| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9986824769433466, | |
| "eval_steps": 100, | |
| "global_step": 379, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.002635046113306983, | |
| "grad_norm": 1701.3284378032004, | |
| "learning_rate": 1.3157894736842104e-08, | |
| "logits/chosen": -4.685327529907227, | |
| "logits/rejected": -4.87608528137207, | |
| "logps/chosen": -207.7137451171875, | |
| "logps/rejected": -145.5098114013672, | |
| "loss": 0.6931, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.026350461133069828, | |
| "grad_norm": 1584.9120962726856, | |
| "learning_rate": 1.3157894736842104e-07, | |
| "logits/chosen": -4.499300479888916, | |
| "logits/rejected": -4.840802192687988, | |
| "logps/chosen": -223.6631317138672, | |
| "logps/rejected": -160.81097412109375, | |
| "loss": 0.7136, | |
| "rewards/accuracies": 0.4444444477558136, | |
| "rewards/chosen": 0.004151582717895508, | |
| "rewards/margins": 0.002958830911666155, | |
| "rewards/rejected": 0.001192751806229353, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.052700922266139656, | |
| "grad_norm": 955.29292445595, | |
| "learning_rate": 2.631578947368421e-07, | |
| "logits/chosen": -4.518028259277344, | |
| "logits/rejected": -4.817793846130371, | |
| "logps/chosen": -220.1512451171875, | |
| "logps/rejected": -172.69322204589844, | |
| "loss": 0.4939, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": 0.9042474627494812, | |
| "rewards/margins": 0.9315685033798218, | |
| "rewards/rejected": -0.02732105180621147, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.07905138339920949, | |
| "grad_norm": 1078.6380127502835, | |
| "learning_rate": 3.9473684210526315e-07, | |
| "logits/chosen": -4.58280086517334, | |
| "logits/rejected": -4.878857612609863, | |
| "logps/chosen": -212.6325225830078, | |
| "logps/rejected": -183.79238891601562, | |
| "loss": 0.3906, | |
| "rewards/accuracies": 0.8374999761581421, | |
| "rewards/chosen": 3.010448932647705, | |
| "rewards/margins": 2.6640021800994873, | |
| "rewards/rejected": 0.346446692943573, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.10540184453227931, | |
| "grad_norm": 1034.6942446334156, | |
| "learning_rate": 4.999575626062319e-07, | |
| "logits/chosen": -4.496463298797607, | |
| "logits/rejected": -4.832797050476074, | |
| "logps/chosen": -225.2804412841797, | |
| "logps/rejected": -177.3509063720703, | |
| "loss": 0.4804, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": 4.241217136383057, | |
| "rewards/margins": 4.989673614501953, | |
| "rewards/rejected": -0.7484563589096069, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.13175230566534915, | |
| "grad_norm": 1292.8962171873402, | |
| "learning_rate": 4.984737660598186e-07, | |
| "logits/chosen": -4.517908573150635, | |
| "logits/rejected": -4.786294937133789, | |
| "logps/chosen": -214.01718139648438, | |
| "logps/rejected": -174.09678649902344, | |
| "loss": 0.5163, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": 3.7782645225524902, | |
| "rewards/margins": 5.338944435119629, | |
| "rewards/rejected": -1.5606796741485596, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.15810276679841898, | |
| "grad_norm": 1081.6576652859876, | |
| "learning_rate": 4.948824853131236e-07, | |
| "logits/chosen": -4.719171047210693, | |
| "logits/rejected": -4.9818878173828125, | |
| "logps/chosen": -215.00869750976562, | |
| "logps/rejected": -180.1785430908203, | |
| "loss": 0.4821, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": 3.4492969512939453, | |
| "rewards/margins": 6.232473373413086, | |
| "rewards/rejected": -2.7831759452819824, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.1844532279314888, | |
| "grad_norm": 780.7579097796144, | |
| "learning_rate": 4.892141805936084e-07, | |
| "logits/chosen": -4.691411018371582, | |
| "logits/rejected": -4.9814910888671875, | |
| "logps/chosen": -222.72189331054688, | |
| "logps/rejected": -186.29788208007812, | |
| "loss": 0.5005, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": 3.7675652503967285, | |
| "rewards/margins": 6.836648464202881, | |
| "rewards/rejected": -3.0690836906433105, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.21080368906455862, | |
| "grad_norm": 1751.5015463679365, | |
| "learning_rate": 4.81516928858564e-07, | |
| "logits/chosen": -4.691437721252441, | |
| "logits/rejected": -4.943267822265625, | |
| "logps/chosen": -219.7614288330078, | |
| "logps/rejected": -179.60899353027344, | |
| "loss": 0.5615, | |
| "rewards/accuracies": 0.831250011920929, | |
| "rewards/chosen": 3.4195969104766846, | |
| "rewards/margins": 6.783135890960693, | |
| "rewards/rejected": -3.363539457321167, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.23715415019762845, | |
| "grad_norm": 1233.3333710128843, | |
| "learning_rate": 4.7185601601995784e-07, | |
| "logits/chosen": -4.569981575012207, | |
| "logits/rejected": -4.954944610595703, | |
| "logps/chosen": -212.4117431640625, | |
| "logps/rejected": -173.86911010742188, | |
| "loss": 0.4989, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": 3.7934730052948, | |
| "rewards/margins": 7.21230936050415, | |
| "rewards/rejected": -3.4188361167907715, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.2635046113306983, | |
| "grad_norm": 1244.999056136522, | |
| "learning_rate": 4.603133832077953e-07, | |
| "logits/chosen": -4.672577381134033, | |
| "logits/rejected": -4.9231157302856445, | |
| "logps/chosen": -206.11221313476562, | |
| "logps/rejected": -176.262451171875, | |
| "loss": 0.5909, | |
| "rewards/accuracies": 0.8374999761581421, | |
| "rewards/chosen": 3.6933536529541016, | |
| "rewards/margins": 6.304307460784912, | |
| "rewards/rejected": -2.6109542846679688, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.2635046113306983, | |
| "eval_logits/chosen": -4.612120628356934, | |
| "eval_logits/rejected": -4.85081672668457, | |
| "eval_logps/chosen": -396.4924011230469, | |
| "eval_logps/rejected": -512.6068115234375, | |
| "eval_loss": 6.653407096862793, | |
| "eval_rewards/accuracies": 0.27734375, | |
| "eval_rewards/chosen": -3.0089728832244873, | |
| "eval_rewards/margins": -5.8993988037109375, | |
| "eval_rewards/rejected": 2.890425682067871, | |
| "eval_runtime": 98.318, | |
| "eval_samples_per_second": 20.342, | |
| "eval_steps_per_second": 0.325, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.2898550724637681, | |
| "grad_norm": 870.1595065644425, | |
| "learning_rate": 4.4698693176863316e-07, | |
| "logits/chosen": -4.71740198135376, | |
| "logits/rejected": -4.952963829040527, | |
| "logps/chosen": -208.0568084716797, | |
| "logps/rejected": -172.92605590820312, | |
| "loss": 0.4837, | |
| "rewards/accuracies": 0.8374999761581421, | |
| "rewards/chosen": 3.114009380340576, | |
| "rewards/margins": 7.352834224700928, | |
| "rewards/rejected": -4.238823890686035, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.31620553359683795, | |
| "grad_norm": 1230.5660125025618, | |
| "learning_rate": 4.319896928940505e-07, | |
| "logits/chosen": -4.8249382972717285, | |
| "logits/rejected": -5.139795780181885, | |
| "logps/chosen": -204.9242401123047, | |
| "logps/rejected": -181.16221618652344, | |
| "loss": 0.6657, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": 4.0227861404418945, | |
| "rewards/margins": 6.973240852355957, | |
| "rewards/rejected": -2.950455904006958, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.3425559947299078, | |
| "grad_norm": 1400.2090737238004, | |
| "learning_rate": 4.1544886892205354e-07, | |
| "logits/chosen": -4.680369853973389, | |
| "logits/rejected": -4.979363441467285, | |
| "logps/chosen": -218.03897094726562, | |
| "logps/rejected": -192.44757080078125, | |
| "loss": 0.4732, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": 3.1923911571502686, | |
| "rewards/margins": 8.124491691589355, | |
| "rewards/rejected": -4.932101249694824, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.3689064558629776, | |
| "grad_norm": 1284.6303869923338, | |
| "learning_rate": 3.975047544428254e-07, | |
| "logits/chosen": -4.738985061645508, | |
| "logits/rejected": -4.995828628540039, | |
| "logps/chosen": -200.0480499267578, | |
| "logps/rejected": -173.74069213867188, | |
| "loss": 0.7249, | |
| "rewards/accuracies": 0.8687499761581421, | |
| "rewards/chosen": 3.118265151977539, | |
| "rewards/margins": 6.879487037658691, | |
| "rewards/rejected": -3.7612221240997314, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.3952569169960474, | |
| "grad_norm": 1087.9280552765, | |
| "learning_rate": 3.78309546359696e-07, | |
| "logits/chosen": -4.816591739654541, | |
| "logits/rejected": -5.046236991882324, | |
| "logps/chosen": -197.75784301757812, | |
| "logps/rejected": -187.12631225585938, | |
| "loss": 0.6191, | |
| "rewards/accuracies": 0.918749988079071, | |
| "rewards/chosen": 1.7183030843734741, | |
| "rewards/margins": 8.374044418334961, | |
| "rewards/rejected": -6.655740261077881, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.42160737812911725, | |
| "grad_norm": 870.0436658351372, | |
| "learning_rate": 3.580260529980584e-07, | |
| "logits/chosen": -4.622679233551025, | |
| "logits/rejected": -4.90102481842041, | |
| "logps/chosen": -210.0752410888672, | |
| "logps/rejected": -182.36378479003906, | |
| "loss": 0.5743, | |
| "rewards/accuracies": 0.8374999761581421, | |
| "rewards/chosen": 2.5026469230651855, | |
| "rewards/margins": 6.609257698059082, | |
| "rewards/rejected": -4.106611251831055, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.4479578392621871, | |
| "grad_norm": 1109.7545725706777, | |
| "learning_rate": 3.36826313211205e-07, | |
| "logits/chosen": -4.862623691558838, | |
| "logits/rejected": -5.1108198165893555, | |
| "logps/chosen": -195.3279266357422, | |
| "logps/rejected": -170.62802124023438, | |
| "loss": 0.6483, | |
| "rewards/accuracies": 0.8187500238418579, | |
| "rewards/chosen": 2.5267701148986816, | |
| "rewards/margins": 7.295065402984619, | |
| "rewards/rejected": -4.768294811248779, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.4743083003952569, | |
| "grad_norm": 1174.4955529247104, | |
| "learning_rate": 3.14890137195437e-07, | |
| "logits/chosen": -4.831478595733643, | |
| "logits/rejected": -5.037031173706055, | |
| "logps/chosen": -211.2603302001953, | |
| "logps/rejected": -190.32931518554688, | |
| "loss": 0.5829, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": 1.8433873653411865, | |
| "rewards/margins": 7.003817558288574, | |
| "rewards/rejected": -5.160429954528809, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.5006587615283268, | |
| "grad_norm": 1405.2874598551575, | |
| "learning_rate": 2.9240358139084013e-07, | |
| "logits/chosen": -4.9419050216674805, | |
| "logits/rejected": -5.200289726257324, | |
| "logps/chosen": -209.5860595703125, | |
| "logps/rejected": -188.95069885253906, | |
| "loss": 0.5485, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": 1.218836784362793, | |
| "rewards/margins": 8.241748809814453, | |
| "rewards/rejected": -7.022912502288818, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.5270092226613966, | |
| "grad_norm": 1006.939051019727, | |
| "learning_rate": 2.695573704031885e-07, | |
| "logits/chosen": -4.7473955154418945, | |
| "logits/rejected": -5.000621795654297, | |
| "logps/chosen": -217.6959686279297, | |
| "logps/rejected": -190.3670196533203, | |
| "loss": 0.7239, | |
| "rewards/accuracies": 0.8374999761581421, | |
| "rewards/chosen": 3.8086190223693848, | |
| "rewards/margins": 6.961747646331787, | |
| "rewards/rejected": -3.1531288623809814, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.5270092226613966, | |
| "eval_logits/chosen": -4.771502494812012, | |
| "eval_logits/rejected": -4.989596843719482, | |
| "eval_logps/chosen": -396.1126708984375, | |
| "eval_logps/rejected": -511.37469482421875, | |
| "eval_loss": 8.072031021118164, | |
| "eval_rewards/accuracies": 0.2734375, | |
| "eval_rewards/chosen": -2.8191022872924805, | |
| "eval_rewards/margins": -6.325590133666992, | |
| "eval_rewards/rejected": 3.5064878463745117, | |
| "eval_runtime": 97.9327, | |
| "eval_samples_per_second": 20.422, | |
| "eval_steps_per_second": 0.327, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.5533596837944664, | |
| "grad_norm": 1103.8948285514384, | |
| "learning_rate": 2.465452793317865e-07, | |
| "logits/chosen": -4.7866034507751465, | |
| "logits/rejected": -5.081458568572998, | |
| "logps/chosen": -228.73922729492188, | |
| "logps/rejected": -201.91409301757812, | |
| "loss": 0.5681, | |
| "rewards/accuracies": 0.8687499761581421, | |
| "rewards/chosen": 2.7976126670837402, | |
| "rewards/margins": 7.8736419677734375, | |
| "rewards/rejected": -5.076028823852539, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.5797101449275363, | |
| "grad_norm": 715.6592810091468, | |
| "learning_rate": 2.2356249022388789e-07, | |
| "logits/chosen": -4.758913516998291, | |
| "logits/rejected": -5.04398250579834, | |
| "logps/chosen": -205.92568969726562, | |
| "logps/rejected": -170.3586883544922, | |
| "loss": 0.4895, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": 4.249292373657227, | |
| "rewards/margins": 7.541260719299316, | |
| "rewards/rejected": -3.291968822479248, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.6060606060606061, | |
| "grad_norm": 944.8747719910025, | |
| "learning_rate": 2.0080393659578038e-07, | |
| "logits/chosen": -4.764594078063965, | |
| "logits/rejected": -5.100175857543945, | |
| "logps/chosen": -210.0486297607422, | |
| "logps/rejected": -179.91928100585938, | |
| "loss": 0.5197, | |
| "rewards/accuracies": 0.8687499761581421, | |
| "rewards/chosen": 5.489472389221191, | |
| "rewards/margins": 8.828967094421387, | |
| "rewards/rejected": -3.3394947052001953, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.6324110671936759, | |
| "grad_norm": 1267.6123006349214, | |
| "learning_rate": 1.7846265006183976e-07, | |
| "logits/chosen": -4.821089744567871, | |
| "logits/rejected": -5.0871124267578125, | |
| "logps/chosen": -211.7589874267578, | |
| "logps/rejected": -181.33819580078125, | |
| "loss": 0.5444, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -0.23131565749645233, | |
| "rewards/margins": 6.660643577575684, | |
| "rewards/rejected": -6.891959190368652, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.6587615283267457, | |
| "grad_norm": 911.8592996107284, | |
| "learning_rate": 1.5672812309497722e-07, | |
| "logits/chosen": -4.781493663787842, | |
| "logits/rejected": -5.052074432373047, | |
| "logps/chosen": -201.82290649414062, | |
| "logps/rejected": -175.06719970703125, | |
| "loss": 0.5887, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": 2.856235980987549, | |
| "rewards/margins": 8.613394737243652, | |
| "rewards/rejected": -5.757159233093262, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.6851119894598156, | |
| "grad_norm": 840.8020636445922, | |
| "learning_rate": 1.357847018050843e-07, | |
| "logits/chosen": -4.674568176269531, | |
| "logits/rejected": -4.961749076843262, | |
| "logps/chosen": -232.63119506835938, | |
| "logps/rejected": -204.2981719970703, | |
| "loss": 0.6421, | |
| "rewards/accuracies": 0.8062499761581421, | |
| "rewards/chosen": 2.906367778778076, | |
| "rewards/margins": 7.688606262207031, | |
| "rewards/rejected": -4.782238483428955, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.7114624505928854, | |
| "grad_norm": 787.2425305884327, | |
| "learning_rate": 1.1581002236747328e-07, | |
| "logits/chosen": -4.637770652770996, | |
| "logits/rejected": -4.975251197814941, | |
| "logps/chosen": -190.36253356933594, | |
| "logps/rejected": -166.21788024902344, | |
| "loss": 0.5113, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": 2.2253878116607666, | |
| "rewards/margins": 8.868097305297852, | |
| "rewards/rejected": -6.6427106857299805, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.7378129117259552, | |
| "grad_norm": 1064.0700290859438, | |
| "learning_rate": 9.697350436308427e-08, | |
| "logits/chosen": -4.585692405700684, | |
| "logits/rejected": -4.871885776519775, | |
| "logps/chosen": -229.0143585205078, | |
| "logps/rejected": -194.08761596679688, | |
| "loss": 0.5252, | |
| "rewards/accuracies": 0.8374999761581421, | |
| "rewards/chosen": 3.0419421195983887, | |
| "rewards/margins": 8.473891258239746, | |
| "rewards/rejected": -5.431949138641357, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.764163372859025, | |
| "grad_norm": 839.1911415542294, | |
| "learning_rate": 7.943491380952188e-08, | |
| "logits/chosen": -4.828024864196777, | |
| "logits/rejected": -5.060397148132324, | |
| "logps/chosen": -196.42611694335938, | |
| "logps/rejected": -171.01446533203125, | |
| "loss": 0.4822, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": 4.435191631317139, | |
| "rewards/margins": 8.396050453186035, | |
| "rewards/rejected": -3.9608588218688965, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.7905138339920948, | |
| "grad_norm": 887.4606387846492, | |
| "learning_rate": 6.334300807088508e-08, | |
| "logits/chosen": -4.608688831329346, | |
| "logits/rejected": -4.9593987464904785, | |
| "logps/chosen": -195.76962280273438, | |
| "logps/rejected": -164.97048950195312, | |
| "loss": 0.5556, | |
| "rewards/accuracies": 0.893750011920929, | |
| "rewards/chosen": 3.875316619873047, | |
| "rewards/margins": 8.369561195373535, | |
| "rewards/rejected": -4.4942450523376465, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.7905138339920948, | |
| "eval_logits/chosen": -4.660388469696045, | |
| "eval_logits/rejected": -4.890798091888428, | |
| "eval_logps/chosen": -399.5094909667969, | |
| "eval_logps/rejected": -517.530029296875, | |
| "eval_loss": 6.923000335693359, | |
| "eval_rewards/accuracies": 0.31640625, | |
| "eval_rewards/chosen": -4.517519474029541, | |
| "eval_rewards/margins": -4.946353435516357, | |
| "eval_rewards/rejected": 0.42883408069610596, | |
| "eval_runtime": 99.1917, | |
| "eval_samples_per_second": 20.163, | |
| "eval_steps_per_second": 0.323, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.8168642951251647, | |
| "grad_norm": 1030.5057638496492, | |
| "learning_rate": 4.8834274139883084e-08, | |
| "logits/chosen": -4.663365840911865, | |
| "logits/rejected": -5.023074150085449, | |
| "logps/chosen": -201.54336547851562, | |
| "logps/rejected": -171.71324157714844, | |
| "loss": 0.533, | |
| "rewards/accuracies": 0.8812500238418579, | |
| "rewards/chosen": 4.074584007263184, | |
| "rewards/margins": 8.798391342163086, | |
| "rewards/rejected": -4.723808765411377, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.8432147562582345, | |
| "grad_norm": 1279.040936993484, | |
| "learning_rate": 3.60317709937693e-08, | |
| "logits/chosen": -4.707262992858887, | |
| "logits/rejected": -5.011012077331543, | |
| "logps/chosen": -223.0199432373047, | |
| "logps/rejected": -183.51393127441406, | |
| "loss": 0.5239, | |
| "rewards/accuracies": 0.8187500238418579, | |
| "rewards/chosen": 3.1946961879730225, | |
| "rewards/margins": 7.443505764007568, | |
| "rewards/rejected": -4.248808860778809, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.8695652173913043, | |
| "grad_norm": 790.9511672313881, | |
| "learning_rate": 2.5044085842905683e-08, | |
| "logits/chosen": -4.707150459289551, | |
| "logits/rejected": -4.967694282531738, | |
| "logps/chosen": -208.9776153564453, | |
| "logps/rejected": -184.1087188720703, | |
| "loss": 0.5674, | |
| "rewards/accuracies": 0.918749988079071, | |
| "rewards/chosen": 4.723270893096924, | |
| "rewards/margins": 9.567978858947754, | |
| "rewards/rejected": -4.844708442687988, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.8959156785243741, | |
| "grad_norm": 1075.929120672392, | |
| "learning_rate": 1.5964413124758493e-08, | |
| "logits/chosen": -4.645999431610107, | |
| "logits/rejected": -4.947402477264404, | |
| "logps/chosen": -212.19015502929688, | |
| "logps/rejected": -185.76678466796875, | |
| "loss": 0.5578, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": 4.216076374053955, | |
| "rewards/margins": 8.17945671081543, | |
| "rewards/rejected": -3.9633796215057373, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.922266139657444, | |
| "grad_norm": 1055.0113319220882, | |
| "learning_rate": 8.869764055041501e-09, | |
| "logits/chosen": -4.714905738830566, | |
| "logits/rejected": -4.937991619110107, | |
| "logps/chosen": -215.5680389404297, | |
| "logps/rejected": -200.6318359375, | |
| "loss": 0.5113, | |
| "rewards/accuracies": 0.793749988079071, | |
| "rewards/chosen": 2.9399425983428955, | |
| "rewards/margins": 7.131015777587891, | |
| "rewards/rejected": -4.191073417663574, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.9486166007905138, | |
| "grad_norm": 1003.3923870155849, | |
| "learning_rate": 3.82031344036729e-09, | |
| "logits/chosen": -4.614500999450684, | |
| "logits/rejected": -4.9195661544799805, | |
| "logps/chosen": -216.197998046875, | |
| "logps/rejected": -186.8596649169922, | |
| "loss": 0.4994, | |
| "rewards/accuracies": 0.8187500238418579, | |
| "rewards/chosen": 2.650672197341919, | |
| "rewards/margins": 6.149393081665039, | |
| "rewards/rejected": -3.49872088432312, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.9749670619235836, | |
| "grad_norm": 1021.842139855702, | |
| "learning_rate": 8.588892925590063e-10, | |
| "logits/chosen": -4.69917631149292, | |
| "logits/rejected": -5.054296016693115, | |
| "logps/chosen": -212.79714965820312, | |
| "logps/rejected": -171.7442169189453, | |
| "loss": 0.5535, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": 3.2063796520233154, | |
| "rewards/margins": 8.728368759155273, | |
| "rewards/rejected": -5.521987438201904, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.9986824769433466, | |
| "step": 379, | |
| "total_flos": 0.0, | |
| "train_loss": 0.5517442987587962, | |
| "train_runtime": 6181.8185, | |
| "train_samples_per_second": 7.85, | |
| "train_steps_per_second": 0.061 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 379, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |