| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 273, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 498.0, | |
| "learning_rate": 5.000000000000001e-07, | |
| "log_odds_chosen": 0.2863905131816864, | |
| "log_odds_ratio": -0.5870538949966431, | |
| "logits/chosen": -2.007096767425537, | |
| "logits/rejected": -2.008589029312134, | |
| "logps/chosen": -0.9226220846176147, | |
| "logps/rejected": -1.105508804321289, | |
| "loss": 3.8837, | |
| "nll_loss": 4.084702491760254, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.04613110423088074, | |
| "rewards/margins": 0.009144341573119164, | |
| "rewards/rejected": -0.05527544766664505, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 50.5, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "log_odds_chosen": 0.3382885456085205, | |
| "log_odds_ratio": -0.5765059590339661, | |
| "logits/chosen": -2.6738319396972656, | |
| "logits/rejected": -2.6777594089508057, | |
| "logps/chosen": -0.902550220489502, | |
| "logps/rejected": -1.1331396102905273, | |
| "loss": 1.6578, | |
| "nll_loss": 1.7452325820922852, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -0.0451275110244751, | |
| "rewards/margins": 0.011529467068612576, | |
| "rewards/rejected": -0.05665697902441025, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 1.5e-06, | |
| "log_odds_chosen": 0.2815089225769043, | |
| "log_odds_ratio": -0.5875880122184753, | |
| "logits/chosen": -2.437102794647217, | |
| "logits/rejected": -2.4390525817871094, | |
| "logps/chosen": -0.8905347585678101, | |
| "logps/rejected": -1.0546678304672241, | |
| "loss": 0.5865, | |
| "nll_loss": 0.5374206900596619, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.044526733458042145, | |
| "rewards/margins": 0.008206663653254509, | |
| "rewards/rejected": -0.0527334026992321, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "log_odds_chosen": 0.3029056787490845, | |
| "log_odds_ratio": -0.6036669611930847, | |
| "logits/chosen": -2.1445510387420654, | |
| "logits/rejected": -2.166757583618164, | |
| "logps/chosen": -0.9048384428024292, | |
| "logps/rejected": -1.0757750272750854, | |
| "loss": 0.5633, | |
| "nll_loss": 0.5419524908065796, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.04524192586541176, | |
| "rewards/margins": 0.008546831086277962, | |
| "rewards/rejected": -0.05378875881433487, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 2.5e-06, | |
| "log_odds_chosen": 0.2378097027540207, | |
| "log_odds_ratio": -0.630806565284729, | |
| "logits/chosen": -1.8973627090454102, | |
| "logits/rejected": -1.934752106666565, | |
| "logps/chosen": -0.8510879278182983, | |
| "logps/rejected": -1.0084228515625, | |
| "loss": 0.5498, | |
| "nll_loss": 0.47691044211387634, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.042554400861263275, | |
| "rewards/margins": 0.007866747677326202, | |
| "rewards/rejected": -0.050421155989170074, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 3e-06, | |
| "log_odds_chosen": 0.34561508893966675, | |
| "log_odds_ratio": -0.5882634520530701, | |
| "logits/chosen": -1.789703369140625, | |
| "logits/rejected": -1.8299936056137085, | |
| "logps/chosen": -0.8602263331413269, | |
| "logps/rejected": -1.0266495943069458, | |
| "loss": 0.5348, | |
| "nll_loss": 0.4759138524532318, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -0.043011318892240524, | |
| "rewards/margins": 0.008321164175868034, | |
| "rewards/rejected": -0.05133248120546341, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 1.5703125, | |
| "learning_rate": 3.5e-06, | |
| "log_odds_chosen": 0.306732177734375, | |
| "log_odds_ratio": -0.6226531267166138, | |
| "logits/chosen": -1.8786418437957764, | |
| "logits/rejected": -1.9659067392349243, | |
| "logps/chosen": -0.8097027540206909, | |
| "logps/rejected": -0.9237287640571594, | |
| "loss": 0.5199, | |
| "nll_loss": 0.47153204679489136, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.040485139936208725, | |
| "rewards/margins": 0.005701296962797642, | |
| "rewards/rejected": -0.04618643969297409, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 1.5703125, | |
| "learning_rate": 4.000000000000001e-06, | |
| "log_odds_chosen": 0.46479716897010803, | |
| "log_odds_ratio": -0.5484603047370911, | |
| "logits/chosen": -1.894330382347107, | |
| "logits/rejected": -1.9072799682617188, | |
| "logps/chosen": -0.8353139758110046, | |
| "logps/rejected": -1.0923199653625488, | |
| "loss": 0.5272, | |
| "nll_loss": 0.49910029768943787, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.04176569730043411, | |
| "rewards/margins": 0.012850301340222359, | |
| "rewards/rejected": -0.05461599677801132, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 1.5703125, | |
| "learning_rate": 4.5e-06, | |
| "log_odds_chosen": 0.46429938077926636, | |
| "log_odds_ratio": -0.561553418636322, | |
| "logits/chosen": -1.9992767572402954, | |
| "logits/rejected": -2.0206494331359863, | |
| "logps/chosen": -0.7664617300033569, | |
| "logps/rejected": -0.9973196983337402, | |
| "loss": 0.5236, | |
| "nll_loss": 0.47519445419311523, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.038323089480400085, | |
| "rewards/margins": 0.01154289674013853, | |
| "rewards/rejected": -0.04986598342657089, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 1.75, | |
| "learning_rate": 5e-06, | |
| "log_odds_chosen": 0.357390433549881, | |
| "log_odds_ratio": -0.5652999877929688, | |
| "logits/chosen": -1.8920485973358154, | |
| "logits/rejected": -1.9230598211288452, | |
| "logps/chosen": -0.7654845118522644, | |
| "logps/rejected": -0.9835049510002136, | |
| "loss": 0.4967, | |
| "nll_loss": 0.4664408266544342, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.03827423229813576, | |
| "rewards/margins": 0.010901018045842648, | |
| "rewards/rejected": -0.04917524382472038, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "grad_norm": 1.6015625, | |
| "learning_rate": 4.767312946227961e-06, | |
| "log_odds_chosen": 0.2798970341682434, | |
| "log_odds_ratio": -0.5895387530326843, | |
| "logits/chosen": -1.9187400341033936, | |
| "logits/rejected": -1.9244420528411865, | |
| "logps/chosen": -0.8067512512207031, | |
| "logps/rejected": -0.9561241269111633, | |
| "loss": 0.4901, | |
| "nll_loss": 0.6010779738426208, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.040337562561035156, | |
| "rewards/margins": 0.007468645926564932, | |
| "rewards/rejected": -0.047806207090616226, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 4.564354645876385e-06, | |
| "log_odds_chosen": 0.6895818114280701, | |
| "log_odds_ratio": -0.45160675048828125, | |
| "logits/chosen": -1.7748291492462158, | |
| "logits/rejected": -1.8056682348251343, | |
| "logps/chosen": -0.7309473752975464, | |
| "logps/rejected": -1.0976765155792236, | |
| "loss": 0.4873, | |
| "nll_loss": 0.4398137629032135, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": -0.03654736652970314, | |
| "rewards/margins": 0.018336456269025803, | |
| "rewards/rejected": -0.054883819073438644, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 4.385290096535147e-06, | |
| "log_odds_chosen": 0.658761203289032, | |
| "log_odds_ratio": -0.47310584783554077, | |
| "logits/chosen": -1.9612312316894531, | |
| "logits/rejected": -1.9696747064590454, | |
| "logps/chosen": -0.7409031987190247, | |
| "logps/rejected": -1.0737724304199219, | |
| "loss": 0.4918, | |
| "nll_loss": 0.43778449296951294, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": -0.037045158445835114, | |
| "rewards/margins": 0.016643470153212547, | |
| "rewards/rejected": -0.05368863418698311, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "grad_norm": 1.75, | |
| "learning_rate": 4.2257712736425835e-06, | |
| "log_odds_chosen": 0.5739403367042542, | |
| "log_odds_ratio": -0.5177870988845825, | |
| "logits/chosen": -1.7142318487167358, | |
| "logits/rejected": -1.7519795894622803, | |
| "logps/chosen": -0.724389374256134, | |
| "logps/rejected": -1.0089597702026367, | |
| "loss": 0.475, | |
| "nll_loss": 0.4226527214050293, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": -0.03621947020292282, | |
| "rewards/margins": 0.014228522777557373, | |
| "rewards/rejected": -0.050447989255189896, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 1.53125, | |
| "learning_rate": 4.082482904638631e-06, | |
| "log_odds_chosen": 0.437107652425766, | |
| "log_odds_ratio": -0.5386956334114075, | |
| "logits/chosen": -1.970740556716919, | |
| "logits/rejected": -2.0762734413146973, | |
| "logps/chosen": -0.736059308052063, | |
| "logps/rejected": -0.9635001420974731, | |
| "loss": 0.4783, | |
| "nll_loss": 0.4815542697906494, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.03680296614766121, | |
| "rewards/margins": 0.011372040025889874, | |
| "rewards/rejected": -0.04817500710487366, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 3.952847075210474e-06, | |
| "log_odds_chosen": 0.25039786100387573, | |
| "log_odds_ratio": -0.6173754334449768, | |
| "logits/chosen": -1.8762115240097046, | |
| "logits/rejected": -1.9119548797607422, | |
| "logps/chosen": -0.8681584596633911, | |
| "logps/rejected": -1.0069737434387207, | |
| "loss": 0.4902, | |
| "nll_loss": 0.4779680371284485, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.043407924473285675, | |
| "rewards/margins": 0.006940762512385845, | |
| "rewards/rejected": -0.050348687916994095, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 3.834824944236852e-06, | |
| "log_odds_chosen": 0.8294051885604858, | |
| "log_odds_ratio": -0.4564356803894043, | |
| "logits/chosen": -1.9998172521591187, | |
| "logits/rejected": -2.012346029281616, | |
| "logps/chosen": -0.7080036997795105, | |
| "logps/rejected": -1.075842022895813, | |
| "loss": 0.4728, | |
| "nll_loss": 0.4948364198207855, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -0.035400185734033585, | |
| "rewards/margins": 0.018391914665699005, | |
| "rewards/rejected": -0.05379210039973259, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "grad_norm": 1.5234375, | |
| "learning_rate": 3.72677996249965e-06, | |
| "log_odds_chosen": 0.6013724207878113, | |
| "log_odds_ratio": -0.5433425307273865, | |
| "logits/chosen": -1.8251736164093018, | |
| "logits/rejected": -1.8962123394012451, | |
| "logps/chosen": -0.7299692034721375, | |
| "logps/rejected": -1.0021026134490967, | |
| "loss": 0.4729, | |
| "nll_loss": 0.419732004404068, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.036498457193374634, | |
| "rewards/margins": 0.013606672175228596, | |
| "rewards/rejected": -0.05010513216257095, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.09, | |
| "grad_norm": 1.9921875, | |
| "learning_rate": 3.6273812505500587e-06, | |
| "log_odds_chosen": 0.9894807934761047, | |
| "log_odds_ratio": -0.3992963433265686, | |
| "logits/chosen": -1.8533046245574951, | |
| "logits/rejected": -1.8780378103256226, | |
| "logps/chosen": -0.6669403314590454, | |
| "logps/rejected": -1.1133853197097778, | |
| "loss": 0.441, | |
| "nll_loss": 0.40508294105529785, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": -0.03334701806306839, | |
| "rewards/margins": 0.02232225239276886, | |
| "rewards/rejected": -0.05566927045583725, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "grad_norm": 1.6015625, | |
| "learning_rate": 3.5355339059327378e-06, | |
| "log_odds_chosen": 0.7237910032272339, | |
| "log_odds_ratio": -0.43176889419555664, | |
| "logits/chosen": -1.7843163013458252, | |
| "logits/rejected": -1.8297704458236694, | |
| "logps/chosen": -0.6735831499099731, | |
| "logps/rejected": -1.0497777462005615, | |
| "loss": 0.4287, | |
| "nll_loss": 0.4101743698120117, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": -0.033679161220788956, | |
| "rewards/margins": 0.01880973018705845, | |
| "rewards/rejected": -0.052488893270492554, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.31, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 3.450327796711771e-06, | |
| "log_odds_chosen": 0.6548705101013184, | |
| "log_odds_ratio": -0.4936867654323578, | |
| "logits/chosen": -1.9768335819244385, | |
| "logits/rejected": -1.964470624923706, | |
| "logps/chosen": -0.7660457491874695, | |
| "logps/rejected": -1.0849239826202393, | |
| "loss": 0.4251, | |
| "nll_loss": 0.4697703421115875, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -0.03830229118466377, | |
| "rewards/margins": 0.01594390906393528, | |
| "rewards/rejected": -0.0542462058365345, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.42, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 3.3709993123162106e-06, | |
| "log_odds_chosen": 0.9413236379623413, | |
| "log_odds_ratio": -0.3946041464805603, | |
| "logits/chosen": -1.8952887058258057, | |
| "logits/rejected": -1.9475895166397095, | |
| "logps/chosen": -0.6552693247795105, | |
| "logps/rejected": -1.1159121990203857, | |
| "loss": 0.4313, | |
| "nll_loss": 0.44144296646118164, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": -0.032763466238975525, | |
| "rewards/margins": 0.023032139986753464, | |
| "rewards/rejected": -0.05579560995101929, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 3.296902366978936e-06, | |
| "log_odds_chosen": 0.8275907635688782, | |
| "log_odds_ratio": -0.4250548481941223, | |
| "logits/chosen": -1.8575712442398071, | |
| "logits/rejected": -1.9162076711654663, | |
| "logps/chosen": -0.6322061419487, | |
| "logps/rejected": -1.0946836471557617, | |
| "loss": 0.4229, | |
| "nll_loss": 0.3766014575958252, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -0.03161030635237694, | |
| "rewards/margins": 0.023123880848288536, | |
| "rewards/rejected": -0.05473419278860092, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 3.2274861218395142e-06, | |
| "log_odds_chosen": 0.7520133852958679, | |
| "log_odds_ratio": -0.45257338881492615, | |
| "logits/chosen": -1.8115984201431274, | |
| "logits/rejected": -1.8446438312530518, | |
| "logps/chosen": -0.7262736558914185, | |
| "logps/rejected": -1.1011757850646973, | |
| "loss": 0.423, | |
| "nll_loss": 0.4082041382789612, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": -0.03631368279457092, | |
| "rewards/margins": 0.01874510571360588, | |
| "rewards/rejected": -0.055058788508176804, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.75, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 3.1622776601683796e-06, | |
| "log_odds_chosen": 0.9655596017837524, | |
| "log_odds_ratio": -0.42737776041030884, | |
| "logits/chosen": -1.9757699966430664, | |
| "logits/rejected": -1.9998546838760376, | |
| "logps/chosen": -0.6736587285995483, | |
| "logps/rejected": -1.1431043148040771, | |
| "loss": 0.4281, | |
| "nll_loss": 0.4197885990142822, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": -0.0336829349398613, | |
| "rewards/margins": 0.023472286760807037, | |
| "rewards/rejected": -0.057155221700668335, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.86, | |
| "grad_norm": 1.953125, | |
| "learning_rate": 3.1008683647302113e-06, | |
| "log_odds_chosen": 1.018965482711792, | |
| "log_odds_ratio": -0.3666786551475525, | |
| "logits/chosen": -1.9215749502182007, | |
| "logits/rejected": -1.9601131677627563, | |
| "logps/chosen": -0.5990187525749207, | |
| "logps/rejected": -1.1184070110321045, | |
| "loss": 0.421, | |
| "nll_loss": 0.39652419090270996, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": -0.0299509409815073, | |
| "rewards/margins": 0.025969412177801132, | |
| "rewards/rejected": -0.055920351296663284, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.97, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 3.0429030972509227e-06, | |
| "log_odds_chosen": 0.7740264534950256, | |
| "log_odds_ratio": -0.44708919525146484, | |
| "logits/chosen": -1.8729135990142822, | |
| "logits/rejected": -1.9034210443496704, | |
| "logps/chosen": -0.6478686332702637, | |
| "logps/rejected": -0.9832038879394531, | |
| "loss": 0.4251, | |
| "nll_loss": 0.4181249141693115, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": -0.0323934331536293, | |
| "rewards/margins": 0.016766764223575592, | |
| "rewards/rejected": -0.0491601899266243, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 273, | |
| "total_flos": 0.0, | |
| "train_loss": 0.649145109749539, | |
| "train_runtime": 4268.1705, | |
| "train_samples_per_second": 8.104, | |
| "train_steps_per_second": 0.064 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 273, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "total_flos": 0.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |