking_idealized": 0.550000011920929, | |
| "ranking_idealized_expo": 0.550000011920929, | |
| "ranking_simple": 0.5333333611488342, | |
| "regularize": 1580.7845458984375, | |
| "step": 125 | |
| }, | |
| { | |
| "dpo_loss": 830.5226440429688, | |
| "epoch": 0.7368918280585735, | |
| "grad_norm": 76215.9858570087, | |
| "learning_rate": 4.965385884295467e-06, | |
| "logits": -1.0607830286026, | |
| "logps": -75.56150817871094, | |
| "loss": 1535.5706, | |
| "objective": 1588.941650390625, | |
| "ranking_idealized": 0.44999998807907104, | |
| "ranking_idealized_expo": 0.44999998807907104, | |
| "ranking_simple": 0.4749999940395355, | |
| "regularize": 1588.941650390625, | |
| "step": 130 | |
| }, | |
| { | |
| "dpo_loss": 714.103271484375, | |
| "epoch": 0.7652338214454416, | |
| "grad_norm": 72632.27368489993, | |
| "learning_rate": 4.956679162840646e-06, | |
| "logits": -0.9477968215942383, | |
| "logps": -78.04027557373047, | |
| "loss": 1558.5909, | |
| "objective": 1568.9949951171875, | |
| "ranking_idealized": 0.5291666388511658, | |
| "ranking_idealized_expo": 0.5291666388511658, | |
| "ranking_simple": 0.5166666507720947, | |
| "regularize": 1568.9949951171875, | |
| "step": 135 | |
| }, | |
| { | |
| "dpo_loss": 747.6431884765625, | |
| "epoch": 0.7935758148323099, | |
| "grad_norm": 70212.75101987133, | |
| "learning_rate": 4.947006115536947e-06, | |
| "logits": -0.7389863729476929, | |
| "logps": -78.2565689086914, | |
| "loss": 1619.474, | |
| "objective": 1611.41845703125, | |
| "ranking_idealized": 0.5041666626930237, | |
| "ranking_idealized_expo": 0.5041666626930237, | |
| "ranking_simple": 0.5083333253860474, | |
| "regularize": 1611.41845703125, | |
| "step": 140 | |
| }, | |
| { | |
| "dpo_loss": 722.9232788085938, | |
| "epoch": 0.821917808219178, | |
| "grad_norm": 68650.23095680359, | |
| "learning_rate": 4.9363705472424825e-06, | |
| "logits": -0.7260185480117798, | |
| "logps": -78.47148132324219, | |
| "loss": 1489.2668, | |
| "objective": 1463.8128662109375, | |
| "ranking_idealized": 0.4833333194255829, | |
| "ranking_idealized_expo": 0.4833333194255829, | |
| "ranking_simple": 0.4749999940395355, | |
| "regularize": 1463.8128662109375, | |
| "step": 145 | |
| }, | |
| { | |
| "dpo_loss": 799.6663818359375, | |
| "epoch": 0.8502598016060463, | |
| "grad_norm": 70266.02356268882, | |
| "learning_rate": 4.924776641419513e-06, | |
| "logits": -0.7246825695037842, | |
| "logps": -77.73950958251953, | |
| "loss": 1544.2886, | |
| "objective": 1557.9066162109375, | |
| "ranking_idealized": 0.5375000238418579, | |
| "ranking_idealized_expo": 0.5375000238418579, | |
| "ranking_simple": 0.5375000238418579, | |
| "regularize": 1557.9066162109375, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.8502598016060463, | |
| "eval_dpo_loss": 936.6275634765625, | |
| "eval_logits": -0.8486141562461853, | |
| "eval_logps": -83.39346313476562, | |
| "eval_loss": 1796.87939453125, | |
| "eval_objective": 1837.2159423828125, | |
| "eval_ranking_idealized": 0.5092975497245789, | |
| "eval_ranking_idealized_expo": 0.5092975497245789, | |
| "eval_ranking_simple": 0.49896693229675293, | |
| "eval_regularize": 1837.2159423828125, | |
| "eval_runtime": 259.8051, | |
| "eval_samples_per_second": 22.286, | |
| "eval_steps_per_second": 0.931, | |
| "step": 150 | |
| }, | |
| { | |
| "dpo_loss": 733.1316528320312, | |
| "epoch": 0.8786017949929145, | |
| "grad_norm": 70225.42078139378, | |
| "learning_rate": 4.9122289584888926e-06, | |
| "logits": -0.8346138596534729, | |
| "logps": -76.66129302978516, | |
| "loss": 1527.9433, | |
| "objective": 1578.56982421875, | |
| "ranking_idealized": 0.512499988079071, | |
| "ranking_idealized_expo": 0.512499988079071, | |
| "ranking_simple": 0.5208333134651184, | |
| "regularize": 1578.56982421875, | |
| "step": 155 | |
| }, | |
| { | |
| "dpo_loss": 784.9417724609375, | |
| "epoch": 0.9069437883797827, | |
| "grad_norm": 74170.24944845453, | |
| "learning_rate": 4.8987324340362445e-06, | |
| "logits": -0.8054352402687073, | |
| "logps": -76.35670471191406, | |
| "loss": 1490.3339, | |
| "objective": 1511.6434326171875, | |
| "ranking_idealized": 0.4958333373069763, | |
| "ranking_idealized_expo": 0.4958333373069763, | |
| "ranking_simple": 0.49166667461395264, | |
| "regularize": 1511.6434326171875, | |
| "step": 160 | |
| }, | |
| { | |
| "dpo_loss": 794.072998046875, | |
| "epoch": 0.9352857817666509, | |
| "grad_norm": 72453.13771553633, | |
| "learning_rate": 4.884292376870567e-06, | |
| "logits": -0.7948398590087891, | |
| "logps": -77.8802490234375, | |
| "loss": 1440.4338, | |
| "objective": 1524.854248046875, | |
| "ranking_idealized": 0.5208333134651184, | |
| "ranking_idealized_expo": 0.5208333134651184, | |
| "ranking_simple": 0.512499988079071, | |
| "regularize": 1524.854248046875, | |
| "step": 165 | |
| }, | |
| { | |
| "dpo_loss": 855.9963989257812, | |
| "epoch": 0.9636277751535192, | |
| "grad_norm": 72841.32203749125, | |
| "learning_rate": 4.868914466936038e-06, | |
| "logits": -0.7872663140296936, | |
| "logps": -78.33912658691406, | |
| "loss": 1525.1104, | |
| "objective": 1553.3717041015625, | |
| "ranking_idealized": 0.5083333253860474, | |
| "ranking_idealized_expo": 0.5083333253860474, | |
| "ranking_simple": 0.5291666388511658, | |
| "regularize": 1553.3717041015625, | |
| "step": 170 | |
| }, | |
| { | |
| "dpo_loss": 774.2244262695312, | |
| "epoch": 0.9919697685403873, | |
| "grad_norm": 70428.90742393199, | |
| "learning_rate": 4.8526047530778175e-06, | |
| "logits": -0.7365032434463501, | |
| "logps": -77.8334732055664, | |
| "loss": 1490.3975, | |
| "objective": 1444.81201171875, | |
| "ranking_idealized": 0.574999988079071, | |
| "ranking_idealized_expo": 0.574999988079071, | |
| "ranking_simple": 0.5666666626930237, | |
| "regularize": 1444.81201171875, | |
| "step": 175 | |
| }, | |
| { | |
| "dpo_loss": 819.1224975585938, | |
| "epoch": 1.0203117619272555, | |
| "grad_norm": 68862.05104644371, | |
| "learning_rate": 4.835369650662767e-06, | |
| "logits": -0.7541021108627319, | |
| "logps": -79.26795959472656, | |
| "loss": 1489.1468, | |
| "objective": 1596.475830078125, | |
| "ranking_idealized": 0.5416666865348816, | |
| "ranking_idealized_expo": 0.5416666865348816, | |
| "ranking_simple": 0.5458333492279053, | |
| "regularize": 1596.475830078125, | |
| "step": 180 | |
| }, | |
| { | |
| "dpo_loss": 750.1921997070312, | |
| "epoch": 1.0486537553141237, | |
| "grad_norm": 68584.6258680744, | |
| "learning_rate": 4.817215939055984e-06, | |
| "logits": -0.6492658853530884, | |
| "logps": -76.89971160888672, | |
| "loss": 1358.0238, | |
| "objective": 1361.669921875, | |
| "ranking_idealized": 0.4833333194255829, | |
| "ranking_idealized_expo": 0.4833333194255829, | |
| "ranking_simple": 0.47083333134651184, | |
| "regularize": 1361.669921875, | |
| "step": 185 | |
| }, | |
| { | |
| "dpo_loss": 857.0948486328125, | |
| "epoch": 1.076995748700992, | |
| "grad_norm": 66212.13495810713, | |
| "learning_rate": 4.798150758954164e-06, | |
| "logits": -0.7034427523612976, | |
| "logps": -77.8251953125, | |
| "loss": 1414.8312, | |
| "objective": 1527.893310546875, | |
| "ranking_idealized": 0.550000011920929, | |
| "ranking_idealized_expo": 0.550000011920929, | |
| "ranking_simple": 0.5625, | |
| "regularize": 1527.893310546875, | |
| "step": 190 | |
| }, | |
| { | |
| "dpo_loss": 726.6162719726562, | |
| "epoch": 1.10533774208786, | |
| "grad_norm": 69617.2929524415, | |
| "learning_rate": 4.778181609576832e-06, | |
| "logits": -0.7142274379730225, | |
| "logps": -76.12057495117188, | |
| "loss": 1356.5107, | |
| "objective": 1371.98046875, | |
| "ranking_idealized": 0.5666666626930237, | |
| "ranking_idealized_expo": 0.5666666626930237, | |
| "ranking_simple": 0.5708333253860474, | |
| "regularize": 1371.98046875, | |
| "step": 195 | |
| }, | |
| { | |
| "dpo_loss": 723.1171875, | |
| "epoch": 1.1336797354747283, | |
| "grad_norm": 68913.9279715263, | |
| "learning_rate": 4.757316345716554e-06, | |
| "logits": -0.7645713686943054, | |
| "logps": -76.01897430419922, | |
| "loss": 1387.1779, | |
| "objective": 1441.4329833984375, | |
| "ranking_idealized": 0.5083333253860474, | |
| "ranking_idealized_expo": 0.5083333253860474, | |
| "ranking_simple": 0.5, | |
| "regularize": 1441.4329833984375, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.1336797354747283, | |
| "eval_dpo_loss": 1018.1060180664062, | |
| "eval_logits": -0.801040768623352, | |
| "eval_logps": -81.00394439697266, | |
| "eval_loss": 1946.64453125, | |
| "eval_objective": 1988.5870361328125, | |
| "eval_ranking_idealized": 0.5092975497245789, | |
| "eval_ranking_idealized_expo": 0.5092975497245789, | |
| "eval_ranking_simple": 0.5010330677032471, | |
| "eval_regularize": 1988.5870361328125, | |
| "eval_runtime": 261.3337, | |
| "eval_samples_per_second": 22.156, | |
| "eval_steps_per_second": 0.926, | |
| "step": 200 | |
| }, | |
| { | |
| "dpo_loss": 628.9264526367188, | |
| "epoch": 1.1620217288615966, | |
| "grad_norm": 63479.38452604848, | |
| "learning_rate": 4.735563174649278e-06, | |
| "logits": -0.7396198511123657, | |
| "logps": -76.6708755493164, | |
| "loss": 1375.7494, | |
| "objective": 1379.516845703125, | |
| "ranking_idealized": 0.48750001192092896, | |
| "ranking_idealized_expo": 0.48750001192092896, | |
| "ranking_simple": 0.512499988079071, | |
| "regularize": 1379.516845703125, | |
| "step": 205 | |
| }, | |
| { | |
| "dpo_loss": 673.1471557617188, | |
| "epoch": 1.1903637222484649, | |
| "grad_norm": 64539.915055366386, | |
| "learning_rate": 4.7129306529060415e-06, | |
| "logits": -0.721174955368042, | |
| "logps": -77.625732421875, | |
| "loss": 1383.7545, | |
| "objective": 1384.804443359375, | |
| "ranking_idealized": 0.4958333373069763, | |
| "ranking_idealized_expo": 0.4958333373069763, | |
| "ranking_simple": 0.5083333253860474, | |
| "regularize": 1384.804443359375, | |
| "step": 210 | |
| }, | |
| { | |
| "dpo_loss": 747.7578125, | |
| "epoch": 1.2187057156353331, | |
| "grad_norm": 62227.15430087378, | |
| "learning_rate": 4.68942768290728e-06, | |
| "logits": -0.6516799926757812, | |
| "logps": -77.21683502197266, | |
| "loss": 1315.6651, | |
| "objective": 1423.86669921875, | |
| "ranking_idealized": 0.4833333194255829, | |
| "ranking_idealized_expo": 0.4833333194255829, | |
| "ranking_simple": 0.4749999940395355, | |
| "regularize": 1423.86669921875, | |
| "step": 215 | |
| }, | |
| { | |
| "dpo_loss": 679.1942749023438, | |
| "epoch": 1.2470477090222012, | |
| "grad_norm": 74804.77662673105, | |
| "learning_rate": 4.665063509461098e-06, | |
| "logits": -0.5166414976119995, | |
| "logps": -77.51664733886719, | |
| "loss": 1282.1909, | |
| "objective": 1365.6038818359375, | |
| "ranking_idealized": 0.6000000238418579, | |
| "ranking_idealized_expo": 0.6000000238418579, | |
| "ranking_simple": 0.5833333134651184, | |
| "regularize": 1365.6038818359375, | |
| "step": 220 | |
| }, | |
| { | |
| "dpo_loss": 573.4988403320312, | |
| "epoch": 1.2753897024090695, | |
| "grad_norm": 68013.96758028683, | |
| "learning_rate": 4.639847716126855e-06, | |
| "logits": -0.6740496754646301, | |
| "logps": -78.64627075195312, | |
| "loss": 1277.235, | |
| "objective": 1197.0047607421875, | |
| "ranking_idealized": 0.5291666388511658, | |
| "ranking_idealized_expo": 0.5291666388511658, | |
| "ranking_simple": 0.5208333134651184, | |
| "regularize": 1197.0047607421875, | |
| "step": 225 | |
| }, | |
| { | |
| "dpo_loss": 539.927978515625, | |
| "epoch": 1.3037316957959377, | |
| "grad_norm": 63535.71644157263, | |
| "learning_rate": 4.613790221445511e-06, | |
| "logits": -0.649699866771698, | |
| "logps": -77.72300720214844, | |
| "loss": 1240.6593, | |
| "objective": 1214.3502197265625, | |
| "ranking_idealized": 0.5083333253860474, | |
| "ranking_idealized_expo": 0.5083333253860474, | |
| "ranking_simple": 0.5166666507720947, | |
| "regularize": 1214.3502197265625, | |
| "step": 230 | |
| }, | |
| { | |
| "dpo_loss": 614.890380859375, | |
| "epoch": 1.3320736891828058, | |
| "grad_norm": 63907.397620999705, | |
| "learning_rate": 4.586901275038201e-06, | |
| "logits": -0.7215268611907959, | |
| "logps": -75.93678283691406, | |
| "loss": 1294.443, | |
| "objective": 1246.745849609375, | |
| "ranking_idealized": 0.4833333194255829, | |
| "ranking_idealized_expo": 0.4833333194255829, | |
| "ranking_simple": 0.5041666626930237, | |
| "regularize": 1246.745849609375, | |
| "step": 235 | |
| }, | |
| { | |
| "dpo_loss": 674.669189453125, | |
| "epoch": 1.360415682569674, | |
| "grad_norm": 58847.80898438854, | |
| "learning_rate": 4.559191453574582e-06, | |
| "logits": -0.6891920566558838, | |
| "logps": -76.02350616455078, | |
| "loss": 1277.4793, | |
| "objective": 1229.0069580078125, | |
| "ranking_idealized": 0.4791666567325592, | |
| "ranking_idealized_expo": 0.4791666567325592, | |
| "ranking_simple": 0.4791666567325592, | |
| "regularize": 1229.0069580078125, | |
| "step": 240 | |
| }, | |
| { | |
| "dpo_loss": 682.5062866210938, | |
| "epoch": 1.3887576759565423, | |
| "grad_norm": 61032.25600688672, | |
| "learning_rate": 4.530671656612544e-06, | |
| "logits": -0.7168589234352112, | |
| "logps": -76.41120910644531, | |
| "loss": 1283.5139, | |
| "objective": 1295.427490234375, | |
| "ranking_idealized": 0.5166666507720947, | |
| "ranking_idealized_expo": 0.5166666507720947, | |
| "ranking_simple": 0.5208333134651184, | |
| "regularize": 1295.427490234375, | |
| "step": 245 | |
| }, | |
| { | |
| "dpo_loss": 529.6734008789062, | |
| "epoch": 1.4170996693434104, | |
| "grad_norm": 64433.112029992735, | |
| "learning_rate": 4.501353102310901e-06, | |
| "logits": -0.7715837955474854, | |
| "logps": -74.85303497314453, | |
| "loss": 1257.5858, | |
| "objective": 1227.3660888671875, | |
| "ranking_idealized": 0.47083333134651184, | |
| "ranking_idealized_expo": 0.47083333134651184, | |
| "ranking_simple": 0.4791666567325592, | |
| "regularize": 1227.3660888671875, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.4170996693434104, | |
| "eval_dpo_loss": 1059.197265625, | |
| "eval_logits": -0.8937114477157593, | |
| "eval_logps": -79.56282043457031, | |
| "eval_loss": 2056.783447265625, | |
| "eval_objective": 2078.739990234375, | |
| "eval_ranking_idealized": 0.5092975497245789, | |
| "eval_ranking_idealized_expo": 0.5092975497245789, | |
| "eval_ranking_simple": 0.5030992031097412, | |
| "eval_regularize": 2078.739990234375, | |
| "eval_runtime": 260.3832, | |
| "eval_samples_per_second": 22.236, | |
| "eval_steps_per_second": 0.929, | |
| "step": 250 | |
| }, | |
| { | |
| "dpo_loss": 606.7435913085938, | |
| "epoch": 1.4454416627302786, | |
| "grad_norm": 60772.290582903894, | |
| "learning_rate": 4.4712473230167775e-06, | |
| "logits": -0.7840810418128967, | |
| "logps": -76.13771057128906, | |
| "loss": 1227.542, | |
| "objective": 1233.7110595703125, | |
| "ranking_idealized": 0.5458333492279053, | |
| "ranking_idealized_expo": 0.5458333492279053, | |
| "ranking_simple": 0.5416666865348816, | |
| "regularize": 1233.7110595703125, | |
| "step": 255 | |
| }, | |
| { | |
| "dpo_loss": 675.6203002929688, | |
| "epoch": 1.473783656117147, | |
| "grad_norm": 61075.39269116474, | |
| "learning_rate": 4.440366160729393e-06, | |
| "logits": -0.7470805644989014, | |
| "logps": -77.35476684570312, | |
| "loss": 1195.1969, | |
| "objective": 1320.491455078125, | |
| "ranking_idealized": 0.48750001192092896, | |
| "ranking_idealized_expo": 0.48750001192092896, | |
| "ranking_simple": 0.5, | |
| "regularize": 1320.491455078125, | |
| "step": 260 | |
| }, | |
| { | |
| "dpo_loss": 581.8402099609375, | |
| "epoch": 1.5021256495040152, | |
| "grad_norm": 63860.91658867988, | |
| "learning_rate": 4.4087217624420595e-06, | |
| "logits": -0.7068949937820435, | |
| "logps": -75.44217681884766, | |
| "loss": 1146.6147, | |
| "objective": 1054.690673828125, | |
| "ranking_idealized": 0.5041666626930237, | |
| "ranking_idealized_expo": 0.5041666626930237, | |
| "ranking_simple": 0.4958333373069763, | |
| "regularize": 1054.690673828125, | |
| "step": 265 | |
| }, | |
| { | |
| "dpo_loss": 614.632080078125, | |
| "epoch": 1.5304676428908834, | |
| "grad_norm": 60024.06688450634, | |
| "learning_rate": 4.376326575364206e-06, | |
| "logits": -0.7246025204658508, | |
| "logps": -75.80039978027344, | |
| "loss": 1165.9641, | |
| "objective": 1283.4388427734375, | |
| "ranking_idealized": 0.5041666626930237, | |
| "ranking_idealized_expo": 0.5041666626930237, | |
| "ranking_simple": 0.5208333134651184, | |
| "regularize": 1283.4388427734375, | |
| "step": 270 | |
| }, | |
| { | |
| "dpo_loss": 554.4395751953125, | |
| "epoch": 1.5588096362777515, | |
| "grad_norm": 60695.49838031304, | |
| "learning_rate": 4.34319334202531e-06, | |
| "logits": -0.7526471018791199, | |
| "logps": -74.80017852783203, | |
| "loss": 1087.4088, | |
| "objective": 1118.8092041015625, | |
| "ranking_idealized": 0.5249999761581421, | |
| "ranking_idealized_expo": 0.5249999761581421, | |
| "ranking_simple": 0.5208333134651184, | |
| "regularize": 1118.8092041015625, | |
| "step": 275 | |
| }, | |
| { | |
| "dpo_loss": 546.6119995117188, | |
| "epoch": 1.5871516296646198, | |
| "grad_norm": 60429.09566571946, | |
| "learning_rate": 4.309335095262675e-06, | |
| "logits": -0.673143208026886, | |
| "logps": -72.87140655517578, | |
| "loss": 1104.0904, | |
| "objective": 1092.2620849609375, | |
| "ranking_idealized": 0.5041666626930237, | |
| "ranking_idealized_expo": 0.5041666626930237, | |
| "ranking_simple": 0.4958333373069763, | |
| "regularize": 1092.2620849609375, | |
| "step": 280 | |
| }, | |
| { | |
| "dpo_loss": 584.6941528320312, | |
| "epoch": 1.615493623051488, | |
| "grad_norm": 61625.27955622327, | |
| "learning_rate": 4.274765153095008e-06, | |
| "logits": -0.687822163105011, | |
| "logps": -74.89695739746094, | |
| "loss": 1180.4073, | |
| "objective": 1178.606689453125, | |
| "ranking_idealized": 0.5166666507720947, | |
| "ranking_idealized_expo": 0.5166666507720947, | |
| "ranking_simple": 0.512499988079071, | |
| "regularize": 1178.606689453125, | |
| "step": 285 | |
| }, | |
| { | |
| "dpo_loss": 505.9402770996094, | |
| "epoch": 1.643835616438356, | |
| "grad_norm": 57020.12431127765, | |
| "learning_rate": 4.239497113483819e-06, | |
| "logits": -0.6894593834877014, | |
| "logps": -73.80435943603516, | |
| "loss": 1135.5879, | |
| "objective": 1057.634765625, | |
| "ranking_idealized": 0.5625, | |
| "ranking_idealized_expo": 0.5625, | |
| "ranking_simple": 0.574999988079071, | |
| "regularize": 1057.634765625, | |
| "step": 290 | |
| }, | |
| { | |
| "dpo_loss": 584.1075439453125, | |
| "epoch": 1.6721776098252243, | |
| "grad_norm": 57611.18104908386, | |
| "learning_rate": 4.203544848984729e-06, | |
| "logits": -0.6624549031257629, | |
| "logps": -72.43326568603516, | |
| "loss": 1093.2101, | |
| "objective": 1054.095458984375, | |
| "ranking_idealized": 0.4791666567325592, | |
| "ranking_idealized_expo": 0.4791666567325592, | |
| "ranking_simple": 0.49166667461395264, | |
| "regularize": 1054.095458984375, | |
| "step": 295 | |
| }, | |
| { | |
| "dpo_loss": 577.9215698242188, | |
| "epoch": 1.7005196032120926, | |
| "grad_norm": 64211.25163104252, | |
| "learning_rate": 4.16692250129073e-06, | |
| "logits": -0.675134003162384, | |
| "logps": -74.83804321289062, | |
| "loss": 1062.9078, | |
| "objective": 1094.4080810546875, | |
| "ranking_idealized": 0.574999988079071, | |
| "ranking_idealized_expo": 0.574999988079071, | |
| "ranking_simple": 0.550000011920929, | |
| "regularize": 1094.4080810546875, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.7005196032120926, | |
| "eval_dpo_loss": 1115.767822265625, | |
| "eval_logits": -0.7208537459373474, | |
| "eval_logps": -79.72731018066406, | |
| "eval_loss": 2170.694580078125, | |
| "eval_objective": 2202.780517578125, | |
| "eval_ranking_idealized": 0.5092975497245789, | |
| "eval_ranking_idealized_expo": 0.5092975497245789, | |
| "eval_ranking_simple": 0.5030992031097412, | |
| "eval_regularize": 2202.780517578125, | |
| "eval_runtime": 260.4557, | |
| "eval_samples_per_second": 22.23, | |
| "eval_steps_per_second": 0.929, | |
| "step": 300 | |
| }, | |
| { | |
| "dpo_loss": 499.92578125, | |
| "epoch": 1.7288615965989607, | |
| "grad_norm": 59997.71496488459, | |
| "learning_rate": 4.129644475669617e-06, | |
| "logits": -0.6386987566947937, | |
| "logps": -76.04856872558594, | |
| "loss": 1068.0816, | |
| "objective": 1115.4102783203125, | |
| "ranking_idealized": 0.5041666626930237, | |
| "ranking_idealized_expo": 0.5041666626930237, | |
| "ranking_simple": 0.5083333253860474, | |
| "regularize": 1115.4102783203125, | |
| "step": 305 | |
| }, | |
| { | |
| "dpo_loss": 489.96710205078125, | |
| "epoch": 1.7572035899858292, | |
| "grad_norm": 64953.71255296531, | |
| "learning_rate": 4.091725435297721e-06, | |
| "logits": -0.6985257267951965, | |
| "logps": -73.54243469238281, | |
| "loss": 1091.0647, | |
| "objective": 1090.1666259765625, | |
| "ranking_idealized": 0.5208333134651184, | |
| "ranking_idealized_expo": 0.5208333134651184, | |
| "ranking_simple": 0.5166666507720947, | |
| "regularize": 1090.1666259765625, | |
| "step": 310 | |
| }, | |
| { | |
| "dpo_loss": 550.898681640625, | |
| "epoch": 1.7855455833726972, | |
| "grad_norm": 60495.13771709575, | |
| "learning_rate": 4.053180295492203e-06, | |
| "logits": -0.6273626685142517, | |
| "logps": -75.3206558227539, | |
| "loss": 1086.04, | |
| "objective": 1067.1929931640625, | |
| "ranking_idealized": 0.5083333253860474, | |
| "ranking_idealized_expo": 0.5083333253860474, | |
| "ranking_simple": 0.512499988079071, | |
| "regularize": 1067.1929931640625, | |
| "step": 315 | |
| }, | |
| { | |
| "dpo_loss": 710.1861572265625, | |
| "epoch": 1.8138875767595655, | |
| "grad_norm": 62915.79210205411, | |
| "learning_rate": 4.014024217844167e-06, | |
| "logits": -0.6406947374343872, | |
| "logps": -76.75931549072266, | |
| "loss": 1107.8994, | |
| "objective": 1157.3203125, | |
| "ranking_idealized": 0.5041666626930237, | |
| "ranking_idealized_expo": 0.5041666626930237, | |
| "ranking_simple": 0.4833333194255829, | |
| "regularize": 1157.3203125, | |
| "step": 320 | |
| }, | |
| { | |
| "dpo_loss": 698.9453125, | |
| "epoch": 1.8422295701464337, | |
| "grad_norm": 66764.29984408473, | |
| "learning_rate": 3.974272604254906e-06, | |
| "logits": -0.6487582921981812, | |
| "logps": -78.7199935913086, | |
| "loss": 1102.0305, | |
| "objective": 1170.1241455078125, | |
| "ranking_idealized": 0.5166666507720947, | |
| "ranking_idealized_expo": 0.5166666507720947, | |
| "ranking_simple": 0.550000011920929, | |
| "regularize": 1170.1241455078125, | |
| "step": 325 | |
| }, | |
| { | |
| "dpo_loss": 565.6978759765625, | |
| "epoch": 1.8705715635333018, | |
| "grad_norm": 61211.42685047196, | |
| "learning_rate": 3.933941090877615e-06, | |
| "logits": -0.547908365726471, | |
| "logps": -75.97573852539062, | |
| "loss": 1043.7004, | |
| "objective": 1092.4935302734375, | |
| "ranking_idealized": 0.47083333134651184, | |
| "ranking_idealized_expo": 0.47083333134651184, | |
| "ranking_simple": 0.4541666805744171, | |
| "regularize": 1092.4935302734375, | |
| "step": 330 | |
| }, | |
| { | |
| "dpo_loss": 442.16241455078125, | |
| "epoch": 1.89891355692017, | |
| "grad_norm": 59387.646813099236, | |
| "learning_rate": 3.893045541966975e-06, | |
| "logits": -0.6546408534049988, | |
| "logps": -75.68651580810547, | |
| "loss": 1035.5035, | |
| "objective": 1078.9815673828125, | |
| "ranking_idealized": 0.4958333373069763, | |
| "ranking_idealized_expo": 0.4958333373069763, | |
| "ranking_simple": 0.5083333253860474, | |
| "regularize": 1078.9815673828125, | |
| "step": 335 | |
| }, | |
| { | |
| "dpo_loss": 504.0133361816406, | |
| "epoch": 1.9272555503070383, | |
| "grad_norm": 65019.32677125802, | |
| "learning_rate": 3.8516020436389945e-06, | |
| "logits": -0.611504852771759, | |
| "logps": -79.22772979736328, | |
| "loss": 1007.5374, | |
| "objective": 1054.186767578125, | |
| "ranking_idealized": 0.5416666865348816, | |
| "ranking_idealized_expo": 0.5416666865348816, | |
| "ranking_simple": 0.5333333611488342, | |
| "regularize": 1054.186767578125, | |
| "step": 340 | |
| }, | |
| { | |
| "dpo_loss": 512.2960815429688, | |
| "epoch": 1.9555975436939064, | |
| "grad_norm": 62729.616628268755, | |
| "learning_rate": 3.8096268975436045e-06, | |
| "logits": -0.6399423480033875, | |
| "logps": -77.9858627319336, | |
| "loss": 990.7236, | |
| "objective": 1057.446533203125, | |
| "ranking_idealized": 0.5333333611488342, | |
| "ranking_idealized_expo": 0.5333333611488342, | |
| "ranking_simple": 0.5249999761581421, | |
| "regularize": 1057.446533203125, | |
| "step": 345 | |
| }, | |
| { | |
| "dpo_loss": 415.81988525390625, | |
| "epoch": 1.9839395370807746, | |
| "grad_norm": 63823.03198972447, | |
| "learning_rate": 3.767136614452458e-06, | |
| "logits": -0.5935099124908447, | |
| "logps": -79.65575408935547, | |
| "loss": 1015.0369, | |
| "objective": 949.8970336914062, | |
| "ranking_idealized": 0.48750001192092896, | |
| "ranking_idealized_expo": 0.48750001192092896, | |
| "ranking_simple": 0.47083333134651184, | |
| "regularize": 949.8970336914062, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.9839395370807746, | |
| "eval_dpo_loss": 1156.4827880859375, | |
| "eval_logits": -0.673879086971283, | |
| "eval_logps": -83.59508514404297, | |
| "eval_loss": 2227.17138671875, | |
| "eval_objective": 2262.3740234375, | |
| "eval_ranking_idealized": 0.5092975497245789, | |
| "eval_ranking_idealized_expo": 0.5092975497245789, | |
| "eval_ranking_simple": 0.5123966932296753, | |
| "eval_regularize": 2262.3740234375, | |
| "eval_runtime": 260.1877, | |
| "eval_samples_per_second": 22.253, | |
| "eval_steps_per_second": 0.93, | |
| "step": 350 | |
| }, | |
| { | |
| "dpo_loss": 531.9288940429688, | |
| "epoch": 2.012281530467643, | |
| "grad_norm": 59707.22724571865, | |
| "learning_rate": 3.724147907764478e-06, | |
| "logits": -0.5535954236984253, | |
| "logps": -79.02854919433594, | |
| "loss": 973.0249, | |
| "objective": 1016.3368530273438, | |
| "ranking_idealized": 0.5083333253860474, | |
| "ranking_idealized_expo": 0.5083333253860474, | |
| "ranking_simple": 0.5083333253860474, | |
| "regularize": 1016.3368530273438, | |
| "step": 355 | |
| }, | |
| { | |
| "dpo_loss": 468.62908935546875, | |
| "epoch": 2.040623523854511, | |
| "grad_norm": 57289.87868431793, | |
| "learning_rate": 3.6806776869317074e-06, | |
| "logits": -0.5899576544761658, | |
| "logps": -77.1968002319336, | |
| "loss": 955.7221, | |
| "objective": 970.0728759765625, | |
| "ranking_idealized": 0.5291666388511658, | |
| "ranking_idealized_expo": 0.5291666388511658, | |
| "ranking_simple": 0.550000011920929, | |
| "regularize": 970.0728759765625, | |
| "step": 360 | |
| }, | |
| { | |
| "dpo_loss": 416.2799072265625, | |
| "epoch": 2.0689655172413794, | |
| "grad_norm": 62539.981653631265, | |
| "learning_rate": 3.6367430508080283e-06, | |
| "logits": -0.6331656575202942, | |
| "logps": -80.54222106933594, | |
| "loss": 919.5296, | |
| "objective": 905.774169921875, | |
| "ranking_idealized": 0.5041666626930237, | |
| "ranking_idealized_expo": 0.5041666626930237, | |
| "ranking_simple": 0.5208333134651184, | |
| "regularize": 905.774169921875, | |
| "step": 365 | |
| }, | |
| { | |
| "dpo_loss": 487.22967529296875, | |
| "epoch": 2.0973075106282475, | |
| "grad_norm": 57687.52155159706, | |
| "learning_rate": 3.5923612809233987e-06, | |
| "logits": -0.5682953000068665, | |
| "logps": -77.53422546386719, | |
| "loss": 867.4084, | |
| "objective": 892.9639892578125, | |
| "ranking_idealized": 0.4833333194255829, | |
| "ranking_idealized_expo": 0.4833333194255829, | |
| "ranking_simple": 0.49166667461395264, | |
| "regularize": 892.9639892578125, | |
| "step": 370 | |
| }, | |
| { | |
| "dpo_loss": 387.1273193359375, | |
| "epoch": 2.1256495040151155, | |
| "grad_norm": 58841.64643600781, | |
| "learning_rate": 3.547549834686222e-06, | |
| "logits": -0.5675191879272461, | |
| "logps": -79.55015563964844, | |
| "loss": 896.5979, | |
| "objective": 880.7286376953125, | |
| "ranking_idealized": 0.5874999761581421, | |
| "ranking_idealized_expo": 0.5874999761581421, | |
| "ranking_simple": 0.5833333134651184, | |
| "regularize": 880.7286376953125, | |
| "step": 375 | |
| }, | |
| { | |
| "dpo_loss": 408.3785095214844, | |
| "epoch": 2.153991497401984, | |
| "grad_norm": 58397.78972496958, | |
| "learning_rate": 3.5023263385165346e-06, | |
| "logits": -0.5271536707878113, | |
| "logps": -78.26580047607422, | |
| "loss": 892.9402, | |
| "objective": 895.1597900390625, | |
| "ranking_idealized": 0.5291666388511658, | |
| "ranking_idealized_expo": 0.5291666388511658, | |
| "ranking_simple": 0.5333333611488342, | |
| "regularize": 895.1597900390625, | |
| "step": 380 | |
| }, | |
| { | |
| "dpo_loss": 378.6316833496094, | |
| "epoch": 2.182333490788852, | |
| "grad_norm": 57845.20228394267, | |
| "learning_rate": 3.4567085809127247e-06, | |
| "logits": -0.5461028218269348, | |
| "logps": -81.12430572509766, | |
| "loss": 854.0151, | |
| "objective": 875.6107177734375, | |
| "ranking_idealized": 0.5, | |
| "ranking_idealized_expo": 0.5, | |
| "ranking_simple": 0.5249999761581421, | |
| "regularize": 875.6107177734375, | |
| "step": 385 | |
| }, | |
| { | |
| "dpo_loss": 431.4305114746094, | |
| "epoch": 2.21067548417572, | |
| "grad_norm": 58847.02033627762, | |
| "learning_rate": 3.410714505454486e-06, | |
| "logits": -0.5168668627738953, | |
| "logps": -79.27171325683594, | |
| "loss": 854.3453, | |
| "objective": 838.873779296875, | |
| "ranking_idealized": 0.4833333194255829, | |
| "ranking_idealized_expo": 0.4833333194255829, | |
| "ranking_simple": 0.47083333134651184, | |
| "regularize": 838.873779296875, | |
| "step": 390 | |
| }, | |
| { | |
| "dpo_loss": 492.15777587890625, | |
| "epoch": 2.2390174775625886, | |
| "grad_norm": 58173.621993324545, | |
| "learning_rate": 3.364362203744777e-06, | |
| "logits": -0.5718950629234314, | |
| "logps": -79.84632110595703, | |
| "loss": 891.3756, | |
| "objective": 879.3973388671875, | |
| "ranking_idealized": 0.5166666507720947, | |
| "ranking_idealized_expo": 0.5166666507720947, | |
| "ranking_simple": 0.5166666507720947, | |
| "regularize": 879.3973388671875, | |
| "step": 395 | |
| }, | |
| { | |
| "dpo_loss": 420.82257080078125, | |
| "epoch": 2.2673594709494567, | |
| "grad_norm": 56625.7986223953, | |
| "learning_rate": 3.3176699082935546e-06, | |
| "logits": -0.6329799294471741, | |
| "logps": -81.30545806884766, | |
| "loss": 849.8354, | |
| "objective": 825.5804443359375, | |
| "ranking_idealized": 0.574999988079071, | |
| "ranking_idealized_expo": 0.574999988079071, | |
| "ranking_simple": 0.5958333611488342, | |
| "regularize": 825.5804443359375, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.2673594709494567, | |
| "eval_dpo_loss": 1124.7908935546875, | |
| "eval_logits": -0.6953743100166321, | |
| "eval_logps": -83.39958190917969, | |
| "eval_loss": 2210.667236328125, | |
| "eval_objective": 2238.018798828125, | |
| "eval_ranking_idealized": 0.5092975497245789, | |
| "eval_ranking_idealized_expo": 0.5092975497245789, | |
| "eval_ranking_simple": 0.5154958963394165, | |
| "eval_regularize": 2238.018798828125, | |
| "eval_runtime": 260.8763, | |
| "eval_samples_per_second": 22.194, | |
| "eval_steps_per_second": 0.928, | |
| "step": 400 | |
| }, | |
| { | |
| "dpo_loss": 455.4064636230469, | |
| "epoch": 2.295701464336325, | |
| "grad_norm": 59678.323118492524, | |
| "learning_rate": 3.2706559853460818e-06, | |
| "logits": -0.6739230751991272, | |
| "logps": -79.11216735839844, | |
| "loss": 848.5012, | |
| "objective": 833.8142700195312, | |
| "ranking_idealized": 0.5583333373069763, | |
| "ranking_idealized_expo": 0.5583333373069763, | |
| "ranking_simple": 0.5708333253860474, | |
| "regularize": 833.8142700195312, | |
| "step": 405 | |
| }, | |
| { | |
| "dpo_loss": 401.0616455078125, | |
| "epoch": 2.324043457723193, | |
| "grad_norm": 65764.3857558314, | |
| "learning_rate": 3.2233389276586325e-06, | |
| "logits": -0.596556544303894, | |
| "logps": -78.09807586669922, | |
| "loss": 830.7495, | |
| "objective": 797.6709594726562, | |
| "ranking_idealized": 0.4749999940395355, | |
| "ranking_idealized_expo": 0.4749999940395355, | |
| "ranking_simple": 0.4749999940395355, | |
| "regularize": 797.6709594726562, | |
| "step": 410 | |
| }, | |
| { | |
| "dpo_loss": 449.0657653808594, | |
| "epoch": 2.3523854511100613, | |
| "grad_norm": 60576.92516292515, | |
| "learning_rate": 3.1757373472244324e-06, | |
| "logits": -0.6169402003288269, | |
| "logps": -77.96237182617188, | |
| "loss": 830.681, | |
| "objective": 843.7548217773438, | |
| "ranking_idealized": 0.5291666388511658, | |
| "ranking_idealized_expo": 0.5291666388511658, | |
| "ranking_simple": 0.5291666388511658, | |
| "regularize": 843.7548217773438, | |
| "step": 415 | |
| }, | |
| { | |
| "dpo_loss": 407.09564208984375, | |
| "epoch": 2.3807274444969297, | |
| "grad_norm": 59864.322780894625, | |
| "learning_rate": 3.127869967952698e-06, | |
| "logits": -0.597426176071167, | |
| "logps": -78.8098373413086, | |
| "loss": 802.8715, | |
| "objective": 764.6652221679688, | |
| "ranking_idealized": 0.47083333134651184, | |
| "ranking_idealized_expo": 0.47083333134651184, | |
| "ranking_simple": 0.4583333432674408, | |
| "regularize": 764.6652221679688, | |
| "step": 420 | |
| }, | |
| { | |
| "dpo_loss": 370.85565185546875, | |
| "epoch": 2.409069437883798, | |
| "grad_norm": 59177.15286602958, | |
| "learning_rate": 3.0797556183036582e-06, | |
| "logits": -0.6467216610908508, | |
| "logps": -77.39534759521484, | |
| "loss": 813.4841, | |
| "objective": 817.958251953125, | |
| "ranking_idealized": 0.4958333373069763, | |
| "ranking_idealized_expo": 0.4958333373069763, | |
| "ranking_simple": 0.5083333253860474, | |
| "regularize": 817.958251953125, | |
| "step": 425 | |
| }, | |
| { | |
| "dpo_loss": 354.9871826171875, | |
| "epoch": 2.4374114312706663, | |
| "grad_norm": 59260.79608193281, | |
| "learning_rate": 3.0314132238824416e-06, | |
| "logits": -0.6649688482284546, | |
| "logps": -77.8988037109375, | |
| "loss": 765.4834, | |
| "objective": 690.4028930664062, | |
| "ranking_idealized": 0.5291666388511658, | |
| "ranking_idealized_expo": 0.5291666388511658, | |
| "ranking_simple": 0.5375000238418579, | |
| "regularize": 690.4028930664062, | |
| "step": 430 | |
| }, | |
| { | |
| "dpo_loss": 338.4743347167969, | |
| "epoch": 2.4657534246575343, | |
| "grad_norm": 55351.556007813095, | |
| "learning_rate": 2.9828617999947647e-06, | |
| "logits": -0.6696223616600037, | |
| "logps": -77.22901916503906, | |
| "loss": 774.5823, | |
| "objective": 722.385009765625, | |
| "ranking_idealized": 0.5083333253860474, | |
| "ranking_idealized_expo": 0.5083333253860474, | |
| "ranking_simple": 0.512499988079071, | |
| "regularize": 722.385009765625, | |
| "step": 435 | |
| }, | |
| { | |
| "dpo_loss": 500.3384704589844, | |
| "epoch": 2.4940954180444024, | |
| "grad_norm": 61581.25711417744, | |
| "learning_rate": 2.9341204441673267e-06, | |
| "logits": -0.6167859435081482, | |
| "logps": -76.51101684570312, | |
| "loss": 785.1207, | |
| "objective": 844.1969604492188, | |
| "ranking_idealized": 0.5375000238418579, | |
| "ranking_idealized_expo": 0.5333333611488342, | |
| "ranking_simple": 0.5291666388511658, | |
| "regularize": 844.1969604492188, | |
| "step": 440 | |
| }, | |
| { | |
| "dpo_loss": 445.2440185546875, | |
| "epoch": 2.5224374114312704, | |
| "grad_norm": 59474.235147403386, | |
| "learning_rate": 2.8852083286358647e-06, | |
| "logits": -0.6244394779205322, | |
| "logps": -74.75468444824219, | |
| "loss": 735.9972, | |
| "objective": 698.7183227539062, | |
| "ranking_idealized": 0.5166666507720947, | |
| "ranking_idealized_expo": 0.5166666507720947, | |
| "ranking_simple": 0.4958333373069763, | |
| "regularize": 698.7183227539062, | |
| "step": 445 | |
| }, | |
| { | |
| "dpo_loss": 383.2066650390625, | |
| "epoch": 2.550779404818139, | |
| "grad_norm": 63593.63871540052, | |
| "learning_rate": 2.8361446928038298e-06, | |
| "logits": -0.6291276216506958, | |
| "logps": -76.71273040771484, | |
| "loss": 749.1392, | |
| "objective": 726.7490234375, | |
| "ranking_idealized": 0.5, | |
| "ranking_idealized_expo": 0.5, | |
| "ranking_simple": 0.49166667461395264, | |
| "regularize": 726.7490234375, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 2.550779404818139, | |
| "eval_dpo_loss": 1157.103515625, | |
| "eval_logits": -0.6204014420509338, | |
| "eval_logps": -80.84976196289062, | |
| "eval_loss": 2232.329833984375, | |
| "eval_objective": 2283.406982421875, | |
| "eval_ranking_idealized": 0.5092975497245789, | |
| "eval_ranking_idealized_expo": 0.5092975497245789, | |
| "eval_ranking_simple": 0.5134297609329224, | |
| "eval_regularize": 2283.406982421875, | |
| "eval_runtime": 260.1477, | |
| "eval_samples_per_second": 22.257, | |
| "eval_steps_per_second": 0.93, | |
| "step": 450 | |
| }, | |
| { | |
| "dpo_loss": 370.9847106933594, | |
| "epoch": 2.579121398205007, | |
| "grad_norm": 58175.776810504, | |
| "learning_rate": 2.7869488356746344e-06, | |
| "logits": -0.5754743814468384, | |
| "logps": -77.64400482177734, | |
| "loss": 764.5896, | |
| "objective": 809.1742553710938, | |
| "ranking_idealized": 0.5, | |
| "ranking_idealized_expo": 0.5, | |
| "ranking_simple": 0.5083333253860474, | |
| "regularize": 809.1742553710938, | |
| "step": 455 | |
| }, | |
| { | |
| "dpo_loss": 425.83538818359375, | |
| "epoch": 2.6074633915918755, | |
| "grad_norm": 56477.01767094451, | |
| "learning_rate": 2.7376401082604563e-06, | |
| "logits": -0.5983397364616394, | |
| "logps": -77.55906677246094, | |
| "loss": 744.7457, | |
| "objective": 810.4805908203125, | |
| "ranking_idealized": 0.512499988079071, | |
| "ranking_idealized_expo": 0.512499988079071, | |
| "ranking_simple": 0.5333333611488342, | |
| "regularize": 810.4805908203125, | |
| "step": 460 | |
| }, | |
| { | |
| "dpo_loss": 349.92388916015625, | |
| "epoch": 2.6358053849787435, | |
| "grad_norm": 58522.655824615445, | |
| "learning_rate": 2.6882379059705953e-06, | |
| "logits": -0.5105490684509277, | |
| "logps": -77.51533508300781, | |
| "loss": 699.5353, | |
| "objective": 748.3710327148438, | |
| "ranking_idealized": 0.4958333373069763, | |
| "ranking_idealized_expo": 0.4958333373069763, | |
| "ranking_simple": 0.5083333253860474, | |
| "regularize": 748.3710327148438, | |
| "step": 465 | |
| }, | |
| { | |
| "dpo_loss": 312.98828125, | |
| "epoch": 2.6641473783656116, | |
| "grad_norm": 58397.15682461157, | |
| "learning_rate": 2.6387616609823506e-06, | |
| "logits": -0.5410483479499817, | |
| "logps": -76.95088195800781, | |
| "loss": 705.2115, | |
| "objective": 725.9006958007812, | |
| "ranking_idealized": 0.550000011920929, | |
| "ranking_idealized_expo": 0.550000011920929, | |
| "ranking_simple": 0.5458333492279053, | |
| "regularize": 725.9006958007812, | |
| "step": 470 | |
| }, | |
| { | |
| "dpo_loss": 282.96087646484375, | |
| "epoch": 2.69248937175248, | |
| "grad_norm": 60283.35669741736, | |
| "learning_rate": 2.5892308345974517e-06, | |
| "logits": -0.5889594554901123, | |
| "logps": -76.49653625488281, | |
| "loss": 705.286, | |
| "objective": 678.2340698242188, | |
| "ranking_idealized": 0.4958333373069763, | |
| "ranking_idealized_expo": 0.4958333373069763, | |
| "ranking_simple": 0.49166667461395264, | |
| "regularize": 678.2340698242188, | |
| "step": 475 | |
| }, | |
| { | |
| "dpo_loss": 387.2418518066406, | |
| "epoch": 2.720831365139348, | |
| "grad_norm": 58143.27861764097, | |
| "learning_rate": 2.53966490958702e-06, | |
| "logits": -0.6299055814743042, | |
| "logps": -76.01927185058594, | |
| "loss": 694.6192, | |
| "objective": 725.1319580078125, | |
| "ranking_idealized": 0.5, | |
| "ranking_idealized_expo": 0.5, | |
| "ranking_simple": 0.5, | |
| "regularize": 725.1319580078125, | |
| "step": 480 | |
| }, | |
| { | |
| "dpo_loss": 366.7168884277344, | |
| "epoch": 2.7491733585262166, | |
| "grad_norm": 57418.84427310581, | |
| "learning_rate": 2.490083382528097e-06, | |
| "logits": -0.5646944642066956, | |
| "logps": -78.08667755126953, | |
| "loss": 658.8851, | |
| "objective": 672.9464721679688, | |
| "ranking_idealized": 0.5208333134651184, | |
| "ranking_idealized_expo": 0.5208333134651184, | |
| "ranking_simple": 0.5166666507720947, | |
| "regularize": 672.9464721679688, | |
| "step": 485 | |
| }, | |
| { | |
| "dpo_loss": 299.7682800292969, | |
| "epoch": 2.7775153519130846, | |
| "grad_norm": 57617.14262224855, | |
| "learning_rate": 2.440505756134732e-06, | |
| "logits": -0.5018786787986755, | |
| "logps": -76.6084976196289, | |
| "loss": 710.2159, | |
| "objective": 683.9637451171875, | |
| "ranking_idealized": 0.47083333134651184, | |
| "ranking_idealized_expo": 0.47083333134651184, | |
| "ranking_simple": 0.47083333134651184, | |
| "regularize": 683.9637451171875, | |
| "step": 490 | |
| }, | |
| { | |
| "dpo_loss": 372.53076171875, | |
| "epoch": 2.8058573452999527, | |
| "grad_norm": 57177.66074623478, | |
| "learning_rate": 2.3909515315866606e-06, | |
| "logits": -0.543052077293396, | |
| "logps": -74.97837829589844, | |
| "loss": 666.0089, | |
| "objective": 707.5836181640625, | |
| "ranking_idealized": 0.4791666567325592, | |
| "ranking_idealized_expo": 0.4791666567325592, | |
| "ranking_simple": 0.4625000059604645, | |
| "regularize": 707.5836181640625, | |
| "step": 495 | |
| }, | |
| { | |
| "dpo_loss": 402.10662841796875, | |
| "epoch": 2.8341993386868207, | |
| "grad_norm": 57068.622360461784, | |
| "learning_rate": 2.341440200858589e-06, | |
| "logits": -0.5944234132766724, | |
| "logps": -74.34591674804688, | |
| "loss": 663.6063, | |
| "objective": 650.8515625, | |
| "ranking_idealized": 0.5208333134651184, | |
| "ranking_idealized_expo": 0.5208333134651184, | |
| "ranking_simple": 0.512499988079071, | |
| "regularize": 650.8515625, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.8341993386868207, | |
| "eval_dpo_loss": 1152.4822998046875, | |
| "eval_logits": -0.6463432908058167, | |
| "eval_logps": -81.1036148071289, | |
| "eval_loss": 2235.325439453125, | |
| "eval_objective": 2277.773681640625, | |
| "eval_ranking_idealized": 0.5092975497245789, | |
| "eval_ranking_idealized_expo": 0.5092975497245789, | |
| "eval_ranking_simple": 0.5082644820213318, | |
| "eval_regularize": 2277.773681640625, | |
| "eval_runtime": 262.0167, | |
| "eval_samples_per_second": 22.098, | |
| "eval_steps_per_second": 0.924, | |
| "step": 500 | |
| }, | |
| { | |
| "dpo_loss": 277.9737854003906, | |
| "epoch": 2.862541332073689, | |
| "grad_norm": 57899.221094649656, | |
| "learning_rate": 2.2919912390530945e-06, | |
| "logits": -0.5871773362159729, | |
| "logps": -76.18864440917969, | |
| "loss": 671.9541, | |
| "objective": 656.1866455078125, | |
| "ranking_idealized": 0.5083333253860474, | |
| "ranking_idealized_expo": 0.5083333253860474, | |
| "ranking_simple": 0.5249999761581421, | |
| "regularize": 656.1866455078125, | |
| "step": 505 | |
| }, | |
| { | |
| "dpo_loss": 308.4158935546875, | |
| "epoch": 2.8908833254605573, | |
| "grad_norm": 61442.9981374047, | |
| "learning_rate": 2.242624096740164e-06, | |
| "logits": -0.6056591272354126, | |
| "logps": -76.62039947509766, | |
| "loss": 651.3555, | |
| "objective": 650.9279174804688, | |
| "ranking_idealized": 0.5249999761581421, | |
| "ranking_idealized_expo": 0.5249999761581421, | |
| "ranking_simple": 0.5375000238418579, | |
| "regularize": 650.9279174804688, | |
| "step": 510 | |
| }, | |
| { | |
| "dpo_loss": 333.2612609863281, | |
| "epoch": 2.9192253188474258, | |
| "grad_norm": 65234.87039662293, | |
| "learning_rate": 2.193358192306384e-06, | |
| "logits": -0.677362322807312, | |
| "logps": -76.61473846435547, | |
| "loss": 644.7175, | |
| "objective": 633.654296875, | |
| "ranking_idealized": 0.4791666567325592, | |
| "ranking_idealized_expo": 0.4791666567325592, | |
| "ranking_simple": 0.4791666567325592, | |
| "regularize": 633.654296875, | |
| "step": 515 | |
| }, | |
| { | |
| "dpo_loss": 290.90093994140625, | |
| "epoch": 2.947567312234294, | |
| "grad_norm": 62487.68797546038, | |
| "learning_rate": 2.1442129043167877e-06, | |
| "logits": -0.6403795480728149, | |
| "logps": -78.69046020507812, | |
| "loss": 617.916, | |
| "objective": 615.7841186523438, | |
| "ranking_idealized": 0.5083333253860474, | |
| "ranking_idealized_expo": 0.5083333253860474, | |
| "ranking_simple": 0.5041666626930237, | |
| "regularize": 615.7841186523438, | |
| "step": 520 | |
| }, | |
| { | |
| "dpo_loss": 285.1398010253906, | |
| "epoch": 2.975909305621162, | |
| "grad_norm": 57603.254740385855, | |
| "learning_rate": 2.0952075638923656e-06, | |
| "logits": -0.631269633769989, | |
| "logps": -77.06267547607422, | |
| "loss": 600.514, | |
| "objective": 611.5770874023438, | |
| "ranking_idealized": 0.5666666626930237, | |
| "ranking_idealized_expo": 0.5666666626930237, | |
| "ranking_simple": 0.5541666746139526, | |
| "regularize": 611.5770874023438, | |
| "step": 525 | |
| }, | |
| { | |
| "dpo_loss": 257.36700439453125, | |
| "epoch": 3.0042512990080303, | |
| "grad_norm": 58200.63176299685, | |
| "learning_rate": 2.046361447106244e-06, | |
| "logits": -0.58321613073349, | |
| "logps": -76.48860168457031, | |
| "loss": 561.4095, | |
| "objective": 531.5440063476562, | |
| "ranking_idealized": 0.512499988079071, | |
| "ranking_idealized_expo": 0.512499988079071, | |
| "ranking_simple": 0.5083333253860474, | |
| "regularize": 531.5440063476562, | |
| "step": 530 | |
| }, | |
| { | |
| "dpo_loss": 268.19036865234375, | |
| "epoch": 3.0325932923948984, | |
| "grad_norm": 56509.28719745372, | |
| "learning_rate": 1.997693767401503e-06, | |
| "logits": -0.5703064203262329, | |
| "logps": -78.15198516845703, | |
| "loss": 537.4105, | |
| "objective": 540.3373413085938, | |
| "ranking_idealized": 0.5416666865348816, | |
| "ranking_idealized_expo": 0.5416666865348816, | |
| "ranking_simple": 0.5375000238418579, | |
| "regularize": 540.3373413085938, | |
| "step": 535 | |
| }, | |
| { | |
| "dpo_loss": 260.3810119628906, | |
| "epoch": 3.0609352857817664, | |
| "grad_norm": 56574.095934264005, | |
| "learning_rate": 1.9492236680336486e-06, | |
| "logits": -0.6156631112098694, | |
| "logps": -76.84761810302734, | |
| "loss": 528.5064, | |
| "objective": 518.1077880859375, | |
| "ranking_idealized": 0.5333333611488342, | |
| "ranking_idealized_expo": 0.5333333611488342, | |
| "ranking_simple": 0.550000011920929, | |
| "regularize": 518.1077880859375, | |
| "step": 540 | |
| }, | |
| { | |
| "dpo_loss": 294.6708679199219, | |
| "epoch": 3.089277279168635, | |
| "grad_norm": 57329.00701868531, | |
| "learning_rate": 1.9009702145406728e-06, | |
| "logits": -0.6262565851211548, | |
| "logps": -77.80619812011719, | |
| "loss": 534.8658, | |
| "objective": 547.5757446289062, | |
| "ranking_idealized": 0.49166667461395264, | |
| "ranking_idealized_expo": 0.49166667461395264, | |
| "ranking_simple": 0.4958333373069763, | |
| "regularize": 547.5757446289062, | |
| "step": 545 | |
| }, | |
| { | |
| "dpo_loss": 304.8855285644531, | |
| "epoch": 3.117619272555503, | |
| "grad_norm": 58804.14825774506, | |
| "learning_rate": 1.852952387243698e-06, | |
| "logits": -0.566185712814331, | |
| "logps": -78.45851135253906, | |
| "loss": 547.2687, | |
| "objective": 564.5450439453125, | |
| "ranking_idealized": 0.5375000238418579, | |
| "ranking_idealized_expo": 0.5375000238418579, | |
| "ranking_simple": 0.5458333492279053, | |
| "regularize": 564.5450439453125, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 3.117619272555503, | |
| "eval_dpo_loss": 1133.89697265625, | |
| "eval_logits": -0.662289023399353, | |
| "eval_logps": -81.35185241699219, | |
| "eval_loss": 2247.691650390625, | |
| "eval_objective": 2265.5048828125, | |
| "eval_ranking_idealized": 0.5092975497245789, | |
| "eval_ranking_idealized_expo": 0.5092975497245789, | |
| "eval_ranking_simple": 0.5144628286361694, | |
| "eval_regularize": 2265.5048828125, | |
| "eval_runtime": 259.4387, | |
| "eval_samples_per_second": 22.317, | |
| "eval_steps_per_second": 0.933, | |
| "step": 550 | |
| }, | |
| { | |
| "dpo_loss": 281.4118957519531, | |
| "epoch": 3.1459612659423715, | |
| "grad_norm": 59417.51403592292, | |
| "learning_rate": 1.8051890737811395e-06, | |
| "logits": -0.5252026915550232, | |
| "logps": -78.36255645751953, | |
| "loss": 532.1873, | |
| "objective": 550.6338500976562, | |
| "ranking_idealized": 0.550000011920929, | |
| "ranking_idealized_expo": 0.550000011920929, | |
| "ranking_simple": 0.5458333492279053, | |
| "regularize": 550.6338500976562, | |
| "step": 555 | |
| }, | |
| { | |
| "dpo_loss": 236.99884033203125, | |
| "epoch": 3.1743032593292395, | |
| "grad_norm": 59018.08983892326, | |
| "learning_rate": 1.7576990616793139e-06, | |
| "logits": -0.5617191195487976, | |
| "logps": -75.77594757080078, | |
| "loss": 505.1167, | |
| "objective": 438.2338562011719, | |
| "ranking_idealized": 0.5041666626930237, | |
| "ranking_idealized_expo": 0.5041666626930237, | |
| "ranking_simple": 0.5, | |
| "regularize": 438.2338562011719, | |
| "step": 560 | |
| }, | |
| { | |
| "dpo_loss": 246.71607971191406, | |
| "epoch": 3.2026452527161076, | |
| "grad_norm": 58636.618971388525, | |
| "learning_rate": 1.7105010309624381e-06, | |
| "logits": -0.5584675669670105, | |
| "logps": -77.66993713378906, | |
| "loss": 493.3785, | |
| "objective": 472.7521667480469, | |
| "ranking_idealized": 0.5416666865348816, | |
| "ranking_idealized_expo": 0.5416666865348816, | |
| "ranking_simple": 0.5375000238418579, | |
| "regularize": 472.7521667480469, | |
| "step": 565 | |
| }, | |
| { | |
| "dpo_loss": 265.2983703613281, | |
| "epoch": 3.230987246102976, | |
| "grad_norm": 55898.06189925703, | |
| "learning_rate": 1.6636135468049122e-06, | |
| "logits": -0.5392119288444519, | |
| "logps": -77.28318786621094, | |
| "loss": 497.3006, | |
| "objective": 500.74493408203125, | |
| "ranking_idealized": 0.4791666567325592, | |
| "ranking_idealized_expo": 0.4791666567325592, | |
| "ranking_simple": 0.48750001192092896, | |
| "regularize": 500.74493408203125, | |
| "step": 570 | |
| }, | |
| { | |
| "dpo_loss": 231.4838104248047, | |
| "epoch": 3.259329239489844, | |
| "grad_norm": 55759.490082320495, | |
| "learning_rate": 1.617055052228768e-06, | |
| "logits": -0.6065218448638916, | |
| "logps": -77.56108856201172, | |
| "loss": 497.8869, | |
| "objective": 468.4706115722656, | |
| "ranking_idealized": 0.5, | |
| "ranking_idealized_expo": 0.5, | |
| "ranking_simple": 0.4958333373069763, | |
| "regularize": 468.4706115722656, | |
| "step": 575 | |
| }, | |
| { | |
| "dpo_loss": 225.2176055908203, | |
| "epoch": 3.287671232876712, | |
| "grad_norm": 61004.08082189617, | |
| "learning_rate": 1.5708438608491816e-06, | |
| "logits": -0.6214988827705383, | |
| "logps": -78.54485321044922, | |
| "loss": 473.5735, | |
| "objective": 457.38262939453125, | |
| "ranking_idealized": 0.5541666746139526, | |
| "ranking_idealized_expo": 0.5541666746139526, | |
| "ranking_simple": 0.5541666746139526, | |
| "regularize": 457.38262939453125, | |
| "step": 580 | |
| }, | |
| { | |
| "dpo_loss": 232.1999969482422, | |
| "epoch": 3.3160132262635806, | |
| "grad_norm": 60725.76092889549, | |
| "learning_rate": 1.524998149670871e-06, | |
| "logits": -0.6042133569717407, | |
| "logps": -79.63090515136719, | |
| "loss": 487.6306, | |
| "objective": 457.5376281738281, | |
| "ranking_idealized": 0.5291666388511658, | |
| "ranking_idealized_expo": 0.5291666388511658, | |
| "ranking_simple": 0.5249999761581421, | |
| "regularize": 457.5376281738281, | |
| "step": 585 | |
| }, | |
| { | |
| "dpo_loss": 272.1759033203125, | |
| "epoch": 3.3443552196504487, | |
| "grad_norm": 56118.05633186447, | |
| "learning_rate": 1.479535951938243e-06, | |
| "logits": -0.6058977246284485, | |
| "logps": -79.4846420288086, | |
| "loss": 480.4184, | |
| "objective": 515.1324462890625, | |
| "ranking_idealized": 0.48750001192092896, | |
| "ranking_idealized_expo": 0.48750001192092896, | |
| "ranking_simple": 0.48750001192092896, | |
| "regularize": 515.1324462890625, | |
| "step": 590 | |
| }, | |
| { | |
| "dpo_loss": 245.82069396972656, | |
| "epoch": 3.372697213037317, | |
| "grad_norm": 55867.37155318488, | |
| "learning_rate": 1.43447515004208e-06, | |
| "logits": -0.5171046853065491, | |
| "logps": -79.08625030517578, | |
| "loss": 482.6767, | |
| "objective": 483.4396667480469, | |
| "ranking_idealized": 0.5249999761581421, | |
| "ranking_idealized_expo": 0.5249999761581421, | |
| "ranking_simple": 0.5291666388511658, | |
| "regularize": 483.4396667480469, | |
| "step": 595 | |
| }, | |
| { | |
| "dpo_loss": 196.64842224121094, | |
| "epoch": 3.4010392064241852, | |
| "grad_norm": 56516.07590190884, | |
| "learning_rate": 1.3898334684855647e-06, | |
| "logits": -0.5405918955802917, | |
| "logps": -78.04920959472656, | |
| "loss": 451.9043, | |
| "objective": 474.75640869140625, | |
| "ranking_idealized": 0.5041666626930237, | |
| "ranking_idealized_expo": 0.5041666626930237, | |
| "ranking_simple": 0.512499988079071, | |
| "regularize": 474.75640869140625, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 3.4010392064241852, | |
| "eval_dpo_loss": 1143.4464111328125, | |
| "eval_logits": -0.6081129908561707, | |
| "eval_logps": -81.8093032836914, | |
| "eval_loss": 2235.049072265625, | |
| "eval_objective": 2263.895751953125, | |
| "eval_ranking_idealized": 0.5092975497245789, | |
| "eval_ranking_idealized_expo": 0.5092975497245789, | |
| "eval_ranking_simple": 0.5113636255264282, | |
| "eval_regularize": 2263.895751953125, | |
| "eval_runtime": 263.0772, | |
| "eval_samples_per_second": 22.009, | |
| "eval_steps_per_second": 0.92, | |
| "step": 600 | |
| }, | |
| { | |
| "dpo_loss": 200.47567749023438, | |
| "epoch": 3.4293811998110533, | |
| "grad_norm": 61245.11887519878, | |
| "learning_rate": 1.3456284669124159e-06, | |
| "logits": -0.5445960164070129, | |
| "logps": -80.08841705322266, | |
| "loss": 443.7833, | |
| "objective": 392.1485900878906, | |
| "ranking_idealized": 0.5333333611488342, | |
| "ranking_idealized_expo": 0.5333333611488342, | |
| "ranking_simple": 0.5375000238418579, | |
| "regularize": 392.1485900878906, | |
| "step": 605 | |
| }, | |
| { | |
| "dpo_loss": 224.0120086669922, | |
| "epoch": 3.4577231931979218, | |
| "grad_norm": 55703.47222096951, | |
| "learning_rate": 1.301877533199859e-06, | |
| "logits": -0.5698374509811401, | |
| "logps": -78.10750579833984, | |
| "loss": 453.1188, | |
| "objective": 450.0478210449219, | |
| "ranking_idealized": 0.512499988079071, | |
| "ranking_idealized_expo": 0.512499988079071, | |
| "ranking_simple": 0.512499988079071, | |
| "regularize": 450.0478210449219, | |
| "step": 610 | |
| }, | |
| { | |
| "dpo_loss": 201.32762145996094, | |
| "epoch": 3.48606518658479, | |
| "grad_norm": 56298.35092268398, | |
| "learning_rate": 1.2585978766191726e-06, | |
| "logits": -0.6067613959312439, | |
| "logps": -78.46564483642578, | |
| "loss": 443.3495, | |
| "objective": 431.0823059082031, | |
| "ranking_idealized": 0.46666666865348816, | |
| "ranking_idealized_expo": 0.46666666865348816, | |
| "ranking_simple": 0.47083333134651184, | |
| "regularize": 431.0823059082031, | |
| "step": 615 | |
| }, | |
| { | |
| "dpo_loss": 206.5738983154297, | |
| "epoch": 3.514407179971658, | |
| "grad_norm": 63836.369713544904, | |
| "learning_rate": 1.2158065210664848e-06, | |
| "logits": -0.5364049077033997, | |
| "logps": -77.92042541503906, | |
| "loss": 413.5914, | |
| "objective": 407.43011474609375, | |
| "ranking_idealized": 0.4958333373069763, | |
| "ranking_idealized_expo": 0.4958333373069763, | |
| "ranking_simple": 0.4833333194255829, | |
| "regularize": 407.43011474609375, | |
| "step": 620 | |
| }, | |
| { | |
| "dpo_loss": 262.5015869140625, | |
| "epoch": 3.5427491733585263, | |
| "grad_norm": 57117.17093128659, | |
| "learning_rate": 1.1735202983664803e-06, | |
| "logits": -0.5198719501495361, | |
| "logps": -77.65792083740234, | |
| "loss": 434.5029, | |
| "objective": 441.78729248046875, | |
| "ranking_idealized": 0.5208333134651184, | |
| "ranking_idealized_expo": 0.5208333134651184, | |
| "ranking_simple": 0.5291666388511658, | |
| "regularize": 441.78729248046875, | |
| "step": 625 | |
| }, | |
| { | |
| "dpo_loss": 175.2751007080078, | |
| "epoch": 3.5710911667453944, | |
| "grad_norm": 57300.8422558694, | |
| "learning_rate": 1.1317558416516696e-06, | |
| "logits": -0.5618345737457275, | |
| "logps": -77.11524200439453, | |
| "loss": 418.5029, | |
| "objective": 366.7109375, | |
| "ranking_idealized": 0.512499988079071, | |
| "ranking_idealized_expo": 0.512499988079071, | |
| "ranking_simple": 0.5166666507720947, | |
| "regularize": 366.7109375, | |
| "step": 630 | |
| }, | |
| { | |
| "dpo_loss": 247.67637634277344, | |
| "epoch": 3.5994331601322624, | |
| "grad_norm": 57342.11408252111, | |
| "learning_rate": 1.0905295788197993e-06, | |
| "logits": -0.5630912184715271, | |
| "logps": -77.43748474121094, | |
| "loss": 390.8204, | |
| "objective": 448.0376281738281, | |
| "ranking_idealized": 0.48750001192092896, | |
| "ranking_idealized_expo": 0.48750001192092896, | |
| "ranking_simple": 0.4749999940395355, | |
| "regularize": 448.0376281738281, | |
| "step": 635 | |
| }, | |
| { | |
| "dpo_loss": 221.48199462890625, | |
| "epoch": 3.627775153519131, | |
| "grad_norm": 57029.256971048184, | |
| "learning_rate": 1.049857726072005e-06, | |
| "logits": -0.5471541285514832, | |
| "logps": -77.80196380615234, | |
| "loss": 394.0431, | |
| "objective": 420.2522888183594, | |
| "ranking_idealized": 0.4625000059604645, | |
| "ranking_idealized_expo": 0.4625000059604645, | |
| "ranking_simple": 0.48750001192092896, | |
| "regularize": 420.2522888183594, | |
| "step": 640 | |
| }, | |
| { | |
| "dpo_loss": 210.2812957763672, | |
| "epoch": 3.656117146905999, | |
| "grad_norm": 56760.30754643641, | |
| "learning_rate": 1.0097562815342215e-06, | |
| "logits": -0.5718862414360046, | |
| "logps": -76.699951171875, | |
| "loss": 382.7592, | |
| "objective": 373.0257263183594, | |
| "ranking_idealized": 0.5249999761581421, | |
| "ranking_idealized_expo": 0.5249999761581421, | |
| "ranking_simple": 0.5166666507720947, | |
| "regularize": 373.0257263183594, | |
| "step": 645 | |
| }, | |
| { | |
| "dpo_loss": 190.9052276611328, | |
| "epoch": 3.6844591402928675, | |
| "grad_norm": 61051.71098872104, | |
| "learning_rate": 9.702410189643838e-07, | |
| "logits": -0.5754767060279846, | |
| "logps": -77.90436553955078, | |
| "loss": 383.0005, | |
| "objective": 371.94171142578125, | |
| "ranking_idealized": 0.5041666626930237, | |
| "ranking_idealized_expo": 0.5041666626930237, | |
| "ranking_simple": 0.48750001192092896, | |
| "regularize": 371.94171142578125, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 3.6844591402928675, | |
| "eval_dpo_loss": 1148.96923828125, | |
| "eval_logits": -0.6417236328125, | |
| "eval_logps": -81.60212707519531, | |
| "eval_loss": 2233.306640625, | |
| "eval_objective": 2277.599365234375, | |
| "eval_ranking_idealized": 0.5092975497245789, | |
| "eval_ranking_idealized_expo": 0.5092975497245789, | |
| "eval_ranking_simple": 0.5123966932296753, | |
| "eval_regularize": 2277.599365234375, | |
| "eval_runtime": 259.78, | |
| "eval_samples_per_second": 22.288, | |
| "eval_steps_per_second": 0.932, | |
| "step": 650 | |
| }, | |
| { | |
| "dpo_loss": 196.76412963867188, | |
| "epoch": 3.7128011336797355, | |
| "grad_norm": 58071.414511462746, | |
| "learning_rate": 9.313274815478698e-07, | |
| "logits": -0.56723552942276, | |
| "logps": -76.70596313476562, | |
| "loss": 379.5748, | |
| "objective": 366.3069763183594, | |
| "ranking_idealized": 0.5208333134651184, | |
| "ranking_idealized_expo": 0.5208333134651184, | |
| "ranking_simple": 0.512499988079071, | |
| "regularize": 366.3069763183594, | |
| "step": 655 | |
| }, | |
| { | |
| "dpo_loss": 194.60183715820312, | |
| "epoch": 3.7411431270666036, | |
| "grad_norm": 60809.553574292404, | |
| "learning_rate": 8.930309757836517e-07, | |
| "logits": -0.6155992150306702, | |
| "logps": -77.6422348022461, | |
| "loss": 361.61, | |
| "objective": 359.7954406738281, | |
| "ranking_idealized": 0.5541666746139526, | |
| "ranking_idealized_expo": 0.5541666746139526, | |
| "ranking_simple": 0.5541666746139526, | |
| "regularize": 359.7954406738281, | |
| "step": 660 | |
| }, | |
| { | |
| "dpo_loss": 192.55584716796875, | |
| "epoch": 3.769485120453472, | |
| "grad_norm": 59289.71674690953, | |
| "learning_rate": 8.553665654635343e-07, | |
| "logits": -0.6106492280960083, | |
| "logps": -78.0013198852539, | |
| "loss": 360.7243, | |
| "objective": 351.4775695800781, | |
| "ranking_idealized": 0.5166666507720947, | |
| "ranking_idealized_expo": 0.5166666507720947, | |
| "ranking_simple": 0.5041666626930237, | |
| "regularize": 351.4775695800781, | |
| "step": 665 | |
| }, | |
| { | |
| "dpo_loss": 144.13723754882812, | |
| "epoch": 3.79782711384034, | |
| "grad_norm": 57262.66850621469, | |
| "learning_rate": 8.183490657468687e-07, | |
| "logits": -0.6628882884979248, | |
| "logps": -77.76026916503906, | |
| "loss": 349.6578, | |
| "objective": 297.4377746582031, | |
| "ranking_idealized": 0.5916666388511658, | |
| "ranking_idealized_expo": 0.5916666388511658, | |
| "ranking_simple": 0.5833333134651184, | |
| "regularize": 297.4377746582031, | |
| "step": 670 | |
| }, | |
| { | |
| "dpo_loss": 152.4486846923828, | |
| "epoch": 3.826169107227208, | |
| "grad_norm": 56825.12670236396, | |
| "learning_rate": 7.819930373330669e-07, | |
| "logits": -0.6416188478469849, | |
| "logps": -76.90306091308594, | |
| "loss": 353.6488, | |
| "objective": 341.40216064453125, | |
| "ranking_idealized": 0.5375000238418579, | |
| "ranking_idealized_expo": 0.5375000238418579, | |
| "ranking_simple": 0.550000011920929, | |
| "regularize": 341.40216064453125, | |
| "step": 675 | |
| }, | |
| { | |
| "dpo_loss": 184.26919555664062, | |
| "epoch": 3.8545111006140766, | |
| "grad_norm": 53705.32969410845, | |
| "learning_rate": 7.463127807341966e-07, | |
| "logits": -0.5924732089042664, | |
| "logps": -77.391357421875, | |
| "loss": 333.8282, | |
| "objective": 344.5659484863281, | |
| "ranking_idealized": 0.5083333253860474, | |
| "ranking_idealized_expo": 0.5083333253860474, | |
| "ranking_simple": 0.5, | |
| "regularize": 344.5659484863281, | |
| "step": 680 | |
| }, | |
| { | |
| "dpo_loss": 148.51840209960938, | |
| "epoch": 3.8828530940009447, | |
| "grad_norm": 59401.195526544805, | |
| "learning_rate": 7.113223306499336e-07, | |
| "logits": -0.59928297996521, | |
| "logps": -77.14978790283203, | |
| "loss": 326.404, | |
| "objective": 304.36529541015625, | |
| "ranking_idealized": 0.5166666507720947, | |
| "ranking_idealized_expo": 0.5166666507720947, | |
| "ranking_simple": 0.5208333134651184, | |
| "regularize": 304.36529541015625, | |
| "step": 685 | |
| }, | |
| { | |
| "dpo_loss": 139.26687622070312, | |
| "epoch": 3.9111950873878127, | |
| "grad_norm": 55908.406363835, | |
| "learning_rate": 6.770354504470575e-07, | |
| "logits": -0.6397600769996643, | |
| "logps": -77.13099670410156, | |
| "loss": 327.1117, | |
| "objective": 284.6099853515625, | |
| "ranking_idealized": 0.5458333492279053, | |
| "ranking_idealized_expo": 0.5458333492279053, | |
| "ranking_simple": 0.5541666746139526, | |
| "regularize": 284.6099853515625, | |
| "step": 690 | |
| }, | |
| { | |
| "dpo_loss": 149.27659606933594, | |
| "epoch": 3.9395370807746812, | |
| "grad_norm": 56202.2834304956, | |
| "learning_rate": 6.434656267456843e-07, | |
| "logits": -0.5855825543403625, | |
| "logps": -78.09581756591797, | |
| "loss": 313.6309, | |
| "objective": 333.0680847167969, | |
| "ranking_idealized": 0.46666666865348816, | |
| "ranking_idealized_expo": 0.46666666865348816, | |
| "ranking_simple": 0.46666666865348816, | |
| "regularize": 333.0680847167969, | |
| "step": 695 | |
| }, | |
| { | |
| "dpo_loss": 192.27964782714844, | |
| "epoch": 3.9678790741615493, | |
| "grad_norm": 54921.578630640375, | |
| "learning_rate": 6.106260641143547e-07, | |
| "logits": -0.5845039486885071, | |
| "logps": -77.96537780761719, | |
| "loss": 316.0834, | |
| "objective": 346.9200134277344, | |
| "ranking_idealized": 0.4833333194255829, | |
| "ranking_idealized_expo": 0.4833333194255829, | |
| "ranking_simple": 0.4833333194255829, | |
| "regularize": 346.9200134277344, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 3.9678790741615493, | |
| "eval_dpo_loss": 1143.5379638671875, | |
| "eval_logits": -0.6440603733062744, | |
| "eval_logps": -82.07394409179688, | |
| "eval_loss": 2236.5556640625, | |
| "eval_objective": 2269.068115234375, | |
| "eval_ranking_idealized": 0.5092975497245789, | |
| "eval_ranking_idealized_expo": 0.5092975497245789, | |
| "eval_ranking_simple": 0.5134297609329224, | |
| "eval_regularize": 2269.068115234375, | |
| "eval_runtime": 260.028, | |
| "eval_samples_per_second": 22.267, | |
| "eval_steps_per_second": 0.931, | |
| "step": 700 | |
| }, | |
| { | |
| "dpo_loss": 133.92913818359375, | |
| "epoch": 3.9962210675484178, | |
| "grad_norm": 53821.890841527515, | |
| "learning_rate": 5.785296798760601e-07, | |
| "logits": -0.5399293303489685, | |
| "logps": -76.87466430664062, | |
| "loss": 297.6826, | |
| "objective": 305.3105163574219, | |
| "ranking_idealized": 0.5249999761581421, | |
| "ranking_idealized_expo": 0.5249999761581421, | |
| "ranking_simple": 0.5333333611488342, | |
| "regularize": 305.3105163574219, | |
| "step": 705 | |
| }, | |
| { | |
| "dpo_loss": 136.2448272705078, | |
| "epoch": 4.024563060935286, | |
| "grad_norm": 56556.65117729612, | |
| "learning_rate": 5.471890990272666e-07, | |
| "logits": -0.5713457465171814, | |
| "logps": -78.32731628417969, | |
| "loss": 257.5077, | |
| "objective": 262.51849365234375, | |
| "ranking_idealized": 0.4791666567325592, | |
| "ranking_idealized_expo": 0.4791666567325592, | |
| "ranking_simple": 0.4791666567325592, | |
| "regularize": 262.51849365234375, | |
| "step": 710 | |
| }, | |
| { | |
| "dpo_loss": 146.4274444580078, | |
| "epoch": 4.052905054322154, | |
| "grad_norm": 56529.950131775324, | |
| "learning_rate": 5.166166492719124e-07, | |
| "logits": -0.5606953501701355, | |
| "logps": -77.65556335449219, | |
| "loss": 260.5169, | |
| "objective": 290.5565490722656, | |
| "ranking_idealized": 0.5333333611488342, | |
| "ranking_idealized_expo": 0.5333333611488342, | |
| "ranking_simple": 0.5416666865348816, | |
| "regularize": 290.5565490722656, | |
| "step": 715 | |
| }, | |
| { | |
| "dpo_loss": 120.44002532958984, | |
| "epoch": 4.081247047709022, | |
| "grad_norm": 57683.14038076662, | |
| "learning_rate": 4.868243561723535e-07, | |
| "logits": -0.5295476913452148, | |
| "logps": -79.97325134277344, | |
| "loss": 242.1283, | |
| "objective": 255.8209991455078, | |
| "ranking_idealized": 0.5208333134651184, | |
| "ranking_idealized_expo": 0.5208333134651184, | |
| "ranking_simple": 0.5166666507720947, | |
| "regularize": 255.8209991455078, | |
| "step": 720 | |
| }, | |
| { | |
| "dpo_loss": 118.35858917236328, | |
| "epoch": 4.109589041095891, | |
| "grad_norm": 55672.19278003207, | |
| "learning_rate": 4.57823938419153e-07, | |
| "logits": -0.5605884790420532, | |
| "logps": -77.58100128173828, | |
| "loss": 248.1455, | |
| "objective": 227.64146423339844, | |
| "ranking_idealized": 0.5291666388511658, | |
| "ranking_idealized_expo": 0.5291666388511658, | |
| "ranking_simple": 0.5291666388511658, | |
| "regularize": 227.64146423339844, | |
| "step": 725 | |
| }, | |
| { | |
| "dpo_loss": 126.79460906982422, | |
| "epoch": 4.137931034482759, | |
| "grad_norm": 54558.816031980285, | |
| "learning_rate": 4.2962680322157335e-07, | |
| "logits": -0.6369396448135376, | |
| "logps": -77.99376678466797, | |
| "loss": 251.7915, | |
| "objective": 234.14012145996094, | |
| "ranking_idealized": 0.5375000238418579, | |
| "ranking_idealized_expo": 0.5375000238418579, | |
| "ranking_simple": 0.5375000238418579, | |
| "regularize": 234.14012145996094, | |
| "step": 730 | |
| }, | |
| { | |
| "dpo_loss": 105.81147766113281, | |
| "epoch": 4.166273027869627, | |
| "grad_norm": 57552.01210460219, | |
| "learning_rate": 4.0224404182059443e-07, | |
| "logits": -0.556169867515564, | |
| "logps": -79.20640563964844, | |
| "loss": 244.4964, | |
| "objective": 241.80613708496094, | |
| "ranking_idealized": 0.5249999761581421, | |
| "ranking_idealized_expo": 0.5249999761581421, | |
| "ranking_simple": 0.5291666388511658, | |
| "regularize": 241.80613708496094, | |
| "step": 735 | |
| }, | |
| { | |
| "dpo_loss": 122.31700897216797, | |
| "epoch": 4.194615021256495, | |
| "grad_norm": 55106.91910179694, | |
| "learning_rate": 3.756864251262143e-07, | |
| "logits": -0.5948054790496826, | |
| "logps": -78.53282928466797, | |
| "loss": 247.5535, | |
| "objective": 229.94223022460938, | |
| "ranking_idealized": 0.5208333134651184, | |
| "ranking_idealized_expo": 0.5208333134651184, | |
| "ranking_simple": 0.5249999761581421, | |
| "regularize": 229.94223022460938, | |
| "step": 740 | |
| }, | |
| { | |
| "dpo_loss": 112.44102478027344, | |
| "epoch": 4.222957014643363, | |
| "grad_norm": 56632.58905444662, | |
| "learning_rate": 3.499643994807486e-07, | |
| "logits": -0.6257311105728149, | |
| "logps": -76.18108367919922, | |
| "loss": 229.3078, | |
| "objective": 219.73800659179688, | |
| "ranking_idealized": 0.48750001192092896, | |
| "ranking_idealized_expo": 0.48750001192092896, | |
| "ranking_simple": 0.4958333373069763, | |
| "regularize": 219.73800659179688, | |
| "step": 745 | |
| }, | |
| { | |
| "dpo_loss": 108.38746643066406, | |
| "epoch": 4.251299008030231, | |
| "grad_norm": 57857.865386387704, | |
| "learning_rate": 3.250880825498026e-07, | |
| "logits": -0.651103675365448, | |
| "logps": -78.09248352050781, | |
| "loss": 230.1662, | |
| "objective": 248.08090209960938, | |
| "ranking_idealized": 0.5249999761581421, | |
| "ranking_idealized_expo": 0.5249999761581421, | |
| "ranking_simple": 0.5291666388511658, | |
| "regularize": 248.08090209960938, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 4.251299008030231, | |
| "eval_dpo_loss": 1146.1785888671875, | |
| "eval_logits": -0.6514174342155457, | |
| "eval_logps": -82.18943786621094, | |
| "eval_loss": 2241.186279296875, | |
| "eval_objective": 2272.541748046875, | |
| "eval_ranking_idealized": 0.5092975497245789, | |
| "eval_ranking_idealized_expo": 0.5092975497245789, | |
| "eval_ranking_simple": 0.5123966932296753, | |
| "eval_regularize": 2272.541748046875, | |
| "eval_runtime": 260.6171, | |
| "eval_samples_per_second": 22.217, | |
| "eval_steps_per_second": 0.929, | |
| "step": 750 | |
| }, | |
| { | |
| "dpo_loss": 94.8583755493164, | |
| "epoch": 4.2796410014171, | |
| "grad_norm": 57171.85697996976, | |
| "learning_rate": 3.0106725934252095e-07, | |
| "logits": -0.617469847202301, | |
| "logps": -77.90315246582031, | |
| "loss": 235.9205, | |
| "objective": 197.84715270996094, | |
| "ranking_idealized": 0.5333333611488342, | |
| "ranking_idealized_expo": 0.5333333611488342, | |
| "ranking_simple": 0.5333333611488342, | |
| "regularize": 197.84715270996094, | |
| "step": 755 | |
| }, | |
| { | |
| "dpo_loss": 123.18533325195312, | |
| "epoch": 4.307982994803968, | |
| "grad_norm": 54580.774915438706, | |
| "learning_rate": 2.779113783626916e-07, | |
| "logits": -0.6015483736991882, | |
| "logps": -77.99730682373047, | |
| "loss": 219.0399, | |
| "objective": 203.16455078125, | |
| "ranking_idealized": 0.5625, | |
| "ranking_idealized_expo": 0.5625, | |
| "ranking_simple": 0.5541666746139526, | |
| "regularize": 203.16455078125, | |
| "step": 760 | |
| }, | |
| { | |
| "dpo_loss": 110.12825012207031, | |
| "epoch": 4.336324988190836, | |
| "grad_norm": 59828.89205734499, | |
| "learning_rate": 2.5562954789221164e-07, | |
| "logits": -0.6103938817977905, | |
| "logps": -78.54331970214844, | |
| "loss": 228.1548, | |
| "objective": 236.6626739501953, | |
| "ranking_idealized": 0.5249999761581421, | |
| "ranking_idealized_expo": 0.5249999761581421, | |
| "ranking_simple": 0.5249999761581421, | |
| "regularize": 236.6626739501953, | |
| "step": 765 | |
| }, | |
| { | |
| "dpo_loss": 123.61617279052734, | |
| "epoch": 4.364666981577704, | |
| "grad_norm": 55023.44732573502, | |
| "learning_rate": 2.3423053240837518e-07, | |
| "logits": -0.5958681702613831, | |
| "logps": -77.1136703491211, | |
| "loss": 230.3628, | |
| "objective": 218.6300811767578, | |
| "ranking_idealized": 0.5041666626930237, | |
| "ranking_idealized_expo": 0.5041666626930237, | |
| "ranking_simple": 0.512499988079071, | |
| "regularize": 218.6300811767578, | |
| "step": 770 | |
| }, | |
| { | |
| "dpo_loss": 98.90251159667969, | |
| "epoch": 4.393008974964572, | |
| "grad_norm": 54732.23444299214, | |
| "learning_rate": 2.137227491364016e-07, | |
| "logits": -0.5972557663917542, | |
| "logps": -78.07060241699219, | |
| "loss": 216.1991, | |
| "objective": 207.3569793701172, | |
| "ranking_idealized": 0.5083333253860474, | |
| "ranking_idealized_expo": 0.5083333253860474, | |
| "ranking_simple": 0.5166666507720947, | |
| "regularize": 207.3569793701172, | |
| "step": 775 | |
| }, | |
| { | |
| "dpo_loss": 109.3499526977539, | |
| "epoch": 4.42135096835144, | |
| "grad_norm": 54081.47665517887, | |
| "learning_rate": 1.941142647385469e-07, | |
| "logits": -0.6113469004631042, | |
| "logps": -77.00961303710938, | |
| "loss": 210.2967, | |
| "objective": 211.32273864746094, | |
| "ranking_idealized": 0.5, | |
| "ranking_idealized_expo": 0.5, | |
| "ranking_simple": 0.5083333253860474, | |
| "regularize": 211.32273864746094, | |
| "step": 780 | |
| }, | |
| { | |
| "dpo_loss": 129.6833953857422, | |
| "epoch": 4.449692961738309, | |
| "grad_norm": 56904.662928690464, | |
| "learning_rate": 1.7541279214111277e-07, | |
| "logits": -0.6509999632835388, | |
| "logps": -76.9924087524414, | |
| "loss": 202.5833, | |
| "objective": 214.20053100585938, | |
| "ranking_idealized": 0.5291666388511658, | |
| "ranking_idealized_expo": 0.5291666388511658, | |
| "ranking_simple": 0.5249999761581421, | |
| "regularize": 214.20053100585938, | |
| "step": 785 | |
| }, | |
| { | |
| "dpo_loss": 105.77576446533203, | |
| "epoch": 4.478034955125177, | |
| "grad_norm": 59495.74269597329, | |
| "learning_rate": 1.5762568750059604e-07, | |
| "logits": -0.6297593712806702, | |
| "logps": -79.3594970703125, | |
| "loss": 213.9758, | |
| "objective": 195.52975463867188, | |
| "ranking_idealized": 0.5541666746139526, | |
| "ranking_idealized_expo": 0.5541666746139526, | |
| "ranking_simple": 0.5541666746139526, | |
| "regularize": 195.52975463867188, | |
| "step": 790 | |
| }, | |
| { | |
| "dpo_loss": 126.51469421386719, | |
| "epoch": 4.506376948512045, | |
| "grad_norm": 57426.690295786095, | |
| "learning_rate": 1.4075994731016895e-07, | |
| "logits": -0.5589895248413086, | |
| "logps": -78.96713256835938, | |
| "loss": 208.3141, | |
| "objective": 228.62591552734375, | |
| "ranking_idealized": 0.550000011920929, | |
| "ranking_idealized_expo": 0.550000011920929, | |
| "ranking_simple": 0.5541666746139526, | |
| "regularize": 228.62591552734375, | |
| "step": 795 | |
| }, | |
| { | |
| "dpo_loss": 99.8931655883789, | |
| "epoch": 4.534718941898913, | |
| "grad_norm": 54974.75424590611, | |
| "learning_rate": 1.2482220564763669e-07, | |
| "logits": -0.5342037677764893, | |
| "logps": -78.09777069091797, | |
| "loss": 198.8015, | |
| "objective": 194.4702911376953, | |
| "ranking_idealized": 0.5083333253860474, | |
| "ranking_idealized_expo": 0.5083333253860474, | |
| "ranking_simple": 0.5166666507720947, | |
| "regularize": 194.4702911376953, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 4.534718941898913, | |
| "eval_dpo_loss": 1141.6485595703125, | |
| "eval_logits": -0.6624540090560913, | |
| "eval_logps": -82.07605743408203, | |
| "eval_loss": 2236.1728515625, | |
| "eval_objective": 2266.98193359375, | |
| "eval_ranking_idealized": 0.5092975497245789, | |
| "eval_ranking_idealized_expo": 0.5092975497245789, | |
| "eval_ranking_simple": 0.5134297609329224, | |
| "eval_regularize": 2266.98193359375, | |
| "eval_runtime": 260.1873, | |
| "eval_samples_per_second": 22.253, | |
| "eval_steps_per_second": 0.93, | |
| "step": 800 | |
| }, | |
| { | |
| "dpo_loss": 98.73998260498047, | |
| "epoch": 4.563060935285781, | |
| "grad_norm": 57388.01438250827, | |
| "learning_rate": 1.0981873156594381e-07, | |
| "logits": -0.6012968420982361, | |
| "logps": -77.73635864257812, | |
| "loss": 187.937, | |
| "objective": 201.25303649902344, | |
| "ranking_idealized": 0.4583333432674408, | |
| "ranking_idealized_expo": 0.4583333432674408, | |
| "ranking_simple": 0.4625000059604645, | |
| "regularize": 201.25303649902344, | |
| "step": 805 | |
| }, | |
| { | |
| "dpo_loss": 111.82048797607422, | |
| "epoch": 4.59140292867265, | |
| "grad_norm": 55981.29882559336, | |
| "learning_rate": 9.575542662726756e-08, | |
| "logits": -0.6273016333580017, | |
| "logps": -76.8697738647461, | |
| "loss": 192.8168, | |
| "objective": 208.81210327148438, | |
| "ranking_idealized": 0.4583333432674408, | |
| "ranking_idealized_expo": 0.4583333432674408, | |
| "ranking_simple": 0.4583333432674408, | |
| "regularize": 208.81210327148438, | |
| "step": 810 | |
| }, | |
| { | |
| "dpo_loss": 91.50080871582031, | |
| "epoch": 4.619744922059518, | |
| "grad_norm": 56005.94901608643, | |
| "learning_rate": 8.26378225816582e-08, | |
| "logits": -0.5446544885635376, | |
| "logps": -78.59085083007812, | |
| "loss": 189.4629, | |
| "objective": 190.6603240966797, | |
| "ranking_idealized": 0.5874999761581421, | |
| "ranking_idealized_expo": 0.5874999761581421, | |
| "ranking_simple": 0.6000000238418579, | |
| "regularize": 190.6603240966797, | |
| "step": 815 | |
| }, | |
| { | |
| "dpo_loss": 85.91332244873047, | |
| "epoch": 4.648086915446386, | |
| "grad_norm": 57321.07122409185, | |
| "learning_rate": 7.047107919114588e-08, | |
| "logits": -0.6059623956680298, | |
| "logps": -77.87149047851562, | |
| "loss": 196.4597, | |
| "objective": 188.6878204345703, | |
| "ranking_idealized": 0.5625, | |
| "ranking_idealized_expo": 0.5625, | |
| "ranking_simple": 0.5625, | |
| "regularize": 188.6878204345703, | |
| "step": 820 | |
| }, | |
| { | |
| "dpo_loss": 100.03099060058594, | |
| "epoch": 4.6764289088332545, | |
| "grad_norm": 56144.87510056358, | |
| "learning_rate": 5.92599822001666e-08, | |
| "logits": -0.5770820379257202, | |
| "logps": -76.37511444091797, | |
| "loss": 190.9932, | |
| "objective": 200.11782836914062, | |
| "ranking_idealized": 0.5, | |
| "ranking_idealized_expo": 0.5, | |
| "ranking_simple": 0.5041666626930237, | |
| "regularize": 200.11782836914062, | |
| "step": 825 | |
| }, | |
| { | |
| "dpo_loss": 130.17276000976562, | |
| "epoch": 4.7047709022201225, | |
| "grad_norm": 56205.70525916846, | |
| "learning_rate": 4.9008941453107527e-08, | |
| "logits": -0.6312499642372131, | |
| "logps": -78.51829528808594, | |
| "loss": 199.7126, | |
| "objective": 203.79873657226562, | |
| "ranking_idealized": 0.512499988079071, | |
| "ranking_idealized_expo": 0.512499988079071, | |
| "ranking_simple": 0.5208333134651184, | |
| "regularize": 203.79873657226562, | |
| "step": 830 | |
| }, | |
| { | |
| "dpo_loss": 97.63799285888672, | |
| "epoch": 4.733112895606991, | |
| "grad_norm": 52585.06236070743, | |
| "learning_rate": 3.972198915970976e-08, | |
| "logits": -0.5862113833427429, | |
| "logps": -77.74220275878906, | |
| "loss": 193.6844, | |
| "objective": 188.87692260742188, | |
| "ranking_idealized": 0.48750001192092896, | |
| "ranking_idealized_expo": 0.48750001192092896, | |
| "ranking_simple": 0.4958333373069763, | |
| "regularize": 188.87692260742188, | |
| "step": 835 | |
| }, | |
| { | |
| "dpo_loss": 104.34432220458984, | |
| "epoch": 4.7614548889938595, | |
| "grad_norm": 54158.02776349709, | |
| "learning_rate": 3.1402778309014284e-08, | |
| "logits": -0.5883139967918396, | |
| "logps": -78.25198364257812, | |
| "loss": 182.0377, | |
| "objective": 192.16619873046875, | |
| "ranking_idealized": 0.5541666746139526, | |
| "ranking_idealized_expo": 0.5541666746139526, | |
| "ranking_simple": 0.5541666746139526, | |
| "regularize": 192.16619873046875, | |
| "step": 840 | |
| }, | |
| { | |
| "dpo_loss": 85.43553161621094, | |
| "epoch": 4.7897968823807275, | |
| "grad_norm": 55251.77271941104, | |
| "learning_rate": 2.4054581232470785e-08, | |
| "logits": -0.5956212878227234, | |
| "logps": -78.41481018066406, | |
| "loss": 182.2349, | |
| "objective": 189.16375732421875, | |
| "ranking_idealized": 0.4541666805744171, | |
| "ranking_idealized_expo": 0.4541666805744171, | |
| "ranking_simple": 0.4541666805744171, | |
| "regularize": 189.16375732421875, | |
| "step": 845 | |
| }, | |
| { | |
| "dpo_loss": 105.7170639038086, | |
| "epoch": 4.818138875767596, | |
| "grad_norm": 54494.83204693859, | |
| "learning_rate": 1.768028831677926e-08, | |
| "logits": -0.6094614863395691, | |
| "logps": -77.0250244140625, | |
| "loss": 189.8097, | |
| "objective": 175.89796447753906, | |
| "ranking_idealized": 0.5291666388511658, | |
| "ranking_idealized_expo": 0.5291666388511658, | |
| "ranking_simple": 0.5249999761581421, | |
| "regularize": 175.89796447753906, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 4.818138875767596, | |
| "eval_dpo_loss": 1141.93798828125, | |
| "eval_logits": -0.6598728895187378, | |
| "eval_logps": -82.0995101928711, | |
| "eval_loss": 2234.33984375, | |
| "eval_objective": 2266.176025390625, | |
| "eval_ranking_idealized": 0.5092975497245789, | |
| "eval_ranking_idealized_expo": 0.5092975497245789, | |
| "eval_ranking_simple": 0.5123966932296753, | |
| "eval_regularize": 2266.176025390625, | |
| "eval_runtime": 259.4687, | |
| "eval_samples_per_second": 22.315, | |
| "eval_steps_per_second": 0.933, | |
| "step": 850 | |
| }, | |
| { | |
| "dpo_loss": 105.15518188476562, | |
| "epoch": 4.846480869154464, | |
| "grad_norm": 53663.63564822801, | |
| "learning_rate": 1.2282406866966078e-08, | |
| "logits": -0.5728623270988464, | |
| "logps": -77.2236099243164, | |
| "loss": 189.0291, | |
| "objective": 193.78550720214844, | |
| "ranking_idealized": 0.5375000238418579, | |
| "ranking_idealized_expo": 0.5375000238418579, | |
| "ranking_simple": 0.5291666388511658, | |
| "regularize": 193.78550720214844, | |
| "step": 855 | |
| }, | |
| { | |
| "dpo_loss": 80.22781372070312, | |
| "epoch": 4.874822862541333, | |
| "grad_norm": 57241.34026283923, | |
| "learning_rate": 7.863060120144316e-09, | |
| "logits": -0.5664411783218384, | |
| "logps": -77.44783782958984, | |
| "loss": 191.157, | |
| "objective": 193.0424346923828, | |
| "ranking_idealized": 0.550000011920929, | |
| "ranking_idealized_expo": 0.550000011920929, | |
| "ranking_simple": 0.550000011920929, | |
| "regularize": 193.0424346923828, | |
| "step": 860 | |
| }, | |
| { | |
| "dpo_loss": 81.85896301269531, | |
| "epoch": 4.903164855928201, | |
| "grad_norm": 56266.32691569455, | |
| "learning_rate": 4.423986410346526e-09, | |
| "logits": -0.6029998064041138, | |
| "logps": -76.39881896972656, | |
| "loss": 175.239, | |
| "objective": 165.09597778320312, | |
| "ranking_idealized": 0.4958333373069763, | |
| "ranking_idealized_expo": 0.4958333373069763, | |
| "ranking_simple": 0.48750001192092896, | |
| "regularize": 165.09597778320312, | |
| "step": 865 | |
| }, | |
| { | |
| "dpo_loss": 81.69947814941406, | |
| "epoch": 4.931506849315069, | |
| "grad_norm": 56846.595860039386, | |
| "learning_rate": 1.9665384847583622e-09, | |
| "logits": -0.6089357733726501, | |
| "logps": -78.06834411621094, | |
| "loss": 181.1643, | |
| "objective": 200.40501403808594, | |
| "ranking_idealized": 0.5083333253860474, | |
| "ranking_idealized_expo": 0.5083333253860474, | |
| "ranking_simple": 0.512499988079071, | |
| "regularize": 200.40501403808594, | |
| "step": 870 | |
| }, | |
| { | |
| "dpo_loss": 107.40243530273438, | |
| "epoch": 4.959848842701937, | |
| "grad_norm": 59409.659983453515, | |
| "learning_rate": 4.916829716183901e-10, | |
| "logits": -0.5824072957038879, | |
| "logps": -77.70771026611328, | |
| "loss": 184.499, | |
| "objective": 193.38832092285156, | |
| "ranking_idealized": 0.5458333492279053, | |
| "ranking_idealized_expo": 0.5458333492279053, | |
| "ranking_simple": 0.5458333492279053, | |
| "regularize": 193.38832092285156, | |
| "step": 875 | |
| }, | |
| { | |
| "dpo_loss": 100.64095306396484, | |
| "epoch": 4.988190836088805, | |
| "grad_norm": 57665.05790656053, | |
| "learning_rate": 0.0, | |
| "logits": -0.6364601850509644, | |
| "logps": -78.62317657470703, | |
| "loss": 185.8119, | |
| "objective": 193.60719299316406, | |
| "ranking_idealized": 0.5291666388511658, | |
| "ranking_idealized_expo": 0.5291666388511658, | |
| "ranking_simple": 0.5208333134651184, | |
| "regularize": 193.60719299316406, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 4.988190836088805, | |
| "step": 880, | |
| "total_flos": 0.0, | |
| "train_loss": 743.9546574996276, | |
| "train_runtime": 35228.1764, | |
| "train_samples_per_second": 7.21, | |
| "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 | |
| } | |