| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9485384932070811, | |
| "eval_steps": 500, | |
| "global_step": 72, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.013174145738987238, | |
| "grad_norm": 0.538081705570221, | |
| "learning_rate": 1.25e-07, | |
| "logits/chosen": 10.038352012634277, | |
| "logits/rejected": 10.592904090881348, | |
| "logps/chosen": -0.6228358745574951, | |
| "logps/rejected": -0.6871199011802673, | |
| "loss": 1.342, | |
| "rewards/accuracies": 0.578125, | |
| "rewards/chosen": -1.2456717491149902, | |
| "rewards/margins": 0.12856802344322205, | |
| "rewards/rejected": -1.3742398023605347, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.026348291477974475, | |
| "grad_norm": 0.6521235108375549, | |
| "learning_rate": 2.5e-07, | |
| "logits/chosen": 10.320584297180176, | |
| "logits/rejected": 10.721946716308594, | |
| "logps/chosen": -0.7115719318389893, | |
| "logps/rejected": -0.788784384727478, | |
| "loss": 1.3147, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -1.4231438636779785, | |
| "rewards/margins": 0.15442489087581635, | |
| "rewards/rejected": -1.577568769454956, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.03952243721696171, | |
| "grad_norm": 0.8797138929367065, | |
| "learning_rate": 3.75e-07, | |
| "logits/chosen": 9.899504661560059, | |
| "logits/rejected": 10.505952835083008, | |
| "logps/chosen": -0.8225007057189941, | |
| "logps/rejected": -0.8832307457923889, | |
| "loss": 1.3674, | |
| "rewards/accuracies": 0.5078125, | |
| "rewards/chosen": -1.6450014114379883, | |
| "rewards/margins": 0.12146000564098358, | |
| "rewards/rejected": -1.7664614915847778, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.05269658295594895, | |
| "grad_norm": 1.9139935970306396, | |
| "learning_rate": 5e-07, | |
| "logits/chosen": 10.082985877990723, | |
| "logits/rejected": 10.576549530029297, | |
| "logps/chosen": -0.6892099976539612, | |
| "logps/rejected": -0.7180394530296326, | |
| "loss": 1.4038, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -1.3784199953079224, | |
| "rewards/margins": 0.05765870213508606, | |
| "rewards/rejected": -1.4360789060592651, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.06587072869493618, | |
| "grad_norm": 0.8647859692573547, | |
| "learning_rate": 6.249999999999999e-07, | |
| "logits/chosen": 10.318564414978027, | |
| "logits/rejected": 11.072587966918945, | |
| "logps/chosen": -0.6658570766448975, | |
| "logps/rejected": -0.6663312911987305, | |
| "loss": 1.4062, | |
| "rewards/accuracies": 0.5703125, | |
| "rewards/chosen": -1.331714153289795, | |
| "rewards/margins": 0.0009482596069574356, | |
| "rewards/rejected": -1.332662582397461, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.07904487443392343, | |
| "grad_norm": 0.7906696796417236, | |
| "learning_rate": 7.5e-07, | |
| "logits/chosen": 10.802580833435059, | |
| "logits/rejected": 11.333773612976074, | |
| "logps/chosen": -0.7257988452911377, | |
| "logps/rejected": -0.7839725017547607, | |
| "loss": 1.3781, | |
| "rewards/accuracies": 0.6015625, | |
| "rewards/chosen": -1.4515976905822754, | |
| "rewards/margins": 0.11634734272956848, | |
| "rewards/rejected": -1.5679450035095215, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.09221902017291066, | |
| "grad_norm": 0.724219799041748, | |
| "learning_rate": 8.75e-07, | |
| "logits/chosen": 9.928263664245605, | |
| "logits/rejected": 10.422144889831543, | |
| "logps/chosen": -0.5926575660705566, | |
| "logps/rejected": -0.6688517928123474, | |
| "loss": 1.314, | |
| "rewards/accuracies": 0.5859375, | |
| "rewards/chosen": -1.1853151321411133, | |
| "rewards/margins": 0.15238842368125916, | |
| "rewards/rejected": -1.3377035856246948, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.1053931659118979, | |
| "grad_norm": 0.558660089969635, | |
| "learning_rate": 1e-06, | |
| "logits/chosen": 10.657012939453125, | |
| "logits/rejected": 11.171004295349121, | |
| "logps/chosen": -0.6659789681434631, | |
| "logps/rejected": -0.7012848258018494, | |
| "loss": 1.365, | |
| "rewards/accuracies": 0.578125, | |
| "rewards/chosen": -1.3319579362869263, | |
| "rewards/margins": 0.07061176747083664, | |
| "rewards/rejected": -1.4025696516036987, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.11856731165088513, | |
| "grad_norm": 0.7675488591194153, | |
| "learning_rate": 9.994504457428556e-07, | |
| "logits/chosen": 10.544637680053711, | |
| "logits/rejected": 10.839460372924805, | |
| "logps/chosen": -0.814159095287323, | |
| "logps/rejected": -0.7815468907356262, | |
| "loss": 1.4888, | |
| "rewards/accuracies": 0.515625, | |
| "rewards/chosen": -1.628318190574646, | |
| "rewards/margins": -0.0652243047952652, | |
| "rewards/rejected": -1.5630937814712524, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.13174145738987236, | |
| "grad_norm": 1.1157082319259644, | |
| "learning_rate": 9.97802991010949e-07, | |
| "logits/chosen": 10.10114574432373, | |
| "logits/rejected": 10.555818557739258, | |
| "logps/chosen": -0.673196017742157, | |
| "logps/rejected": -0.6864349246025085, | |
| "loss": 1.4279, | |
| "rewards/accuracies": 0.5546875, | |
| "rewards/chosen": -1.346392035484314, | |
| "rewards/margins": 0.02647773176431656, | |
| "rewards/rejected": -1.372869849205017, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.14491560312885962, | |
| "grad_norm": 1.2121509313583374, | |
| "learning_rate": 9.950612572673255e-07, | |
| "logits/chosen": 10.158075332641602, | |
| "logits/rejected": 10.813385009765625, | |
| "logps/chosen": -0.7734582424163818, | |
| "logps/rejected": -0.8254096508026123, | |
| "loss": 1.3538, | |
| "rewards/accuracies": 0.5546875, | |
| "rewards/chosen": -1.5469164848327637, | |
| "rewards/margins": 0.10390281677246094, | |
| "rewards/rejected": -1.6508193016052246, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.15808974886784685, | |
| "grad_norm": 0.921929657459259, | |
| "learning_rate": 9.912312714377879e-07, | |
| "logits/chosen": 10.2570161819458, | |
| "logits/rejected": 10.633421897888184, | |
| "logps/chosen": -0.7107410430908203, | |
| "logps/rejected": -0.7390152812004089, | |
| "loss": 1.3791, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -1.4214820861816406, | |
| "rewards/margins": 0.056548528373241425, | |
| "rewards/rejected": -1.4780305624008179, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.17126389460683408, | |
| "grad_norm": 1.0088778734207153, | |
| "learning_rate": 9.863214526624063e-07, | |
| "logits/chosen": 9.808923721313477, | |
| "logits/rejected": 10.603569984436035, | |
| "logps/chosen": -0.7246454954147339, | |
| "logps/rejected": -0.8062013387680054, | |
| "loss": 1.3588, | |
| "rewards/accuracies": 0.5234375, | |
| "rewards/chosen": -1.4492909908294678, | |
| "rewards/margins": 0.1631116420030594, | |
| "rewards/rejected": -1.6124026775360107, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.1844380403458213, | |
| "grad_norm": 1.2618002891540527, | |
| "learning_rate": 9.8034259378842e-07, | |
| "logits/chosen": 10.618194580078125, | |
| "logits/rejected": 11.330740928649902, | |
| "logps/chosen": -0.6871081590652466, | |
| "logps/rejected": -0.752004086971283, | |
| "loss": 1.3318, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -1.3742163181304932, | |
| "rewards/margins": 0.1297919750213623, | |
| "rewards/rejected": -1.504008173942566, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.19761218608480857, | |
| "grad_norm": 0.3426913917064667, | |
| "learning_rate": 9.73307837645217e-07, | |
| "logits/chosen": 9.945769309997559, | |
| "logits/rejected": 10.648375511169434, | |
| "logps/chosen": -0.6394751071929932, | |
| "logps/rejected": -0.7090991139411926, | |
| "loss": 1.3241, | |
| "rewards/accuracies": 0.546875, | |
| "rewards/chosen": -1.2789502143859863, | |
| "rewards/margins": 0.13924814760684967, | |
| "rewards/rejected": -1.4181982278823853, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.2107863318237958, | |
| "grad_norm": 1.1439241170883179, | |
| "learning_rate": 9.652326481535433e-07, | |
| "logits/chosen": 10.895740509033203, | |
| "logits/rejected": 11.243146896362305, | |
| "logps/chosen": -0.6249470710754395, | |
| "logps/rejected": -0.6594743132591248, | |
| "loss": 1.3593, | |
| "rewards/accuracies": 0.4765625, | |
| "rewards/chosen": -1.249894142150879, | |
| "rewards/margins": 0.06905444711446762, | |
| "rewards/rejected": -1.3189486265182495, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.22396047756278303, | |
| "grad_norm": 0.6084972023963928, | |
| "learning_rate": 9.561347763324483e-07, | |
| "logits/chosen": 10.294602394104004, | |
| "logits/rejected": 10.75848388671875, | |
| "logps/chosen": -0.6647714972496033, | |
| "logps/rejected": -0.6630449891090393, | |
| "loss": 1.4144, | |
| "rewards/accuracies": 0.5703125, | |
| "rewards/chosen": -1.3295429944992065, | |
| "rewards/margins": -0.0034531853161752224, | |
| "rewards/rejected": -1.3260899782180786, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.23713462330177026, | |
| "grad_norm": 1.20784330368042, | |
| "learning_rate": 9.460342212786932e-07, | |
| "logits/chosen": 10.368552207946777, | |
| "logits/rejected": 11.025596618652344, | |
| "logps/chosen": -0.8019055724143982, | |
| "logps/rejected": -0.7324545979499817, | |
| "loss": 1.5474, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -1.6038111448287964, | |
| "rewards/margins": -0.13890185952186584, | |
| "rewards/rejected": -1.4649091958999634, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.2503087690407575, | |
| "grad_norm": 0.6674954891204834, | |
| "learning_rate": 9.349531862043951e-07, | |
| "logits/chosen": 10.525771141052246, | |
| "logits/rejected": 10.864877700805664, | |
| "logps/chosen": -0.6636431217193604, | |
| "logps/rejected": -0.690964937210083, | |
| "loss": 1.3837, | |
| "rewards/accuracies": 0.640625, | |
| "rewards/chosen": -1.3272862434387207, | |
| "rewards/margins": 0.05464361608028412, | |
| "rewards/rejected": -1.381929874420166, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.2634829147797447, | |
| "grad_norm": 0.9430455565452576, | |
| "learning_rate": 9.229160296295487e-07, | |
| "logits/chosen": 10.509737014770508, | |
| "logits/rejected": 11.218311309814453, | |
| "logps/chosen": -0.7572717666625977, | |
| "logps/rejected": -0.7445338368415833, | |
| "loss": 1.4599, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -1.5145435333251953, | |
| "rewards/margins": -0.02547581121325493, | |
| "rewards/rejected": -1.4890676736831665, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.276657060518732, | |
| "grad_norm": 0.5352253317832947, | |
| "learning_rate": 9.099492118367122e-07, | |
| "logits/chosen": 9.723971366882324, | |
| "logits/rejected": 10.554994583129883, | |
| "logps/chosen": -0.706066906452179, | |
| "logps/rejected": -0.7981133460998535, | |
| "loss": 1.3079, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -1.412133812904358, | |
| "rewards/margins": 0.18409286439418793, | |
| "rewards/rejected": -1.596226692199707, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.28983120625771924, | |
| "grad_norm": 1.4943097829818726, | |
| "learning_rate": 8.960812367055646e-07, | |
| "logits/chosen": 10.025705337524414, | |
| "logits/rejected": 10.595457077026367, | |
| "logps/chosen": -0.7785383462905884, | |
| "logps/rejected": -0.7879722118377686, | |
| "loss": 1.4168, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -1.5570766925811768, | |
| "rewards/margins": 0.01886790432035923, | |
| "rewards/rejected": -1.575944423675537, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.3030053519967065, | |
| "grad_norm": 0.4453490972518921, | |
| "learning_rate": 8.813425890551909e-07, | |
| "logits/chosen": 10.01511001586914, | |
| "logits/rejected": 10.817825317382812, | |
| "logps/chosen": -0.7974975109100342, | |
| "logps/rejected": -0.7531540393829346, | |
| "loss": 1.5107, | |
| "rewards/accuracies": 0.5078125, | |
| "rewards/chosen": -1.5949950218200684, | |
| "rewards/margins": -0.08868695795536041, | |
| "rewards/rejected": -1.5063080787658691, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.3161794977356937, | |
| "grad_norm": 0.8292773962020874, | |
| "learning_rate": 8.657656676318345e-07, | |
| "logits/chosen": 10.253622055053711, | |
| "logits/rejected": 10.707040786743164, | |
| "logps/chosen": -0.6545951962471008, | |
| "logps/rejected": -0.6871030330657959, | |
| "loss": 1.3837, | |
| "rewards/accuracies": 0.546875, | |
| "rewards/chosen": -1.3091903924942017, | |
| "rewards/margins": 0.06501554697751999, | |
| "rewards/rejected": -1.3742060661315918, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.32935364347468093, | |
| "grad_norm": 3.5230870246887207, | |
| "learning_rate": 8.493847138894208e-07, | |
| "logits/chosen": 10.166427612304688, | |
| "logits/rejected": 10.720212936401367, | |
| "logps/chosen": -0.6964614391326904, | |
| "logps/rejected": -0.7443124055862427, | |
| "loss": 1.3755, | |
| "rewards/accuracies": 0.5546875, | |
| "rewards/chosen": -1.3929228782653809, | |
| "rewards/margins": 0.09570197016000748, | |
| "rewards/rejected": -1.4886248111724854, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.34252778921366817, | |
| "grad_norm": 1.1598762273788452, | |
| "learning_rate": 8.322357367194108e-07, | |
| "logits/chosen": 10.49393367767334, | |
| "logits/rejected": 11.039409637451172, | |
| "logps/chosen": -0.7815979719161987, | |
| "logps/rejected": -0.8116061687469482, | |
| "loss": 1.3981, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -1.5631959438323975, | |
| "rewards/margins": 0.06001615524291992, | |
| "rewards/rejected": -1.6232123374938965, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.3557019349526554, | |
| "grad_norm": 2.6781487464904785, | |
| "learning_rate": 8.143564332954425e-07, | |
| "logits/chosen": 10.239036560058594, | |
| "logits/rejected": 11.069831848144531, | |
| "logps/chosen": -0.7489575147628784, | |
| "logps/rejected": -0.7497468590736389, | |
| "loss": 1.4293, | |
| "rewards/accuracies": 0.4921875, | |
| "rewards/chosen": -1.4979150295257568, | |
| "rewards/margins": 0.0015787146985530853, | |
| "rewards/rejected": -1.4994937181472778, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.3688760806916426, | |
| "grad_norm": 0.4053559899330139, | |
| "learning_rate": 7.957861062067612e-07, | |
| "logits/chosen": 9.529670715332031, | |
| "logits/rejected": 10.38549518585205, | |
| "logps/chosen": -0.7035645246505737, | |
| "logps/rejected": -0.7123870253562927, | |
| "loss": 1.4258, | |
| "rewards/accuracies": 0.546875, | |
| "rewards/chosen": -1.4071290493011475, | |
| "rewards/margins": 0.017645111307501793, | |
| "rewards/rejected": -1.4247740507125854, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.3820502264306299, | |
| "grad_norm": 0.6468765735626221, | |
| "learning_rate": 7.765655770625996e-07, | |
| "logits/chosen": 10.515686988830566, | |
| "logits/rejected": 10.88726806640625, | |
| "logps/chosen": -0.6611747145652771, | |
| "logps/rejected": -0.7040727734565735, | |
| "loss": 1.368, | |
| "rewards/accuracies": 0.6328125, | |
| "rewards/chosen": -1.3223494291305542, | |
| "rewards/margins": 0.08579609543085098, | |
| "rewards/rejected": -1.408145546913147, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.39522437216961714, | |
| "grad_norm": 0.6986653208732605, | |
| "learning_rate": 7.567370967574209e-07, | |
| "logits/chosen": 10.111821174621582, | |
| "logits/rejected": 11.31137466430664, | |
| "logps/chosen": -0.6936085820198059, | |
| "logps/rejected": -0.6943917274475098, | |
| "loss": 1.4297, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -1.3872171640396118, | |
| "rewards/margins": 0.0015661753714084625, | |
| "rewards/rejected": -1.3887834548950195, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.4083985179086044, | |
| "grad_norm": 0.4217478334903717, | |
| "learning_rate": 7.363442525942826e-07, | |
| "logits/chosen": 9.911711692810059, | |
| "logits/rejected": 10.704763412475586, | |
| "logps/chosen": -0.6641364097595215, | |
| "logps/rejected": -0.6581347584724426, | |
| "loss": 1.4344, | |
| "rewards/accuracies": 0.546875, | |
| "rewards/chosen": -1.328272819519043, | |
| "rewards/margins": -0.012003323063254356, | |
| "rewards/rejected": -1.3162695169448853, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.4215726636475916, | |
| "grad_norm": 0.6227706670761108, | |
| "learning_rate": 7.154318724704851e-07, | |
| "logits/chosen": 9.962979316711426, | |
| "logits/rejected": 10.975818634033203, | |
| "logps/chosen": -0.6416503190994263, | |
| "logps/rejected": -0.6450206637382507, | |
| "loss": 1.4274, | |
| "rewards/accuracies": 0.4765625, | |
| "rewards/chosen": -1.2833006381988525, | |
| "rewards/margins": 0.006740846671164036, | |
| "rewards/rejected": -1.2900413274765015, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.43474680938657884, | |
| "grad_norm": 0.9526137709617615, | |
| "learning_rate": 6.940459263361248e-07, | |
| "logits/chosen": 10.01764965057373, | |
| "logits/rejected": 10.877232551574707, | |
| "logps/chosen": -0.7044752836227417, | |
| "logps/rejected": -0.6560808420181274, | |
| "loss": 1.5103, | |
| "rewards/accuracies": 0.4765625, | |
| "rewards/chosen": -1.4089505672454834, | |
| "rewards/margins": -0.09678899496793747, | |
| "rewards/rejected": -1.3121616840362549, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.44792095512556607, | |
| "grad_norm": 0.7180893421173096, | |
| "learning_rate": 6.722334251421664e-07, | |
| "logits/chosen": 9.873922348022461, | |
| "logits/rejected": 10.598325729370117, | |
| "logps/chosen": -0.7634012699127197, | |
| "logps/rejected": -0.7632129192352295, | |
| "loss": 1.4281, | |
| "rewards/accuracies": 0.5859375, | |
| "rewards/chosen": -1.5268025398254395, | |
| "rewards/margins": -0.00037665292620658875, | |
| "rewards/rejected": -1.526425838470459, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.4610951008645533, | |
| "grad_norm": 1.9519481658935547, | |
| "learning_rate": 6.500423175001703e-07, | |
| "logits/chosen": 10.66737174987793, | |
| "logits/rejected": 11.170511245727539, | |
| "logps/chosen": -0.7307225465774536, | |
| "logps/rejected": -0.7907586097717285, | |
| "loss": 1.3918, | |
| "rewards/accuracies": 0.5390625, | |
| "rewards/chosen": -1.4614450931549072, | |
| "rewards/margins": 0.12007206678390503, | |
| "rewards/rejected": -1.581517219543457, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.47426924660354053, | |
| "grad_norm": 0.7991163730621338, | |
| "learning_rate": 6.275213842808382e-07, | |
| "logits/chosen": 10.241875648498535, | |
| "logits/rejected": 10.639405250549316, | |
| "logps/chosen": -0.6178256869316101, | |
| "logps/rejected": -0.6322227120399475, | |
| "loss": 1.3976, | |
| "rewards/accuracies": 0.4921875, | |
| "rewards/chosen": -1.2356513738632202, | |
| "rewards/margins": 0.028794117271900177, | |
| "rewards/rejected": -1.264445424079895, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.4874433923425278, | |
| "grad_norm": 0.6113852262496948, | |
| "learning_rate": 6.047201313830723e-07, | |
| "logits/chosen": 10.230165481567383, | |
| "logits/rejected": 10.763371467590332, | |
| "logps/chosen": -0.6862261295318604, | |
| "logps/rejected": -0.7486332654953003, | |
| "loss": 1.3326, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -1.3724522590637207, | |
| "rewards/margins": 0.12481416761875153, | |
| "rewards/rejected": -1.4972665309906006, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.500617538081515, | |
| "grad_norm": 1.0607084035873413, | |
| "learning_rate": 5.816886809092651e-07, | |
| "logits/chosen": 10.049585342407227, | |
| "logits/rejected": 10.759794235229492, | |
| "logps/chosen": -0.6145029067993164, | |
| "logps/rejected": -0.6638337969779968, | |
| "loss": 1.3483, | |
| "rewards/accuracies": 0.5859375, | |
| "rewards/chosen": -1.2290058135986328, | |
| "rewards/margins": 0.0986616313457489, | |
| "rewards/rejected": -1.3276675939559937, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.5137916838205022, | |
| "grad_norm": 1.5060722827911377, | |
| "learning_rate": 5.584776609860413e-07, | |
| "logits/chosen": 10.381738662719727, | |
| "logits/rejected": 10.864119529724121, | |
| "logps/chosen": -0.7063203454017639, | |
| "logps/rejected": -0.6916000247001648, | |
| "loss": 1.4498, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -1.4126406908035278, | |
| "rewards/margins": -0.029440607875585556, | |
| "rewards/rejected": -1.3832000494003296, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.5269658295594895, | |
| "grad_norm": 2.536262273788452, | |
| "learning_rate": 5.351380944726465e-07, | |
| "logits/chosen": 10.201406478881836, | |
| "logits/rejected": 11.195016860961914, | |
| "logps/chosen": -0.6529728174209595, | |
| "logps/rejected": -0.695421576499939, | |
| "loss": 1.3726, | |
| "rewards/accuracies": 0.640625, | |
| "rewards/chosen": -1.305945634841919, | |
| "rewards/margins": 0.08489762991666794, | |
| "rewards/rejected": -1.390843152999878, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.5401399752984768, | |
| "grad_norm": 3.44834303855896, | |
| "learning_rate": 5.117212868016303e-07, | |
| "logits/chosen": 10.57003402709961, | |
| "logits/rejected": 10.89995288848877, | |
| "logps/chosen": -0.6615808606147766, | |
| "logps/rejected": -0.6666579842567444, | |
| "loss": 1.4101, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -1.3231617212295532, | |
| "rewards/margins": 0.010154157876968384, | |
| "rewards/rejected": -1.3333159685134888, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.553314121037464, | |
| "grad_norm": 0.3262707591056824, | |
| "learning_rate": 4.882787131983697e-07, | |
| "logits/chosen": 9.684130668640137, | |
| "logits/rejected": 10.468993186950684, | |
| "logps/chosen": -0.6019183993339539, | |
| "logps/rejected": -0.6411232948303223, | |
| "loss": 1.3798, | |
| "rewards/accuracies": 0.5078125, | |
| "rewards/chosen": -1.2038367986679077, | |
| "rewards/margins": 0.07840971648693085, | |
| "rewards/rejected": -1.2822465896606445, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.5664882667764513, | |
| "grad_norm": 1.150577187538147, | |
| "learning_rate": 4.648619055273537e-07, | |
| "logits/chosen": 9.844969749450684, | |
| "logits/rejected": 10.612713813781738, | |
| "logps/chosen": -0.5501813292503357, | |
| "logps/rejected": -0.591031551361084, | |
| "loss": 1.3513, | |
| "rewards/accuracies": 0.6171875, | |
| "rewards/chosen": -1.1003626585006714, | |
| "rewards/margins": 0.08170032501220703, | |
| "rewards/rejected": -1.182063102722168, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.5796624125154385, | |
| "grad_norm": 0.3965461850166321, | |
| "learning_rate": 4.4152233901395875e-07, | |
| "logits/chosen": 9.765134811401367, | |
| "logits/rejected": 10.664865493774414, | |
| "logps/chosen": -0.689943253993988, | |
| "logps/rejected": -0.6770355701446533, | |
| "loss": 1.4345, | |
| "rewards/accuracies": 0.5703125, | |
| "rewards/chosen": -1.379886507987976, | |
| "rewards/margins": -0.025815514847636223, | |
| "rewards/rejected": -1.3540711402893066, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.5928365582544257, | |
| "grad_norm": 0.5376319289207458, | |
| "learning_rate": 4.183113190907348e-07, | |
| "logits/chosen": 10.018446922302246, | |
| "logits/rejected": 10.538941383361816, | |
| "logps/chosen": -0.6233353614807129, | |
| "logps/rejected": -0.633882999420166, | |
| "loss": 1.4166, | |
| "rewards/accuracies": 0.5390625, | |
| "rewards/chosen": -1.2466707229614258, | |
| "rewards/margins": 0.02109522372484207, | |
| "rewards/rejected": -1.267765998840332, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.606010703993413, | |
| "grad_norm": 3.684044122695923, | |
| "learning_rate": 3.9527986861692785e-07, | |
| "logits/chosen": 10.023819923400879, | |
| "logits/rejected": 10.466320037841797, | |
| "logps/chosen": -0.644775927066803, | |
| "logps/rejected": -0.6808078289031982, | |
| "loss": 1.3706, | |
| "rewards/accuracies": 0.5703125, | |
| "rewards/chosen": -1.289551854133606, | |
| "rewards/margins": 0.07206393778324127, | |
| "rewards/rejected": -1.3616156578063965, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.6191848497324002, | |
| "grad_norm": 0.5732494592666626, | |
| "learning_rate": 3.724786157191618e-07, | |
| "logits/chosen": 10.115565299987793, | |
| "logits/rejected": 10.92066764831543, | |
| "logps/chosen": -0.7901978492736816, | |
| "logps/rejected": -0.7942913174629211, | |
| "loss": 1.4251, | |
| "rewards/accuracies": 0.515625, | |
| "rewards/chosen": -1.5803956985473633, | |
| "rewards/margins": 0.008186978287994862, | |
| "rewards/rejected": -1.5885826349258423, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.6323589954713874, | |
| "grad_norm": 0.6679599285125732, | |
| "learning_rate": 3.499576824998297e-07, | |
| "logits/chosen": 10.542475700378418, | |
| "logits/rejected": 11.161402702331543, | |
| "logps/chosen": -0.6402223706245422, | |
| "logps/rejected": -0.6876631379127502, | |
| "loss": 1.3518, | |
| "rewards/accuracies": 0.6328125, | |
| "rewards/chosen": -1.2804447412490845, | |
| "rewards/margins": 0.09488136321306229, | |
| "rewards/rejected": -1.3753262758255005, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.6455331412103746, | |
| "grad_norm": 0.45617973804473877, | |
| "learning_rate": 3.2776657485783356e-07, | |
| "logits/chosen": 10.640542030334473, | |
| "logits/rejected": 11.157126426696777, | |
| "logps/chosen": -0.6611472368240356, | |
| "logps/rejected": -0.6986327767372131, | |
| "loss": 1.3795, | |
| "rewards/accuracies": 0.5390625, | |
| "rewards/chosen": -1.3222944736480713, | |
| "rewards/margins": 0.07497115433216095, | |
| "rewards/rejected": -1.3972655534744263, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.6587072869493619, | |
| "grad_norm": 0.718313992023468, | |
| "learning_rate": 3.0595407366387506e-07, | |
| "logits/chosen": 10.246100425720215, | |
| "logits/rejected": 10.788708686828613, | |
| "logps/chosen": -0.5527799725532532, | |
| "logps/rejected": -0.5676759481430054, | |
| "loss": 1.383, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -1.1055599451065063, | |
| "rewards/margins": 0.02979196421802044, | |
| "rewards/rejected": -1.1353518962860107, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.6718814326883491, | |
| "grad_norm": 0.18990406394004822, | |
| "learning_rate": 2.845681275295148e-07, | |
| "logits/chosen": 10.160649299621582, | |
| "logits/rejected": 10.608582496643066, | |
| "logps/chosen": -0.5916914343833923, | |
| "logps/rejected": -0.5999540090560913, | |
| "loss": 1.4021, | |
| "rewards/accuracies": 0.5078125, | |
| "rewards/chosen": -1.1833828687667847, | |
| "rewards/margins": 0.016525164246559143, | |
| "rewards/rejected": -1.1999080181121826, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.6850555784273363, | |
| "grad_norm": 0.48461204767227173, | |
| "learning_rate": 2.636557474057173e-07, | |
| "logits/chosen": 9.736953735351562, | |
| "logits/rejected": 10.655252456665039, | |
| "logps/chosen": -0.6059479117393494, | |
| "logps/rejected": -0.65425705909729, | |
| "loss": 1.3626, | |
| "rewards/accuracies": 0.5546875, | |
| "rewards/chosen": -1.2118958234786987, | |
| "rewards/margins": 0.09661829471588135, | |
| "rewards/rejected": -1.30851411819458, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.6982297241663236, | |
| "grad_norm": 0.33438029885292053, | |
| "learning_rate": 2.432629032425789e-07, | |
| "logits/chosen": 10.174696922302246, | |
| "logits/rejected": 10.733561515808105, | |
| "logps/chosen": -0.6094481348991394, | |
| "logps/rejected": -0.6400697827339172, | |
| "loss": 1.3683, | |
| "rewards/accuracies": 0.5078125, | |
| "rewards/chosen": -1.2188962697982788, | |
| "rewards/margins": 0.06124337017536163, | |
| "rewards/rejected": -1.2801395654678345, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.7114038699053108, | |
| "grad_norm": 0.3054359257221222, | |
| "learning_rate": 2.2343442293740028e-07, | |
| "logits/chosen": 10.520360946655273, | |
| "logits/rejected": 11.13870620727539, | |
| "logps/chosen": -0.6996334195137024, | |
| "logps/rejected": -0.6852084994316101, | |
| "loss": 1.4469, | |
| "rewards/accuracies": 0.484375, | |
| "rewards/chosen": -1.3992668390274048, | |
| "rewards/margins": -0.028849666938185692, | |
| "rewards/rejected": -1.3704169988632202, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.724578015644298, | |
| "grad_norm": 1.6864748001098633, | |
| "learning_rate": 2.0421389379323877e-07, | |
| "logits/chosen": 10.444706916809082, | |
| "logits/rejected": 10.993450164794922, | |
| "logps/chosen": -0.6644651293754578, | |
| "logps/rejected": -0.732700765132904, | |
| "loss": 1.3438, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -1.3289302587509155, | |
| "rewards/margins": 0.1364712119102478, | |
| "rewards/rejected": -1.465401530265808, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.7377521613832853, | |
| "grad_norm": 1.1142092943191528, | |
| "learning_rate": 1.8564356670455767e-07, | |
| "logits/chosen": 10.173879623413086, | |
| "logits/rejected": 10.971562385559082, | |
| "logps/chosen": -0.6234769225120544, | |
| "logps/rejected": -0.682326078414917, | |
| "loss": 1.3395, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.2469538450241089, | |
| "rewards/margins": 0.11769835650920868, | |
| "rewards/rejected": -1.364652156829834, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.7509263071222725, | |
| "grad_norm": 1.3884321451187134, | |
| "learning_rate": 1.6776426328058919e-07, | |
| "logits/chosen": 10.79382038116455, | |
| "logits/rejected": 11.301587104797363, | |
| "logps/chosen": -0.6458045244216919, | |
| "logps/rejected": -0.6631965041160583, | |
| "loss": 1.3818, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -1.2916090488433838, | |
| "rewards/margins": 0.03478388115763664, | |
| "rewards/rejected": -1.3263930082321167, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.7641004528612598, | |
| "grad_norm": 0.9691101312637329, | |
| "learning_rate": 1.5061528611057915e-07, | |
| "logits/chosen": 10.396957397460938, | |
| "logits/rejected": 10.836028099060059, | |
| "logps/chosen": -0.6224152445793152, | |
| "logps/rejected": -0.6705679893493652, | |
| "loss": 1.3657, | |
| "rewards/accuracies": 0.5859375, | |
| "rewards/chosen": -1.2448304891586304, | |
| "rewards/margins": 0.09630556404590607, | |
| "rewards/rejected": -1.3411359786987305, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.7772745986002471, | |
| "grad_norm": 0.5600712895393372, | |
| "learning_rate": 1.3423433236816562e-07, | |
| "logits/chosen": 10.439901351928711, | |
| "logits/rejected": 11.005784034729004, | |
| "logps/chosen": -0.6858406662940979, | |
| "logps/rejected": -0.7000098824501038, | |
| "loss": 1.4016, | |
| "rewards/accuracies": 0.5390625, | |
| "rewards/chosen": -1.3716813325881958, | |
| "rewards/margins": 0.028338586911559105, | |
| "rewards/rejected": -1.4000197649002075, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.7904487443392343, | |
| "grad_norm": 0.4134702980518341, | |
| "learning_rate": 1.1865741094480908e-07, | |
| "logits/chosen": 10.502270698547363, | |
| "logits/rejected": 11.183271408081055, | |
| "logps/chosen": -0.606414794921875, | |
| "logps/rejected": -0.6565254926681519, | |
| "loss": 1.333, | |
| "rewards/accuracies": 0.6171875, | |
| "rewards/chosen": -1.21282958984375, | |
| "rewards/margins": 0.10022131353616714, | |
| "rewards/rejected": -1.3130509853363037, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.8036228900782215, | |
| "grad_norm": 1.311030626296997, | |
| "learning_rate": 1.0391876329443533e-07, | |
| "logits/chosen": 10.570302963256836, | |
| "logits/rejected": 11.035686492919922, | |
| "logps/chosen": -0.6471387147903442, | |
| "logps/rejected": -0.6374362707138062, | |
| "loss": 1.4409, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -1.2942774295806885, | |
| "rewards/margins": -0.01940501108765602, | |
| "rewards/rejected": -1.2748725414276123, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.8167970358172087, | |
| "grad_norm": 0.7308284044265747, | |
| "learning_rate": 9.00507881632877e-08, | |
| "logits/chosen": 10.05208969116211, | |
| "logits/rejected": 10.694860458374023, | |
| "logps/chosen": -0.565160870552063, | |
| "logps/rejected": -0.5756160616874695, | |
| "loss": 1.3922, | |
| "rewards/accuracies": 0.5546875, | |
| "rewards/chosen": -1.130321741104126, | |
| "rewards/margins": 0.020910188555717468, | |
| "rewards/rejected": -1.151232123374939, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.829971181556196, | |
| "grad_norm": 0.6492581367492676, | |
| "learning_rate": 7.708397037045128e-08, | |
| "logits/chosen": 10.530025482177734, | |
| "logits/rejected": 11.059508323669434, | |
| "logps/chosen": -0.7004156112670898, | |
| "logps/rejected": -0.6723081469535828, | |
| "loss": 1.4625, | |
| "rewards/accuracies": 0.484375, | |
| "rewards/chosen": -1.4008312225341797, | |
| "rewards/margins": -0.05621491000056267, | |
| "rewards/rejected": -1.3446162939071655, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.8431453272951832, | |
| "grad_norm": 0.6271635293960571, | |
| "learning_rate": 6.504681379560489e-08, | |
| "logits/chosen": 10.55631160736084, | |
| "logits/rejected": 11.027376174926758, | |
| "logps/chosen": -0.664703905582428, | |
| "logps/rejected": -0.6908209323883057, | |
| "loss": 1.3858, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -1.329407811164856, | |
| "rewards/margins": 0.0522342287003994, | |
| "rewards/rejected": -1.3816418647766113, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.8563194730341704, | |
| "grad_norm": 0.5429289937019348, | |
| "learning_rate": 5.396577872130675e-08, | |
| "logits/chosen": 10.111677169799805, | |
| "logits/rejected": 10.929055213928223, | |
| "logps/chosen": -0.5984624028205872, | |
| "logps/rejected": -0.6411637663841248, | |
| "loss": 1.3614, | |
| "rewards/accuracies": 0.515625, | |
| "rewards/chosen": -1.1969248056411743, | |
| "rewards/margins": 0.08540263772010803, | |
| "rewards/rejected": -1.2823275327682495, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.8694936187731577, | |
| "grad_norm": 0.5303859710693359, | |
| "learning_rate": 4.3865223667551686e-08, | |
| "logits/chosen": 10.389065742492676, | |
| "logits/rejected": 10.994528770446777, | |
| "logps/chosen": -0.7132977843284607, | |
| "logps/rejected": -0.7277129888534546, | |
| "loss": 1.3985, | |
| "rewards/accuracies": 0.578125, | |
| "rewards/chosen": -1.4265955686569214, | |
| "rewards/margins": 0.02883046492934227, | |
| "rewards/rejected": -1.4554259777069092, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.8826677645121449, | |
| "grad_norm": 0.8996275067329407, | |
| "learning_rate": 3.476735184645674e-08, | |
| "logits/chosen": 10.195171356201172, | |
| "logits/rejected": 10.991616249084473, | |
| "logps/chosen": -0.6438397765159607, | |
| "logps/rejected": -0.6550954580307007, | |
| "loss": 1.4138, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -1.2876795530319214, | |
| "rewards/margins": 0.02251136302947998, | |
| "rewards/rejected": -1.3101909160614014, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.8958419102511321, | |
| "grad_norm": 0.48269936442375183, | |
| "learning_rate": 2.6692162354782943e-08, | |
| "logits/chosen": 10.73905086517334, | |
| "logits/rejected": 11.05859661102295, | |
| "logps/chosen": -0.6217143535614014, | |
| "logps/rejected": -0.6485203504562378, | |
| "loss": 1.3768, | |
| "rewards/accuracies": 0.609375, | |
| "rewards/chosen": -1.2434287071228027, | |
| "rewards/margins": 0.053611982613801956, | |
| "rewards/rejected": -1.2970407009124756, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.9090160559901194, | |
| "grad_norm": 0.6163386702537537, | |
| "learning_rate": 1.9657406211579962e-08, | |
| "logits/chosen": 10.56104850769043, | |
| "logits/rejected": 11.248091697692871, | |
| "logps/chosen": -0.6829113364219666, | |
| "logps/rejected": -0.7346217036247253, | |
| "loss": 1.3653, | |
| "rewards/accuracies": 0.5859375, | |
| "rewards/chosen": -1.365822672843933, | |
| "rewards/margins": 0.10342076420783997, | |
| "rewards/rejected": -1.4692434072494507, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.9221902017291066, | |
| "grad_norm": 0.9636749625205994, | |
| "learning_rate": 1.3678547337593494e-08, | |
| "logits/chosen": 10.036705017089844, | |
| "logits/rejected": 10.577312469482422, | |
| "logps/chosen": -0.6028749942779541, | |
| "logps/rejected": -0.6518182754516602, | |
| "loss": 1.3437, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -1.2057499885559082, | |
| "rewards/margins": 0.09788656234741211, | |
| "rewards/rejected": -1.3036365509033203, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.9353643474680938, | |
| "grad_norm": 0.7640946507453918, | |
| "learning_rate": 8.768728562211946e-09, | |
| "logits/chosen": 10.159095764160156, | |
| "logits/rejected": 11.175093650817871, | |
| "logps/chosen": -0.6173404455184937, | |
| "logps/rejected": -0.662550151348114, | |
| "loss": 1.3734, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -1.2346808910369873, | |
| "rewards/margins": 0.09041938930749893, | |
| "rewards/rejected": -1.325100302696228, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.9485384932070811, | |
| "grad_norm": 0.5827455520629883, | |
| "learning_rate": 4.938742732674528e-09, | |
| "logits/chosen": 10.687435150146484, | |
| "logits/rejected": 11.472260475158691, | |
| "logps/chosen": -0.48376959562301636, | |
| "logps/rejected": -0.5084850192070007, | |
| "loss": 1.3663, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -0.9675391912460327, | |
| "rewards/margins": 0.04943079501390457, | |
| "rewards/rejected": -1.0169700384140015, | |
| "step": 72 | |
| } | |
| ], | |
| "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 | |
| } | |