| { | |
| "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": -3.1403026580810547, | |
| "logits/rejected": -2.4295239448547363, | |
| "logps/chosen": -392.2955017089844, | |
| "logps/rejected": -613.5740966796875, | |
| "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": -2.442929267883301, | |
| "logits/rejected": -2.321220636367798, | |
| "logps/chosen": -261.98870849609375, | |
| "logps/rejected": -221.4100799560547, | |
| "loss": 0.6931, | |
| "rewards/accuracies": 0.4722222089767456, | |
| "rewards/chosen": 7.1744441811461e-05, | |
| "rewards/margins": 5.5226653785211965e-05, | |
| "rewards/rejected": 1.6517798940185457e-05, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 8.000000000000001e-07, | |
| "logits/chosen": -2.42722225189209, | |
| "logits/rejected": -2.285809278488159, | |
| "logps/chosen": -244.7586212158203, | |
| "logps/rejected": -254.1322784423828, | |
| "loss": 0.6931, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -4.93793522764463e-05, | |
| "rewards/margins": 0.00013055391900707036, | |
| "rewards/rejected": -0.00017993326764553785, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 1.2000000000000002e-06, | |
| "logits/chosen": -2.468625068664551, | |
| "logits/rejected": -2.3896145820617676, | |
| "logps/chosen": -251.3707733154297, | |
| "logps/rejected": -300.4630126953125, | |
| "loss": 0.6931, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": 0.00025913942954503, | |
| "rewards/margins": 9.763217531144619e-05, | |
| "rewards/rejected": 0.00016150726878549904, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 1.6000000000000001e-06, | |
| "logits/chosen": -2.483097553253174, | |
| "logits/rejected": -2.135282516479492, | |
| "logps/chosen": -252.2518310546875, | |
| "logps/rejected": -206.49740600585938, | |
| "loss": 0.693, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": 0.0007439021137543023, | |
| "rewards/margins": 0.0003426038019824773, | |
| "rewards/rejected": 0.00040129836997948587, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "logits/chosen": -2.6477389335632324, | |
| "logits/rejected": -2.486539363861084, | |
| "logps/chosen": -245.1098175048828, | |
| "logps/rejected": -234.5915985107422, | |
| "loss": 0.6928, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": 0.0010015963343903422, | |
| "rewards/margins": 0.00047645685845054686, | |
| "rewards/rejected": 0.0005251394468359649, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 2.4000000000000003e-06, | |
| "logits/chosen": -2.5867533683776855, | |
| "logits/rejected": -1.8461745977401733, | |
| "logps/chosen": -288.46343994140625, | |
| "logps/rejected": -191.20741271972656, | |
| "loss": 0.6924, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": 0.0025676097720861435, | |
| "rewards/margins": 0.001473221811465919, | |
| "rewards/rejected": 0.0010943879606202245, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 2.8000000000000003e-06, | |
| "logits/chosen": -2.4274649620056152, | |
| "logits/rejected": -2.21500825881958, | |
| "logps/chosen": -206.21188354492188, | |
| "logps/rejected": -188.66172790527344, | |
| "loss": 0.6918, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": 0.004373534582555294, | |
| "rewards/margins": 0.0019022986525669694, | |
| "rewards/rejected": 0.0024712360464036465, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 3.2000000000000003e-06, | |
| "logits/chosen": -2.272698163986206, | |
| "logits/rejected": -2.2475826740264893, | |
| "logps/chosen": -218.1708984375, | |
| "logps/rejected": -247.9695281982422, | |
| "loss": 0.6919, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": 0.00605297926813364, | |
| "rewards/margins": 0.002590904710814357, | |
| "rewards/rejected": 0.0034620738588273525, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 3.6000000000000003e-06, | |
| "logits/chosen": -2.5878186225891113, | |
| "logits/rejected": -2.174198627471924, | |
| "logps/chosen": -284.7269287109375, | |
| "logps/rejected": -222.1537322998047, | |
| "loss": 0.6908, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": 0.00458324421197176, | |
| "rewards/margins": 0.005124131683260202, | |
| "rewards/rejected": -0.0005408872966654599, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 4.000000000000001e-06, | |
| "logits/chosen": -2.4876809120178223, | |
| "logits/rejected": -2.304762363433838, | |
| "logps/chosen": -258.6788330078125, | |
| "logps/rejected": -230.6868133544922, | |
| "loss": 0.6878, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": 0.0009092213585972786, | |
| "rewards/margins": 0.013424595817923546, | |
| "rewards/rejected": -0.012515375390648842, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 4.4e-06, | |
| "logits/chosen": -2.24761700630188, | |
| "logits/rejected": -2.0283989906311035, | |
| "logps/chosen": -242.88485717773438, | |
| "logps/rejected": -253.8578643798828, | |
| "loss": 0.6832, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.02561577782034874, | |
| "rewards/margins": 0.02318538725376129, | |
| "rewards/rejected": -0.04880116507411003, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 4.800000000000001e-06, | |
| "logits/chosen": -2.0988376140594482, | |
| "logits/rejected": -2.193089008331299, | |
| "logps/chosen": -242.8556365966797, | |
| "logps/rejected": -317.71917724609375, | |
| "loss": 0.6753, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.047444917261600494, | |
| "rewards/margins": 0.03924765810370445, | |
| "rewards/rejected": -0.08669257164001465, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 4.999756310023261e-06, | |
| "logits/chosen": -2.241037607192993, | |
| "logits/rejected": -2.157954692840576, | |
| "logps/chosen": -331.423828125, | |
| "logps/rejected": -348.91046142578125, | |
| "loss": 0.6688, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.08836604654788971, | |
| "rewards/margins": 0.0334697850048542, | |
| "rewards/rejected": -0.12183582782745361, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 4.997807075247147e-06, | |
| "logits/chosen": -1.7352321147918701, | |
| "logits/rejected": -1.5876859426498413, | |
| "logps/chosen": -398.1210021972656, | |
| "logps/rejected": -428.30328369140625, | |
| "loss": 0.6519, | |
| "rewards/accuracies": 0.44999998807907104, | |
| "rewards/chosen": -0.22779341042041779, | |
| "rewards/margins": 0.0445188470184803, | |
| "rewards/rejected": -0.2723122239112854, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 4.993910125649561e-06, | |
| "logits/chosen": -1.735691785812378, | |
| "logits/rejected": -1.7642030715942383, | |
| "logps/chosen": -525.7499389648438, | |
| "logps/rejected": -695.8289184570312, | |
| "loss": 0.6354, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.30290618538856506, | |
| "rewards/margins": 0.11975699663162231, | |
| "rewards/rejected": -0.422663152217865, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 4.988068499954578e-06, | |
| "logits/chosen": -2.004002332687378, | |
| "logits/rejected": -1.7766345739364624, | |
| "logps/chosen": -464.7952575683594, | |
| "logps/rejected": -630.8657836914062, | |
| "loss": 0.6461, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.23786354064941406, | |
| "rewards/margins": 0.18317463994026184, | |
| "rewards/rejected": -0.4210381507873535, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 4.980286753286196e-06, | |
| "logits/chosen": -2.0650768280029297, | |
| "logits/rejected": -1.572537899017334, | |
| "logps/chosen": -591.57421875, | |
| "logps/rejected": -780.8646240234375, | |
| "loss": 0.607, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.2891489565372467, | |
| "rewards/margins": 0.27657198905944824, | |
| "rewards/rejected": -0.5657209157943726, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 4.970570953616383e-06, | |
| "logits/chosen": -1.4938254356384277, | |
| "logits/rejected": -1.661007285118103, | |
| "logps/chosen": -764.5491943359375, | |
| "logps/rejected": -1000.9505615234375, | |
| "loss": 0.6384, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.5677521228790283, | |
| "rewards/margins": 0.19688071310520172, | |
| "rewards/rejected": -0.7646328806877136, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 4.958928677033465e-06, | |
| "logits/chosen": -1.6295101642608643, | |
| "logits/rejected": -1.3617414236068726, | |
| "logps/chosen": -720.4210815429688, | |
| "logps/rejected": -925.54638671875, | |
| "loss": 0.6042, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.4372948706150055, | |
| "rewards/margins": 0.2413189709186554, | |
| "rewards/rejected": -0.6786138415336609, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 4.9453690018345144e-06, | |
| "logits/chosen": -1.502726674079895, | |
| "logits/rejected": -1.3334269523620605, | |
| "logps/chosen": -752.0069580078125, | |
| "logps/rejected": -1068.231201171875, | |
| "loss": 0.64, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.5011454820632935, | |
| "rewards/margins": 0.2965936064720154, | |
| "rewards/rejected": -0.7977390289306641, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 4.9299025014463665e-06, | |
| "logits/chosen": -1.1884076595306396, | |
| "logits/rejected": -1.0619341135025024, | |
| "logps/chosen": -890.1341552734375, | |
| "logps/rejected": -1238.742431640625, | |
| "loss": 0.6007, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.6316002607345581, | |
| "rewards/margins": 0.3692142069339752, | |
| "rewards/rejected": -1.000814437866211, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 4.912541236180779e-06, | |
| "logits/chosen": -1.0146244764328003, | |
| "logits/rejected": -0.9316056966781616, | |
| "logps/chosen": -979.2394409179688, | |
| "logps/rejected": -1465.3270263671875, | |
| "loss": 0.6018, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.7414819002151489, | |
| "rewards/margins": 0.49276018142700195, | |
| "rewards/rejected": -1.2342422008514404, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 4.893298743830168e-06, | |
| "logits/chosen": -1.2160202264785767, | |
| "logits/rejected": -1.0113346576690674, | |
| "logps/chosen": -1194.5838623046875, | |
| "logps/rejected": -1710.5823974609375, | |
| "loss": 0.5866, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.9506030082702637, | |
| "rewards/margins": 0.5406419634819031, | |
| "rewards/rejected": -1.491244912147522, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 4.8721900291112415e-06, | |
| "logits/chosen": -1.7798312902450562, | |
| "logits/rejected": -1.6643425226211548, | |
| "logps/chosen": -1057.5977783203125, | |
| "logps/rejected": -1005.8741455078125, | |
| "loss": 0.694, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.7734619379043579, | |
| "rewards/margins": -0.0247958954423666, | |
| "rewards/rejected": -0.7486661076545715, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 4.849231551964771e-06, | |
| "logits/chosen": -2.0089643001556396, | |
| "logits/rejected": -1.7859022617340088, | |
| "logps/chosen": -399.1302795410156, | |
| "logps/rejected": -577.6984252929688, | |
| "loss": 0.6424, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.1754738688468933, | |
| "rewards/margins": 0.13275322318077087, | |
| "rewards/rejected": -0.3082270920276642, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 4.824441214720629e-06, | |
| "logits/chosen": -1.7344192266464233, | |
| "logits/rejected": -1.632328748703003, | |
| "logps/chosen": -632.4493408203125, | |
| "logps/rejected": -760.2273559570312, | |
| "loss": 0.6252, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.41359105706214905, | |
| "rewards/margins": 0.12629616260528564, | |
| "rewards/rejected": -0.5398872494697571, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 4.7978383481380865e-06, | |
| "logits/chosen": -1.5548350811004639, | |
| "logits/rejected": -1.5831079483032227, | |
| "logps/chosen": -598.45556640625, | |
| "logps/rejected": -712.7095336914062, | |
| "loss": 0.6096, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.3854503631591797, | |
| "rewards/margins": 0.13576054573059082, | |
| "rewards/rejected": -0.5212109088897705, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 4.769443696332272e-06, | |
| "logits/chosen": -2.0036838054656982, | |
| "logits/rejected": -1.7330118417739868, | |
| "logps/chosen": -426.128662109375, | |
| "logps/rejected": -617.08984375, | |
| "loss": 0.6208, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.2103346288204193, | |
| "rewards/margins": 0.19718368351459503, | |
| "rewards/rejected": -0.40751829743385315, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 4.7392794005985324e-06, | |
| "logits/chosen": -1.8511890172958374, | |
| "logits/rejected": -1.5459794998168945, | |
| "logps/chosen": -737.9730224609375, | |
| "logps/rejected": -961.6851806640625, | |
| "loss": 0.6214, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.48743873834609985, | |
| "rewards/margins": 0.20717349648475647, | |
| "rewards/rejected": -0.6946121454238892, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 4.707368982147318e-06, | |
| "logits/chosen": -1.7424571514129639, | |
| "logits/rejected": -1.4730459451675415, | |
| "logps/chosen": -833.6031494140625, | |
| "logps/rejected": -950.7183837890625, | |
| "loss": 0.5805, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.5092605352401733, | |
| "rewards/margins": 0.1874401867389679, | |
| "rewards/rejected": -0.6967006921768188, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 4.673737323763048e-06, | |
| "logits/chosen": -1.373366355895996, | |
| "logits/rejected": -1.2923896312713623, | |
| "logps/chosen": -975.0471801757812, | |
| "logps/rejected": -1144.96533203125, | |
| "loss": 0.6091, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.7069741487503052, | |
| "rewards/margins": 0.21132870018482208, | |
| "rewards/rejected": -0.9183028340339661, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 4.638410650401267e-06, | |
| "logits/chosen": -1.477074384689331, | |
| "logits/rejected": -1.2639570236206055, | |
| "logps/chosen": -1050.070556640625, | |
| "logps/rejected": -1336.6949462890625, | |
| "loss": 0.5927, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.8159879446029663, | |
| "rewards/margins": 0.2922418415546417, | |
| "rewards/rejected": -1.1082298755645752, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 4.601416508739211e-06, | |
| "logits/chosen": -1.5806219577789307, | |
| "logits/rejected": -1.5797499418258667, | |
| "logps/chosen": -1069.904296875, | |
| "logps/rejected": -1329.1131591796875, | |
| "loss": 0.6329, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.8489059209823608, | |
| "rewards/margins": 0.2538915276527405, | |
| "rewards/rejected": -1.102797508239746, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 4.562783745695738e-06, | |
| "logits/chosen": -1.578185796737671, | |
| "logits/rejected": -1.3778866529464722, | |
| "logps/chosen": -898.2579956054688, | |
| "logps/rejected": -1295.5048828125, | |
| "loss": 0.5816, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.6616091728210449, | |
| "rewards/margins": 0.3867344856262207, | |
| "rewards/rejected": -1.0483437776565552, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 4.522542485937369e-06, | |
| "logits/chosen": -1.3196460008621216, | |
| "logits/rejected": -1.0029346942901611, | |
| "logps/chosen": -1030.9923095703125, | |
| "logps/rejected": -1271.06494140625, | |
| "loss": 0.6062, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.7833272218704224, | |
| "rewards/margins": 0.2811380922794342, | |
| "rewards/rejected": -1.0644652843475342, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 4.4807241083879774e-06, | |
| "logits/chosen": -1.4315626621246338, | |
| "logits/rejected": -1.1289265155792236, | |
| "logps/chosen": -1395.56640625, | |
| "logps/rejected": -2238.73681640625, | |
| "loss": 0.5686, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.1033650636672974, | |
| "rewards/margins": 0.8617695569992065, | |
| "rewards/rejected": -1.965134620666504, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 4.437361221760449e-06, | |
| "logits/chosen": -1.0882983207702637, | |
| "logits/rejected": -0.9719398617744446, | |
| "logps/chosen": -1259.433349609375, | |
| "logps/rejected": -1656.2109375, | |
| "loss": 0.5973, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.069411277770996, | |
| "rewards/margins": 0.3754344880580902, | |
| "rewards/rejected": -1.4448457956314087, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 4.3924876391293915e-06, | |
| "logits/chosen": -0.9243530035018921, | |
| "logits/rejected": -0.9347097277641296, | |
| "logps/chosen": -1276.580810546875, | |
| "logps/rejected": -1660.425048828125, | |
| "loss": 0.5932, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.1309508085250854, | |
| "rewards/margins": 0.35707855224609375, | |
| "rewards/rejected": -1.4880293607711792, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 4.346138351564711e-06, | |
| "logits/chosen": -1.2673766613006592, | |
| "logits/rejected": -1.1589628458023071, | |
| "logps/chosen": -1206.547607421875, | |
| "logps/rejected": -1615.82568359375, | |
| "loss": 0.5588, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.9714127779006958, | |
| "rewards/margins": 0.4280439019203186, | |
| "rewards/rejected": -1.39945650100708, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 4.2983495008466285e-06, | |
| "logits/chosen": -1.5774317979812622, | |
| "logits/rejected": -1.2660281658172607, | |
| "logps/chosen": -1082.2581787109375, | |
| "logps/rejected": -1570.060791015625, | |
| "loss": 0.5484, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.8231463432312012, | |
| "rewards/margins": 0.5096359848976135, | |
| "rewards/rejected": -1.3327823877334595, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 4.249158351283414e-06, | |
| "logits/chosen": -1.4447729587554932, | |
| "logits/rejected": -1.1427133083343506, | |
| "logps/chosen": -1174.7425537109375, | |
| "logps/rejected": -1466.5479736328125, | |
| "loss": 0.5432, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.9159573316574097, | |
| "rewards/margins": 0.33625325560569763, | |
| "rewards/rejected": -1.2522106170654297, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 4.198603260653792e-06, | |
| "logits/chosen": -0.7784031629562378, | |
| "logits/rejected": -0.5429641008377075, | |
| "logps/chosen": -2375.2294921875, | |
| "logps/rejected": -2860.4833984375, | |
| "loss": 0.5023, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -2.1455845832824707, | |
| "rewards/margins": 0.4987269341945648, | |
| "rewards/rejected": -2.6443114280700684, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 4.146723650296701e-06, | |
| "logits/chosen": -0.8310529589653015, | |
| "logits/rejected": -0.47391629219055176, | |
| "logps/chosen": -3060.8125, | |
| "logps/rejected": -3744.44140625, | |
| "loss": 0.6076, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -2.840507984161377, | |
| "rewards/margins": 0.725513219833374, | |
| "rewards/rejected": -3.56602144241333, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 4.093559974371725e-06, | |
| "logits/chosen": -1.8024406433105469, | |
| "logits/rejected": -1.4066709280014038, | |
| "logps/chosen": -1206.559326171875, | |
| "logps/rejected": -1958.577392578125, | |
| "loss": 0.5517, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -0.948940098285675, | |
| "rewards/margins": 0.76356440782547, | |
| "rewards/rejected": -1.7125043869018555, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 4.039153688314146e-06, | |
| "logits/chosen": -1.7638174295425415, | |
| "logits/rejected": -1.6592317819595337, | |
| "logps/chosen": -888.7286376953125, | |
| "logps/rejected": -1274.7813720703125, | |
| "loss": 0.5681, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.6799342036247253, | |
| "rewards/margins": 0.4119555950164795, | |
| "rewards/rejected": -1.0918899774551392, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 3.983547216509254e-06, | |
| "logits/chosen": -1.8817472457885742, | |
| "logits/rejected": -1.766122817993164, | |
| "logps/chosen": -1327.4205322265625, | |
| "logps/rejected": -1799.326904296875, | |
| "loss": 0.5546, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.9942048192024231, | |
| "rewards/margins": 0.49855145812034607, | |
| "rewards/rejected": -1.4927562475204468, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 3.92678391921108e-06, | |
| "logits/chosen": -1.842795968055725, | |
| "logits/rejected": -1.6158469915390015, | |
| "logps/chosen": -1413.307373046875, | |
| "logps/rejected": -1754.1558837890625, | |
| "loss": 0.6049, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.0906217098236084, | |
| "rewards/margins": 0.3714645802974701, | |
| "rewards/rejected": -1.4620864391326904, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 3.868908058731376e-06, | |
| "logits/chosen": -1.6520286798477173, | |
| "logits/rejected": -1.5457327365875244, | |
| "logps/chosen": -1587.302490234375, | |
| "logps/rejected": -1965.464599609375, | |
| "loss": 0.5766, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.3615858554840088, | |
| "rewards/margins": 0.39060306549072266, | |
| "rewards/rejected": -1.7521886825561523, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 3.8099647649251984e-06, | |
| "logits/chosen": -1.8340473175048828, | |
| "logits/rejected": -1.6425994634628296, | |
| "logps/chosen": -1667.3792724609375, | |
| "logps/rejected": -2022.701904296875, | |
| "loss": 0.5243, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.4147528409957886, | |
| "rewards/margins": 0.4026387333869934, | |
| "rewards/rejected": -1.8173916339874268, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 3.7500000000000005e-06, | |
| "logits/chosen": -1.565735936164856, | |
| "logits/rejected": -1.4348723888397217, | |
| "logps/chosen": -1869.2935791015625, | |
| "logps/rejected": -2379.262939453125, | |
| "loss": 0.5181, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.658975601196289, | |
| "rewards/margins": 0.5309830904006958, | |
| "rewards/rejected": -2.1899585723876953, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 3.689060522675689e-06, | |
| "logits/chosen": -1.8067477941513062, | |
| "logits/rejected": -1.4334272146224976, | |
| "logps/chosen": -1798.0201416015625, | |
| "logps/rejected": -2005.1083984375, | |
| "loss": 0.5859, | |
| "rewards/accuracies": 0.4749999940395355, | |
| "rewards/chosen": -1.5768072605133057, | |
| "rewards/margins": 0.24011941254138947, | |
| "rewards/rejected": -1.8169265985488892, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 3.627193851723577e-06, | |
| "logits/chosen": -1.9058891534805298, | |
| "logits/rejected": -1.5708621740341187, | |
| "logps/chosen": -1344.331787109375, | |
| "logps/rejected": -2066.19677734375, | |
| "loss": 0.5302, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -1.0564631223678589, | |
| "rewards/margins": 0.7303067445755005, | |
| "rewards/rejected": -1.7867698669433594, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 3.564448228912682e-06, | |
| "logits/chosen": -1.8703407049179077, | |
| "logits/rejected": -1.8959243297576904, | |
| "logps/chosen": -1418.962890625, | |
| "logps/rejected": -1918.7298583984375, | |
| "loss": 0.5437, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.2395734786987305, | |
| "rewards/margins": 0.4607962965965271, | |
| "rewards/rejected": -1.7003698348999023, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 3.5008725813922383e-06, | |
| "logits/chosen": -1.4778034687042236, | |
| "logits/rejected": -1.42378830909729, | |
| "logps/chosen": -2291.814453125, | |
| "logps/rejected": -2982.93212890625, | |
| "loss": 0.4967, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -2.091310977935791, | |
| "rewards/margins": 0.6848040819168091, | |
| "rewards/rejected": -2.7761149406433105, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 3.436516483539781e-06, | |
| "logits/chosen": -1.495721459388733, | |
| "logits/rejected": -1.364970326423645, | |
| "logps/chosen": -2159.25634765625, | |
| "logps/rejected": -2941.31982421875, | |
| "loss": 0.5244, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.9415247440338135, | |
| "rewards/margins": 0.7436372637748718, | |
| "rewards/rejected": -2.685161828994751, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 3.3714301183045382e-06, | |
| "logits/chosen": -1.7252031564712524, | |
| "logits/rejected": -1.3431731462478638, | |
| "logps/chosen": -2346.15673828125, | |
| "logps/rejected": -2880.50341796875, | |
| "loss": 0.5818, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -2.072269916534424, | |
| "rewards/margins": 0.5818148851394653, | |
| "rewards/rejected": -2.6540849208831787, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 3.3056642380762783e-06, | |
| "logits/chosen": -1.7475759983062744, | |
| "logits/rejected": -1.6887413263320923, | |
| "logps/chosen": -2365.16650390625, | |
| "logps/rejected": -3232.527587890625, | |
| "loss": 0.4838, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -2.112750291824341, | |
| "rewards/margins": 0.815180778503418, | |
| "rewards/rejected": -2.927931070327759, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 3.2392701251101172e-06, | |
| "logits/chosen": -1.6227779388427734, | |
| "logits/rejected": -1.371058702468872, | |
| "logps/chosen": -2093.24365234375, | |
| "logps/rejected": -2805.071533203125, | |
| "loss": 0.5206, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -1.8952558040618896, | |
| "rewards/margins": 0.6814407706260681, | |
| "rewards/rejected": -2.5766968727111816, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 3.1722995515381644e-06, | |
| "logits/chosen": -1.6335232257843018, | |
| "logits/rejected": -1.2477694749832153, | |
| "logps/chosen": -2436.821044921875, | |
| "logps/rejected": -3250.9990234375, | |
| "loss": 0.5081, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -2.2077012062072754, | |
| "rewards/margins": 0.8523558378219604, | |
| "rewards/rejected": -3.0600569248199463, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 3.1048047389991693e-06, | |
| "logits/chosen": -1.898289442062378, | |
| "logits/rejected": -1.5713775157928467, | |
| "logps/chosen": -1884.5660400390625, | |
| "logps/rejected": -2746.454833984375, | |
| "loss": 0.5623, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.6871427297592163, | |
| "rewards/margins": 0.8711158633232117, | |
| "rewards/rejected": -2.5582587718963623, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 3.0368383179176584e-06, | |
| "logits/chosen": -1.4297007322311401, | |
| "logits/rejected": -1.3948471546173096, | |
| "logps/chosen": -2035.6175537109375, | |
| "logps/rejected": -3012.0087890625, | |
| "loss": 0.5395, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -1.860030174255371, | |
| "rewards/margins": 0.9378849267959595, | |
| "rewards/rejected": -2.797914981842041, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 2.9684532864643123e-06, | |
| "logits/chosen": -1.7777831554412842, | |
| "logits/rejected": -1.3341039419174194, | |
| "logps/chosen": -2097.790283203125, | |
| "logps/rejected": -2818.270263671875, | |
| "loss": 0.4943, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -1.8172571659088135, | |
| "rewards/margins": 0.7645599246025085, | |
| "rewards/rejected": -2.581817150115967, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 2.8997029692295875e-06, | |
| "logits/chosen": -1.7123359441757202, | |
| "logits/rejected": -1.4930963516235352, | |
| "logps/chosen": -1850.494873046875, | |
| "logps/rejected": -2223.23779296875, | |
| "loss": 0.5415, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.5890159606933594, | |
| "rewards/margins": 0.4220260679721832, | |
| "rewards/rejected": -2.0110421180725098, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 2.8306409756428067e-06, | |
| "logits/chosen": -1.7620747089385986, | |
| "logits/rejected": -1.273874282836914, | |
| "logps/chosen": -1751.422119140625, | |
| "logps/rejected": -2349.684814453125, | |
| "loss": 0.5267, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -1.5236599445343018, | |
| "rewards/margins": 0.6361467838287354, | |
| "rewards/rejected": -2.159806728363037, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 2.761321158169134e-06, | |
| "logits/chosen": -1.8131462335586548, | |
| "logits/rejected": -1.548789381980896, | |
| "logps/chosen": -1939.831298828125, | |
| "logps/rejected": -2459.304443359375, | |
| "loss": 0.5526, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.7154333591461182, | |
| "rewards/margins": 0.5310146808624268, | |
| "rewards/rejected": -2.246448040008545, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 2.6917975703170466e-06, | |
| "logits/chosen": -1.4764631986618042, | |
| "logits/rejected": -1.2985259294509888, | |
| "logps/chosen": -2074.831787109375, | |
| "logps/rejected": -2988.404296875, | |
| "loss": 0.5177, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.8811687231063843, | |
| "rewards/margins": 0.8929063677787781, | |
| "rewards/rejected": -2.7740752696990967, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 2.6221244244890336e-06, | |
| "logits/chosen": -1.606785774230957, | |
| "logits/rejected": -1.4506292343139648, | |
| "logps/chosen": -2944.038330078125, | |
| "logps/rejected": -3770.83203125, | |
| "loss": 0.5202, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -2.648715019226074, | |
| "rewards/margins": 0.8418272733688354, | |
| "rewards/rejected": -3.4905428886413574, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 2.5523560497083927e-06, | |
| "logits/chosen": -1.7594387531280518, | |
| "logits/rejected": -1.4435147047042847, | |
| "logps/chosen": -2691.794921875, | |
| "logps/rejected": -2907.418212890625, | |
| "loss": 0.6347, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -2.4359123706817627, | |
| "rewards/margins": 0.3157985210418701, | |
| "rewards/rejected": -2.751710891723633, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 2.482546849255096e-06, | |
| "logits/chosen": -1.6512796878814697, | |
| "logits/rejected": -1.2861438989639282, | |
| "logps/chosen": -2340.228515625, | |
| "logps/rejected": -3065.31591796875, | |
| "loss": 0.4919, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": -2.0771026611328125, | |
| "rewards/margins": 0.6874440908432007, | |
| "rewards/rejected": -2.7645466327667236, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 2.4127512582437486e-06, | |
| "logits/chosen": -1.7903190851211548, | |
| "logits/rejected": -1.5637561082839966, | |
| "logps/chosen": -1961.098876953125, | |
| "logps/rejected": -2661.05126953125, | |
| "loss": 0.5352, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -1.727064847946167, | |
| "rewards/margins": 0.7038751840591431, | |
| "rewards/rejected": -2.4309401512145996, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 2.3430237011767166e-06, | |
| "logits/chosen": -1.5963449478149414, | |
| "logits/rejected": -1.5752345323562622, | |
| "logps/chosen": -2412.08447265625, | |
| "logps/rejected": -2860.287109375, | |
| "loss": 0.5364, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -2.1989083290100098, | |
| "rewards/margins": 0.4225144386291504, | |
| "rewards/rejected": -2.62142276763916, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 2.2734185495055503e-06, | |
| "logits/chosen": -1.581274390220642, | |
| "logits/rejected": -1.338945746421814, | |
| "logps/chosen": -2689.8212890625, | |
| "logps/rejected": -3143.3369140625, | |
| "loss": 0.5331, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -2.4156579971313477, | |
| "rewards/margins": 0.4918283522129059, | |
| "rewards/rejected": -2.9074864387512207, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 2.2039900792337477e-06, | |
| "logits/chosen": -1.6093494892120361, | |
| "logits/rejected": -1.2763807773590088, | |
| "logps/chosen": -1861.92578125, | |
| "logps/rejected": -2657.68212890625, | |
| "loss": 0.4712, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -1.6057929992675781, | |
| "rewards/margins": 0.847273051738739, | |
| "rewards/rejected": -2.453066110610962, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 2.134792428593971e-06, | |
| "logits/chosen": -1.3526313304901123, | |
| "logits/rejected": -1.3224549293518066, | |
| "logps/chosen": -2195.27197265625, | |
| "logps/rejected": -2851.3291015625, | |
| "loss": 0.5536, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -2.023761749267578, | |
| "rewards/margins": 0.6121652722358704, | |
| "rewards/rejected": -2.635927200317383, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 2.0658795558326745e-06, | |
| "logits/chosen": -1.6000654697418213, | |
| "logits/rejected": -1.4151827096939087, | |
| "logps/chosen": -2148.247802734375, | |
| "logps/rejected": -2729.99853515625, | |
| "loss": 0.5222, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.903270959854126, | |
| "rewards/margins": 0.5996490716934204, | |
| "rewards/rejected": -2.5029196739196777, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 1.997305197135089e-06, | |
| "logits/chosen": -1.8795032501220703, | |
| "logits/rejected": -1.8359801769256592, | |
| "logps/chosen": -2000.689208984375, | |
| "logps/rejected": -2926.07275390625, | |
| "loss": 0.5271, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -1.7533314228057861, | |
| "rewards/margins": 0.9173868298530579, | |
| "rewards/rejected": -2.6707184314727783, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 1.9291228247233607e-06, | |
| "logits/chosen": -1.6307891607284546, | |
| "logits/rejected": -1.496146321296692, | |
| "logps/chosen": -1922.004150390625, | |
| "logps/rejected": -2854.74365234375, | |
| "loss": 0.4909, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -1.7167818546295166, | |
| "rewards/margins": 0.9143074750900269, | |
| "rewards/rejected": -2.631089448928833, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 1.8613856051605242e-06, | |
| "logits/chosen": -1.5307934284210205, | |
| "logits/rejected": -1.4985663890838623, | |
| "logps/chosen": -1941.616455078125, | |
| "logps/rejected": -2720.62109375, | |
| "loss": 0.495, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.7477505207061768, | |
| "rewards/margins": 0.7529212236404419, | |
| "rewards/rejected": -2.500671863555908, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 1.7941463578928088e-06, | |
| "logits/chosen": -1.6022170782089233, | |
| "logits/rejected": -1.5732767581939697, | |
| "logps/chosen": -2065.362548828125, | |
| "logps/rejected": -2565.924560546875, | |
| "loss": 0.5311, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.8574340343475342, | |
| "rewards/margins": 0.4975149631500244, | |
| "rewards/rejected": -2.3549489974975586, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 1.7274575140626318e-06, | |
| "logits/chosen": -1.5523189306259155, | |
| "logits/rejected": -1.2900121212005615, | |
| "logps/chosen": -2249.78466796875, | |
| "logps/rejected": -3071.79541015625, | |
| "loss": 0.5322, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -2.0411088466644287, | |
| "rewards/margins": 0.8322113752365112, | |
| "rewards/rejected": -2.8733203411102295, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 1.661371075624363e-06, | |
| "logits/chosen": -1.750195860862732, | |
| "logits/rejected": -1.4256199598312378, | |
| "logps/chosen": -2680.12548828125, | |
| "logps/rejected": -3339.82177734375, | |
| "loss": 0.5582, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -2.413327693939209, | |
| "rewards/margins": 0.6952977180480957, | |
| "rewards/rejected": -3.108625650405884, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 1.5959385747947697e-06, | |
| "logits/chosen": -1.6673352718353271, | |
| "logits/rejected": -1.4309087991714478, | |
| "logps/chosen": -2934.502685546875, | |
| "logps/rejected": -3557.366455078125, | |
| "loss": 0.5826, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -2.6931636333465576, | |
| "rewards/margins": 0.6061044931411743, | |
| "rewards/rejected": -3.2992682456970215, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 1.5312110338697427e-06, | |
| "logits/chosen": -1.4265161752700806, | |
| "logits/rejected": -1.3840219974517822, | |
| "logps/chosen": -2902.10205078125, | |
| "logps/rejected": -3211.46484375, | |
| "loss": 0.5793, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -2.7132956981658936, | |
| "rewards/margins": 0.2958894670009613, | |
| "rewards/rejected": -3.009185314178467, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 1.467238925438646e-06, | |
| "logits/chosen": -1.93195378780365, | |
| "logits/rejected": -1.7828657627105713, | |
| "logps/chosen": -2659.49609375, | |
| "logps/rejected": -3121.871826171875, | |
| "loss": 0.5751, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -2.427194118499756, | |
| "rewards/margins": 0.47958964109420776, | |
| "rewards/rejected": -2.9067838191986084, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 1.4040721330273063e-06, | |
| "logits/chosen": -1.6783872842788696, | |
| "logits/rejected": -1.3621320724487305, | |
| "logps/chosen": -2476.935791015625, | |
| "logps/rejected": -3240.056640625, | |
| "loss": 0.5022, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -2.256714344024658, | |
| "rewards/margins": 0.7921364307403564, | |
| "rewards/rejected": -3.0488505363464355, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 1.3417599122003464e-06, | |
| "logits/chosen": -1.631744146347046, | |
| "logits/rejected": -1.4881222248077393, | |
| "logps/chosen": -2611.767333984375, | |
| "logps/rejected": -3342.803466796875, | |
| "loss": 0.5239, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -2.361072301864624, | |
| "rewards/margins": 0.6957711577415466, | |
| "rewards/rejected": -3.0568435192108154, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 1.280350852153168e-06, | |
| "logits/chosen": -1.6476237773895264, | |
| "logits/rejected": -1.2135014533996582, | |
| "logps/chosen": -2210.89697265625, | |
| "logps/rejected": -3153.556884765625, | |
| "loss": 0.49, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": -1.9529176950454712, | |
| "rewards/margins": 0.9891131520271301, | |
| "rewards/rejected": -2.942031145095825, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 1.2198928378235717e-06, | |
| "logits/chosen": -1.8185428380966187, | |
| "logits/rejected": -1.4744179248809814, | |
| "logps/chosen": -1982.8447265625, | |
| "logps/rejected": -2986.81591796875, | |
| "loss": 0.4908, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.720696210861206, | |
| "rewards/margins": 1.0247818231582642, | |
| "rewards/rejected": -2.7454781532287598, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 1.160433012552508e-06, | |
| "logits/chosen": -1.9028065204620361, | |
| "logits/rejected": -1.5971095561981201, | |
| "logps/chosen": -1988.5980224609375, | |
| "logps/rejected": -2843.60302734375, | |
| "loss": 0.5298, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.695246934890747, | |
| "rewards/margins": 0.9005265235900879, | |
| "rewards/rejected": -2.595773458480835, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 1.1020177413231334e-06, | |
| "logits/chosen": -1.7520854473114014, | |
| "logits/rejected": -1.4778826236724854, | |
| "logps/chosen": -1867.814453125, | |
| "logps/rejected": -2722.72900390625, | |
| "loss": 0.5195, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.6399002075195312, | |
| "rewards/margins": 0.8625130653381348, | |
| "rewards/rejected": -2.502413272857666, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 1.0446925746067768e-06, | |
| "logits/chosen": -1.6180340051651, | |
| "logits/rejected": -1.4504057168960571, | |
| "logps/chosen": -2156.244140625, | |
| "logps/rejected": -3364.479736328125, | |
| "loss": 0.4724, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -1.9639689922332764, | |
| "rewards/margins": 1.1746752262115479, | |
| "rewards/rejected": -3.138643980026245, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 9.88502212844063e-07, | |
| "logits/chosen": -1.8366491794586182, | |
| "logits/rejected": -1.353752851486206, | |
| "logps/chosen": -1940.987060546875, | |
| "logps/rejected": -2697.123779296875, | |
| "loss": 0.506, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.714422583580017, | |
| "rewards/margins": 0.7562412023544312, | |
| "rewards/rejected": -2.4706637859344482, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 9.334904715888496e-07, | |
| "logits/chosen": -1.8136298656463623, | |
| "logits/rejected": -1.5429127216339111, | |
| "logps/chosen": -2025.2021484375, | |
| "logps/rejected": -2842.939453125, | |
| "loss": 0.5274, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.792414903640747, | |
| "rewards/margins": 0.8433243036270142, | |
| "rewards/rejected": -2.63573956489563, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 8.797002473421729e-07, | |
| "logits/chosen": -1.6221929788589478, | |
| "logits/rejected": -1.536675214767456, | |
| "logps/chosen": -2179.556884765625, | |
| "logps/rejected": -2724.71142578125, | |
| "loss": 0.5293, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -1.8998562097549438, | |
| "rewards/margins": 0.5943135023117065, | |
| "rewards/rejected": -2.4941699504852295, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 8.271734841028553e-07, | |
| "logits/chosen": -1.599360704421997, | |
| "logits/rejected": -1.4945180416107178, | |
| "logps/chosen": -2247.677001953125, | |
| "logps/rejected": -2832.541015625, | |
| "loss": 0.5645, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -2.041861057281494, | |
| "rewards/margins": 0.5804768800735474, | |
| "rewards/rejected": -2.622338056564331, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 7.759511406608255e-07, | |
| "logits/chosen": -1.9042466878890991, | |
| "logits/rejected": -1.5334478616714478, | |
| "logps/chosen": -1946.2919921875, | |
| "logps/rejected": -2741.82861328125, | |
| "loss": 0.466, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.713287115097046, | |
| "rewards/margins": 0.8301644325256348, | |
| "rewards/rejected": -2.5434517860412598, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 7.260731586586983e-07, | |
| "logits/chosen": -1.5018055438995361, | |
| "logits/rejected": -1.4937880039215088, | |
| "logps/chosen": -2611.099609375, | |
| "logps/rejected": -3151.30810546875, | |
| "loss": 0.5775, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -2.3841938972473145, | |
| "rewards/margins": 0.5148266553878784, | |
| "rewards/rejected": -2.8990206718444824, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 6.775784314464717e-07, | |
| "logits/chosen": -1.6764729022979736, | |
| "logits/rejected": -1.5057140588760376, | |
| "logps/chosen": -2298.930908203125, | |
| "logps/rejected": -2712.99169921875, | |
| "loss": 0.556, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -2.0648088455200195, | |
| "rewards/margins": 0.4065318703651428, | |
| "rewards/rejected": -2.4713406562805176, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 6.305047737536707e-07, | |
| "logits/chosen": -1.7027795314788818, | |
| "logits/rejected": -1.3025496006011963, | |
| "logps/chosen": -2075.1337890625, | |
| "logps/rejected": -2931.57080078125, | |
| "loss": 0.4584, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.846689224243164, | |
| "rewards/margins": 0.8770244717597961, | |
| "rewards/rejected": -2.7237138748168945, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 5.848888922025553e-07, | |
| "logits/chosen": -1.8110439777374268, | |
| "logits/rejected": -1.4326590299606323, | |
| "logps/chosen": -2325.86474609375, | |
| "logps/rejected": -3089.902099609375, | |
| "loss": 0.5249, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -2.067896604537964, | |
| "rewards/margins": 0.7900792360305786, | |
| "rewards/rejected": -2.857975482940674, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 5.407663566854008e-07, | |
| "logits/chosen": -1.6088895797729492, | |
| "logits/rejected": -1.4155261516571045, | |
| "logps/chosen": -2324.50927734375, | |
| "logps/rejected": -3341.44970703125, | |
| "loss": 0.525, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -2.1127965450286865, | |
| "rewards/margins": 0.9637571573257446, | |
| "rewards/rejected": -3.0765535831451416, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 4.981715726281666e-07, | |
| "logits/chosen": -1.6132185459136963, | |
| "logits/rejected": -1.2942755222320557, | |
| "logps/chosen": -2351.848388671875, | |
| "logps/rejected": -3291.189453125, | |
| "loss": 0.5183, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -2.1332767009735107, | |
| "rewards/margins": 0.9448688626289368, | |
| "rewards/rejected": -3.0781455039978027, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 4.5713775416217884e-07, | |
| "logits/chosen": -1.770582914352417, | |
| "logits/rejected": -1.515246868133545, | |
| "logps/chosen": -2241.62109375, | |
| "logps/rejected": -3231.28125, | |
| "loss": 0.4636, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": -1.9664819240570068, | |
| "rewards/margins": 1.0160348415374756, | |
| "rewards/rejected": -2.9825170040130615, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 4.1769689822475147e-07, | |
| "logits/chosen": -1.5879560708999634, | |
| "logits/rejected": -1.5036201477050781, | |
| "logps/chosen": -1936.926025390625, | |
| "logps/rejected": -2708.7578125, | |
| "loss": 0.4947, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.7738593816757202, | |
| "rewards/margins": 0.7560992240905762, | |
| "rewards/rejected": -2.529958486557007, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 3.798797596089351e-07, | |
| "logits/chosen": -1.4568753242492676, | |
| "logits/rejected": -1.477905511856079, | |
| "logps/chosen": -2428.654296875, | |
| "logps/rejected": -3266.05712890625, | |
| "loss": 0.5201, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -2.2049851417541504, | |
| "rewards/margins": 0.8197803497314453, | |
| "rewards/rejected": -3.0247654914855957, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 3.4371582698185636e-07, | |
| "logits/chosen": -1.6677932739257812, | |
| "logits/rejected": -1.5246306657791138, | |
| "logps/chosen": -2556.70751953125, | |
| "logps/rejected": -3194.56884765625, | |
| "loss": 0.5391, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -2.3062329292297363, | |
| "rewards/margins": 0.6159250140190125, | |
| "rewards/rejected": -2.9221577644348145, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 3.092332998903416e-07, | |
| "logits/chosen": -1.7090551853179932, | |
| "logits/rejected": -1.4194791316986084, | |
| "logps/chosen": -2150.77392578125, | |
| "logps/rejected": -2783.882568359375, | |
| "loss": 0.5195, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.9338016510009766, | |
| "rewards/margins": 0.6355168223381042, | |
| "rewards/rejected": -2.5693182945251465, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 2.764590667717562e-07, | |
| "logits/chosen": -1.606188178062439, | |
| "logits/rejected": -1.4986306428909302, | |
| "logps/chosen": -2124.823974609375, | |
| "logps/rejected": -2895.18310546875, | |
| "loss": 0.4628, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.912660837173462, | |
| "rewards/margins": 0.7604308724403381, | |
| "rewards/rejected": -2.673091411590576, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 2.454186839872158e-07, | |
| "logits/chosen": -1.5546796321868896, | |
| "logits/rejected": -1.550550103187561, | |
| "logps/chosen": -2096.05078125, | |
| "logps/rejected": -2596.4248046875, | |
| "loss": 0.5839, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.9206135272979736, | |
| "rewards/margins": 0.4739322066307068, | |
| "rewards/rejected": -2.394545793533325, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 2.1613635589349756e-07, | |
| "logits/chosen": -1.657184362411499, | |
| "logits/rejected": -1.3406727313995361, | |
| "logps/chosen": -2227.37841796875, | |
| "logps/rejected": -3146.83349609375, | |
| "loss": 0.459, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -1.9884710311889648, | |
| "rewards/margins": 0.9205359220504761, | |
| "rewards/rejected": -2.9090065956115723, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "learning_rate": 1.8863491596921745e-07, | |
| "logits/chosen": -1.4567536115646362, | |
| "logits/rejected": -1.1964176893234253, | |
| "logps/chosen": -2488.02099609375, | |
| "logps/rejected": -3093.317626953125, | |
| "loss": 0.5213, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -2.2580440044403076, | |
| "rewards/margins": 0.6173666715621948, | |
| "rewards/rejected": -2.875410556793213, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 1.629358090099639e-07, | |
| "logits/chosen": -1.8676036596298218, | |
| "logits/rejected": -1.5324196815490723, | |
| "logps/chosen": -2476.497314453125, | |
| "logps/rejected": -3140.97607421875, | |
| "loss": 0.5292, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -2.2105040550231934, | |
| "rewards/margins": 0.6873103380203247, | |
| "rewards/rejected": -2.8978142738342285, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 1.3905907440629752e-07, | |
| "logits/chosen": -1.5062824487686157, | |
| "logits/rejected": -1.7191336154937744, | |
| "logps/chosen": -2101.674560546875, | |
| "logps/rejected": -2914.34228515625, | |
| "loss": 0.4889, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.8925731182098389, | |
| "rewards/margins": 0.7410818338394165, | |
| "rewards/rejected": -2.633654832839966, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 1.1702333051763271e-07, | |
| "logits/chosen": -1.885053277015686, | |
| "logits/rejected": -1.7132413387298584, | |
| "logps/chosen": -2271.51806640625, | |
| "logps/rejected": -3019.444580078125, | |
| "loss": 0.4746, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -2.0059704780578613, | |
| "rewards/margins": 0.8035289645195007, | |
| "rewards/rejected": -2.809499740600586, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 9.684576015420277e-08, | |
| "logits/chosen": -1.7674312591552734, | |
| "logits/rejected": -1.7866573333740234, | |
| "logps/chosen": -2304.083740234375, | |
| "logps/rejected": -3042.02783203125, | |
| "loss": 0.5235, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -2.0314338207244873, | |
| "rewards/margins": 0.7121332883834839, | |
| "rewards/rejected": -2.7435669898986816, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 7.854209717842231e-08, | |
| "logits/chosen": -1.6722157001495361, | |
| "logits/rejected": -1.738372564315796, | |
| "logps/chosen": -2533.139404296875, | |
| "logps/rejected": -3169.639892578125, | |
| "loss": 0.508, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -2.2765398025512695, | |
| "rewards/margins": 0.6401499509811401, | |
| "rewards/rejected": -2.91668963432312, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 6.212661423609184e-08, | |
| "logits/chosen": -1.6138194799423218, | |
| "logits/rejected": -1.4168260097503662, | |
| "logps/chosen": -2651.40087890625, | |
| "logps/rejected": -2953.78564453125, | |
| "loss": 0.5389, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -2.369499683380127, | |
| "rewards/margins": 0.3613986372947693, | |
| "rewards/rejected": -2.730898380279541, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 4.761211162702117e-08, | |
| "logits/chosen": -1.8268349170684814, | |
| "logits/rejected": -1.389704942703247, | |
| "logps/chosen": -2308.206298828125, | |
| "logps/rejected": -3084.591796875, | |
| "loss": 0.4561, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -2.0564706325531006, | |
| "rewards/margins": 0.8305153846740723, | |
| "rewards/rejected": -2.8869857788085938, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 3.5009907323737826e-08, | |
| "logits/chosen": -1.5044516324996948, | |
| "logits/rejected": -1.3144376277923584, | |
| "logps/chosen": -2185.82275390625, | |
| "logps/rejected": -3079.09130859375, | |
| "loss": 0.4552, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -1.9693256616592407, | |
| "rewards/margins": 0.8803497552871704, | |
| "rewards/rejected": -2.8496756553649902, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 2.4329828146074096e-08, | |
| "logits/chosen": -1.5807933807373047, | |
| "logits/rejected": -1.4424628019332886, | |
| "logps/chosen": -2456.582763671875, | |
| "logps/rejected": -2916.781982421875, | |
| "loss": 0.5377, | |
| "rewards/accuracies": 0.4749999940395355, | |
| "rewards/chosen": -2.223078489303589, | |
| "rewards/margins": 0.4909304976463318, | |
| "rewards/rejected": -2.7140090465545654, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "learning_rate": 1.5580202098509078e-08, | |
| "logits/chosen": -1.4222395420074463, | |
| "logits/rejected": -1.4579776525497437, | |
| "logps/chosen": -2549.04052734375, | |
| "logps/rejected": -3383.83203125, | |
| "loss": 0.4934, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -2.3147170543670654, | |
| "rewards/margins": 0.7796028852462769, | |
| "rewards/rejected": -3.094320058822632, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 8.767851876239075e-09, | |
| "logits/chosen": -1.9085347652435303, | |
| "logits/rejected": -1.5333627462387085, | |
| "logps/chosen": -2555.34228515625, | |
| "logps/rejected": -3125.24853515625, | |
| "loss": 0.5644, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -2.2714717388153076, | |
| "rewards/margins": 0.6149500012397766, | |
| "rewards/rejected": -2.8864216804504395, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 3.8980895450474455e-09, | |
| "logits/chosen": -1.6516891717910767, | |
| "logits/rejected": -1.3506808280944824, | |
| "logps/chosen": -2481.84228515625, | |
| "logps/rejected": -3454.458984375, | |
| "loss": 0.4893, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -2.2125697135925293, | |
| "rewards/margins": 1.0183699131011963, | |
| "rewards/rejected": -3.2309398651123047, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 9.747123991141193e-10, | |
| "logits/chosen": -1.2762700319290161, | |
| "logits/rejected": -1.1082122325897217, | |
| "logps/chosen": -2387.552490234375, | |
| "logps/rejected": -3240.255615234375, | |
| "loss": 0.4634, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -2.163398265838623, | |
| "rewards/margins": 0.8387178182601929, | |
| "rewards/rejected": -3.0021164417266846, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 0.0, | |
| "logits/chosen": -1.7946316003799438, | |
| "logits/rejected": -1.654803991317749, | |
| "logps/chosen": -2349.287109375, | |
| "logps/rejected": -2943.7216796875, | |
| "loss": 0.547, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -2.1279029846191406, | |
| "rewards/margins": 0.613613486289978, | |
| "rewards/rejected": -2.74151611328125, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 1250, | |
| "total_flos": 0.0, | |
| "train_loss": 0.5607085178375244, | |
| "train_runtime": 13233.1545, | |
| "train_samples_per_second": 1.134, | |
| "train_steps_per_second": 0.094 | |
| } | |
| ], | |
| "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 | |
| } | |