| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9984827495217362, | |
| "eval_steps": 400, | |
| "global_step": 473, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.002110957187149548, | |
| "grad_norm": 4.06041781261902, | |
| "learning_rate": 1.0416666666666666e-08, | |
| "logits/chosen": -0.9878771901130676, | |
| "logits/rejected": -0.7230668663978577, | |
| "logps/chosen": -251.34963989257812, | |
| "logps/rejected": -287.15838623046875, | |
| "loss": 0.6931, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.01055478593574774, | |
| "grad_norm": 4.105776204209711, | |
| "learning_rate": 5.208333333333333e-08, | |
| "logits/chosen": -0.9975427985191345, | |
| "logits/rejected": -0.6888133883476257, | |
| "logps/chosen": -272.86993408203125, | |
| "logps/rejected": -286.1126708984375, | |
| "loss": 0.6932, | |
| "rewards/accuracies": 0.359375, | |
| "rewards/chosen": -0.00011302110215183347, | |
| "rewards/margins": -0.00016400158347096294, | |
| "rewards/rejected": 5.0980423111468554e-05, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.02110957187149548, | |
| "grad_norm": 4.582740122598074, | |
| "learning_rate": 1.0416666666666667e-07, | |
| "logits/chosen": -1.0306963920593262, | |
| "logits/rejected": -0.7614760398864746, | |
| "logps/chosen": -274.9240417480469, | |
| "logps/rejected": -293.0102233886719, | |
| "loss": 0.6933, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.0007906880346126854, | |
| "rewards/margins": 0.00033544833422638476, | |
| "rewards/rejected": 0.00045523978769779205, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03166435780724322, | |
| "grad_norm": 4.2774752538157, | |
| "learning_rate": 1.5624999999999999e-07, | |
| "logits/chosen": -0.9785528182983398, | |
| "logits/rejected": -0.6860870122909546, | |
| "logps/chosen": -253.0674285888672, | |
| "logps/rejected": -277.63006591796875, | |
| "loss": 0.693, | |
| "rewards/accuracies": 0.44999998807907104, | |
| "rewards/chosen": 0.00013054809824097902, | |
| "rewards/margins": -4.017539322376251e-05, | |
| "rewards/rejected": 0.0001707235351204872, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.04221914374299096, | |
| "grad_norm": 4.154426876518502, | |
| "learning_rate": 2.0833333333333333e-07, | |
| "logits/chosen": -0.8861902356147766, | |
| "logits/rejected": -0.7466350793838501, | |
| "logps/chosen": -303.89990234375, | |
| "logps/rejected": -317.44354248046875, | |
| "loss": 0.6929, | |
| "rewards/accuracies": 0.48750001192092896, | |
| "rewards/chosen": -0.0017729544779285789, | |
| "rewards/margins": 0.0007360944291576743, | |
| "rewards/rejected": -0.0025090486742556095, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0527739296787387, | |
| "grad_norm": 4.391172268221911, | |
| "learning_rate": 2.604166666666667e-07, | |
| "logits/chosen": -0.9347459077835083, | |
| "logits/rejected": -0.7482324242591858, | |
| "logps/chosen": -257.277099609375, | |
| "logps/rejected": -276.6146240234375, | |
| "loss": 0.6925, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.0021017056424170732, | |
| "rewards/margins": 0.001351111801341176, | |
| "rewards/rejected": -0.003452816978096962, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.06332871561448644, | |
| "grad_norm": 4.197922988205761, | |
| "learning_rate": 3.1249999999999997e-07, | |
| "logits/chosen": -0.9495643377304077, | |
| "logits/rejected": -0.722493052482605, | |
| "logps/chosen": -271.043212890625, | |
| "logps/rejected": -304.54779052734375, | |
| "loss": 0.6909, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.007976134307682514, | |
| "rewards/margins": 0.006574218161404133, | |
| "rewards/rejected": -0.014550352469086647, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07388350155023418, | |
| "grad_norm": 4.146928571575294, | |
| "learning_rate": 3.645833333333333e-07, | |
| "logits/chosen": -0.9740797281265259, | |
| "logits/rejected": -0.6511734127998352, | |
| "logps/chosen": -269.66558837890625, | |
| "logps/rejected": -282.5665588378906, | |
| "loss": 0.6901, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.014239413663744926, | |
| "rewards/margins": 0.006516980938613415, | |
| "rewards/rejected": -0.020756395533680916, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.08443828748598192, | |
| "grad_norm": 4.434625304551974, | |
| "learning_rate": 4.1666666666666667e-07, | |
| "logits/chosen": -0.9476411938667297, | |
| "logits/rejected": -0.8075205087661743, | |
| "logps/chosen": -254.89315795898438, | |
| "logps/rejected": -269.08843994140625, | |
| "loss": 0.6872, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.025963688269257545, | |
| "rewards/margins": 0.010111861862242222, | |
| "rewards/rejected": -0.03607555106282234, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.09499307342172966, | |
| "grad_norm": 4.334333725960549, | |
| "learning_rate": 4.6874999999999996e-07, | |
| "logits/chosen": -0.9668118357658386, | |
| "logits/rejected": -0.7818160057067871, | |
| "logps/chosen": -262.55804443359375, | |
| "logps/rejected": -279.33416748046875, | |
| "loss": 0.6841, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.03984779864549637, | |
| "rewards/margins": 0.01882680132985115, | |
| "rewards/rejected": -0.05867459625005722, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.1055478593574774, | |
| "grad_norm": 4.672589574927753, | |
| "learning_rate": 4.999726797933858e-07, | |
| "logits/chosen": -1.1188008785247803, | |
| "logits/rejected": -0.7495776414871216, | |
| "logps/chosen": -274.8739318847656, | |
| "logps/rejected": -298.7273254394531, | |
| "loss": 0.6736, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": -0.0697299912571907, | |
| "rewards/margins": 0.04492691531777382, | |
| "rewards/rejected": -0.11465690284967422, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.11610264529322514, | |
| "grad_norm": 6.181545467894853, | |
| "learning_rate": 4.99665396039775e-07, | |
| "logits/chosen": -1.1051629781723022, | |
| "logits/rejected": -0.9252668619155884, | |
| "logps/chosen": -270.87841796875, | |
| "logps/rejected": -289.1105651855469, | |
| "loss": 0.6576, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -0.14251390099525452, | |
| "rewards/margins": 0.10730250179767609, | |
| "rewards/rejected": -0.2498163878917694, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.12665743122897288, | |
| "grad_norm": 7.558458036407823, | |
| "learning_rate": 4.99017099386437e-07, | |
| "logits/chosen": -1.3753994703292847, | |
| "logits/rejected": -1.1374595165252686, | |
| "logps/chosen": -306.73138427734375, | |
| "logps/rejected": -341.1016845703125, | |
| "loss": 0.6281, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.37672942876815796, | |
| "rewards/margins": 0.20342817902565002, | |
| "rewards/rejected": -0.5801576375961304, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.13721221716472062, | |
| "grad_norm": 66.9078697860869, | |
| "learning_rate": 4.980286753286194e-07, | |
| "logits/chosen": -1.5234780311584473, | |
| "logits/rejected": -1.2897632122039795, | |
| "logps/chosen": -359.58074951171875, | |
| "logps/rejected": -414.0733947753906, | |
| "loss": 0.6211, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -1.063786506652832, | |
| "rewards/margins": 0.41773301362991333, | |
| "rewards/rejected": -1.4815195798873901, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.14776700310046836, | |
| "grad_norm": 12.864645326069963, | |
| "learning_rate": 4.967014739346915e-07, | |
| "logits/chosen": -1.4892756938934326, | |
| "logits/rejected": -1.3145514726638794, | |
| "logps/chosen": -410.1991271972656, | |
| "logps/rejected": -467.57623291015625, | |
| "loss": 0.6578, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -1.4497530460357666, | |
| "rewards/margins": 0.4113241136074066, | |
| "rewards/rejected": -1.8610769510269165, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.1583217890362161, | |
| "grad_norm": 12.52823747211519, | |
| "learning_rate": 4.950373080021136e-07, | |
| "logits/chosen": -1.2871811389923096, | |
| "logits/rejected": -1.2107694149017334, | |
| "logps/chosen": -333.7770080566406, | |
| "logps/rejected": -356.77728271484375, | |
| "loss": 0.5936, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.579529881477356, | |
| "rewards/margins": 0.2453722506761551, | |
| "rewards/rejected": -0.8249020576477051, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.16887657497196384, | |
| "grad_norm": 17.45417402457315, | |
| "learning_rate": 4.930384505813737e-07, | |
| "logits/chosen": -1.4544992446899414, | |
| "logits/rejected": -1.3903002738952637, | |
| "logps/chosen": -392.7482604980469, | |
| "logps/rejected": -472.0469665527344, | |
| "loss": 0.5588, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -1.0880345106124878, | |
| "rewards/margins": 0.6102110147476196, | |
| "rewards/rejected": -1.6982454061508179, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.17943136090771158, | |
| "grad_norm": 36.09231440231977, | |
| "learning_rate": 4.907076318712738e-07, | |
| "logits/chosen": -1.5770976543426514, | |
| "logits/rejected": -1.4604318141937256, | |
| "logps/chosen": -478.90264892578125, | |
| "logps/rejected": -560.0563354492188, | |
| "loss": 0.557, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.8632774353027344, | |
| "rewards/margins": 0.6757477521896362, | |
| "rewards/rejected": -2.53902530670166, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.18998614684345932, | |
| "grad_norm": 12.93821128098639, | |
| "learning_rate": 4.88048035489807e-07, | |
| "logits/chosen": -1.5192902088165283, | |
| "logits/rejected": -1.3578670024871826, | |
| "logps/chosen": -401.8139343261719, | |
| "logps/rejected": -476.27557373046875, | |
| "loss": 0.5413, | |
| "rewards/accuracies": 0.7875000238418579, | |
| "rewards/chosen": -1.2336255311965942, | |
| "rewards/margins": 0.6196562051773071, | |
| "rewards/rejected": -1.8532816171646118, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.20054093277920707, | |
| "grad_norm": 10.174846260495245, | |
| "learning_rate": 4.85063294125718e-07, | |
| "logits/chosen": -1.4184257984161377, | |
| "logits/rejected": -1.349448561668396, | |
| "logps/chosen": -405.37030029296875, | |
| "logps/rejected": -465.87835693359375, | |
| "loss": 0.584, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -1.2496789693832397, | |
| "rewards/margins": 0.4809112548828125, | |
| "rewards/rejected": -1.7305902242660522, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.2110957187149548, | |
| "grad_norm": 11.934921621793764, | |
| "learning_rate": 4.817574845766874e-07, | |
| "logits/chosen": -1.6645218133926392, | |
| "logits/rejected": -1.4362655878067017, | |
| "logps/chosen": -419.6376953125, | |
| "logps/rejected": -511.97955322265625, | |
| "loss": 0.5179, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -1.6008678674697876, | |
| "rewards/margins": 0.7449867129325867, | |
| "rewards/rejected": -2.3458542823791504, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.22165050465070255, | |
| "grad_norm": 11.911318908319176, | |
| "learning_rate": 4.781351221809166e-07, | |
| "logits/chosen": -1.6439392566680908, | |
| "logits/rejected": -1.5432502031326294, | |
| "logps/chosen": -462.12249755859375, | |
| "logps/rejected": -561.0224609375, | |
| "loss": 0.563, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -2.0325913429260254, | |
| "rewards/margins": 0.8397369384765625, | |
| "rewards/rejected": -2.872328281402588, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.23220529058645029, | |
| "grad_norm": 12.017204037189543, | |
| "learning_rate": 4.742011546497182e-07, | |
| "logits/chosen": -1.5051389932632446, | |
| "logits/rejected": -1.4414231777191162, | |
| "logps/chosen": -449.30657958984375, | |
| "logps/rejected": -550.5023803710938, | |
| "loss": 0.5068, | |
| "rewards/accuracies": 0.8062499761581421, | |
| "rewards/chosen": -1.5960687398910522, | |
| "rewards/margins": 0.8259360194206238, | |
| "rewards/rejected": -2.4220046997070312, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.24276007652219803, | |
| "grad_norm": 16.38172474746581, | |
| "learning_rate": 4.6996095530953875e-07, | |
| "logits/chosen": -1.6456964015960693, | |
| "logits/rejected": -1.4991674423217773, | |
| "logps/chosen": -492.7826232910156, | |
| "logps/rejected": -631.8049926757812, | |
| "loss": 0.5084, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -2.1061949729919434, | |
| "rewards/margins": 1.1719015836715698, | |
| "rewards/rejected": -3.2780966758728027, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.25331486245794577, | |
| "grad_norm": 19.153944069009537, | |
| "learning_rate": 4.654203157626399e-07, | |
| "logits/chosen": -1.7168292999267578, | |
| "logits/rejected": -1.5979254245758057, | |
| "logps/chosen": -499.25775146484375, | |
| "logps/rejected": -663.107177734375, | |
| "loss": 0.5023, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -2.420952558517456, | |
| "rewards/margins": 1.3582748174667358, | |
| "rewards/rejected": -3.7792270183563232, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.26386964839369353, | |
| "grad_norm": 15.209663016177657, | |
| "learning_rate": 4.605854379764673e-07, | |
| "logits/chosen": -1.529827356338501, | |
| "logits/rejected": -1.4732040166854858, | |
| "logps/chosen": -422.29827880859375, | |
| "logps/rejected": -505.83154296875, | |
| "loss": 0.4999, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": -1.6309483051300049, | |
| "rewards/margins": 0.74617999792099, | |
| "rewards/rejected": -2.3771283626556396, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.27442443432944125, | |
| "grad_norm": 11.779884218050361, | |
| "learning_rate": 4.5546292581250857e-07, | |
| "logits/chosen": -1.5948470830917358, | |
| "logits/rejected": -1.4933321475982666, | |
| "logps/chosen": -455.269287109375, | |
| "logps/rejected": -543.4944458007812, | |
| "loss": 0.4868, | |
| "rewards/accuracies": 0.8062499761581421, | |
| "rewards/chosen": -1.7508251667022705, | |
| "rewards/margins": 0.8269698023796082, | |
| "rewards/rejected": -2.5777950286865234, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.284979220265189, | |
| "grad_norm": 19.10994211444829, | |
| "learning_rate": 4.5005977600621275e-07, | |
| "logits/chosen": -1.7589390277862549, | |
| "logits/rejected": -1.5872992277145386, | |
| "logps/chosen": -568.4318237304688, | |
| "logps/rejected": -727.857666015625, | |
| "loss": 0.4781, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -2.8650763034820557, | |
| "rewards/margins": 1.3692649602890015, | |
| "rewards/rejected": -4.234341621398926, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.2955340062009367, | |
| "grad_norm": 15.492507693472131, | |
| "learning_rate": 4.443833686102919e-07, | |
| "logits/chosen": -1.6332670450210571, | |
| "logits/rejected": -1.4034180641174316, | |
| "logps/chosen": -461.40826416015625, | |
| "logps/rejected": -608.6381225585938, | |
| "loss": 0.4705, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -1.9456875324249268, | |
| "rewards/margins": 1.1761558055877686, | |
| "rewards/rejected": -3.1218433380126953, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.3060887921366845, | |
| "grad_norm": 16.867870056444893, | |
| "learning_rate": 4.384414569144561e-07, | |
| "logits/chosen": -1.5909126996994019, | |
| "logits/rejected": -1.5102109909057617, | |
| "logps/chosen": -441.18670654296875, | |
| "logps/rejected": -565.0006713867188, | |
| "loss": 0.5053, | |
| "rewards/accuracies": 0.793749988079071, | |
| "rewards/chosen": -1.909717321395874, | |
| "rewards/margins": 1.0274746417999268, | |
| "rewards/rejected": -2.937191963195801, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.3166435780724322, | |
| "grad_norm": 12.53246744632567, | |
| "learning_rate": 4.3224215685535287e-07, | |
| "logits/chosen": -1.6234986782073975, | |
| "logits/rejected": -1.4319360256195068, | |
| "logps/chosen": -540.7374267578125, | |
| "logps/rejected": -702.7229614257812, | |
| "loss": 0.4621, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -2.5035765171051025, | |
| "rewards/margins": 1.4901249408721924, | |
| "rewards/rejected": -3.993701457977295, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.32719836400818, | |
| "grad_norm": 15.04185556896974, | |
| "learning_rate": 4.2579393593117364e-07, | |
| "logits/chosen": -1.533140778541565, | |
| "logits/rejected": -1.3760929107666016, | |
| "logps/chosen": -481.59954833984375, | |
| "logps/rejected": -652.427490234375, | |
| "loss": 0.4563, | |
| "rewards/accuracies": 0.8187500238418579, | |
| "rewards/chosen": -2.2860851287841797, | |
| "rewards/margins": 1.5269566774368286, | |
| "rewards/rejected": -3.8130416870117188, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.3377531499439277, | |
| "grad_norm": 13.12290477739756, | |
| "learning_rate": 4.191056016360699e-07, | |
| "logits/chosen": -1.6271283626556396, | |
| "logits/rejected": -1.5437796115875244, | |
| "logps/chosen": -519.0302734375, | |
| "logps/rejected": -663.0972900390625, | |
| "loss": 0.4474, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": -2.57169771194458, | |
| "rewards/margins": 1.3303985595703125, | |
| "rewards/rejected": -3.9020965099334717, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.34830793587967546, | |
| "grad_norm": 15.784620667841397, | |
| "learning_rate": 4.121862894301754e-07, | |
| "logits/chosen": -1.5655263662338257, | |
| "logits/rejected": -1.5315742492675781, | |
| "logps/chosen": -580.234375, | |
| "logps/rejected": -725.3365478515625, | |
| "loss": 0.4634, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": -2.9336700439453125, | |
| "rewards/margins": 1.4053281545639038, | |
| "rewards/rejected": -4.338997840881348, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.35886272181542317, | |
| "grad_norm": 18.950750373982537, | |
| "learning_rate": 4.050454502616667e-07, | |
| "logits/chosen": -1.7045695781707764, | |
| "logits/rejected": -1.6517536640167236, | |
| "logps/chosen": -554.7913208007812, | |
| "logps/rejected": -716.6380615234375, | |
| "loss": 0.4487, | |
| "rewards/accuracies": 0.8062499761581421, | |
| "rewards/chosen": -2.823065757751465, | |
| "rewards/margins": 1.4535109996795654, | |
| "rewards/rejected": -4.276576519012451, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.36941750775117094, | |
| "grad_norm": 17.91495880964947, | |
| "learning_rate": 3.976928376579047e-07, | |
| "logits/chosen": -1.6950843334197998, | |
| "logits/rejected": -1.582262396812439, | |
| "logps/chosen": -520.1227416992188, | |
| "logps/rejected": -664.7024536132812, | |
| "loss": 0.445, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -2.5456645488739014, | |
| "rewards/margins": 1.2352759838104248, | |
| "rewards/rejected": -3.780940532684326, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.37997229368691865, | |
| "grad_norm": 15.17521580280272, | |
| "learning_rate": 3.9013849440328945e-07, | |
| "logits/chosen": -1.6940500736236572, | |
| "logits/rejected": -1.5467922687530518, | |
| "logps/chosen": -520.3582763671875, | |
| "logps/rejected": -680.7635498046875, | |
| "loss": 0.4665, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": -2.551764726638794, | |
| "rewards/margins": 1.372143268585205, | |
| "rewards/rejected": -3.92390775680542, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.3905270796226664, | |
| "grad_norm": 14.308326454629483, | |
| "learning_rate": 3.8239273882202473e-07, | |
| "logits/chosen": -1.6619869470596313, | |
| "logits/rejected": -1.5419933795928955, | |
| "logps/chosen": -506.9136657714844, | |
| "logps/rejected": -639.9857177734375, | |
| "loss": 0.4574, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -2.3956522941589355, | |
| "rewards/margins": 1.23929762840271, | |
| "rewards/rejected": -3.6349494457244873, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.40108186555841413, | |
| "grad_norm": 15.767200038387838, | |
| "learning_rate": 3.7446615068452804e-07, | |
| "logits/chosen": -1.6123485565185547, | |
| "logits/rejected": -1.4534804821014404, | |
| "logps/chosen": -510.5772399902344, | |
| "logps/rejected": -681.8480224609375, | |
| "loss": 0.4326, | |
| "rewards/accuracies": 0.831250011920929, | |
| "rewards/chosen": -2.398505687713623, | |
| "rewards/margins": 1.5248647928237915, | |
| "rewards/rejected": -3.923370838165283, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.4116366514941619, | |
| "grad_norm": 16.282324670439472, | |
| "learning_rate": 3.6636955675673743e-07, | |
| "logits/chosen": -1.6264712810516357, | |
| "logits/rejected": -1.4750279188156128, | |
| "logps/chosen": -514.038330078125, | |
| "logps/rejected": -647.3781127929688, | |
| "loss": 0.4404, | |
| "rewards/accuracies": 0.8062499761581421, | |
| "rewards/chosen": -2.37794828414917, | |
| "rewards/margins": 1.2105674743652344, | |
| "rewards/rejected": -3.5885162353515625, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.4221914374299096, | |
| "grad_norm": 21.312257244757074, | |
| "learning_rate": 3.5811401601205093e-07, | |
| "logits/chosen": -1.5191190242767334, | |
| "logits/rejected": -1.5369209051132202, | |
| "logps/chosen": -559.904052734375, | |
| "logps/rejected": -721.0767211914062, | |
| "loss": 0.4493, | |
| "rewards/accuracies": 0.8374999761581421, | |
| "rewards/chosen": -2.823049545288086, | |
| "rewards/margins": 1.5707520246505737, | |
| "rewards/rejected": -4.393801689147949, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.4327462233656574, | |
| "grad_norm": 18.951174802021, | |
| "learning_rate": 3.497108045260995e-07, | |
| "logits/chosen": -1.6123138666152954, | |
| "logits/rejected": -1.5209693908691406, | |
| "logps/chosen": -508.78436279296875, | |
| "logps/rejected": -680.7193603515625, | |
| "loss": 0.4287, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": -2.3978936672210693, | |
| "rewards/margins": 1.566540002822876, | |
| "rewards/rejected": -3.9644336700439453, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.4433010093014051, | |
| "grad_norm": 28.219925216993275, | |
| "learning_rate": 3.411714000749838e-07, | |
| "logits/chosen": -1.661116361618042, | |
| "logits/rejected": -1.5592620372772217, | |
| "logps/chosen": -547.822998046875, | |
| "logps/rejected": -737.5689697265625, | |
| "loss": 0.4248, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": -2.826833724975586, | |
| "rewards/margins": 1.727709412574768, | |
| "rewards/rejected": -4.5545430183410645, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.45385579523715286, | |
| "grad_norm": 27.445717175281757, | |
| "learning_rate": 3.3250746645801287e-07, | |
| "logits/chosen": -1.6852436065673828, | |
| "logits/rejected": -1.6252915859222412, | |
| "logps/chosen": -603.220947265625, | |
| "logps/rejected": -795.3065185546875, | |
| "loss": 0.4409, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": -3.4767067432403564, | |
| "rewards/margins": 1.7523845434188843, | |
| "rewards/rejected": -5.229090690612793, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.46441058117290057, | |
| "grad_norm": 16.007752631407985, | |
| "learning_rate": 3.237308375663571e-07, | |
| "logits/chosen": -1.762291669845581, | |
| "logits/rejected": -1.5295162200927734, | |
| "logps/chosen": -531.2833862304688, | |
| "logps/rejected": -725.7946166992188, | |
| "loss": 0.3819, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": -2.7329602241516113, | |
| "rewards/margins": 1.7636291980743408, | |
| "rewards/rejected": -4.496589660644531, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.47496536710864834, | |
| "grad_norm": 18.506003391111292, | |
| "learning_rate": 3.148535012193767e-07, | |
| "logits/chosen": -1.754020094871521, | |
| "logits/rejected": -1.6287786960601807, | |
| "logps/chosen": -556.3197631835938, | |
| "logps/rejected": -728.1083984375, | |
| "loss": 0.3855, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": -2.739107370376587, | |
| "rewards/margins": 1.6264499425888062, | |
| "rewards/rejected": -4.3655571937561035, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.48552015304439605, | |
| "grad_norm": 20.68896480009483, | |
| "learning_rate": 3.0588758279070183e-07, | |
| "logits/chosen": -1.6177418231964111, | |
| "logits/rejected": -1.5763094425201416, | |
| "logps/chosen": -562.428955078125, | |
| "logps/rejected": -753.3512573242188, | |
| "loss": 0.4092, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -2.9177470207214355, | |
| "rewards/margins": 1.7720897197723389, | |
| "rewards/rejected": -4.6898369789123535, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.4960749389801438, | |
| "grad_norm": 16.381736251716195, | |
| "learning_rate": 2.968453286464312e-07, | |
| "logits/chosen": -1.4933600425720215, | |
| "logits/rejected": -1.5784003734588623, | |
| "logps/chosen": -504.9471740722656, | |
| "logps/rejected": -648.5574951171875, | |
| "loss": 0.4179, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -2.378976345062256, | |
| "rewards/margins": 1.3860208988189697, | |
| "rewards/rejected": -3.7649970054626465, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.5066297249158915, | |
| "grad_norm": 22.296205809275865, | |
| "learning_rate": 2.8773908941806877e-07, | |
| "logits/chosen": -1.6254231929779053, | |
| "logits/rejected": -1.621469259262085, | |
| "logps/chosen": -599.214599609375, | |
| "logps/rejected": -809.9522705078125, | |
| "loss": 0.39, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": -3.3130722045898438, | |
| "rewards/margins": 1.939295768737793, | |
| "rewards/rejected": -5.2523674964904785, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5171845108516393, | |
| "grad_norm": 17.974418608052964, | |
| "learning_rate": 2.785813031330473e-07, | |
| "logits/chosen": -1.6836649179458618, | |
| "logits/rejected": -1.6946824789047241, | |
| "logps/chosen": -636.5640869140625, | |
| "logps/rejected": -862.1500244140625, | |
| "loss": 0.407, | |
| "rewards/accuracies": 0.8187500238418579, | |
| "rewards/chosen": -3.709088087081909, | |
| "rewards/margins": 2.079369068145752, | |
| "rewards/rejected": -5.788456916809082, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.5277392967873871, | |
| "grad_norm": 21.869860477038394, | |
| "learning_rate": 2.693844782258779e-07, | |
| "logits/chosen": -1.597246766090393, | |
| "logits/rejected": -1.4808999300003052, | |
| "logps/chosen": -554.4249877929688, | |
| "logps/rejected": -735.31103515625, | |
| "loss": 0.377, | |
| "rewards/accuracies": 0.8374999761581421, | |
| "rewards/chosen": -2.8226776123046875, | |
| "rewards/margins": 1.673805832862854, | |
| "rewards/rejected": -4.49648380279541, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.5382940827231347, | |
| "grad_norm": 35.342954704070486, | |
| "learning_rate": 2.601611764531342e-07, | |
| "logits/chosen": -1.599726676940918, | |
| "logits/rejected": -1.5378262996673584, | |
| "logps/chosen": -625.1596069335938, | |
| "logps/rejected": -838.8173828125, | |
| "loss": 0.3907, | |
| "rewards/accuracies": 0.831250011920929, | |
| "rewards/chosen": -3.5563435554504395, | |
| "rewards/margins": 1.9609460830688477, | |
| "rewards/rejected": -5.517289638519287, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.5488488686588825, | |
| "grad_norm": 17.9297074159325, | |
| "learning_rate": 2.5092399573560323e-07, | |
| "logits/chosen": -1.6411758661270142, | |
| "logits/rejected": -1.6522302627563477, | |
| "logps/chosen": -658.8689575195312, | |
| "logps/rejected": -871.1637573242188, | |
| "loss": 0.4168, | |
| "rewards/accuracies": 0.8187500238418579, | |
| "rewards/chosen": -3.866881847381592, | |
| "rewards/margins": 2.113030195236206, | |
| "rewards/rejected": -5.979912757873535, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5594036545946303, | |
| "grad_norm": 16.443040045807887, | |
| "learning_rate": 2.4168555295104124e-07, | |
| "logits/chosen": -1.5852059125900269, | |
| "logits/rejected": -1.5837304592132568, | |
| "logps/chosen": -594.8711547851562, | |
| "logps/rejected": -801.4606323242188, | |
| "loss": 0.4093, | |
| "rewards/accuracies": 0.8374999761581421, | |
| "rewards/chosen": -3.3196072578430176, | |
| "rewards/margins": 1.8402855396270752, | |
| "rewards/rejected": -5.159893035888672, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.569958440530378, | |
| "grad_norm": 15.035798283991243, | |
| "learning_rate": 2.3245846670103626e-07, | |
| "logits/chosen": -1.5612332820892334, | |
| "logits/rejected": -1.5125606060028076, | |
| "logps/chosen": -579.2890014648438, | |
| "logps/rejected": -773.1954956054688, | |
| "loss": 0.3769, | |
| "rewards/accuracies": 0.8374999761581421, | |
| "rewards/chosen": -3.0682718753814697, | |
| "rewards/margins": 1.7783126831054688, | |
| "rewards/rejected": -4.846584320068359, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.5805132264661257, | |
| "grad_norm": 20.761167206822886, | |
| "learning_rate": 2.232553400755159e-07, | |
| "logits/chosen": -1.6236953735351562, | |
| "logits/rejected": -1.5238358974456787, | |
| "logps/chosen": -613.8088989257812, | |
| "logps/rejected": -867.7283325195312, | |
| "loss": 0.3697, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": -3.327275037765503, | |
| "rewards/margins": 2.372307538986206, | |
| "rewards/rejected": -5.699582576751709, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.5910680124018735, | |
| "grad_norm": 20.416122820074975, | |
| "learning_rate": 2.1408874343844294e-07, | |
| "logits/chosen": -1.6903560161590576, | |
| "logits/rejected": -1.5515328645706177, | |
| "logps/chosen": -648.6663818359375, | |
| "logps/rejected": -949.74609375, | |
| "loss": 0.3701, | |
| "rewards/accuracies": 0.8374999761581421, | |
| "rewards/chosen": -3.67596435546875, | |
| "rewards/margins": 2.625474214553833, | |
| "rewards/rejected": -6.301438808441162, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.6016227983376212, | |
| "grad_norm": 20.954462877817495, | |
| "learning_rate": 2.049711972582101e-07, | |
| "logits/chosen": -1.7400896549224854, | |
| "logits/rejected": -1.6318342685699463, | |
| "logps/chosen": -647.7393798828125, | |
| "logps/rejected": -899.9304809570312, | |
| "loss": 0.3719, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": -3.6654484272003174, | |
| "rewards/margins": 2.371953010559082, | |
| "rewards/rejected": -6.03740119934082, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.612177584273369, | |
| "grad_norm": 17.14928736018049, | |
| "learning_rate": 1.9591515500618588e-07, | |
| "logits/chosen": -1.6016016006469727, | |
| "logits/rejected": -1.5183677673339844, | |
| "logps/chosen": -670.8876342773438, | |
| "logps/rejected": -865.2810668945312, | |
| "loss": 0.4464, | |
| "rewards/accuracies": 0.793749988079071, | |
| "rewards/chosen": -3.901280641555786, | |
| "rewards/margins": 1.737006425857544, | |
| "rewards/rejected": -5.638287544250488, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6227323702091166, | |
| "grad_norm": 15.631817925073218, | |
| "learning_rate": 1.8693298614677112e-07, | |
| "logits/chosen": -1.4730761051177979, | |
| "logits/rejected": -1.3837854862213135, | |
| "logps/chosen": -539.0037231445312, | |
| "logps/rejected": -730.4061279296875, | |
| "loss": 0.3841, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": -2.58305025100708, | |
| "rewards/margins": 1.7921463251113892, | |
| "rewards/rejected": -4.37519645690918, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.6332871561448644, | |
| "grad_norm": 17.780553626895177, | |
| "learning_rate": 1.7803695924219814e-07, | |
| "logits/chosen": -1.4669979810714722, | |
| "logits/rejected": -1.4254872798919678, | |
| "logps/chosen": -584.629150390625, | |
| "logps/rejected": -784.3005981445312, | |
| "loss": 0.4051, | |
| "rewards/accuracies": 0.831250011920929, | |
| "rewards/chosen": -3.1204633712768555, | |
| "rewards/margins": 1.9275195598602295, | |
| "rewards/rejected": -5.047983169555664, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6438419420806122, | |
| "grad_norm": 19.866976638480853, | |
| "learning_rate": 1.6923922519515067e-07, | |
| "logits/chosen": -1.4443576335906982, | |
| "logits/rejected": -1.4355580806732178, | |
| "logps/chosen": -523.0612182617188, | |
| "logps/rejected": -729.1703491210938, | |
| "loss": 0.4069, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": -2.749462842941284, | |
| "rewards/margins": 1.8047136068344116, | |
| "rewards/rejected": -4.554176330566406, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.65439672801636, | |
| "grad_norm": 19.42045257459303, | |
| "learning_rate": 1.605518006520924e-07, | |
| "logits/chosen": -1.4894784688949585, | |
| "logits/rejected": -1.4559067487716675, | |
| "logps/chosen": -589.7485961914062, | |
| "logps/rejected": -808.5824584960938, | |
| "loss": 0.3816, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -3.1640734672546387, | |
| "rewards/margins": 2.0186519622802734, | |
| "rewards/rejected": -5.182725429534912, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.6649515139521076, | |
| "grad_norm": 34.470860837004935, | |
| "learning_rate": 1.519865515899731e-07, | |
| "logits/chosen": -1.5242574214935303, | |
| "logits/rejected": -1.3928359746932983, | |
| "logps/chosen": -588.0692138671875, | |
| "logps/rejected": -841.7396240234375, | |
| "loss": 0.3858, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": -3.3201167583465576, | |
| "rewards/margins": 2.1101181507110596, | |
| "rewards/rejected": -5.430234432220459, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.6755062998878554, | |
| "grad_norm": 30.007742834102533, | |
| "learning_rate": 1.4355517710873182e-07, | |
| "logits/chosen": -1.546661615371704, | |
| "logits/rejected": -1.4665791988372803, | |
| "logps/chosen": -627.7368774414062, | |
| "logps/rejected": -869.39453125, | |
| "loss": 0.3914, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": -3.6285042762756348, | |
| "rewards/margins": 2.2051749229431152, | |
| "rewards/rejected": -5.83367919921875, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.6860610858236031, | |
| "grad_norm": 34.25630759725382, | |
| "learning_rate": 1.3526919345173318e-07, | |
| "logits/chosen": -1.5128138065338135, | |
| "logits/rejected": -1.4411523342132568, | |
| "logps/chosen": -602.6571044921875, | |
| "logps/rejected": -849.2667846679688, | |
| "loss": 0.4211, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": -3.323326826095581, | |
| "rewards/margins": 2.333648681640625, | |
| "rewards/rejected": -5.656975269317627, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.6966158717593509, | |
| "grad_norm": 28.233731363069086, | |
| "learning_rate": 1.2713991827596443e-07, | |
| "logits/chosen": -1.495444655418396, | |
| "logits/rejected": -1.4902544021606445, | |
| "logps/chosen": -605.8360595703125, | |
| "logps/rejected": -796.5411376953125, | |
| "loss": 0.3791, | |
| "rewards/accuracies": 0.8187500238418579, | |
| "rewards/chosen": -3.165205478668213, | |
| "rewards/margins": 1.914367914199829, | |
| "rewards/rejected": -5.079573631286621, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.7071706576950986, | |
| "grad_norm": 28.532009996830972, | |
| "learning_rate": 1.191784551934773e-07, | |
| "logits/chosen": -1.5311321020126343, | |
| "logits/rejected": -1.444746971130371, | |
| "logps/chosen": -515.9117431640625, | |
| "logps/rejected": -714.9118041992188, | |
| "loss": 0.3985, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": -2.741084575653076, | |
| "rewards/margins": 1.8353731632232666, | |
| "rewards/rejected": -4.5764570236206055, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.7177254436308463, | |
| "grad_norm": 34.722496076575624, | |
| "learning_rate": 1.1139567860518953e-07, | |
| "logits/chosen": -1.3724250793457031, | |
| "logits/rejected": -1.3445093631744385, | |
| "logps/chosen": -516.8666381835938, | |
| "logps/rejected": -690.0352172851562, | |
| "loss": 0.4182, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": -2.5948445796966553, | |
| "rewards/margins": 1.631908655166626, | |
| "rewards/rejected": -4.226753234863281, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.7282802295665941, | |
| "grad_norm": 29.539595863207374, | |
| "learning_rate": 1.0380221884776128e-07, | |
| "logits/chosen": -1.4173920154571533, | |
| "logits/rejected": -1.3776549100875854, | |
| "logps/chosen": -542.7964477539062, | |
| "logps/rejected": -722.7305297851562, | |
| "loss": 0.4204, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": -2.7417469024658203, | |
| "rewards/margins": 1.6534277200698853, | |
| "rewards/rejected": -4.395174980163574, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.7388350155023419, | |
| "grad_norm": 17.47998143760053, | |
| "learning_rate": 9.640844767383405e-08, | |
| "logits/chosen": -1.3427600860595703, | |
| "logits/rejected": -1.2479599714279175, | |
| "logps/chosen": -540.9366455078125, | |
| "logps/rejected": -741.2656860351562, | |
| "loss": 0.4206, | |
| "rewards/accuracies": 0.831250011920929, | |
| "rewards/chosen": -2.7238881587982178, | |
| "rewards/margins": 1.8479011058807373, | |
| "rewards/rejected": -4.571789741516113, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.7493898014380895, | |
| "grad_norm": 22.53665974937468, | |
| "learning_rate": 8.922446408546378e-08, | |
| "logits/chosen": -1.3614610433578491, | |
| "logits/rejected": -1.3304545879364014, | |
| "logps/chosen": -564.0736694335938, | |
| "logps/rejected": -796.0197143554688, | |
| "loss": 0.4234, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": -2.8989734649658203, | |
| "rewards/margins": 2.1152186393737793, | |
| "rewards/rejected": -5.0141921043396, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.7599445873738373, | |
| "grad_norm": 34.016623930856696, | |
| "learning_rate": 8.22600805400994e-08, | |
| "logits/chosen": -1.4026951789855957, | |
| "logits/rejected": -1.3086416721343994, | |
| "logps/chosen": -557.1563720703125, | |
| "logps/rejected": -775.2821044921875, | |
| "loss": 0.3918, | |
| "rewards/accuracies": 0.8687499761581421, | |
| "rewards/chosen": -2.755949020385742, | |
| "rewards/margins": 1.9373924732208252, | |
| "rewards/rejected": -4.6933417320251465, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.7704993733095851, | |
| "grad_norm": 19.583973928513963, | |
| "learning_rate": 7.552480954794558e-08, | |
| "logits/chosen": -1.4313266277313232, | |
| "logits/rejected": -1.3103513717651367, | |
| "logps/chosen": -564.1690673828125, | |
| "logps/rejected": -756.8692626953125, | |
| "loss": 0.4048, | |
| "rewards/accuracies": 0.7875000238418579, | |
| "rewards/chosen": -2.88138484954834, | |
| "rewards/margins": 1.8533366918563843, | |
| "rewards/rejected": -4.7347211837768555, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.7810541592453328, | |
| "grad_norm": 18.206536925706505, | |
| "learning_rate": 6.902785067901854e-08, | |
| "logits/chosen": -1.3645613193511963, | |
| "logits/rejected": -1.3341350555419922, | |
| "logps/chosen": -579.579833984375, | |
| "logps/rejected": -779.4511108398438, | |
| "loss": 0.3792, | |
| "rewards/accuracies": 0.8187500238418579, | |
| "rewards/chosen": -3.0597000122070312, | |
| "rewards/margins": 1.8546708822250366, | |
| "rewards/rejected": -4.914370536804199, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.7916089451810805, | |
| "grad_norm": 25.99340540536939, | |
| "learning_rate": 6.277807799763973e-08, | |
| "logits/chosen": -1.5334607362747192, | |
| "logits/rejected": -1.4499049186706543, | |
| "logps/chosen": -563.6980590820312, | |
| "logps/rejected": -789.72119140625, | |
| "loss": 0.3903, | |
| "rewards/accuracies": 0.8687499761581421, | |
| "rewards/chosen": -3.1820693016052246, | |
| "rewards/margins": 1.9633289575576782, | |
| "rewards/rejected": -5.145398139953613, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.8021637311168283, | |
| "grad_norm": 27.363349380647865, | |
| "learning_rate": 5.678402794153145e-08, | |
| "logits/chosen": -1.490678071975708, | |
| "logits/rejected": -1.4118311405181885, | |
| "logps/chosen": -635.3906860351562, | |
| "logps/rejected": -841.986328125, | |
| "loss": 0.3978, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -3.6473312377929688, | |
| "rewards/margins": 1.8795642852783203, | |
| "rewards/rejected": -5.526895523071289, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.812718517052576, | |
| "grad_norm": 22.024184023637442, | |
| "learning_rate": 5.105388766206969e-08, | |
| "logits/chosen": -1.6372134685516357, | |
| "logits/rejected": -1.4456651210784912, | |
| "logps/chosen": -670.1497802734375, | |
| "logps/rejected": -900.1561279296875, | |
| "loss": 0.4107, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": -3.8800835609436035, | |
| "rewards/margins": 2.1356186866760254, | |
| "rewards/rejected": -6.015702724456787, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.8232733029883238, | |
| "grad_norm": 18.957756630133876, | |
| "learning_rate": 4.5595483841620484e-08, | |
| "logits/chosen": -1.4949450492858887, | |
| "logits/rejected": -1.466933250427246, | |
| "logps/chosen": -592.1412963867188, | |
| "logps/rejected": -814.1785888671875, | |
| "loss": 0.3793, | |
| "rewards/accuracies": 0.8374999761581421, | |
| "rewards/chosen": -3.3024775981903076, | |
| "rewards/margins": 1.9800523519515991, | |
| "rewards/rejected": -5.282530307769775, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.8338280889240715, | |
| "grad_norm": 24.130033467348618, | |
| "learning_rate": 4.0416272003232526e-08, | |
| "logits/chosen": -1.5286778211593628, | |
| "logits/rejected": -1.4018694162368774, | |
| "logps/chosen": -614.2500610351562, | |
| "logps/rejected": -848.65673828125, | |
| "loss": 0.4233, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": -3.3793911933898926, | |
| "rewards/margins": 2.2501933574676514, | |
| "rewards/rejected": -5.629584312438965, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.8443828748598192, | |
| "grad_norm": 21.924780448040742, | |
| "learning_rate": 3.552332632729041e-08, | |
| "logits/chosen": -1.3315622806549072, | |
| "logits/rejected": -1.4189374446868896, | |
| "logps/chosen": -591.8238525390625, | |
| "logps/rejected": -745.0551147460938, | |
| "loss": 0.4269, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -3.1886703968048096, | |
| "rewards/margins": 1.4928715229034424, | |
| "rewards/rejected": -4.68154239654541, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8443828748598192, | |
| "eval_logits/chosen": -1.3921160697937012, | |
| "eval_logits/rejected": -1.360321283340454, | |
| "eval_logps/chosen": -602.09814453125, | |
| "eval_logps/rejected": -797.6878051757812, | |
| "eval_loss": 0.38372838497161865, | |
| "eval_rewards/accuracies": 0.8643724918365479, | |
| "eval_rewards/chosen": -3.2510859966278076, | |
| "eval_rewards/margins": 1.8691294193267822, | |
| "eval_rewards/rejected": -5.120214939117432, | |
| "eval_runtime": 313.5439, | |
| "eval_samples_per_second": 6.302, | |
| "eval_steps_per_second": 1.576, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.854937660795567, | |
| "grad_norm": 19.316578134763, | |
| "learning_rate": 3.092332998903416e-08, | |
| "logits/chosen": -1.4394410848617554, | |
| "logits/rejected": -1.3752410411834717, | |
| "logps/chosen": -601.6661987304688, | |
| "logps/rejected": -783.7593994140625, | |
| "loss": 0.3863, | |
| "rewards/accuracies": 0.8187500238418579, | |
| "rewards/chosen": -3.3254780769348145, | |
| "rewards/margins": 1.7059745788574219, | |
| "rewards/rejected": -5.031452655792236, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.8654924467313148, | |
| "grad_norm": 28.757209433347523, | |
| "learning_rate": 2.6622566030146455e-08, | |
| "logits/chosen": -1.4320390224456787, | |
| "logits/rejected": -1.413570523262024, | |
| "logps/chosen": -565.2530517578125, | |
| "logps/rejected": -754.3922729492188, | |
| "loss": 0.4213, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": -2.987544298171997, | |
| "rewards/margins": 1.7588831186294556, | |
| "rewards/rejected": -4.7464280128479, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.8760472326670625, | |
| "grad_norm": 18.43535305501894, | |
| "learning_rate": 2.26269087768734e-08, | |
| "logits/chosen": -1.4583203792572021, | |
| "logits/rejected": -1.3761074542999268, | |
| "logps/chosen": -577.720703125, | |
| "logps/rejected": -806.8515014648438, | |
| "loss": 0.3829, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": -3.2120673656463623, | |
| "rewards/margins": 2.084867477416992, | |
| "rewards/rejected": -5.296935081481934, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.8866020186028102, | |
| "grad_norm": 20.96371705639774, | |
| "learning_rate": 1.894181581640106e-08, | |
| "logits/chosen": -1.384445071220398, | |
| "logits/rejected": -1.4241831302642822, | |
| "logps/chosen": -569.5962524414062, | |
| "logps/rejected": -749.0470581054688, | |
| "loss": 0.3819, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": -3.081413745880127, | |
| "rewards/margins": 1.7126514911651611, | |
| "rewards/rejected": -4.794064998626709, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.897156804538558, | |
| "grad_norm": 26.21197116684321, | |
| "learning_rate": 1.5572320542448143e-08, | |
| "logits/chosen": -1.3703296184539795, | |
| "logits/rejected": -1.3327410221099854, | |
| "logps/chosen": -617.3011474609375, | |
| "logps/rejected": -836.2391357421875, | |
| "loss": 0.3825, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": -3.3691182136535645, | |
| "rewards/margins": 2.073474407196045, | |
| "rewards/rejected": -5.442592620849609, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.9077115904743057, | |
| "grad_norm": 22.75331582162892, | |
| "learning_rate": 1.2523025280255729e-08, | |
| "logits/chosen": -1.4123306274414062, | |
| "logits/rejected": -1.4244548082351685, | |
| "logps/chosen": -593.9744262695312, | |
| "logps/rejected": -802.8662719726562, | |
| "loss": 0.3422, | |
| "rewards/accuracies": 0.8687499761581421, | |
| "rewards/chosen": -3.253312349319458, | |
| "rewards/margins": 1.9519180059432983, | |
| "rewards/rejected": -5.205230712890625, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.9182663764100535, | |
| "grad_norm": 17.924586145441417, | |
| "learning_rate": 9.798095000364214e-09, | |
| "logits/chosen": -1.5641670227050781, | |
| "logits/rejected": -1.4084519147872925, | |
| "logps/chosen": -600.7659912109375, | |
| "logps/rejected": -869.5779418945312, | |
| "loss": 0.3553, | |
| "rewards/accuracies": 0.8812500238418579, | |
| "rewards/chosen": -3.2969956398010254, | |
| "rewards/margins": 2.371945381164551, | |
| "rewards/rejected": -5.668941020965576, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.9288211623458011, | |
| "grad_norm": 25.268001512531786, | |
| "learning_rate": 7.401251629764876e-09, | |
| "logits/chosen": -1.5503554344177246, | |
| "logits/rejected": -1.3530725240707397, | |
| "logps/chosen": -560.983154296875, | |
| "logps/rejected": -772.5062255859375, | |
| "loss": 0.4149, | |
| "rewards/accuracies": 0.8374999761581421, | |
| "rewards/chosen": -3.1528074741363525, | |
| "rewards/margins": 1.8247524499893188, | |
| "rewards/rejected": -4.977559566497803, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.9393759482815489, | |
| "grad_norm": 18.848350951492684, | |
| "learning_rate": 5.335768968195098e-09, | |
| "logits/chosen": -1.4744082689285278, | |
| "logits/rejected": -1.370416283607483, | |
| "logps/chosen": -589.2518920898438, | |
| "logps/rejected": -806.16943359375, | |
| "loss": 0.4069, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": -3.1618690490722656, | |
| "rewards/margins": 2.018306255340576, | |
| "rewards/rejected": -5.180174827575684, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.9499307342172967, | |
| "grad_norm": 14.44527353219914, | |
| "learning_rate": 3.604468216521883e-09, | |
| "logits/chosen": -1.3630057573318481, | |
| "logits/rejected": -1.2414896488189697, | |
| "logps/chosen": -591.0068359375, | |
| "logps/rejected": -788.7886352539062, | |
| "loss": 0.371, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": -3.234240770339966, | |
| "rewards/margins": 1.8283990621566772, | |
| "rewards/rejected": -5.0626397132873535, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.9604855201530444, | |
| "grad_norm": 17.894278742200918, | |
| "learning_rate": 2.2097141233206884e-09, | |
| "logits/chosen": -1.5407589673995972, | |
| "logits/rejected": -1.4432313442230225, | |
| "logps/chosen": -616.9603271484375, | |
| "logps/rejected": -858.2979736328125, | |
| "loss": 0.3764, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": -3.5265698432922363, | |
| "rewards/margins": 2.2084403038024902, | |
| "rewards/rejected": -5.735010623931885, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.9710403060887921, | |
| "grad_norm": 18.751457793725724, | |
| "learning_rate": 1.1534117549133472e-09, | |
| "logits/chosen": -1.46907639503479, | |
| "logits/rejected": -1.3426740169525146, | |
| "logps/chosen": -629.7129516601562, | |
| "logps/rejected": -850.1477661132812, | |
| "loss": 0.3512, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": -3.58038330078125, | |
| "rewards/margins": 2.005390167236328, | |
| "rewards/rejected": -5.585773468017578, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.9815950920245399, | |
| "grad_norm": 29.39806840137654, | |
| "learning_rate": 4.3700389327672173e-10, | |
| "logits/chosen": -1.469868779182434, | |
| "logits/rejected": -1.3952000141143799, | |
| "logps/chosen": -604.9054565429688, | |
| "logps/rejected": -832.4953002929688, | |
| "loss": 0.3768, | |
| "rewards/accuracies": 0.8687499761581421, | |
| "rewards/chosen": -3.387165069580078, | |
| "rewards/margins": 2.0868310928344727, | |
| "rewards/rejected": -5.473996162414551, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.9921498779602876, | |
| "grad_norm": 22.359148428267968, | |
| "learning_rate": 6.146906537587982e-11, | |
| "logits/chosen": -1.4432518482208252, | |
| "logits/rejected": -1.4169013500213623, | |
| "logps/chosen": -600.5186157226562, | |
| "logps/rejected": -811.341796875, | |
| "loss": 0.3939, | |
| "rewards/accuracies": 0.8062499761581421, | |
| "rewards/chosen": -3.3816261291503906, | |
| "rewards/margins": 2.0300629138946533, | |
| "rewards/rejected": -5.411688804626465, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.9984827495217362, | |
| "step": 473, | |
| "total_flos": 0.0, | |
| "train_loss": 0.4645486564767285, | |
| "train_runtime": 24049.7915, | |
| "train_samples_per_second": 2.521, | |
| "train_steps_per_second": 0.02 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 473, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 1000000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |