| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.9765925925925925, | |
| "eval_steps": 500, | |
| "global_step": 315, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.047407407407407405, | |
| "grad_norm": 677.5082397460938, | |
| "learning_rate": 7.8125e-06, | |
| "log_odds_chosen": 4.972014427185059, | |
| "log_odds_ratio": -9.77397632598877, | |
| "logits/chosen": 138.50723266601562, | |
| "logits/rejected": 153.07931518554688, | |
| "logps/chosen": -20.706424713134766, | |
| "logps/rejected": -25.678659439086914, | |
| "loss": 295.6143, | |
| "nll_loss": 8.335786819458008, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -1.0353213548660278, | |
| "rewards/margins": 0.24861177802085876, | |
| "rewards/rejected": -1.2839329242706299, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.09481481481481481, | |
| "grad_norm": 424.99365234375, | |
| "learning_rate": 1.5625e-05, | |
| "log_odds_chosen": 2.362799644470215, | |
| "log_odds_ratio": -7.1902971267700195, | |
| "logits/chosen": 135.88121032714844, | |
| "logits/rejected": 160.3680419921875, | |
| "logps/chosen": -15.390497207641602, | |
| "logps/rejected": -17.75222396850586, | |
| "loss": 244.5906, | |
| "nll_loss": 7.254161834716797, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -0.7695248126983643, | |
| "rewards/margins": 0.1180863231420517, | |
| "rewards/rejected": -0.8876112103462219, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.14222222222222222, | |
| "grad_norm": 270.6084899902344, | |
| "learning_rate": 2.34375e-05, | |
| "log_odds_chosen": 4.370176792144775, | |
| "log_odds_ratio": -7.611123085021973, | |
| "logits/chosen": 126.235595703125, | |
| "logits/rejected": 153.06771850585938, | |
| "logps/chosen": -19.671573638916016, | |
| "logps/rejected": -24.04007339477539, | |
| "loss": 286.111, | |
| "nll_loss": 8.265220642089844, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.9835788011550903, | |
| "rewards/margins": 0.218424990773201, | |
| "rewards/rejected": -1.2020035982131958, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.18962962962962962, | |
| "grad_norm": 2947.77685546875, | |
| "learning_rate": 3.125e-05, | |
| "log_odds_chosen": -0.3353792726993561, | |
| "log_odds_ratio": -5.916926383972168, | |
| "logits/chosen": 159.31063842773438, | |
| "logits/rejected": 163.8131561279297, | |
| "logps/chosen": -13.794464111328125, | |
| "logps/rejected": -13.460695266723633, | |
| "loss": 223.2703, | |
| "nll_loss": 6.456460475921631, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -0.6897231936454773, | |
| "rewards/margins": -0.016688454896211624, | |
| "rewards/rejected": -0.6730347871780396, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.23703703703703705, | |
| "grad_norm": 194.74801635742188, | |
| "learning_rate": 3.90625e-05, | |
| "log_odds_chosen": 1.4447908401489258, | |
| "log_odds_ratio": -1.7767770290374756, | |
| "logits/chosen": 184.77932739257812, | |
| "logits/rejected": 192.9888916015625, | |
| "logps/chosen": -4.297823429107666, | |
| "logps/rejected": -5.7312822341918945, | |
| "loss": 82.643, | |
| "nll_loss": 2.5885708332061768, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.21489116549491882, | |
| "rewards/margins": 0.0716729462146759, | |
| "rewards/rejected": -0.2865641117095947, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.28444444444444444, | |
| "grad_norm": 331.9227600097656, | |
| "learning_rate": 4.6875e-05, | |
| "log_odds_chosen": 0.1392047107219696, | |
| "log_odds_ratio": -0.8184011578559875, | |
| "logits/chosen": 195.76513671875, | |
| "logits/rejected": 218.49557495117188, | |
| "logps/chosen": -1.6848766803741455, | |
| "logps/rejected": -1.8006423711776733, | |
| "loss": 62.6908, | |
| "nll_loss": 1.9485574960708618, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.08424384146928787, | |
| "rewards/margins": 0.005788279231637716, | |
| "rewards/rejected": -0.09003211557865143, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.33185185185185184, | |
| "grad_norm": 115.5135498046875, | |
| "learning_rate": 4.998613757348784e-05, | |
| "log_odds_chosen": 0.006599183194339275, | |
| "log_odds_ratio": -0.8610709309577942, | |
| "logits/chosen": 240.92092895507812, | |
| "logits/rejected": 237.75436401367188, | |
| "logps/chosen": -1.599082112312317, | |
| "logps/rejected": -1.6070514917373657, | |
| "loss": 56.7162, | |
| "nll_loss": 1.7848625183105469, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -0.0799541026353836, | |
| "rewards/margins": 0.00039846962317824364, | |
| "rewards/rejected": -0.08035256713628769, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.37925925925925924, | |
| "grad_norm": 102.91368865966797, | |
| "learning_rate": 4.990147841143462e-05, | |
| "log_odds_chosen": 0.20803086459636688, | |
| "log_odds_ratio": -0.7011880874633789, | |
| "logits/chosen": 223.38040161132812, | |
| "logits/rejected": 223.494873046875, | |
| "logps/chosen": -1.3678431510925293, | |
| "logps/rejected": -1.5363763570785522, | |
| "loss": 53.8445, | |
| "nll_loss": 1.6675764322280884, | |
| "rewards/accuracies": 0.48750001192092896, | |
| "rewards/chosen": -0.06839215010404587, | |
| "rewards/margins": 0.008426659740507603, | |
| "rewards/rejected": -0.07681881636381149, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.4266666666666667, | |
| "grad_norm": 72.4631576538086, | |
| "learning_rate": 4.97401218720448e-05, | |
| "log_odds_chosen": 0.17005965113639832, | |
| "log_odds_ratio": -0.780523419380188, | |
| "logits/chosen": 195.12318420410156, | |
| "logits/rejected": 192.85745239257812, | |
| "logps/chosen": -1.381034255027771, | |
| "logps/rejected": -1.5216766595840454, | |
| "loss": 51.6844, | |
| "nll_loss": 1.6356455087661743, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -0.06905172020196915, | |
| "rewards/margins": 0.007032120134681463, | |
| "rewards/rejected": -0.07608383893966675, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.4740740740740741, | |
| "grad_norm": 189.4646759033203, | |
| "learning_rate": 4.9502564938797946e-05, | |
| "log_odds_chosen": 0.17814292013645172, | |
| "log_odds_ratio": -0.7335196733474731, | |
| "logits/chosen": 187.52923583984375, | |
| "logits/rejected": 187.25880432128906, | |
| "logps/chosen": -1.3315727710723877, | |
| "logps/rejected": -1.4765068292617798, | |
| "loss": 52.9646, | |
| "nll_loss": 1.67325758934021, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.06657864153385162, | |
| "rewards/margins": 0.0072467029094696045, | |
| "rewards/rejected": -0.07382534444332123, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.5214814814814814, | |
| "grad_norm": 109.14909362792969, | |
| "learning_rate": 4.918953929490768e-05, | |
| "log_odds_chosen": 0.11172562837600708, | |
| "log_odds_ratio": -0.7358081340789795, | |
| "logits/chosen": 181.12258911132812, | |
| "logits/rejected": 176.92349243164062, | |
| "logps/chosen": -1.2745424509048462, | |
| "logps/rejected": -1.3694889545440674, | |
| "loss": 48.8487, | |
| "nll_loss": 1.5679895877838135, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -0.06372712552547455, | |
| "rewards/margins": 0.004747326020151377, | |
| "rewards/rejected": -0.06847444176673889, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.5688888888888889, | |
| "grad_norm": 174.18399047851562, | |
| "learning_rate": 4.88020090697132e-05, | |
| "log_odds_chosen": 0.21670827269554138, | |
| "log_odds_ratio": -0.6783889532089233, | |
| "logits/chosen": 178.7930908203125, | |
| "logits/rejected": 172.34970092773438, | |
| "logps/chosen": -1.1696356534957886, | |
| "logps/rejected": -1.3248437643051147, | |
| "loss": 47.1801, | |
| "nll_loss": 1.456142783164978, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.05848178267478943, | |
| "rewards/margins": 0.007760416716337204, | |
| "rewards/rejected": -0.06624220311641693, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.6162962962962963, | |
| "grad_norm": 140.9652557373047, | |
| "learning_rate": 4.834116786912897e-05, | |
| "log_odds_chosen": 0.18169096112251282, | |
| "log_odds_ratio": -0.7172664403915405, | |
| "logits/chosen": 180.29550170898438, | |
| "logits/rejected": 182.17674255371094, | |
| "logps/chosen": -1.2322423458099365, | |
| "logps/rejected": -1.3746764659881592, | |
| "loss": 46.7575, | |
| "nll_loss": 1.4799208641052246, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -0.06161211058497429, | |
| "rewards/margins": 0.007121713366359472, | |
| "rewards/rejected": -0.0687338262796402, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.6637037037037037, | |
| "grad_norm": 222.3861083984375, | |
| "learning_rate": 4.7808435099299045e-05, | |
| "log_odds_chosen": 0.2717851996421814, | |
| "log_odds_ratio": -0.6738412380218506, | |
| "logits/chosen": 187.9241943359375, | |
| "logits/rejected": 175.0797576904297, | |
| "logps/chosen": -1.1564773321151733, | |
| "logps/rejected": -1.3804590702056885, | |
| "loss": 46.2865, | |
| "nll_loss": 1.4590284824371338, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.05782386660575867, | |
| "rewards/margins": 0.01119908131659031, | |
| "rewards/rejected": -0.06902293860912323, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.7111111111111111, | |
| "grad_norm": 71.39551544189453, | |
| "learning_rate": 4.720545159477922e-05, | |
| "log_odds_chosen": 0.2626160979270935, | |
| "log_odds_ratio": -0.6702799201011658, | |
| "logits/chosen": 178.911865234375, | |
| "logits/rejected": 184.33798217773438, | |
| "logps/chosen": -1.1193958520889282, | |
| "logps/rejected": -1.3148605823516846, | |
| "loss": 45.1045, | |
| "nll_loss": 1.4315224885940552, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.05596979334950447, | |
| "rewards/margins": 0.009773238562047482, | |
| "rewards/rejected": -0.06574302911758423, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.7585185185185185, | |
| "grad_norm": 77.85042572021484, | |
| "learning_rate": 4.653407456471222e-05, | |
| "log_odds_chosen": 0.25611892342567444, | |
| "log_odds_ratio": -0.7023987770080566, | |
| "logits/chosen": 186.37899780273438, | |
| "logits/rejected": 181.420654296875, | |
| "logps/chosen": -1.1519978046417236, | |
| "logps/rejected": -1.3404672145843506, | |
| "loss": 44.5742, | |
| "nll_loss": 1.4464070796966553, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.05759989097714424, | |
| "rewards/margins": 0.009423469193279743, | |
| "rewards/rejected": -0.06702335178852081, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.8059259259259259, | |
| "grad_norm": 131.22291564941406, | |
| "learning_rate": 4.579637187256222e-05, | |
| "log_odds_chosen": 0.21689875423908234, | |
| "log_odds_ratio": -0.6867625713348389, | |
| "logits/chosen": 193.07936096191406, | |
| "logits/rejected": 179.24037170410156, | |
| "logps/chosen": -1.0784428119659424, | |
| "logps/rejected": -1.2363873720169067, | |
| "loss": 44.1766, | |
| "nll_loss": 1.3691766262054443, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.053922139108181, | |
| "rewards/margins": 0.007897228002548218, | |
| "rewards/rejected": -0.061819374561309814, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.8533333333333334, | |
| "grad_norm": 102.12734985351562, | |
| "learning_rate": 4.499461566702685e-05, | |
| "log_odds_chosen": 0.28654736280441284, | |
| "log_odds_ratio": -0.6533454060554504, | |
| "logits/chosen": 198.9041290283203, | |
| "logits/rejected": 193.521728515625, | |
| "logps/chosen": -1.0521138906478882, | |
| "logps/rejected": -1.2458927631378174, | |
| "loss": 45.3582, | |
| "nll_loss": 1.402296781539917, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.05260569974780083, | |
| "rewards/margins": 0.00968894176185131, | |
| "rewards/rejected": -0.06229463964700699, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.9007407407407407, | |
| "grad_norm": 66.06409454345703, | |
| "learning_rate": 4.413127538374411e-05, | |
| "log_odds_chosen": 0.35550013184547424, | |
| "log_odds_ratio": -0.6025335192680359, | |
| "logits/chosen": 186.05125427246094, | |
| "logits/rejected": 179.599609375, | |
| "logps/chosen": -0.9773751497268677, | |
| "logps/rejected": -1.2151715755462646, | |
| "loss": 43.5128, | |
| "nll_loss": 1.3024979829788208, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -0.048868756741285324, | |
| "rewards/margins": 0.011889824643731117, | |
| "rewards/rejected": -0.06075857952237129, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.9481481481481482, | |
| "grad_norm": 54.939449310302734, | |
| "learning_rate": 4.320901013934887e-05, | |
| "log_odds_chosen": 0.1582535207271576, | |
| "log_odds_ratio": -0.7311251759529114, | |
| "logits/chosen": 187.7299346923828, | |
| "logits/rejected": 176.6575927734375, | |
| "logps/chosen": -1.0948975086212158, | |
| "logps/rejected": -1.2327477931976318, | |
| "loss": 43.5056, | |
| "nll_loss": 1.4032866954803467, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -0.05474487692117691, | |
| "rewards/margins": 0.006892507430166006, | |
| "rewards/rejected": -0.061637382954359055, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.9955555555555555, | |
| "grad_norm": 107.38319396972656, | |
| "learning_rate": 4.223066054130568e-05, | |
| "log_odds_chosen": 0.27941060066223145, | |
| "log_odds_ratio": -0.6358767151832581, | |
| "logits/chosen": 184.79893493652344, | |
| "logits/rejected": 176.64239501953125, | |
| "logps/chosen": -1.011204481124878, | |
| "logps/rejected": -1.2123634815216064, | |
| "loss": 42.5508, | |
| "nll_loss": 1.3267438411712646, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.050560224801301956, | |
| "rewards/margins": 0.010057945735752583, | |
| "rewards/rejected": -0.06061817333102226, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.037925925925926, | |
| "grad_norm": 69.81970977783203, | |
| "learning_rate": 4.1199239938743797e-05, | |
| "log_odds_chosen": 0.506846010684967, | |
| "log_odds_ratio": -0.5906091332435608, | |
| "logits/chosen": 173.7986297607422, | |
| "logits/rejected": 167.03512573242188, | |
| "logps/chosen": -0.878280758857727, | |
| "logps/rejected": -1.224012851715088, | |
| "loss": 33.7043, | |
| "nll_loss": 1.1471017599105835, | |
| "rewards/accuracies": 0.6643356680870056, | |
| "rewards/chosen": -0.04391403868794441, | |
| "rewards/margins": 0.01728660613298416, | |
| "rewards/rejected": -0.061200641095638275, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.0853333333333333, | |
| "grad_norm": 69.83367156982422, | |
| "learning_rate": 4.0117925141242174e-05, | |
| "log_odds_chosen": 0.726813018321991, | |
| "log_odds_ratio": -0.48568326234817505, | |
| "logits/chosen": 172.2089385986328, | |
| "logits/rejected": 163.337646484375, | |
| "logps/chosen": -0.841424286365509, | |
| "logps/rejected": -1.3027280569076538, | |
| "loss": 36.2183, | |
| "nll_loss": 1.106263279914856, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": -0.04207121208310127, | |
| "rewards/margins": 0.023065198212862015, | |
| "rewards/rejected": -0.06513641774654388, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.1327407407407408, | |
| "grad_norm": 99.912841796875, | |
| "learning_rate": 3.899004663415084e-05, | |
| "log_odds_chosen": 0.6838423013687134, | |
| "log_odds_ratio": -0.5168099403381348, | |
| "logits/chosen": 165.5643310546875, | |
| "logits/rejected": 164.38926696777344, | |
| "logps/chosen": -0.8310489654541016, | |
| "logps/rejected": -1.2339271306991577, | |
| "loss": 36.5605, | |
| "nll_loss": 1.1273092031478882, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.04155245050787926, | |
| "rewards/margins": 0.020143909379839897, | |
| "rewards/rejected": -0.061696358025074005, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.1801481481481482, | |
| "grad_norm": 73.86228942871094, | |
| "learning_rate": 3.781907832058587e-05, | |
| "log_odds_chosen": 0.6109746694564819, | |
| "log_odds_ratio": -0.5379955172538757, | |
| "logits/chosen": 163.86184692382812, | |
| "logits/rejected": 161.99966430664062, | |
| "logps/chosen": -0.8782113790512085, | |
| "logps/rejected": -1.2740932703018188, | |
| "loss": 34.8236, | |
| "nll_loss": 1.1110111474990845, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.043910570442676544, | |
| "rewards/margins": 0.019794095307588577, | |
| "rewards/rejected": -0.06370466202497482, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.2275555555555555, | |
| "grad_norm": 105.75894165039062, | |
| "learning_rate": 3.660862682169282e-05, | |
| "log_odds_chosen": 0.5121253728866577, | |
| "log_odds_ratio": -0.5994678735733032, | |
| "logits/chosen": 166.21932983398438, | |
| "logits/rejected": 163.54107666015625, | |
| "logps/chosen": -0.8976671099662781, | |
| "logps/rejected": -1.1910202503204346, | |
| "loss": 34.8684, | |
| "nll_loss": 1.1224849224090576, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.0448833629488945, | |
| "rewards/margins": 0.01466765720397234, | |
| "rewards/rejected": -0.059551019221544266, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.274962962962963, | |
| "grad_norm": 97.21427154541016, | |
| "learning_rate": 3.5362420368134356e-05, | |
| "log_odds_chosen": 0.5886248350143433, | |
| "log_odds_ratio": -0.5262036323547363, | |
| "logits/chosen": 165.6793670654297, | |
| "logits/rejected": 166.072509765625, | |
| "logps/chosen": -0.7888522148132324, | |
| "logps/rejected": -1.1330220699310303, | |
| "loss": 35.0844, | |
| "nll_loss": 1.073457956314087, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -0.03944260999560356, | |
| "rewards/margins": 0.017208490520715714, | |
| "rewards/rejected": -0.056651096791028976, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.3223703703703704, | |
| "grad_norm": 49.054542541503906, | |
| "learning_rate": 3.408429731701635e-05, | |
| "log_odds_chosen": 0.6223124265670776, | |
| "log_odds_ratio": -0.554074764251709, | |
| "logits/chosen": 161.89114379882812, | |
| "logits/rejected": 166.47528076171875, | |
| "logps/chosen": -0.8866279721260071, | |
| "logps/rejected": -1.2842514514923096, | |
| "loss": 36.4437, | |
| "nll_loss": 1.141160011291504, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.04433140158653259, | |
| "rewards/margins": 0.019881170243024826, | |
| "rewards/rejected": -0.06421257555484772, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.3697777777777778, | |
| "grad_norm": 41.25300216674805, | |
| "learning_rate": 3.2778194329621104e-05, | |
| "log_odds_chosen": 0.6255282163619995, | |
| "log_odds_ratio": -0.5217418670654297, | |
| "logits/chosen": 167.75601196289062, | |
| "logits/rejected": 155.63418579101562, | |
| "logps/chosen": -0.8616455793380737, | |
| "logps/rejected": -1.2749989032745361, | |
| "loss": 35.9371, | |
| "nll_loss": 1.1240451335906982, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.043082281947135925, | |
| "rewards/margins": 0.02066767029464245, | |
| "rewards/rejected": -0.06374995410442352, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.417185185185185, | |
| "grad_norm": 43.59724426269531, | |
| "learning_rate": 3.144813424636031e-05, | |
| "log_odds_chosen": 0.5113286972045898, | |
| "log_odds_ratio": -0.5680266618728638, | |
| "logits/chosen": 165.27484130859375, | |
| "logits/rejected": 169.46823120117188, | |
| "logps/chosen": -0.8289966583251953, | |
| "logps/rejected": -1.1291935443878174, | |
| "loss": 34.8101, | |
| "nll_loss": 1.084172010421753, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.041449833661317825, | |
| "rewards/margins": 0.01500984001904726, | |
| "rewards/rejected": -0.05645967274904251, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.4645925925925927, | |
| "grad_norm": 42.254207611083984, | |
| "learning_rate": 3.0098213696293542e-05, | |
| "log_odds_chosen": 0.5680588483810425, | |
| "log_odds_ratio": -0.5456587672233582, | |
| "logits/chosen": 173.2566375732422, | |
| "logits/rejected": 164.67221069335938, | |
| "logps/chosen": -0.8088580369949341, | |
| "logps/rejected": -1.1407465934753418, | |
| "loss": 35.4576, | |
| "nll_loss": 1.1147387027740479, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.040442898869514465, | |
| "rewards/margins": 0.016594436019659042, | |
| "rewards/rejected": -0.057037342339754105, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.512, | |
| "grad_norm": 45.57976531982422, | |
| "learning_rate": 2.8732590479375165e-05, | |
| "log_odds_chosen": 0.540373682975769, | |
| "log_odds_ratio": -0.543350338935852, | |
| "logits/chosen": 172.65396118164062, | |
| "logits/rejected": 173.95144653320312, | |
| "logps/chosen": -0.910565972328186, | |
| "logps/rejected": -1.2549046277999878, | |
| "loss": 35.9476, | |
| "nll_loss": 1.1549952030181885, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.045528292655944824, | |
| "rewards/margins": 0.017216943204402924, | |
| "rewards/rejected": -0.06274524331092834, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.5594074074074074, | |
| "grad_norm": 54.967220306396484, | |
| "learning_rate": 2.7355470760292956e-05, | |
| "log_odds_chosen": 0.6586408019065857, | |
| "log_odds_ratio": -0.5135148763656616, | |
| "logits/chosen": 164.9529266357422, | |
| "logits/rejected": 176.58729553222656, | |
| "logps/chosen": -0.8376239538192749, | |
| "logps/rejected": -1.241012454032898, | |
| "loss": 36.1099, | |
| "nll_loss": 1.1062064170837402, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -0.041881196200847626, | |
| "rewards/margins": 0.020169423893094063, | |
| "rewards/rejected": -0.062050629407167435, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.6068148148148147, | |
| "grad_norm": 48.87372970581055, | |
| "learning_rate": 2.597109611334169e-05, | |
| "log_odds_chosen": 0.6633528470993042, | |
| "log_odds_ratio": -0.5211383104324341, | |
| "logits/chosen": 169.59494018554688, | |
| "logits/rejected": 169.11012268066406, | |
| "logps/chosen": -0.8359187245368958, | |
| "logps/rejected": -1.2383381128311157, | |
| "loss": 35.4707, | |
| "nll_loss": 1.0976989269256592, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -0.041795939207077026, | |
| "rewards/margins": 0.02012096717953682, | |
| "rewards/rejected": -0.06191689893603325, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.6542222222222223, | |
| "grad_norm": 38.77116012573242, | |
| "learning_rate": 2.458373045823404e-05, | |
| "log_odds_chosen": 0.6313592195510864, | |
| "log_odds_ratio": -0.519364058971405, | |
| "logits/chosen": 174.59616088867188, | |
| "logits/rejected": 163.9171600341797, | |
| "logps/chosen": -0.826526939868927, | |
| "logps/rejected": -1.2227591276168823, | |
| "loss": 35.0289, | |
| "nll_loss": 1.1160304546356201, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -0.04132634773850441, | |
| "rewards/margins": 0.019811619073152542, | |
| "rewards/rejected": -0.06113796308636665, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.7016296296296296, | |
| "grad_norm": 38.79117965698242, | |
| "learning_rate": 2.3197646927086697e-05, | |
| "log_odds_chosen": 0.3873888850212097, | |
| "log_odds_ratio": -0.6197336912155151, | |
| "logits/chosen": 164.18601989746094, | |
| "logits/rejected": 166.98605346679688, | |
| "logps/chosen": -0.8658819198608398, | |
| "logps/rejected": -1.0931422710418701, | |
| "loss": 35.4427, | |
| "nll_loss": 1.1255178451538086, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.043294090777635574, | |
| "rewards/margins": 0.011363019235432148, | |
| "rewards/rejected": -0.054657112807035446, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.749037037037037, | |
| "grad_norm": 41.68294906616211, | |
| "learning_rate": 2.1817114703032176e-05, | |
| "log_odds_chosen": 0.5725528001785278, | |
| "log_odds_ratio": -0.5303586721420288, | |
| "logits/chosen": 168.3577880859375, | |
| "logits/rejected": 167.2732391357422, | |
| "logps/chosen": -0.8526233434677124, | |
| "logps/rejected": -1.2170125246047974, | |
| "loss": 35.5527, | |
| "nll_loss": 1.1156741380691528, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.04263117164373398, | |
| "rewards/margins": 0.01821945607662201, | |
| "rewards/rejected": -0.06085062772035599, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.7964444444444445, | |
| "grad_norm": 57.764461517333984, | |
| "learning_rate": 2.0446385870993467e-05, | |
| "log_odds_chosen": 0.6254726648330688, | |
| "log_odds_ratio": -0.5300140976905823, | |
| "logits/chosen": 176.73265075683594, | |
| "logits/rejected": 166.2118377685547, | |
| "logps/chosen": -0.8647019267082214, | |
| "logps/rejected": -1.2612402439117432, | |
| "loss": 35.7179, | |
| "nll_loss": 1.1093218326568604, | |
| "rewards/accuracies": 0.7875000238418579, | |
| "rewards/chosen": -0.04323510080575943, | |
| "rewards/margins": 0.019826913252472878, | |
| "rewards/rejected": -0.06306201219558716, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.8438518518518519, | |
| "grad_norm": 57.696022033691406, | |
| "learning_rate": 1.9089682321121834e-05, | |
| "log_odds_chosen": 0.7172800302505493, | |
| "log_odds_ratio": -0.5064095258712769, | |
| "logits/chosen": 168.99301147460938, | |
| "logits/rejected": 158.69894409179688, | |
| "logps/chosen": -0.8419252634048462, | |
| "logps/rejected": -1.2953702211380005, | |
| "loss": 35.0776, | |
| "nll_loss": 1.1046844720840454, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.04209625720977783, | |
| "rewards/margins": 0.022672254592180252, | |
| "rewards/rejected": -0.06476850807666779, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.8912592592592592, | |
| "grad_norm": 38.26820373535156, | |
| "learning_rate": 1.775118274523545e-05, | |
| "log_odds_chosen": 0.5704585313796997, | |
| "log_odds_ratio": -0.5714629888534546, | |
| "logits/chosen": 170.9229736328125, | |
| "logits/rejected": 165.9208526611328, | |
| "logps/chosen": -0.8371521830558777, | |
| "logps/rejected": -1.18575119972229, | |
| "loss": 35.9511, | |
| "nll_loss": 1.0667616128921509, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.041857607662677765, | |
| "rewards/margins": 0.017429957166314125, | |
| "rewards/rejected": -0.05928756669163704, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.9386666666666668, | |
| "grad_norm": 37.01579666137695, | |
| "learning_rate": 1.643500976631037e-05, | |
| "log_odds_chosen": 0.5442375540733337, | |
| "log_odds_ratio": -0.5546382665634155, | |
| "logits/chosen": 175.0882110595703, | |
| "logits/rejected": 163.4840850830078, | |
| "logps/chosen": -0.8371946215629578, | |
| "logps/rejected": -1.1808607578277588, | |
| "loss": 34.5686, | |
| "nll_loss": 1.0792796611785889, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -0.04185972735285759, | |
| "rewards/margins": 0.01718330569565296, | |
| "rewards/rejected": -0.0590430311858654, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.986074074074074, | |
| "grad_norm": 58.07009506225586, | |
| "learning_rate": 1.514521724066537e-05, | |
| "log_odds_chosen": 0.5513701438903809, | |
| "log_odds_ratio": -0.5487216114997864, | |
| "logits/chosen": 160.052734375, | |
| "logits/rejected": 165.99490356445312, | |
| "logps/chosen": -0.804582953453064, | |
| "logps/rejected": -1.125427484512329, | |
| "loss": 33.0603, | |
| "nll_loss": 1.0380481481552124, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.04022914916276932, | |
| "rewards/margins": 0.01604221947491169, | |
| "rewards/rejected": -0.056271374225616455, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.0284444444444443, | |
| "grad_norm": 46.680702209472656, | |
| "learning_rate": 1.3885777771950348e-05, | |
| "log_odds_chosen": 0.9105411171913147, | |
| "log_odds_ratio": -0.4484807252883911, | |
| "logits/chosen": 155.98167419433594, | |
| "logits/rejected": 156.97645568847656, | |
| "logps/chosen": -0.7199848890304565, | |
| "logps/rejected": -1.2084262371063232, | |
| "loss": 27.155, | |
| "nll_loss": 0.9470744132995605, | |
| "rewards/accuracies": 0.8251748085021973, | |
| "rewards/chosen": -0.03599924594163895, | |
| "rewards/margins": 0.024422070011496544, | |
| "rewards/rejected": -0.06042131036520004, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.075851851851852, | |
| "grad_norm": 47.282562255859375, | |
| "learning_rate": 1.2660570475395683e-05, | |
| "log_odds_chosen": 1.1580384969711304, | |
| "log_odds_ratio": -0.39033469557762146, | |
| "logits/chosen": 142.10780334472656, | |
| "logits/rejected": 143.4950408935547, | |
| "logps/chosen": -0.6312204599380493, | |
| "logps/rejected": -1.257129192352295, | |
| "loss": 26.9195, | |
| "nll_loss": 0.8665180206298828, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -0.031561024487018585, | |
| "rewards/margins": 0.03129544109106064, | |
| "rewards/rejected": -0.06285646557807922, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.1232592592592594, | |
| "grad_norm": 44.752628326416016, | |
| "learning_rate": 1.1473369030008974e-05, | |
| "log_odds_chosen": 1.052072286605835, | |
| "log_odds_ratio": -0.412436306476593, | |
| "logits/chosen": 148.757568359375, | |
| "logits/rejected": 149.47320556640625, | |
| "logps/chosen": -0.6186680793762207, | |
| "logps/rejected": -1.1820106506347656, | |
| "loss": 28.038, | |
| "nll_loss": 0.87525475025177, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": -0.030933406203985214, | |
| "rewards/margins": 0.0281671229749918, | |
| "rewards/rejected": -0.05910053104162216, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.1706666666666665, | |
| "grad_norm": 45.039642333984375, | |
| "learning_rate": 1.0327830055518842e-05, | |
| "log_odds_chosen": 1.1950232982635498, | |
| "log_odds_ratio": -0.3665129542350769, | |
| "logits/chosen": 148.42233276367188, | |
| "logits/rejected": 144.0989227294922, | |
| "logps/chosen": -0.6338512897491455, | |
| "logps/rejected": -1.2817251682281494, | |
| "loss": 26.4916, | |
| "nll_loss": 0.8120902180671692, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": -0.031692564487457275, | |
| "rewards/margins": 0.03239370137453079, | |
| "rewards/rejected": -0.06408625841140747, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.218074074074074, | |
| "grad_norm": 44.143611907958984, | |
| "learning_rate": 9.227481849865235e-06, | |
| "log_odds_chosen": 1.0343241691589355, | |
| "log_odds_ratio": -0.41224583983421326, | |
| "logits/chosen": 142.4004364013672, | |
| "logits/rejected": 147.73538208007812, | |
| "logps/chosen": -0.6527639627456665, | |
| "logps/rejected": -1.1827428340911865, | |
| "loss": 27.6183, | |
| "nll_loss": 0.8605834245681763, | |
| "rewards/accuracies": 0.8374999761581421, | |
| "rewards/chosen": -0.032638195902109146, | |
| "rewards/margins": 0.026498943567276, | |
| "rewards/rejected": -0.059137143194675446, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 2.2654814814814817, | |
| "grad_norm": 41.318138122558594, | |
| "learning_rate": 8.175713521924978e-06, | |
| "log_odds_chosen": 1.1422860622406006, | |
| "log_odds_ratio": -0.3735355734825134, | |
| "logits/chosen": 145.45065307617188, | |
| "logits/rejected": 151.5826416015625, | |
| "logps/chosen": -0.6259297132492065, | |
| "logps/rejected": -1.2332427501678467, | |
| "loss": 27.1915, | |
| "nll_loss": 0.8626583218574524, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": -0.03129648417234421, | |
| "rewards/margins": 0.030365657061338425, | |
| "rewards/rejected": -0.06166214868426323, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.3128888888888888, | |
| "grad_norm": 44.362037658691406, | |
| "learning_rate": 7.1757645529443665e-06, | |
| "log_odds_chosen": 1.273262858390808, | |
| "log_odds_ratio": -0.35038676857948303, | |
| "logits/chosen": 142.28067016601562, | |
| "logits/rejected": 141.07127380371094, | |
| "logps/chosen": -0.6050504446029663, | |
| "logps/rejected": -1.2771079540252686, | |
| "loss": 26.2994, | |
| "nll_loss": 0.849342942237854, | |
| "rewards/accuracies": 0.8687499761581421, | |
| "rewards/chosen": -0.030252521857619286, | |
| "rewards/margins": 0.03360287845134735, | |
| "rewards/rejected": -0.06385539472103119, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 2.3602962962962963, | |
| "grad_norm": 37.62168884277344, | |
| "learning_rate": 6.230714818829733e-06, | |
| "log_odds_chosen": 1.244945764541626, | |
| "log_odds_ratio": -0.36415261030197144, | |
| "logits/chosen": 138.27842712402344, | |
| "logits/rejected": 143.33724975585938, | |
| "logps/chosen": -0.6403986811637878, | |
| "logps/rejected": -1.302079200744629, | |
| "loss": 26.4474, | |
| "nll_loss": 0.8333737254142761, | |
| "rewards/accuracies": 0.831250011920929, | |
| "rewards/chosen": -0.03201993554830551, | |
| "rewards/margins": 0.03308403491973877, | |
| "rewards/rejected": -0.06510396301746368, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.407703703703704, | |
| "grad_norm": 40.77067947387695, | |
| "learning_rate": 5.343475104027743e-06, | |
| "log_odds_chosen": 1.3023749589920044, | |
| "log_odds_ratio": -0.3504185378551483, | |
| "logits/chosen": 136.38796997070312, | |
| "logits/rejected": 143.35763549804688, | |
| "logps/chosen": -0.5728102326393127, | |
| "logps/rejected": -1.279425859451294, | |
| "loss": 26.3726, | |
| "nll_loss": 0.8203259706497192, | |
| "rewards/accuracies": 0.8687499761581421, | |
| "rewards/chosen": -0.028640514239668846, | |
| "rewards/margins": 0.03533077985048294, | |
| "rewards/rejected": -0.06397129595279694, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 2.455111111111111, | |
| "grad_norm": 45.096004486083984, | |
| "learning_rate": 4.516778136213037e-06, | |
| "log_odds_chosen": 1.3704315423965454, | |
| "log_odds_ratio": -0.3305371403694153, | |
| "logits/chosen": 137.95346069335938, | |
| "logits/rejected": 139.55905151367188, | |
| "logps/chosen": -0.591722309589386, | |
| "logps/rejected": -1.3149653673171997, | |
| "loss": 26.1383, | |
| "nll_loss": 0.8089566230773926, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": -0.02958611212670803, | |
| "rewards/margins": 0.036162156611680984, | |
| "rewards/rejected": -0.06574826687574387, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.5025185185185186, | |
| "grad_norm": 47.061092376708984, | |
| "learning_rate": 3.7531701693965554e-06, | |
| "log_odds_chosen": 1.1507422924041748, | |
| "log_odds_ratio": -0.408168226480484, | |
| "logits/chosen": 135.37014770507812, | |
| "logits/rejected": 141.3232421875, | |
| "logps/chosen": -0.6197280883789062, | |
| "logps/rejected": -1.2372686862945557, | |
| "loss": 25.815, | |
| "nll_loss": 0.8066200017929077, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": -0.030986404046416283, | |
| "rewards/margins": 0.03087702952325344, | |
| "rewards/rejected": -0.061863433569669724, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 2.549925925925926, | |
| "grad_norm": 44.70289611816406, | |
| "learning_rate": 3.055003141378948e-06, | |
| "log_odds_chosen": 1.2243850231170654, | |
| "log_odds_ratio": -0.36160460114479065, | |
| "logits/chosen": 140.19070434570312, | |
| "logits/rejected": 143.959716796875, | |
| "logps/chosen": -0.632489025592804, | |
| "logps/rejected": -1.2881219387054443, | |
| "loss": 25.857, | |
| "nll_loss": 0.8426779508590698, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": -0.031624455004930496, | |
| "rewards/margins": 0.03278164193034172, | |
| "rewards/rejected": -0.06440608948469162, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.5973333333333333, | |
| "grad_norm": 40.76127243041992, | |
| "learning_rate": 2.424427429704365e-06, | |
| "log_odds_chosen": 1.3216700553894043, | |
| "log_odds_ratio": -0.3369961977005005, | |
| "logits/chosen": 138.0738067626953, | |
| "logits/rejected": 139.0469970703125, | |
| "logps/chosen": -0.6244256496429443, | |
| "logps/rejected": -1.3335908651351929, | |
| "loss": 27.0255, | |
| "nll_loss": 0.8462891578674316, | |
| "rewards/accuracies": 0.893750011920929, | |
| "rewards/chosen": -0.031221281737089157, | |
| "rewards/margins": 0.035458266735076904, | |
| "rewards/rejected": -0.06667955219745636, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 2.644740740740741, | |
| "grad_norm": 42.352970123291016, | |
| "learning_rate": 1.8633852284264508e-06, | |
| "log_odds_chosen": 1.190826654434204, | |
| "log_odds_ratio": -0.3770466446876526, | |
| "logits/chosen": 140.8355255126953, | |
| "logits/rejected": 142.74783325195312, | |
| "logps/chosen": -0.6085912585258484, | |
| "logps/rejected": -1.2392776012420654, | |
| "loss": 25.9705, | |
| "nll_loss": 0.7998742461204529, | |
| "rewards/accuracies": 0.893750011920929, | |
| "rewards/chosen": -0.03042956255376339, | |
| "rewards/margins": 0.031534306704998016, | |
| "rewards/rejected": -0.061963874846696854, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.6921481481481484, | |
| "grad_norm": 40.149749755859375, | |
| "learning_rate": 1.3736045660864034e-06, | |
| "log_odds_chosen": 1.1410518884658813, | |
| "log_odds_ratio": -0.3837069869041443, | |
| "logits/chosen": 147.14105224609375, | |
| "logits/rejected": 141.97027587890625, | |
| "logps/chosen": -0.6317769289016724, | |
| "logps/rejected": -1.2591302394866943, | |
| "loss": 26.657, | |
| "nll_loss": 0.8469123840332031, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": -0.0315888449549675, | |
| "rewards/margins": 0.03136766701936722, | |
| "rewards/rejected": -0.06295650452375412, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 2.7395555555555555, | |
| "grad_norm": 60.75810623168945, | |
| "learning_rate": 9.565939833279192e-07, | |
| "log_odds_chosen": 1.31714928150177, | |
| "log_odds_ratio": -0.3334599435329437, | |
| "logits/chosen": 140.9696807861328, | |
| "logits/rejected": 147.40383911132812, | |
| "logps/chosen": -0.5831328630447388, | |
| "logps/rejected": -1.263795018196106, | |
| "loss": 27.04, | |
| "nll_loss": 0.8344683647155762, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": -0.029156645759940147, | |
| "rewards/margins": 0.03403311222791672, | |
| "rewards/rejected": -0.06318975985050201, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.786962962962963, | |
| "grad_norm": 43.854427337646484, | |
| "learning_rate": 6.136378865420872e-07, | |
| "log_odds_chosen": 1.1479493379592896, | |
| "log_odds_ratio": -0.4157779812812805, | |
| "logits/chosen": 139.2587890625, | |
| "logits/rejected": 147.88906860351562, | |
| "logps/chosen": -0.6736854314804077, | |
| "logps/rejected": -1.3039884567260742, | |
| "loss": 26.5112, | |
| "nll_loss": 0.8539054989814758, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": -0.033684272319078445, | |
| "rewards/margins": 0.03151514753699303, | |
| "rewards/rejected": -0.06519941985607147, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 2.83437037037037, | |
| "grad_norm": 52.22172164916992, | |
| "learning_rate": 3.45792591853214e-07, | |
| "log_odds_chosen": 1.2029494047164917, | |
| "log_odds_ratio": -0.38027894496917725, | |
| "logits/chosen": 151.04905700683594, | |
| "logits/rejected": 142.887939453125, | |
| "logps/chosen": -0.6414423584938049, | |
| "logps/rejected": -1.3054555654525757, | |
| "loss": 26.8389, | |
| "nll_loss": 0.8671246767044067, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": -0.032072119414806366, | |
| "rewards/margins": 0.033200670033693314, | |
| "rewards/rejected": -0.06527277827262878, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.8817777777777778, | |
| "grad_norm": 47.876564025878906, | |
| "learning_rate": 1.538830716302092e-07, | |
| "log_odds_chosen": 1.184543490409851, | |
| "log_odds_ratio": -0.3588692843914032, | |
| "logits/chosen": 142.39617919921875, | |
| "logits/rejected": 142.94223022460938, | |
| "logps/chosen": -0.6531465649604797, | |
| "logps/rejected": -1.2773756980895996, | |
| "loss": 26.2345, | |
| "nll_loss": 0.8135266304016113, | |
| "rewards/accuracies": 0.8687499761581421, | |
| "rewards/chosen": -0.03265732526779175, | |
| "rewards/margins": 0.031211454421281815, | |
| "rewards/rejected": -0.06386878341436386, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 2.9291851851851853, | |
| "grad_norm": 41.12287902832031, | |
| "learning_rate": 3.8500413544415025e-08, | |
| "log_odds_chosen": 1.222886323928833, | |
| "log_odds_ratio": -0.37252649664878845, | |
| "logits/chosen": 142.91061401367188, | |
| "logits/rejected": 140.1123046875, | |
| "logps/chosen": -0.613469123840332, | |
| "logps/rejected": -1.2194663286209106, | |
| "loss": 26.3091, | |
| "nll_loss": 0.8243390917778015, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": -0.03067345917224884, | |
| "rewards/margins": 0.030299866572022438, | |
| "rewards/rejected": -0.06097331643104553, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.9765925925925925, | |
| "grad_norm": 49.64106369018555, | |
| "learning_rate": 0.0, | |
| "log_odds_chosen": 1.233667016029358, | |
| "log_odds_ratio": -0.3490303158760071, | |
| "logits/chosen": 141.09365844726562, | |
| "logits/rejected": 138.0718231201172, | |
| "logps/chosen": -0.6048657894134521, | |
| "logps/rejected": -1.2514102458953857, | |
| "loss": 25.6403, | |
| "nll_loss": 0.8243409395217896, | |
| "rewards/accuracies": 0.8812500238418579, | |
| "rewards/chosen": -0.030243292450904846, | |
| "rewards/margins": 0.03232722356915474, | |
| "rewards/rejected": -0.06257051229476929, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 2.9765925925925925, | |
| "step": 315, | |
| "total_flos": 0.0, | |
| "train_loss": 50.92685067313058, | |
| "train_runtime": 8087.8942, | |
| "train_samples_per_second": 2.504, | |
| "train_steps_per_second": 0.039 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 315, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 100000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |