| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9980806142034548, | |
| "eval_steps": 1000000, | |
| "global_step": 390, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0025591810620601407, | |
| "grad_norm": 709.6319834867974, | |
| "learning_rate": 5.128205128205128e-09, | |
| "logits/chosen": -2.5583817958831787, | |
| "logits/rejected": -2.4487552642822266, | |
| "logps/chosen": -258.1644592285156, | |
| "logps/rejected": -216.25729370117188, | |
| "loss": 0.6931, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.025591810620601407, | |
| "grad_norm": 670.9725237886468, | |
| "learning_rate": 5.128205128205127e-08, | |
| "logits/chosen": -2.606004238128662, | |
| "logits/rejected": -2.553126573562622, | |
| "logps/chosen": -267.5910339355469, | |
| "logps/rejected": -217.67140197753906, | |
| "loss": 0.6976, | |
| "rewards/accuracies": 0.4513888955116272, | |
| "rewards/chosen": -0.013270225375890732, | |
| "rewards/margins": 0.0032119054812937975, | |
| "rewards/rejected": -0.01648213155567646, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.05118362124120281, | |
| "grad_norm": 590.5735450897452, | |
| "learning_rate": 1.0256410256410255e-07, | |
| "logits/chosen": -2.6245453357696533, | |
| "logits/rejected": -2.562206745147705, | |
| "logps/chosen": -261.142333984375, | |
| "logps/rejected": -207.0738067626953, | |
| "loss": 0.6676, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": 0.08092932403087616, | |
| "rewards/margins": 0.09077399969100952, | |
| "rewards/rejected": -0.009844672866165638, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.07677543186180422, | |
| "grad_norm": 644.2505273786248, | |
| "learning_rate": 1.5384615384615385e-07, | |
| "logits/chosen": -2.6289420127868652, | |
| "logits/rejected": -2.5566790103912354, | |
| "logps/chosen": -253.9628143310547, | |
| "logps/rejected": -198.890380859375, | |
| "loss": 0.5156, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": 0.5922147035598755, | |
| "rewards/margins": 0.6031575798988342, | |
| "rewards/rejected": -0.010942881926894188, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.10236724248240563, | |
| "grad_norm": 428.1589759637874, | |
| "learning_rate": 1.999959945379852e-07, | |
| "logits/chosen": -2.637456178665161, | |
| "logits/rejected": -2.5609130859375, | |
| "logps/chosen": -247.48666381835938, | |
| "logps/rejected": -193.59280395507812, | |
| "loss": 0.3866, | |
| "rewards/accuracies": 0.8062499761581421, | |
| "rewards/chosen": 1.4313427209854126, | |
| "rewards/margins": 1.3846019506454468, | |
| "rewards/rejected": 0.04674070328474045, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.12795905310300704, | |
| "grad_norm": 366.0325785868742, | |
| "learning_rate": 1.9951572723309917e-07, | |
| "logits/chosen": -2.684983730316162, | |
| "logits/rejected": -2.618076801300049, | |
| "logps/chosen": -262.3794250488281, | |
| "logps/rejected": -204.0382843017578, | |
| "loss": 0.3488, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": 2.9288697242736816, | |
| "rewards/margins": 2.3735203742980957, | |
| "rewards/rejected": 0.5553494095802307, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.15355086372360843, | |
| "grad_norm": 375.0458504104559, | |
| "learning_rate": 1.9823877374156647e-07, | |
| "logits/chosen": -2.669403553009033, | |
| "logits/rejected": -2.6018548011779785, | |
| "logps/chosen": -262.41534423828125, | |
| "logps/rejected": -206.5387420654297, | |
| "loss": 0.311, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": 3.2061855792999268, | |
| "rewards/margins": 2.446310520172119, | |
| "rewards/rejected": 0.7598745822906494, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.17914267434420986, | |
| "grad_norm": 371.9136724850062, | |
| "learning_rate": 1.9617535688178958e-07, | |
| "logits/chosen": -2.6535415649414062, | |
| "logits/rejected": -2.590367078781128, | |
| "logps/chosen": -290.74871826171875, | |
| "logps/rejected": -224.1929168701172, | |
| "loss": 0.3051, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": 3.5809733867645264, | |
| "rewards/margins": 3.0821049213409424, | |
| "rewards/rejected": 0.4988683760166168, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.20473448496481125, | |
| "grad_norm": 381.95411975681657, | |
| "learning_rate": 1.9334199560765839e-07, | |
| "logits/chosen": -2.6085634231567383, | |
| "logits/rejected": -2.5423429012298584, | |
| "logps/chosen": -260.9337463378906, | |
| "logps/rejected": -203.12725830078125, | |
| "loss": 0.2569, | |
| "rewards/accuracies": 0.8687499761581421, | |
| "rewards/chosen": 3.3122222423553467, | |
| "rewards/margins": 3.3377063274383545, | |
| "rewards/rejected": -0.025484371930360794, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.23032629558541268, | |
| "grad_norm": 331.9831865016541, | |
| "learning_rate": 1.897613727639014e-07, | |
| "logits/chosen": -2.610903739929199, | |
| "logits/rejected": -2.556959867477417, | |
| "logps/chosen": -256.61138916015625, | |
| "logps/rejected": -215.09701538085938, | |
| "loss": 0.2658, | |
| "rewards/accuracies": 0.8687499761581421, | |
| "rewards/chosen": 2.5669326782226562, | |
| "rewards/margins": 2.867710590362549, | |
| "rewards/rejected": -0.3007778227329254, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.2559181062060141, | |
| "grad_norm": 634.1345493995642, | |
| "learning_rate": 1.8546215349560202e-07, | |
| "logits/chosen": -2.6384997367858887, | |
| "logits/rejected": -2.563520908355713, | |
| "logps/chosen": -236.5784912109375, | |
| "logps/rejected": -199.6844940185547, | |
| "loss": 0.3117, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": 2.7952113151550293, | |
| "rewards/margins": 3.302074909210205, | |
| "rewards/rejected": -0.5068637132644653, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.28150991682661547, | |
| "grad_norm": 319.85079714910466, | |
| "learning_rate": 1.8047875576562553e-07, | |
| "logits/chosen": -2.630192518234253, | |
| "logits/rejected": -2.549891710281372, | |
| "logps/chosen": -258.26971435546875, | |
| "logps/rejected": -208.4285430908203, | |
| "loss": 0.282, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": 3.702361583709717, | |
| "rewards/margins": 3.704643726348877, | |
| "rewards/rejected": -0.002281466033309698, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.30710172744721687, | |
| "grad_norm": 376.18943633320504, | |
| "learning_rate": 1.748510748171101e-07, | |
| "logits/chosen": -2.6130287647247314, | |
| "logits/rejected": -2.543454170227051, | |
| "logps/chosen": -262.194580078125, | |
| "logps/rejected": -206.72488403320312, | |
| "loss": 0.2802, | |
| "rewards/accuracies": 0.8812500238418579, | |
| "rewards/chosen": 3.633894443511963, | |
| "rewards/margins": 3.8131375312805176, | |
| "rewards/rejected": -0.17924347519874573, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.3326935380678183, | |
| "grad_norm": 258.08224292776333, | |
| "learning_rate": 1.6862416378687336e-07, | |
| "logits/chosen": -2.587902069091797, | |
| "logits/rejected": -2.5218377113342285, | |
| "logps/chosen": -250.2627410888672, | |
| "logps/rejected": -193.98731994628906, | |
| "loss": 0.2644, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": 3.5202510356903076, | |
| "rewards/margins": 3.647292375564575, | |
| "rewards/rejected": -0.12704159319400787, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.3582853486884197, | |
| "grad_norm": 312.9816016514414, | |
| "learning_rate": 1.6184787302662547e-07, | |
| "logits/chosen": -2.6076598167419434, | |
| "logits/rejected": -2.539602279663086, | |
| "logps/chosen": -267.7660217285156, | |
| "logps/rejected": -213.330078125, | |
| "loss": 0.3055, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": 3.705143690109253, | |
| "rewards/margins": 3.5916450023651123, | |
| "rewards/rejected": 0.11349865049123764, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.3838771593090211, | |
| "grad_norm": 298.63492430036257, | |
| "learning_rate": 1.5457645101945046e-07, | |
| "logits/chosen": -2.588223457336426, | |
| "logits/rejected": -2.521327495574951, | |
| "logps/chosen": -262.39776611328125, | |
| "logps/rejected": -210.2900848388672, | |
| "loss": 0.2675, | |
| "rewards/accuracies": 0.887499988079071, | |
| "rewards/chosen": 4.096975803375244, | |
| "rewards/margins": 4.044339656829834, | |
| "rewards/rejected": 0.05263558775186539, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.4094689699296225, | |
| "grad_norm": 277.3876418755778, | |
| "learning_rate": 1.4686811008647035e-07, | |
| "logits/chosen": -2.5860092639923096, | |
| "logits/rejected": -2.5171661376953125, | |
| "logps/chosen": -241.29501342773438, | |
| "logps/rejected": -192.94955444335938, | |
| "loss": 0.2544, | |
| "rewards/accuracies": 0.887499988079071, | |
| "rewards/chosen": 3.8214523792266846, | |
| "rewards/margins": 3.785486936569214, | |
| "rewards/rejected": 0.0359656922519207, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.4350607805502239, | |
| "grad_norm": 339.61682629601955, | |
| "learning_rate": 1.387845603604855e-07, | |
| "logits/chosen": -2.5826542377471924, | |
| "logits/rejected": -2.51751446723938, | |
| "logps/chosen": -262.99298095703125, | |
| "logps/rejected": -206.5457000732422, | |
| "loss": 0.2334, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": 3.8493824005126953, | |
| "rewards/margins": 3.933476209640503, | |
| "rewards/rejected": -0.08409398794174194, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.46065259117082535, | |
| "grad_norm": 293.5783730044579, | |
| "learning_rate": 1.3039051575742468e-07, | |
| "logits/chosen": -2.6330299377441406, | |
| "logits/rejected": -2.547426700592041, | |
| "logps/chosen": -235.4442596435547, | |
| "logps/rejected": -183.63687133789062, | |
| "loss": 0.2871, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": 3.294564723968506, | |
| "rewards/margins": 3.4877796173095703, | |
| "rewards/rejected": -0.193215012550354, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.48624440179142675, | |
| "grad_norm": 298.0737945063753, | |
| "learning_rate": 1.2175317590061675e-07, | |
| "logits/chosen": -2.5894675254821777, | |
| "logits/rejected": -2.51374888420105, | |
| "logps/chosen": -257.42132568359375, | |
| "logps/rejected": -199.62106323242188, | |
| "loss": 0.239, | |
| "rewards/accuracies": 0.887499988079071, | |
| "rewards/chosen": 3.699636936187744, | |
| "rewards/margins": 3.845207929611206, | |
| "rewards/rejected": -0.1455712914466858, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.5118362124120281, | |
| "grad_norm": 326.3771289852159, | |
| "learning_rate": 1.1294168814540553e-07, | |
| "logits/chosen": -2.625087261199951, | |
| "logits/rejected": -2.5350747108459473, | |
| "logps/chosen": -271.9561462402344, | |
| "logps/rejected": -214.1556854248047, | |
| "loss": 0.2317, | |
| "rewards/accuracies": 0.90625, | |
| "rewards/chosen": 4.283473014831543, | |
| "rewards/margins": 4.552158832550049, | |
| "rewards/rejected": -0.2686860263347626, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.5374280230326296, | |
| "grad_norm": 510.1343443372893, | |
| "learning_rate": 1.0402659401094151e-07, | |
| "logits/chosen": -2.602785587310791, | |
| "logits/rejected": -2.5288333892822266, | |
| "logps/chosen": -264.21099853515625, | |
| "logps/rejected": -207.3443145751953, | |
| "loss": 0.3153, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": 3.421428680419922, | |
| "rewards/margins": 3.9131996631622314, | |
| "rewards/rejected": -0.4917708933353424, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.5630198336532309, | |
| "grad_norm": 362.61099626331173, | |
| "learning_rate": 9.507926445081218e-08, | |
| "logits/chosen": -2.607333183288574, | |
| "logits/rejected": -2.5410573482513428, | |
| "logps/chosen": -255.5386199951172, | |
| "logps/rejected": -210.58261108398438, | |
| "loss": 0.2657, | |
| "rewards/accuracies": 0.893750011920929, | |
| "rewards/chosen": 3.069124698638916, | |
| "rewards/margins": 3.4770140647888184, | |
| "rewards/rejected": -0.4078896641731262, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.5886116442738324, | |
| "grad_norm": 270.86067735708986, | |
| "learning_rate": 8.61713284835267e-08, | |
| "logits/chosen": -2.6318671703338623, | |
| "logits/rejected": -2.5431408882141113, | |
| "logps/chosen": -265.3260803222656, | |
| "logps/rejected": -212.6390380859375, | |
| "loss": 0.249, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": 4.038520812988281, | |
| "rewards/margins": 3.9009087085723877, | |
| "rewards/rejected": 0.1376120150089264, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.6142034548944337, | |
| "grad_norm": 496.26361322510223, | |
| "learning_rate": 7.73740997570278e-08, | |
| "logits/chosen": -2.6172757148742676, | |
| "logits/rejected": -2.566117763519287, | |
| "logps/chosen": -259.39569091796875, | |
| "logps/rejected": -214.13717651367188, | |
| "loss": 0.2362, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": 4.167778968811035, | |
| "rewards/margins": 4.450715065002441, | |
| "rewards/rejected": -0.28293582797050476, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.6397952655150352, | |
| "grad_norm": 385.39306248287744, | |
| "learning_rate": 6.875800563794424e-08, | |
| "logits/chosen": -2.6301958560943604, | |
| "logits/rejected": -2.5562362670898438, | |
| "logps/chosen": -264.5022277832031, | |
| "logps/rejected": -217.06472778320312, | |
| "loss": 0.2541, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": 3.638233184814453, | |
| "rewards/margins": 3.9798378944396973, | |
| "rewards/rejected": -0.34160494804382324, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.6653870761356366, | |
| "grad_norm": 246.0647695237683, | |
| "learning_rate": 6.039202339608431e-08, | |
| "logits/chosen": -2.6222901344299316, | |
| "logits/rejected": -2.5577120780944824, | |
| "logps/chosen": -258.84637451171875, | |
| "logps/rejected": -203.7962188720703, | |
| "loss": 0.2317, | |
| "rewards/accuracies": 0.893750011920929, | |
| "rewards/chosen": 3.5067996978759766, | |
| "rewards/margins": 3.801957607269287, | |
| "rewards/rejected": -0.295158326625824, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.690978886756238, | |
| "grad_norm": 382.6551210350461, | |
| "learning_rate": 5.2343127997869205e-08, | |
| "logits/chosen": -2.602179765701294, | |
| "logits/rejected": -2.5085558891296387, | |
| "logps/chosen": -270.6271057128906, | |
| "logps/rejected": -222.81655883789062, | |
| "loss": 0.2926, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": 2.9784107208251953, | |
| "rewards/margins": 4.063758850097656, | |
| "rewards/rejected": -1.085348129272461, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.7165706973768394, | |
| "grad_norm": 291.15530026271324, | |
| "learning_rate": 4.4675755929468636e-08, | |
| "logits/chosen": -2.627225160598755, | |
| "logits/rejected": -2.565744638442993, | |
| "logps/chosen": -266.7916259765625, | |
| "logps/rejected": -214.30709838867188, | |
| "loss": 0.2375, | |
| "rewards/accuracies": 0.90625, | |
| "rewards/chosen": 3.4573512077331543, | |
| "rewards/margins": 4.138546943664551, | |
| "rewards/rejected": -0.6811951398849487, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.7421625079974408, | |
| "grad_norm": 302.9770657235692, | |
| "learning_rate": 3.745128934207224e-08, | |
| "logits/chosen": -2.590620279312134, | |
| "logits/rejected": -2.5388922691345215, | |
| "logps/chosen": -265.51971435546875, | |
| "logps/rejected": -212.6427764892578, | |
| "loss": 0.2286, | |
| "rewards/accuracies": 0.893750011920929, | |
| "rewards/chosen": 3.2131683826446533, | |
| "rewards/margins": 3.975499391555786, | |
| "rewards/rejected": -0.7623313665390015, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.7677543186180422, | |
| "grad_norm": 339.1306333767976, | |
| "learning_rate": 3.0727564649040063e-08, | |
| "logits/chosen": -2.6111063957214355, | |
| "logits/rejected": -2.5518836975097656, | |
| "logps/chosen": -271.9001159667969, | |
| "logps/rejected": -216.2710723876953, | |
| "loss": 0.2446, | |
| "rewards/accuracies": 0.9125000238418579, | |
| "rewards/chosen": 3.3372998237609863, | |
| "rewards/margins": 3.9620718955993652, | |
| "rewards/rejected": -0.6247718930244446, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.7933461292386437, | |
| "grad_norm": 235.34557136252968, | |
| "learning_rate": 2.4558409508920985e-08, | |
| "logits/chosen": -2.5742857456207275, | |
| "logits/rejected": -2.499851703643799, | |
| "logps/chosen": -275.28167724609375, | |
| "logps/rejected": -211.0914764404297, | |
| "loss": 0.2598, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": 3.4961440563201904, | |
| "rewards/margins": 4.247152805328369, | |
| "rewards/rejected": -0.7510083913803101, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.818937939859245, | |
| "grad_norm": 368.2414347720992, | |
| "learning_rate": 1.899321190108335e-08, | |
| "logits/chosen": -2.59656023979187, | |
| "logits/rejected": -2.545151472091675, | |
| "logps/chosen": -258.7973327636719, | |
| "logps/rejected": -200.56219482421875, | |
| "loss": 0.2329, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": 3.0572078227996826, | |
| "rewards/margins": 3.9228408336639404, | |
| "rewards/rejected": -0.8656327128410339, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.8445297504798465, | |
| "grad_norm": 412.2800478150093, | |
| "learning_rate": 1.4076524743778317e-08, | |
| "logits/chosen": -2.6195101737976074, | |
| "logits/rejected": -2.55584716796875, | |
| "logps/chosen": -262.2277526855469, | |
| "logps/rejected": -213.35403442382812, | |
| "loss": 0.231, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": 3.7354748249053955, | |
| "rewards/margins": 4.0478105545043945, | |
| "rewards/rejected": -0.31233564019203186, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.8701215611004478, | |
| "grad_norm": 291.1936087568274, | |
| "learning_rate": 9.847709219918398e-09, | |
| "logits/chosen": -2.6071531772613525, | |
| "logits/rejected": -2.5460381507873535, | |
| "logps/chosen": -257.75164794921875, | |
| "logps/rejected": -206.60986328125, | |
| "loss": 0.221, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": 3.5297904014587402, | |
| "rewards/margins": 4.399306297302246, | |
| "rewards/rejected": -0.8695155382156372, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.8957133717210493, | |
| "grad_norm": 301.99507828681374, | |
| "learning_rate": 6.340619665972846e-09, | |
| "logits/chosen": -2.617502212524414, | |
| "logits/rejected": -2.5671114921569824, | |
| "logps/chosen": -267.4635009765625, | |
| "logps/rejected": -222.5786895751953, | |
| "loss": 0.2579, | |
| "rewards/accuracies": 0.8687499761581421, | |
| "rewards/chosen": 3.393681287765503, | |
| "rewards/margins": 3.9400668144226074, | |
| "rewards/rejected": -0.546385645866394, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.9213051823416507, | |
| "grad_norm": 348.99526280716225, | |
| "learning_rate": 3.583332546643769e-09, | |
| "logits/chosen": -2.6180646419525146, | |
| "logits/rejected": -2.5445971488952637, | |
| "logps/chosen": -272.35406494140625, | |
| "logps/rejected": -214.8975067138672, | |
| "loss": 0.2825, | |
| "rewards/accuracies": 0.893750011920929, | |
| "rewards/chosen": 3.431410551071167, | |
| "rewards/margins": 4.0157270431518555, | |
| "rewards/rejected": -0.584316074848175, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.946896992962252, | |
| "grad_norm": 251.0460885716357, | |
| "learning_rate": 1.5979216850509847e-09, | |
| "logits/chosen": -2.5794007778167725, | |
| "logits/rejected": -2.5391972064971924, | |
| "logps/chosen": -271.82513427734375, | |
| "logps/rejected": -225.1968231201172, | |
| "loss": 0.2529, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": 3.722885847091675, | |
| "rewards/margins": 4.0348358154296875, | |
| "rewards/rejected": -0.3119499981403351, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.9724888035828535, | |
| "grad_norm": 201.79043052952954, | |
| "learning_rate": 4.002815478505006e-10, | |
| "logits/chosen": -2.6285715103149414, | |
| "logits/rejected": -2.585494041442871, | |
| "logps/chosen": -261.1199951171875, | |
| "logps/rejected": -226.61105346679688, | |
| "loss": 0.255, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": 3.6326937675476074, | |
| "rewards/margins": 4.036477088928223, | |
| "rewards/rejected": -0.40378379821777344, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.9980806142034548, | |
| "grad_norm": 458.24518941114127, | |
| "learning_rate": 0.0, | |
| "logits/chosen": -2.641213893890381, | |
| "logits/rejected": -2.575150966644287, | |
| "logps/chosen": -248.17904663085938, | |
| "logps/rejected": -206.1395721435547, | |
| "loss": 0.2365, | |
| "rewards/accuracies": 0.918749988079071, | |
| "rewards/chosen": 3.4077067375183105, | |
| "rewards/margins": 4.0728960037231445, | |
| "rewards/rejected": -0.6651893854141235, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.9980806142034548, | |
| "step": 390, | |
| "total_flos": 0.0, | |
| "train_loss": 0.2947830175742125, | |
| "train_runtime": 6391.2384, | |
| "train_samples_per_second": 7.823, | |
| "train_steps_per_second": 0.061 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 390, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "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": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |