| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.985781990521327, | |
| "eval_steps": 500, | |
| "global_step": 315, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.04739336492890995, | |
| "grad_norm": 720.0, | |
| "learning_rate": 1.5625e-06, | |
| "log_odds_chosen": 3.653895616531372, | |
| "log_odds_ratio": -11.719749450683594, | |
| "logits/chosen": 278.66485595703125, | |
| "logits/rejected": 294.8902282714844, | |
| "logps/chosen": -22.453229904174805, | |
| "logps/rejected": -26.106348037719727, | |
| "loss": 103.8003, | |
| "nll_loss": 6.5329999923706055, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -11.226614952087402, | |
| "rewards/margins": 1.8265600204467773, | |
| "rewards/rejected": -13.053174018859863, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0947867298578199, | |
| "grad_norm": 304.0, | |
| "learning_rate": 3.125e-06, | |
| "log_odds_chosen": 3.6431682109832764, | |
| "log_odds_ratio": -7.441667079925537, | |
| "logits/chosen": 269.0749206542969, | |
| "logits/rejected": 282.71282958984375, | |
| "logps/chosen": -18.20242691040039, | |
| "logps/rejected": -21.8449649810791, | |
| "loss": 99.1968, | |
| "nll_loss": 5.850668907165527, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -9.101213455200195, | |
| "rewards/margins": 1.8212683200836182, | |
| "rewards/rejected": -10.92248249053955, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.14218009478672985, | |
| "grad_norm": 139.0, | |
| "learning_rate": 4.6875000000000004e-06, | |
| "log_odds_chosen": 5.687544345855713, | |
| "log_odds_ratio": -7.769516944885254, | |
| "logits/chosen": 264.29461669921875, | |
| "logits/rejected": 291.05987548828125, | |
| "logps/chosen": -18.975749969482422, | |
| "logps/rejected": -24.661663055419922, | |
| "loss": 95.6825, | |
| "nll_loss": 6.186770439147949, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -9.487874984741211, | |
| "rewards/margins": 2.8429577350616455, | |
| "rewards/rejected": -12.330831527709961, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.1895734597156398, | |
| "grad_norm": 153.0, | |
| "learning_rate": 6.25e-06, | |
| "log_odds_chosen": -1.2323758602142334, | |
| "log_odds_ratio": -10.211746215820312, | |
| "logits/chosen": 292.7019958496094, | |
| "logits/rejected": 290.5914001464844, | |
| "logps/chosen": -20.584659576416016, | |
| "logps/rejected": -19.353229522705078, | |
| "loss": 99.4907, | |
| "nll_loss": 6.084651470184326, | |
| "rewards/accuracies": 0.4749999940395355, | |
| "rewards/chosen": -10.292329788208008, | |
| "rewards/margins": -0.6157160997390747, | |
| "rewards/rejected": -9.676614761352539, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.23696682464454977, | |
| "grad_norm": 81.5, | |
| "learning_rate": 7.8125e-06, | |
| "log_odds_chosen": 0.21437835693359375, | |
| "log_odds_ratio": -9.132922172546387, | |
| "logits/chosen": 278.7820129394531, | |
| "logits/rejected": 288.18951416015625, | |
| "logps/chosen": -18.19257354736328, | |
| "logps/rejected": -18.40748405456543, | |
| "loss": 88.5475, | |
| "nll_loss": 4.977797508239746, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -9.09628677368164, | |
| "rewards/margins": 0.10745634883642197, | |
| "rewards/rejected": -9.203742027282715, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.2843601895734597, | |
| "grad_norm": 124.0, | |
| "learning_rate": 9.375000000000001e-06, | |
| "log_odds_chosen": 1.75958251953125, | |
| "log_odds_ratio": -9.482071876525879, | |
| "logits/chosen": 280.55633544921875, | |
| "logits/rejected": 301.93743896484375, | |
| "logps/chosen": -19.764301300048828, | |
| "logps/rejected": -21.52188491821289, | |
| "loss": 98.2752, | |
| "nll_loss": 6.01733922958374, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -9.882150650024414, | |
| "rewards/margins": 0.8787924647331238, | |
| "rewards/rejected": -10.760942459106445, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.33175355450236965, | |
| "grad_norm": 239.0, | |
| "learning_rate": 9.997227514697568e-06, | |
| "log_odds_chosen": -0.15938568115234375, | |
| "log_odds_ratio": -7.877626895904541, | |
| "logits/chosen": 297.74163818359375, | |
| "logits/rejected": 288.8265075683594, | |
| "logps/chosen": -17.699235916137695, | |
| "logps/rejected": -17.539152145385742, | |
| "loss": 83.3604, | |
| "nll_loss": 5.2883453369140625, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -8.849617958068848, | |
| "rewards/margins": -0.08004142343997955, | |
| "rewards/rejected": -8.769576072692871, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.3791469194312796, | |
| "grad_norm": 410.0, | |
| "learning_rate": 9.980295682286924e-06, | |
| "log_odds_chosen": 2.3873629570007324, | |
| "log_odds_ratio": -3.3676114082336426, | |
| "logits/chosen": 300.3001708984375, | |
| "logits/rejected": 310.92047119140625, | |
| "logps/chosen": -10.030499458312988, | |
| "logps/rejected": -12.416150093078613, | |
| "loss": 73.3518, | |
| "nll_loss": 4.223599433898926, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -5.015249729156494, | |
| "rewards/margins": 1.1928250789642334, | |
| "rewards/rejected": -6.208075046539307, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.4265402843601896, | |
| "grad_norm": 308.0, | |
| "learning_rate": 9.94802437440896e-06, | |
| "log_odds_chosen": 1.004201889038086, | |
| "log_odds_ratio": -3.536766767501831, | |
| "logits/chosen": 309.2627258300781, | |
| "logits/rejected": 307.8229675292969, | |
| "logps/chosen": -8.605690002441406, | |
| "logps/rejected": -9.609885215759277, | |
| "loss": 52.2283, | |
| "nll_loss": 3.156938076019287, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -4.302845001220703, | |
| "rewards/margins": 0.5020972490310669, | |
| "rewards/rejected": -4.804942607879639, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.47393364928909953, | |
| "grad_norm": 112.0, | |
| "learning_rate": 9.90051298775959e-06, | |
| "log_odds_chosen": 0.10521616786718369, | |
| "log_odds_ratio": -1.6550161838531494, | |
| "logits/chosen": 334.4619445800781, | |
| "logits/rejected": 343.36077880859375, | |
| "logps/chosen": -3.9240214824676514, | |
| "logps/rejected": -4.02620792388916, | |
| "loss": 36.9834, | |
| "nll_loss": 2.3400511741638184, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -1.9620107412338257, | |
| "rewards/margins": 0.05109361559152603, | |
| "rewards/rejected": -2.01310396194458, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.5213270142180095, | |
| "grad_norm": 50.25, | |
| "learning_rate": 9.837907858981536e-06, | |
| "log_odds_chosen": 0.29174837470054626, | |
| "log_odds_ratio": -0.8331424593925476, | |
| "logits/chosen": 334.2044677734375, | |
| "logits/rejected": 351.3958435058594, | |
| "logps/chosen": -1.8196113109588623, | |
| "logps/rejected": -2.076477527618408, | |
| "loss": 28.4647, | |
| "nll_loss": 1.7829310894012451, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.9098056554794312, | |
| "rewards/margins": 0.1284329891204834, | |
| "rewards/rejected": -1.038238763809204, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.5687203791469194, | |
| "grad_norm": 47.75, | |
| "learning_rate": 9.760401813942641e-06, | |
| "log_odds_chosen": 0.35578861832618713, | |
| "log_odds_ratio": -0.7586129903793335, | |
| "logits/chosen": 327.1716613769531, | |
| "logits/rejected": 352.7845764160156, | |
| "logps/chosen": -1.6458431482315063, | |
| "logps/rejected": -1.9578958749771118, | |
| "loss": 26.7026, | |
| "nll_loss": 1.681125283241272, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.8229215741157532, | |
| "rewards/margins": 0.156026229262352, | |
| "rewards/rejected": -0.9789479374885559, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.6161137440758294, | |
| "grad_norm": 34.5, | |
| "learning_rate": 9.668233573825794e-06, | |
| "log_odds_chosen": 0.31834372878074646, | |
| "log_odds_ratio": -0.7771207094192505, | |
| "logits/chosen": 322.9273681640625, | |
| "logits/rejected": 341.0210266113281, | |
| "logps/chosen": -1.5688014030456543, | |
| "logps/rejected": -1.8273004293441772, | |
| "loss": 26.4611, | |
| "nll_loss": 1.6627897024154663, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.7844007015228271, | |
| "rewards/margins": 0.12924957275390625, | |
| "rewards/rejected": -0.9136502146720886, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.6635071090047393, | |
| "grad_norm": 47.25, | |
| "learning_rate": 9.56168701985981e-06, | |
| "log_odds_chosen": 0.31223994493484497, | |
| "log_odds_ratio": -0.7746738791465759, | |
| "logits/chosen": 353.25103759765625, | |
| "logits/rejected": 372.4833984375, | |
| "logps/chosen": -1.557328701019287, | |
| "logps/rejected": -1.8247381448745728, | |
| "loss": 25.781, | |
| "nll_loss": 1.557205319404602, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.7786643505096436, | |
| "rewards/margins": 0.13370472192764282, | |
| "rewards/rejected": -0.9123690724372864, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.7109004739336493, | |
| "grad_norm": 36.0, | |
| "learning_rate": 9.441090318955843e-06, | |
| "log_odds_chosen": 0.3869365155696869, | |
| "log_odds_ratio": -0.778424084186554, | |
| "logits/chosen": 346.4082336425781, | |
| "logits/rejected": 360.0977783203125, | |
| "logps/chosen": -1.5176082849502563, | |
| "logps/rejected": -1.8383760452270508, | |
| "loss": 25.2293, | |
| "nll_loss": 1.5534436702728271, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -0.7588041424751282, | |
| "rewards/margins": 0.1603839099407196, | |
| "rewards/rejected": -0.9191880226135254, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.7582938388625592, | |
| "grad_norm": 83.0, | |
| "learning_rate": 9.306814912942445e-06, | |
| "log_odds_chosen": 0.3693597912788391, | |
| "log_odds_ratio": -0.7350634336471558, | |
| "logits/chosen": 346.6322937011719, | |
| "logits/rejected": 358.0809326171875, | |
| "logps/chosen": -1.4770724773406982, | |
| "logps/rejected": -1.7809489965438843, | |
| "loss": 24.6783, | |
| "nll_loss": 1.5355550050735474, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.7385362386703491, | |
| "rewards/margins": 0.1519382894039154, | |
| "rewards/rejected": -0.8904744982719421, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.8056872037914692, | |
| "grad_norm": 28.5, | |
| "learning_rate": 9.159274374512444e-06, | |
| "log_odds_chosen": 0.38480785489082336, | |
| "log_odds_ratio": -0.7371683120727539, | |
| "logits/chosen": 356.5403137207031, | |
| "logits/rejected": 372.90631103515625, | |
| "logps/chosen": -1.5246539115905762, | |
| "logps/rejected": -1.854161024093628, | |
| "loss": 24.3913, | |
| "nll_loss": 1.5450783967971802, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.7623269557952881, | |
| "rewards/margins": 0.16475361585617065, | |
| "rewards/rejected": -0.927080512046814, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.8530805687203792, | |
| "grad_norm": 44.75, | |
| "learning_rate": 8.99892313340537e-06, | |
| "log_odds_chosen": 0.2174159735441208, | |
| "log_odds_ratio": -0.8837092518806458, | |
| "logits/chosen": 354.29132080078125, | |
| "logits/rejected": 361.8697814941406, | |
| "logps/chosen": -1.6351244449615479, | |
| "logps/rejected": -1.8084831237792969, | |
| "loss": 25.1868, | |
| "nll_loss": 1.6281230449676514, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -0.8175622224807739, | |
| "rewards/margins": 0.08667941391468048, | |
| "rewards/rejected": -0.9042415618896484, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.9004739336492891, | |
| "grad_norm": 41.0, | |
| "learning_rate": 8.826255076748823e-06, | |
| "log_odds_chosen": 0.14011089503765106, | |
| "log_odds_ratio": -0.7930618524551392, | |
| "logits/chosen": 362.77874755859375, | |
| "logits/rejected": 373.4800720214844, | |
| "logps/chosen": -1.4461233615875244, | |
| "logps/rejected": -1.5512562990188599, | |
| "loss": 24.0704, | |
| "nll_loss": 1.4283101558685303, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.7230616807937622, | |
| "rewards/margins": 0.05256646126508713, | |
| "rewards/rejected": -0.7756281495094299, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.9478672985781991, | |
| "grad_norm": 70.5, | |
| "learning_rate": 8.641802027869774e-06, | |
| "log_odds_chosen": 0.2718258798122406, | |
| "log_odds_ratio": -0.7795127630233765, | |
| "logits/chosen": 367.4668273925781, | |
| "logits/rejected": 367.6637878417969, | |
| "logps/chosen": -1.4626038074493408, | |
| "logps/rejected": -1.709529161453247, | |
| "loss": 24.1424, | |
| "nll_loss": 1.538220763206482, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.7313019037246704, | |
| "rewards/margins": 0.12346267700195312, | |
| "rewards/rejected": -0.8547645807266235, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.995260663507109, | |
| "grad_norm": 31.75, | |
| "learning_rate": 8.446132108261136e-06, | |
| "log_odds_chosen": 0.2905386984348297, | |
| "log_odds_ratio": -0.6851338744163513, | |
| "logits/chosen": 360.0987548828125, | |
| "logits/rejected": 378.06414794921875, | |
| "logps/chosen": -1.3334792852401733, | |
| "logps/rejected": -1.5764299631118774, | |
| "loss": 23.5124, | |
| "nll_loss": 1.4705913066864014, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.6667396426200867, | |
| "rewards/margins": 0.12147532403469086, | |
| "rewards/rejected": -0.7882149815559387, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.042654028436019, | |
| "grad_norm": 30.125, | |
| "learning_rate": 8.23984798774876e-06, | |
| "log_odds_chosen": 0.2926952838897705, | |
| "log_odds_ratio": -0.7383573055267334, | |
| "logits/chosen": 356.43695068359375, | |
| "logits/rejected": 380.3526611328125, | |
| "logps/chosen": -1.3705774545669556, | |
| "logps/rejected": -1.615526556968689, | |
| "loss": 22.8403, | |
| "nll_loss": 1.443694829940796, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.6852887272834778, | |
| "rewards/margins": 0.12247464805841446, | |
| "rewards/rejected": -0.8077632784843445, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.0900473933649288, | |
| "grad_norm": 36.25, | |
| "learning_rate": 8.023585028248435e-06, | |
| "log_odds_chosen": 0.456248939037323, | |
| "log_odds_ratio": -0.6926681399345398, | |
| "logits/chosen": 356.6784973144531, | |
| "logits/rejected": 381.8028259277344, | |
| "logps/chosen": -1.3445045948028564, | |
| "logps/rejected": -1.689343810081482, | |
| "loss": 22.9507, | |
| "nll_loss": 1.427918553352356, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.6722522974014282, | |
| "rewards/margins": 0.17241965234279633, | |
| "rewards/rejected": -0.844671905040741, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.1374407582938388, | |
| "grad_norm": 32.0, | |
| "learning_rate": 7.798009326830167e-06, | |
| "log_odds_chosen": 0.44028574228286743, | |
| "log_odds_ratio": -0.6239734888076782, | |
| "logits/chosen": 351.1431579589844, | |
| "logits/rejected": 363.55499267578125, | |
| "logps/chosen": -1.2665112018585205, | |
| "logps/rejected": -1.6185169219970703, | |
| "loss": 22.7249, | |
| "nll_loss": 1.449618935585022, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.6332556009292603, | |
| "rewards/margins": 0.1760028749704361, | |
| "rewards/rejected": -0.8092584609985352, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.1848341232227488, | |
| "grad_norm": 30.375, | |
| "learning_rate": 7.563815664117173e-06, | |
| "log_odds_chosen": 0.3645080029964447, | |
| "log_odds_ratio": -0.6768008470535278, | |
| "logits/chosen": 369.084716796875, | |
| "logits/rejected": 371.45977783203125, | |
| "logps/chosen": -1.303069829940796, | |
| "logps/rejected": -1.5789194107055664, | |
| "loss": 21.767, | |
| "nll_loss": 1.3582165241241455, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.651534914970398, | |
| "rewards/margins": 0.13792480528354645, | |
| "rewards/rejected": -0.7894597053527832, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.2322274881516588, | |
| "grad_norm": 31.125, | |
| "learning_rate": 7.321725364338566e-06, | |
| "log_odds_chosen": 0.15869663655757904, | |
| "log_odds_ratio": -0.7532464861869812, | |
| "logits/chosen": 353.90032958984375, | |
| "logits/rejected": 370.54364013671875, | |
| "logps/chosen": -1.2809131145477295, | |
| "logps/rejected": -1.387822151184082, | |
| "loss": 21.6046, | |
| "nll_loss": 1.3660101890563965, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.6404565572738647, | |
| "rewards/margins": 0.053454458713531494, | |
| "rewards/rejected": -0.693911075592041, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.2796208530805688, | |
| "grad_norm": 51.25, | |
| "learning_rate": 7.072484073626872e-06, | |
| "log_odds_chosen": 0.4641496241092682, | |
| "log_odds_ratio": -0.6374386548995972, | |
| "logits/chosen": 360.2814636230469, | |
| "logits/rejected": 370.50323486328125, | |
| "logps/chosen": -1.2486190795898438, | |
| "logps/rejected": -1.6152652502059937, | |
| "loss": 22.2528, | |
| "nll_loss": 1.3934818506240845, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.6243095397949219, | |
| "rewards/margins": 0.18332314491271973, | |
| "rewards/rejected": -0.8076326251029968, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.3270142180094786, | |
| "grad_norm": 50.25, | |
| "learning_rate": 6.816859463403271e-06, | |
| "log_odds_chosen": 0.2742091715335846, | |
| "log_odds_ratio": -0.7456444501876831, | |
| "logits/chosen": 355.84222412109375, | |
| "logits/rejected": 354.6243591308594, | |
| "logps/chosen": -1.2655051946640015, | |
| "logps/rejected": -1.4896043539047241, | |
| "loss": 22.9608, | |
| "nll_loss": 1.401302695274353, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.6327525973320007, | |
| "rewards/margins": 0.11204960197210312, | |
| "rewards/rejected": -0.7448021769523621, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.3744075829383886, | |
| "grad_norm": 42.75, | |
| "learning_rate": 6.555638865924221e-06, | |
| "log_odds_chosen": 0.5057348012924194, | |
| "log_odds_ratio": -0.5890580415725708, | |
| "logits/chosen": 345.1250915527344, | |
| "logits/rejected": 369.5255432128906, | |
| "logps/chosen": -1.169845461845398, | |
| "logps/rejected": -1.5732532739639282, | |
| "loss": 22.323, | |
| "nll_loss": 1.3889384269714355, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.584922730922699, | |
| "rewards/margins": 0.20170390605926514, | |
| "rewards/rejected": -0.7866266369819641, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.4218009478672986, | |
| "grad_norm": 61.25, | |
| "learning_rate": 6.289626849272062e-06, | |
| "log_odds_chosen": 0.21882423758506775, | |
| "log_odds_ratio": -0.7196656465530396, | |
| "logits/chosen": 353.15582275390625, | |
| "logits/rejected": 339.25982666015625, | |
| "logps/chosen": -1.1948503255844116, | |
| "logps/rejected": -1.3711706399917603, | |
| "loss": 22.3161, | |
| "nll_loss": 1.4074368476867676, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.5974251627922058, | |
| "rewards/margins": 0.08816017955541611, | |
| "rewards/rejected": -0.6855853199958801, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.4691943127962086, | |
| "grad_norm": 36.25, | |
| "learning_rate": 6.0196427392587085e-06, | |
| "log_odds_chosen": 0.34388962388038635, | |
| "log_odds_ratio": -0.6600432395935059, | |
| "logits/chosen": 352.5672302246094, | |
| "logits/rejected": 373.4999084472656, | |
| "logps/chosen": -1.137880563735962, | |
| "logps/rejected": -1.3981521129608154, | |
| "loss": 21.9751, | |
| "nll_loss": 1.3336918354034424, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.568940281867981, | |
| "rewards/margins": 0.13013575971126556, | |
| "rewards/rejected": -0.6990760564804077, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.5165876777251186, | |
| "grad_norm": 42.0, | |
| "learning_rate": 5.746518095875033e-06, | |
| "log_odds_chosen": 0.24596929550170898, | |
| "log_odds_ratio": -0.7056238651275635, | |
| "logits/chosen": 355.90606689453125, | |
| "logits/rejected": 360.6635437011719, | |
| "logps/chosen": -1.2390286922454834, | |
| "logps/rejected": -1.4282642602920532, | |
| "loss": 22.5868, | |
| "nll_loss": 1.4504239559173584, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.6195143461227417, | |
| "rewards/margins": 0.09461767971515656, | |
| "rewards/rejected": -0.7141321301460266, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.5639810426540284, | |
| "grad_norm": 34.75, | |
| "learning_rate": 5.471094152058592e-06, | |
| "log_odds_chosen": 0.29543933272361755, | |
| "log_odds_ratio": -0.6770726442337036, | |
| "logits/chosen": 362.8612060546875, | |
| "logits/rejected": 357.1744384765625, | |
| "logps/chosen": -1.2012829780578613, | |
| "logps/rejected": -1.4275243282318115, | |
| "loss": 21.8669, | |
| "nll_loss": 1.353208303451538, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.6006414890289307, | |
| "rewards/margins": 0.11312057822942734, | |
| "rewards/rejected": -0.7137621641159058, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.6113744075829384, | |
| "grad_norm": 30.625, | |
| "learning_rate": 5.1942192226683385e-06, | |
| "log_odds_chosen": 0.359115868806839, | |
| "log_odds_ratio": -0.6642639636993408, | |
| "logits/chosen": 354.8602294921875, | |
| "logits/rejected": 367.0655212402344, | |
| "logps/chosen": -1.2128181457519531, | |
| "logps/rejected": -1.4822323322296143, | |
| "loss": 21.8466, | |
| "nll_loss": 1.3282759189605713, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -0.6064090728759766, | |
| "rewards/margins": 0.13470709323883057, | |
| "rewards/rejected": -0.7411161661148071, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.6587677725118484, | |
| "grad_norm": 38.75, | |
| "learning_rate": 4.916746091646808e-06, | |
| "log_odds_chosen": 0.20625083148479462, | |
| "log_odds_ratio": -0.7107352018356323, | |
| "logits/chosen": 358.31890869140625, | |
| "logits/rejected": 366.37408447265625, | |
| "logps/chosen": -1.1632264852523804, | |
| "logps/rejected": -1.3307292461395264, | |
| "loss": 21.8714, | |
| "nll_loss": 1.3562901020050049, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.5816132426261902, | |
| "rewards/margins": 0.08375142514705658, | |
| "rewards/rejected": -0.6653646230697632, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.7061611374407581, | |
| "grad_norm": 63.25, | |
| "learning_rate": 4.6395293854173395e-06, | |
| "log_odds_chosen": 0.19339993596076965, | |
| "log_odds_ratio": -0.7509890198707581, | |
| "logits/chosen": 347.43951416015625, | |
| "logits/rejected": 358.0630798339844, | |
| "logps/chosen": -1.2565109729766846, | |
| "logps/rejected": -1.391413688659668, | |
| "loss": 22.1212, | |
| "nll_loss": 1.416812539100647, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.6282554864883423, | |
| "rewards/margins": 0.06745139509439468, | |
| "rewards/rejected": -0.695706844329834, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.7535545023696684, | |
| "grad_norm": 27.375, | |
| "learning_rate": 4.363422940606435e-06, | |
| "log_odds_chosen": 0.4669272005558014, | |
| "log_odds_ratio": -0.6223559975624084, | |
| "logits/chosen": 350.216796875, | |
| "logits/rejected": 363.3144836425781, | |
| "logps/chosen": -1.1427295207977295, | |
| "logps/rejected": -1.5226044654846191, | |
| "loss": 21.106, | |
| "nll_loss": 1.2993601560592651, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -0.5713647603988647, | |
| "rewards/margins": 0.189937561750412, | |
| "rewards/rejected": -0.7613022327423096, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.8009478672985781, | |
| "grad_norm": 31.0, | |
| "learning_rate": 4.089277174198694e-06, | |
| "log_odds_chosen": 0.23544028401374817, | |
| "log_odds_ratio": -0.722697913646698, | |
| "logits/chosen": 353.28558349609375, | |
| "logits/rejected": 369.5932922363281, | |
| "logps/chosen": -1.1963495016098022, | |
| "logps/rejected": -1.3738784790039062, | |
| "loss": 21.9431, | |
| "nll_loss": 1.3263076543807983, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.5981747508049011, | |
| "rewards/margins": 0.08876445889472961, | |
| "rewards/rejected": -0.6869392395019531, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.8483412322274881, | |
| "grad_norm": 37.5, | |
| "learning_rate": 3.817936464224367e-06, | |
| "log_odds_chosen": 0.41035833954811096, | |
| "log_odds_ratio": -0.6357004642486572, | |
| "logits/chosen": 342.2323303222656, | |
| "logits/rejected": 361.4287414550781, | |
| "logps/chosen": -1.1266499757766724, | |
| "logps/rejected": -1.4432952404022217, | |
| "loss": 21.7598, | |
| "nll_loss": 1.3171429634094238, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.5633249878883362, | |
| "rewards/margins": 0.15832264721393585, | |
| "rewards/rejected": -0.7216476202011108, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.8957345971563981, | |
| "grad_norm": 33.5, | |
| "learning_rate": 3.55023654904709e-06, | |
| "log_odds_chosen": 0.1588168442249298, | |
| "log_odds_ratio": -0.7843244671821594, | |
| "logits/chosen": 346.522216796875, | |
| "logits/rejected": 360.4036865234375, | |
| "logps/chosen": -1.248828649520874, | |
| "logps/rejected": -1.3542238473892212, | |
| "loss": 22.3502, | |
| "nll_loss": 1.417950987815857, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -0.624414324760437, | |
| "rewards/margins": 0.05269758030772209, | |
| "rewards/rejected": -0.6771119236946106, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.943127962085308, | |
| "grad_norm": 38.25, | |
| "learning_rate": 3.2870019532620744e-06, | |
| "log_odds_chosen": 0.19328053295612335, | |
| "log_odds_ratio": -0.6914501786231995, | |
| "logits/chosen": 339.40057373046875, | |
| "logits/rejected": 364.0176696777344, | |
| "logps/chosen": -1.1170837879180908, | |
| "logps/rejected": -1.2566817998886108, | |
| "loss": 20.5383, | |
| "nll_loss": 1.269911527633667, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.5585418939590454, | |
| "rewards/margins": 0.06979899108409882, | |
| "rewards/rejected": -0.6283408999443054, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.9905213270142181, | |
| "grad_norm": 28.25, | |
| "learning_rate": 3.0290434481330746e-06, | |
| "log_odds_chosen": 0.2612162232398987, | |
| "log_odds_ratio": -0.672803521156311, | |
| "logits/chosen": 355.0847473144531, | |
| "logits/rejected": 357.27752685546875, | |
| "logps/chosen": -1.1057822704315186, | |
| "logps/rejected": -1.2985315322875977, | |
| "loss": 20.8159, | |
| "nll_loss": 1.2727569341659546, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.5528911352157593, | |
| "rewards/margins": 0.09637459367513657, | |
| "rewards/rejected": -0.6492657661437988, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.037914691943128, | |
| "grad_norm": 28.875, | |
| "learning_rate": 2.77715555439007e-06, | |
| "log_odds_chosen": 0.3204149305820465, | |
| "log_odds_ratio": -0.6769343018531799, | |
| "logits/chosen": 358.9181823730469, | |
| "logits/rejected": 361.39923095703125, | |
| "logps/chosen": -1.1455085277557373, | |
| "logps/rejected": -1.3854711055755615, | |
| "loss": 20.8727, | |
| "nll_loss": 1.2813211679458618, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.5727542638778687, | |
| "rewards/margins": 0.1199813038110733, | |
| "rewards/rejected": -0.6927355527877808, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.085308056872038, | |
| "grad_norm": 33.25, | |
| "learning_rate": 2.532114095079137e-06, | |
| "log_odds_chosen": 0.41654521226882935, | |
| "log_odds_ratio": -0.6568773984909058, | |
| "logits/chosen": 366.5960388183594, | |
| "logits/rejected": 352.90496826171875, | |
| "logps/chosen": -1.1304197311401367, | |
| "logps/rejected": -1.4528872966766357, | |
| "loss": 20.8523, | |
| "nll_loss": 1.349096655845642, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -0.5652098655700684, | |
| "rewards/margins": 0.1612338125705719, | |
| "rewards/rejected": -0.7264436483383179, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.132701421800948, | |
| "grad_norm": 26.75, | |
| "learning_rate": 2.2946738060017947e-06, | |
| "log_odds_chosen": 0.2243175506591797, | |
| "log_odds_ratio": -0.7143479585647583, | |
| "logits/chosen": 332.3900451660156, | |
| "logits/rejected": 345.49566650390625, | |
| "logps/chosen": -1.181205153465271, | |
| "logps/rejected": -1.3491504192352295, | |
| "loss": 20.5804, | |
| "nll_loss": 1.364635705947876, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.5906025767326355, | |
| "rewards/margins": 0.08397253602743149, | |
| "rewards/rejected": -0.6745752096176147, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.1800947867298577, | |
| "grad_norm": 31.75, | |
| "learning_rate": 2.0655660111037685e-06, | |
| "log_odds_chosen": 0.3350989818572998, | |
| "log_odds_ratio": -0.6724013090133667, | |
| "logits/chosen": 369.9901123046875, | |
| "logits/rejected": 377.08782958984375, | |
| "logps/chosen": -1.1567586660385132, | |
| "logps/rejected": -1.401942253112793, | |
| "loss": 20.7933, | |
| "nll_loss": 1.311959981918335, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.5783793330192566, | |
| "rewards/margins": 0.12259165942668915, | |
| "rewards/rejected": -0.7009711265563965, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.227488151658768, | |
| "grad_norm": 33.25, | |
| "learning_rate": 1.8454963699730471e-06, | |
| "log_odds_chosen": 0.2978705167770386, | |
| "log_odds_ratio": -0.6846665143966675, | |
| "logits/chosen": 344.5165100097656, | |
| "logits/rejected": 354.6488342285156, | |
| "logps/chosen": -1.1545495986938477, | |
| "logps/rejected": -1.3718467950820923, | |
| "loss": 20.9976, | |
| "nll_loss": 1.339095950126648, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.5772747993469238, | |
| "rewards/margins": 0.10864856094121933, | |
| "rewards/rejected": -0.6859233975410461, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 2.2748815165876777, | |
| "grad_norm": 30.75, | |
| "learning_rate": 1.6351427043849955e-06, | |
| "log_odds_chosen": 0.3817955255508423, | |
| "log_odds_ratio": -0.6393663287162781, | |
| "logits/chosen": 349.63629150390625, | |
| "logits/rejected": 357.9341735839844, | |
| "logps/chosen": -1.0905182361602783, | |
| "logps/rejected": -1.3862955570220947, | |
| "loss": 20.6181, | |
| "nll_loss": 1.3354963064193726, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.5452591180801392, | |
| "rewards/margins": 0.14788874983787537, | |
| "rewards/rejected": -0.6931477785110474, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.322274881516588, | |
| "grad_norm": 26.875, | |
| "learning_rate": 1.4351529105888735e-06, | |
| "log_odds_chosen": 0.31271064281463623, | |
| "log_odds_ratio": -0.6530163884162903, | |
| "logits/chosen": 331.0179138183594, | |
| "logits/rejected": 351.15350341796875, | |
| "logps/chosen": -1.0642945766448975, | |
| "logps/rejected": -1.2827186584472656, | |
| "loss": 20.275, | |
| "nll_loss": 1.2534034252166748, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.5321472883224487, | |
| "rewards/margins": 0.1092119961977005, | |
| "rewards/rejected": -0.6413593292236328, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 2.3696682464454977, | |
| "grad_norm": 28.625, | |
| "learning_rate": 1.2461429637659466e-06, | |
| "log_odds_chosen": 0.5228301882743835, | |
| "log_odds_ratio": -0.573235273361206, | |
| "logits/chosen": 358.1142883300781, | |
| "logits/rejected": 357.08905029296875, | |
| "logps/chosen": -1.084540843963623, | |
| "logps/rejected": -1.4777616262435913, | |
| "loss": 19.7075, | |
| "nll_loss": 1.2512117624282837, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -0.5422704219818115, | |
| "rewards/margins": 0.19661036133766174, | |
| "rewards/rejected": -0.7388808131217957, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.4170616113744074, | |
| "grad_norm": 26.75, | |
| "learning_rate": 1.0686950208055486e-06, | |
| "log_odds_chosen": 0.421553373336792, | |
| "log_odds_ratio": -0.6212750673294067, | |
| "logits/chosen": 343.3612976074219, | |
| "logits/rejected": 357.06671142578125, | |
| "logps/chosen": -1.0726759433746338, | |
| "logps/rejected": -1.3855555057525635, | |
| "loss": 20.2572, | |
| "nll_loss": 1.2650914192199707, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.5363379716873169, | |
| "rewards/margins": 0.15643975138664246, | |
| "rewards/rejected": -0.6927777528762817, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 2.4644549763033177, | |
| "grad_norm": 27.75, | |
| "learning_rate": 9.033556272426075e-07, | |
| "log_odds_chosen": 0.4636826515197754, | |
| "log_odds_ratio": -0.6046215891838074, | |
| "logits/chosen": 346.56195068359375, | |
| "logits/rejected": 366.7275390625, | |
| "logps/chosen": -1.0248494148254395, | |
| "logps/rejected": -1.3430243730545044, | |
| "loss": 20.2488, | |
| "nll_loss": 1.2966502904891968, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -0.5124247074127197, | |
| "rewards/margins": 0.1590874344110489, | |
| "rewards/rejected": -0.6715121865272522, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.5118483412322274, | |
| "grad_norm": 27.875, | |
| "learning_rate": 7.506340338793111e-07, | |
| "log_odds_chosen": 0.3095242977142334, | |
| "log_odds_ratio": -0.6864131689071655, | |
| "logits/chosen": 350.9102783203125, | |
| "logits/rejected": 345.47369384765625, | |
| "logps/chosen": -1.0581636428833008, | |
| "logps/rejected": -1.2898659706115723, | |
| "loss": 19.9452, | |
| "nll_loss": 1.1947921514511108, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -0.5290818214416504, | |
| "rewards/margins": 0.11585123836994171, | |
| "rewards/rejected": -0.6449329853057861, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 2.5592417061611377, | |
| "grad_norm": 26.125, | |
| "learning_rate": 6.110006282757897e-07, | |
| "log_odds_chosen": 0.3585304915904999, | |
| "log_odds_ratio": -0.6409192681312561, | |
| "logits/chosen": 346.3677673339844, | |
| "logits/rejected": 351.1196594238281, | |
| "logps/chosen": -1.0988967418670654, | |
| "logps/rejected": -1.3483343124389648, | |
| "loss": 20.3609, | |
| "nll_loss": 1.2682546377182007, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.5494483709335327, | |
| "rewards/margins": 0.1247188076376915, | |
| "rewards/rejected": -0.6741671562194824, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.6066350710900474, | |
| "grad_norm": 32.25, | |
| "learning_rate": 4.848854859408731e-07, | |
| "log_odds_chosen": 0.3773255944252014, | |
| "log_odds_ratio": -0.6349105834960938, | |
| "logits/chosen": 331.9405212402344, | |
| "logits/rejected": 360.78240966796875, | |
| "logps/chosen": -1.0606693029403687, | |
| "logps/rejected": -1.3333818912506104, | |
| "loss": 20.7172, | |
| "nll_loss": 1.2634260654449463, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.5303346514701843, | |
| "rewards/margins": 0.13635624945163727, | |
| "rewards/rejected": -0.6666909456253052, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 2.654028436018957, | |
| "grad_norm": 35.5, | |
| "learning_rate": 3.7267704568529015e-07, | |
| "log_odds_chosen": 0.3883630633354187, | |
| "log_odds_ratio": -0.6591242551803589, | |
| "logits/chosen": 358.720458984375, | |
| "logits/rejected": 362.07733154296875, | |
| "logps/chosen": -1.1015071868896484, | |
| "logps/rejected": -1.3767952919006348, | |
| "loss": 20.1116, | |
| "nll_loss": 1.2957426309585571, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.5507535934448242, | |
| "rewards/margins": 0.13764409720897675, | |
| "rewards/rejected": -0.6883976459503174, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.7014218009478674, | |
| "grad_norm": 27.125, | |
| "learning_rate": 2.7472091321728067e-07, | |
| "log_odds_chosen": 0.4095282554626465, | |
| "log_odds_ratio": -0.6598828434944153, | |
| "logits/chosen": 338.1583557128906, | |
| "logits/rejected": 358.77423095703125, | |
| "logps/chosen": -1.065172553062439, | |
| "logps/rejected": -1.3705942630767822, | |
| "loss": 21.1264, | |
| "nll_loss": 1.314082384109497, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.5325862765312195, | |
| "rewards/margins": 0.15271088480949402, | |
| "rewards/rejected": -0.6852971315383911, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 2.748815165876777, | |
| "grad_norm": 28.125, | |
| "learning_rate": 1.9131879666558385e-07, | |
| "log_odds_chosen": 0.3623240888118744, | |
| "log_odds_ratio": -0.6543713212013245, | |
| "logits/chosen": 346.6483459472656, | |
| "logits/rejected": 349.0867614746094, | |
| "logps/chosen": -1.1220111846923828, | |
| "logps/rejected": -1.3769835233688354, | |
| "loss": 20.8844, | |
| "nll_loss": 1.2857048511505127, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.5610055923461914, | |
| "rewards/margins": 0.1274861991405487, | |
| "rewards/rejected": -0.6884917616844177, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.7962085308056874, | |
| "grad_norm": 29.875, | |
| "learning_rate": 1.2272757730841744e-07, | |
| "log_odds_chosen": 0.4991677701473236, | |
| "log_odds_ratio": -0.5991695523262024, | |
| "logits/chosen": 338.85333251953125, | |
| "logits/rejected": 345.8002624511719, | |
| "logps/chosen": -1.084472417831421, | |
| "logps/rejected": -1.4532811641693115, | |
| "loss": 20.8087, | |
| "nll_loss": 1.2487828731536865, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.5422362089157104, | |
| "rewards/margins": 0.1844043731689453, | |
| "rewards/rejected": -0.7266405820846558, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 2.843601895734597, | |
| "grad_norm": 30.5, | |
| "learning_rate": 6.91585183706428e-08, | |
| "log_odds_chosen": 0.4421129822731018, | |
| "log_odds_ratio": -0.6238647699356079, | |
| "logits/chosen": 352.36297607421875, | |
| "logits/rejected": 363.74932861328125, | |
| "logps/chosen": -1.0848851203918457, | |
| "logps/rejected": -1.4058189392089844, | |
| "loss": 20.2124, | |
| "nll_loss": 1.2451623678207397, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.5424425601959229, | |
| "rewards/margins": 0.16046686470508575, | |
| "rewards/rejected": -0.7029094696044922, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.890995260663507, | |
| "grad_norm": 29.0, | |
| "learning_rate": 3.077661432604184e-08, | |
| "log_odds_chosen": 0.5017856359481812, | |
| "log_odds_ratio": -0.5793642401695251, | |
| "logits/chosen": 331.6553649902344, | |
| "logits/rejected": 360.7071838378906, | |
| "logps/chosen": -1.0543787479400635, | |
| "logps/rejected": -1.3993524312973022, | |
| "loss": 19.7715, | |
| "nll_loss": 1.265842318534851, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.5271893739700317, | |
| "rewards/margins": 0.1724867820739746, | |
| "rewards/rejected": -0.6996762156486511, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 2.938388625592417, | |
| "grad_norm": 94.0, | |
| "learning_rate": 7.700082708883006e-09, | |
| "log_odds_chosen": 0.4347440302371979, | |
| "log_odds_ratio": -0.60276859998703, | |
| "logits/chosen": 349.830810546875, | |
| "logits/rejected": 379.3188781738281, | |
| "logps/chosen": -1.1171703338623047, | |
| "logps/rejected": -1.4191040992736816, | |
| "loss": 20.9093, | |
| "nll_loss": 1.3094590902328491, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.5585851669311523, | |
| "rewards/margins": 0.15096691250801086, | |
| "rewards/rejected": -0.7095520496368408, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.985781990521327, | |
| "grad_norm": 30.125, | |
| "learning_rate": 0.0, | |
| "log_odds_chosen": 0.4408469796180725, | |
| "log_odds_ratio": -0.6020691990852356, | |
| "logits/chosen": 353.28302001953125, | |
| "logits/rejected": 368.32501220703125, | |
| "logps/chosen": -1.073335886001587, | |
| "logps/rejected": -1.3923397064208984, | |
| "loss": 20.0639, | |
| "nll_loss": 1.2285845279693604, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -0.5366679430007935, | |
| "rewards/margins": 0.15950192511081696, | |
| "rewards/rejected": -0.6961698532104492, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 2.985781990521327, | |
| "step": 315, | |
| "total_flos": 0.0, | |
| "train_loss": 31.780521017407615, | |
| "train_runtime": 7128.5338, | |
| "train_samples_per_second": 2.841, | |
| "train_steps_per_second": 0.044 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 315, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 100000, | |
| "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": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |