| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.47426924660354053, | |
| "eval_steps": 500, | |
| "global_step": 36, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.013174145738987238, | |
| "grad_norm": 0.5080101490020752, | |
| "learning_rate": 1.25e-07, | |
| "logits/chosen": 10.32492733001709, | |
| "logits/rejected": 10.282785415649414, | |
| "logps/chosen": -1.0583709478378296, | |
| "logps/rejected": -1.1253927946090698, | |
| "loss": 1.3887, | |
| "rewards/accuracies": 0.5390625, | |
| "rewards/chosen": -2.116741895675659, | |
| "rewards/margins": 0.1340436041355133, | |
| "rewards/rejected": -2.2507855892181396, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.026348291477974475, | |
| "grad_norm": 0.5300341844558716, | |
| "learning_rate": 2.5e-07, | |
| "logits/chosen": 10.530074119567871, | |
| "logits/rejected": 10.672085762023926, | |
| "logps/chosen": -1.1093882322311401, | |
| "logps/rejected": -1.2041704654693604, | |
| "loss": 1.3086, | |
| "rewards/accuracies": 0.6171875, | |
| "rewards/chosen": -2.2187764644622803, | |
| "rewards/margins": 0.18956437706947327, | |
| "rewards/rejected": -2.4083409309387207, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.03952243721696171, | |
| "grad_norm": 1.1035377979278564, | |
| "learning_rate": 3.75e-07, | |
| "logits/chosen": 10.240287780761719, | |
| "logits/rejected": 10.40180492401123, | |
| "logps/chosen": -1.1902998685836792, | |
| "logps/rejected": -1.311903715133667, | |
| "loss": 1.3245, | |
| "rewards/accuracies": 0.6484375, | |
| "rewards/chosen": -2.3805997371673584, | |
| "rewards/margins": 0.24320749938488007, | |
| "rewards/rejected": -2.623807430267334, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.05269658295594895, | |
| "grad_norm": 1.8411834239959717, | |
| "learning_rate": 5e-07, | |
| "logits/chosen": 10.358713150024414, | |
| "logits/rejected": 10.405998229980469, | |
| "logps/chosen": -1.0821034908294678, | |
| "logps/rejected": -1.1429979801177979, | |
| "loss": 1.3768, | |
| "rewards/accuracies": 0.578125, | |
| "rewards/chosen": -2.1642069816589355, | |
| "rewards/margins": 0.12178920209407806, | |
| "rewards/rejected": -2.2859959602355957, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.06587072869493618, | |
| "grad_norm": 0.5591928362846375, | |
| "learning_rate": 6.249999999999999e-07, | |
| "logits/chosen": 10.748247146606445, | |
| "logits/rejected": 10.783531188964844, | |
| "logps/chosen": -1.1161627769470215, | |
| "logps/rejected": -1.1165733337402344, | |
| "loss": 1.4573, | |
| "rewards/accuracies": 0.5859375, | |
| "rewards/chosen": -2.232325553894043, | |
| "rewards/margins": 0.0008210651576519012, | |
| "rewards/rejected": -2.2331466674804688, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.07904487443392343, | |
| "grad_norm": 1.0748578310012817, | |
| "learning_rate": 7.5e-07, | |
| "logits/chosen": 10.895101547241211, | |
| "logits/rejected": 10.935002326965332, | |
| "logps/chosen": -1.051721453666687, | |
| "logps/rejected": -1.1617565155029297, | |
| "loss": 1.3341, | |
| "rewards/accuracies": 0.609375, | |
| "rewards/chosen": -2.103442907333374, | |
| "rewards/margins": 0.22007031738758087, | |
| "rewards/rejected": -2.3235130310058594, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.09221902017291066, | |
| "grad_norm": 1.2474150657653809, | |
| "learning_rate": 8.75e-07, | |
| "logits/chosen": 10.060023307800293, | |
| "logits/rejected": 10.135562896728516, | |
| "logps/chosen": -1.0141205787658691, | |
| "logps/rejected": -1.1586568355560303, | |
| "loss": 1.3246, | |
| "rewards/accuracies": 0.6328125, | |
| "rewards/chosen": -2.0282411575317383, | |
| "rewards/margins": 0.28907278180122375, | |
| "rewards/rejected": -2.3173136711120605, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.1053931659118979, | |
| "grad_norm": 1.7001811265945435, | |
| "learning_rate": 1e-06, | |
| "logits/chosen": 10.837862014770508, | |
| "logits/rejected": 10.750317573547363, | |
| "logps/chosen": -1.0156888961791992, | |
| "logps/rejected": -1.1156163215637207, | |
| "loss": 1.314, | |
| "rewards/accuracies": 0.671875, | |
| "rewards/chosen": -2.0313777923583984, | |
| "rewards/margins": 0.19985482096672058, | |
| "rewards/rejected": -2.2312326431274414, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.11856731165088513, | |
| "grad_norm": 1.0888780355453491, | |
| "learning_rate": 9.994504457428556e-07, | |
| "logits/chosen": 10.434420585632324, | |
| "logits/rejected": 10.604791641235352, | |
| "logps/chosen": -1.1994280815124512, | |
| "logps/rejected": -1.1771385669708252, | |
| "loss": 1.4926, | |
| "rewards/accuracies": 0.5234375, | |
| "rewards/chosen": -2.3988561630249023, | |
| "rewards/margins": -0.04457877576351166, | |
| "rewards/rejected": -2.3542771339416504, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.13174145738987236, | |
| "grad_norm": 0.9426828622817993, | |
| "learning_rate": 9.97802991010949e-07, | |
| "logits/chosen": 10.330013275146484, | |
| "logits/rejected": 10.423635482788086, | |
| "logps/chosen": -1.1276183128356934, | |
| "logps/rejected": -1.1535909175872803, | |
| "loss": 1.4741, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -2.2552366256713867, | |
| "rewards/margins": 0.05194506794214249, | |
| "rewards/rejected": -2.3071818351745605, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.14491560312885962, | |
| "grad_norm": 1.174481987953186, | |
| "learning_rate": 9.950612572673255e-07, | |
| "logits/chosen": 10.557825088500977, | |
| "logits/rejected": 10.500927925109863, | |
| "logps/chosen": -1.2195600271224976, | |
| "logps/rejected": -1.3650047779083252, | |
| "loss": 1.3192, | |
| "rewards/accuracies": 0.640625, | |
| "rewards/chosen": -2.439120054244995, | |
| "rewards/margins": 0.2908894717693329, | |
| "rewards/rejected": -2.7300095558166504, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.15808974886784685, | |
| "grad_norm": 0.9938077330589294, | |
| "learning_rate": 9.912312714377879e-07, | |
| "logits/chosen": 10.221735000610352, | |
| "logits/rejected": 10.339948654174805, | |
| "logps/chosen": -1.072077989578247, | |
| "logps/rejected": -1.114700198173523, | |
| "loss": 1.3846, | |
| "rewards/accuracies": 0.6015625, | |
| "rewards/chosen": -2.144155979156494, | |
| "rewards/margins": 0.08524461090564728, | |
| "rewards/rejected": -2.229400396347046, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.17126389460683408, | |
| "grad_norm": 0.9858765006065369, | |
| "learning_rate": 9.863214526624063e-07, | |
| "logits/chosen": 10.290239334106445, | |
| "logits/rejected": 10.385334014892578, | |
| "logps/chosen": -1.1751586198806763, | |
| "logps/rejected": -1.3674362897872925, | |
| "loss": 1.2879, | |
| "rewards/accuracies": 0.6015625, | |
| "rewards/chosen": -2.3503172397613525, | |
| "rewards/margins": 0.38455531001091003, | |
| "rewards/rejected": -2.734872579574585, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.1844380403458213, | |
| "grad_norm": 1.6941860914230347, | |
| "learning_rate": 9.8034259378842e-07, | |
| "logits/chosen": 10.756081581115723, | |
| "logits/rejected": 10.947944641113281, | |
| "logps/chosen": -1.0922513008117676, | |
| "logps/rejected": -1.1827645301818848, | |
| "loss": 1.3465, | |
| "rewards/accuracies": 0.5859375, | |
| "rewards/chosen": -2.184502601623535, | |
| "rewards/margins": 0.18102645874023438, | |
| "rewards/rejected": -2.3655290603637695, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.19761218608480857, | |
| "grad_norm": 0.33364373445510864, | |
| "learning_rate": 9.73307837645217e-07, | |
| "logits/chosen": 10.299134254455566, | |
| "logits/rejected": 10.412195205688477, | |
| "logps/chosen": -1.1260679960250854, | |
| "logps/rejected": -1.2360444068908691, | |
| "loss": 1.3164, | |
| "rewards/accuracies": 0.609375, | |
| "rewards/chosen": -2.252135992050171, | |
| "rewards/margins": 0.219952791929245, | |
| "rewards/rejected": -2.4720888137817383, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.2107863318237958, | |
| "grad_norm": 0.7868668437004089, | |
| "learning_rate": 9.652326481535433e-07, | |
| "logits/chosen": 10.79407024383545, | |
| "logits/rejected": 10.986239433288574, | |
| "logps/chosen": -1.0521718263626099, | |
| "logps/rejected": -1.0945489406585693, | |
| "loss": 1.415, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -2.1043436527252197, | |
| "rewards/margins": 0.08475431054830551, | |
| "rewards/rejected": -2.1890978813171387, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.22396047756278303, | |
| "grad_norm": 0.6708123683929443, | |
| "learning_rate": 9.561347763324483e-07, | |
| "logits/chosen": 10.472009658813477, | |
| "logits/rejected": 10.457694053649902, | |
| "logps/chosen": -1.0896780490875244, | |
| "logps/rejected": -1.1595150232315063, | |
| "loss": 1.3736, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -2.179356098175049, | |
| "rewards/margins": 0.13967394828796387, | |
| "rewards/rejected": -2.3190300464630127, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.23713462330177026, | |
| "grad_norm": 0.9609116911888123, | |
| "learning_rate": 9.460342212786932e-07, | |
| "logits/chosen": 10.532232284545898, | |
| "logits/rejected": 10.626241683959961, | |
| "logps/chosen": -1.1494747400283813, | |
| "logps/rejected": -1.1549882888793945, | |
| "loss": 1.4623, | |
| "rewards/accuracies": 0.6171875, | |
| "rewards/chosen": -2.2989494800567627, | |
| "rewards/margins": 0.011027364060282707, | |
| "rewards/rejected": -2.309976577758789, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.2503087690407575, | |
| "grad_norm": 0.49065420031547546, | |
| "learning_rate": 9.349531862043951e-07, | |
| "logits/chosen": 10.552773475646973, | |
| "logits/rejected": 10.44497013092041, | |
| "logps/chosen": -1.123961329460144, | |
| "logps/rejected": -1.2145094871520996, | |
| "loss": 1.3068, | |
| "rewards/accuracies": 0.671875, | |
| "rewards/chosen": -2.247922658920288, | |
| "rewards/margins": 0.18109644949436188, | |
| "rewards/rejected": -2.429018974304199, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.2634829147797447, | |
| "grad_norm": 1.4662220478057861, | |
| "learning_rate": 9.229160296295487e-07, | |
| "logits/chosen": 10.48397445678711, | |
| "logits/rejected": 10.557746887207031, | |
| "logps/chosen": -1.0979560613632202, | |
| "logps/rejected": -1.1902183294296265, | |
| "loss": 1.355, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -2.1959121227264404, | |
| "rewards/margins": 0.1845243275165558, | |
| "rewards/rejected": -2.380436658859253, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.276657060518732, | |
| "grad_norm": 1.0163906812667847, | |
| "learning_rate": 9.099492118367122e-07, | |
| "logits/chosen": 10.452715873718262, | |
| "logits/rejected": 10.512972831726074, | |
| "logps/chosen": -1.0967323780059814, | |
| "logps/rejected": -1.2304078340530396, | |
| "loss": 1.2888, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -2.193464756011963, | |
| "rewards/margins": 0.26735079288482666, | |
| "rewards/rejected": -2.460815668106079, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.28983120625771924, | |
| "grad_norm": 3.908339738845825, | |
| "learning_rate": 8.960812367055646e-07, | |
| "logits/chosen": 10.526410102844238, | |
| "logits/rejected": 10.603084564208984, | |
| "logps/chosen": -1.224048376083374, | |
| "logps/rejected": -1.2536935806274414, | |
| "loss": 1.4348, | |
| "rewards/accuracies": 0.578125, | |
| "rewards/chosen": -2.448096752166748, | |
| "rewards/margins": 0.0592900887131691, | |
| "rewards/rejected": -2.507387161254883, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.3030053519967065, | |
| "grad_norm": 0.31968942284584045, | |
| "learning_rate": 8.813425890551909e-07, | |
| "logits/chosen": 10.193231582641602, | |
| "logits/rejected": 10.474884986877441, | |
| "logps/chosen": -1.1644542217254639, | |
| "logps/rejected": -1.1897979974746704, | |
| "loss": 1.4224, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -2.3289084434509277, | |
| "rewards/margins": 0.05068742483854294, | |
| "rewards/rejected": -2.379595994949341, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.3161794977356937, | |
| "grad_norm": 1.0304734706878662, | |
| "learning_rate": 8.657656676318345e-07, | |
| "logits/chosen": 10.281815528869629, | |
| "logits/rejected": 10.21823501586914, | |
| "logps/chosen": -1.0919454097747803, | |
| "logps/rejected": -1.210888385772705, | |
| "loss": 1.325, | |
| "rewards/accuracies": 0.5703125, | |
| "rewards/chosen": -2.1838908195495605, | |
| "rewards/margins": 0.23788578808307648, | |
| "rewards/rejected": -2.42177677154541, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.32935364347468093, | |
| "grad_norm": 2.0577356815338135, | |
| "learning_rate": 8.493847138894208e-07, | |
| "logits/chosen": 10.638153076171875, | |
| "logits/rejected": 10.773796081542969, | |
| "logps/chosen": -1.0134081840515137, | |
| "logps/rejected": -1.0593600273132324, | |
| "loss": 1.372, | |
| "rewards/accuracies": 0.5703125, | |
| "rewards/chosen": -2.0268163681030273, | |
| "rewards/margins": 0.09190365672111511, | |
| "rewards/rejected": -2.118720054626465, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.34252778921366817, | |
| "grad_norm": 0.6051430106163025, | |
| "learning_rate": 8.322357367194108e-07, | |
| "logits/chosen": 10.498373031616211, | |
| "logits/rejected": 10.679786682128906, | |
| "logps/chosen": -1.0168992280960083, | |
| "logps/rejected": -1.0872042179107666, | |
| "loss": 1.3499, | |
| "rewards/accuracies": 0.578125, | |
| "rewards/chosen": -2.0337984561920166, | |
| "rewards/margins": 0.14060987532138824, | |
| "rewards/rejected": -2.174408435821533, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.3557019349526554, | |
| "grad_norm": 1.769278645515442, | |
| "learning_rate": 8.143564332954425e-07, | |
| "logits/chosen": 10.561027526855469, | |
| "logits/rejected": 10.846573829650879, | |
| "logps/chosen": -1.114155650138855, | |
| "logps/rejected": -1.1514461040496826, | |
| "loss": 1.399, | |
| "rewards/accuracies": 0.5703125, | |
| "rewards/chosen": -2.22831130027771, | |
| "rewards/margins": 0.07458095252513885, | |
| "rewards/rejected": -2.3028922080993652, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.3688760806916426, | |
| "grad_norm": 0.6603056192398071, | |
| "learning_rate": 7.957861062067612e-07, | |
| "logits/chosen": 10.477982521057129, | |
| "logits/rejected": 10.683720588684082, | |
| "logps/chosen": -1.0256245136260986, | |
| "logps/rejected": -1.1278630495071411, | |
| "loss": 1.3446, | |
| "rewards/accuracies": 0.578125, | |
| "rewards/chosen": -2.0512490272521973, | |
| "rewards/margins": 0.20447733998298645, | |
| "rewards/rejected": -2.2557260990142822, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.3820502264306299, | |
| "grad_norm": 1.5685954093933105, | |
| "learning_rate": 7.765655770625996e-07, | |
| "logits/chosen": 10.337364196777344, | |
| "logits/rejected": 10.332595825195312, | |
| "logps/chosen": -1.043882966041565, | |
| "logps/rejected": -1.0915210247039795, | |
| "loss": 1.3965, | |
| "rewards/accuracies": 0.6796875, | |
| "rewards/chosen": -2.08776593208313, | |
| "rewards/margins": 0.09527605772018433, | |
| "rewards/rejected": -2.183042049407959, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.39522437216961714, | |
| "grad_norm": 0.5102595686912537, | |
| "learning_rate": 7.567370967574209e-07, | |
| "logits/chosen": 10.61121654510498, | |
| "logits/rejected": 10.84305191040039, | |
| "logps/chosen": -1.0729877948760986, | |
| "logps/rejected": -1.1098031997680664, | |
| "loss": 1.4371, | |
| "rewards/accuracies": 0.578125, | |
| "rewards/chosen": -2.1459755897521973, | |
| "rewards/margins": 0.07363072782754898, | |
| "rewards/rejected": -2.219606399536133, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.4083985179086044, | |
| "grad_norm": 0.8290932774543762, | |
| "learning_rate": 7.363442525942826e-07, | |
| "logits/chosen": 10.494096755981445, | |
| "logits/rejected": 10.498456954956055, | |
| "logps/chosen": -0.9797660708427429, | |
| "logps/rejected": -1.0840959548950195, | |
| "loss": 1.3044, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -1.9595321416854858, | |
| "rewards/margins": 0.2086598128080368, | |
| "rewards/rejected": -2.168191909790039, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.4215726636475916, | |
| "grad_norm": 0.36728209257125854, | |
| "learning_rate": 7.154318724704851e-07, | |
| "logits/chosen": 10.502325057983398, | |
| "logits/rejected": 10.551267623901367, | |
| "logps/chosen": -1.040996789932251, | |
| "logps/rejected": -1.0938708782196045, | |
| "loss": 1.3946, | |
| "rewards/accuracies": 0.5703125, | |
| "rewards/chosen": -2.081993579864502, | |
| "rewards/margins": 0.10574813187122345, | |
| "rewards/rejected": -2.187741756439209, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.43474680938657884, | |
| "grad_norm": 0.5727468132972717, | |
| "learning_rate": 6.940459263361248e-07, | |
| "logits/chosen": 10.425085067749023, | |
| "logits/rejected": 10.45986270904541, | |
| "logps/chosen": -1.0259983539581299, | |
| "logps/rejected": -1.064608097076416, | |
| "loss": 1.4179, | |
| "rewards/accuracies": 0.6015625, | |
| "rewards/chosen": -2.0519967079162598, | |
| "rewards/margins": 0.07721954584121704, | |
| "rewards/rejected": -2.129216194152832, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.44792095512556607, | |
| "grad_norm": 0.8594540953636169, | |
| "learning_rate": 6.722334251421664e-07, | |
| "logits/chosen": 10.195260047912598, | |
| "logits/rejected": 10.263075828552246, | |
| "logps/chosen": -1.0975958108901978, | |
| "logps/rejected": -1.1462163925170898, | |
| "loss": 1.4471, | |
| "rewards/accuracies": 0.5703125, | |
| "rewards/chosen": -2.1951916217803955, | |
| "rewards/margins": 0.09724146127700806, | |
| "rewards/rejected": -2.2924327850341797, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.4610951008645533, | |
| "grad_norm": 2.0125956535339355, | |
| "learning_rate": 6.500423175001703e-07, | |
| "logits/chosen": 10.623847007751465, | |
| "logits/rejected": 10.824085235595703, | |
| "logps/chosen": -1.1299512386322021, | |
| "logps/rejected": -1.1615726947784424, | |
| "loss": 1.4721, | |
| "rewards/accuracies": 0.578125, | |
| "rewards/chosen": -2.2599024772644043, | |
| "rewards/margins": 0.06324289739131927, | |
| "rewards/rejected": -2.3231453895568848, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.47426924660354053, | |
| "grad_norm": 0.9820675253868103, | |
| "learning_rate": 6.275213842808382e-07, | |
| "logits/chosen": 10.475167274475098, | |
| "logits/rejected": 10.816570281982422, | |
| "logps/chosen": -1.0255820751190186, | |
| "logps/rejected": -1.0681825876235962, | |
| "loss": 1.3712, | |
| "rewards/accuracies": 0.578125, | |
| "rewards/chosen": -2.051164150238037, | |
| "rewards/margins": 0.08520102500915527, | |
| "rewards/rejected": -2.1363651752471924, | |
| "step": 36 | |
| } | |
| ], | |
| "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 | |
| } | |