| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 100, | |
| "global_step": 385, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 1.282051282051282e-07, | |
| "logits/chosen": -1.7278180122375488, | |
| "logits/rejected": -1.7377450466156006, | |
| "logps/chosen": -29.553977966308594, | |
| "logps/rejected": -42.813133239746094, | |
| "loss": 1.0, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 1.282051282051282e-06, | |
| "logits/chosen": -1.866492748260498, | |
| "logits/rejected": -1.87080979347229, | |
| "logps/chosen": -36.97657775878906, | |
| "logps/rejected": -33.65824890136719, | |
| "loss": 0.9236, | |
| "rewards/accuracies": 0.5277777910232544, | |
| "rewards/chosen": 0.015009618364274502, | |
| "rewards/margins": 0.02909613959491253, | |
| "rewards/rejected": -0.014086521230638027, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 2.564102564102564e-06, | |
| "logits/chosen": -1.9977840185165405, | |
| "logits/rejected": -2.000425100326538, | |
| "logps/chosen": -29.640512466430664, | |
| "logps/rejected": -29.048751831054688, | |
| "loss": 1.0528, | |
| "rewards/accuracies": 0.4375, | |
| "rewards/chosen": 0.0008407801506109536, | |
| "rewards/margins": -0.0065057664178311825, | |
| "rewards/rejected": 0.007346546743065119, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 3.846153846153847e-06, | |
| "logits/chosen": -1.9210313558578491, | |
| "logits/rejected": -1.9183330535888672, | |
| "logps/chosen": -31.377187728881836, | |
| "logps/rejected": -33.214942932128906, | |
| "loss": 0.976, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": 0.019451653584837914, | |
| "rewards/margins": 0.01769269071519375, | |
| "rewards/rejected": 0.001758962869644165, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 4.999896948438434e-06, | |
| "logits/chosen": -2.0173258781433105, | |
| "logits/rejected": -2.008592128753662, | |
| "logps/chosen": -32.55642318725586, | |
| "logps/rejected": -32.49436569213867, | |
| "loss": 1.0336, | |
| "rewards/accuracies": 0.4749999940395355, | |
| "rewards/chosen": 0.010191624984145164, | |
| "rewards/margins": 0.001039800001308322, | |
| "rewards/rejected": 0.009151825681328773, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 4.987541037542187e-06, | |
| "logits/chosen": -1.8626506328582764, | |
| "logits/rejected": -1.8518873453140259, | |
| "logps/chosen": -33.54867172241211, | |
| "logps/rejected": -35.45621109008789, | |
| "loss": 1.0318, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 0.004401391837745905, | |
| "rewards/margins": 0.006334079895168543, | |
| "rewards/rejected": -0.001932688057422638, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 4.954691471941119e-06, | |
| "logits/chosen": -1.940718650817871, | |
| "logits/rejected": -1.9426720142364502, | |
| "logps/chosen": -32.53395462036133, | |
| "logps/rejected": -33.20496368408203, | |
| "loss": 0.9445, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": 0.03311632573604584, | |
| "rewards/margins": 0.04990752786397934, | |
| "rewards/rejected": -0.016791202127933502, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 4.901618883413549e-06, | |
| "logits/chosen": -2.0724740028381348, | |
| "logits/rejected": -2.077458381652832, | |
| "logps/chosen": -33.9911994934082, | |
| "logps/rejected": -36.61388397216797, | |
| "loss": 1.1022, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.001358801149763167, | |
| "rewards/margins": 0.021143654361367226, | |
| "rewards/rejected": -0.022502455860376358, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 4.828760511501322e-06, | |
| "logits/chosen": -1.936197280883789, | |
| "logits/rejected": -1.9393237829208374, | |
| "logps/chosen": -34.28167724609375, | |
| "logps/rejected": -34.63819885253906, | |
| "loss": 0.8186, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": 0.07306724786758423, | |
| "rewards/margins": 0.09596274793148041, | |
| "rewards/rejected": -0.02289549633860588, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 4.7367166013034295e-06, | |
| "logits/chosen": -1.9451515674591064, | |
| "logits/rejected": -1.949669599533081, | |
| "logps/chosen": -32.39059066772461, | |
| "logps/rejected": -32.346839904785156, | |
| "loss": 1.0017, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": 0.03841521218419075, | |
| "rewards/margins": 0.03238191828131676, | |
| "rewards/rejected": 0.006033292505890131, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 4.626245458345211e-06, | |
| "logits/chosen": -2.042168617248535, | |
| "logits/rejected": -2.040160655975342, | |
| "logps/chosen": -32.12788772583008, | |
| "logps/rejected": -31.280298233032227, | |
| "loss": 0.8581, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": 0.05647433549165726, | |
| "rewards/margins": 0.06351961940526962, | |
| "rewards/rejected": -0.007045289967209101, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "eval_logits/chosen": -2.237440824508667, | |
| "eval_logits/rejected": -2.232595682144165, | |
| "eval_logps/chosen": -34.017024993896484, | |
| "eval_logps/rejected": -37.50282287597656, | |
| "eval_loss": 1.0730067491531372, | |
| "eval_rewards/accuracies": 0.5199335813522339, | |
| "eval_rewards/chosen": 0.00876238290220499, | |
| "eval_rewards/margins": 0.001864485559053719, | |
| "eval_rewards/rejected": 0.0068978965282440186, | |
| "eval_runtime": 146.0042, | |
| "eval_samples_per_second": 2.349, | |
| "eval_steps_per_second": 0.295, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 4.498257201263691e-06, | |
| "logits/chosen": -1.9984264373779297, | |
| "logits/rejected": -1.9960410594940186, | |
| "logps/chosen": -33.10862350463867, | |
| "logps/rejected": -34.00126266479492, | |
| "loss": 1.1665, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": 0.0675957053899765, | |
| "rewards/margins": 0.05354113504290581, | |
| "rewards/rejected": 0.014054578728973866, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 4.353806263777678e-06, | |
| "logits/chosen": -2.0099892616271973, | |
| "logits/rejected": -2.001642942428589, | |
| "logps/chosen": -32.312686920166016, | |
| "logps/rejected": -32.10304641723633, | |
| "loss": 0.9851, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": 0.06610965728759766, | |
| "rewards/margins": 0.03688037022948265, | |
| "rewards/rejected": 0.029229288920760155, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 4.1940827077152755e-06, | |
| "logits/chosen": -2.0365984439849854, | |
| "logits/rejected": -2.0286123752593994, | |
| "logps/chosen": -30.3278865814209, | |
| "logps/rejected": -32.04685592651367, | |
| "loss": 0.9819, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.07391555607318878, | |
| "rewards/margins": 0.07084138691425323, | |
| "rewards/rejected": 0.0030741647351533175, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 4.0204024186666215e-06, | |
| "logits/chosen": -1.9668314456939697, | |
| "logits/rejected": -1.9770755767822266, | |
| "logps/chosen": -31.2120304107666, | |
| "logps/rejected": -32.57902526855469, | |
| "loss": 0.8178, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": 0.10381648689508438, | |
| "rewards/margins": 0.12588787078857422, | |
| "rewards/rejected": -0.02207140065729618, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 3.834196265035119e-06, | |
| "logits/chosen": -1.8799388408660889, | |
| "logits/rejected": -1.881087064743042, | |
| "logps/chosen": -33.97100830078125, | |
| "logps/rejected": -34.84876251220703, | |
| "loss": 0.8351, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": 0.11084038019180298, | |
| "rewards/margins": 0.15659382939338684, | |
| "rewards/rejected": -0.04575346037745476, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 3.636998309800573e-06, | |
| "logits/chosen": -1.9326753616333008, | |
| "logits/rejected": -1.9292488098144531, | |
| "logps/chosen": -36.040306091308594, | |
| "logps/rejected": -32.739051818847656, | |
| "loss": 0.8539, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": 0.0656842365860939, | |
| "rewards/margins": 0.07383431494235992, | |
| "rewards/rejected": -0.008150083012878895, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 3.4304331721118078e-06, | |
| "logits/chosen": -2.0338892936706543, | |
| "logits/rejected": -2.026510238647461, | |
| "logps/chosen": -33.518821716308594, | |
| "logps/rejected": -31.37355613708496, | |
| "loss": 0.7141, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": 0.13951851427555084, | |
| "rewards/margins": 0.15037165582180023, | |
| "rewards/rejected": -0.010853144340217113, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 3.2162026428305436e-06, | |
| "logits/chosen": -2.039998769760132, | |
| "logits/rejected": -2.045238733291626, | |
| "logps/chosen": -32.28400421142578, | |
| "logps/rejected": -32.450523376464844, | |
| "loss": 0.7652, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": 0.13039177656173706, | |
| "rewards/margins": 0.11308407783508301, | |
| "rewards/rejected": 0.017307698726654053, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 2.996071664294641e-06, | |
| "logits/chosen": -2.0408711433410645, | |
| "logits/rejected": -2.0380892753601074, | |
| "logps/chosen": -31.287479400634766, | |
| "logps/rejected": -31.33124351501465, | |
| "loss": 0.8251, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": 0.10073776543140411, | |
| "rewards/margins": 0.10725078731775284, | |
| "rewards/rejected": -0.0065130265429615974, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 2.7718537898066833e-06, | |
| "logits/chosen": -1.9110311269760132, | |
| "logits/rejected": -1.915704369544983, | |
| "logps/chosen": -31.336145401000977, | |
| "logps/rejected": -32.791221618652344, | |
| "loss": 0.899, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": 0.13506832718849182, | |
| "rewards/margins": 0.13521215319633484, | |
| "rewards/rejected": -0.0001438349427189678, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "eval_logits/chosen": -2.2354750633239746, | |
| "eval_logits/rejected": -2.2306265830993652, | |
| "eval_logps/chosen": -34.03763198852539, | |
| "eval_logps/rejected": -37.534156799316406, | |
| "eval_loss": 1.0715795755386353, | |
| "eval_rewards/accuracies": 0.5070598125457764, | |
| "eval_rewards/chosen": -0.0015398082323372364, | |
| "eval_rewards/margins": 0.007228231523185968, | |
| "eval_rewards/rejected": -0.008768039755523205, | |
| "eval_runtime": 145.8306, | |
| "eval_samples_per_second": 2.352, | |
| "eval_steps_per_second": 0.295, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 2.5453962426402006e-06, | |
| "logits/chosen": -2.0229578018188477, | |
| "logits/rejected": -2.033618450164795, | |
| "logps/chosen": -31.776050567626953, | |
| "logps/rejected": -33.90400314331055, | |
| "loss": 0.7465, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": 0.09762789309024811, | |
| "rewards/margins": 0.11827856302261353, | |
| "rewards/rejected": -0.020650675520300865, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 2.3185646976551794e-06, | |
| "logits/chosen": -1.9135267734527588, | |
| "logits/rejected": -1.9282987117767334, | |
| "logps/chosen": -29.896175384521484, | |
| "logps/rejected": -31.5633487701416, | |
| "loss": 0.75, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": 0.10768184810876846, | |
| "rewards/margins": 0.11026783287525177, | |
| "rewards/rejected": -0.002585983369499445, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 2.0932279108998323e-06, | |
| "logits/chosen": -1.9711973667144775, | |
| "logits/rejected": -1.9751732349395752, | |
| "logps/chosen": -33.15174102783203, | |
| "logps/rejected": -31.605077743530273, | |
| "loss": 0.7191, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": 0.1328631192445755, | |
| "rewards/margins": 0.1607515811920166, | |
| "rewards/rejected": -0.027888456359505653, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 1.8712423238279358e-06, | |
| "logits/chosen": -1.969957709312439, | |
| "logits/rejected": -1.9480478763580322, | |
| "logps/chosen": -33.9122200012207, | |
| "logps/rejected": -35.02121353149414, | |
| "loss": 0.6955, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 0.12076646089553833, | |
| "rewards/margins": 0.15597540140151978, | |
| "rewards/rejected": -0.03520893678069115, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 1.6544367689701824e-06, | |
| "logits/chosen": -2.0103070735931396, | |
| "logits/rejected": -2.006990909576416, | |
| "logps/chosen": -32.72673797607422, | |
| "logps/rejected": -36.23841094970703, | |
| "loss": 0.8178, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": 0.09916610270738602, | |
| "rewards/margins": 0.10590960830450058, | |
| "rewards/rejected": -0.006743511650711298, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 1.4445974030621963e-06, | |
| "logits/chosen": -1.8776795864105225, | |
| "logits/rejected": -1.875245451927185, | |
| "logps/chosen": -34.003971099853516, | |
| "logps/rejected": -35.510765075683594, | |
| "loss": 0.889, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": 0.09221391379833221, | |
| "rewards/margins": 0.09528535604476929, | |
| "rewards/rejected": -0.0030714483000338078, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 1.243452991757889e-06, | |
| "logits/chosen": -1.863521933555603, | |
| "logits/rejected": -1.8610206842422485, | |
| "logps/chosen": -34.20132827758789, | |
| "logps/rejected": -31.76943016052246, | |
| "loss": 0.8268, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": 0.09813599288463593, | |
| "rewards/margins": 0.09718601405620575, | |
| "rewards/rejected": 0.000949984765611589, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 1.0526606671603523e-06, | |
| "logits/chosen": -1.9676616191864014, | |
| "logits/rejected": -1.957082748413086, | |
| "logps/chosen": -35.020606994628906, | |
| "logps/rejected": -31.831247329711914, | |
| "loss": 0.6669, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": 0.168013796210289, | |
| "rewards/margins": 0.16314153373241425, | |
| "rewards/rejected": 0.00487226527184248, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 8.737922755071455e-07, | |
| "logits/chosen": -2.0636610984802246, | |
| "logits/rejected": -2.0486764907836914, | |
| "logps/chosen": -30.683029174804688, | |
| "logps/rejected": -32.61827850341797, | |
| "loss": 0.894, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": 0.12054909765720367, | |
| "rewards/margins": 0.10377003997564316, | |
| "rewards/rejected": 0.016779040917754173, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 7.08321427484816e-07, | |
| "logits/chosen": -1.9357779026031494, | |
| "logits/rejected": -1.9332023859024048, | |
| "logps/chosen": -32.54056930541992, | |
| "logps/rejected": -30.850332260131836, | |
| "loss": 0.6105, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": 0.1932690441608429, | |
| "rewards/margins": 0.21003055572509766, | |
| "rewards/rejected": -0.01676151715219021, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "eval_logits/chosen": -2.233783006668091, | |
| "eval_logits/rejected": -2.228926658630371, | |
| "eval_logps/chosen": -34.01414108276367, | |
| "eval_logps/rejected": -37.53648376464844, | |
| "eval_loss": 1.0314662456512451, | |
| "eval_rewards/accuracies": 0.560215950012207, | |
| "eval_rewards/chosen": 0.010205330327153206, | |
| "eval_rewards/margins": 0.020134516060352325, | |
| "eval_rewards/rejected": -0.009929186664521694, | |
| "eval_runtime": 145.6316, | |
| "eval_samples_per_second": 2.355, | |
| "eval_steps_per_second": 0.295, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 5.576113578589035e-07, | |
| "logits/chosen": -1.9184128046035767, | |
| "logits/rejected": -1.9151279926300049, | |
| "logps/chosen": -31.325061798095703, | |
| "logps/rejected": -33.77220916748047, | |
| "loss": 0.7283, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 0.13914386928081512, | |
| "rewards/margins": 0.15417365729808807, | |
| "rewards/rejected": -0.01502978801727295, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 4.229036944380913e-07, | |
| "logits/chosen": -1.968726396560669, | |
| "logits/rejected": -1.9564485549926758, | |
| "logps/chosen": -34.366207122802734, | |
| "logps/rejected": -33.61689376831055, | |
| "loss": 0.6804, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": 0.11493507772684097, | |
| "rewards/margins": 0.1539594680070877, | |
| "rewards/rejected": -0.03902440145611763, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 3.053082288996112e-07, | |
| "logits/chosen": -2.0041136741638184, | |
| "logits/rejected": -2.002657651901245, | |
| "logps/chosen": -33.25464630126953, | |
| "logps/rejected": -32.49077606201172, | |
| "loss": 0.7724, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 0.11557211726903915, | |
| "rewards/margins": 0.12116815894842148, | |
| "rewards/rejected": -0.005596047732979059, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 2.0579377374915805e-07, | |
| "logits/chosen": -2.0917208194732666, | |
| "logits/rejected": -2.0759525299072266, | |
| "logps/chosen": -33.83209991455078, | |
| "logps/rejected": -33.08992004394531, | |
| "loss": 0.7606, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 0.17364642024040222, | |
| "rewards/margins": 0.15958845615386963, | |
| "rewards/rejected": 0.014057991094887257, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 1.2518018074041684e-07, | |
| "logits/chosen": -1.9635206460952759, | |
| "logits/rejected": -1.9625988006591797, | |
| "logps/chosen": -32.91681671142578, | |
| "logps/rejected": -32.46485137939453, | |
| "loss": 0.7913, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": 0.18264031410217285, | |
| "rewards/margins": 0.182492196559906, | |
| "rewards/rejected": 0.00014809667482040823, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 6.41315865106129e-08, | |
| "logits/chosen": -1.9189882278442383, | |
| "logits/rejected": -1.9293220043182373, | |
| "logps/chosen": -31.9406681060791, | |
| "logps/rejected": -35.306640625, | |
| "loss": 0.7574, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": 0.1320098638534546, | |
| "rewards/margins": 0.1409510374069214, | |
| "rewards/rejected": -0.008941170759499073, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 2.3150941078050325e-08, | |
| "logits/chosen": -2.0582680702209473, | |
| "logits/rejected": -2.051753520965576, | |
| "logps/chosen": -33.39839553833008, | |
| "logps/rejected": -29.18343734741211, | |
| "loss": 0.7663, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": 0.12678876519203186, | |
| "rewards/margins": 0.1052960604429245, | |
| "rewards/rejected": 0.02149270847439766, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 2.575864278703266e-09, | |
| "logits/chosen": -1.9175922870635986, | |
| "logits/rejected": -1.9197555780410767, | |
| "logps/chosen": -33.878089904785156, | |
| "logps/rejected": -30.871530532836914, | |
| "loss": 0.6999, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": 0.18047982454299927, | |
| "rewards/margins": 0.17993128299713135, | |
| "rewards/rejected": 0.0005485474830493331, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 385, | |
| "total_flos": 0.0, | |
| "train_loss": 0.8446982934877469, | |
| "train_runtime": 3252.3399, | |
| "train_samples_per_second": 0.947, | |
| "train_steps_per_second": 0.118 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 385, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "total_flos": 0.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |