| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9987943737441393, | |
| "eval_steps": 400, | |
| "global_step": 466, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.010716677829872739, | |
| "grad_norm": 38.485095124933046, | |
| "learning_rate": 5.3191489361702123e-08, | |
| "logits/chosen": -2.398301601409912, | |
| "logits/rejected": -2.3920369148254395, | |
| "logps/chosen": -0.5667670965194702, | |
| "logps/rejected": -0.5555010437965393, | |
| "loss": 1.5468, | |
| "rewards/accuracies": 0.4375, | |
| "rewards/chosen": -0.5667670965194702, | |
| "rewards/margins": -0.011266030371189117, | |
| "rewards/rejected": -0.5555010437965393, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.021433355659745478, | |
| "grad_norm": 18.322790262213086, | |
| "learning_rate": 1.0638297872340425e-07, | |
| "logits/chosen": -2.4026951789855957, | |
| "logits/rejected": -2.3731987476348877, | |
| "logps/chosen": -0.5517200231552124, | |
| "logps/rejected": -0.5781593918800354, | |
| "loss": 1.554, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.5517200231552124, | |
| "rewards/margins": 0.026439335197210312, | |
| "rewards/rejected": -0.5781593918800354, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.032150033489618215, | |
| "grad_norm": 17.062192600734857, | |
| "learning_rate": 1.5957446808510638e-07, | |
| "logits/chosen": -2.4439353942871094, | |
| "logits/rejected": -2.449899196624756, | |
| "logps/chosen": -0.5637143850326538, | |
| "logps/rejected": -0.567409336566925, | |
| "loss": 1.5617, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.5637143850326538, | |
| "rewards/margins": 0.0036949075292795897, | |
| "rewards/rejected": -0.567409336566925, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.042866711319490956, | |
| "grad_norm": 17.524849197474367, | |
| "learning_rate": 2.127659574468085e-07, | |
| "logits/chosen": -2.384251117706299, | |
| "logits/rejected": -2.3945975303649902, | |
| "logps/chosen": -0.5456732511520386, | |
| "logps/rejected": -0.5427749752998352, | |
| "loss": 1.5321, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.5456732511520386, | |
| "rewards/margins": -0.002898297505453229, | |
| "rewards/rejected": -0.5427749752998352, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0535833891493637, | |
| "grad_norm": 14.321312880487431, | |
| "learning_rate": 2.659574468085106e-07, | |
| "logits/chosen": -2.279252529144287, | |
| "logits/rejected": -2.2811872959136963, | |
| "logps/chosen": -0.5260757803916931, | |
| "logps/rejected": -0.543364405632019, | |
| "loss": 1.53, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.5260757803916931, | |
| "rewards/margins": 0.017288658767938614, | |
| "rewards/rejected": -0.543364405632019, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.06430006697923643, | |
| "grad_norm": 19.544853384077058, | |
| "learning_rate": 3.1914893617021275e-07, | |
| "logits/chosen": -2.3907198905944824, | |
| "logits/rejected": -2.403733015060425, | |
| "logps/chosen": -0.5487326979637146, | |
| "logps/rejected": -0.5386630892753601, | |
| "loss": 1.5515, | |
| "rewards/accuracies": 0.46875, | |
| "rewards/chosen": -0.5487326979637146, | |
| "rewards/margins": -0.010069593787193298, | |
| "rewards/rejected": -0.5386630892753601, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07501674480910918, | |
| "grad_norm": 24.103155081211728, | |
| "learning_rate": 3.7234042553191484e-07, | |
| "logits/chosen": -2.452573299407959, | |
| "logits/rejected": -2.423945665359497, | |
| "logps/chosen": -0.5365626811981201, | |
| "logps/rejected": -0.5920503735542297, | |
| "loss": 1.5478, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -0.5365626811981201, | |
| "rewards/margins": 0.055487751960754395, | |
| "rewards/rejected": -0.5920503735542297, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.08573342263898191, | |
| "grad_norm": 21.06293839287325, | |
| "learning_rate": 4.25531914893617e-07, | |
| "logits/chosen": -2.3676323890686035, | |
| "logits/rejected": -2.349027156829834, | |
| "logps/chosen": -0.5672295093536377, | |
| "logps/rejected": -0.5381959676742554, | |
| "loss": 1.5581, | |
| "rewards/accuracies": 0.42500001192092896, | |
| "rewards/chosen": -0.5672295093536377, | |
| "rewards/margins": -0.02903353050351143, | |
| "rewards/rejected": -0.5381959676742554, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.09645010046885466, | |
| "grad_norm": 17.138121888291746, | |
| "learning_rate": 4.787234042553192e-07, | |
| "logits/chosen": -2.334918260574341, | |
| "logits/rejected": -2.3297195434570312, | |
| "logps/chosen": -0.5420322418212891, | |
| "logps/rejected": -0.560600757598877, | |
| "loss": 1.5426, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.5420322418212891, | |
| "rewards/margins": 0.01856859028339386, | |
| "rewards/rejected": -0.560600757598877, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.1071667782987274, | |
| "grad_norm": 16.650052527651823, | |
| "learning_rate": 4.999367579541646e-07, | |
| "logits/chosen": -2.3611702919006348, | |
| "logits/rejected": -2.3566925525665283, | |
| "logps/chosen": -0.542779803276062, | |
| "logps/rejected": -0.5264804363250732, | |
| "loss": 1.5287, | |
| "rewards/accuracies": 0.45625001192092896, | |
| "rewards/chosen": -0.542779803276062, | |
| "rewards/margins": -0.016299355775117874, | |
| "rewards/rejected": -0.5264804363250732, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.11788345612860013, | |
| "grad_norm": 25.231872930806986, | |
| "learning_rate": 4.99550394651012e-07, | |
| "logits/chosen": -2.4198124408721924, | |
| "logits/rejected": -2.4228649139404297, | |
| "logps/chosen": -0.5739614367485046, | |
| "logps/rejected": -0.5911067128181458, | |
| "loss": 1.5407, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -0.5739614367485046, | |
| "rewards/margins": 0.017145240679383278, | |
| "rewards/rejected": -0.5911067128181458, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.12860013395847286, | |
| "grad_norm": 27.654893603033948, | |
| "learning_rate": 4.988133448023071e-07, | |
| "logits/chosen": -2.3333356380462646, | |
| "logits/rejected": -2.342374324798584, | |
| "logps/chosen": -0.505652129650116, | |
| "logps/rejected": -0.52570641040802, | |
| "loss": 1.5384, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.505652129650116, | |
| "rewards/margins": 0.020054226741194725, | |
| "rewards/rejected": -0.52570641040802, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.13931681178834562, | |
| "grad_norm": 18.301707545724692, | |
| "learning_rate": 4.977266441646379e-07, | |
| "logits/chosen": -2.3869385719299316, | |
| "logits/rejected": -2.391303777694702, | |
| "logps/chosen": -0.5198084712028503, | |
| "logps/rejected": -0.5502195358276367, | |
| "loss": 1.5466, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.5198084712028503, | |
| "rewards/margins": 0.030411023646593094, | |
| "rewards/rejected": -0.5502195358276367, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.15003348961821836, | |
| "grad_norm": 19.461976097246225, | |
| "learning_rate": 4.962918198495654e-07, | |
| "logits/chosen": -2.2925384044647217, | |
| "logits/rejected": -2.296604633331299, | |
| "logps/chosen": -0.5172106027603149, | |
| "logps/rejected": -0.5665396451950073, | |
| "loss": 1.5322, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.5172106027603149, | |
| "rewards/margins": 0.04932904243469238, | |
| "rewards/rejected": -0.5665396451950073, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.1607501674480911, | |
| "grad_norm": 16.227392616134246, | |
| "learning_rate": 4.945108881776135e-07, | |
| "logits/chosen": -2.3317081928253174, | |
| "logits/rejected": -2.3258421421051025, | |
| "logps/chosen": -0.48806238174438477, | |
| "logps/rejected": -0.5191291570663452, | |
| "loss": 1.5323, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.48806238174438477, | |
| "rewards/margins": 0.031066760420799255, | |
| "rewards/rejected": -0.5191291570663452, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.17146684527796383, | |
| "grad_norm": 18.668922205376074, | |
| "learning_rate": 4.923863518447879e-07, | |
| "logits/chosen": -2.304633617401123, | |
| "logits/rejected": -2.3267812728881836, | |
| "logps/chosen": -0.4988403916358948, | |
| "logps/rejected": -0.5690027475357056, | |
| "loss": 1.5304, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.4988403916358948, | |
| "rewards/margins": 0.07016237080097198, | |
| "rewards/rejected": -0.5690027475357056, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.18218352310783656, | |
| "grad_norm": 21.78609949315303, | |
| "learning_rate": 4.899211964056035e-07, | |
| "logits/chosen": -2.338728427886963, | |
| "logits/rejected": -2.331895112991333, | |
| "logps/chosen": -0.5135633945465088, | |
| "logps/rejected": -0.5660439729690552, | |
| "loss": 1.5365, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.5135633945465088, | |
| "rewards/margins": 0.052480556070804596, | |
| "rewards/rejected": -0.5660439729690552, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.19290020093770932, | |
| "grad_norm": 18.81785585703593, | |
| "learning_rate": 4.871188860775643e-07, | |
| "logits/chosen": -2.288607120513916, | |
| "logits/rejected": -2.292832374572754, | |
| "logps/chosen": -0.5013225674629211, | |
| "logps/rejected": -0.5400410890579224, | |
| "loss": 1.5369, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -0.5013225674629211, | |
| "rewards/margins": 0.03871860355138779, | |
| "rewards/rejected": -0.5400410890579224, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.20361687876758205, | |
| "grad_norm": 18.284986731616815, | |
| "learning_rate": 4.839833588729897e-07, | |
| "logits/chosen": -2.33103609085083, | |
| "logits/rejected": -2.3427205085754395, | |
| "logps/chosen": -0.5798801779747009, | |
| "logps/rejected": -0.5994023680686951, | |
| "loss": 1.5235, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -0.5798801779747009, | |
| "rewards/margins": 0.01952212117612362, | |
| "rewards/rejected": -0.5994023680686951, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.2143335565974548, | |
| "grad_norm": 24.996484906718788, | |
| "learning_rate": 4.80519021065031e-07, | |
| "logits/chosen": -2.3483448028564453, | |
| "logits/rejected": -2.3573968410491943, | |
| "logps/chosen": -0.5851219296455383, | |
| "logps/rejected": -0.6033455729484558, | |
| "loss": 1.5464, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -0.5851219296455383, | |
| "rewards/margins": 0.018223673105239868, | |
| "rewards/rejected": -0.6033455729484558, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.22505023442732752, | |
| "grad_norm": 15.841869984614812, | |
| "learning_rate": 4.767307409956528e-07, | |
| "logits/chosen": -2.3424832820892334, | |
| "logits/rejected": -2.342660903930664, | |
| "logps/chosen": -0.5200175046920776, | |
| "logps/rejected": -0.5671975612640381, | |
| "loss": 1.5177, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -0.5200175046920776, | |
| "rewards/margins": 0.04718000069260597, | |
| "rewards/rejected": -0.5671975612640381, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.23576691225720026, | |
| "grad_norm": 26.162342495475848, | |
| "learning_rate": 4.726238422342805e-07, | |
| "logits/chosen": -2.370981216430664, | |
| "logits/rejected": -2.377699375152588, | |
| "logps/chosen": -0.5119336247444153, | |
| "logps/rejected": -0.5585616230964661, | |
| "loss": 1.5421, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.5119336247444153, | |
| "rewards/margins": 0.04662800580263138, | |
| "rewards/rejected": -0.5585616230964661, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.24648359008707302, | |
| "grad_norm": 22.223934331525054, | |
| "learning_rate": 4.6820409609673027e-07, | |
| "logits/chosen": -2.445955276489258, | |
| "logits/rejected": -2.439049005508423, | |
| "logps/chosen": -0.507788360118866, | |
| "logps/rejected": -0.5285546779632568, | |
| "loss": 1.573, | |
| "rewards/accuracies": 0.48750001192092896, | |
| "rewards/chosen": -0.507788360118866, | |
| "rewards/margins": 0.02076634392142296, | |
| "rewards/rejected": -0.5285546779632568, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.2572002679169457, | |
| "grad_norm": 17.337249756046997, | |
| "learning_rate": 4.6347771353493173e-07, | |
| "logits/chosen": -2.507343053817749, | |
| "logits/rejected": -2.4980244636535645, | |
| "logps/chosen": -0.5571922063827515, | |
| "logps/rejected": -0.5488871335983276, | |
| "loss": 1.5546, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -0.5571922063827515, | |
| "rewards/margins": -0.008305097930133343, | |
| "rewards/rejected": -0.5488871335983276, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.2679169457468185, | |
| "grad_norm": 18.61479199807285, | |
| "learning_rate": 4.5845133640884217e-07, | |
| "logits/chosen": -2.496476173400879, | |
| "logits/rejected": -2.5013513565063477, | |
| "logps/chosen": -0.5598434209823608, | |
| "logps/rejected": -0.5717306733131409, | |
| "loss": 1.5472, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.5598434209823608, | |
| "rewards/margins": 0.01188725233078003, | |
| "rewards/rejected": -0.5717306733131409, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.27863362357669125, | |
| "grad_norm": 24.242840152077708, | |
| "learning_rate": 4.531320281528169e-07, | |
| "logits/chosen": -2.577277660369873, | |
| "logits/rejected": -2.5614867210388184, | |
| "logps/chosen": -0.5257032513618469, | |
| "logps/rejected": -0.5607699155807495, | |
| "loss": 1.5206, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.5257032513618469, | |
| "rewards/margins": 0.03506666049361229, | |
| "rewards/rejected": -0.5607699155807495, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.289350301406564, | |
| "grad_norm": 28.052518647744463, | |
| "learning_rate": 4.4752726384955293e-07, | |
| "logits/chosen": -2.5903048515319824, | |
| "logits/rejected": -2.6012890338897705, | |
| "logps/chosen": -0.5207422971725464, | |
| "logps/rejected": -0.5546741485595703, | |
| "loss": 1.5394, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.5207422971725464, | |
| "rewards/margins": 0.03393186628818512, | |
| "rewards/rejected": -0.5546741485595703, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.3000669792364367, | |
| "grad_norm": 20.236382854852724, | |
| "learning_rate": 4.4164491972555294e-07, | |
| "logits/chosen": -2.6124188899993896, | |
| "logits/rejected": -2.5999808311462402, | |
| "logps/chosen": -0.5296235084533691, | |
| "logps/rejected": -0.5590516924858093, | |
| "loss": 1.5202, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -0.5296235084533691, | |
| "rewards/margins": 0.02942819520831108, | |
| "rewards/rejected": -0.5590516924858093, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.31078365706630945, | |
| "grad_norm": 18.380103954877008, | |
| "learning_rate": 4.3549326208287393e-07, | |
| "logits/chosen": -2.710564613342285, | |
| "logits/rejected": -2.68730092048645, | |
| "logps/chosen": -0.5258200764656067, | |
| "logps/rejected": -0.5745820999145508, | |
| "loss": 1.5158, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -0.5258200764656067, | |
| "rewards/margins": 0.0487620048224926, | |
| "rewards/rejected": -0.5745820999145508, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.3215003348961822, | |
| "grad_norm": 19.20157418312748, | |
| "learning_rate": 4.2908093568271187e-07, | |
| "logits/chosen": -2.70224666595459, | |
| "logits/rejected": -2.694885730743408, | |
| "logps/chosen": -0.5079634189605713, | |
| "logps/rejected": -0.5393082499504089, | |
| "loss": 1.5455, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.5079634189605713, | |
| "rewards/margins": 0.03134483844041824, | |
| "rewards/rejected": -0.5393082499504089, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3322170127260549, | |
| "grad_norm": 17.41360540258233, | |
| "learning_rate": 4.2241695159714843e-07, | |
| "logits/chosen": -2.793240785598755, | |
| "logits/rejected": -2.801339626312256, | |
| "logps/chosen": -0.5376655459403992, | |
| "logps/rejected": -0.5706537365913391, | |
| "loss": 1.5278, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.5376655459403992, | |
| "rewards/margins": 0.03298817574977875, | |
| "rewards/rejected": -0.5706537365913391, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.34293369055592765, | |
| "grad_norm": 21.649135642065247, | |
| "learning_rate": 4.1551067454613074e-07, | |
| "logits/chosen": -2.802827835083008, | |
| "logits/rejected": -2.78732967376709, | |
| "logps/chosen": -0.558406412601471, | |
| "logps/rejected": -0.6224438548088074, | |
| "loss": 1.5318, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.558406412601471, | |
| "rewards/margins": 0.06403747946023941, | |
| "rewards/rejected": -0.6224438548088074, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.3536503683858004, | |
| "grad_norm": 20.48435877400994, | |
| "learning_rate": 4.0837180973747876e-07, | |
| "logits/chosen": -2.9101483821868896, | |
| "logits/rejected": -2.8818156719207764, | |
| "logps/chosen": -0.5387953519821167, | |
| "logps/rejected": -0.571646511554718, | |
| "loss": 1.5233, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.5387953519821167, | |
| "rewards/margins": 0.032851073890924454, | |
| "rewards/rejected": -0.571646511554718, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.3643670462156731, | |
| "grad_norm": 17.62876339336723, | |
| "learning_rate": 4.0101038922841455e-07, | |
| "logits/chosen": -2.9059550762176514, | |
| "logits/rejected": -2.9104220867156982, | |
| "logps/chosen": -0.5236160159111023, | |
| "logps/rejected": -0.5999457836151123, | |
| "loss": 1.5112, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.5236160159111023, | |
| "rewards/margins": 0.07632973045110703, | |
| "rewards/rejected": -0.5999457836151123, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.3750837240455459, | |
| "grad_norm": 20.453674701254528, | |
| "learning_rate": 3.934367578277783e-07, | |
| "logits/chosen": -2.9715123176574707, | |
| "logits/rejected": -2.953988790512085, | |
| "logps/chosen": -0.5574652552604675, | |
| "logps/rejected": -0.5864735841751099, | |
| "loss": 1.5348, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.5574652552604675, | |
| "rewards/margins": 0.029008299112319946, | |
| "rewards/rejected": -0.5864735841751099, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.38580040187541864, | |
| "grad_norm": 22.621519510630307, | |
| "learning_rate": 3.856615585587434e-07, | |
| "logits/chosen": -3.067772388458252, | |
| "logits/rejected": -3.0693717002868652, | |
| "logps/chosen": -0.5790558457374573, | |
| "logps/rejected": -0.6112508177757263, | |
| "loss": 1.545, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.5790558457374573, | |
| "rewards/margins": 0.03219496086239815, | |
| "rewards/rejected": -0.6112508177757263, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.3965170797052914, | |
| "grad_norm": 27.765398609724794, | |
| "learning_rate": 3.776957177024581e-07, | |
| "logits/chosen": -3.070247173309326, | |
| "logits/rejected": -3.0919833183288574, | |
| "logps/chosen": -0.5750337839126587, | |
| "logps/rejected": -0.6032044887542725, | |
| "loss": 1.5218, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.5750337839126587, | |
| "rewards/margins": 0.028170805424451828, | |
| "rewards/rejected": -0.6032044887542725, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.4072337575351641, | |
| "grad_norm": 19.099968753214323, | |
| "learning_rate": 3.69550429443633e-07, | |
| "logits/chosen": -2.944047451019287, | |
| "logits/rejected": -2.959052324295044, | |
| "logps/chosen": -0.5526596307754517, | |
| "logps/rejected": -0.7156730890274048, | |
| "loss": 1.5268, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.5526596307754517, | |
| "rewards/margins": 0.16301342844963074, | |
| "rewards/rejected": -0.7156730890274048, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.41795043536503684, | |
| "grad_norm": 24.556600066756868, | |
| "learning_rate": 3.612371401396502e-07, | |
| "logits/chosen": -3.195617198944092, | |
| "logits/rejected": -3.168480396270752, | |
| "logps/chosen": -0.515397846698761, | |
| "logps/rejected": -0.5484700798988342, | |
| "loss": 1.5305, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.515397846698761, | |
| "rewards/margins": 0.03307227045297623, | |
| "rewards/rejected": -0.5484700798988342, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.4286671131949096, | |
| "grad_norm": 20.445564921711252, | |
| "learning_rate": 3.527675322353011e-07, | |
| "logits/chosen": -3.071345567703247, | |
| "logits/rejected": -3.0912113189697266, | |
| "logps/chosen": -0.4999586045742035, | |
| "logps/rejected": -0.5286678075790405, | |
| "loss": 1.5201, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -0.4999586045742035, | |
| "rewards/margins": 0.028709203004837036, | |
| "rewards/rejected": -0.5286678075790405, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.4393837910247823, | |
| "grad_norm": 28.017877985161014, | |
| "learning_rate": 3.4415350784575693e-07, | |
| "logits/chosen": -3.2119345664978027, | |
| "logits/rejected": -3.2098147869110107, | |
| "logps/chosen": -0.48476672172546387, | |
| "logps/rejected": -0.5175997018814087, | |
| "loss": 1.5335, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.48476672172546387, | |
| "rewards/margins": 0.032833002507686615, | |
| "rewards/rejected": -0.5175997018814087, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.45010046885465504, | |
| "grad_norm": 20.446366205479023, | |
| "learning_rate": 3.3540717203084224e-07, | |
| "logits/chosen": -3.2167842388153076, | |
| "logits/rejected": -3.209538698196411, | |
| "logps/chosen": -0.5124386548995972, | |
| "logps/rejected": -0.5561140179634094, | |
| "loss": 1.5035, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.5124386548995972, | |
| "rewards/margins": 0.04367532581090927, | |
| "rewards/rejected": -0.5561140179634094, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.4608171466845278, | |
| "grad_norm": 20.249310261680897, | |
| "learning_rate": 3.265408157841153e-07, | |
| "logits/chosen": -3.2260100841522217, | |
| "logits/rejected": -3.2222793102264404, | |
| "logps/chosen": -0.5612623691558838, | |
| "logps/rejected": -0.6017634272575378, | |
| "loss": 1.4952, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -0.5612623691558838, | |
| "rewards/margins": 0.040501099079847336, | |
| "rewards/rejected": -0.6017634272575378, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.4715338245144005, | |
| "grad_norm": 28.781450630078922, | |
| "learning_rate": 3.1756689876066187e-07, | |
| "logits/chosen": -3.26216459274292, | |
| "logits/rejected": -3.2409801483154297, | |
| "logps/chosen": -0.5308222770690918, | |
| "logps/rejected": -0.609308660030365, | |
| "loss": 1.5377, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.5308222770690918, | |
| "rewards/margins": 0.078486368060112, | |
| "rewards/rejected": -0.609308660030365, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.4822505023442733, | |
| "grad_norm": 20.9210927889319, | |
| "learning_rate": 3.0849803176787184e-07, | |
| "logits/chosen": -3.3171279430389404, | |
| "logits/rejected": -3.3153088092803955, | |
| "logps/chosen": -0.5533649921417236, | |
| "logps/rejected": -0.6165200471878052, | |
| "loss": 1.5082, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.5533649921417236, | |
| "rewards/margins": 0.06315504014492035, | |
| "rewards/rejected": -0.6165200471878052, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.49296718017414604, | |
| "grad_norm": 26.953714991186636, | |
| "learning_rate": 2.993469590438065e-07, | |
| "logits/chosen": -3.2146403789520264, | |
| "logits/rejected": -3.2017455101013184, | |
| "logps/chosen": -0.539149284362793, | |
| "logps/rejected": -0.5770600438117981, | |
| "loss": 1.4968, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -0.539149284362793, | |
| "rewards/margins": 0.03791072219610214, | |
| "rewards/rejected": -0.5770600438117981, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.5036838580040187, | |
| "grad_norm": 23.063142701753478, | |
| "learning_rate": 2.901265403480597e-07, | |
| "logits/chosen": -3.2051620483398438, | |
| "logits/rejected": -3.1935036182403564, | |
| "logps/chosen": -0.5151172876358032, | |
| "logps/rejected": -0.6125957369804382, | |
| "loss": 1.508, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.5151172876358032, | |
| "rewards/margins": 0.09747843444347382, | |
| "rewards/rejected": -0.6125957369804382, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.5144005358338914, | |
| "grad_norm": 28.679041181758652, | |
| "learning_rate": 2.808497328902782e-07, | |
| "logits/chosen": -3.152073621749878, | |
| "logits/rejected": -3.15959095954895, | |
| "logps/chosen": -0.5507190823554993, | |
| "logps/rejected": -0.6257890462875366, | |
| "loss": 1.5193, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.5507190823554993, | |
| "rewards/margins": 0.07506996393203735, | |
| "rewards/rejected": -0.6257890462875366, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5251172136637642, | |
| "grad_norm": 24.21410810508721, | |
| "learning_rate": 2.715295731217396e-07, | |
| "logits/chosen": -3.1899075508117676, | |
| "logits/rejected": -3.1687207221984863, | |
| "logps/chosen": -0.5879708528518677, | |
| "logps/rejected": -0.6203640103340149, | |
| "loss": 1.5033, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -0.5879708528518677, | |
| "rewards/margins": 0.032393116503953934, | |
| "rewards/rejected": -0.6203640103340149, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.535833891493637, | |
| "grad_norm": 24.944015170343583, | |
| "learning_rate": 2.621791584155725e-07, | |
| "logits/chosen": -3.186570167541504, | |
| "logits/rejected": -3.1771557331085205, | |
| "logps/chosen": -0.610142171382904, | |
| "logps/rejected": -0.604741096496582, | |
| "loss": 1.5341, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.610142171382904, | |
| "rewards/margins": -0.005401057191193104, | |
| "rewards/rejected": -0.604741096496582, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.5465505693235098, | |
| "grad_norm": 22.133595287604955, | |
| "learning_rate": 2.52811628661367e-07, | |
| "logits/chosen": -3.2132976055145264, | |
| "logits/rejected": -3.21321439743042, | |
| "logps/chosen": -0.5207762122154236, | |
| "logps/rejected": -0.5928905606269836, | |
| "loss": 1.4997, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.5207762122154236, | |
| "rewards/margins": 0.07211437076330185, | |
| "rewards/rejected": -0.5928905606269836, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.5572672471533825, | |
| "grad_norm": 21.31026102730006, | |
| "learning_rate": 2.434401478000353e-07, | |
| "logits/chosen": -3.141242504119873, | |
| "logits/rejected": -3.1591105461120605, | |
| "logps/chosen": -0.526002049446106, | |
| "logps/rejected": -0.6240478754043579, | |
| "loss": 1.5259, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.526002049446106, | |
| "rewards/margins": 0.09804578125476837, | |
| "rewards/rejected": -0.6240478754043579, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5679839249832552, | |
| "grad_norm": 25.465212058173062, | |
| "learning_rate": 2.340778853248759e-07, | |
| "logits/chosen": -3.1647770404815674, | |
| "logits/rejected": -3.161611557006836, | |
| "logps/chosen": -0.5447332859039307, | |
| "logps/rejected": -0.6431838870048523, | |
| "loss": 1.496, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.5447332859039307, | |
| "rewards/margins": 0.098450668156147, | |
| "rewards/rejected": -0.6431838870048523, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.578700602813128, | |
| "grad_norm": 29.79126766513931, | |
| "learning_rate": 2.247379977748339e-07, | |
| "logits/chosen": -3.1814353466033936, | |
| "logits/rejected": -3.1744346618652344, | |
| "logps/chosen": -0.5134252905845642, | |
| "logps/rejected": -0.5546069145202637, | |
| "loss": 1.5172, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.5134252905845642, | |
| "rewards/margins": 0.04118159040808678, | |
| "rewards/rejected": -0.5546069145202637, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.5894172806430007, | |
| "grad_norm": 20.911132725985905, | |
| "learning_rate": 2.1543361024596573e-07, | |
| "logits/chosen": -3.2813096046447754, | |
| "logits/rejected": -3.2904067039489746, | |
| "logps/chosen": -0.478159099817276, | |
| "logps/rejected": -0.5237951278686523, | |
| "loss": 1.52, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.478159099817276, | |
| "rewards/margins": 0.045636050403118134, | |
| "rewards/rejected": -0.5237951278686523, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.6001339584728734, | |
| "grad_norm": 22.184303638777845, | |
| "learning_rate": 2.0617779794709082e-07, | |
| "logits/chosen": -3.227949857711792, | |
| "logits/rejected": -3.199709415435791, | |
| "logps/chosen": -0.5574294924736023, | |
| "logps/rejected": -0.6131659746170044, | |
| "loss": 1.5077, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.5574294924736023, | |
| "rewards/margins": 0.05573650076985359, | |
| "rewards/rejected": -0.6131659746170044, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.6108506363027462, | |
| "grad_norm": 23.16627725406288, | |
| "learning_rate": 1.9698356782554687e-07, | |
| "logits/chosen": -3.2686588764190674, | |
| "logits/rejected": -3.27409291267395, | |
| "logps/chosen": -0.6056652069091797, | |
| "logps/rejected": -0.6142518520355225, | |
| "loss": 1.529, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.6056652069091797, | |
| "rewards/margins": 0.008586685173213482, | |
| "rewards/rejected": -0.6142518520355225, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.6215673141326189, | |
| "grad_norm": 43.63471821872903, | |
| "learning_rate": 1.878638402888727e-07, | |
| "logits/chosen": -3.3371078968048096, | |
| "logits/rejected": -3.3204472064971924, | |
| "logps/chosen": -0.5225822925567627, | |
| "logps/rejected": -0.5866876840591431, | |
| "loss": 1.5024, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -0.5225822925567627, | |
| "rewards/margins": 0.06410545110702515, | |
| "rewards/rejected": -0.5866876840591431, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6322839919624916, | |
| "grad_norm": 25.701019654379248, | |
| "learning_rate": 1.7883143104810223e-07, | |
| "logits/chosen": -3.2457165718078613, | |
| "logits/rejected": -3.2686493396759033, | |
| "logps/chosen": -0.5675607919692993, | |
| "logps/rejected": -0.6349104642868042, | |
| "loss": 1.4993, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.5675607919692993, | |
| "rewards/margins": 0.06734969466924667, | |
| "rewards/rejected": -0.6349104642868042, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.6430006697923644, | |
| "grad_norm": 23.619470082480703, | |
| "learning_rate": 1.6989903310818548e-07, | |
| "logits/chosen": -3.3820648193359375, | |
| "logits/rejected": -3.3857345581054688, | |
| "logps/chosen": -0.5343354344367981, | |
| "logps/rejected": -0.6022351980209351, | |
| "loss": 1.4951, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -0.5343354344367981, | |
| "rewards/margins": 0.06789974123239517, | |
| "rewards/rejected": -0.6022351980209351, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6537173476222371, | |
| "grad_norm": 22.753889982587108, | |
| "learning_rate": 1.610791989308466e-07, | |
| "logits/chosen": -3.3234715461730957, | |
| "logits/rejected": -3.2950637340545654, | |
| "logps/chosen": -0.5346772074699402, | |
| "logps/rejected": -0.5783089995384216, | |
| "loss": 1.5037, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.5346772074699402, | |
| "rewards/margins": 0.04363187402486801, | |
| "rewards/rejected": -0.5783089995384216, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.6644340254521098, | |
| "grad_norm": 29.970511411150166, | |
| "learning_rate": 1.523843227949418e-07, | |
| "logits/chosen": -3.3544983863830566, | |
| "logits/rejected": -3.355292558670044, | |
| "logps/chosen": -0.5303682088851929, | |
| "logps/rejected": -0.5744400024414062, | |
| "loss": 1.4936, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.5303682088851929, | |
| "rewards/margins": 0.04407178610563278, | |
| "rewards/rejected": -0.5744400024414062, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.6751507032819826, | |
| "grad_norm": 19.95585897697304, | |
| "learning_rate": 1.4382662337910867e-07, | |
| "logits/chosen": -3.291701078414917, | |
| "logits/rejected": -3.3025760650634766, | |
| "logps/chosen": -0.5352946519851685, | |
| "logps/rejected": -0.5887657403945923, | |
| "loss": 1.502, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.5352946519851685, | |
| "rewards/margins": 0.053471069782972336, | |
| "rewards/rejected": -0.5887657403945923, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.6858673811118553, | |
| "grad_norm": 22.149800280030288, | |
| "learning_rate": 1.3541812659118104e-07, | |
| "logits/chosen": -3.334624767303467, | |
| "logits/rejected": -3.306361675262451, | |
| "logps/chosen": -0.5542501211166382, | |
| "logps/rejected": -0.5587230920791626, | |
| "loss": 1.524, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.5542501211166382, | |
| "rewards/margins": 0.004472990520298481, | |
| "rewards/rejected": -0.5587230920791626, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.696584058941728, | |
| "grad_norm": 25.604347826249015, | |
| "learning_rate": 1.2717064866850047e-07, | |
| "logits/chosen": -3.3316922187805176, | |
| "logits/rejected": -3.342952251434326, | |
| "logps/chosen": -0.5008823275566101, | |
| "logps/rejected": -0.5843838453292847, | |
| "loss": 1.4797, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -0.5008823275566101, | |
| "rewards/margins": 0.08350151777267456, | |
| "rewards/rejected": -0.5843838453292847, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.7073007367716008, | |
| "grad_norm": 24.090724072052794, | |
| "learning_rate": 1.1909577957287117e-07, | |
| "logits/chosen": -3.3231868743896484, | |
| "logits/rejected": -3.2961513996124268, | |
| "logps/chosen": -0.5285135507583618, | |
| "logps/rejected": -0.582491397857666, | |
| "loss": 1.4972, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.5285135507583618, | |
| "rewards/margins": 0.05397792533040047, | |
| "rewards/rejected": -0.582491397857666, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.7180174146014735, | |
| "grad_norm": 36.87118130194036, | |
| "learning_rate": 1.1120486670349441e-07, | |
| "logits/chosen": -3.4078221321105957, | |
| "logits/rejected": -3.3980777263641357, | |
| "logps/chosen": -0.5307896733283997, | |
| "logps/rejected": -0.6150525808334351, | |
| "loss": 1.5019, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.5307896733283997, | |
| "rewards/margins": 0.08426298201084137, | |
| "rewards/rejected": -0.6150525808334351, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.7287340924313462, | |
| "grad_norm": 25.52897276938915, | |
| "learning_rate": 1.0350899895076948e-07, | |
| "logits/chosen": -3.3364510536193848, | |
| "logits/rejected": -3.321132183074951, | |
| "logps/chosen": -0.5735017657279968, | |
| "logps/rejected": -0.6429983973503113, | |
| "loss": 1.521, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -0.5735017657279968, | |
| "rewards/margins": 0.06949658691883087, | |
| "rewards/rejected": -0.6429983973503113, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.739450770261219, | |
| "grad_norm": 20.331652657630677, | |
| "learning_rate": 9.601899111337069e-08, | |
| "logits/chosen": -3.367867946624756, | |
| "logits/rejected": -3.3775951862335205, | |
| "logps/chosen": -0.5771843194961548, | |
| "logps/rejected": -0.6704375147819519, | |
| "loss": 1.496, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.5771843194961548, | |
| "rewards/margins": 0.09325330704450607, | |
| "rewards/rejected": -0.6704375147819519, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.7501674480910918, | |
| "grad_norm": 22.274114959082404, | |
| "learning_rate": 8.874536870049765e-08, | |
| "logits/chosen": -3.3537139892578125, | |
| "logits/rejected": -3.378490924835205, | |
| "logps/chosen": -0.5365322828292847, | |
| "logps/rejected": -0.6033391952514648, | |
| "loss": 1.4971, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -0.5365322828292847, | |
| "rewards/margins": 0.06680694222450256, | |
| "rewards/rejected": -0.6033391952514648, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.7608841259209645, | |
| "grad_norm": 24.177285206790465, | |
| "learning_rate": 8.169835314065663e-08, | |
| "logits/chosen": -3.3511605262756348, | |
| "logits/rejected": -3.3401741981506348, | |
| "logps/chosen": -0.5807870626449585, | |
| "logps/rejected": -0.6009158492088318, | |
| "loss": 1.499, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -0.5807870626449585, | |
| "rewards/margins": 0.020128777250647545, | |
| "rewards/rejected": -0.6009158492088318, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.7716008037508373, | |
| "grad_norm": 22.05484832098868, | |
| "learning_rate": 7.488784741775816e-08, | |
| "logits/chosen": -3.3435721397399902, | |
| "logits/rejected": -3.33024525642395, | |
| "logps/chosen": -0.5081290006637573, | |
| "logps/rejected": -0.5656172037124634, | |
| "loss": 1.5079, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.5081290006637573, | |
| "rewards/margins": 0.057488154619932175, | |
| "rewards/rejected": -0.5656172037124634, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.78231748158071, | |
| "grad_norm": 24.600234649150252, | |
| "learning_rate": 6.832342215471715e-08, | |
| "logits/chosen": -3.3981361389160156, | |
| "logits/rejected": -3.39165997505188, | |
| "logps/chosen": -0.5507721304893494, | |
| "logps/rejected": -0.6153653860092163, | |
| "loss": 1.5147, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.5507721304893494, | |
| "rewards/margins": 0.06459329277276993, | |
| "rewards/rejected": -0.6153653860092163, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.7930341594105828, | |
| "grad_norm": 31.85175227986763, | |
| "learning_rate": 6.201430216410985e-08, | |
| "logits/chosen": -3.391742706298828, | |
| "logits/rejected": -3.413292646408081, | |
| "logps/chosen": -0.5624402761459351, | |
| "logps/rejected": -0.6359676122665405, | |
| "loss": 1.5149, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.5624402761459351, | |
| "rewards/margins": 0.07352729141712189, | |
| "rewards/rejected": -0.6359676122665405, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.8037508372404555, | |
| "grad_norm": 23.503804219453023, | |
| "learning_rate": 5.596935348479029e-08, | |
| "logits/chosen": -3.3563759326934814, | |
| "logits/rejected": -3.338204860687256, | |
| "logps/chosen": -0.6166857481002808, | |
| "logps/rejected": -0.700817883014679, | |
| "loss": 1.5014, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.6166857481002808, | |
| "rewards/margins": 0.08413223922252655, | |
| "rewards/rejected": -0.700817883014679, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.8144675150703282, | |
| "grad_norm": 26.081587652629086, | |
| "learning_rate": 5.01970709226808e-08, | |
| "logits/chosen": -3.443342924118042, | |
| "logits/rejected": -3.4111335277557373, | |
| "logps/chosen": -0.5469595193862915, | |
| "logps/rejected": -0.6322815418243408, | |
| "loss": 1.5083, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.5469595193862915, | |
| "rewards/margins": 0.08532204478979111, | |
| "rewards/rejected": -0.6322815418243408, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.825184192900201, | |
| "grad_norm": 24.73716667501714, | |
| "learning_rate": 4.4705566113246635e-08, | |
| "logits/chosen": -3.349184513092041, | |
| "logits/rejected": -3.359894275665283, | |
| "logps/chosen": -0.5703479647636414, | |
| "logps/rejected": -0.6489995718002319, | |
| "loss": 1.4984, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.5703479647636414, | |
| "rewards/margins": 0.07865166664123535, | |
| "rewards/rejected": -0.6489995718002319, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.8359008707300737, | |
| "grad_norm": 24.765231706983972, | |
| "learning_rate": 3.950255612243042e-08, | |
| "logits/chosen": -3.3900883197784424, | |
| "logits/rejected": -3.3493354320526123, | |
| "logps/chosen": -0.6082748174667358, | |
| "logps/rejected": -0.6323951482772827, | |
| "loss": 1.5319, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.6082748174667358, | |
| "rewards/margins": 0.024120373651385307, | |
| "rewards/rejected": -0.6323951482772827, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.8466175485599464, | |
| "grad_norm": 23.65912516379117, | |
| "learning_rate": 3.459535260206384e-08, | |
| "logits/chosen": -3.3917059898376465, | |
| "logits/rejected": -3.3759236335754395, | |
| "logps/chosen": -0.5519801378250122, | |
| "logps/rejected": -0.5884855389595032, | |
| "loss": 1.5143, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.5519801378250122, | |
| "rewards/margins": 0.036505263298749924, | |
| "rewards/rejected": -0.5884855389595032, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.8573342263898192, | |
| "grad_norm": 28.498252560499747, | |
| "learning_rate": 2.999085151499764e-08, | |
| "logits/chosen": -3.3536014556884766, | |
| "logits/rejected": -3.322979688644409, | |
| "logps/chosen": -0.5050632953643799, | |
| "logps/rejected": -0.6183533072471619, | |
| "loss": 1.4908, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.5050632953643799, | |
| "rewards/margins": 0.11329002678394318, | |
| "rewards/rejected": -0.6183533072471619, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8573342263898192, | |
| "eval_logits/chosen": -3.4685490131378174, | |
| "eval_logits/rejected": -3.4632692337036133, | |
| "eval_logps/chosen": -0.5601765513420105, | |
| "eval_logps/rejected": -0.6162815093994141, | |
| "eval_loss": 1.522051215171814, | |
| "eval_rewards/accuracies": 0.5797872543334961, | |
| "eval_rewards/chosen": -0.5601765513420105, | |
| "eval_rewards/margins": 0.05610497295856476, | |
| "eval_rewards/rejected": -0.6162815093994141, | |
| "eval_runtime": 433.0685, | |
| "eval_samples_per_second": 6.913, | |
| "eval_steps_per_second": 0.434, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8680509042196919, | |
| "grad_norm": 23.705058603217022, | |
| "learning_rate": 2.5695523444387747e-08, | |
| "logits/chosen": -3.364602565765381, | |
| "logits/rejected": -3.350884199142456, | |
| "logps/chosen": -0.5609386563301086, | |
| "logps/rejected": -0.605690598487854, | |
| "loss": 1.5111, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.5609386563301086, | |
| "rewards/margins": 0.044751983135938644, | |
| "rewards/rejected": -0.605690598487854, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.8787675820495646, | |
| "grad_norm": 22.37221809399767, | |
| "learning_rate": 2.1715404500756875e-08, | |
| "logits/chosen": -3.412458896636963, | |
| "logits/rejected": -3.3824081420898438, | |
| "logps/chosen": -0.5534378290176392, | |
| "logps/rejected": -0.5746613144874573, | |
| "loss": 1.4932, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -0.5534378290176392, | |
| "rewards/margins": 0.021223513409495354, | |
| "rewards/rejected": -0.5746613144874573, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.8894842598794374, | |
| "grad_norm": 26.32210656993751, | |
| "learning_rate": 1.8056087839608546e-08, | |
| "logits/chosen": -3.332536220550537, | |
| "logits/rejected": -3.3019680976867676, | |
| "logps/chosen": -0.5494991540908813, | |
| "logps/rejected": -0.6182190179824829, | |
| "loss": 1.4885, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.5494991540908813, | |
| "rewards/margins": 0.06871981918811798, | |
| "rewards/rejected": -0.6182190179824829, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.9002009377093101, | |
| "grad_norm": 27.43028943315468, | |
| "learning_rate": 1.4722715801513935e-08, | |
| "logits/chosen": -3.3295421600341797, | |
| "logits/rejected": -3.336806058883667, | |
| "logps/chosen": -0.5279422998428345, | |
| "logps/rejected": -0.5645946860313416, | |
| "loss": 1.5217, | |
| "rewards/accuracies": 0.4937500059604645, | |
| "rewards/chosen": -0.5279422998428345, | |
| "rewards/margins": 0.03665238991379738, | |
| "rewards/rejected": -0.5645946860313416, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.9109176155391828, | |
| "grad_norm": 25.425067386156872, | |
| "learning_rate": 1.1719972685717395e-08, | |
| "logits/chosen": -3.337613582611084, | |
| "logits/rejected": -3.3247127532958984, | |
| "logps/chosen": -0.5188074111938477, | |
| "logps/rejected": -0.6191619634628296, | |
| "loss": 1.4926, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.5188074111938477, | |
| "rewards/margins": 0.10035452991724014, | |
| "rewards/rejected": -0.6191619634628296, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.9216342933690556, | |
| "grad_norm": 35.62756758017028, | |
| "learning_rate": 9.05207816741485e-09, | |
| "logits/chosen": -3.3888893127441406, | |
| "logits/rejected": -3.4012131690979004, | |
| "logps/chosen": -0.6197360754013062, | |
| "logps/rejected": -0.6971450448036194, | |
| "loss": 1.5011, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.6197360754013062, | |
| "rewards/margins": 0.07740898430347443, | |
| "rewards/rejected": -0.6971450448036194, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.9323509711989283, | |
| "grad_norm": 27.22699922034511, | |
| "learning_rate": 6.722781367956481e-09, | |
| "logits/chosen": -3.455385208129883, | |
| "logits/rejected": -3.4597179889678955, | |
| "logps/chosen": -0.5144860744476318, | |
| "logps/rejected": -0.6305041909217834, | |
| "loss": 1.4974, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.5144860744476318, | |
| "rewards/margins": 0.1160181388258934, | |
| "rewards/rejected": -0.6305041909217834, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.943067649028801, | |
| "grad_norm": 23.43687138581986, | |
| "learning_rate": 4.735355586305861e-09, | |
| "logits/chosen": -3.2910914421081543, | |
| "logits/rejected": -3.288029193878174, | |
| "logps/chosen": -0.5328935980796814, | |
| "logps/rejected": -0.6186143755912781, | |
| "loss": 1.4731, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.5328935980796814, | |
| "rewards/margins": 0.08572079241275787, | |
| "rewards/rejected": -0.6186143755912781, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.9537843268586738, | |
| "grad_norm": 27.25503683317219, | |
| "learning_rate": 3.0925936991598454e-09, | |
| "logits/chosen": -3.320718765258789, | |
| "logits/rejected": -3.319967746734619, | |
| "logps/chosen": -0.5472747683525085, | |
| "logps/rejected": -0.6223639845848083, | |
| "loss": 1.493, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.5472747683525085, | |
| "rewards/margins": 0.07508920133113861, | |
| "rewards/rejected": -0.6223639845848083, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.9645010046885466, | |
| "grad_norm": 23.669876406776968, | |
| "learning_rate": 1.7968042361930225e-09, | |
| "logits/chosen": -3.354473114013672, | |
| "logits/rejected": -3.3409972190856934, | |
| "logps/chosen": -0.5448306202888489, | |
| "logps/rejected": -0.6130356788635254, | |
| "loss": 1.4876, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -0.5448306202888489, | |
| "rewards/margins": 0.06820502877235413, | |
| "rewards/rejected": -0.6130356788635254, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.9752176825184193, | |
| "grad_norm": 27.737663302659318, | |
| "learning_rate": 8.498081359423759e-10, | |
| "logits/chosen": -3.331641435623169, | |
| "logits/rejected": -3.3293700218200684, | |
| "logps/chosen": -0.5374922156333923, | |
| "logps/rejected": -0.6145657300949097, | |
| "loss": 1.4953, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.5374922156333923, | |
| "rewards/margins": 0.07707355916500092, | |
| "rewards/rejected": -0.6145657300949097, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.9859343603482921, | |
| "grad_norm": 29.309243600698153, | |
| "learning_rate": 2.5293618689056774e-10, | |
| "logits/chosen": -3.3501858711242676, | |
| "logits/rejected": -3.3541958332061768, | |
| "logps/chosen": -0.5475316643714905, | |
| "logps/rejected": -0.6366786360740662, | |
| "loss": 1.5039, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.5475316643714905, | |
| "rewards/margins": 0.08914700895547867, | |
| "rewards/rejected": -0.6366786360740662, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.9966510381781648, | |
| "grad_norm": 21.403219882354602, | |
| "learning_rate": 7.02715734424153e-12, | |
| "logits/chosen": -3.3826191425323486, | |
| "logits/rejected": -3.36454701423645, | |
| "logps/chosen": -0.5860726833343506, | |
| "logps/rejected": -0.6269819736480713, | |
| "loss": 1.5076, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.5860726833343506, | |
| "rewards/margins": 0.040909312665462494, | |
| "rewards/rejected": -0.6269819736480713, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.9987943737441393, | |
| "step": 466, | |
| "total_flos": 0.0, | |
| "train_loss": 1.5188358733070766, | |
| "train_runtime": 19352.2078, | |
| "train_samples_per_second": 3.086, | |
| "train_steps_per_second": 0.024 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 466, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 1000000, | |
| "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": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |