| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9997382884061764, | |
| "eval_steps": 100, | |
| "global_step": 1910, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.005234231876472127, | |
| "grad_norm": 57.10669050657411, | |
| "learning_rate": 2.094240837696335e-08, | |
| "logits/chosen": -0.9002814292907715, | |
| "logits/rejected": -0.9369659423828125, | |
| "logps/chosen": -1.2799328565597534, | |
| "logps/rejected": -1.224565863609314, | |
| "loss": 1.3867, | |
| "rewards/accuracies": 0.4625000059604645, | |
| "rewards/chosen": -0.002279318869113922, | |
| "rewards/margins": 0.01926611177623272, | |
| "rewards/rejected": -0.02154543064534664, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.010468463752944255, | |
| "grad_norm": 205.46189868419341, | |
| "learning_rate": 4.18848167539267e-08, | |
| "logits/chosen": -0.9780851602554321, | |
| "logits/rejected": -0.9534770250320435, | |
| "logps/chosen": -1.2819010019302368, | |
| "logps/rejected": -1.324920415878296, | |
| "loss": 1.3854, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.018570536747574806, | |
| "rewards/margins": 0.012391218915581703, | |
| "rewards/rejected": 0.006179317831993103, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.015702695629416383, | |
| "grad_norm": 424.88597498836, | |
| "learning_rate": 6.282722513089005e-08, | |
| "logits/chosen": -0.9846030473709106, | |
| "logits/rejected": -1.001585841178894, | |
| "logps/chosen": -1.3621526956558228, | |
| "logps/rejected": -1.4822871685028076, | |
| "loss": 1.3862, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": 0.006618998944759369, | |
| "rewards/margins": 0.005857313051819801, | |
| "rewards/rejected": 0.0007616858929395676, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.02093692750588851, | |
| "grad_norm": 111.87617735661307, | |
| "learning_rate": 8.37696335078534e-08, | |
| "logits/chosen": -0.9275991320610046, | |
| "logits/rejected": -1.0113765001296997, | |
| "logps/chosen": -1.279705286026001, | |
| "logps/rejected": -1.6348555088043213, | |
| "loss": 1.3855, | |
| "rewards/accuracies": 0.4375, | |
| "rewards/chosen": -0.006214768625795841, | |
| "rewards/margins": -0.005492387805134058, | |
| "rewards/rejected": -0.0007223807042464614, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.02617115938236064, | |
| "grad_norm": 73.6668063718139, | |
| "learning_rate": 1.0471204188481675e-07, | |
| "logits/chosen": -0.8899806141853333, | |
| "logits/rejected": -0.9631911516189575, | |
| "logps/chosen": -1.2982203960418701, | |
| "logps/rejected": -1.5544034242630005, | |
| "loss": 1.3849, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": 0.00998431071639061, | |
| "rewards/margins": 0.011886270716786385, | |
| "rewards/rejected": -0.0019019600003957748, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.031405391258832765, | |
| "grad_norm": 116.30393708461156, | |
| "learning_rate": 1.256544502617801e-07, | |
| "logits/chosen": -0.9441679120063782, | |
| "logits/rejected": -0.965167224407196, | |
| "logps/chosen": -1.3546206951141357, | |
| "logps/rejected": -1.517407774925232, | |
| "loss": 1.3838, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": 0.027701353654265404, | |
| "rewards/margins": 0.03188776224851608, | |
| "rewards/rejected": -0.004186409059911966, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.036639623135304895, | |
| "grad_norm": 2187.360162764326, | |
| "learning_rate": 1.4659685863874346e-07, | |
| "logits/chosen": -0.8137859106063843, | |
| "logits/rejected": -0.9285731315612793, | |
| "logps/chosen": -1.251236081123352, | |
| "logps/rejected": -1.5110037326812744, | |
| "loss": 1.3836, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.01799890398979187, | |
| "rewards/margins": 0.022558193653821945, | |
| "rewards/rejected": -0.040557097643613815, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.04187385501177702, | |
| "grad_norm": 63.631796686270974, | |
| "learning_rate": 1.675392670157068e-07, | |
| "logits/chosen": -0.9329401850700378, | |
| "logits/rejected": -0.9414553642272949, | |
| "logps/chosen": -1.2446186542510986, | |
| "logps/rejected": -1.329633355140686, | |
| "loss": 1.3799, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.034679971635341644, | |
| "rewards/margins": 0.031586576253175735, | |
| "rewards/rejected": -0.06626654416322708, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.04710808688824915, | |
| "grad_norm": 70.57364025424204, | |
| "learning_rate": 1.8848167539267015e-07, | |
| "logits/chosen": -0.975296676158905, | |
| "logits/rejected": -0.975991427898407, | |
| "logps/chosen": -1.411871314048767, | |
| "logps/rejected": -1.4917197227478027, | |
| "loss": 1.3838, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.03906696289777756, | |
| "rewards/margins": 0.0577811673283577, | |
| "rewards/rejected": -0.09684813022613525, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.05234231876472128, | |
| "grad_norm": 12165.55527932566, | |
| "learning_rate": 2.094240837696335e-07, | |
| "logits/chosen": -0.9536153078079224, | |
| "logits/rejected": -1.0614324808120728, | |
| "logps/chosen": -1.387503743171692, | |
| "logps/rejected": -1.6376310586929321, | |
| "loss": 1.3855, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.07099077105522156, | |
| "rewards/margins": 0.1324595957994461, | |
| "rewards/rejected": -0.20345036685466766, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.05234231876472128, | |
| "eval_logits/chosen": -1.0141676664352417, | |
| "eval_logits/rejected": -1.0460058450698853, | |
| "eval_logps/chosen": -1.2700459957122803, | |
| "eval_logps/rejected": -1.4935321807861328, | |
| "eval_loss": 1.3827202320098877, | |
| "eval_rewards/accuracies": 0.6448412537574768, | |
| "eval_rewards/chosen": 0.01650167442858219, | |
| "eval_rewards/margins": 0.07835288345813751, | |
| "eval_rewards/rejected": -0.06185121089220047, | |
| "eval_runtime": 264.6269, | |
| "eval_samples_per_second": 7.558, | |
| "eval_steps_per_second": 0.238, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.05757655064119341, | |
| "grad_norm": 191.0826517024898, | |
| "learning_rate": 2.3036649214659686e-07, | |
| "logits/chosen": -0.8402193188667297, | |
| "logits/rejected": -0.9604326486587524, | |
| "logps/chosen": -1.3015209436416626, | |
| "logps/rejected": -1.4067150354385376, | |
| "loss": 1.3805, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": 0.010574941523373127, | |
| "rewards/margins": 0.0737369954586029, | |
| "rewards/rejected": -0.06316206604242325, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.06281078251766553, | |
| "grad_norm": 105.26964983422378, | |
| "learning_rate": 2.513089005235602e-07, | |
| "logits/chosen": -0.8950595855712891, | |
| "logits/rejected": -0.9546734690666199, | |
| "logps/chosen": -1.2874231338500977, | |
| "logps/rejected": -1.275660753250122, | |
| "loss": 1.4167, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": 0.02625320851802826, | |
| "rewards/margins": 0.05933423712849617, | |
| "rewards/rejected": -0.03308102488517761, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.06804501439413765, | |
| "grad_norm": 98.23082684726371, | |
| "learning_rate": 2.7225130890052355e-07, | |
| "logits/chosen": -0.8970452547073364, | |
| "logits/rejected": -0.9690724611282349, | |
| "logps/chosen": -1.2717323303222656, | |
| "logps/rejected": -1.5841439962387085, | |
| "loss": 1.3711, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.016156885772943497, | |
| "rewards/margins": 0.16786156594753265, | |
| "rewards/rejected": -0.15170469880104065, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.07327924627060979, | |
| "grad_norm": 105.57197537487045, | |
| "learning_rate": 2.931937172774869e-07, | |
| "logits/chosen": -1.0700782537460327, | |
| "logits/rejected": -1.0342130661010742, | |
| "logps/chosen": -1.399418830871582, | |
| "logps/rejected": -1.4947900772094727, | |
| "loss": 1.3747, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.007722643204033375, | |
| "rewards/margins": 0.1320706158876419, | |
| "rewards/rejected": -0.1397932469844818, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.07851347814708191, | |
| "grad_norm": 92.95205585650491, | |
| "learning_rate": 3.1413612565445027e-07, | |
| "logits/chosen": -0.8677465319633484, | |
| "logits/rejected": -0.9042151570320129, | |
| "logps/chosen": -1.2097841501235962, | |
| "logps/rejected": -1.4593037366867065, | |
| "loss": 1.3624, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.037995148450136185, | |
| "rewards/margins": 0.11828543990850449, | |
| "rewards/rejected": -0.15628059208393097, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.08374771002355404, | |
| "grad_norm": 60.18689578296943, | |
| "learning_rate": 3.350785340314136e-07, | |
| "logits/chosen": -0.8861294984817505, | |
| "logits/rejected": -0.954685389995575, | |
| "logps/chosen": -1.2184758186340332, | |
| "logps/rejected": -1.4243767261505127, | |
| "loss": 1.3659, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.012581342831254005, | |
| "rewards/margins": 0.19156894087791443, | |
| "rewards/rejected": -0.2041502743959427, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.08898194190002617, | |
| "grad_norm": 162.74421304335266, | |
| "learning_rate": 3.56020942408377e-07, | |
| "logits/chosen": -0.9357515573501587, | |
| "logits/rejected": -0.9549610018730164, | |
| "logps/chosen": -1.3333146572113037, | |
| "logps/rejected": -1.5601656436920166, | |
| "loss": 1.3878, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.03334398195147514, | |
| "rewards/margins": 0.2373563051223755, | |
| "rewards/rejected": -0.20401231944561005, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.0942161737764983, | |
| "grad_norm": 303.36378167257243, | |
| "learning_rate": 3.769633507853403e-07, | |
| "logits/chosen": -1.0010288953781128, | |
| "logits/rejected": -1.0621322393417358, | |
| "logps/chosen": -1.3325862884521484, | |
| "logps/rejected": -1.6137030124664307, | |
| "loss": 1.3917, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": 0.10068617761135101, | |
| "rewards/margins": 0.23769037425518036, | |
| "rewards/rejected": -0.13700421154499054, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.09945040565297043, | |
| "grad_norm": 677.0727123683995, | |
| "learning_rate": 3.9790575916230365e-07, | |
| "logits/chosen": -0.9208686947822571, | |
| "logits/rejected": -1.0320428609848022, | |
| "logps/chosen": -1.3446584939956665, | |
| "logps/rejected": -1.5973131656646729, | |
| "loss": 1.3741, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": 0.060563139617443085, | |
| "rewards/margins": 0.27492621541023254, | |
| "rewards/rejected": -0.21436305344104767, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.10468463752944256, | |
| "grad_norm": 362.00771368392924, | |
| "learning_rate": 3.9997294651491985e-07, | |
| "logits/chosen": -0.8778219223022461, | |
| "logits/rejected": -0.8942776918411255, | |
| "logps/chosen": -1.2816615104675293, | |
| "logps/rejected": -1.3702523708343506, | |
| "loss": 1.4322, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": 0.046035002917051315, | |
| "rewards/margins": 0.17283782362937927, | |
| "rewards/rejected": -0.12680283188819885, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.10468463752944256, | |
| "eval_logits/chosen": -1.00635826587677, | |
| "eval_logits/rejected": -1.0356963872909546, | |
| "eval_logps/chosen": -1.2760363817214966, | |
| "eval_logps/rejected": -1.509326696395874, | |
| "eval_loss": 1.3940050601959229, | |
| "eval_rewards/accuracies": 0.7003968358039856, | |
| "eval_rewards/chosen": -0.073353111743927, | |
| "eval_rewards/margins": 0.2254164069890976, | |
| "eval_rewards/rejected": -0.2987695634365082, | |
| "eval_runtime": 263.6711, | |
| "eval_samples_per_second": 7.585, | |
| "eval_steps_per_second": 0.239, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.10991886940591468, | |
| "grad_norm": 777.6134451437069, | |
| "learning_rate": 3.9987943769122714e-07, | |
| "logits/chosen": -0.9393421411514282, | |
| "logits/rejected": -0.9751921892166138, | |
| "logps/chosen": -1.2527530193328857, | |
| "logps/rejected": -1.512480616569519, | |
| "loss": 1.4172, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.14303788542747498, | |
| "rewards/margins": 0.22242093086242676, | |
| "rewards/rejected": -0.36545881628990173, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.11515310128238682, | |
| "grad_norm": 325.44094651913406, | |
| "learning_rate": 3.997191707590292e-07, | |
| "logits/chosen": -0.9741231799125671, | |
| "logits/rejected": -0.9654073715209961, | |
| "logps/chosen": -1.4455256462097168, | |
| "logps/rejected": -1.4085519313812256, | |
| "loss": 1.3963, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": 0.06883502006530762, | |
| "rewards/margins": 0.3231067359447479, | |
| "rewards/rejected": -0.2542716860771179, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.12038733315885894, | |
| "grad_norm": 619.1952484265515, | |
| "learning_rate": 3.9949219924617967e-07, | |
| "logits/chosen": -0.9289252161979675, | |
| "logits/rejected": -0.9729310870170593, | |
| "logps/chosen": -1.2874866724014282, | |
| "logps/rejected": -1.5892341136932373, | |
| "loss": 1.4192, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.08541570603847504, | |
| "rewards/margins": 0.2191784679889679, | |
| "rewards/rejected": -0.3045941889286041, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.12562156503533106, | |
| "grad_norm": 361.72441654521475, | |
| "learning_rate": 3.9919859895932e-07, | |
| "logits/chosen": -0.9237004518508911, | |
| "logits/rejected": -1.0107048749923706, | |
| "logps/chosen": -1.3676064014434814, | |
| "logps/rejected": -1.7672977447509766, | |
| "loss": 1.4555, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": 0.15573230385780334, | |
| "rewards/margins": 0.4071730673313141, | |
| "rewards/rejected": -0.25144073367118835, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.13085579691180318, | |
| "grad_norm": 107.37570680791188, | |
| "learning_rate": 3.988384679585609e-07, | |
| "logits/chosen": -0.9424523115158081, | |
| "logits/rejected": -0.9789683222770691, | |
| "logps/chosen": -1.3719688653945923, | |
| "logps/rejected": -1.476319670677185, | |
| "loss": 1.4147, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": 0.12189966440200806, | |
| "rewards/margins": 0.20880040526390076, | |
| "rewards/rejected": -0.0869007408618927, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.1360900287882753, | |
| "grad_norm": 95.00301311323727, | |
| "learning_rate": 3.9841192652473133e-07, | |
| "logits/chosen": -0.9747894406318665, | |
| "logits/rejected": -1.0530925989151, | |
| "logps/chosen": -1.266790509223938, | |
| "logps/rejected": -1.4995120763778687, | |
| "loss": 1.43, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": 0.008727884851396084, | |
| "rewards/margins": 0.2547326385974884, | |
| "rewards/rejected": -0.24600473046302795, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.14132426066474746, | |
| "grad_norm": 297.54395880852655, | |
| "learning_rate": 3.979191171192052e-07, | |
| "logits/chosen": -0.9566612243652344, | |
| "logits/rejected": -0.978779673576355, | |
| "logps/chosen": -1.352007508277893, | |
| "logps/rejected": -1.5359153747558594, | |
| "loss": 1.4378, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 0.09225025027990341, | |
| "rewards/margins": 0.26082533597946167, | |
| "rewards/rejected": -0.16857507824897766, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.14655849254121958, | |
| "grad_norm": 152.14462193643473, | |
| "learning_rate": 3.973602043363207e-07, | |
| "logits/chosen": -1.0354989767074585, | |
| "logits/rejected": -1.0463378429412842, | |
| "logps/chosen": -1.391145944595337, | |
| "logps/rejected": -1.4438835382461548, | |
| "loss": 1.4181, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": 0.12363862991333008, | |
| "rewards/margins": 0.4137144684791565, | |
| "rewards/rejected": -0.2900758385658264, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.1517927244176917, | |
| "grad_norm": 150.78955202749123, | |
| "learning_rate": 3.9673537484840704e-07, | |
| "logits/chosen": -0.956143856048584, | |
| "logits/rejected": -1.0358890295028687, | |
| "logps/chosen": -1.398080587387085, | |
| "logps/rejected": -1.4760212898254395, | |
| "loss": 1.4314, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -0.0525251105427742, | |
| "rewards/margins": 0.2129652500152588, | |
| "rewards/rejected": -0.2654903531074524, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.15702695629416383, | |
| "grad_norm": 175.4681648073153, | |
| "learning_rate": 3.960448373434375e-07, | |
| "logits/chosen": -0.9618331789970398, | |
| "logits/rejected": -1.0131045579910278, | |
| "logps/chosen": -1.219518780708313, | |
| "logps/rejected": -1.6010315418243408, | |
| "loss": 1.4033, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 0.10120991617441177, | |
| "rewards/margins": 0.5264266729354858, | |
| "rewards/rejected": -0.42521676421165466, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.15702695629416383, | |
| "eval_logits/chosen": -1.0152956247329712, | |
| "eval_logits/rejected": -1.0432747602462769, | |
| "eval_logps/chosen": -1.2684417963027954, | |
| "eval_logps/rejected": -1.5052709579467773, | |
| "eval_loss": 1.407413363456726, | |
| "eval_rewards/accuracies": 0.7103174328804016, | |
| "eval_rewards/chosen": 0.04056532308459282, | |
| "eval_rewards/margins": 0.27849799394607544, | |
| "eval_rewards/rejected": -0.23793263733386993, | |
| "eval_runtime": 264.1297, | |
| "eval_samples_per_second": 7.572, | |
| "eval_steps_per_second": 0.239, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.16226118817063595, | |
| "grad_norm": 249.4339583636602, | |
| "learning_rate": 3.9528882245532945e-07, | |
| "logits/chosen": -0.9384096264839172, | |
| "logits/rejected": -0.9961991310119629, | |
| "logps/chosen": -1.3366284370422363, | |
| "logps/rejected": -1.5762803554534912, | |
| "loss": 1.4132, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": 0.06849001348018646, | |
| "rewards/margins": 0.22860321402549744, | |
| "rewards/rejected": -0.16011318564414978, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.16749542004710807, | |
| "grad_norm": 100.66710833696462, | |
| "learning_rate": 3.9446758268691394e-07, | |
| "logits/chosen": -0.9083954095840454, | |
| "logits/rejected": -1.0078755617141724, | |
| "logps/chosen": -1.3704053163528442, | |
| "logps/rejected": -1.5062158107757568, | |
| "loss": 1.4061, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": 0.11326809227466583, | |
| "rewards/margins": 0.3582002520561218, | |
| "rewards/rejected": -0.2449321746826172, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.17272965192358022, | |
| "grad_norm": 93.62113614305316, | |
| "learning_rate": 3.935813923256026e-07, | |
| "logits/chosen": -0.9598302841186523, | |
| "logits/rejected": -1.1092036962509155, | |
| "logps/chosen": -1.197092890739441, | |
| "logps/rejected": -1.5259929895401, | |
| "loss": 1.4512, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": 0.29086190462112427, | |
| "rewards/margins": 0.46881595253944397, | |
| "rewards/rejected": -0.1779540628194809, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.17796388380005235, | |
| "grad_norm": 1259.8554969874265, | |
| "learning_rate": 3.9263054735177724e-07, | |
| "logits/chosen": -0.9662519693374634, | |
| "logits/rejected": -0.9433367848396301, | |
| "logps/chosen": -1.3681367635726929, | |
| "logps/rejected": -1.514211654663086, | |
| "loss": 1.4474, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.054476749151945114, | |
| "rewards/margins": 0.17750175297260284, | |
| "rewards/rejected": -0.23197850584983826, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.18319811567652447, | |
| "grad_norm": 157.20223175549822, | |
| "learning_rate": 3.916153653399351e-07, | |
| "logits/chosen": -0.9618955850601196, | |
| "logits/rejected": -0.965499758720398, | |
| "logps/chosen": -1.2811925411224365, | |
| "logps/rejected": -1.559472680091858, | |
| "loss": 1.4054, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -0.09336249530315399, | |
| "rewards/margins": 0.42978915572166443, | |
| "rewards/rejected": -0.5231517553329468, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.1884323475529966, | |
| "grad_norm": 1304.5710427907864, | |
| "learning_rate": 3.9053618535262144e-07, | |
| "logits/chosen": -0.9978113174438477, | |
| "logits/rejected": -1.0920830965042114, | |
| "logps/chosen": -1.295592188835144, | |
| "logps/rejected": -1.5009433031082153, | |
| "loss": 1.4453, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.11287063360214233, | |
| "rewards/margins": 0.2577170133590698, | |
| "rewards/rejected": -0.14484639465808868, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.19366657942946872, | |
| "grad_norm": 418.65708160055283, | |
| "learning_rate": 3.893933678271856e-07, | |
| "logits/chosen": -1.0005239248275757, | |
| "logits/rejected": -1.0983339548110962, | |
| "logps/chosen": -1.2382128238677979, | |
| "logps/rejected": -1.5115060806274414, | |
| "loss": 1.4038, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": 0.09283250570297241, | |
| "rewards/margins": 0.5117446780204773, | |
| "rewards/rejected": -0.4189121723175049, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.19890081130594087, | |
| "grad_norm": 193.74911912257855, | |
| "learning_rate": 3.881872944553976e-07, | |
| "logits/chosen": -1.0515316724777222, | |
| "logits/rejected": -1.0202170610427856, | |
| "logps/chosen": -1.5801050662994385, | |
| "logps/rejected": -1.6806576251983643, | |
| "loss": 1.4325, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": 0.18667231500148773, | |
| "rewards/margins": 0.5009941458702087, | |
| "rewards/rejected": -0.31432193517684937, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.204135043182413, | |
| "grad_norm": 164.76136107173468, | |
| "learning_rate": 3.869183680559662e-07, | |
| "logits/chosen": -0.9525713920593262, | |
| "logits/rejected": -1.025831699371338, | |
| "logps/chosen": -1.4257429838180542, | |
| "logps/rejected": -1.517421841621399, | |
| "loss": 1.465, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": 0.1658320128917694, | |
| "rewards/margins": 0.2568032145500183, | |
| "rewards/rejected": -0.09097117930650711, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.2093692750588851, | |
| "grad_norm": 132.36602040912516, | |
| "learning_rate": 3.8558701244000107e-07, | |
| "logits/chosen": -1.0163413286209106, | |
| "logits/rejected": -1.12177312374115, | |
| "logps/chosen": -1.2447032928466797, | |
| "logps/rejected": -1.5292352437973022, | |
| "loss": 1.4072, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.006441372446715832, | |
| "rewards/margins": 0.40031346678733826, | |
| "rewards/rejected": -0.40675482153892517, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.2093692750588851, | |
| "eval_logits/chosen": -1.0346137285232544, | |
| "eval_logits/rejected": -1.0629608631134033, | |
| "eval_logps/chosen": -1.2605527639389038, | |
| "eval_logps/rejected": -1.4991511106491089, | |
| "eval_loss": 1.4382668733596802, | |
| "eval_rewards/accuracies": 0.7222222089767456, | |
| "eval_rewards/chosen": 0.15890030562877655, | |
| "eval_rewards/margins": 0.3050336241722107, | |
| "eval_rewards/rejected": -0.14613336324691772, | |
| "eval_runtime": 269.1306, | |
| "eval_samples_per_second": 7.431, | |
| "eval_steps_per_second": 0.234, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.21460350693535724, | |
| "grad_norm": 277.3029855825092, | |
| "learning_rate": 3.841936722694628e-07, | |
| "logits/chosen": -0.9349578619003296, | |
| "logits/rejected": -1.000216007232666, | |
| "logps/chosen": -1.3645018339157104, | |
| "logps/rejected": -1.5343906879425049, | |
| "loss": 1.4552, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.2753569781780243, | |
| "rewards/margins": 0.29162487387657166, | |
| "rewards/rejected": -0.01626790128648281, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.21983773881182936, | |
| "grad_norm": 323.6630242480252, | |
| "learning_rate": 3.8273881290864986e-07, | |
| "logits/chosen": -0.9792436361312866, | |
| "logits/rejected": -1.0122894048690796, | |
| "logps/chosen": -1.2483856678009033, | |
| "logps/rejected": -1.410635232925415, | |
| "loss": 1.4225, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.008017847314476967, | |
| "rewards/margins": 0.28319111466407776, | |
| "rewards/rejected": -0.2912089228630066, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.22507197068830148, | |
| "grad_norm": 86.80607700867975, | |
| "learning_rate": 3.812229202687705e-07, | |
| "logits/chosen": -0.9529207944869995, | |
| "logits/rejected": -1.045686960220337, | |
| "logps/chosen": -1.2583253383636475, | |
| "logps/rejected": -1.5180001258850098, | |
| "loss": 1.4296, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": 0.12871792912483215, | |
| "rewards/margins": 0.5155263543128967, | |
| "rewards/rejected": -0.3868083655834198, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.23030620256477363, | |
| "grad_norm": 915.1354551049579, | |
| "learning_rate": 3.796465006456523e-07, | |
| "logits/chosen": -1.0024458169937134, | |
| "logits/rejected": -1.068861961364746, | |
| "logps/chosen": -1.277616024017334, | |
| "logps/rejected": -1.4234261512756348, | |
| "loss": 1.4226, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": 0.05915598198771477, | |
| "rewards/margins": 0.5244419574737549, | |
| "rewards/rejected": -0.4652860760688782, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.23554043444124576, | |
| "grad_norm": 218.08119707377622, | |
| "learning_rate": 3.7801008055064363e-07, | |
| "logits/chosen": -0.9519055485725403, | |
| "logits/rejected": -0.963157057762146, | |
| "logps/chosen": -1.3588078022003174, | |
| "logps/rejected": -1.5339311361312866, | |
| "loss": 1.3993, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": 0.24848918616771698, | |
| "rewards/margins": 0.4637584686279297, | |
| "rewards/rejected": -0.2152692824602127, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.24077466631771788, | |
| "grad_norm": 99.08327892137484, | |
| "learning_rate": 3.7631420653476316e-07, | |
| "logits/chosen": -0.9202947616577148, | |
| "logits/rejected": -1.0152450799942017, | |
| "logps/chosen": -1.2867461442947388, | |
| "logps/rejected": -1.4801595211029053, | |
| "loss": 1.4462, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -0.1470358669757843, | |
| "rewards/margins": 0.279130756855011, | |
| "rewards/rejected": -0.4261665940284729, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.24600889819419, | |
| "grad_norm": 228.85559080922738, | |
| "learning_rate": 3.74559445006156e-07, | |
| "logits/chosen": -0.9630579948425293, | |
| "logits/rejected": -0.9651592373847961, | |
| "logps/chosen": -1.393817663192749, | |
| "logps/rejected": -1.5304372310638428, | |
| "loss": 1.3859, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": 0.1274794042110443, | |
| "rewards/margins": 0.317279577255249, | |
| "rewards/rejected": -0.1898001730442047, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.2512431300706621, | |
| "grad_norm": 202.93750821371427, | |
| "learning_rate": 3.727463820409182e-07, | |
| "logits/chosen": -0.9812225103378296, | |
| "logits/rejected": -1.044806718826294, | |
| "logps/chosen": -1.4120748043060303, | |
| "logps/rejected": -1.6228997707366943, | |
| "loss": 1.4094, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": 0.017949607223272324, | |
| "rewards/margins": 0.22776713967323303, | |
| "rewards/rejected": -0.2098175287246704, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.2564773619471343, | |
| "grad_norm": 166.1560506458322, | |
| "learning_rate": 3.7087562318735215e-07, | |
| "logits/chosen": -0.9065971374511719, | |
| "logits/rejected": -0.9788573384284973, | |
| "logps/chosen": -1.269921064376831, | |
| "logps/rejected": -1.3954648971557617, | |
| "loss": 1.4468, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": 0.12124637514352798, | |
| "rewards/margins": 0.16167902946472168, | |
| "rewards/rejected": -0.04043266549706459, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.26171159382360637, | |
| "grad_norm": 198.70593003367702, | |
| "learning_rate": 3.6894779326371806e-07, | |
| "logits/chosen": -0.9271315336227417, | |
| "logits/rejected": -0.9706541299819946, | |
| "logps/chosen": -1.3781875371932983, | |
| "logps/rejected": -1.5902111530303955, | |
| "loss": 1.4887, | |
| "rewards/accuracies": 0.7875000238418579, | |
| "rewards/chosen": 0.14317944645881653, | |
| "rewards/margins": 0.4533039927482605, | |
| "rewards/rejected": -0.3101245164871216, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.26171159382360637, | |
| "eval_logits/chosen": -1.058478593826294, | |
| "eval_logits/rejected": -1.0874103307724, | |
| "eval_logps/chosen": -1.2464978694915771, | |
| "eval_logps/rejected": -1.4891070127487183, | |
| "eval_loss": 1.5037912130355835, | |
| "eval_rewards/accuracies": 0.704365074634552, | |
| "eval_rewards/chosen": 0.3697235584259033, | |
| "eval_rewards/margins": 0.36519646644592285, | |
| "eval_rewards/rejected": 0.00452705891802907, | |
| "eval_runtime": 265.1463, | |
| "eval_samples_per_second": 7.543, | |
| "eval_steps_per_second": 0.238, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.2669458257000785, | |
| "grad_norm": 262.1381649197113, | |
| "learning_rate": 3.669635361495502e-07, | |
| "logits/chosen": -1.0345722436904907, | |
| "logits/rejected": -1.1255292892456055, | |
| "logps/chosen": -1.2064478397369385, | |
| "logps/rejected": -1.574259877204895, | |
| "loss": 1.4465, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": 0.039454132318496704, | |
| "rewards/margins": 0.6132072806358337, | |
| "rewards/rejected": -0.5737532377243042, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.2721800575765506, | |
| "grad_norm": 89.6307110887034, | |
| "learning_rate": 3.6492351457060587e-07, | |
| "logits/chosen": -0.9336854219436646, | |
| "logits/rejected": -0.965878963470459, | |
| "logps/chosen": -1.1347582340240479, | |
| "logps/rejected": -1.5124342441558838, | |
| "loss": 1.4152, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.026898790150880814, | |
| "rewards/margins": 0.5120420455932617, | |
| "rewards/rejected": -0.5389407873153687, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.27741428945302277, | |
| "grad_norm": 1388.4179719079036, | |
| "learning_rate": 3.6282840987752065e-07, | |
| "logits/chosen": -1.0223872661590576, | |
| "logits/rejected": -1.0572293996810913, | |
| "logps/chosen": -1.3843199014663696, | |
| "logps/rejected": -1.625836968421936, | |
| "loss": 1.475, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": 0.04534434527158737, | |
| "rewards/margins": 0.48546886444091797, | |
| "rewards/rejected": -0.4401245713233948, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.2826485213294949, | |
| "grad_norm": 195.53670379895084, | |
| "learning_rate": 3.606789218182429e-07, | |
| "logits/chosen": -1.0764001607894897, | |
| "logits/rejected": -1.1225221157073975, | |
| "logps/chosen": -1.5096803903579712, | |
| "logps/rejected": -1.679340124130249, | |
| "loss": 1.431, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.07530850172042847, | |
| "rewards/margins": 0.32085931301116943, | |
| "rewards/rejected": -0.3961678147315979, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.287882753205967, | |
| "grad_norm": 127.17060631430301, | |
| "learning_rate": 3.584757683043235e-07, | |
| "logits/chosen": -0.9253309965133667, | |
| "logits/rejected": -0.9592300653457642, | |
| "logps/chosen": -1.2050528526306152, | |
| "logps/rejected": -1.4214028120040894, | |
| "loss": 1.4257, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": 0.09262965619564056, | |
| "rewards/margins": 0.39863070845603943, | |
| "rewards/rejected": -0.30600103735923767, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.29311698508243916, | |
| "grad_norm": 119.21787519480398, | |
| "learning_rate": 3.5621968517113905e-07, | |
| "logits/chosen": -1.011732816696167, | |
| "logits/rejected": -1.074186086654663, | |
| "logps/chosen": -1.367538571357727, | |
| "logps/rejected": -1.4972589015960693, | |
| "loss": 1.4233, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": 0.16185715794563293, | |
| "rewards/margins": 0.49168023467063904, | |
| "rewards/rejected": -0.3298230767250061, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.29835121695891126, | |
| "grad_norm": 131.29687077414553, | |
| "learning_rate": 3.5391142593212927e-07, | |
| "logits/chosen": -0.9979642629623413, | |
| "logits/rejected": -1.066590666770935, | |
| "logps/chosen": -1.3047949075698853, | |
| "logps/rejected": -1.441133737564087, | |
| "loss": 1.4295, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": 0.03215108811855316, | |
| "rewards/margins": 0.302137553691864, | |
| "rewards/rejected": -0.26998645067214966, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.3035854488353834, | |
| "grad_norm": 794.2192746772438, | |
| "learning_rate": 3.515517615271293e-07, | |
| "logits/chosen": -0.9547045826911926, | |
| "logits/rejected": -1.0005202293395996, | |
| "logps/chosen": -1.1787316799163818, | |
| "logps/rejected": -1.4146497249603271, | |
| "loss": 1.4159, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": 0.23103637993335724, | |
| "rewards/margins": 0.48520785570144653, | |
| "rewards/rejected": -0.2541714906692505, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.30881968071185556, | |
| "grad_norm": 114.48798800720019, | |
| "learning_rate": 3.4914148006488197e-07, | |
| "logits/chosen": -0.9228054881095886, | |
| "logits/rejected": -0.9928166270256042, | |
| "logps/chosen": -1.314188003540039, | |
| "logps/rejected": -1.7351295948028564, | |
| "loss": 1.4898, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.02696867287158966, | |
| "rewards/margins": -0.19175395369529724, | |
| "rewards/rejected": 0.1647852510213852, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.31405391258832765, | |
| "grad_norm": 128.958271520903, | |
| "learning_rate": 3.466813865598163e-07, | |
| "logits/chosen": -0.9575554132461548, | |
| "logits/rejected": -1.0143183469772339, | |
| "logps/chosen": -1.3230020999908447, | |
| "logps/rejected": -1.4859802722930908, | |
| "loss": 1.4435, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": 0.171253964304924, | |
| "rewards/margins": 0.0992809310555458, | |
| "rewards/rejected": 0.07197302579879761, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.31405391258832765, | |
| "eval_logits/chosen": -1.0527211427688599, | |
| "eval_logits/rejected": -1.0806084871292114, | |
| "eval_logps/chosen": -1.2730886936187744, | |
| "eval_logps/rejected": -1.5170137882232666, | |
| "eval_loss": 1.4242910146713257, | |
| "eval_rewards/accuracies": 0.72817462682724, | |
| "eval_rewards/chosen": -0.029136493802070618, | |
| "eval_rewards/margins": 0.3849383294582367, | |
| "eval_rewards/rejected": -0.4140748083591461, | |
| "eval_runtime": 266.4809, | |
| "eval_samples_per_second": 7.505, | |
| "eval_steps_per_second": 0.236, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.3192881444647998, | |
| "grad_norm": 221.61630470246598, | |
| "learning_rate": 3.4417230266317886e-07, | |
| "logits/chosen": -0.9701067805290222, | |
| "logits/rejected": -1.0397741794586182, | |
| "logps/chosen": -1.228003740310669, | |
| "logps/rejected": -1.4498927593231201, | |
| "loss": 1.4383, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.038508955389261246, | |
| "rewards/margins": 0.4932937026023865, | |
| "rewards/rejected": -0.5318026542663574, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.3245223763412719, | |
| "grad_norm": 233.4642222088906, | |
| "learning_rate": 3.41615066388609e-07, | |
| "logits/chosen": -0.9511051177978516, | |
| "logits/rejected": -1.0637654066085815, | |
| "logps/chosen": -1.2569276094436646, | |
| "logps/rejected": -1.5339549779891968, | |
| "loss": 1.4488, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": 0.14977118372917175, | |
| "rewards/margins": 0.43153300881385803, | |
| "rewards/rejected": -0.2817618250846863, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.32975660821774405, | |
| "grad_norm": 340.4994744679967, | |
| "learning_rate": 3.390105318322492e-07, | |
| "logits/chosen": -1.0287960767745972, | |
| "logits/rejected": -1.0511208772659302, | |
| "logps/chosen": -1.3860498666763306, | |
| "logps/rejected": -1.5678269863128662, | |
| "loss": 1.414, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.0474742166697979, | |
| "rewards/margins": 0.32056504487991333, | |
| "rewards/rejected": -0.36803925037384033, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.33499084009421615, | |
| "grad_norm": 96.74481880578419, | |
| "learning_rate": 3.3635956888748385e-07, | |
| "logits/chosen": -0.988872230052948, | |
| "logits/rejected": -1.0875409841537476, | |
| "logps/chosen": -1.3919366598129272, | |
| "logps/rejected": -1.6267982721328735, | |
| "loss": 1.4391, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": 0.08910200744867325, | |
| "rewards/margins": 0.730194628238678, | |
| "rewards/rejected": -0.6410925984382629, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.3402250719706883, | |
| "grad_norm": 106.35251243553931, | |
| "learning_rate": 3.336630629544019e-07, | |
| "logits/chosen": -0.9367281794548035, | |
| "logits/rejected": -0.9902782440185547, | |
| "logps/chosen": -1.3781791925430298, | |
| "logps/rejected": -1.411145567893982, | |
| "loss": 1.4227, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": 0.07712526619434357, | |
| "rewards/margins": 0.3716704249382019, | |
| "rewards/rejected": -0.29454511404037476, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.34545930384716045, | |
| "grad_norm": 36.1355257607673, | |
| "learning_rate": 3.3092191464408037e-07, | |
| "logits/chosen": -0.9591676592826843, | |
| "logits/rejected": -1.0145364999771118, | |
| "logps/chosen": -1.2052505016326904, | |
| "logps/rejected": -1.494354009628296, | |
| "loss": 1.3849, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": 0.10244836658239365, | |
| "rewards/margins": 0.40993595123291016, | |
| "rewards/rejected": -0.3074875473976135, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.35069353572363254, | |
| "grad_norm": 605.954181910212, | |
| "learning_rate": 3.281370394777878e-07, | |
| "logits/chosen": -0.916668713092804, | |
| "logits/rejected": -0.9587345123291016, | |
| "logps/chosen": -1.2746045589447021, | |
| "logps/rejected": -1.5481473207473755, | |
| "loss": 1.4423, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": 0.026984428986907005, | |
| "rewards/margins": 0.21059663593769073, | |
| "rewards/rejected": -0.18361221253871918, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.3559277676001047, | |
| "grad_norm": 165.59804456052152, | |
| "learning_rate": 3.2530936758120725e-07, | |
| "logits/chosen": -0.8825721740722656, | |
| "logits/rejected": -0.9537725448608398, | |
| "logps/chosen": -1.3265260457992554, | |
| "logps/rejected": -1.4398428201675415, | |
| "loss": 1.4304, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.1924048364162445, | |
| "rewards/margins": 0.5167059898376465, | |
| "rewards/rejected": -0.3243011236190796, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.3611619994765768, | |
| "grad_norm": 149.3409994530682, | |
| "learning_rate": 3.224398433737821e-07, | |
| "logits/chosen": -0.9888921976089478, | |
| "logits/rejected": -1.0671908855438232, | |
| "logps/chosen": -1.249463438987732, | |
| "logps/rejected": -1.6025193929672241, | |
| "loss": 1.4495, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": 0.06590111553668976, | |
| "rewards/margins": 0.3165399134159088, | |
| "rewards/rejected": -0.25063878297805786, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.36639623135304894, | |
| "grad_norm": 39.76420565368717, | |
| "learning_rate": 3.195294252532876e-07, | |
| "logits/chosen": -0.9943802952766418, | |
| "logits/rejected": -1.0306392908096313, | |
| "logps/chosen": -1.3779702186584473, | |
| "logps/rejected": -1.5936956405639648, | |
| "loss": 1.3754, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.06695530563592911, | |
| "rewards/margins": 0.3752886950969696, | |
| "rewards/rejected": -0.4422439932823181, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.36639623135304894, | |
| "eval_logits/chosen": -1.0582585334777832, | |
| "eval_logits/rejected": -1.0870941877365112, | |
| "eval_logps/chosen": -1.2654900550842285, | |
| "eval_logps/rejected": -1.511101484298706, | |
| "eval_loss": 1.4125109910964966, | |
| "eval_rewards/accuracies": 0.7440476417541504, | |
| "eval_rewards/chosen": 0.08484180271625519, | |
| "eval_rewards/margins": 0.4102318584918976, | |
| "eval_rewards/rejected": -0.3253900408744812, | |
| "eval_runtime": 263.46, | |
| "eval_samples_per_second": 7.591, | |
| "eval_steps_per_second": 0.239, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.3716304632295211, | |
| "grad_norm": 64.77721558069398, | |
| "learning_rate": 3.165790852757337e-07, | |
| "logits/chosen": -1.0020430088043213, | |
| "logits/rejected": -1.0420172214508057, | |
| "logps/chosen": -1.356451392173767, | |
| "logps/rejected": -1.4674928188323975, | |
| "loss": 1.3739, | |
| "rewards/accuracies": 0.8374999761581421, | |
| "rewards/chosen": 0.20629188418388367, | |
| "rewards/margins": 0.4772399067878723, | |
| "rewards/rejected": -0.27094796299934387, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.3768646951059932, | |
| "grad_norm": 1320.2371661660036, | |
| "learning_rate": 3.135898088307064e-07, | |
| "logits/chosen": -0.9764865636825562, | |
| "logits/rejected": -0.9581249952316284, | |
| "logps/chosen": -1.360630750656128, | |
| "logps/rejected": -1.4430283308029175, | |
| "loss": 1.4128, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": 0.09529178589582443, | |
| "rewards/margins": 0.38423871994018555, | |
| "rewards/rejected": -0.2889469563961029, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.38209892698246534, | |
| "grad_norm": 271.7523715869803, | |
| "learning_rate": 3.1056259431225556e-07, | |
| "logits/chosen": -1.0364816188812256, | |
| "logits/rejected": -1.0083050727844238, | |
| "logps/chosen": -1.3824083805084229, | |
| "logps/rejected": -1.5307719707489014, | |
| "loss": 1.4463, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": 0.052348293364048004, | |
| "rewards/margins": 0.6019971966743469, | |
| "rewards/rejected": -0.5496489405632019, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.38733315885893743, | |
| "grad_norm": 80.88419376322062, | |
| "learning_rate": 3.074984527854392e-07, | |
| "logits/chosen": -0.9187393188476562, | |
| "logits/rejected": -0.9706932902336121, | |
| "logps/chosen": -1.3978350162506104, | |
| "logps/rejected": -1.6177310943603516, | |
| "loss": 1.4341, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.2012447863817215, | |
| "rewards/margins": 0.3044939339160919, | |
| "rewards/rejected": -0.10324916988611221, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.3925673907354096, | |
| "grad_norm": 90.01276163951067, | |
| "learning_rate": 3.043984076486364e-07, | |
| "logits/chosen": -1.0107219219207764, | |
| "logits/rejected": -0.9986340403556824, | |
| "logps/chosen": -1.273086428642273, | |
| "logps/rejected": -1.4610965251922607, | |
| "loss": 1.513, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": 0.2087746411561966, | |
| "rewards/margins": 0.4228358864784241, | |
| "rewards/rejected": -0.21406126022338867, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.39780162261188173, | |
| "grad_norm": 2597.0418080116397, | |
| "learning_rate": 3.0126349429174023e-07, | |
| "logits/chosen": -1.0616979598999023, | |
| "logits/rejected": -1.0581673383712769, | |
| "logps/chosen": -1.5037667751312256, | |
| "logps/rejected": -1.6452052593231201, | |
| "loss": 1.5244, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.08855433762073517, | |
| "rewards/margins": 0.4076352119445801, | |
| "rewards/rejected": -0.49618959426879883, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.40303585448835383, | |
| "grad_norm": 162.29078450187296, | |
| "learning_rate": 2.9809475975034583e-07, | |
| "logits/chosen": -0.9451099634170532, | |
| "logits/rejected": -0.9363230466842651, | |
| "logps/chosen": -1.2786595821380615, | |
| "logps/rejected": -1.4589165449142456, | |
| "loss": 1.4723, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": 0.0040801106952130795, | |
| "rewards/margins": 0.1179959774017334, | |
| "rewards/rejected": -0.11391584575176239, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.408270086364826, | |
| "grad_norm": 110.30330182049372, | |
| "learning_rate": 2.948932623560495e-07, | |
| "logits/chosen": -0.9574085474014282, | |
| "logits/rejected": -0.9982038736343384, | |
| "logps/chosen": -1.3518567085266113, | |
| "logps/rejected": -1.5835750102996826, | |
| "loss": 1.4165, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.035115428268909454, | |
| "rewards/margins": 0.5685119032859802, | |
| "rewards/rejected": -0.6036273837089539, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.4135043182412981, | |
| "grad_norm": 60.51171656648762, | |
| "learning_rate": 2.916600713829742e-07, | |
| "logits/chosen": -1.0429871082305908, | |
| "logits/rejected": -1.0647838115692139, | |
| "logps/chosen": -1.4226644039154053, | |
| "logps/rejected": -1.5080631971359253, | |
| "loss": 1.4199, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": 0.059725116938352585, | |
| "rewards/margins": 0.1583862155675888, | |
| "rewards/rejected": -0.09866108745336533, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.4187385501177702, | |
| "grad_norm": 290.41341566251265, | |
| "learning_rate": 2.8839626669064067e-07, | |
| "logits/chosen": -0.9505065679550171, | |
| "logits/rejected": -1.0394840240478516, | |
| "logps/chosen": -1.3534128665924072, | |
| "logps/rejected": -1.6843515634536743, | |
| "loss": 1.4283, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 0.08606131374835968, | |
| "rewards/margins": 0.5727912187576294, | |
| "rewards/rejected": -0.4867299497127533, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.4187385501177702, | |
| "eval_logits/chosen": -1.0662418603897095, | |
| "eval_logits/rejected": -1.0950087308883667, | |
| "eval_logps/chosen": -1.2765921354293823, | |
| "eval_logps/rejected": -1.5239523649215698, | |
| "eval_loss": 1.4428484439849854, | |
| "eval_rewards/accuracies": 0.726190447807312, | |
| "eval_rewards/chosen": -0.08169105648994446, | |
| "eval_rewards/margins": 0.4364626109600067, | |
| "eval_rewards/rejected": -0.5181536674499512, | |
| "eval_runtime": 264.4094, | |
| "eval_samples_per_second": 7.564, | |
| "eval_steps_per_second": 0.238, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.4239727819942423, | |
| "grad_norm": 136.61664077267568, | |
| "learning_rate": 2.8510293836330317e-07, | |
| "logits/chosen": -0.9569026827812195, | |
| "logits/rejected": -1.0604639053344727, | |
| "logps/chosen": -1.1840898990631104, | |
| "logps/rejected": -1.5309747457504272, | |
| "loss": 1.4795, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.0819721594452858, | |
| "rewards/margins": 0.2086147964000702, | |
| "rewards/rejected": -0.2905869781970978, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.42920701387071447, | |
| "grad_norm": 157.90172498376174, | |
| "learning_rate": 2.8178118634587043e-07, | |
| "logits/chosen": -1.030447244644165, | |
| "logits/rejected": -1.0620393753051758, | |
| "logps/chosen": -1.3706046342849731, | |
| "logps/rejected": -1.6218140125274658, | |
| "loss": 1.4986, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": 0.20854957401752472, | |
| "rewards/margins": 0.1404011845588684, | |
| "rewards/rejected": 0.06814839690923691, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.4344412457471866, | |
| "grad_norm": 145.68423411902808, | |
| "learning_rate": 2.7843212007653255e-07, | |
| "logits/chosen": -0.9203750491142273, | |
| "logits/rejected": -0.9779027104377747, | |
| "logps/chosen": -1.2530797719955444, | |
| "logps/rejected": -1.425672173500061, | |
| "loss": 1.4582, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": 0.06837911903858185, | |
| "rewards/margins": 0.17998042702674866, | |
| "rewards/rejected": -0.1116013303399086, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.4396754776236587, | |
| "grad_norm": 68.18620996905915, | |
| "learning_rate": 2.750568581162179e-07, | |
| "logits/chosen": -0.9847718477249146, | |
| "logits/rejected": -1.110650658607483, | |
| "logps/chosen": -1.2794532775878906, | |
| "logps/rejected": -1.498471975326538, | |
| "loss": 1.4343, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": 0.10923127084970474, | |
| "rewards/margins": 0.44351276755332947, | |
| "rewards/rejected": -0.3342815339565277, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.44490970950013087, | |
| "grad_norm": 242.48291318354657, | |
| "learning_rate": 2.7165652777500305e-07, | |
| "logits/chosen": -1.017841100692749, | |
| "logits/rejected": -1.0413384437561035, | |
| "logps/chosen": -1.3309098482131958, | |
| "logps/rejected": -1.5017926692962646, | |
| "loss": 1.44, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -0.07142224162817001, | |
| "rewards/margins": 0.2956584095954895, | |
| "rewards/rejected": -0.3670806884765625, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.45014394137660296, | |
| "grad_norm": 109.43096475182443, | |
| "learning_rate": 2.682322647355999e-07, | |
| "logits/chosen": -0.998466968536377, | |
| "logits/rejected": -1.0665570497512817, | |
| "logps/chosen": -1.2548444271087646, | |
| "logps/rejected": -1.7455116510391235, | |
| "loss": 1.4151, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.12004120647907257, | |
| "rewards/margins": 0.5363563299179077, | |
| "rewards/rejected": -0.6563974618911743, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.4553781732530751, | |
| "grad_norm": 124.43882606046388, | |
| "learning_rate": 2.6478521267404725e-07, | |
| "logits/chosen": -0.911393940448761, | |
| "logits/rejected": -0.9683974385261536, | |
| "logps/chosen": -1.299457311630249, | |
| "logps/rejected": -1.5499870777130127, | |
| "loss": 1.4252, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.025322074070572853, | |
| "rewards/margins": 0.29149001836776733, | |
| "rewards/rejected": -0.26616793870925903, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.46061240512954726, | |
| "grad_norm": 78.64863181721877, | |
| "learning_rate": 2.613165228777323e-07, | |
| "logits/chosen": -1.0315337181091309, | |
| "logits/rejected": -1.059273600578308, | |
| "logps/chosen": -1.3746845722198486, | |
| "logps/rejected": -1.4652302265167236, | |
| "loss": 1.386, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": 0.06165723875164986, | |
| "rewards/margins": 0.20029735565185547, | |
| "rewards/rejected": -0.13864010572433472, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.46584663700601936, | |
| "grad_norm": 72.31012788663783, | |
| "learning_rate": 2.578273538608695e-07, | |
| "logits/chosen": -1.0157678127288818, | |
| "logits/rejected": -0.9929217100143433, | |
| "logps/chosen": -1.3486502170562744, | |
| "logps/rejected": -1.329793930053711, | |
| "loss": 1.422, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.019388969987630844, | |
| "rewards/margins": 0.2866728901863098, | |
| "rewards/rejected": -0.30606183409690857, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.4710808688824915, | |
| "grad_norm": 168.71230864025074, | |
| "learning_rate": 2.5431887097756707e-07, | |
| "logits/chosen": -0.9708254933357239, | |
| "logits/rejected": -1.0406501293182373, | |
| "logps/chosen": -1.3819133043289185, | |
| "logps/rejected": -1.7021602392196655, | |
| "loss": 1.4394, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": 0.25760143995285034, | |
| "rewards/margins": 0.4438048303127289, | |
| "rewards/rejected": -0.18620333075523376, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.4710808688824915, | |
| "eval_logits/chosen": -1.0399378538131714, | |
| "eval_logits/rejected": -1.067466139793396, | |
| "eval_logps/chosen": -1.25888991355896, | |
| "eval_logps/rejected": -1.5051480531692505, | |
| "eval_loss": 1.4415701627731323, | |
| "eval_rewards/accuracies": 0.7420634627342224, | |
| "eval_rewards/chosen": 0.18384411931037903, | |
| "eval_rewards/margins": 0.419933021068573, | |
| "eval_rewards/rejected": -0.23608890175819397, | |
| "eval_runtime": 265.0578, | |
| "eval_samples_per_second": 7.546, | |
| "eval_steps_per_second": 0.238, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.4763151007589636, | |
| "grad_norm": 107.01879296821734, | |
| "learning_rate": 2.507922460326075e-07, | |
| "logits/chosen": -0.9058791995048523, | |
| "logits/rejected": -0.931871235370636, | |
| "logps/chosen": -1.2475355863571167, | |
| "logps/rejected": -1.5984818935394287, | |
| "loss": 1.5115, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": 0.04505655914545059, | |
| "rewards/margins": 0.017883723601698875, | |
| "rewards/rejected": 0.027172747999429703, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.48154933263543576, | |
| "grad_norm": 72.74505143056861, | |
| "learning_rate": 2.4724865689007444e-07, | |
| "logits/chosen": -0.9704592823982239, | |
| "logits/rejected": -1.0217134952545166, | |
| "logps/chosen": -1.2369471788406372, | |
| "logps/rejected": -1.5074516534805298, | |
| "loss": 1.433, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": 0.1680927574634552, | |
| "rewards/margins": 0.4736298620700836, | |
| "rewards/rejected": -0.3055371642112732, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.48678356451190785, | |
| "grad_norm": 114.65949067860906, | |
| "learning_rate": 2.436892870799559e-07, | |
| "logits/chosen": -0.9633470773696899, | |
| "logits/rejected": -0.9601320028305054, | |
| "logps/chosen": -1.268119215965271, | |
| "logps/rejected": -1.3131425380706787, | |
| "loss": 1.3956, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.11231119930744171, | |
| "rewards/margins": 0.2431112825870514, | |
| "rewards/rejected": -0.13080011308193207, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.49201779638838, | |
| "grad_norm": 101.41199727533493, | |
| "learning_rate": 2.4011532540285447e-07, | |
| "logits/chosen": -1.023418664932251, | |
| "logits/rejected": -1.0831656455993652, | |
| "logps/chosen": -1.4656355381011963, | |
| "logps/rejected": -1.5011208057403564, | |
| "loss": 1.4291, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": 0.01922314427793026, | |
| "rewards/margins": 0.3083404004573822, | |
| "rewards/rejected": -0.2891172766685486, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.49725202826485215, | |
| "grad_norm": 89.65607411754935, | |
| "learning_rate": 2.3652796553293793e-07, | |
| "logits/chosen": -0.9452352523803711, | |
| "logits/rejected": -1.002318024635315, | |
| "logps/chosen": -1.3159103393554688, | |
| "logps/rejected": -1.7574964761734009, | |
| "loss": 1.4089, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": 0.028190921992063522, | |
| "rewards/margins": 0.4952470362186432, | |
| "rewards/rejected": -0.4670560956001282, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.5024862601413242, | |
| "grad_norm": 382.85539202651387, | |
| "learning_rate": 2.3292840561926163e-07, | |
| "logits/chosen": -0.9121773838996887, | |
| "logits/rejected": -0.9947258830070496, | |
| "logps/chosen": -1.3057327270507812, | |
| "logps/rejected": -1.6355533599853516, | |
| "loss": 1.4497, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": 0.17302857339382172, | |
| "rewards/margins": 0.6395506858825684, | |
| "rewards/rejected": -0.46652212738990784, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.5077204920177963, | |
| "grad_norm": 104.65497018703128, | |
| "learning_rate": 2.2931784788559626e-07, | |
| "logits/chosen": -1.0272817611694336, | |
| "logits/rejected": -1.0622152090072632, | |
| "logps/chosen": -1.3166601657867432, | |
| "logps/rejected": -1.7344980239868164, | |
| "loss": 1.4224, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.005081593990325928, | |
| "rewards/margins": 0.5101627111434937, | |
| "rewards/rejected": -0.5152442455291748, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.5129547238942685, | |
| "grad_norm": 203.5247230819424, | |
| "learning_rate": 2.2569749822889524e-07, | |
| "logits/chosen": -0.9486488103866577, | |
| "logits/rejected": -1.042937994003296, | |
| "logps/chosen": -1.2337818145751953, | |
| "logps/rejected": -1.4374665021896362, | |
| "loss": 1.4641, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": 0.06478907912969589, | |
| "rewards/margins": 0.2325349748134613, | |
| "rewards/rejected": -0.1677459180355072, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.5181889557707406, | |
| "grad_norm": 133.03376715975915, | |
| "learning_rate": 2.220685658165347e-07, | |
| "logits/chosen": -0.963657021522522, | |
| "logits/rejected": -1.0291000604629517, | |
| "logps/chosen": -1.1978986263275146, | |
| "logps/rejected": -1.4739136695861816, | |
| "loss": 1.3969, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": 0.08724164962768555, | |
| "rewards/margins": 0.43057960271835327, | |
| "rewards/rejected": -0.3433380126953125, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.5234231876472127, | |
| "grad_norm": 58.99833399504203, | |
| "learning_rate": 2.1843226268246133e-07, | |
| "logits/chosen": -0.9091464281082153, | |
| "logits/rejected": -0.9449722170829773, | |
| "logps/chosen": -1.191294550895691, | |
| "logps/rejected": -1.4181480407714844, | |
| "loss": 1.3847, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": 0.02543627843260765, | |
| "rewards/margins": 0.339578777551651, | |
| "rewards/rejected": -0.31414246559143066, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.5234231876472127, | |
| "eval_logits/chosen": -1.0453258752822876, | |
| "eval_logits/rejected": -1.0717276334762573, | |
| "eval_logps/chosen": -1.263582706451416, | |
| "eval_logps/rejected": -1.5113520622253418, | |
| "eval_loss": 1.41469407081604, | |
| "eval_rewards/accuracies": 0.7440476417541504, | |
| "eval_rewards/chosen": 0.11345059424638748, | |
| "eval_rewards/margins": 0.44259801506996155, | |
| "eval_rewards/rejected": -0.32914745807647705, | |
| "eval_runtime": 263.0472, | |
| "eval_samples_per_second": 7.603, | |
| "eval_steps_per_second": 0.24, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.528657419523685, | |
| "grad_norm": 175.82571938691842, | |
| "learning_rate": 2.1478980332238308e-07, | |
| "logits/chosen": -1.0793265104293823, | |
| "logits/rejected": -1.0983974933624268, | |
| "logps/chosen": -1.3436429500579834, | |
| "logps/rejected": -1.510811686515808, | |
| "loss": 1.4792, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": 0.012947812676429749, | |
| "rewards/margins": 0.24009516835212708, | |
| "rewards/rejected": -0.22714734077453613, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.533891651400157, | |
| "grad_norm": 173.3542667625851, | |
| "learning_rate": 2.1114240428813748e-07, | |
| "logits/chosen": -1.024076223373413, | |
| "logits/rejected": -1.10614013671875, | |
| "logps/chosen": -1.3571223020553589, | |
| "logps/rejected": -1.4241881370544434, | |
| "loss": 1.3932, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": 0.04782426357269287, | |
| "rewards/margins": 0.3030202388763428, | |
| "rewards/rejected": -0.2551959455013275, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.5391258832766291, | |
| "grad_norm": 170.6198273419075, | |
| "learning_rate": 2.074912837813728e-07, | |
| "logits/chosen": -1.0137916803359985, | |
| "logits/rejected": -1.0654577016830444, | |
| "logps/chosen": -1.3849023580551147, | |
| "logps/rejected": -1.4623501300811768, | |
| "loss": 1.3837, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 0.2025812566280365, | |
| "rewards/margins": 0.5163997411727905, | |
| "rewards/rejected": -0.31381842494010925, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.5443601151531012, | |
| "grad_norm": 169.2969270431579, | |
| "learning_rate": 2.0383766124667928e-07, | |
| "logits/chosen": -1.017490267753601, | |
| "logits/rejected": -0.980577826499939, | |
| "logps/chosen": -1.4890670776367188, | |
| "logps/rejected": -1.468117594718933, | |
| "loss": 1.4238, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": 0.07746653258800507, | |
| "rewards/margins": 0.3936583995819092, | |
| "rewards/rejected": -0.3161918818950653, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.5495943470295734, | |
| "grad_norm": 124.53171403879527, | |
| "learning_rate": 2.001827569643039e-07, | |
| "logits/chosen": -0.9910699725151062, | |
| "logits/rejected": -1.0238853693008423, | |
| "logps/chosen": -1.331132173538208, | |
| "logps/rejected": -1.5334141254425049, | |
| "loss": 1.4448, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": 0.05304880812764168, | |
| "rewards/margins": 0.41380447149276733, | |
| "rewards/rejected": -0.36075565218925476, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.5548285789060455, | |
| "grad_norm": 199.67308013058255, | |
| "learning_rate": 1.9652779164258702e-07, | |
| "logits/chosen": -1.0257099866867065, | |
| "logits/rejected": -1.0186676979064941, | |
| "logps/chosen": -1.3845534324645996, | |
| "logps/rejected": -1.4948935508728027, | |
| "loss": 1.3531, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": 0.1899380087852478, | |
| "rewards/margins": 0.5664941072463989, | |
| "rewards/rejected": -0.37655606865882874, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.5600628107825176, | |
| "grad_norm": 142.96112752249948, | |
| "learning_rate": 1.928739860102556e-07, | |
| "logits/chosen": -1.033825159072876, | |
| "logits/rejected": -1.0962693691253662, | |
| "logps/chosen": -1.3696187734603882, | |
| "logps/rejected": -1.540083408355713, | |
| "loss": 1.389, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": 0.26318567991256714, | |
| "rewards/margins": 0.5921273827552795, | |
| "rewards/rejected": -0.3289416432380676, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.5652970426589898, | |
| "grad_norm": 89.70184971339678, | |
| "learning_rate": 1.8922256040870999e-07, | |
| "logits/chosen": -0.981185793876648, | |
| "logits/rejected": -1.0736793279647827, | |
| "logps/chosen": -1.3201899528503418, | |
| "logps/rejected": -1.4359652996063232, | |
| "loss": 1.4041, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": 0.05177130550146103, | |
| "rewards/margins": 0.3131803870201111, | |
| "rewards/rejected": -0.26140910387039185, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.5705312745354619, | |
| "grad_norm": 196.20946597505844, | |
| "learning_rate": 1.8557473438443928e-07, | |
| "logits/chosen": -0.8916726112365723, | |
| "logits/rejected": -0.9935215711593628, | |
| "logps/chosen": -1.2833975553512573, | |
| "logps/rejected": -1.6577835083007812, | |
| "loss": 1.4167, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": 0.01740972325205803, | |
| "rewards/margins": 0.34133031964302063, | |
| "rewards/rejected": -0.3239206075668335, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.575765506411934, | |
| "grad_norm": 143.17390397505113, | |
| "learning_rate": 1.819317262817032e-07, | |
| "logits/chosen": -0.9497382044792175, | |
| "logits/rejected": -1.0630282163619995, | |
| "logps/chosen": -1.2121965885162354, | |
| "logps/rejected": -1.4094746112823486, | |
| "loss": 1.4128, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": 0.16705064475536346, | |
| "rewards/margins": 0.38941216468811035, | |
| "rewards/rejected": -0.22236153483390808, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.575765506411934, | |
| "eval_logits/chosen": -1.0532737970352173, | |
| "eval_logits/rejected": -1.0803874731063843, | |
| "eval_logps/chosen": -1.25905179977417, | |
| "eval_logps/rejected": -1.5050337314605713, | |
| "eval_loss": 1.4049805402755737, | |
| "eval_rewards/accuracies": 0.7341269850730896, | |
| "eval_rewards/chosen": 0.1814154088497162, | |
| "eval_rewards/margins": 0.415788859128952, | |
| "eval_rewards/rejected": -0.23437345027923584, | |
| "eval_runtime": 264.5999, | |
| "eval_samples_per_second": 7.559, | |
| "eval_steps_per_second": 0.238, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.5809997382884062, | |
| "grad_norm": 252.28706540848293, | |
| "learning_rate": 1.7829475283561475e-07, | |
| "logits/chosen": -1.101175308227539, | |
| "logits/rejected": -1.095895767211914, | |
| "logps/chosen": -1.4018386602401733, | |
| "logps/rejected": -1.5792086124420166, | |
| "loss": 1.3957, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": 0.19977723062038422, | |
| "rewards/margins": 0.4205760359764099, | |
| "rewards/rejected": -0.2207988053560257, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.5862339701648783, | |
| "grad_norm": 87.90505007580005, | |
| "learning_rate": 1.7466502876576064e-07, | |
| "logits/chosen": -1.0454351902008057, | |
| "logits/rejected": -1.1120624542236328, | |
| "logps/chosen": -1.349515676498413, | |
| "logps/rejected": -1.5461986064910889, | |
| "loss": 1.369, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": 0.16857576370239258, | |
| "rewards/margins": 0.5384188890457153, | |
| "rewards/rejected": -0.369843065738678, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.5914682020413504, | |
| "grad_norm": 144.50955438385066, | |
| "learning_rate": 1.7104376637049473e-07, | |
| "logits/chosen": -1.0022087097167969, | |
| "logits/rejected": -0.9819334149360657, | |
| "logps/chosen": -1.2998772859573364, | |
| "logps/rejected": -1.5478651523590088, | |
| "loss": 1.3787, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.08974303305149078, | |
| "rewards/margins": 0.4815033972263336, | |
| "rewards/rejected": -0.5712464451789856, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.5967024339178225, | |
| "grad_norm": 442.03933803012757, | |
| "learning_rate": 1.6743217512204052e-07, | |
| "logits/chosen": -0.9362661242485046, | |
| "logits/rejected": -1.0156729221343994, | |
| "logps/chosen": -1.2051947116851807, | |
| "logps/rejected": -1.3971805572509766, | |
| "loss": 1.4071, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": 0.11900795996189117, | |
| "rewards/margins": 0.34501004219055176, | |
| "rewards/rejected": -0.2260020673274994, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.6019366657942947, | |
| "grad_norm": 109.24460216819094, | |
| "learning_rate": 1.6383146126253681e-07, | |
| "logits/chosen": -1.0144506692886353, | |
| "logits/rejected": -1.0540077686309814, | |
| "logps/chosen": -1.3877553939819336, | |
| "logps/rejected": -1.6135002374649048, | |
| "loss": 1.3792, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": 0.021001553162932396, | |
| "rewards/margins": 0.2840999960899353, | |
| "rewards/rejected": -0.26309841871261597, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.6071708976707668, | |
| "grad_norm": 278.3453630329118, | |
| "learning_rate": 1.60242827401163e-07, | |
| "logits/chosen": -1.0601884126663208, | |
| "logits/rejected": -1.0491106510162354, | |
| "logps/chosen": -1.3804155588150024, | |
| "logps/rejected": -1.478729486465454, | |
| "loss": 1.4281, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": 0.07954071462154388, | |
| "rewards/margins": 0.320735901594162, | |
| "rewards/rejected": -0.2411951720714569, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.6124051295472389, | |
| "grad_norm": 191.92209965945904, | |
| "learning_rate": 1.5666747211247708e-07, | |
| "logits/chosen": -0.9765886068344116, | |
| "logits/rejected": -1.0292718410491943, | |
| "logps/chosen": -1.3733490705490112, | |
| "logps/rejected": -1.5234708786010742, | |
| "loss": 1.4206, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 0.011603465303778648, | |
| "rewards/margins": 0.39544984698295593, | |
| "rewards/rejected": -0.38384637236595154, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.6176393614237111, | |
| "grad_norm": 508.0527712691584, | |
| "learning_rate": 1.5310658953610188e-07, | |
| "logits/chosen": -0.8951209783554077, | |
| "logits/rejected": -0.9667248725891113, | |
| "logps/chosen": -1.3316763639450073, | |
| "logps/rejected": -1.4773919582366943, | |
| "loss": 1.3855, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": 0.10079389810562134, | |
| "rewards/margins": 0.26270270347595215, | |
| "rewards/rejected": -0.161908820271492, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.6228735933001832, | |
| "grad_norm": 122.17829279586468, | |
| "learning_rate": 1.4956136897789153e-07, | |
| "logits/chosen": -1.054325819015503, | |
| "logits/rejected": -1.1099879741668701, | |
| "logps/chosen": -1.3022596836090088, | |
| "logps/rejected": -1.4433825016021729, | |
| "loss": 1.3658, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": 0.12086385488510132, | |
| "rewards/margins": 0.40718135237693787, | |
| "rewards/rejected": -0.28631752729415894, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.6281078251766553, | |
| "grad_norm": 529.7173443798092, | |
| "learning_rate": 1.4603299451271378e-07, | |
| "logits/chosen": -0.945719838142395, | |
| "logits/rejected": -0.953068733215332, | |
| "logps/chosen": -1.3291584253311157, | |
| "logps/rejected": -1.4614441394805908, | |
| "loss": 1.4134, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": 0.010808942839503288, | |
| "rewards/margins": 0.28808557987213135, | |
| "rewards/rejected": -0.2772766053676605, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.6281078251766553, | |
| "eval_logits/chosen": -1.0341262817382812, | |
| "eval_logits/rejected": -1.0602811574935913, | |
| "eval_logps/chosen": -1.267260193824768, | |
| "eval_logps/rejected": -1.5152653455734253, | |
| "eval_loss": 1.3929948806762695, | |
| "eval_rewards/accuracies": 0.738095223903656, | |
| "eval_rewards/chosen": 0.0582895502448082, | |
| "eval_rewards/margins": 0.4461366534233093, | |
| "eval_rewards/rejected": -0.3878471553325653, | |
| "eval_runtime": 268.653, | |
| "eval_samples_per_second": 7.445, | |
| "eval_steps_per_second": 0.235, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.6333420570531274, | |
| "grad_norm": 103.87458468332127, | |
| "learning_rate": 1.4252264458897765e-07, | |
| "logits/chosen": -1.0020216703414917, | |
| "logits/rejected": -1.003204345703125, | |
| "logps/chosen": -1.3405869007110596, | |
| "logps/rejected": -1.4487977027893066, | |
| "loss": 1.4199, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 0.15262803435325623, | |
| "rewards/margins": 0.41794830560684204, | |
| "rewards/rejected": -0.26532021164894104, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.6385762889295996, | |
| "grad_norm": 126.9034714680501, | |
| "learning_rate": 1.390314916350422e-07, | |
| "logits/chosen": -1.0031870603561401, | |
| "logits/rejected": -1.025270700454712, | |
| "logps/chosen": -1.4444220066070557, | |
| "logps/rejected": -1.5348130464553833, | |
| "loss": 1.3962, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -0.06764872372150421, | |
| "rewards/margins": 0.563138484954834, | |
| "rewards/rejected": -0.6307872533798218, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.6438105208060717, | |
| "grad_norm": 276.6912026252489, | |
| "learning_rate": 1.3556070166763415e-07, | |
| "logits/chosen": -0.9344841241836548, | |
| "logits/rejected": -1.0246646404266357, | |
| "logps/chosen": -1.3276522159576416, | |
| "logps/rejected": -1.4863455295562744, | |
| "loss": 1.4041, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": 0.14306631684303284, | |
| "rewards/margins": 0.48536840081214905, | |
| "rewards/rejected": -0.3423021733760834, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.6490447526825438, | |
| "grad_norm": 332.7158437087833, | |
| "learning_rate": 1.321114339024084e-07, | |
| "logits/chosen": -0.9632574915885925, | |
| "logits/rejected": -1.0173401832580566, | |
| "logps/chosen": -1.2917425632476807, | |
| "logps/rejected": -1.5060694217681885, | |
| "loss": 1.3616, | |
| "rewards/accuracies": 0.7875000238418579, | |
| "rewards/chosen": 0.10876176506280899, | |
| "rewards/margins": 0.5349315404891968, | |
| "rewards/rejected": -0.4261697232723236, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.654278984559016, | |
| "grad_norm": 97.38808549564327, | |
| "learning_rate": 1.2868484036677894e-07, | |
| "logits/chosen": -1.0494537353515625, | |
| "logits/rejected": -1.0573270320892334, | |
| "logps/chosen": -1.312703013420105, | |
| "logps/rejected": -1.3993957042694092, | |
| "loss": 1.419, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.14204749464988708, | |
| "rewards/margins": 0.43673020601272583, | |
| "rewards/rejected": -0.29468271136283875, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.6595132164354881, | |
| "grad_norm": 76.07649121820532, | |
| "learning_rate": 1.2528206551515154e-07, | |
| "logits/chosen": -1.0372177362442017, | |
| "logits/rejected": -1.07349693775177, | |
| "logps/chosen": -1.5161725282669067, | |
| "logps/rejected": -1.4368443489074707, | |
| "loss": 1.3813, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": 0.14961332082748413, | |
| "rewards/margins": 0.501705527305603, | |
| "rewards/rejected": -0.35209211707115173, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.6647474483119602, | |
| "grad_norm": 68.86700640108432, | |
| "learning_rate": 1.2190424584668462e-07, | |
| "logits/chosen": -0.9774061441421509, | |
| "logits/rejected": -1.022447943687439, | |
| "logps/chosen": -1.3168667554855347, | |
| "logps/rejected": -1.5211305618286133, | |
| "loss": 1.36, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": 0.06608399748802185, | |
| "rewards/margins": 0.4469234347343445, | |
| "rewards/rejected": -0.38083943724632263, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.6699816801884323, | |
| "grad_norm": 114.4568415859116, | |
| "learning_rate": 1.185525095257085e-07, | |
| "logits/chosen": -1.0163053274154663, | |
| "logits/rejected": -1.0360127687454224, | |
| "logps/chosen": -1.306158423423767, | |
| "logps/rejected": -1.535194993019104, | |
| "loss": 1.4034, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 0.10055939853191376, | |
| "rewards/margins": 0.4471976161003113, | |
| "rewards/rejected": -0.34663820266723633, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.6752159120649045, | |
| "grad_norm": 54.90434249157224, | |
| "learning_rate": 1.1522797600492707e-07, | |
| "logits/chosen": -0.9692333936691284, | |
| "logits/rejected": -1.0122019052505493, | |
| "logps/chosen": -1.2191261053085327, | |
| "logps/rejected": -1.5494043827056885, | |
| "loss": 1.3791, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": 0.07996504008769989, | |
| "rewards/margins": 0.43224668502807617, | |
| "rewards/rejected": -0.3522816598415375, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.6804501439413766, | |
| "grad_norm": 108.18282033188903, | |
| "learning_rate": 1.1193175565153017e-07, | |
| "logits/chosen": -0.9120146036148071, | |
| "logits/rejected": -0.9983510971069336, | |
| "logps/chosen": -1.3106328248977661, | |
| "logps/rejected": -1.4724805355072021, | |
| "loss": 1.3657, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": 0.011367501690983772, | |
| "rewards/margins": 0.22781343758106232, | |
| "rewards/rejected": -0.2164459526538849, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.6804501439413766, | |
| "eval_logits/chosen": -1.0687713623046875, | |
| "eval_logits/rejected": -1.0977264642715454, | |
| "eval_logps/chosen": -1.2662289142608643, | |
| "eval_logps/rejected": -1.5134109258651733, | |
| "eval_loss": 1.3927397727966309, | |
| "eval_rewards/accuracies": 0.7361111044883728, | |
| "eval_rewards/chosen": 0.07375912368297577, | |
| "eval_rewards/margins": 0.4337916970252991, | |
| "eval_rewards/rejected": -0.3600325584411621, | |
| "eval_runtime": 263.1627, | |
| "eval_samples_per_second": 7.6, | |
| "eval_steps_per_second": 0.239, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.6856843758178487, | |
| "grad_norm": 156.5057273737964, | |
| "learning_rate": 1.0866494937633952e-07, | |
| "logits/chosen": -0.9764137268066406, | |
| "logits/rejected": -1.0053216218948364, | |
| "logps/chosen": -1.370274305343628, | |
| "logps/rejected": -1.550768494606018, | |
| "loss": 1.3795, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": 0.14111948013305664, | |
| "rewards/margins": 0.499011367559433, | |
| "rewards/rejected": -0.35789191722869873, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.6909186076943209, | |
| "grad_norm": 183.97189217976666, | |
| "learning_rate": 1.0542864826611373e-07, | |
| "logits/chosen": -0.9627429246902466, | |
| "logits/rejected": -0.980324387550354, | |
| "logps/chosen": -1.3693149089813232, | |
| "logps/rejected": -1.5022780895233154, | |
| "loss": 1.3761, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": 0.18044526875019073, | |
| "rewards/margins": 0.5759294033050537, | |
| "rewards/rejected": -0.39548414945602417, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.696152839570793, | |
| "grad_norm": 123.36310936062844, | |
| "learning_rate": 1.0222393321913405e-07, | |
| "logits/chosen": -0.9242043495178223, | |
| "logits/rejected": -0.9338275194168091, | |
| "logps/chosen": -1.349094271659851, | |
| "logps/rejected": -1.579219102859497, | |
| "loss": 1.3722, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.06291097402572632, | |
| "rewards/margins": 0.4077533781528473, | |
| "rewards/rejected": -0.4706643223762512, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.7013870714472651, | |
| "grad_norm": 166.04682689360567, | |
| "learning_rate": 9.905187458419342e-08, | |
| "logits/chosen": -0.852869987487793, | |
| "logits/rejected": -0.9787343144416809, | |
| "logps/chosen": -1.232936143875122, | |
| "logps/rejected": -1.542320966720581, | |
| "loss": 1.4051, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": 0.18137158453464508, | |
| "rewards/margins": 0.23298697173595428, | |
| "rewards/rejected": -0.051615405827760696, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.7066213033237373, | |
| "grad_norm": 95.47739919094735, | |
| "learning_rate": 9.591353180310812e-08, | |
| "logits/chosen": -0.9522558450698853, | |
| "logits/rejected": -1.0096293687820435, | |
| "logps/chosen": -1.3504010438919067, | |
| "logps/rejected": -1.3914040327072144, | |
| "loss": 1.4006, | |
| "rewards/accuracies": 0.7875000238418579, | |
| "rewards/chosen": 0.19880017638206482, | |
| "rewards/margins": 0.5202890038490295, | |
| "rewards/rejected": -0.32148876786231995, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.7118555352002094, | |
| "grad_norm": 128.51259333091, | |
| "learning_rate": 9.280995305687343e-08, | |
| "logits/chosen": -0.9708206057548523, | |
| "logits/rejected": -1.021451711654663, | |
| "logps/chosen": -1.3550317287445068, | |
| "logps/rejected": -1.5246636867523193, | |
| "loss": 1.4348, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.06314614415168762, | |
| "rewards/margins": 0.2832590341567993, | |
| "rewards/rejected": -0.22011291980743408, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.7170897670766815, | |
| "grad_norm": 94.91641424091182, | |
| "learning_rate": 8.974217491557916e-08, | |
| "logits/chosen": -1.0391184091567993, | |
| "logits/rejected": -1.0677263736724854, | |
| "logps/chosen": -1.3149784803390503, | |
| "logps/rejected": -1.5607545375823975, | |
| "loss": 1.3456, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": 0.1072906106710434, | |
| "rewards/margins": 0.558443546295166, | |
| "rewards/rejected": -0.4511529505252838, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.7223239989531536, | |
| "grad_norm": 288.16991161199616, | |
| "learning_rate": 8.6711221992204e-08, | |
| "logits/chosen": -0.9859504699707031, | |
| "logits/rejected": -1.0607242584228516, | |
| "logps/chosen": -1.2985832691192627, | |
| "logps/rejected": -1.4298102855682373, | |
| "loss": 1.3638, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": 0.14652974903583527, | |
| "rewards/margins": 0.583271861076355, | |
| "rewards/rejected": -0.4367421269416809, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.7275582308296258, | |
| "grad_norm": 147.08196145784484, | |
| "learning_rate": 8.371810660040286e-08, | |
| "logits/chosen": -0.9275096654891968, | |
| "logits/rejected": -1.0270216464996338, | |
| "logps/chosen": -1.4230481386184692, | |
| "logps/rejected": -1.7390727996826172, | |
| "loss": 1.4108, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.008330064825713634, | |
| "rewards/margins": 0.352591872215271, | |
| "rewards/rejected": -0.3609219491481781, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.7327924627060979, | |
| "grad_norm": 44.78410651121718, | |
| "learning_rate": 8.076382841640277e-08, | |
| "logits/chosen": -1.0107920169830322, | |
| "logits/rejected": -1.0693024396896362, | |
| "logps/chosen": -1.3128068447113037, | |
| "logps/rejected": -1.4866211414337158, | |
| "loss": 1.3569, | |
| "rewards/accuracies": 0.7875000238418579, | |
| "rewards/chosen": 0.04387504979968071, | |
| "rewards/margins": 0.42374491691589355, | |
| "rewards/rejected": -0.37986987829208374, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.7327924627060979, | |
| "eval_logits/chosen": -1.0332118272781372, | |
| "eval_logits/rejected": -1.0589492321014404, | |
| "eval_logps/chosen": -1.2619318962097168, | |
| "eval_logps/rejected": -1.5095207691192627, | |
| "eval_loss": 1.401206135749817, | |
| "eval_rewards/accuracies": 0.738095223903656, | |
| "eval_rewards/chosen": 0.13821402192115784, | |
| "eval_rewards/margins": 0.4398939311504364, | |
| "eval_rewards/rejected": -0.30167993903160095, | |
| "eval_runtime": 269.235, | |
| "eval_samples_per_second": 7.428, | |
| "eval_steps_per_second": 0.234, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.73802669458257, | |
| "grad_norm": 105.62122860453555, | |
| "learning_rate": 7.784937414511845e-08, | |
| "logits/chosen": -1.001814365386963, | |
| "logits/rejected": -1.0847572088241577, | |
| "logps/chosen": -1.3243391513824463, | |
| "logps/rejected": -1.6870197057724, | |
| "loss": 1.4171, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": 0.1925734281539917, | |
| "rewards/margins": 0.6328744292259216, | |
| "rewards/rejected": -0.4403010904788971, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.7432609264590422, | |
| "grad_norm": 51.61690842370248, | |
| "learning_rate": 7.497571719060176e-08, | |
| "logits/chosen": -0.9514597654342651, | |
| "logits/rejected": -1.020627737045288, | |
| "logps/chosen": -1.2951841354370117, | |
| "logps/rejected": -1.4787520170211792, | |
| "loss": 1.3945, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": 0.0273969117552042, | |
| "rewards/margins": 0.40641704201698303, | |
| "rewards/rejected": -0.37902015447616577, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.7484951583355143, | |
| "grad_norm": 155.75874519048327, | |
| "learning_rate": 7.214381733093155e-08, | |
| "logits/chosen": -0.98955237865448, | |
| "logits/rejected": -1.0787885189056396, | |
| "logps/chosen": -1.2784019708633423, | |
| "logps/rejected": -1.4226205348968506, | |
| "loss": 1.4072, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": 0.33527031540870667, | |
| "rewards/margins": 0.6042425632476807, | |
| "rewards/rejected": -0.2689722776412964, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.7537293902119864, | |
| "grad_norm": 48.85149863967685, | |
| "learning_rate": 6.935462039765676e-08, | |
| "logits/chosen": -0.9959059953689575, | |
| "logits/rejected": -1.0643514394760132, | |
| "logps/chosen": -1.2606055736541748, | |
| "logps/rejected": -1.371964454650879, | |
| "loss": 1.3895, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": 0.2062605619430542, | |
| "rewards/margins": 0.3812524676322937, | |
| "rewards/rejected": -0.17499187588691711, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.7589636220884585, | |
| "grad_norm": 120.78360298158358, | |
| "learning_rate": 6.660905795989545e-08, | |
| "logits/chosen": -1.0321362018585205, | |
| "logits/rejected": -1.0730319023132324, | |
| "logps/chosen": -1.4630558490753174, | |
| "logps/rejected": -1.6992202997207642, | |
| "loss": 1.3719, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": 0.18504472076892853, | |
| "rewards/margins": 0.6389643549919128, | |
| "rewards/rejected": -0.4539197087287903, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.7641978539649307, | |
| "grad_norm": 111.414446510109, | |
| "learning_rate": 6.39080470131989e-08, | |
| "logits/chosen": -0.9712077975273132, | |
| "logits/rejected": -1.0402915477752686, | |
| "logps/chosen": -1.3074638843536377, | |
| "logps/rejected": -1.4820187091827393, | |
| "loss": 1.3939, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": 0.018328525125980377, | |
| "rewards/margins": 0.3735812306404114, | |
| "rewards/rejected": -0.3552526831626892, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.7694320858414028, | |
| "grad_norm": 421.6141362022388, | |
| "learning_rate": 6.125248967328198e-08, | |
| "logits/chosen": -0.9822176694869995, | |
| "logits/rejected": -1.059078335762024, | |
| "logps/chosen": -1.365307092666626, | |
| "logps/rejected": -1.535718560218811, | |
| "loss": 1.3764, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.1095791831612587, | |
| "rewards/margins": 0.26515552401542664, | |
| "rewards/rejected": -0.15557633340358734, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.7746663177178749, | |
| "grad_norm": 93.39309030522578, | |
| "learning_rate": 5.8643272874724504e-08, | |
| "logits/chosen": -0.9487727284431458, | |
| "logits/rejected": -1.0384645462036133, | |
| "logps/chosen": -1.2958605289459229, | |
| "logps/rejected": -1.4537372589111328, | |
| "loss": 1.3613, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.055318187922239304, | |
| "rewards/margins": 0.5227149724960327, | |
| "rewards/rejected": -0.4673967957496643, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.7799005495943471, | |
| "grad_norm": 162.31192195360663, | |
| "learning_rate": 5.608126807474145e-08, | |
| "logits/chosen": -0.9301995038986206, | |
| "logits/rejected": -0.9804097414016724, | |
| "logps/chosen": -1.415575623512268, | |
| "logps/rejected": -1.5492489337921143, | |
| "loss": 1.4203, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.015231065452098846, | |
| "rewards/margins": 0.22754482924938202, | |
| "rewards/rejected": -0.24277588725090027, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.7851347814708192, | |
| "grad_norm": 136.2356214233426, | |
| "learning_rate": 5.356733096212422e-08, | |
| "logits/chosen": -1.0096968412399292, | |
| "logits/rejected": -1.0772438049316406, | |
| "logps/chosen": -1.5461193323135376, | |
| "logps/rejected": -1.8158565759658813, | |
| "loss": 1.4025, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": 0.11093674600124359, | |
| "rewards/margins": 0.6151713132858276, | |
| "rewards/rejected": -0.5042346119880676, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.7851347814708192, | |
| "eval_logits/chosen": -1.050403118133545, | |
| "eval_logits/rejected": -1.0775498151779175, | |
| "eval_logps/chosen": -1.2663925886154175, | |
| "eval_logps/rejected": -1.513696312904358, | |
| "eval_loss": 1.3905304670333862, | |
| "eval_rewards/accuracies": 0.7460317611694336, | |
| "eval_rewards/chosen": 0.07130717486143112, | |
| "eval_rewards/margins": 0.4356210231781006, | |
| "eval_rewards/rejected": -0.36431384086608887, | |
| "eval_runtime": 266.5396, | |
| "eval_samples_per_second": 7.504, | |
| "eval_steps_per_second": 0.236, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.7903690133472913, | |
| "grad_norm": 130.03419321106188, | |
| "learning_rate": 5.1102301171446824e-08, | |
| "logits/chosen": -1.013892412185669, | |
| "logits/rejected": -1.0700061321258545, | |
| "logps/chosen": -1.3864901065826416, | |
| "logps/rejected": -1.503999948501587, | |
| "loss": 1.3938, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.011791354976594448, | |
| "rewards/margins": 0.44438084959983826, | |
| "rewards/rejected": -0.45617228746414185, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.7956032452237635, | |
| "grad_norm": 221.10773316045186, | |
| "learning_rate": 4.8687002002635204e-08, | |
| "logits/chosen": -0.9744777679443359, | |
| "logits/rejected": -1.0147373676300049, | |
| "logps/chosen": -1.3872696161270142, | |
| "logps/rejected": -1.5349359512329102, | |
| "loss": 1.3795, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": 0.1630631983280182, | |
| "rewards/margins": 0.39758509397506714, | |
| "rewards/rejected": -0.23452191054821014, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.8008374771002356, | |
| "grad_norm": 38.83280676048386, | |
| "learning_rate": 4.632224014599151e-08, | |
| "logits/chosen": -1.0282185077667236, | |
| "logits/rejected": -1.0352448225021362, | |
| "logps/chosen": -1.3574353456497192, | |
| "logps/rejected": -1.47411048412323, | |
| "loss": 1.3567, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": 0.10859794914722443, | |
| "rewards/margins": 0.5248547792434692, | |
| "rewards/rejected": -0.4162568151950836, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.8060717089767077, | |
| "grad_norm": 66.73720875935125, | |
| "learning_rate": 4.400880541276608e-08, | |
| "logits/chosen": -1.018883228302002, | |
| "logits/rejected": -0.9882569313049316, | |
| "logps/chosen": -1.3592469692230225, | |
| "logps/rejected": -1.4263975620269775, | |
| "loss": 1.38, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": 0.13372251391410828, | |
| "rewards/margins": 0.3992050290107727, | |
| "rewards/rejected": -0.26548251509666443, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.8113059408531798, | |
| "grad_norm": 126.07795147064398, | |
| "learning_rate": 4.1747470471367066e-08, | |
| "logits/chosen": -0.9707099795341492, | |
| "logits/rejected": -1.043423056602478, | |
| "logps/chosen": -1.257214903831482, | |
| "logps/rejected": -1.4531395435333252, | |
| "loss": 1.3925, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": 0.08239827305078506, | |
| "rewards/margins": 0.43017762899398804, | |
| "rewards/rejected": -0.3477793335914612, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.816540172729652, | |
| "grad_norm": 239.23092115114255, | |
| "learning_rate": 3.953899058929542e-08, | |
| "logits/chosen": -0.9740544557571411, | |
| "logits/rejected": -1.002071738243103, | |
| "logps/chosen": -1.36293625831604, | |
| "logps/rejected": -1.4228624105453491, | |
| "loss": 1.3994, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": 0.1495702564716339, | |
| "rewards/margins": 0.3382664620876312, | |
| "rewards/rejected": -0.18869620561599731, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.821774404606124, | |
| "grad_norm": 271.7775399259313, | |
| "learning_rate": 3.738410338089149e-08, | |
| "logits/chosen": -1.0329101085662842, | |
| "logits/rejected": -1.0802018642425537, | |
| "logps/chosen": -1.3035242557525635, | |
| "logps/rejected": -1.4981635808944702, | |
| "loss": 1.4136, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": 0.19920700788497925, | |
| "rewards/margins": 0.6002500653266907, | |
| "rewards/rejected": -0.40104299783706665, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 0.8270086364825961, | |
| "grad_norm": 112.4116696366104, | |
| "learning_rate": 3.528352856097816e-08, | |
| "logits/chosen": -1.06103515625, | |
| "logits/rejected": -1.077307105064392, | |
| "logps/chosen": -1.341233730316162, | |
| "logps/rejected": -1.4782545566558838, | |
| "loss": 1.3651, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.10534757375717163, | |
| "rewards/margins": 0.4684361517429352, | |
| "rewards/rejected": -0.36308857798576355, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.8322428683590684, | |
| "grad_norm": 112.88065021967937, | |
| "learning_rate": 3.323796770448157e-08, | |
| "logits/chosen": -0.891954243183136, | |
| "logits/rejected": -1.009387731552124, | |
| "logps/chosen": -1.2871875762939453, | |
| "logps/rejected": -1.667616844177246, | |
| "loss": 1.3924, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": 0.18675394356250763, | |
| "rewards/margins": 0.7343874573707581, | |
| "rewards/rejected": -0.5476335287094116, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.8374771002355405, | |
| "grad_norm": 274.35787721935316, | |
| "learning_rate": 3.1248104012111085e-08, | |
| "logits/chosen": -1.0275261402130127, | |
| "logits/rejected": -1.0268418788909912, | |
| "logps/chosen": -1.4677902460098267, | |
| "logps/rejected": -1.5933417081832886, | |
| "loss": 1.4056, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.0058928607031702995, | |
| "rewards/margins": 0.3813323378562927, | |
| "rewards/rejected": -0.3872251808643341, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.8374771002355405, | |
| "eval_logits/chosen": -1.0522223711013794, | |
| "eval_logits/rejected": -1.0792322158813477, | |
| "eval_logps/chosen": -1.262178659439087, | |
| "eval_logps/rejected": -1.5092347860336304, | |
| "eval_loss": 1.3950434923171997, | |
| "eval_rewards/accuracies": 0.7341269850730896, | |
| "eval_rewards/chosen": 0.13451193273067474, | |
| "eval_rewards/margins": 0.4319010376930237, | |
| "eval_rewards/rejected": -0.2973891496658325, | |
| "eval_runtime": 263.6942, | |
| "eval_samples_per_second": 7.585, | |
| "eval_steps_per_second": 0.239, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.8427113321120125, | |
| "grad_norm": 138.0710133394622, | |
| "learning_rate": 2.931460208217562e-08, | |
| "logits/chosen": -0.9226272702217102, | |
| "logits/rejected": -1.032504677772522, | |
| "logps/chosen": -1.2919390201568604, | |
| "logps/rejected": -1.6714973449707031, | |
| "loss": 1.3897, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": 0.11851558834314346, | |
| "rewards/margins": 0.38856449723243713, | |
| "rewards/rejected": -0.27004891633987427, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 0.8479455639884846, | |
| "grad_norm": 374.96362049867355, | |
| "learning_rate": 2.743810768861341e-08, | |
| "logits/chosen": -0.9746836423873901, | |
| "logits/rejected": -1.0413103103637695, | |
| "logps/chosen": -1.255078673362732, | |
| "logps/rejected": -1.5265228748321533, | |
| "loss": 1.3898, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.0367145761847496, | |
| "rewards/margins": 0.3037044405937195, | |
| "rewards/rejected": -0.26698988676071167, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.8531797958649568, | |
| "grad_norm": 1561.6988123064714, | |
| "learning_rate": 2.5619247565308444e-08, | |
| "logits/chosen": -1.0172195434570312, | |
| "logits/rejected": -1.0282325744628906, | |
| "logps/chosen": -1.3624012470245361, | |
| "logps/rejected": -1.428806185722351, | |
| "loss": 1.4127, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.02823961339890957, | |
| "rewards/margins": 0.31183555722236633, | |
| "rewards/rejected": -0.34007519483566284, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 0.8584140277414289, | |
| "grad_norm": 432.62083647251535, | |
| "learning_rate": 2.3858629196766845e-08, | |
| "logits/chosen": -0.9970697164535522, | |
| "logits/rejected": -1.0459386110305786, | |
| "logps/chosen": -1.39687979221344, | |
| "logps/rejected": -1.453002691268921, | |
| "loss": 1.382, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.012905001640319824, | |
| "rewards/margins": 0.27340295910835266, | |
| "rewards/rejected": -0.2863079607486725, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.863648259617901, | |
| "grad_norm": 107.41089429393227, | |
| "learning_rate": 2.2156840615221563e-08, | |
| "logits/chosen": -1.0009286403656006, | |
| "logits/rejected": -1.008098840713501, | |
| "logps/chosen": -1.3218780755996704, | |
| "logps/rejected": -1.461322546005249, | |
| "loss": 1.4188, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": 0.14685675501823425, | |
| "rewards/margins": 0.3876083493232727, | |
| "rewards/rejected": -0.24075157940387726, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.8688824914943732, | |
| "grad_norm": 139.7533229551783, | |
| "learning_rate": 2.0514450204234724e-08, | |
| "logits/chosen": -1.0012562274932861, | |
| "logits/rejected": -1.0368671417236328, | |
| "logps/chosen": -1.3768261671066284, | |
| "logps/rejected": -1.5115500688552856, | |
| "loss": 1.3544, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.0796336680650711, | |
| "rewards/margins": 0.47576791048049927, | |
| "rewards/rejected": -0.39613422751426697, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.8741167233708453, | |
| "grad_norm": 100.1098611400833, | |
| "learning_rate": 1.8932006508861865e-08, | |
| "logits/chosen": -0.9033932685852051, | |
| "logits/rejected": -1.0143510103225708, | |
| "logps/chosen": -1.2409793138504028, | |
| "logps/rejected": -1.513612985610962, | |
| "loss": 1.3784, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": 0.053215838968753815, | |
| "rewards/margins": 0.46576422452926636, | |
| "rewards/rejected": -0.41254839301109314, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 0.8793509552473174, | |
| "grad_norm": 164.31866508664524, | |
| "learning_rate": 1.7410038052442633e-08, | |
| "logits/chosen": -0.9056793451309204, | |
| "logits/rejected": -0.9717051386833191, | |
| "logps/chosen": -1.3354060649871826, | |
| "logps/rejected": -1.5312366485595703, | |
| "loss": 1.3525, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": 0.1547504961490631, | |
| "rewards/margins": 0.549881637096405, | |
| "rewards/rejected": -0.3951311409473419, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.8845851871237895, | |
| "grad_norm": 781.1957844328009, | |
| "learning_rate": 1.5949053160077974e-08, | |
| "logits/chosen": -1.0040967464447021, | |
| "logits/rejected": -1.042490839958191, | |
| "logps/chosen": -1.3877414464950562, | |
| "logps/rejected": -1.6622329950332642, | |
| "loss": 1.4005, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 0.12258515506982803, | |
| "rewards/margins": 0.5700903534889221, | |
| "rewards/rejected": -0.4475051760673523, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 0.8898194190002617, | |
| "grad_norm": 156.43790654355806, | |
| "learning_rate": 1.4549539788853981e-08, | |
| "logits/chosen": -0.9561022520065308, | |
| "logits/rejected": -1.0446237325668335, | |
| "logps/chosen": -1.3862628936767578, | |
| "logps/rejected": -1.4567193984985352, | |
| "loss": 1.3963, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": 0.06578576564788818, | |
| "rewards/margins": 0.31272581219673157, | |
| "rewards/rejected": -0.246940016746521, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.8898194190002617, | |
| "eval_logits/chosen": -1.0522090196609497, | |
| "eval_logits/rejected": -1.0791518688201904, | |
| "eval_logps/chosen": -1.266129732131958, | |
| "eval_logps/rejected": -1.5129594802856445, | |
| "eval_loss": 1.3915550708770752, | |
| "eval_rewards/accuracies": 0.7400793433189392, | |
| "eval_rewards/chosen": 0.07524589449167252, | |
| "eval_rewards/margins": 0.42850303649902344, | |
| "eval_rewards/rejected": -0.3532571494579315, | |
| "eval_runtime": 270.4262, | |
| "eval_samples_per_second": 7.396, | |
| "eval_steps_per_second": 0.233, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.8950536508767338, | |
| "grad_norm": 133.3233493091381, | |
| "learning_rate": 1.3211965364867906e-08, | |
| "logits/chosen": -0.9714654684066772, | |
| "logits/rejected": -1.0295698642730713, | |
| "logps/chosen": -1.28077232837677, | |
| "logps/rejected": -1.6242740154266357, | |
| "loss": 1.3713, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": 0.012086862698197365, | |
| "rewards/margins": 0.3643137514591217, | |
| "rewards/rejected": -0.3522268831729889, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 0.9002878827532059, | |
| "grad_norm": 162.10930120150212, | |
| "learning_rate": 1.1936776627111789e-08, | |
| "logits/chosen": -0.9902065396308899, | |
| "logits/rejected": -1.016322135925293, | |
| "logps/chosen": -1.3354121446609497, | |
| "logps/rejected": -1.4340362548828125, | |
| "loss": 1.3901, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": 0.08343851566314697, | |
| "rewards/margins": 0.25945615768432617, | |
| "rewards/rejected": -0.1760176420211792, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 0.9055221146296781, | |
| "grad_norm": 79.88141762385948, | |
| "learning_rate": 1.072439947826531e-08, | |
| "logits/chosen": -0.9664213061332703, | |
| "logits/rejected": -1.0111840963363647, | |
| "logps/chosen": -1.2742303609848022, | |
| "logps/rejected": -1.4425808191299438, | |
| "loss": 1.3581, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.03463122993707657, | |
| "rewards/margins": 0.39741796255111694, | |
| "rewards/rejected": -0.3627867102622986, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 0.9107563465061502, | |
| "grad_norm": 98.56849565071181, | |
| "learning_rate": 9.575238842447686e-09, | |
| "logits/chosen": -1.0055774450302124, | |
| "logits/rejected": -1.024350881576538, | |
| "logps/chosen": -1.4313799142837524, | |
| "logps/rejected": -1.6486520767211914, | |
| "loss": 1.3729, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.043267179280519485, | |
| "rewards/margins": 0.36729955673217773, | |
| "rewards/rejected": -0.4105667173862457, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.9159905783826223, | |
| "grad_norm": 50.17096951913472, | |
| "learning_rate": 8.489678529976242e-09, | |
| "logits/chosen": -0.9870138168334961, | |
| "logits/rejected": -1.058259129524231, | |
| "logps/chosen": -1.3640022277832031, | |
| "logps/rejected": -1.4488258361816406, | |
| "loss": 1.378, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.029376666992902756, | |
| "rewards/margins": 0.24675174057483673, | |
| "rewards/rejected": -0.2173750400543213, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.9212248102590945, | |
| "grad_norm": 120.75312862313243, | |
| "learning_rate": 7.468081109177027e-09, | |
| "logits/chosen": -1.0199341773986816, | |
| "logits/rejected": -1.0588958263397217, | |
| "logps/chosen": -1.4964534044265747, | |
| "logps/rejected": -1.6024501323699951, | |
| "loss": 1.3929, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": 0.12446703016757965, | |
| "rewards/margins": 0.37628185749053955, | |
| "rewards/rejected": -0.2518148422241211, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 0.9264590421355666, | |
| "grad_norm": 61.49650646596077, | |
| "learning_rate": 6.5107877852898176e-09, | |
| "logits/chosen": -0.9167430996894836, | |
| "logits/rejected": -0.9556129574775696, | |
| "logps/chosen": -1.3390752077102661, | |
| "logps/rejected": -1.6009235382080078, | |
| "loss": 1.423, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": 0.4433160424232483, | |
| "rewards/margins": 0.9942270517349243, | |
| "rewards/rejected": -0.5509108901023865, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 0.9316932740120387, | |
| "grad_norm": 71.97694388450074, | |
| "learning_rate": 5.6181182865083996e-09, | |
| "logits/chosen": -0.9885386228561401, | |
| "logits/rejected": -1.063110113143921, | |
| "logps/chosen": -1.2272593975067139, | |
| "logps/rejected": -1.5328699350357056, | |
| "loss": 1.3663, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": 0.1304434984922409, | |
| "rewards/margins": 0.5252709984779358, | |
| "rewards/rejected": -0.3948274552822113, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 0.9369275058885108, | |
| "grad_norm": 119.59418512940066, | |
| "learning_rate": 4.790370757193906e-09, | |
| "logits/chosen": -0.9046109914779663, | |
| "logits/rejected": -0.9860088229179382, | |
| "logps/chosen": -1.3347175121307373, | |
| "logps/rejected": -1.5166120529174805, | |
| "loss": 1.3675, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": 0.14059893786907196, | |
| "rewards/margins": 0.5057962536811829, | |
| "rewards/rejected": -0.36519724130630493, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 0.942161737764983, | |
| "grad_norm": 171.46537540400269, | |
| "learning_rate": 4.0278216582971145e-09, | |
| "logits/chosen": -0.9255016446113586, | |
| "logits/rejected": -0.9924243092536926, | |
| "logps/chosen": -1.1826350688934326, | |
| "logps/rejected": -1.5568852424621582, | |
| "loss": 1.3775, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": 0.08452494442462921, | |
| "rewards/margins": 0.3724609911441803, | |
| "rewards/rejected": -0.28793609142303467, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.942161737764983, | |
| "eval_logits/chosen": -1.0464354753494263, | |
| "eval_logits/rejected": -1.0728554725646973, | |
| "eval_logps/chosen": -1.2658635377883911, | |
| "eval_logps/rejected": -1.5124911069869995, | |
| "eval_loss": 1.3899798393249512, | |
| "eval_rewards/accuracies": 0.7400793433189392, | |
| "eval_rewards/chosen": 0.07923853397369385, | |
| "eval_rewards/margins": 0.42547234892845154, | |
| "eval_rewards/rejected": -0.3462338149547577, | |
| "eval_runtime": 265.6785, | |
| "eval_samples_per_second": 7.528, | |
| "eval_steps_per_second": 0.237, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.9473959696414551, | |
| "grad_norm": 162.20555081217532, | |
| "learning_rate": 3.3307256750225944e-09, | |
| "logits/chosen": -0.9436967968940735, | |
| "logits/rejected": -1.0339114665985107, | |
| "logps/chosen": -1.2900078296661377, | |
| "logps/rejected": -1.6287893056869507, | |
| "loss": 1.3833, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": 0.13000617921352386, | |
| "rewards/margins": 0.5302383303642273, | |
| "rewards/rejected": -0.4002321660518646, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 0.9526302015179272, | |
| "grad_norm": 69.377718963086, | |
| "learning_rate": 2.6993156317660636e-09, | |
| "logits/chosen": -0.9864269495010376, | |
| "logits/rejected": -1.118485450744629, | |
| "logps/chosen": -1.2873159646987915, | |
| "logps/rejected": -1.5646283626556396, | |
| "loss": 1.3589, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 0.2326822727918625, | |
| "rewards/margins": 0.4942099452018738, | |
| "rewards/rejected": -0.2615277171134949, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 0.9578644333943994, | |
| "grad_norm": 263.95656912812524, | |
| "learning_rate": 2.1338024143528142e-09, | |
| "logits/chosen": -0.9538179636001587, | |
| "logits/rejected": -1.0052978992462158, | |
| "logps/chosen": -1.3281259536743164, | |
| "logps/rejected": -1.455165147781372, | |
| "loss": 1.3882, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": 0.04303558170795441, | |
| "rewards/margins": 0.2979881465435028, | |
| "rewards/rejected": -0.2549525201320648, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 0.9630986652708715, | |
| "grad_norm": 39.18625236358022, | |
| "learning_rate": 1.6343748996036077e-09, | |
| "logits/chosen": -0.9945961833000183, | |
| "logits/rejected": -1.0426194667816162, | |
| "logps/chosen": -1.3158668279647827, | |
| "logps/rejected": -1.5129916667938232, | |
| "loss": 1.3731, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 0.048399608582258224, | |
| "rewards/margins": 0.4530261158943176, | |
| "rewards/rejected": -0.4046264588832855, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 0.9683328971473436, | |
| "grad_norm": 1286.9737737396774, | |
| "learning_rate": 1.2011998922513367e-09, | |
| "logits/chosen": -0.908880889415741, | |
| "logits/rejected": -1.0028820037841797, | |
| "logps/chosen": -1.2711695432662964, | |
| "logps/rejected": -1.590257167816162, | |
| "loss": 1.3902, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": 0.03833279758691788, | |
| "rewards/margins": 0.42264699935913086, | |
| "rewards/rejected": -0.3843142092227936, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.9735671290238157, | |
| "grad_norm": 637.3699428656323, | |
| "learning_rate": 8.34422069229701e-10, | |
| "logits/chosen": -0.9810758829116821, | |
| "logits/rejected": -1.034802794456482, | |
| "logps/chosen": -1.3561222553253174, | |
| "logps/rejected": -1.622603416442871, | |
| "loss": 1.3511, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": 0.09475534409284592, | |
| "rewards/margins": 0.5301617383956909, | |
| "rewards/rejected": -0.4354063868522644, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 0.9788013609002879, | |
| "grad_norm": 140.92434283058824, | |
| "learning_rate": 5.341639313521052e-10, | |
| "logits/chosen": -0.9513614773750305, | |
| "logits/rejected": -0.9637888073921204, | |
| "logps/chosen": -1.2574306726455688, | |
| "logps/rejected": -1.5153371095657349, | |
| "loss": 1.4036, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": 0.3424448072910309, | |
| "rewards/margins": 0.3421035706996918, | |
| "rewards/rejected": 0.00034122465876862407, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 0.98403559277676, | |
| "grad_norm": 36.89568780813949, | |
| "learning_rate": 3.005257623974966e-10, | |
| "logits/chosen": -0.9985333681106567, | |
| "logits/rejected": -1.0075907707214355, | |
| "logps/chosen": -1.2374012470245361, | |
| "logps/rejected": -1.4116681814193726, | |
| "loss": 1.367, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": 0.05069660022854805, | |
| "rewards/margins": 0.28373846411705017, | |
| "rewards/rejected": -0.2330418825149536, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 0.9892698246532321, | |
| "grad_norm": 73.10899677501034, | |
| "learning_rate": 1.3358559561642556e-10, | |
| "logits/chosen": -0.9993526339530945, | |
| "logits/rejected": -1.0922927856445312, | |
| "logps/chosen": -1.3391480445861816, | |
| "logps/rejected": -1.56029212474823, | |
| "loss": 1.3802, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": 0.056907374411821365, | |
| "rewards/margins": 0.32270345091819763, | |
| "rewards/rejected": -0.265796035528183, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 0.9945040565297043, | |
| "grad_norm": 121.55231563020173, | |
| "learning_rate": 3.339918766844807e-11, | |
| "logits/chosen": -0.980857253074646, | |
| "logits/rejected": -1.0078576803207397, | |
| "logps/chosen": -1.4365278482437134, | |
| "logps/rejected": -1.5911202430725098, | |
| "loss": 1.3827, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": 0.15399186313152313, | |
| "rewards/margins": 0.4474136233329773, | |
| "rewards/rejected": -0.29342177510261536, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.9945040565297043, | |
| "eval_logits/chosen": -1.0477176904678345, | |
| "eval_logits/rejected": -1.074251651763916, | |
| "eval_logps/chosen": -1.266432285308838, | |
| "eval_logps/rejected": -1.5128096342086792, | |
| "eval_loss": 1.3904120922088623, | |
| "eval_rewards/accuracies": 0.7400793433189392, | |
| "eval_rewards/chosen": 0.07070931047201157, | |
| "eval_rewards/margins": 0.421721875667572, | |
| "eval_rewards/rejected": -0.35101258754730225, | |
| "eval_runtime": 263.0385, | |
| "eval_samples_per_second": 7.603, | |
| "eval_steps_per_second": 0.24, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.9997382884061764, | |
| "grad_norm": 1926.9365225866572, | |
| "learning_rate": 0.0, | |
| "logits/chosen": -1.0182803869247437, | |
| "logits/rejected": -0.9923262596130371, | |
| "logps/chosen": -1.3955562114715576, | |
| "logps/rejected": -1.4240152835845947, | |
| "loss": 1.3808, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": 0.04296259954571724, | |
| "rewards/margins": 0.4077285826206207, | |
| "rewards/rejected": -0.3647659718990326, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 0.9997382884061764, | |
| "step": 1910, | |
| "total_flos": 0.0, | |
| "train_loss": 1.4068586224660824, | |
| "train_runtime": 26391.8957, | |
| "train_samples_per_second": 2.316, | |
| "train_steps_per_second": 0.072 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1910, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |