| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.3161794977356937, | |
| "eval_steps": 500, | |
| "global_step": 24, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.013174145738987238, | |
| "grad_norm": 0.4815484285354614, | |
| "learning_rate": 6.25e-08, | |
| "logits/chosen": 10.088521957397461, | |
| "logits/rejected": 10.263787269592285, | |
| "logps/chosen": -163.12940979003906, | |
| "logps/ref_chosen": -163.12940979003906, | |
| "logps/ref_rejected": -171.48428344726562, | |
| "logps/rejected": -171.48428344726562, | |
| "loss": 0.6931, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.026348291477974475, | |
| "grad_norm": 0.627070426940918, | |
| "learning_rate": 1.25e-07, | |
| "logits/chosen": 10.592972755432129, | |
| "logits/rejected": 10.720216751098633, | |
| "logps/chosen": -155.91574096679688, | |
| "logps/ref_chosen": -155.91574096679688, | |
| "logps/ref_rejected": -161.34078979492188, | |
| "logps/rejected": -161.34078979492188, | |
| "loss": 0.6931, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.03952243721696171, | |
| "grad_norm": 0.4148138165473938, | |
| "learning_rate": 1.875e-07, | |
| "logits/chosen": 10.043272972106934, | |
| "logits/rejected": 10.398024559020996, | |
| "logps/chosen": -158.3568115234375, | |
| "logps/ref_chosen": -157.65640258789062, | |
| "logps/ref_rejected": -168.5882110595703, | |
| "logps/rejected": -168.91085815429688, | |
| "loss": 0.6951, | |
| "rewards/accuracies": 0.3828125, | |
| "rewards/chosen": -0.007004000246524811, | |
| "rewards/margins": -0.0037774655502289534, | |
| "rewards/rejected": -0.003226534929126501, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.05269658295594895, | |
| "grad_norm": 0.7029770612716675, | |
| "learning_rate": 2.5e-07, | |
| "logits/chosen": 10.250253677368164, | |
| "logits/rejected": 10.45008659362793, | |
| "logps/chosen": -164.01119995117188, | |
| "logps/ref_chosen": -162.89878845214844, | |
| "logps/ref_rejected": -168.30462646484375, | |
| "logps/rejected": -169.1818389892578, | |
| "loss": 0.6944, | |
| "rewards/accuracies": 0.453125, | |
| "rewards/chosen": -0.01112416572868824, | |
| "rewards/margins": -0.0023521997500211, | |
| "rewards/rejected": -0.008771965280175209, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.06587072869493618, | |
| "grad_norm": 0.4063253104686737, | |
| "learning_rate": 3.1249999999999997e-07, | |
| "logits/chosen": 10.442557334899902, | |
| "logits/rejected": 10.740192413330078, | |
| "logps/chosen": -156.1859130859375, | |
| "logps/ref_chosen": -156.03257751464844, | |
| "logps/ref_rejected": -165.37911987304688, | |
| "logps/rejected": -165.6518096923828, | |
| "loss": 0.6926, | |
| "rewards/accuracies": 0.5234375, | |
| "rewards/chosen": -0.001533512957394123, | |
| "rewards/margins": 0.001193464733660221, | |
| "rewards/rejected": -0.002726977691054344, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.07904487443392343, | |
| "grad_norm": 0.4845049977302551, | |
| "learning_rate": 3.75e-07, | |
| "logits/chosen": 10.906261444091797, | |
| "logits/rejected": 11.201122283935547, | |
| "logps/chosen": -162.45692443847656, | |
| "logps/ref_chosen": -161.98570251464844, | |
| "logps/ref_rejected": -169.72560119628906, | |
| "logps/rejected": -170.18275451660156, | |
| "loss": 0.6932, | |
| "rewards/accuracies": 0.515625, | |
| "rewards/chosen": -0.004712029360234737, | |
| "rewards/margins": -0.00014030117017682642, | |
| "rewards/rejected": -0.004571728408336639, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.09221902017291066, | |
| "grad_norm": 0.8172655701637268, | |
| "learning_rate": 4.375e-07, | |
| "logits/chosen": 9.883949279785156, | |
| "logits/rejected": 10.030972480773926, | |
| "logps/chosen": -157.43295288085938, | |
| "logps/ref_chosen": -157.26968383789062, | |
| "logps/ref_rejected": -167.37155151367188, | |
| "logps/rejected": -167.53939819335938, | |
| "loss": 0.6931, | |
| "rewards/accuracies": 0.484375, | |
| "rewards/chosen": -0.0016327811172232032, | |
| "rewards/margins": 4.5756096369586885e-05, | |
| "rewards/rejected": -0.0016785369953140616, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.1053931659118979, | |
| "grad_norm": 0.588524341583252, | |
| "learning_rate": 5e-07, | |
| "logits/chosen": 10.633930206298828, | |
| "logits/rejected": 10.81590747833252, | |
| "logps/chosen": -162.8237762451172, | |
| "logps/ref_chosen": -162.2948455810547, | |
| "logps/ref_rejected": -172.98866271972656, | |
| "logps/rejected": -173.56680297851562, | |
| "loss": 0.6929, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -0.005289244465529919, | |
| "rewards/margins": 0.0004922347725369036, | |
| "rewards/rejected": -0.005781479645520449, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.11856731165088513, | |
| "grad_norm": 0.46077635884284973, | |
| "learning_rate": 4.997252228714278e-07, | |
| "logits/chosen": 10.326555252075195, | |
| "logits/rejected": 10.736672401428223, | |
| "logps/chosen": -164.5288543701172, | |
| "logps/ref_chosen": -163.37091064453125, | |
| "logps/ref_rejected": -173.1500701904297, | |
| "logps/rejected": -174.08392333984375, | |
| "loss": 0.6943, | |
| "rewards/accuracies": 0.4375, | |
| "rewards/chosen": -0.011579334735870361, | |
| "rewards/margins": -0.0022407739888876677, | |
| "rewards/rejected": -0.00933856051415205, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.13174145738987236, | |
| "grad_norm": 0.673312783241272, | |
| "learning_rate": 4.989014955054745e-07, | |
| "logits/chosen": 10.325155258178711, | |
| "logits/rejected": 10.473593711853027, | |
| "logps/chosen": -157.8944091796875, | |
| "logps/ref_chosen": -156.87838745117188, | |
| "logps/ref_rejected": -165.17373657226562, | |
| "logps/rejected": -166.20751953125, | |
| "loss": 0.6931, | |
| "rewards/accuracies": 0.546875, | |
| "rewards/chosen": -0.010160216130316257, | |
| "rewards/margins": 0.00017760891932994127, | |
| "rewards/rejected": -0.010337824933230877, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.14491560312885962, | |
| "grad_norm": 0.6500194668769836, | |
| "learning_rate": 4.975306286336627e-07, | |
| "logits/chosen": 10.476134300231934, | |
| "logits/rejected": 10.66375732421875, | |
| "logps/chosen": -161.99935913085938, | |
| "logps/ref_chosen": -160.73855590820312, | |
| "logps/ref_rejected": -173.1862030029297, | |
| "logps/rejected": -174.4076385498047, | |
| "loss": 0.6934, | |
| "rewards/accuracies": 0.484375, | |
| "rewards/chosen": -0.0126079972833395, | |
| "rewards/margins": -0.00039388981531374156, | |
| "rewards/rejected": -0.012214107438921928, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.15808974886784685, | |
| "grad_norm": 0.5539909601211548, | |
| "learning_rate": 4.956156357188939e-07, | |
| "logits/chosen": 10.318845748901367, | |
| "logits/rejected": 10.355680465698242, | |
| "logps/chosen": -167.43121337890625, | |
| "logps/ref_chosen": -165.21177673339844, | |
| "logps/ref_rejected": -170.47381591796875, | |
| "logps/rejected": -172.76483154296875, | |
| "loss": 0.6928, | |
| "rewards/accuracies": 0.5078125, | |
| "rewards/chosen": -0.022194450721144676, | |
| "rewards/margins": 0.0007156741339713335, | |
| "rewards/rejected": -0.02291012369096279, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.17126389460683408, | |
| "grad_norm": 0.5150694251060486, | |
| "learning_rate": 4.931607263312032e-07, | |
| "logits/chosen": 9.89578914642334, | |
| "logits/rejected": 10.236948013305664, | |
| "logps/chosen": -161.29905700683594, | |
| "logps/ref_chosen": -158.68667602539062, | |
| "logps/ref_rejected": -168.57968139648438, | |
| "logps/rejected": -171.51979064941406, | |
| "loss": 0.6916, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.026123855262994766, | |
| "rewards/margins": 0.0032772955019026995, | |
| "rewards/rejected": -0.029401153326034546, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.1844380403458213, | |
| "grad_norm": 0.6596489548683167, | |
| "learning_rate": 4.9017129689421e-07, | |
| "logits/chosen": 10.484810829162598, | |
| "logits/rejected": 10.99763298034668, | |
| "logps/chosen": -157.44769287109375, | |
| "logps/ref_chosen": -153.92340087890625, | |
| "logps/ref_rejected": -167.03564453125, | |
| "logps/rejected": -170.4557647705078, | |
| "loss": 0.6937, | |
| "rewards/accuracies": 0.4609375, | |
| "rewards/chosen": -0.03524318337440491, | |
| "rewards/margins": -0.0010417333105579019, | |
| "rewards/rejected": -0.034201446920633316, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.19761218608480857, | |
| "grad_norm": 0.39550018310546875, | |
| "learning_rate": 4.866539188226085e-07, | |
| "logits/chosen": 10.189282417297363, | |
| "logits/rejected": 10.43722152709961, | |
| "logps/chosen": -166.56544494628906, | |
| "logps/ref_chosen": -162.66110229492188, | |
| "logps/ref_rejected": -168.7485809326172, | |
| "logps/rejected": -172.78038024902344, | |
| "loss": 0.6926, | |
| "rewards/accuracies": 0.4375, | |
| "rewards/chosen": -0.03904342278838158, | |
| "rewards/margins": 0.0012746157590299845, | |
| "rewards/rejected": -0.04031803831458092, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.2107863318237958, | |
| "grad_norm": 0.6276482939720154, | |
| "learning_rate": 4.826163240767716e-07, | |
| "logits/chosen": 10.743437767028809, | |
| "logits/rejected": 11.031370162963867, | |
| "logps/chosen": -166.45135498046875, | |
| "logps/ref_chosen": -163.39239501953125, | |
| "logps/ref_rejected": -172.29183959960938, | |
| "logps/rejected": -175.4534912109375, | |
| "loss": 0.6927, | |
| "rewards/accuracies": 0.5546875, | |
| "rewards/chosen": -0.03058951534330845, | |
| "rewards/margins": 0.0010271857026964426, | |
| "rewards/rejected": -0.0316167026758194, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.22396047756278303, | |
| "grad_norm": 0.516729474067688, | |
| "learning_rate": 4.780673881662242e-07, | |
| "logits/chosen": 10.332087516784668, | |
| "logits/rejected": 10.48865032196045, | |
| "logps/chosen": -157.08522033691406, | |
| "logps/ref_chosen": -153.6072540283203, | |
| "logps/ref_rejected": -161.9541473388672, | |
| "logps/rejected": -165.6874542236328, | |
| "loss": 0.6919, | |
| "rewards/accuracies": 0.5390625, | |
| "rewards/chosen": -0.03477972373366356, | |
| "rewards/margins": 0.0025533493608236313, | |
| "rewards/rejected": -0.03733307495713234, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.23713462330177026, | |
| "grad_norm": 0.70009446144104, | |
| "learning_rate": 4.730171106393466e-07, | |
| "logits/chosen": 10.40684986114502, | |
| "logits/rejected": 10.725347518920898, | |
| "logps/chosen": -158.2038116455078, | |
| "logps/ref_chosen": -154.3197021484375, | |
| "logps/ref_rejected": -161.81753540039062, | |
| "logps/rejected": -165.58631896972656, | |
| "loss": 0.6938, | |
| "rewards/accuracies": 0.484375, | |
| "rewards/chosen": -0.03884127736091614, | |
| "rewards/margins": -0.0011533537181094289, | |
| "rewards/rejected": -0.03768792748451233, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.2503087690407575, | |
| "grad_norm": 0.47613224387168884, | |
| "learning_rate": 4.6747659310219757e-07, | |
| "logits/chosen": 10.489011764526367, | |
| "logits/rejected": 10.455073356628418, | |
| "logps/chosen": -171.99160766601562, | |
| "logps/ref_chosen": -167.8755340576172, | |
| "logps/ref_rejected": -175.09603881835938, | |
| "logps/rejected": -179.593994140625, | |
| "loss": 0.6913, | |
| "rewards/accuracies": 0.5703125, | |
| "rewards/chosen": -0.04116089642047882, | |
| "rewards/margins": 0.003818710334599018, | |
| "rewards/rejected": -0.04497961327433586, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.2634829147797447, | |
| "grad_norm": 0.6483292579650879, | |
| "learning_rate": 4.6145801481477433e-07, | |
| "logits/chosen": 10.415058135986328, | |
| "logits/rejected": 10.774059295654297, | |
| "logps/chosen": -163.5430450439453, | |
| "logps/ref_chosen": -159.07583618164062, | |
| "logps/ref_rejected": -169.23069763183594, | |
| "logps/rejected": -173.73776245117188, | |
| "loss": 0.693, | |
| "rewards/accuracies": 0.515625, | |
| "rewards/chosen": -0.04467229172587395, | |
| "rewards/margins": 0.00039826278225518763, | |
| "rewards/rejected": -0.04507055878639221, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.276657060518732, | |
| "grad_norm": 0.5634174942970276, | |
| "learning_rate": 4.549746059183561e-07, | |
| "logits/chosen": 10.342830657958984, | |
| "logits/rejected": 10.680377960205078, | |
| "logps/chosen": -163.2490997314453, | |
| "logps/ref_chosen": -159.25521850585938, | |
| "logps/ref_rejected": -169.57681274414062, | |
| "logps/rejected": -173.69276428222656, | |
| "loss": 0.6926, | |
| "rewards/accuracies": 0.578125, | |
| "rewards/chosen": -0.03993882238864899, | |
| "rewards/margins": 0.0012204290833324194, | |
| "rewards/rejected": -0.0411592535674572, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.28983120625771924, | |
| "grad_norm": 0.3794897198677063, | |
| "learning_rate": 4.480406183527823e-07, | |
| "logits/chosen": 10.29517936706543, | |
| "logits/rejected": 10.647847175598145, | |
| "logps/chosen": -161.54783630371094, | |
| "logps/ref_chosen": -157.0568084716797, | |
| "logps/ref_rejected": -163.96209716796875, | |
| "logps/rejected": -168.51736450195312, | |
| "loss": 0.6929, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.04491012915968895, | |
| "rewards/margins": 0.0006425387691706419, | |
| "rewards/rejected": -0.04555266723036766, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.3030053519967065, | |
| "grad_norm": 0.4016757607460022, | |
| "learning_rate": 4.4067129452759546e-07, | |
| "logits/chosen": 10.357274055480957, | |
| "logits/rejected": 10.63122844696045, | |
| "logps/chosen": -162.94578552246094, | |
| "logps/ref_chosen": -158.10250854492188, | |
| "logps/ref_rejected": -169.85250854492188, | |
| "logps/rejected": -174.79525756835938, | |
| "loss": 0.6927, | |
| "rewards/accuracies": 0.5234375, | |
| "rewards/chosen": -0.04843292012810707, | |
| "rewards/margins": 0.0009945080382749438, | |
| "rewards/rejected": -0.04942742735147476, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.3161794977356937, | |
| "grad_norm": 0.46131113171577454, | |
| "learning_rate": 4.3288283381591725e-07, | |
| "logits/chosen": 10.260627746582031, | |
| "logits/rejected": 10.424566268920898, | |
| "logps/chosen": -163.2139129638672, | |
| "logps/ref_chosen": -158.93540954589844, | |
| "logps/ref_rejected": -168.12344360351562, | |
| "logps/rejected": -172.52456665039062, | |
| "loss": 0.6926, | |
| "rewards/accuracies": 0.4921875, | |
| "rewards/chosen": -0.042785100638866425, | |
| "rewards/margins": 0.0012263581156730652, | |
| "rewards/rejected": -0.04401145875453949, | |
| "step": 24 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 75, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 12, | |
| "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": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |