| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 500, | |
| "global_step": 1250, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 4e-08, | |
| "logits/chosen": -2.6577353477478027, | |
| "logits/rejected": -2.043900489807129, | |
| "logps/chosen": -505.98724365234375, | |
| "logps/rejected": -319.40179443359375, | |
| "loss": 0.6931, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 4.0000000000000003e-07, | |
| "logits/chosen": -2.4769232273101807, | |
| "logits/rejected": -2.134289503097534, | |
| "logps/chosen": -285.32098388671875, | |
| "logps/rejected": -191.55812072753906, | |
| "loss": 0.6932, | |
| "rewards/accuracies": 0.3333333432674408, | |
| "rewards/chosen": 0.00036080353311263025, | |
| "rewards/margins": -9.262199455406517e-05, | |
| "rewards/rejected": 0.0004534254549071193, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 8.000000000000001e-07, | |
| "logits/chosen": -2.3606486320495605, | |
| "logits/rejected": -2.1270670890808105, | |
| "logps/chosen": -271.58087158203125, | |
| "logps/rejected": -208.8097381591797, | |
| "loss": 0.6934, | |
| "rewards/accuracies": 0.2750000059604645, | |
| "rewards/chosen": -0.0005878033698536456, | |
| "rewards/margins": -0.0015078920405358076, | |
| "rewards/rejected": 0.0009200887870974839, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 1.2000000000000002e-06, | |
| "logits/chosen": -2.277930498123169, | |
| "logits/rejected": -2.2482991218566895, | |
| "logps/chosen": -269.87554931640625, | |
| "logps/rejected": -288.9063720703125, | |
| "loss": 0.693, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": 0.002121593803167343, | |
| "rewards/margins": 0.0005655083805322647, | |
| "rewards/rejected": 0.0015560855390504003, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 1.6000000000000001e-06, | |
| "logits/chosen": -2.508906841278076, | |
| "logits/rejected": -2.341235637664795, | |
| "logps/chosen": -210.5034942626953, | |
| "logps/rejected": -181.61361694335938, | |
| "loss": 0.6926, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 0.0023279793094843626, | |
| "rewards/margins": 0.0018606262747198343, | |
| "rewards/rejected": 0.0004673528019338846, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "logits/chosen": -2.36310076713562, | |
| "logits/rejected": -2.3391873836517334, | |
| "logps/chosen": -195.63548278808594, | |
| "logps/rejected": -211.3179168701172, | |
| "loss": 0.6924, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 0.0008935689693316817, | |
| "rewards/margins": 0.0021855938248336315, | |
| "rewards/rejected": -0.0012920246226713061, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 2.4000000000000003e-06, | |
| "logits/chosen": -2.4950430393218994, | |
| "logits/rejected": -2.303813934326172, | |
| "logps/chosen": -244.68856811523438, | |
| "logps/rejected": -277.8478698730469, | |
| "loss": 0.6922, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -0.010385877452790737, | |
| "rewards/margins": -0.0004113853501621634, | |
| "rewards/rejected": -0.009974492713809013, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 2.8000000000000003e-06, | |
| "logits/chosen": -2.2493319511413574, | |
| "logits/rejected": -2.266305446624756, | |
| "logps/chosen": -229.941162109375, | |
| "logps/rejected": -217.0909881591797, | |
| "loss": 0.6901, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.011110167019069195, | |
| "rewards/margins": 0.010483531281352043, | |
| "rewards/rejected": -0.021593699231743813, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 3.2000000000000003e-06, | |
| "logits/chosen": -2.399794816970825, | |
| "logits/rejected": -2.393554449081421, | |
| "logps/chosen": -266.1021423339844, | |
| "logps/rejected": -263.08526611328125, | |
| "loss": 0.6861, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.014540717005729675, | |
| "rewards/margins": 0.021736864000558853, | |
| "rewards/rejected": -0.03627758100628853, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 3.6000000000000003e-06, | |
| "logits/chosen": -2.32847261428833, | |
| "logits/rejected": -1.991908073425293, | |
| "logps/chosen": -268.0962219238281, | |
| "logps/rejected": -179.7996826171875, | |
| "loss": 0.6863, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.027069438248872757, | |
| "rewards/margins": 0.021108368411660194, | |
| "rewards/rejected": -0.048177801072597504, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 4.000000000000001e-06, | |
| "logits/chosen": -2.3088860511779785, | |
| "logits/rejected": -2.312539577484131, | |
| "logps/chosen": -274.9145202636719, | |
| "logps/rejected": -250.87716674804688, | |
| "loss": 0.6898, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.031584225594997406, | |
| "rewards/margins": 0.010575409047305584, | |
| "rewards/rejected": -0.042159631848335266, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 4.4e-06, | |
| "logits/chosen": -2.2585253715515137, | |
| "logits/rejected": -2.144835948944092, | |
| "logps/chosen": -211.51986694335938, | |
| "logps/rejected": -179.9736785888672, | |
| "loss": 0.6819, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": 0.005594385787844658, | |
| "rewards/margins": 0.03688768297433853, | |
| "rewards/rejected": -0.03129329904913902, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 4.800000000000001e-06, | |
| "logits/chosen": -2.1611924171447754, | |
| "logits/rejected": -2.171570301055908, | |
| "logps/chosen": -237.06216430664062, | |
| "logps/rejected": -269.0331115722656, | |
| "loss": 0.6703, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.007297643460333347, | |
| "rewards/margins": 0.047815777361392975, | |
| "rewards/rejected": -0.055113423615694046, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 4.999756310023261e-06, | |
| "logits/chosen": -2.448563814163208, | |
| "logits/rejected": -2.3213229179382324, | |
| "logps/chosen": -235.3451690673828, | |
| "logps/rejected": -192.87477111816406, | |
| "loss": 0.6764, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.011573897674679756, | |
| "rewards/margins": 0.052028268575668335, | |
| "rewards/rejected": -0.06360216438770294, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 4.997807075247147e-06, | |
| "logits/chosen": -2.3375329971313477, | |
| "logits/rejected": -2.034670352935791, | |
| "logps/chosen": -199.99537658691406, | |
| "logps/rejected": -180.91787719726562, | |
| "loss": 0.6737, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.08778422325849533, | |
| "rewards/margins": 0.04929418861865997, | |
| "rewards/rejected": -0.1370784044265747, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 4.993910125649561e-06, | |
| "logits/chosen": -2.323729991912842, | |
| "logits/rejected": -2.20963191986084, | |
| "logps/chosen": -229.9593048095703, | |
| "logps/rejected": -198.73062133789062, | |
| "loss": 0.6566, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.018689800053834915, | |
| "rewards/margins": 0.08985014259815216, | |
| "rewards/rejected": -0.10853993892669678, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 4.988068499954578e-06, | |
| "logits/chosen": -2.215893507003784, | |
| "logits/rejected": -2.2889232635498047, | |
| "logps/chosen": -277.0525207519531, | |
| "logps/rejected": -287.1373596191406, | |
| "loss": 0.6778, | |
| "rewards/accuracies": 0.44999998807907104, | |
| "rewards/chosen": -0.24399030208587646, | |
| "rewards/margins": 0.03944157436490059, | |
| "rewards/rejected": -0.28343185782432556, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 4.980286753286196e-06, | |
| "logits/chosen": -2.4302048683166504, | |
| "logits/rejected": -2.032261848449707, | |
| "logps/chosen": -285.89691162109375, | |
| "logps/rejected": -248.88278198242188, | |
| "loss": 0.6419, | |
| "rewards/accuracies": 0.44999998807907104, | |
| "rewards/chosen": -0.18989089131355286, | |
| "rewards/margins": 0.04870374873280525, | |
| "rewards/rejected": -0.23859462141990662, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 4.970570953616383e-06, | |
| "logits/chosen": -2.2379422187805176, | |
| "logits/rejected": -2.230316400527954, | |
| "logps/chosen": -235.77371215820312, | |
| "logps/rejected": -240.2162628173828, | |
| "loss": 0.6573, | |
| "rewards/accuracies": 0.42500001192092896, | |
| "rewards/chosen": -0.0921173095703125, | |
| "rewards/margins": 0.044399768114089966, | |
| "rewards/rejected": -0.13651707768440247, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 4.958928677033465e-06, | |
| "logits/chosen": -2.2377524375915527, | |
| "logits/rejected": -2.1663427352905273, | |
| "logps/chosen": -257.6595764160156, | |
| "logps/rejected": -243.8786163330078, | |
| "loss": 0.6519, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.2135959416627884, | |
| "rewards/margins": 0.13946585357189178, | |
| "rewards/rejected": -0.3530617952346802, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 4.9453690018345144e-06, | |
| "logits/chosen": -2.29764723777771, | |
| "logits/rejected": -2.1510560512542725, | |
| "logps/chosen": -285.13702392578125, | |
| "logps/rejected": -305.1742858886719, | |
| "loss": 0.6381, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.3077496588230133, | |
| "rewards/margins": 0.12263032048940659, | |
| "rewards/rejected": -0.4303800165653229, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 4.9299025014463665e-06, | |
| "logits/chosen": -2.4579226970672607, | |
| "logits/rejected": -2.097339153289795, | |
| "logps/chosen": -314.6923828125, | |
| "logps/rejected": -261.67041015625, | |
| "loss": 0.613, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.4118514657020569, | |
| "rewards/margins": 0.30009493231773376, | |
| "rewards/rejected": -0.7119464874267578, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 4.912541236180779e-06, | |
| "logits/chosen": -2.31429123878479, | |
| "logits/rejected": -1.8840612173080444, | |
| "logps/chosen": -303.4709777832031, | |
| "logps/rejected": -296.21197509765625, | |
| "loss": 0.6429, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.6690406799316406, | |
| "rewards/margins": 0.3159363567829132, | |
| "rewards/rejected": -0.984977126121521, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 4.893298743830168e-06, | |
| "logits/chosen": -2.283473491668701, | |
| "logits/rejected": -2.319681167602539, | |
| "logps/chosen": -240.64303588867188, | |
| "logps/rejected": -283.9293212890625, | |
| "loss": 0.6464, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.4180280268192291, | |
| "rewards/margins": 0.16594772040843964, | |
| "rewards/rejected": -0.5839756727218628, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 4.8721900291112415e-06, | |
| "logits/chosen": -2.2171471118927, | |
| "logits/rejected": -2.035769462585449, | |
| "logps/chosen": -257.6659851074219, | |
| "logps/rejected": -258.67364501953125, | |
| "loss": 0.624, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.5104702711105347, | |
| "rewards/margins": 0.13771632313728333, | |
| "rewards/rejected": -0.6481865644454956, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 4.849231551964771e-06, | |
| "logits/chosen": -2.0292515754699707, | |
| "logits/rejected": -1.9135668277740479, | |
| "logps/chosen": -286.9338684082031, | |
| "logps/rejected": -313.9806213378906, | |
| "loss": 0.6291, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -1.169675588607788, | |
| "rewards/margins": 0.24345961213111877, | |
| "rewards/rejected": -1.4131351709365845, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 4.824441214720629e-06, | |
| "logits/chosen": -2.203101634979248, | |
| "logits/rejected": -1.946657419204712, | |
| "logps/chosen": -382.1039123535156, | |
| "logps/rejected": -404.14892578125, | |
| "loss": 0.5794, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.0869503021240234, | |
| "rewards/margins": 0.3835533559322357, | |
| "rewards/rejected": -1.470503568649292, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 4.7978383481380865e-06, | |
| "logits/chosen": -2.212496519088745, | |
| "logits/rejected": -1.996314287185669, | |
| "logps/chosen": -307.20587158203125, | |
| "logps/rejected": -306.6591796875, | |
| "loss": 0.5909, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.5885969400405884, | |
| "rewards/margins": 0.3659537732601166, | |
| "rewards/rejected": -0.9545507431030273, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 4.769443696332272e-06, | |
| "logits/chosen": -1.9687378406524658, | |
| "logits/rejected": -2.04164719581604, | |
| "logps/chosen": -249.0873565673828, | |
| "logps/rejected": -323.9103088378906, | |
| "loss": 0.5854, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.8271968960762024, | |
| "rewards/margins": 0.3957016170024872, | |
| "rewards/rejected": -1.2228984832763672, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 4.7392794005985324e-06, | |
| "logits/chosen": -2.2899293899536133, | |
| "logits/rejected": -2.000715970993042, | |
| "logps/chosen": -374.6996154785156, | |
| "logps/rejected": -359.8489685058594, | |
| "loss": 0.5851, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.9794802665710449, | |
| "rewards/margins": 0.4062279760837555, | |
| "rewards/rejected": -1.385708212852478, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 4.707368982147318e-06, | |
| "logits/chosen": -2.028409242630005, | |
| "logits/rejected": -1.9832637310028076, | |
| "logps/chosen": -426.76385498046875, | |
| "logps/rejected": -396.494384765625, | |
| "loss": 0.5501, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.3581629991531372, | |
| "rewards/margins": 0.4999775290489197, | |
| "rewards/rejected": -1.8581405878067017, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 4.673737323763048e-06, | |
| "logits/chosen": -1.773838996887207, | |
| "logits/rejected": -1.8148994445800781, | |
| "logps/chosen": -377.904541015625, | |
| "logps/rejected": -456.285400390625, | |
| "loss": 0.6464, | |
| "rewards/accuracies": 0.44999998807907104, | |
| "rewards/chosen": -1.8526346683502197, | |
| "rewards/margins": 0.3698161244392395, | |
| "rewards/rejected": -2.2224507331848145, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 4.638410650401267e-06, | |
| "logits/chosen": -2.000642776489258, | |
| "logits/rejected": -1.6485764980316162, | |
| "logps/chosen": -337.97711181640625, | |
| "logps/rejected": -298.59136962890625, | |
| "loss": 0.6445, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.8390266299247742, | |
| "rewards/margins": 0.39421704411506653, | |
| "rewards/rejected": -1.233243703842163, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 4.601416508739211e-06, | |
| "logits/chosen": -2.2264564037323, | |
| "logits/rejected": -2.249499797821045, | |
| "logps/chosen": -326.7769470214844, | |
| "logps/rejected": -301.43341064453125, | |
| "loss": 0.6091, | |
| "rewards/accuracies": 0.4749999940395355, | |
| "rewards/chosen": -0.5400243997573853, | |
| "rewards/margins": 0.13523522019386292, | |
| "rewards/rejected": -0.6752597093582153, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 4.562783745695738e-06, | |
| "logits/chosen": -2.1335179805755615, | |
| "logits/rejected": -1.8753182888031006, | |
| "logps/chosen": -247.4087677001953, | |
| "logps/rejected": -241.4481201171875, | |
| "loss": 0.6052, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.5356850624084473, | |
| "rewards/margins": 0.29491013288497925, | |
| "rewards/rejected": -0.8305951952934265, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 4.522542485937369e-06, | |
| "logits/chosen": -2.284304141998291, | |
| "logits/rejected": -1.9852168560028076, | |
| "logps/chosen": -362.0438232421875, | |
| "logps/rejected": -358.851806640625, | |
| "loss": 0.5885, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -1.0619395971298218, | |
| "rewards/margins": 0.26577672362327576, | |
| "rewards/rejected": -1.32771635055542, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 4.4807241083879774e-06, | |
| "logits/chosen": -2.22493052482605, | |
| "logits/rejected": -2.0395853519439697, | |
| "logps/chosen": -384.7464904785156, | |
| "logps/rejected": -394.67999267578125, | |
| "loss": 0.541, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.2721855640411377, | |
| "rewards/margins": 0.4639417231082916, | |
| "rewards/rejected": -1.736127495765686, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 4.437361221760449e-06, | |
| "logits/chosen": -2.074812412261963, | |
| "logits/rejected": -2.015082597732544, | |
| "logps/chosen": -367.3848876953125, | |
| "logps/rejected": -382.31573486328125, | |
| "loss": 0.644, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -1.7959085702896118, | |
| "rewards/margins": 0.18800289928913116, | |
| "rewards/rejected": -1.983911156654358, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 4.3924876391293915e-06, | |
| "logits/chosen": -1.972678542137146, | |
| "logits/rejected": -1.9295990467071533, | |
| "logps/chosen": -358.55126953125, | |
| "logps/rejected": -405.638427734375, | |
| "loss": 0.567, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -1.6957597732543945, | |
| "rewards/margins": 0.3802381753921509, | |
| "rewards/rejected": -2.075997829437256, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 4.346138351564711e-06, | |
| "logits/chosen": -2.031592607498169, | |
| "logits/rejected": -1.8613895177841187, | |
| "logps/chosen": -358.3297119140625, | |
| "logps/rejected": -360.80023193359375, | |
| "loss": 0.5674, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.6673065423965454, | |
| "rewards/margins": 0.44297677278518677, | |
| "rewards/rejected": -2.110283374786377, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 4.2983495008466285e-06, | |
| "logits/chosen": -2.104931116104126, | |
| "logits/rejected": -2.1332077980041504, | |
| "logps/chosen": -411.54168701171875, | |
| "logps/rejected": -463.42236328125, | |
| "loss": 0.666, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -1.9785181283950806, | |
| "rewards/margins": 0.3848969638347626, | |
| "rewards/rejected": -2.363415241241455, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 4.249158351283414e-06, | |
| "logits/chosen": -2.211599826812744, | |
| "logits/rejected": -1.9848130941390991, | |
| "logps/chosen": -430.388427734375, | |
| "logps/rejected": -470.9698181152344, | |
| "loss": 0.6333, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.7057955265045166, | |
| "rewards/margins": 0.3454364538192749, | |
| "rewards/rejected": -2.051231861114502, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 4.198603260653792e-06, | |
| "logits/chosen": -2.3403067588806152, | |
| "logits/rejected": -2.1165671348571777, | |
| "logps/chosen": -418.66558837890625, | |
| "logps/rejected": -407.5811462402344, | |
| "loss": 0.5813, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.5130822658538818, | |
| "rewards/margins": 0.13823339343070984, | |
| "rewards/rejected": -1.651315689086914, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 4.146723650296701e-06, | |
| "logits/chosen": -2.2038991451263428, | |
| "logits/rejected": -2.2611026763916016, | |
| "logps/chosen": -443.9969177246094, | |
| "logps/rejected": -509.6358337402344, | |
| "loss": 0.6094, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.4418891668319702, | |
| "rewards/margins": 0.5296822190284729, | |
| "rewards/rejected": -1.9715715646743774, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 4.093559974371725e-06, | |
| "logits/chosen": -2.236635684967041, | |
| "logits/rejected": -1.8922793865203857, | |
| "logps/chosen": -270.41839599609375, | |
| "logps/rejected": -274.03546142578125, | |
| "loss": 0.5395, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.8648042678833008, | |
| "rewards/margins": 0.4120641350746155, | |
| "rewards/rejected": -1.2768683433532715, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 4.039153688314146e-06, | |
| "logits/chosen": -2.2157764434814453, | |
| "logits/rejected": -2.01257586479187, | |
| "logps/chosen": -373.2973937988281, | |
| "logps/rejected": -401.8234558105469, | |
| "loss": 0.6029, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -1.1020699739456177, | |
| "rewards/margins": 0.6321985125541687, | |
| "rewards/rejected": -1.7342685461044312, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 3.983547216509254e-06, | |
| "logits/chosen": -2.215888261795044, | |
| "logits/rejected": -2.1177175045013428, | |
| "logps/chosen": -399.17547607421875, | |
| "logps/rejected": -416.84783935546875, | |
| "loss": 0.5846, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.7063343524932861, | |
| "rewards/margins": 0.4800395965576172, | |
| "rewards/rejected": -2.186373472213745, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 3.92678391921108e-06, | |
| "logits/chosen": -2.060939073562622, | |
| "logits/rejected": -1.9635168313980103, | |
| "logps/chosen": -346.0058898925781, | |
| "logps/rejected": -369.8831481933594, | |
| "loss": 0.6227, | |
| "rewards/accuracies": 0.4749999940395355, | |
| "rewards/chosen": -1.8079555034637451, | |
| "rewards/margins": 0.4140779376029968, | |
| "rewards/rejected": -2.2220332622528076, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 3.868908058731376e-06, | |
| "logits/chosen": -1.974831223487854, | |
| "logits/rejected": -1.8712928295135498, | |
| "logps/chosen": -399.43365478515625, | |
| "logps/rejected": -470.9222717285156, | |
| "loss": 0.5688, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -2.336700439453125, | |
| "rewards/margins": 0.5076223611831665, | |
| "rewards/rejected": -2.844322681427002, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 3.8099647649251984e-06, | |
| "logits/chosen": -2.060171127319336, | |
| "logits/rejected": -1.7853256464004517, | |
| "logps/chosen": -443.10223388671875, | |
| "logps/rejected": -441.1519470214844, | |
| "loss": 0.56, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -1.967047929763794, | |
| "rewards/margins": 0.48686590790748596, | |
| "rewards/rejected": -2.453913927078247, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 3.7500000000000005e-06, | |
| "logits/chosen": -2.0951271057128906, | |
| "logits/rejected": -2.0230164527893066, | |
| "logps/chosen": -356.6705627441406, | |
| "logps/rejected": -427.401611328125, | |
| "loss": 0.5599, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.6239030361175537, | |
| "rewards/margins": 0.6958804726600647, | |
| "rewards/rejected": -2.3197836875915527, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 3.689060522675689e-06, | |
| "logits/chosen": -2.1919496059417725, | |
| "logits/rejected": -2.014500141143799, | |
| "logps/chosen": -428.25701904296875, | |
| "logps/rejected": -452.41729736328125, | |
| "loss": 0.557, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.8657432794570923, | |
| "rewards/margins": 0.5891915559768677, | |
| "rewards/rejected": -2.454935073852539, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 3.627193851723577e-06, | |
| "logits/chosen": -2.0486769676208496, | |
| "logits/rejected": -1.9130027294158936, | |
| "logps/chosen": -471.4190979003906, | |
| "logps/rejected": -506.3304138183594, | |
| "loss": 0.5804, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -2.426065444946289, | |
| "rewards/margins": 0.6026751399040222, | |
| "rewards/rejected": -3.028740406036377, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 3.564448228912682e-06, | |
| "logits/chosen": -2.078453540802002, | |
| "logits/rejected": -2.038710117340088, | |
| "logps/chosen": -485.046875, | |
| "logps/rejected": -521.7474365234375, | |
| "loss": 0.5933, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -2.38494873046875, | |
| "rewards/margins": 0.25780850648880005, | |
| "rewards/rejected": -2.6427571773529053, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 3.5008725813922383e-06, | |
| "logits/chosen": -2.164614200592041, | |
| "logits/rejected": -1.8823134899139404, | |
| "logps/chosen": -410.06988525390625, | |
| "logps/rejected": -389.2319030761719, | |
| "loss": 0.5627, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -1.471362590789795, | |
| "rewards/margins": 0.5164188146591187, | |
| "rewards/rejected": -1.9877815246582031, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 3.436516483539781e-06, | |
| "logits/chosen": -2.1140494346618652, | |
| "logits/rejected": -1.9551842212677002, | |
| "logps/chosen": -301.4757995605469, | |
| "logps/rejected": -336.1775817871094, | |
| "loss": 0.551, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.9967424273490906, | |
| "rewards/margins": 0.5578289031982422, | |
| "rewards/rejected": -1.554571270942688, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 3.3714301183045382e-06, | |
| "logits/chosen": -2.2336459159851074, | |
| "logits/rejected": -2.313737630844116, | |
| "logps/chosen": -379.36846923828125, | |
| "logps/rejected": -424.39483642578125, | |
| "loss": 0.5247, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.8149202466011047, | |
| "rewards/margins": 0.6902198195457458, | |
| "rewards/rejected": -1.5051400661468506, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 3.3056642380762783e-06, | |
| "logits/chosen": -2.109157085418701, | |
| "logits/rejected": -2.1128740310668945, | |
| "logps/chosen": -370.62579345703125, | |
| "logps/rejected": -388.05670166015625, | |
| "loss": 0.5868, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -1.1603527069091797, | |
| "rewards/margins": 0.45087918639183044, | |
| "rewards/rejected": -1.611232042312622, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 3.2392701251101172e-06, | |
| "logits/chosen": -2.079946756362915, | |
| "logits/rejected": -1.9534591436386108, | |
| "logps/chosen": -384.3609313964844, | |
| "logps/rejected": -391.6958312988281, | |
| "loss": 0.5787, | |
| "rewards/accuracies": 0.4749999940395355, | |
| "rewards/chosen": -1.6679966449737549, | |
| "rewards/margins": 0.5511924028396606, | |
| "rewards/rejected": -2.219188928604126, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 3.1722995515381644e-06, | |
| "logits/chosen": -1.974525809288025, | |
| "logits/rejected": -1.8337838649749756, | |
| "logps/chosen": -373.17828369140625, | |
| "logps/rejected": -386.6225280761719, | |
| "loss": 0.5649, | |
| "rewards/accuracies": 0.4749999940395355, | |
| "rewards/chosen": -1.2219445705413818, | |
| "rewards/margins": 0.5704046487808228, | |
| "rewards/rejected": -1.7923492193222046, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 3.1048047389991693e-06, | |
| "logits/chosen": -2.0642426013946533, | |
| "logits/rejected": -1.8965177536010742, | |
| "logps/chosen": -323.34405517578125, | |
| "logps/rejected": -380.21380615234375, | |
| "loss": 0.5979, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.0704076290130615, | |
| "rewards/margins": 0.48811253905296326, | |
| "rewards/rejected": -1.5585200786590576, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 3.0368383179176584e-06, | |
| "logits/chosen": -2.033987283706665, | |
| "logits/rejected": -1.9596669673919678, | |
| "logps/chosen": -305.8987731933594, | |
| "logps/rejected": -335.6326904296875, | |
| "loss": 0.5467, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.6429412961006165, | |
| "rewards/margins": 0.666652500629425, | |
| "rewards/rejected": -1.309593677520752, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 2.9684532864643123e-06, | |
| "logits/chosen": -2.0718424320220947, | |
| "logits/rejected": -1.9492712020874023, | |
| "logps/chosen": -354.63372802734375, | |
| "logps/rejected": -367.26422119140625, | |
| "loss": 0.5716, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.022562026977539, | |
| "rewards/margins": 0.4337197244167328, | |
| "rewards/rejected": -1.4562817811965942, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 2.8997029692295875e-06, | |
| "logits/chosen": -2.2117860317230225, | |
| "logits/rejected": -2.0400938987731934, | |
| "logps/chosen": -381.23028564453125, | |
| "logps/rejected": -429.6178283691406, | |
| "loss": 0.531, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.1089333295822144, | |
| "rewards/margins": 0.7748517990112305, | |
| "rewards/rejected": -1.8837852478027344, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 2.8306409756428067e-06, | |
| "logits/chosen": -2.042997121810913, | |
| "logits/rejected": -1.6112009286880493, | |
| "logps/chosen": -487.6568908691406, | |
| "logps/rejected": -535.9304809570312, | |
| "loss": 0.516, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": -1.8848283290863037, | |
| "rewards/margins": 1.166489601135254, | |
| "rewards/rejected": -3.0513179302215576, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 2.761321158169134e-06, | |
| "logits/chosen": -2.1288044452667236, | |
| "logits/rejected": -1.8480589389801025, | |
| "logps/chosen": -591.8642578125, | |
| "logps/rejected": -573.7616577148438, | |
| "loss": 0.5375, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -3.170348644256592, | |
| "rewards/margins": 0.48750972747802734, | |
| "rewards/rejected": -3.6578586101531982, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 2.6917975703170466e-06, | |
| "logits/chosen": -1.834415078163147, | |
| "logits/rejected": -1.542644739151001, | |
| "logps/chosen": -506.2686462402344, | |
| "logps/rejected": -612.8573608398438, | |
| "loss": 0.5693, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -2.562810182571411, | |
| "rewards/margins": 0.9861109852790833, | |
| "rewards/rejected": -3.5489211082458496, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 2.6221244244890336e-06, | |
| "logits/chosen": -1.8896293640136719, | |
| "logits/rejected": -1.7028945684432983, | |
| "logps/chosen": -479.379638671875, | |
| "logps/rejected": -513.0144653320312, | |
| "loss": 0.5561, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -2.36388897895813, | |
| "rewards/margins": 0.6828749775886536, | |
| "rewards/rejected": -3.0467634201049805, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 2.5523560497083927e-06, | |
| "logits/chosen": -2.059697151184082, | |
| "logits/rejected": -1.8784849643707275, | |
| "logps/chosen": -337.88494873046875, | |
| "logps/rejected": -363.4068298339844, | |
| "loss": 0.6096, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -1.4112889766693115, | |
| "rewards/margins": 0.4129563868045807, | |
| "rewards/rejected": -1.8242454528808594, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 2.482546849255096e-06, | |
| "logits/chosen": -1.9830563068389893, | |
| "logits/rejected": -1.6811624765396118, | |
| "logps/chosen": -361.8495788574219, | |
| "logps/rejected": -382.3643493652344, | |
| "loss": 0.5242, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.473449468612671, | |
| "rewards/margins": 0.5172423720359802, | |
| "rewards/rejected": -1.990692138671875, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 2.4127512582437486e-06, | |
| "logits/chosen": -2.0440447330474854, | |
| "logits/rejected": -1.9845294952392578, | |
| "logps/chosen": -401.8277893066406, | |
| "logps/rejected": -451.3299255371094, | |
| "loss": 0.5288, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -1.1858041286468506, | |
| "rewards/margins": 0.7083054184913635, | |
| "rewards/rejected": -1.8941094875335693, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 2.3430237011767166e-06, | |
| "logits/chosen": -1.833554983139038, | |
| "logits/rejected": -1.6954238414764404, | |
| "logps/chosen": -353.65924072265625, | |
| "logps/rejected": -397.3764343261719, | |
| "loss": 0.531, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -1.3097889423370361, | |
| "rewards/margins": 0.7774924039840698, | |
| "rewards/rejected": -2.0872814655303955, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 2.2734185495055503e-06, | |
| "logits/chosen": -1.6967763900756836, | |
| "logits/rejected": -1.4636828899383545, | |
| "logps/chosen": -397.916748046875, | |
| "logps/rejected": -437.44073486328125, | |
| "loss": 0.4871, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -1.731325387954712, | |
| "rewards/margins": 0.9450262188911438, | |
| "rewards/rejected": -2.676351547241211, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 2.2039900792337477e-06, | |
| "logits/chosen": -1.6748664379119873, | |
| "logits/rejected": -1.4754512310028076, | |
| "logps/chosen": -475.0547790527344, | |
| "logps/rejected": -515.4742431640625, | |
| "loss": 0.5481, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -2.1851890087127686, | |
| "rewards/margins": 0.8229333758354187, | |
| "rewards/rejected": -3.008122205734253, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 2.134792428593971e-06, | |
| "logits/chosen": -1.8436079025268555, | |
| "logits/rejected": -1.525818109512329, | |
| "logps/chosen": -420.80810546875, | |
| "logps/rejected": -455.9149475097656, | |
| "loss": 0.5215, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -1.747849702835083, | |
| "rewards/margins": 0.8033768534660339, | |
| "rewards/rejected": -2.5512263774871826, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 2.0658795558326745e-06, | |
| "logits/chosen": -1.76202392578125, | |
| "logits/rejected": -1.825585126876831, | |
| "logps/chosen": -335.0538635253906, | |
| "logps/rejected": -410.6778259277344, | |
| "loss": 0.5755, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.3501343727111816, | |
| "rewards/margins": 0.7440234422683716, | |
| "rewards/rejected": -2.0941579341888428, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 1.997305197135089e-06, | |
| "logits/chosen": -1.8911097049713135, | |
| "logits/rejected": -1.7050600051879883, | |
| "logps/chosen": -344.654296875, | |
| "logps/rejected": -365.8127136230469, | |
| "loss": 0.5896, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -1.335740089416504, | |
| "rewards/margins": 0.36764174699783325, | |
| "rewards/rejected": -1.7033817768096924, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 1.9291228247233607e-06, | |
| "logits/chosen": -2.199026107788086, | |
| "logits/rejected": -1.8742659091949463, | |
| "logps/chosen": -418.6163024902344, | |
| "logps/rejected": -376.9792175292969, | |
| "loss": 0.5747, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -1.2824668884277344, | |
| "rewards/margins": 0.4385095536708832, | |
| "rewards/rejected": -1.72097647190094, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 1.8613856051605242e-06, | |
| "logits/chosen": -1.913412094116211, | |
| "logits/rejected": -1.6813814640045166, | |
| "logps/chosen": -404.4813232421875, | |
| "logps/rejected": -463.2132263183594, | |
| "loss": 0.5343, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -1.447469711303711, | |
| "rewards/margins": 0.8074652552604675, | |
| "rewards/rejected": -2.2549350261688232, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 1.7941463578928088e-06, | |
| "logits/chosen": -2.0417885780334473, | |
| "logits/rejected": -2.0130181312561035, | |
| "logps/chosen": -379.37066650390625, | |
| "logps/rejected": -394.95867919921875, | |
| "loss": 0.6045, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -1.6197001934051514, | |
| "rewards/margins": 0.29476988315582275, | |
| "rewards/rejected": -1.9144700765609741, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 1.7274575140626318e-06, | |
| "logits/chosen": -1.9285320043563843, | |
| "logits/rejected": -2.0787336826324463, | |
| "logps/chosen": -348.0567321777344, | |
| "logps/rejected": -397.8705139160156, | |
| "loss": 0.522, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.0203365087509155, | |
| "rewards/margins": 0.4765065312385559, | |
| "rewards/rejected": -1.4968432188034058, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 1.661371075624363e-06, | |
| "logits/chosen": -1.9909842014312744, | |
| "logits/rejected": -1.6936451196670532, | |
| "logps/chosen": -265.6180419921875, | |
| "logps/rejected": -281.07952880859375, | |
| "loss": 0.5591, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.7331510782241821, | |
| "rewards/margins": 0.5583804249763489, | |
| "rewards/rejected": -1.2915314435958862, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 1.5959385747947697e-06, | |
| "logits/chosen": -1.9582141637802124, | |
| "logits/rejected": -1.9302053451538086, | |
| "logps/chosen": -336.01959228515625, | |
| "logps/rejected": -410.764404296875, | |
| "loss": 0.4964, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.9309733510017395, | |
| "rewards/margins": 0.613142728805542, | |
| "rewards/rejected": -1.5441160202026367, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 1.5312110338697427e-06, | |
| "logits/chosen": -2.052699327468872, | |
| "logits/rejected": -1.9583276510238647, | |
| "logps/chosen": -289.5865478515625, | |
| "logps/rejected": -312.62725830078125, | |
| "loss": 0.5951, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -0.9973894357681274, | |
| "rewards/margins": 0.37033605575561523, | |
| "rewards/rejected": -1.3677256107330322, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 1.467238925438646e-06, | |
| "logits/chosen": -2.0705270767211914, | |
| "logits/rejected": -1.8725826740264893, | |
| "logps/chosen": -367.77471923828125, | |
| "logps/rejected": -411.2198181152344, | |
| "loss": 0.5386, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.9991909861564636, | |
| "rewards/margins": 0.571709394454956, | |
| "rewards/rejected": -1.5709004402160645, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 1.4040721330273063e-06, | |
| "logits/chosen": -1.7584121227264404, | |
| "logits/rejected": -1.4766103029251099, | |
| "logps/chosen": -463.2286071777344, | |
| "logps/rejected": -484.1163635253906, | |
| "loss": 0.5445, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.6703122854232788, | |
| "rewards/margins": 0.6411558389663696, | |
| "rewards/rejected": -2.3114681243896484, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 1.3417599122003464e-06, | |
| "logits/chosen": -2.1024575233459473, | |
| "logits/rejected": -1.6672672033309937, | |
| "logps/chosen": -391.9578857421875, | |
| "logps/rejected": -422.2921447753906, | |
| "loss": 0.4638, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.2933290004730225, | |
| "rewards/margins": 0.9474877119064331, | |
| "rewards/rejected": -2.240816831588745, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 1.280350852153168e-06, | |
| "logits/chosen": -1.8131592273712158, | |
| "logits/rejected": -1.9417848587036133, | |
| "logps/chosen": -319.12799072265625, | |
| "logps/rejected": -448.4112854003906, | |
| "loss": 0.5363, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.285538911819458, | |
| "rewards/margins": 0.8607152104377747, | |
| "rewards/rejected": -2.146254062652588, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 1.2198928378235717e-06, | |
| "logits/chosen": -1.8834339380264282, | |
| "logits/rejected": -1.7587229013442993, | |
| "logps/chosen": -424.78826904296875, | |
| "logps/rejected": -522.1199340820312, | |
| "loss": 0.5754, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.7636245489120483, | |
| "rewards/margins": 0.8004873991012573, | |
| "rewards/rejected": -2.5641119480133057, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 1.160433012552508e-06, | |
| "logits/chosen": -2.0634958744049072, | |
| "logits/rejected": -2.003753423690796, | |
| "logps/chosen": -335.7600402832031, | |
| "logps/rejected": -375.54852294921875, | |
| "loss": 0.5125, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.9807103872299194, | |
| "rewards/margins": 0.9115110635757446, | |
| "rewards/rejected": -1.892221450805664, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 1.1020177413231334e-06, | |
| "logits/chosen": -2.016789197921753, | |
| "logits/rejected": -1.8195407390594482, | |
| "logps/chosen": -353.8262939453125, | |
| "logps/rejected": -411.1983947753906, | |
| "loss": 0.5023, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.2073825597763062, | |
| "rewards/margins": 0.8694759607315063, | |
| "rewards/rejected": -2.0768585205078125, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 1.0446925746067768e-06, | |
| "logits/chosen": -1.8060710430145264, | |
| "logits/rejected": -1.5303637981414795, | |
| "logps/chosen": -329.4360046386719, | |
| "logps/rejected": -366.3887939453125, | |
| "loss": 0.5347, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.0586979389190674, | |
| "rewards/margins": 0.9634588956832886, | |
| "rewards/rejected": -2.0221567153930664, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 9.88502212844063e-07, | |
| "logits/chosen": -1.9487718343734741, | |
| "logits/rejected": -1.8278251886367798, | |
| "logps/chosen": -424.44256591796875, | |
| "logps/rejected": -527.6069946289062, | |
| "loss": 0.5597, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.5092089176177979, | |
| "rewards/margins": 1.0597232580184937, | |
| "rewards/rejected": -2.568932056427002, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 9.334904715888496e-07, | |
| "logits/chosen": -1.9447358846664429, | |
| "logits/rejected": -1.672228455543518, | |
| "logps/chosen": -440.62774658203125, | |
| "logps/rejected": -548.8284912109375, | |
| "loss": 0.5634, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.6289888620376587, | |
| "rewards/margins": 1.100258469581604, | |
| "rewards/rejected": -2.7292473316192627, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 8.797002473421729e-07, | |
| "logits/chosen": -1.8913949728012085, | |
| "logits/rejected": -1.8936278820037842, | |
| "logps/chosen": -384.93609619140625, | |
| "logps/rejected": -455.0386657714844, | |
| "loss": 0.5326, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.3333874940872192, | |
| "rewards/margins": 0.7129821181297302, | |
| "rewards/rejected": -2.046369791030884, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 8.271734841028553e-07, | |
| "logits/chosen": -1.8880865573883057, | |
| "logits/rejected": -1.6740175485610962, | |
| "logps/chosen": -414.0435485839844, | |
| "logps/rejected": -461.56158447265625, | |
| "loss": 0.4911, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.596765398979187, | |
| "rewards/margins": 0.815388560295105, | |
| "rewards/rejected": -2.412154197692871, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 7.759511406608255e-07, | |
| "logits/chosen": -2.010457754135132, | |
| "logits/rejected": -1.8959916830062866, | |
| "logps/chosen": -383.6888122558594, | |
| "logps/rejected": -452.77105712890625, | |
| "loss": 0.5708, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -1.487220048904419, | |
| "rewards/margins": 0.7800397276878357, | |
| "rewards/rejected": -2.2672595977783203, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 7.260731586586983e-07, | |
| "logits/chosen": -1.7655563354492188, | |
| "logits/rejected": -1.8372739553451538, | |
| "logps/chosen": -259.60394287109375, | |
| "logps/rejected": -331.3665466308594, | |
| "loss": 0.541, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -1.3245407342910767, | |
| "rewards/margins": 0.7111746072769165, | |
| "rewards/rejected": -2.0357155799865723, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 6.775784314464717e-07, | |
| "logits/chosen": -2.1715354919433594, | |
| "logits/rejected": -1.837615728378296, | |
| "logps/chosen": -360.0257873535156, | |
| "logps/rejected": -413.18902587890625, | |
| "loss": 0.4932, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.2116737365722656, | |
| "rewards/margins": 1.0079423189163208, | |
| "rewards/rejected": -2.219616413116455, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 6.305047737536707e-07, | |
| "logits/chosen": -1.8406097888946533, | |
| "logits/rejected": -1.8857877254486084, | |
| "logps/chosen": -375.0958557128906, | |
| "logps/rejected": -486.08197021484375, | |
| "loss": 0.5066, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.7462670803070068, | |
| "rewards/margins": 0.950217604637146, | |
| "rewards/rejected": -2.6964848041534424, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 5.848888922025553e-07, | |
| "logits/chosen": -1.8275880813598633, | |
| "logits/rejected": -2.047853946685791, | |
| "logps/chosen": -441.58966064453125, | |
| "logps/rejected": -461.3450622558594, | |
| "loss": 0.5374, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.767677664756775, | |
| "rewards/margins": 0.5157124996185303, | |
| "rewards/rejected": -2.283390522003174, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 5.407663566854008e-07, | |
| "logits/chosen": -2.1351914405822754, | |
| "logits/rejected": -1.8889188766479492, | |
| "logps/chosen": -394.17120361328125, | |
| "logps/rejected": -436.7420959472656, | |
| "loss": 0.5683, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.3438371419906616, | |
| "rewards/margins": 0.9155893325805664, | |
| "rewards/rejected": -2.2594265937805176, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 4.981715726281666e-07, | |
| "logits/chosen": -2.017271041870117, | |
| "logits/rejected": -1.981236219406128, | |
| "logps/chosen": -441.42254638671875, | |
| "logps/rejected": -580.200927734375, | |
| "loss": 0.4928, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": -1.5878021717071533, | |
| "rewards/margins": 1.3070558309555054, | |
| "rewards/rejected": -2.8948581218719482, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 4.5713775416217884e-07, | |
| "logits/chosen": -1.6851351261138916, | |
| "logits/rejected": -1.7757434844970703, | |
| "logps/chosen": -348.5665588378906, | |
| "logps/rejected": -457.87542724609375, | |
| "loss": 0.5279, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.5111078023910522, | |
| "rewards/margins": 1.0286327600479126, | |
| "rewards/rejected": -2.5397403240203857, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 4.1769689822475147e-07, | |
| "logits/chosen": -1.674856424331665, | |
| "logits/rejected": -1.817155122756958, | |
| "logps/chosen": -318.9717712402344, | |
| "logps/rejected": -434.81475830078125, | |
| "loss": 0.5817, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -1.6043727397918701, | |
| "rewards/margins": 0.6249055862426758, | |
| "rewards/rejected": -2.229278087615967, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 3.798797596089351e-07, | |
| "logits/chosen": -1.953508973121643, | |
| "logits/rejected": -1.7095199823379517, | |
| "logps/chosen": -383.12371826171875, | |
| "logps/rejected": -451.828125, | |
| "loss": 0.5006, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -1.5027226209640503, | |
| "rewards/margins": 0.9227194786071777, | |
| "rewards/rejected": -2.4254419803619385, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 3.4371582698185636e-07, | |
| "logits/chosen": -2.011037826538086, | |
| "logits/rejected": -1.858045220375061, | |
| "logps/chosen": -455.5492248535156, | |
| "logps/rejected": -558.2069091796875, | |
| "loss": 0.4797, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.7510541677474976, | |
| "rewards/margins": 0.8790451884269714, | |
| "rewards/rejected": -2.6300997734069824, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 3.092332998903416e-07, | |
| "logits/chosen": -1.9804046154022217, | |
| "logits/rejected": -2.04115891456604, | |
| "logps/chosen": -395.303466796875, | |
| "logps/rejected": -492.44329833984375, | |
| "loss": 0.5034, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.4553591012954712, | |
| "rewards/margins": 0.8879169225692749, | |
| "rewards/rejected": -2.343275785446167, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 2.764590667717562e-07, | |
| "logits/chosen": -1.8032455444335938, | |
| "logits/rejected": -1.563175916671753, | |
| "logps/chosen": -345.4132385253906, | |
| "logps/rejected": -335.91949462890625, | |
| "loss": 0.5746, | |
| "rewards/accuracies": 0.44999998807907104, | |
| "rewards/chosen": -1.697852373123169, | |
| "rewards/margins": 0.29076820611953735, | |
| "rewards/rejected": -1.988620400428772, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 2.454186839872158e-07, | |
| "logits/chosen": -1.882430076599121, | |
| "logits/rejected": -1.7509571313858032, | |
| "logps/chosen": -354.9794616699219, | |
| "logps/rejected": -450.55560302734375, | |
| "loss": 0.4963, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -1.5439207553863525, | |
| "rewards/margins": 1.0496200323104858, | |
| "rewards/rejected": -2.5935404300689697, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 2.1613635589349756e-07, | |
| "logits/chosen": -1.9673278331756592, | |
| "logits/rejected": -1.735865592956543, | |
| "logps/chosen": -361.7403869628906, | |
| "logps/rejected": -393.9466247558594, | |
| "loss": 0.5859, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -1.3523650169372559, | |
| "rewards/margins": 0.5490896105766296, | |
| "rewards/rejected": -1.9014545679092407, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "learning_rate": 1.8863491596921745e-07, | |
| "logits/chosen": -1.9650977849960327, | |
| "logits/rejected": -1.674034833908081, | |
| "logps/chosen": -398.31817626953125, | |
| "logps/rejected": -418.9993591308594, | |
| "loss": 0.4818, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.6353886127471924, | |
| "rewards/margins": 0.8128952980041504, | |
| "rewards/rejected": -2.4482839107513428, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 1.629358090099639e-07, | |
| "logits/chosen": -1.8651268482208252, | |
| "logits/rejected": -1.7213342189788818, | |
| "logps/chosen": -395.92608642578125, | |
| "logps/rejected": -427.53082275390625, | |
| "loss": 0.5434, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.8581531047821045, | |
| "rewards/margins": 0.6955922842025757, | |
| "rewards/rejected": -2.5537455081939697, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 1.3905907440629752e-07, | |
| "logits/chosen": -1.8218196630477905, | |
| "logits/rejected": -1.603205680847168, | |
| "logps/chosen": -371.84820556640625, | |
| "logps/rejected": -423.8062438964844, | |
| "loss": 0.5957, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -1.6403701305389404, | |
| "rewards/margins": 0.8258684873580933, | |
| "rewards/rejected": -2.466238498687744, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 1.1702333051763271e-07, | |
| "logits/chosen": -1.7774673700332642, | |
| "logits/rejected": -1.5289005041122437, | |
| "logps/chosen": -440.49407958984375, | |
| "logps/rejected": -493.42486572265625, | |
| "loss": 0.5159, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.8031593561172485, | |
| "rewards/margins": 0.9965063333511353, | |
| "rewards/rejected": -2.799665927886963, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 9.684576015420277e-08, | |
| "logits/chosen": -1.7342450618743896, | |
| "logits/rejected": -1.5554310083389282, | |
| "logps/chosen": -407.7359313964844, | |
| "logps/rejected": -490.3565979003906, | |
| "loss": 0.4897, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -1.6332868337631226, | |
| "rewards/margins": 0.9035712480545044, | |
| "rewards/rejected": -2.536858081817627, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 7.854209717842231e-08, | |
| "logits/chosen": -1.728171944618225, | |
| "logits/rejected": -1.6524664163589478, | |
| "logps/chosen": -420.43768310546875, | |
| "logps/rejected": -411.3157653808594, | |
| "loss": 0.6007, | |
| "rewards/accuracies": 0.4749999940395355, | |
| "rewards/chosen": -1.8913564682006836, | |
| "rewards/margins": 0.21912887692451477, | |
| "rewards/rejected": -2.110485553741455, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 6.212661423609184e-08, | |
| "logits/chosen": -1.7257133722305298, | |
| "logits/rejected": -1.6627498865127563, | |
| "logps/chosen": -435.08599853515625, | |
| "logps/rejected": -524.0604248046875, | |
| "loss": 0.6031, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -2.2094435691833496, | |
| "rewards/margins": 0.7494264841079712, | |
| "rewards/rejected": -2.9588701725006104, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 4.761211162702117e-08, | |
| "logits/chosen": -1.8422664403915405, | |
| "logits/rejected": -1.7962735891342163, | |
| "logps/chosen": -370.86566162109375, | |
| "logps/rejected": -408.62042236328125, | |
| "loss": 0.5591, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.4711788892745972, | |
| "rewards/margins": 0.5541782379150391, | |
| "rewards/rejected": -2.025357246398926, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 3.5009907323737826e-08, | |
| "logits/chosen": -1.8273169994354248, | |
| "logits/rejected": -1.549336314201355, | |
| "logps/chosen": -408.63140869140625, | |
| "logps/rejected": -491.55303955078125, | |
| "loss": 0.5183, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -1.7581342458724976, | |
| "rewards/margins": 0.9740850329399109, | |
| "rewards/rejected": -2.7322194576263428, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 2.4329828146074096e-08, | |
| "logits/chosen": -2.1215853691101074, | |
| "logits/rejected": -1.810217261314392, | |
| "logps/chosen": -454.5079040527344, | |
| "logps/rejected": -459.13134765625, | |
| "loss": 0.5396, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.8914794921875, | |
| "rewards/margins": 0.6684591174125671, | |
| "rewards/rejected": -2.559938430786133, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "learning_rate": 1.5580202098509078e-08, | |
| "logits/chosen": -1.9334113597869873, | |
| "logits/rejected": -1.882664680480957, | |
| "logps/chosen": -512.0015869140625, | |
| "logps/rejected": -529.5744018554688, | |
| "loss": 0.5535, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.7347371578216553, | |
| "rewards/margins": 0.7627021074295044, | |
| "rewards/rejected": -2.49743914604187, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 8.767851876239075e-09, | |
| "logits/chosen": -1.6884361505508423, | |
| "logits/rejected": -1.633782148361206, | |
| "logps/chosen": -397.6700134277344, | |
| "logps/rejected": -514.9879150390625, | |
| "loss": 0.5168, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -1.720780611038208, | |
| "rewards/margins": 1.0960266590118408, | |
| "rewards/rejected": -2.816807270050049, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 3.8980895450474455e-09, | |
| "logits/chosen": -1.887984037399292, | |
| "logits/rejected": -1.7204885482788086, | |
| "logps/chosen": -388.229248046875, | |
| "logps/rejected": -438.8857421875, | |
| "loss": 0.5719, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.6957340240478516, | |
| "rewards/margins": 0.7062219381332397, | |
| "rewards/rejected": -2.4019558429718018, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 9.747123991141193e-10, | |
| "logits/chosen": -1.995326042175293, | |
| "logits/rejected": -1.8855457305908203, | |
| "logps/chosen": -367.4112243652344, | |
| "logps/rejected": -460.2139587402344, | |
| "loss": 0.5104, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -1.427626132965088, | |
| "rewards/margins": 1.0081008672714233, | |
| "rewards/rejected": -2.4357268810272217, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 0.0, | |
| "logits/chosen": -1.856300950050354, | |
| "logits/rejected": -1.691680908203125, | |
| "logps/chosen": -407.53253173828125, | |
| "logps/rejected": -473.7496643066406, | |
| "loss": 0.4906, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.637367606163025, | |
| "rewards/margins": 0.8503096699714661, | |
| "rewards/rejected": -2.4876773357391357, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 1250, | |
| "total_flos": 0.0, | |
| "train_loss": 0.37276097068786623, | |
| "train_runtime": 8917.3793, | |
| "train_samples_per_second": 1.682, | |
| "train_steps_per_second": 0.14 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1250, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 20, | |
| "total_flos": 0.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |