| { | |
| "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": 0.06502792239189148, | |
| "logits/rejected": 0.2657509744167328, | |
| "logps/chosen": -767.5213012695312, | |
| "logps/rejected": -534.5206298828125, | |
| "loss": 1.8815, | |
| "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": 0.1509028524160385, | |
| "logits/rejected": 0.21932204067707062, | |
| "logps/chosen": -511.6697998046875, | |
| "logps/rejected": -560.0117797851562, | |
| "loss": 2.1678, | |
| "rewards/accuracies": 0.4236111044883728, | |
| "rewards/chosen": 0.00018176721641793847, | |
| "rewards/margins": 0.00014634148101322353, | |
| "rewards/rejected": 3.5425720852799714e-05, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 8.000000000000001e-07, | |
| "logits/chosen": 0.2015257328748703, | |
| "logits/rejected": 0.23093406856060028, | |
| "logps/chosen": -614.9334106445312, | |
| "logps/rejected": -580.0980224609375, | |
| "loss": 2.1481, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": 0.0002653285046108067, | |
| "rewards/margins": 9.045841579791158e-05, | |
| "rewards/rejected": 0.00017487008881289512, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 1.2000000000000002e-06, | |
| "logits/chosen": 0.15271301567554474, | |
| "logits/rejected": 0.20967456698417664, | |
| "logps/chosen": -615.6820068359375, | |
| "logps/rejected": -616.4979248046875, | |
| "loss": 2.3544, | |
| "rewards/accuracies": 0.44999998807907104, | |
| "rewards/chosen": -0.0001579081581439823, | |
| "rewards/margins": -0.00015658045595046133, | |
| "rewards/rejected": -1.3277137895784108e-06, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 1.6000000000000001e-06, | |
| "logits/chosen": 0.16454997658729553, | |
| "logits/rejected": 0.15700042247772217, | |
| "logps/chosen": -577.5484619140625, | |
| "logps/rejected": -601.3185424804688, | |
| "loss": 2.2042, | |
| "rewards/accuracies": 0.48750001192092896, | |
| "rewards/chosen": -0.0004847920499742031, | |
| "rewards/margins": 7.897525210864842e-05, | |
| "rewards/rejected": -0.0005637672729790211, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "logits/chosen": 0.15456734597682953, | |
| "logits/rejected": 0.230211541056633, | |
| "logps/chosen": -541.5357666015625, | |
| "logps/rejected": -564.4284057617188, | |
| "loss": 2.1489, | |
| "rewards/accuracies": 0.4375, | |
| "rewards/chosen": -0.0010174433700740337, | |
| "rewards/margins": -0.00017645664047449827, | |
| "rewards/rejected": -0.0008409869042225182, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 2.4000000000000003e-06, | |
| "logits/chosen": 0.13179543614387512, | |
| "logits/rejected": 0.23396913707256317, | |
| "logps/chosen": -612.1072387695312, | |
| "logps/rejected": -589.408203125, | |
| "loss": 2.1442, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -0.0011288017267361283, | |
| "rewards/margins": 0.00029289646772667766, | |
| "rewards/rejected": -0.001421698136255145, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 2.8000000000000003e-06, | |
| "logits/chosen": 0.206586092710495, | |
| "logits/rejected": 0.20121100544929504, | |
| "logps/chosen": -562.5863037109375, | |
| "logps/rejected": -549.765380859375, | |
| "loss": 2.0063, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -0.0016541840741410851, | |
| "rewards/margins": 0.0002594069519545883, | |
| "rewards/rejected": -0.001913591055199504, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 3.2000000000000003e-06, | |
| "logits/chosen": 0.15584848821163177, | |
| "logits/rejected": 0.22724422812461853, | |
| "logps/chosen": -616.1813354492188, | |
| "logps/rejected": -597.015625, | |
| "loss": 2.263, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -0.0020384504459798336, | |
| "rewards/margins": 0.0003365904267411679, | |
| "rewards/rejected": -0.002375040901824832, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 3.6000000000000003e-06, | |
| "logits/chosen": 0.13548582792282104, | |
| "logits/rejected": 0.23147103190422058, | |
| "logps/chosen": -588.42919921875, | |
| "logps/rejected": -551.1621704101562, | |
| "loss": 2.2401, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.0012455533724278212, | |
| "rewards/margins": 0.0012538704322651029, | |
| "rewards/rejected": -0.002499423688277602, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 4.000000000000001e-06, | |
| "logits/chosen": 0.15673957765102386, | |
| "logits/rejected": 0.2734481692314148, | |
| "logps/chosen": -574.0126342773438, | |
| "logps/rejected": -630.3812255859375, | |
| "loss": 2.0863, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.002293633297085762, | |
| "rewards/margins": 0.002124248770996928, | |
| "rewards/rejected": -0.004417881835252047, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 4.4e-06, | |
| "logits/chosen": 0.15073809027671814, | |
| "logits/rejected": 0.2886480689048767, | |
| "logps/chosen": -578.6621704101562, | |
| "logps/rejected": -523.6320190429688, | |
| "loss": 2.106, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.0048215631395578384, | |
| "rewards/margins": 0.0025550012942403555, | |
| "rewards/rejected": -0.007376563735306263, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 4.800000000000001e-06, | |
| "logits/chosen": 0.20540550351142883, | |
| "logits/rejected": 0.22172732651233673, | |
| "logps/chosen": -656.3192138671875, | |
| "logps/rejected": -624.5650634765625, | |
| "loss": 2.1297, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.010842840187251568, | |
| "rewards/margins": 0.0032757048029452562, | |
| "rewards/rejected": -0.01411854475736618, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 4.999756310023261e-06, | |
| "logits/chosen": 0.13236024975776672, | |
| "logits/rejected": 0.27400949597358704, | |
| "logps/chosen": -594.3499145507812, | |
| "logps/rejected": -589.4105834960938, | |
| "loss": 2.0564, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -0.01253579556941986, | |
| "rewards/margins": 0.005572236143052578, | |
| "rewards/rejected": -0.018108030781149864, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 4.997807075247147e-06, | |
| "logits/chosen": 0.207227423787117, | |
| "logits/rejected": 0.17817452549934387, | |
| "logps/chosen": -571.6937255859375, | |
| "logps/rejected": -593.8927612304688, | |
| "loss": 2.0839, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.013347337022423744, | |
| "rewards/margins": 0.008908352814614773, | |
| "rewards/rejected": -0.022255688905715942, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 4.993910125649561e-06, | |
| "logits/chosen": 0.1696740686893463, | |
| "logits/rejected": 0.18951769173145294, | |
| "logps/chosen": -596.6219482421875, | |
| "logps/rejected": -637.0972900390625, | |
| "loss": 2.2635, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.022671109065413475, | |
| "rewards/margins": 0.016697829589247704, | |
| "rewards/rejected": -0.03936893865466118, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 4.988068499954578e-06, | |
| "logits/chosen": 0.18265119194984436, | |
| "logits/rejected": 0.27114540338516235, | |
| "logps/chosen": -692.2119140625, | |
| "logps/rejected": -658.11767578125, | |
| "loss": 2.1075, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -0.04613715782761574, | |
| "rewards/margins": 0.007823893800377846, | |
| "rewards/rejected": -0.053961049765348434, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 4.980286753286196e-06, | |
| "logits/chosen": 0.23610782623291016, | |
| "logits/rejected": 0.25577467679977417, | |
| "logps/chosen": -607.8341674804688, | |
| "logps/rejected": -622.9971923828125, | |
| "loss": 2.1709, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.04283341392874718, | |
| "rewards/margins": 0.013730195350944996, | |
| "rewards/rejected": -0.056563615798950195, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 4.970570953616383e-06, | |
| "logits/chosen": 0.2259415090084076, | |
| "logits/rejected": 0.2490004301071167, | |
| "logps/chosen": -619.1112060546875, | |
| "logps/rejected": -711.3251953125, | |
| "loss": 2.1606, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -0.04884644225239754, | |
| "rewards/margins": 0.03266354650259018, | |
| "rewards/rejected": -0.08150999248027802, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 4.958928677033465e-06, | |
| "logits/chosen": 0.21490153670310974, | |
| "logits/rejected": 0.20755355060100555, | |
| "logps/chosen": -661.9138793945312, | |
| "logps/rejected": -686.15869140625, | |
| "loss": 2.1463, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.0876893401145935, | |
| "rewards/margins": 0.05007048696279526, | |
| "rewards/rejected": -0.13775983452796936, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 4.9453690018345144e-06, | |
| "logits/chosen": 0.22380857169628143, | |
| "logits/rejected": 0.34608834981918335, | |
| "logps/chosen": -859.7734375, | |
| "logps/rejected": -910.9814453125, | |
| "loss": 1.8887, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.2093479186296463, | |
| "rewards/margins": 0.08500120043754578, | |
| "rewards/rejected": -0.29434913396835327, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 4.9299025014463665e-06, | |
| "logits/chosen": 0.2813585102558136, | |
| "logits/rejected": 0.33269229531288147, | |
| "logps/chosen": -944.7975463867188, | |
| "logps/rejected": -1029.7379150390625, | |
| "loss": 1.9249, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.3255089223384857, | |
| "rewards/margins": 0.09575793147087097, | |
| "rewards/rejected": -0.4212668836116791, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 4.912541236180779e-06, | |
| "logits/chosen": 0.2951585650444031, | |
| "logits/rejected": 0.49975720047950745, | |
| "logps/chosen": -904.1650390625, | |
| "logps/rejected": -1103.083740234375, | |
| "loss": 1.9609, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.35912585258483887, | |
| "rewards/margins": 0.16983045637607574, | |
| "rewards/rejected": -0.5289562940597534, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 4.893298743830168e-06, | |
| "logits/chosen": 0.4890425205230713, | |
| "logits/rejected": 0.5553490519523621, | |
| "logps/chosen": -935.9508666992188, | |
| "logps/rejected": -1199.338623046875, | |
| "loss": 2.0485, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.3711962103843689, | |
| "rewards/margins": 0.2363450974225998, | |
| "rewards/rejected": -0.6075413823127747, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 4.8721900291112415e-06, | |
| "logits/chosen": 0.3669169545173645, | |
| "logits/rejected": 0.46804532408714294, | |
| "logps/chosen": -862.8795166015625, | |
| "logps/rejected": -937.4802856445312, | |
| "loss": 2.1176, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -0.2683151364326477, | |
| "rewards/margins": 0.10171107202768326, | |
| "rewards/rejected": -0.37002623081207275, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 4.849231551964771e-06, | |
| "logits/chosen": 0.28659749031066895, | |
| "logits/rejected": 0.44282183051109314, | |
| "logps/chosen": -833.5822143554688, | |
| "logps/rejected": -910.1467895507812, | |
| "loss": 1.9678, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.2521842122077942, | |
| "rewards/margins": 0.07951636612415314, | |
| "rewards/rejected": -0.33170050382614136, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 4.824441214720629e-06, | |
| "logits/chosen": 0.40166163444519043, | |
| "logits/rejected": 0.44364985823631287, | |
| "logps/chosen": -788.6900634765625, | |
| "logps/rejected": -991.3458862304688, | |
| "loss": 1.7415, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.2276671826839447, | |
| "rewards/margins": 0.16912813484668732, | |
| "rewards/rejected": -0.3967953026294708, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 4.7978383481380865e-06, | |
| "logits/chosen": 0.3444485664367676, | |
| "logits/rejected": 0.5015489459037781, | |
| "logps/chosen": -825.1337890625, | |
| "logps/rejected": -1013.6962890625, | |
| "loss": 1.8545, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.2368713915348053, | |
| "rewards/margins": 0.19953957200050354, | |
| "rewards/rejected": -0.43641090393066406, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 4.769443696332272e-06, | |
| "logits/chosen": 0.32490837574005127, | |
| "logits/rejected": 0.4952085018157959, | |
| "logps/chosen": -748.92578125, | |
| "logps/rejected": -876.5540161132812, | |
| "loss": 1.9791, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.20112459361553192, | |
| "rewards/margins": 0.12988439202308655, | |
| "rewards/rejected": -0.3310089707374573, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 4.7392794005985324e-06, | |
| "logits/chosen": 0.41541141271591187, | |
| "logits/rejected": 0.5006674528121948, | |
| "logps/chosen": -831.7371215820312, | |
| "logps/rejected": -1032.0704345703125, | |
| "loss": 1.7353, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.24899300932884216, | |
| "rewards/margins": 0.20179975032806396, | |
| "rewards/rejected": -0.45079272985458374, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 4.707368982147318e-06, | |
| "logits/chosen": 0.36991387605667114, | |
| "logits/rejected": 0.48434096574783325, | |
| "logps/chosen": -898.0588989257812, | |
| "logps/rejected": -1055.823486328125, | |
| "loss": 1.9315, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.30798205733299255, | |
| "rewards/margins": 0.15843333303928375, | |
| "rewards/rejected": -0.4664154052734375, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 4.673737323763048e-06, | |
| "logits/chosen": 0.47087937593460083, | |
| "logits/rejected": 0.4668746590614319, | |
| "logps/chosen": -921.783203125, | |
| "logps/rejected": -1135.657470703125, | |
| "loss": 1.9714, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.33973026275634766, | |
| "rewards/margins": 0.1682974398136139, | |
| "rewards/rejected": -0.5080276727676392, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 4.638410650401267e-06, | |
| "logits/chosen": 0.32884734869003296, | |
| "logits/rejected": 0.5667840242385864, | |
| "logps/chosen": -793.4259643554688, | |
| "logps/rejected": -1097.870361328125, | |
| "loss": 1.6468, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.22830717265605927, | |
| "rewards/margins": 0.2654227614402771, | |
| "rewards/rejected": -0.49372991919517517, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 4.601416508739211e-06, | |
| "logits/chosen": 0.3828834891319275, | |
| "logits/rejected": 0.5297464728355408, | |
| "logps/chosen": -948.02197265625, | |
| "logps/rejected": -1136.756103515625, | |
| "loss": 1.8833, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.3334728181362152, | |
| "rewards/margins": 0.1820012331008911, | |
| "rewards/rejected": -0.5154740810394287, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 4.562783745695738e-06, | |
| "logits/chosen": 0.41931548714637756, | |
| "logits/rejected": 0.5291355848312378, | |
| "logps/chosen": -862.7106323242188, | |
| "logps/rejected": -1053.8846435546875, | |
| "loss": 1.7946, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.2459833323955536, | |
| "rewards/margins": 0.19779649376869202, | |
| "rewards/rejected": -0.443779855966568, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 4.522542485937369e-06, | |
| "logits/chosen": 0.3689717650413513, | |
| "logits/rejected": 0.4881824553012848, | |
| "logps/chosen": -875.2266845703125, | |
| "logps/rejected": -1156.237060546875, | |
| "loss": 1.6268, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.257233202457428, | |
| "rewards/margins": 0.27138087153434753, | |
| "rewards/rejected": -0.5286141037940979, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 4.4807241083879774e-06, | |
| "logits/chosen": 0.4062356948852539, | |
| "logits/rejected": 0.4961213171482086, | |
| "logps/chosen": -890.9068603515625, | |
| "logps/rejected": -1214.0831298828125, | |
| "loss": 1.6478, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.3261027932167053, | |
| "rewards/margins": 0.30797919631004333, | |
| "rewards/rejected": -0.6340819597244263, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 4.437361221760449e-06, | |
| "logits/chosen": 0.3728734850883484, | |
| "logits/rejected": 0.39377298951148987, | |
| "logps/chosen": -995.19482421875, | |
| "logps/rejected": -1160.993408203125, | |
| "loss": 2.014, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.40234214067459106, | |
| "rewards/margins": 0.1731390357017517, | |
| "rewards/rejected": -0.5754811763763428, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 4.3924876391293915e-06, | |
| "logits/chosen": 0.357779324054718, | |
| "logits/rejected": 0.48311319947242737, | |
| "logps/chosen": -854.5018310546875, | |
| "logps/rejected": -1101.8800048828125, | |
| "loss": 1.7634, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.2472849190235138, | |
| "rewards/margins": 0.23356270790100098, | |
| "rewards/rejected": -0.48084765672683716, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 4.346138351564711e-06, | |
| "logits/chosen": 0.32345646619796753, | |
| "logits/rejected": 0.4409434199333191, | |
| "logps/chosen": -821.7467651367188, | |
| "logps/rejected": -1076.5284423828125, | |
| "loss": 1.5989, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.2140711098909378, | |
| "rewards/margins": 0.2635394036769867, | |
| "rewards/rejected": -0.4776105284690857, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 4.2983495008466285e-06, | |
| "logits/chosen": 0.36422181129455566, | |
| "logits/rejected": 0.45148295164108276, | |
| "logps/chosen": -922.22705078125, | |
| "logps/rejected": -1089.808837890625, | |
| "loss": 1.7327, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.3074890375137329, | |
| "rewards/margins": 0.18984538316726685, | |
| "rewards/rejected": -0.49733442068099976, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 4.249158351283414e-06, | |
| "logits/chosen": 0.37963396310806274, | |
| "logits/rejected": 0.43684640526771545, | |
| "logps/chosen": -816.420166015625, | |
| "logps/rejected": -1026.3609619140625, | |
| "loss": 1.9326, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.26014500856399536, | |
| "rewards/margins": 0.17839236557483673, | |
| "rewards/rejected": -0.4385373592376709, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 4.198603260653792e-06, | |
| "logits/chosen": 0.3479335308074951, | |
| "logits/rejected": 0.5144117474555969, | |
| "logps/chosen": -808.4892578125, | |
| "logps/rejected": -1022.3094482421875, | |
| "loss": 1.7198, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.24907036125659943, | |
| "rewards/margins": 0.21405251324176788, | |
| "rewards/rejected": -0.4631228446960449, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 4.146723650296701e-06, | |
| "logits/chosen": 0.375163733959198, | |
| "logits/rejected": 0.5701647996902466, | |
| "logps/chosen": -894.6932373046875, | |
| "logps/rejected": -1191.1158447265625, | |
| "loss": 1.7984, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.28425461053848267, | |
| "rewards/margins": 0.2671293616294861, | |
| "rewards/rejected": -0.5513839721679688, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 4.093559974371725e-06, | |
| "logits/chosen": 0.39743003249168396, | |
| "logits/rejected": 0.499051570892334, | |
| "logps/chosen": -769.8292236328125, | |
| "logps/rejected": -1028.2760009765625, | |
| "loss": 1.8098, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.2524164319038391, | |
| "rewards/margins": 0.2162034809589386, | |
| "rewards/rejected": -0.4686198830604553, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 4.039153688314146e-06, | |
| "logits/chosen": 0.37791648507118225, | |
| "logits/rejected": 0.46076732873916626, | |
| "logps/chosen": -876.4410400390625, | |
| "logps/rejected": -1076.7730712890625, | |
| "loss": 1.824, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -0.2770997881889343, | |
| "rewards/margins": 0.2473537176847458, | |
| "rewards/rejected": -0.5244535207748413, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 3.983547216509254e-06, | |
| "logits/chosen": 0.36932164430618286, | |
| "logits/rejected": 0.5779580473899841, | |
| "logps/chosen": -891.6867065429688, | |
| "logps/rejected": -1252.7017822265625, | |
| "loss": 1.5803, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -0.27173933386802673, | |
| "rewards/margins": 0.33879798650741577, | |
| "rewards/rejected": -0.6105372905731201, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 3.92678391921108e-06, | |
| "logits/chosen": 0.376406729221344, | |
| "logits/rejected": 0.46922606229782104, | |
| "logps/chosen": -881.6818237304688, | |
| "logps/rejected": -1075.5714111328125, | |
| "loss": 1.9328, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.3042740225791931, | |
| "rewards/margins": 0.21284916996955872, | |
| "rewards/rejected": -0.5171232223510742, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 3.868908058731376e-06, | |
| "logits/chosen": 0.3659452497959137, | |
| "logits/rejected": 0.47788533568382263, | |
| "logps/chosen": -914.9581909179688, | |
| "logps/rejected": -1192.8797607421875, | |
| "loss": 1.8254, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.333695650100708, | |
| "rewards/margins": 0.27644091844558716, | |
| "rewards/rejected": -0.6101365685462952, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 3.8099647649251984e-06, | |
| "logits/chosen": 0.3602338135242462, | |
| "logits/rejected": 0.4133017063140869, | |
| "logps/chosen": -875.7825927734375, | |
| "logps/rejected": -1087.2510986328125, | |
| "loss": 1.8227, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.3075736165046692, | |
| "rewards/margins": 0.24024859070777893, | |
| "rewards/rejected": -0.547822117805481, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 3.7500000000000005e-06, | |
| "logits/chosen": 0.3213595449924469, | |
| "logits/rejected": 0.4600197374820709, | |
| "logps/chosen": -851.552734375, | |
| "logps/rejected": -1143.1978759765625, | |
| "loss": 1.7429, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.2909749150276184, | |
| "rewards/margins": 0.2939184010028839, | |
| "rewards/rejected": -0.5848933458328247, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 3.689060522675689e-06, | |
| "logits/chosen": 0.3785388469696045, | |
| "logits/rejected": 0.37024298310279846, | |
| "logps/chosen": -865.6300659179688, | |
| "logps/rejected": -1103.9659423828125, | |
| "loss": 1.7996, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.29555124044418335, | |
| "rewards/margins": 0.2511122524738312, | |
| "rewards/rejected": -0.5466634631156921, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 3.627193851723577e-06, | |
| "logits/chosen": 0.35357269644737244, | |
| "logits/rejected": 0.44010990858078003, | |
| "logps/chosen": -922.1881713867188, | |
| "logps/rejected": -1222.6695556640625, | |
| "loss": 1.8295, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.34208646416664124, | |
| "rewards/margins": 0.2784285545349121, | |
| "rewards/rejected": -0.620514988899231, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 3.564448228912682e-06, | |
| "logits/chosen": 0.3053657114505768, | |
| "logits/rejected": 0.4411580562591553, | |
| "logps/chosen": -964.7332763671875, | |
| "logps/rejected": -1235.441650390625, | |
| "loss": 1.7541, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.3641963601112366, | |
| "rewards/margins": 0.26334619522094727, | |
| "rewards/rejected": -0.6275425553321838, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 3.5008725813922383e-06, | |
| "logits/chosen": 0.3650105595588684, | |
| "logits/rejected": 0.35269036889076233, | |
| "logps/chosen": -866.552734375, | |
| "logps/rejected": -1229.647216796875, | |
| "loss": 1.6454, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -0.29515185952186584, | |
| "rewards/margins": 0.3350960910320282, | |
| "rewards/rejected": -0.6302478909492493, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 3.436516483539781e-06, | |
| "logits/chosen": 0.30932605266571045, | |
| "logits/rejected": 0.382567435503006, | |
| "logps/chosen": -870.6939697265625, | |
| "logps/rejected": -1298.612060546875, | |
| "loss": 1.616, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -0.30489498376846313, | |
| "rewards/margins": 0.3668510317802429, | |
| "rewards/rejected": -0.6717459559440613, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 3.3714301183045382e-06, | |
| "logits/chosen": 0.38849467039108276, | |
| "logits/rejected": 0.4396263659000397, | |
| "logps/chosen": -902.1296997070312, | |
| "logps/rejected": -1177.967529296875, | |
| "loss": 1.9176, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.3350718319416046, | |
| "rewards/margins": 0.2506067156791687, | |
| "rewards/rejected": -0.5856784582138062, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 3.3056642380762783e-06, | |
| "logits/chosen": 0.29538553953170776, | |
| "logits/rejected": 0.43000030517578125, | |
| "logps/chosen": -952.9089965820312, | |
| "logps/rejected": -1251.63525390625, | |
| "loss": 1.6891, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -0.36573296785354614, | |
| "rewards/margins": 0.29417911171913147, | |
| "rewards/rejected": -0.6599120497703552, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 3.2392701251101172e-06, | |
| "logits/chosen": 0.26259809732437134, | |
| "logits/rejected": 0.4230431914329529, | |
| "logps/chosen": -981.2227783203125, | |
| "logps/rejected": -1164.7890625, | |
| "loss": 1.8962, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.3732236325740814, | |
| "rewards/margins": 0.22464247047901154, | |
| "rewards/rejected": -0.5978660583496094, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 3.1722995515381644e-06, | |
| "logits/chosen": 0.3150332570075989, | |
| "logits/rejected": 0.4163808226585388, | |
| "logps/chosen": -904.3822021484375, | |
| "logps/rejected": -1273.4002685546875, | |
| "loss": 1.7318, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.3179934620857239, | |
| "rewards/margins": 0.31932058930397034, | |
| "rewards/rejected": -0.6373140215873718, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 3.1048047389991693e-06, | |
| "logits/chosen": 0.2711288630962372, | |
| "logits/rejected": 0.38649213314056396, | |
| "logps/chosen": -961.5247802734375, | |
| "logps/rejected": -1165.8568115234375, | |
| "loss": 1.8103, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.35747116804122925, | |
| "rewards/margins": 0.2142975628376007, | |
| "rewards/rejected": -0.5717687010765076, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 3.0368383179176584e-06, | |
| "logits/chosen": 0.31565985083580017, | |
| "logits/rejected": 0.3596263825893402, | |
| "logps/chosen": -1067.418701171875, | |
| "logps/rejected": -1284.5201416015625, | |
| "loss": 1.7442, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.42869243025779724, | |
| "rewards/margins": 0.23898033797740936, | |
| "rewards/rejected": -0.6676727533340454, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 2.9684532864643123e-06, | |
| "logits/chosen": 0.2723945081233978, | |
| "logits/rejected": 0.42879945039749146, | |
| "logps/chosen": -907.2979736328125, | |
| "logps/rejected": -1193.40625, | |
| "loss": 1.7749, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.37917786836624146, | |
| "rewards/margins": 0.25062739849090576, | |
| "rewards/rejected": -0.6298052668571472, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 2.8997029692295875e-06, | |
| "logits/chosen": 0.2836493253707886, | |
| "logits/rejected": 0.36888745427131653, | |
| "logps/chosen": -958.9281005859375, | |
| "logps/rejected": -1170.929443359375, | |
| "loss": 1.7337, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.36827540397644043, | |
| "rewards/margins": 0.24554581940174103, | |
| "rewards/rejected": -0.613821268081665, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 2.8306409756428067e-06, | |
| "logits/chosen": 0.34693920612335205, | |
| "logits/rejected": 0.33596283197402954, | |
| "logps/chosen": -884.5350341796875, | |
| "logps/rejected": -1140.420654296875, | |
| "loss": 1.7815, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.32642626762390137, | |
| "rewards/margins": 0.22602030634880066, | |
| "rewards/rejected": -0.5524466633796692, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 2.761321158169134e-06, | |
| "logits/chosen": 0.29072973132133484, | |
| "logits/rejected": 0.4021405279636383, | |
| "logps/chosen": -920.5359497070312, | |
| "logps/rejected": -1172.0166015625, | |
| "loss": 1.7447, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.31148773431777954, | |
| "rewards/margins": 0.2528636157512665, | |
| "rewards/rejected": -0.5643514394760132, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 2.6917975703170466e-06, | |
| "logits/chosen": 0.30165717005729675, | |
| "logits/rejected": 0.41581812500953674, | |
| "logps/chosen": -998.71484375, | |
| "logps/rejected": -1180.068115234375, | |
| "loss": 1.7292, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -0.3652830719947815, | |
| "rewards/margins": 0.22447367012500763, | |
| "rewards/rejected": -0.5897567868232727, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 2.6221244244890336e-06, | |
| "logits/chosen": 0.2630220949649811, | |
| "logits/rejected": 0.38998615741729736, | |
| "logps/chosen": -850.4860229492188, | |
| "logps/rejected": -1299.91015625, | |
| "loss": 1.4881, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.32001981139183044, | |
| "rewards/margins": 0.3484157621860504, | |
| "rewards/rejected": -0.6684354543685913, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 2.5523560497083927e-06, | |
| "logits/chosen": 0.28038591146469116, | |
| "logits/rejected": 0.4210067391395569, | |
| "logps/chosen": -989.4406127929688, | |
| "logps/rejected": -1252.375732421875, | |
| "loss": 1.686, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.37420469522476196, | |
| "rewards/margins": 0.259860098361969, | |
| "rewards/rejected": -0.634064793586731, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 2.482546849255096e-06, | |
| "logits/chosen": 0.3020229935646057, | |
| "logits/rejected": 0.3828849792480469, | |
| "logps/chosen": -956.5652465820312, | |
| "logps/rejected": -1176.745849609375, | |
| "loss": 1.6468, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.3711964190006256, | |
| "rewards/margins": 0.2462393045425415, | |
| "rewards/rejected": -0.6174356937408447, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 2.4127512582437486e-06, | |
| "logits/chosen": 0.3271457552909851, | |
| "logits/rejected": 0.3496400713920593, | |
| "logps/chosen": -860.8416137695312, | |
| "logps/rejected": -1315.4486083984375, | |
| "loss": 1.4554, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.3369373083114624, | |
| "rewards/margins": 0.36229121685028076, | |
| "rewards/rejected": -0.6992285251617432, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 2.3430237011767166e-06, | |
| "logits/chosen": 0.32887858152389526, | |
| "logits/rejected": 0.38950568437576294, | |
| "logps/chosen": -962.7874755859375, | |
| "logps/rejected": -1309.929443359375, | |
| "loss": 1.6678, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -0.3625487685203552, | |
| "rewards/margins": 0.3222491443157196, | |
| "rewards/rejected": -0.6847978830337524, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 2.2734185495055503e-06, | |
| "logits/chosen": 0.21860530972480774, | |
| "logits/rejected": 0.3752947151660919, | |
| "logps/chosen": -1050.0916748046875, | |
| "logps/rejected": -1280.526123046875, | |
| "loss": 1.9756, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.42955031991004944, | |
| "rewards/margins": 0.23833701014518738, | |
| "rewards/rejected": -0.6678873896598816, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 2.2039900792337477e-06, | |
| "logits/chosen": 0.3153984248638153, | |
| "logits/rejected": 0.37083715200424194, | |
| "logps/chosen": -918.3059692382812, | |
| "logps/rejected": -1242.553466796875, | |
| "loss": 1.8289, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.34617164731025696, | |
| "rewards/margins": 0.2804465591907501, | |
| "rewards/rejected": -0.6266182065010071, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 2.134792428593971e-06, | |
| "logits/chosen": 0.25939470529556274, | |
| "logits/rejected": 0.31975841522216797, | |
| "logps/chosen": -814.0809326171875, | |
| "logps/rejected": -1025.2806396484375, | |
| "loss": 1.8756, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.257251501083374, | |
| "rewards/margins": 0.21475358307361603, | |
| "rewards/rejected": -0.47200506925582886, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 2.0658795558326745e-06, | |
| "logits/chosen": 0.24114593863487244, | |
| "logits/rejected": 0.3729786276817322, | |
| "logps/chosen": -844.0042724609375, | |
| "logps/rejected": -1063.400634765625, | |
| "loss": 1.701, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.2636208236217499, | |
| "rewards/margins": 0.22609353065490723, | |
| "rewards/rejected": -0.4897143840789795, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 1.997305197135089e-06, | |
| "logits/chosen": 0.26230183243751526, | |
| "logits/rejected": 0.4085041582584381, | |
| "logps/chosen": -874.07568359375, | |
| "logps/rejected": -1172.9375, | |
| "loss": 1.4942, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -0.2855741083621979, | |
| "rewards/margins": 0.28826937079429626, | |
| "rewards/rejected": -0.5738434791564941, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 1.9291228247233607e-06, | |
| "logits/chosen": 0.30484798550605774, | |
| "logits/rejected": 0.4110705256462097, | |
| "logps/chosen": -969.3850708007812, | |
| "logps/rejected": -1283.5828857421875, | |
| "loss": 1.8138, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.34823209047317505, | |
| "rewards/margins": 0.2978925108909607, | |
| "rewards/rejected": -0.6461246013641357, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 1.8613856051605242e-06, | |
| "logits/chosen": 0.26855385303497314, | |
| "logits/rejected": 0.42292389273643494, | |
| "logps/chosen": -959.6076049804688, | |
| "logps/rejected": -1120.3887939453125, | |
| "loss": 1.9167, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.35128024220466614, | |
| "rewards/margins": 0.20051094889640808, | |
| "rewards/rejected": -0.5517911911010742, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 1.7941463578928088e-06, | |
| "logits/chosen": 0.3370608687400818, | |
| "logits/rejected": 0.38068529963493347, | |
| "logps/chosen": -885.8193359375, | |
| "logps/rejected": -1103.631591796875, | |
| "loss": 1.9313, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -0.3268662095069885, | |
| "rewards/margins": 0.2425314486026764, | |
| "rewards/rejected": -0.5693976283073425, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 1.7274575140626318e-06, | |
| "logits/chosen": 0.26554322242736816, | |
| "logits/rejected": 0.3851606249809265, | |
| "logps/chosen": -990.44140625, | |
| "logps/rejected": -1205.0343017578125, | |
| "loss": 1.7604, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.35415148735046387, | |
| "rewards/margins": 0.21467110514640808, | |
| "rewards/rejected": -0.5688225626945496, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 1.661371075624363e-06, | |
| "logits/chosen": 0.2818129062652588, | |
| "logits/rejected": 0.3556861877441406, | |
| "logps/chosen": -850.845703125, | |
| "logps/rejected": -1055.9957275390625, | |
| "loss": 1.7347, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.30765876173973083, | |
| "rewards/margins": 0.22872476279735565, | |
| "rewards/rejected": -0.5363835096359253, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 1.5959385747947697e-06, | |
| "logits/chosen": 0.2711453139781952, | |
| "logits/rejected": 0.38543039560317993, | |
| "logps/chosen": -874.6898193359375, | |
| "logps/rejected": -1049.7301025390625, | |
| "loss": 1.839, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.2933613657951355, | |
| "rewards/margins": 0.20129230618476868, | |
| "rewards/rejected": -0.49465370178222656, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 1.5312110338697427e-06, | |
| "logits/chosen": 0.3340443968772888, | |
| "logits/rejected": 0.3357875943183899, | |
| "logps/chosen": -926.2916870117188, | |
| "logps/rejected": -1192.5069580078125, | |
| "loss": 1.537, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.32535651326179504, | |
| "rewards/margins": 0.2853156626224518, | |
| "rewards/rejected": -0.6106721758842468, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 1.467238925438646e-06, | |
| "logits/chosen": 0.25067275762557983, | |
| "logits/rejected": 0.36189913749694824, | |
| "logps/chosen": -882.9215698242188, | |
| "logps/rejected": -1149.0230712890625, | |
| "loss": 1.7236, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.32856041193008423, | |
| "rewards/margins": 0.2509750723838806, | |
| "rewards/rejected": -0.5795354843139648, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 1.4040721330273063e-06, | |
| "logits/chosen": 0.23793812096118927, | |
| "logits/rejected": 0.3621876537799835, | |
| "logps/chosen": -858.2410278320312, | |
| "logps/rejected": -1095.5906982421875, | |
| "loss": 1.6914, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.2868640124797821, | |
| "rewards/margins": 0.2690005302429199, | |
| "rewards/rejected": -0.5558645129203796, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 1.3417599122003464e-06, | |
| "logits/chosen": 0.34424689412117004, | |
| "logits/rejected": 0.3806430697441101, | |
| "logps/chosen": -927.8732299804688, | |
| "logps/rejected": -1146.036865234375, | |
| "loss": 1.7111, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.30680620670318604, | |
| "rewards/margins": 0.2561938464641571, | |
| "rewards/rejected": -0.5630000829696655, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 1.280350852153168e-06, | |
| "logits/chosen": 0.3122938275337219, | |
| "logits/rejected": 0.2692076861858368, | |
| "logps/chosen": -810.717529296875, | |
| "logps/rejected": -1086.6156005859375, | |
| "loss": 1.7111, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.28373444080352783, | |
| "rewards/margins": 0.24869227409362793, | |
| "rewards/rejected": -0.532426655292511, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 1.2198928378235717e-06, | |
| "logits/chosen": 0.261820912361145, | |
| "logits/rejected": 0.49146905541419983, | |
| "logps/chosen": -936.8932495117188, | |
| "logps/rejected": -1308.829833984375, | |
| "loss": 1.4018, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.31694522500038147, | |
| "rewards/margins": 0.35484033823013306, | |
| "rewards/rejected": -0.6717855334281921, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 1.160433012552508e-06, | |
| "logits/chosen": 0.29470381140708923, | |
| "logits/rejected": 0.3451955020427704, | |
| "logps/chosen": -989.2918090820312, | |
| "logps/rejected": -1229.162353515625, | |
| "loss": 1.6905, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -0.35454291105270386, | |
| "rewards/margins": 0.262814462184906, | |
| "rewards/rejected": -0.6173573732376099, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 1.1020177413231334e-06, | |
| "logits/chosen": 0.3188125193119049, | |
| "logits/rejected": 0.38136720657348633, | |
| "logps/chosen": -893.96240234375, | |
| "logps/rejected": -1130.4246826171875, | |
| "loss": 1.6475, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.31285107135772705, | |
| "rewards/margins": 0.2777191996574402, | |
| "rewards/rejected": -0.5905702710151672, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 1.0446925746067768e-06, | |
| "logits/chosen": 0.28160279989242554, | |
| "logits/rejected": 0.33634817600250244, | |
| "logps/chosen": -903.3307495117188, | |
| "logps/rejected": -1249.0274658203125, | |
| "loss": 1.6275, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.30794739723205566, | |
| "rewards/margins": 0.30408358573913574, | |
| "rewards/rejected": -0.6120310425758362, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 9.88502212844063e-07, | |
| "logits/chosen": 0.2644272446632385, | |
| "logits/rejected": 0.4034818708896637, | |
| "logps/chosen": -837.537109375, | |
| "logps/rejected": -1180.740234375, | |
| "loss": 1.5707, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -0.2741805613040924, | |
| "rewards/margins": 0.31248125433921814, | |
| "rewards/rejected": -0.5866617560386658, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 9.334904715888496e-07, | |
| "logits/chosen": 0.26665693521499634, | |
| "logits/rejected": 0.429934024810791, | |
| "logps/chosen": -959.6316528320312, | |
| "logps/rejected": -1225.3084716796875, | |
| "loss": 1.7093, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.3429884910583496, | |
| "rewards/margins": 0.2681938707828522, | |
| "rewards/rejected": -0.6111823320388794, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 8.797002473421729e-07, | |
| "logits/chosen": 0.311528742313385, | |
| "logits/rejected": 0.3178673982620239, | |
| "logps/chosen": -891.2405395507812, | |
| "logps/rejected": -1275.465087890625, | |
| "loss": 1.4394, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.3114137053489685, | |
| "rewards/margins": 0.3879539370536804, | |
| "rewards/rejected": -0.6993676424026489, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 8.271734841028553e-07, | |
| "logits/chosen": 0.2898734211921692, | |
| "logits/rejected": 0.3742389678955078, | |
| "logps/chosen": -850.2423095703125, | |
| "logps/rejected": -1203.037841796875, | |
| "loss": 1.6261, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.31859657168388367, | |
| "rewards/margins": 0.3100232481956482, | |
| "rewards/rejected": -0.6286198496818542, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 7.759511406608255e-07, | |
| "logits/chosen": 0.30733975768089294, | |
| "logits/rejected": 0.44421592354774475, | |
| "logps/chosen": -934.5499267578125, | |
| "logps/rejected": -1207.3531494140625, | |
| "loss": 1.9113, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.3752412497997284, | |
| "rewards/margins": 0.25361892580986023, | |
| "rewards/rejected": -0.6288601756095886, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 7.260731586586983e-07, | |
| "logits/chosen": 0.3483877182006836, | |
| "logits/rejected": 0.3553027808666229, | |
| "logps/chosen": -893.8357543945312, | |
| "logps/rejected": -1164.4613037109375, | |
| "loss": 1.776, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.35817593336105347, | |
| "rewards/margins": 0.2500895857810974, | |
| "rewards/rejected": -0.6082655191421509, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 6.775784314464717e-07, | |
| "logits/chosen": 0.353769451379776, | |
| "logits/rejected": 0.4078153967857361, | |
| "logps/chosen": -968.4454956054688, | |
| "logps/rejected": -1228.4012451171875, | |
| "loss": 1.716, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.36273428797721863, | |
| "rewards/margins": 0.2496825009584427, | |
| "rewards/rejected": -0.6124168634414673, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 6.305047737536707e-07, | |
| "logits/chosen": 0.30879533290863037, | |
| "logits/rejected": 0.452120304107666, | |
| "logps/chosen": -877.1207275390625, | |
| "logps/rejected": -1282.6204833984375, | |
| "loss": 1.5403, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.30206242203712463, | |
| "rewards/margins": 0.3925813138484955, | |
| "rewards/rejected": -0.6946438550949097, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 5.848888922025553e-07, | |
| "logits/chosen": 0.27114662528038025, | |
| "logits/rejected": 0.45444902777671814, | |
| "logps/chosen": -969.3070068359375, | |
| "logps/rejected": -1175.6612548828125, | |
| "loss": 1.8379, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.3690924048423767, | |
| "rewards/margins": 0.2309758961200714, | |
| "rewards/rejected": -0.6000683903694153, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 5.407663566854008e-07, | |
| "logits/chosen": 0.2744256854057312, | |
| "logits/rejected": 0.41505903005599976, | |
| "logps/chosen": -845.2996826171875, | |
| "logps/rejected": -1316.6790771484375, | |
| "loss": 1.4883, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.27676206827163696, | |
| "rewards/margins": 0.4259292483329773, | |
| "rewards/rejected": -0.702691376209259, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 4.981715726281666e-07, | |
| "logits/chosen": 0.2770433723926544, | |
| "logits/rejected": 0.4135493338108063, | |
| "logps/chosen": -950.1988525390625, | |
| "logps/rejected": -1196.9830322265625, | |
| "loss": 2.0066, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.38483771681785583, | |
| "rewards/margins": 0.2553882300853729, | |
| "rewards/rejected": -0.6402260065078735, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 4.5713775416217884e-07, | |
| "logits/chosen": 0.3525824546813965, | |
| "logits/rejected": 0.35387757420539856, | |
| "logps/chosen": -931.20068359375, | |
| "logps/rejected": -1218.2432861328125, | |
| "loss": 1.7446, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.3306300938129425, | |
| "rewards/margins": 0.3051949739456177, | |
| "rewards/rejected": -0.6358250379562378, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 4.1769689822475147e-07, | |
| "logits/chosen": 0.30737197399139404, | |
| "logits/rejected": 0.4157963693141937, | |
| "logps/chosen": -1038.634033203125, | |
| "logps/rejected": -1335.1549072265625, | |
| "loss": 1.6302, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.37811559438705444, | |
| "rewards/margins": 0.3240962624549866, | |
| "rewards/rejected": -0.7022119760513306, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 3.798797596089351e-07, | |
| "logits/chosen": 0.3381016254425049, | |
| "logits/rejected": 0.4121452867984772, | |
| "logps/chosen": -985.3694458007812, | |
| "logps/rejected": -1181.4552001953125, | |
| "loss": 1.8847, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.37644538283348083, | |
| "rewards/margins": 0.2359016388654709, | |
| "rewards/rejected": -0.6123470067977905, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 3.4371582698185636e-07, | |
| "logits/chosen": 0.2851170003414154, | |
| "logits/rejected": 0.3804434835910797, | |
| "logps/chosen": -922.8662109375, | |
| "logps/rejected": -1263.484375, | |
| "loss": 1.729, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.3330380320549011, | |
| "rewards/margins": 0.33658528327941895, | |
| "rewards/rejected": -0.6696232557296753, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 3.092332998903416e-07, | |
| "logits/chosen": 0.2671842873096466, | |
| "logits/rejected": 0.34879574179649353, | |
| "logps/chosen": -977.2224731445312, | |
| "logps/rejected": -1249.781982421875, | |
| "loss": 1.7384, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -0.3930615484714508, | |
| "rewards/margins": 0.2609555125236511, | |
| "rewards/rejected": -0.6540170907974243, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 2.764590667717562e-07, | |
| "logits/chosen": 0.36944884061813354, | |
| "logits/rejected": 0.3970021903514862, | |
| "logps/chosen": -830.1124877929688, | |
| "logps/rejected": -1250.266357421875, | |
| "loss": 1.5305, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.3065842092037201, | |
| "rewards/margins": 0.34969645738601685, | |
| "rewards/rejected": -0.6562806367874146, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 2.454186839872158e-07, | |
| "logits/chosen": 0.30845582485198975, | |
| "logits/rejected": 0.3318074643611908, | |
| "logps/chosen": -933.2980346679688, | |
| "logps/rejected": -1170.7999267578125, | |
| "loss": 1.8682, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.3466985821723938, | |
| "rewards/margins": 0.2223629206418991, | |
| "rewards/rejected": -0.5690615177154541, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 2.1613635589349756e-07, | |
| "logits/chosen": 0.27249905467033386, | |
| "logits/rejected": 0.4322986602783203, | |
| "logps/chosen": -986.0386962890625, | |
| "logps/rejected": -1213.860107421875, | |
| "loss": 1.792, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.3734912574291229, | |
| "rewards/margins": 0.2489509880542755, | |
| "rewards/rejected": -0.6224422454833984, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "learning_rate": 1.8863491596921745e-07, | |
| "logits/chosen": 0.31418293714523315, | |
| "logits/rejected": 0.39671963453292847, | |
| "logps/chosen": -923.6838989257812, | |
| "logps/rejected": -1139.2310791015625, | |
| "loss": 1.8913, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.3423645496368408, | |
| "rewards/margins": 0.2353101521730423, | |
| "rewards/rejected": -0.5776747465133667, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 1.629358090099639e-07, | |
| "logits/chosen": 0.31794118881225586, | |
| "logits/rejected": 0.4100438058376312, | |
| "logps/chosen": -999.4279174804688, | |
| "logps/rejected": -1165.24169921875, | |
| "loss": 1.7528, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -0.361406147480011, | |
| "rewards/margins": 0.24150419235229492, | |
| "rewards/rejected": -0.6029103994369507, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 1.3905907440629752e-07, | |
| "logits/chosen": 0.28145748376846313, | |
| "logits/rejected": 0.3526201844215393, | |
| "logps/chosen": -896.6658935546875, | |
| "logps/rejected": -1153.307861328125, | |
| "loss": 1.7645, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.33654385805130005, | |
| "rewards/margins": 0.23922619223594666, | |
| "rewards/rejected": -0.5757700204849243, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 1.1702333051763271e-07, | |
| "logits/chosen": 0.30934053659439087, | |
| "logits/rejected": 0.3852478861808777, | |
| "logps/chosen": -949.3416137695312, | |
| "logps/rejected": -1139.066650390625, | |
| "loss": 1.7438, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.33037489652633667, | |
| "rewards/margins": 0.2449624091386795, | |
| "rewards/rejected": -0.575337290763855, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 9.684576015420277e-08, | |
| "logits/chosen": 0.27784961462020874, | |
| "logits/rejected": 0.40725311636924744, | |
| "logps/chosen": -865.6135864257812, | |
| "logps/rejected": -1142.362060546875, | |
| "loss": 1.8127, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.30999353528022766, | |
| "rewards/margins": 0.2895258069038391, | |
| "rewards/rejected": -0.5995193123817444, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 7.854209717842231e-08, | |
| "logits/chosen": 0.28656622767448425, | |
| "logits/rejected": 0.394423246383667, | |
| "logps/chosen": -983.4553833007812, | |
| "logps/rejected": -1191.291748046875, | |
| "loss": 1.897, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.3752153813838959, | |
| "rewards/margins": 0.2264389544725418, | |
| "rewards/rejected": -0.6016542911529541, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 6.212661423609184e-08, | |
| "logits/chosen": 0.33070552349090576, | |
| "logits/rejected": 0.4134686589241028, | |
| "logps/chosen": -823.2513427734375, | |
| "logps/rejected": -1224.76318359375, | |
| "loss": 1.4644, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -0.2675178647041321, | |
| "rewards/margins": 0.37183278799057007, | |
| "rewards/rejected": -0.6393505930900574, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 4.761211162702117e-08, | |
| "logits/chosen": 0.30609947443008423, | |
| "logits/rejected": 0.3751484751701355, | |
| "logps/chosen": -889.3785400390625, | |
| "logps/rejected": -1164.7288818359375, | |
| "loss": 1.9301, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.33601176738739014, | |
| "rewards/margins": 0.24511654675006866, | |
| "rewards/rejected": -0.5811282396316528, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 3.5009907323737826e-08, | |
| "logits/chosen": 0.3077315390110016, | |
| "logits/rejected": 0.32115936279296875, | |
| "logps/chosen": -940.2276611328125, | |
| "logps/rejected": -1252.099853515625, | |
| "loss": 1.6158, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -0.3381388485431671, | |
| "rewards/margins": 0.2894209623336792, | |
| "rewards/rejected": -0.6275597810745239, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 2.4329828146074096e-08, | |
| "logits/chosen": 0.3073035478591919, | |
| "logits/rejected": 0.4198598265647888, | |
| "logps/chosen": -872.7296752929688, | |
| "logps/rejected": -1125.4498291015625, | |
| "loss": 1.7422, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -0.30542269349098206, | |
| "rewards/margins": 0.29255709052085876, | |
| "rewards/rejected": -0.5979797840118408, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "learning_rate": 1.5580202098509078e-08, | |
| "logits/chosen": 0.31831032037734985, | |
| "logits/rejected": 0.4222412109375, | |
| "logps/chosen": -889.1136474609375, | |
| "logps/rejected": -1244.5892333984375, | |
| "loss": 1.7276, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.31571874022483826, | |
| "rewards/margins": 0.3139411509037018, | |
| "rewards/rejected": -0.62965989112854, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 8.767851876239075e-09, | |
| "logits/chosen": 0.2873593270778656, | |
| "logits/rejected": 0.42444127798080444, | |
| "logps/chosen": -910.4513549804688, | |
| "logps/rejected": -1105.3863525390625, | |
| "loss": 1.6767, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -0.29710885882377625, | |
| "rewards/margins": 0.2506183981895447, | |
| "rewards/rejected": -0.5477272272109985, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 3.8980895450474455e-09, | |
| "logits/chosen": 0.32995566725730896, | |
| "logits/rejected": 0.36712411046028137, | |
| "logps/chosen": -900.4547729492188, | |
| "logps/rejected": -1215.38134765625, | |
| "loss": 1.764, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.30327844619750977, | |
| "rewards/margins": 0.2899893820285797, | |
| "rewards/rejected": -0.5932677984237671, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 9.747123991141193e-10, | |
| "logits/chosen": 0.30442702770233154, | |
| "logits/rejected": 0.3750511407852173, | |
| "logps/chosen": -829.1936645507812, | |
| "logps/rejected": -1163.27783203125, | |
| "loss": 1.8085, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.3106636106967926, | |
| "rewards/margins": 0.2759193778038025, | |
| "rewards/rejected": -0.5865829586982727, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 0.0, | |
| "logits/chosen": 0.3020361661911011, | |
| "logits/rejected": 0.362225204706192, | |
| "logps/chosen": -867.2174682617188, | |
| "logps/rejected": -1196.6907958984375, | |
| "loss": 1.7083, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -0.3051598370075226, | |
| "rewards/margins": 0.2913387715816498, | |
| "rewards/rejected": -0.5964986085891724, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 1250, | |
| "total_flos": 0.0, | |
| "train_loss": 1.8189717344284058, | |
| "train_runtime": 11934.9419, | |
| "train_samples_per_second": 1.676, | |
| "train_steps_per_second": 0.105 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1250, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "total_flos": 0.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |