Training in progress, step 94

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:500571fef7582bb6f4a8ca1f7eb8776bec97ef5b92daacd59938af376b95569a
+oid sha256:7cee89f39da97dc062cb36b8f188e9b91c658143ca1a266a82d88c66b2d5eae2
 size 13648432

last-checkpoint/adapter_config.json

@@ -20,8 +20,8 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "v_proj",
-    "q_proj"
+    "q_proj",
+    "v_proj"
   ],
   "task_type": "CAUSAL_LM",
   "use_dora": false,

last-checkpoint/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:500571fef7582bb6f4a8ca1f7eb8776bec97ef5b92daacd59938af376b95569a
+oid sha256:47a08a6e864e2ee1968177f93a78f8afab9ebca87fe279fae45b39b1d4790815
 size 13648432

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b7397a13afbf5516f50f5db50c75e701ecf1160b0474e757b80d8df82a91c915
+oid sha256:7711dccfe7177717c6c0d73a485a5f5d766f81d9dbdd88ae79b64c25ed23c69e
 size 7309882

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6142ee3f8a3ad64f6111a972476e4c1792062e5bb4c04391dd280f3db9c34702
+oid sha256:46e4c05515a68694cd891bd31cda7371a27ef3d19d366250cc2850e7de57780a
 size 1064

last-checkpoint/trainer_state.json

@@ -1,9 +1,9 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.9477707006369427,
+  "epoch": 0.6318471337579618,
   "eval_steps": 500,
-  "global_step": 93,
+  "global_step": 62,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -1123,564 +1123,6 @@
       "rewards/margins": 0.011071886867284775,
       "rewards/rejected": -0.07814602553844452,
       "step": 62
-    },
-    {
-      "epoch": 0.6420382165605095,
-      "grad_norm": 0.637529730796814,
-      "learning_rate": 7.857142857142858e-05,
-      "log_odds_chosen": 0.33163022994995117,
-      "log_odds_ratio": -0.5596598982810974,
-      "logits/chosen": -0.26017192006111145,
-      "logits/rejected": -0.27084898948669434,
-      "logps/chosen": -0.6162149906158447,
-      "logps/rejected": -0.7858166694641113,
-      "loss": 0.7619,
-      "nll_loss": 0.7059016227722168,
-      "rewards/accuracies": 0.71875,
-      "rewards/chosen": -0.06162149831652641,
-      "rewards/margins": 0.016960175707936287,
-      "rewards/rejected": -0.07858168333768845,
-      "step": 63
-    },
-    {
-      "epoch": 0.6522292993630573,
-      "grad_norm": 0.4790436029434204,
-      "learning_rate": 7.823129251700681e-05,
-      "log_odds_chosen": 0.35003671050071716,
-      "log_odds_ratio": -0.5487669110298157,
-      "logits/chosen": -0.1195572018623352,
-      "logits/rejected": -0.18973661959171295,
-      "logps/chosen": -0.5801494121551514,
-      "logps/rejected": -0.7505877017974854,
-      "loss": 0.7456,
-      "nll_loss": 0.6906896829605103,
-      "rewards/accuracies": 0.84375,
-      "rewards/chosen": -0.05801495164632797,
-      "rewards/margins": 0.0170438215136528,
-      "rewards/rejected": -0.07505877315998077,
-      "step": 64
-    },
-    {
-      "epoch": 0.6624203821656051,
-      "grad_norm": 0.47054964303970337,
-      "learning_rate": 7.789115646258504e-05,
-      "log_odds_chosen": 0.2697327733039856,
-      "log_odds_ratio": -0.588991641998291,
-      "logits/chosen": -0.12229520082473755,
-      "logits/rejected": -0.12661772966384888,
-      "logps/chosen": -0.6129696369171143,
-      "logps/rejected": -0.7457621693611145,
-      "loss": 0.7732,
-      "nll_loss": 0.7143112421035767,
-      "rewards/accuracies": 0.65625,
-      "rewards/chosen": -0.061296962201595306,
-      "rewards/margins": 0.01327925268560648,
-      "rewards/rejected": -0.07457621395587921,
-      "step": 65
-    },
-    {
-      "epoch": 0.6726114649681528,
-      "grad_norm": 0.5436939597129822,
-      "learning_rate": 7.755102040816327e-05,
-      "log_odds_chosen": 0.3452521860599518,
-      "log_odds_ratio": -0.5557309985160828,
-      "logits/chosen": -0.08714023232460022,
-      "logits/rejected": -0.1457855999469757,
-      "logps/chosen": -0.5788965225219727,
-      "logps/rejected": -0.7455580234527588,
-      "loss": 0.7397,
-      "nll_loss": 0.6841052174568176,
-      "rewards/accuracies": 0.78125,
-      "rewards/chosen": -0.057889651507139206,
-      "rewards/margins": 0.01666616089642048,
-      "rewards/rejected": -0.07455581426620483,
-      "step": 66
-    },
-    {
-      "epoch": 0.6828025477707006,
-      "grad_norm": 0.5666665434837341,
-      "learning_rate": 7.72108843537415e-05,
-      "log_odds_chosen": 0.3019857406616211,
-      "log_odds_ratio": -0.5755807757377625,
-      "logits/chosen": -0.07147038727998734,
-      "logits/rejected": -0.09695108234882355,
-      "logps/chosen": -0.5981124639511108,
-      "logps/rejected": -0.7489673495292664,
-      "loss": 0.754,
-      "nll_loss": 0.6964120864868164,
-      "rewards/accuracies": 0.78125,
-      "rewards/chosen": -0.059811241924762726,
-      "rewards/margins": 0.015085484832525253,
-      "rewards/rejected": -0.07489673048257828,
-      "step": 67
-    },
-    {
-      "epoch": 0.6929936305732484,
-      "grad_norm": 0.6614330410957336,
-      "learning_rate": 7.687074829931973e-05,
-      "log_odds_chosen": 0.24785318970680237,
-      "log_odds_ratio": -0.5993584990501404,
-      "logits/chosen": -0.1450255811214447,
-      "logits/rejected": -0.1640571653842926,
-      "logps/chosen": -0.6308612823486328,
-      "logps/rejected": -0.7485529184341431,
-      "loss": 0.79,
-      "nll_loss": 0.7300571799278259,
-      "rewards/accuracies": 0.75,
-      "rewards/chosen": -0.06308613717556,
-      "rewards/margins": 0.011769158765673637,
-      "rewards/rejected": -0.07485529780387878,
-      "step": 68
-    },
-    {
-      "epoch": 0.7031847133757961,
-      "grad_norm": 0.5287584662437439,
-      "learning_rate": 7.653061224489796e-05,
-      "log_odds_chosen": 0.2588488459587097,
-      "log_odds_ratio": -0.5906943082809448,
-      "logits/chosen": -0.09049258381128311,
-      "logits/rejected": -0.09309229999780655,
-      "logps/chosen": -0.6527013778686523,
-      "logps/rejected": -0.7823808789253235,
-      "loss": 0.7924,
-      "nll_loss": 0.7333185076713562,
-      "rewards/accuracies": 0.78125,
-      "rewards/chosen": -0.06527013331651688,
-      "rewards/margins": 0.012967949733138084,
-      "rewards/rejected": -0.07823808491230011,
-      "step": 69
-    },
-    {
-      "epoch": 0.7133757961783439,
-      "grad_norm": 0.6998805999755859,
-      "learning_rate": 7.619047619047618e-05,
-      "log_odds_chosen": 0.21674904227256775,
-      "log_odds_ratio": -0.6123349666595459,
-      "logits/chosen": -0.11004475504159927,
-      "logits/rejected": -0.11400632560253143,
-      "logps/chosen": -0.6173443794250488,
-      "logps/rejected": -0.7168747782707214,
-      "loss": 0.768,
-      "nll_loss": 0.7068023085594177,
-      "rewards/accuracies": 0.65625,
-      "rewards/chosen": -0.06173444539308548,
-      "rewards/margins": 0.009953041560947895,
-      "rewards/rejected": -0.0716874822974205,
-      "step": 70
-    },
-    {
-      "epoch": 0.7235668789808917,
-      "grad_norm": 0.742544412612915,
-      "learning_rate": 7.585034013605443e-05,
-      "log_odds_chosen": 0.19548919796943665,
-      "log_odds_ratio": -0.6275305151939392,
-      "logits/chosen": -0.12460804730653763,
-      "logits/rejected": -0.12356013804674149,
-      "logps/chosen": -0.7157447934150696,
-      "logps/rejected": -0.799144983291626,
-      "loss": 0.8176,
-      "nll_loss": 0.7548049688339233,
-      "rewards/accuracies": 0.6875,
-      "rewards/chosen": -0.07157447934150696,
-      "rewards/margins": 0.008340023458003998,
-      "rewards/rejected": -0.07991450279951096,
-      "step": 71
-    },
-    {
-      "epoch": 0.7337579617834394,
-      "grad_norm": 0.5506495237350464,
-      "learning_rate": 7.551020408163266e-05,
-      "log_odds_chosen": 0.2938334345817566,
-      "log_odds_ratio": -0.5719183087348938,
-      "logits/chosen": -0.09276354312896729,
-      "logits/rejected": -0.09323181211948395,
-      "logps/chosen": -0.6002932786941528,
-      "logps/rejected": -0.7459129691123962,
-      "loss": 0.7346,
-      "nll_loss": 0.6773681640625,
-      "rewards/accuracies": 0.75,
-      "rewards/chosen": -0.060029324144124985,
-      "rewards/margins": 0.014561973512172699,
-      "rewards/rejected": -0.07459130138158798,
-      "step": 72
-    },
-    {
-      "epoch": 0.7439490445859872,
-      "grad_norm": 0.5797027349472046,
-      "learning_rate": 7.517006802721089e-05,
-      "log_odds_chosen": 0.1955791562795639,
-      "log_odds_ratio": -0.6199466586112976,
-      "logits/chosen": -0.07904835790395737,
-      "logits/rejected": -0.12583205103874207,
-      "logps/chosen": -0.6262507438659668,
-      "logps/rejected": -0.7190106511116028,
-      "loss": 0.7983,
-      "nll_loss": 0.7363026142120361,
-      "rewards/accuracies": 0.65625,
-      "rewards/chosen": -0.06262507289648056,
-      "rewards/margins": 0.009275996126234531,
-      "rewards/rejected": -0.07190106809139252,
-      "step": 73
-    },
-    {
-      "epoch": 0.7541401273885351,
-      "grad_norm": 0.6255090832710266,
-      "learning_rate": 7.482993197278913e-05,
-      "log_odds_chosen": 0.24435332417488098,
-      "log_odds_ratio": -0.5920006632804871,
-      "logits/chosen": -0.08072115480899811,
-      "logits/rejected": -0.0795532613992691,
-      "logps/chosen": -0.6259615421295166,
-      "logps/rejected": -0.7442739009857178,
-      "loss": 0.7805,
-      "nll_loss": 0.7213104963302612,
-      "rewards/accuracies": 0.8125,
-      "rewards/chosen": -0.0625961497426033,
-      "rewards/margins": 0.011831245385110378,
-      "rewards/rejected": -0.07442739605903625,
-      "step": 74
-    },
-    {
-      "epoch": 0.7643312101910829,
-      "grad_norm": 0.500413715839386,
-      "learning_rate": 7.448979591836736e-05,
-      "log_odds_chosen": 0.1923215687274933,
-      "log_odds_ratio": -0.6174032688140869,
-      "logits/chosen": -0.09215761721134186,
-      "logits/rejected": -0.08426235616207123,
-      "logps/chosen": -0.6419164538383484,
-      "logps/rejected": -0.7315772771835327,
-      "loss": 0.7981,
-      "nll_loss": 0.7363995313644409,
-      "rewards/accuracies": 0.71875,
-      "rewards/chosen": -0.06419164687395096,
-      "rewards/margins": 0.008966086432337761,
-      "rewards/rejected": -0.07315773516893387,
-      "step": 75
-    },
-    {
-      "epoch": 0.7745222929936306,
-      "grad_norm": 0.47041788697242737,
-      "learning_rate": 7.414965986394559e-05,
-      "log_odds_chosen": 0.32406049966812134,
-      "log_odds_ratio": -0.558575451374054,
-      "logits/chosen": -0.0999608039855957,
-      "logits/rejected": -0.10914846509695053,
-      "logps/chosen": -0.6198108196258545,
-      "logps/rejected": -0.7787230014801025,
-      "loss": 0.7178,
-      "nll_loss": 0.6619524359703064,
-      "rewards/accuracies": 0.84375,
-      "rewards/chosen": -0.06198108196258545,
-      "rewards/margins": 0.015891216695308685,
-      "rewards/rejected": -0.07787229120731354,
-      "step": 76
-    },
-    {
-      "epoch": 0.7847133757961784,
-      "grad_norm": 0.5418070554733276,
-      "learning_rate": 7.380952380952382e-05,
-      "log_odds_chosen": 0.1873037964105606,
-      "log_odds_ratio": -0.6218591928482056,
-      "logits/chosen": -0.09966853260993958,
-      "logits/rejected": -0.0808413028717041,
-      "logps/chosen": -0.6605861186981201,
-      "logps/rejected": -0.7546226978302002,
-      "loss": 0.7967,
-      "nll_loss": 0.7345606088638306,
-      "rewards/accuracies": 0.6875,
-      "rewards/chosen": -0.06605862081050873,
-      "rewards/margins": 0.009403654374182224,
-      "rewards/rejected": -0.07546227425336838,
-      "step": 77
-    },
-    {
-      "epoch": 0.7949044585987262,
-      "grad_norm": 0.6250921487808228,
-      "learning_rate": 7.346938775510205e-05,
-      "log_odds_chosen": 0.31786370277404785,
-      "log_odds_ratio": -0.5675632357597351,
-      "logits/chosen": -0.15341174602508545,
-      "logits/rejected": -0.13008305430412292,
-      "logps/chosen": -0.5935506224632263,
-      "logps/rejected": -0.7325422763824463,
-      "loss": 0.7511,
-      "nll_loss": 0.6942938566207886,
-      "rewards/accuracies": 0.75,
-      "rewards/chosen": -0.05935506522655487,
-      "rewards/margins": 0.013899162411689758,
-      "rewards/rejected": -0.07325422763824463,
-      "step": 78
-    },
-    {
-      "epoch": 0.8050955414012739,
-      "grad_norm": 0.40636491775512695,
-      "learning_rate": 7.312925170068027e-05,
-      "log_odds_chosen": 0.30228132009506226,
-      "log_odds_ratio": -0.5691430568695068,
-      "logits/chosen": -0.08937245607376099,
-      "logits/rejected": -0.1251370757818222,
-      "logps/chosen": -0.6316071152687073,
-      "logps/rejected": -0.7746922373771667,
-      "loss": 0.7947,
-      "nll_loss": 0.7377853989601135,
-      "rewards/accuracies": 0.8125,
-      "rewards/chosen": -0.06316071003675461,
-      "rewards/margins": 0.014308515936136246,
-      "rewards/rejected": -0.07746922224760056,
-      "step": 79
-    },
-    {
-      "epoch": 0.8152866242038217,
-      "grad_norm": 0.5820095539093018,
-      "learning_rate": 7.27891156462585e-05,
-      "log_odds_chosen": 0.3547573387622833,
-      "log_odds_ratio": -0.5511735081672668,
-      "logits/chosen": -0.11485984176397324,
-      "logits/rejected": -0.11412106454372406,
-      "logps/chosen": -0.586860716342926,
-      "logps/rejected": -0.7574341297149658,
-      "loss": 0.7343,
-      "nll_loss": 0.679178774356842,
-      "rewards/accuracies": 0.71875,
-      "rewards/chosen": -0.05868607014417648,
-      "rewards/margins": 0.017057344317436218,
-      "rewards/rejected": -0.0757434219121933,
-      "step": 80
-    },
-    {
-      "epoch": 0.8254777070063695,
-      "grad_norm": 0.6103206276893616,
-      "learning_rate": 7.244897959183675e-05,
-      "log_odds_chosen": 0.3428616225719452,
-      "log_odds_ratio": -0.547142505645752,
-      "logits/chosen": -0.17640018463134766,
-      "logits/rejected": -0.1795547753572464,
-      "logps/chosen": -0.5735194683074951,
-      "logps/rejected": -0.7396149039268494,
-      "loss": 0.7013,
-      "nll_loss": 0.6465373635292053,
-      "rewards/accuracies": 0.875,
-      "rewards/chosen": -0.05735195428133011,
-      "rewards/margins": 0.016609538346529007,
-      "rewards/rejected": -0.07396149635314941,
-      "step": 81
-    },
-    {
-      "epoch": 0.8356687898089172,
-      "grad_norm": 0.5069766640663147,
-      "learning_rate": 7.210884353741498e-05,
-      "log_odds_chosen": 0.12126234918832779,
-      "log_odds_ratio": -0.6486403346061707,
-      "logits/chosen": -0.14951294660568237,
-      "logits/rejected": -0.15072469413280487,
-      "logps/chosen": -0.5951024293899536,
-      "logps/rejected": -0.6498846411705017,
-      "loss": 0.7563,
-      "nll_loss": 0.6914424300193787,
-      "rewards/accuracies": 0.65625,
-      "rewards/chosen": -0.059510238468647,
-      "rewards/margins": 0.005478225648403168,
-      "rewards/rejected": -0.06498846411705017,
-      "step": 82
-    },
-    {
-      "epoch": 0.845859872611465,
-      "grad_norm": 0.4635069966316223,
-      "learning_rate": 7.17687074829932e-05,
-      "log_odds_chosen": 0.2604025900363922,
-      "log_odds_ratio": -0.5959426760673523,
-      "logits/chosen": -0.15265314280986786,
-      "logits/rejected": -0.12722516059875488,
-      "logps/chosen": -0.5640627145767212,
-      "logps/rejected": -0.6875459551811218,
-      "loss": 0.72,
-      "nll_loss": 0.6604282259941101,
-      "rewards/accuracies": 0.75,
-      "rewards/chosen": -0.056406278163194656,
-      "rewards/margins": 0.012348320335149765,
-      "rewards/rejected": -0.06875459849834442,
-      "step": 83
-    },
-    {
-      "epoch": 0.8560509554140128,
-      "grad_norm": 0.5292064547538757,
-      "learning_rate": 7.142857142857143e-05,
-      "log_odds_chosen": 0.23624499142169952,
-      "log_odds_ratio": -0.6043432354927063,
-      "logits/chosen": -0.11054065823554993,
-      "logits/rejected": -0.1269894540309906,
-      "logps/chosen": -0.624123215675354,
-      "logps/rejected": -0.7454429864883423,
-      "loss": 0.7638,
-      "nll_loss": 0.703406035900116,
-      "rewards/accuracies": 0.6875,
-      "rewards/chosen": -0.0624123141169548,
-      "rewards/margins": 0.012131982482969761,
-      "rewards/rejected": -0.07454430311918259,
-      "step": 84
-    },
-    {
-      "epoch": 0.8662420382165605,
-      "grad_norm": 0.45075032114982605,
-      "learning_rate": 7.108843537414966e-05,
-      "log_odds_chosen": 0.24057269096374512,
-      "log_odds_ratio": -0.6011200547218323,
-      "logits/chosen": -0.11038406193256378,
-      "logits/rejected": -0.12801669538021088,
-      "logps/chosen": -0.5973632335662842,
-      "logps/rejected": -0.7181129455566406,
-      "loss": 0.717,
-      "nll_loss": 0.6568683385848999,
-      "rewards/accuracies": 0.65625,
-      "rewards/chosen": -0.05973632633686066,
-      "rewards/margins": 0.012074977159500122,
-      "rewards/rejected": -0.07181129604578018,
-      "step": 85
-    },
-    {
-      "epoch": 0.8764331210191083,
-      "grad_norm": 0.577867865562439,
-      "learning_rate": 7.074829931972789e-05,
-      "log_odds_chosen": 0.25693386793136597,
-      "log_odds_ratio": -0.590366542339325,
-      "logits/chosen": -0.0883067399263382,
-      "logits/rejected": -0.0588250458240509,
-      "logps/chosen": -0.6325963735580444,
-      "logps/rejected": -0.7457253932952881,
-      "loss": 0.7601,
-      "nll_loss": 0.7010390758514404,
-      "rewards/accuracies": 0.75,
-      "rewards/chosen": -0.06325963884592056,
-      "rewards/margins": 0.011312908492982388,
-      "rewards/rejected": -0.07457254827022552,
-      "step": 86
-    },
-    {
-      "epoch": 0.8866242038216561,
-      "grad_norm": 0.5341058373451233,
-      "learning_rate": 7.040816326530612e-05,
-      "log_odds_chosen": 0.21396467089653015,
-      "log_odds_ratio": -0.6061536073684692,
-      "logits/chosen": -0.0878664180636406,
-      "logits/rejected": -0.10543867945671082,
-      "logps/chosen": -0.5701617002487183,
-      "logps/rejected": -0.6728689670562744,
-      "loss": 0.751,
-      "nll_loss": 0.6903491616249084,
-      "rewards/accuracies": 0.75,
-      "rewards/chosen": -0.057016175240278244,
-      "rewards/margins": 0.010270730592310429,
-      "rewards/rejected": -0.0672869086265564,
-      "step": 87
-    },
-    {
-      "epoch": 0.8968152866242038,
-      "grad_norm": 0.4483081102371216,
-      "learning_rate": 7.006802721088435e-05,
-      "log_odds_chosen": 0.34266385436058044,
-      "log_odds_ratio": -0.5521624088287354,
-      "logits/chosen": -0.07731188088655472,
-      "logits/rejected": -0.10345152020454407,
-      "logps/chosen": -0.5693691372871399,
-      "logps/rejected": -0.7366178631782532,
-      "loss": 0.7171,
-      "nll_loss": 0.6619335412979126,
-      "rewards/accuracies": 0.8125,
-      "rewards/chosen": -0.05693691596388817,
-      "rewards/margins": 0.01672486774623394,
-      "rewards/rejected": -0.07366178184747696,
-      "step": 88
-    },
-    {
-      "epoch": 0.9070063694267516,
-      "grad_norm": 0.727847158908844,
-      "learning_rate": 6.97278911564626e-05,
-      "log_odds_chosen": 0.2518327534198761,
-      "log_odds_ratio": -0.595729410648346,
-      "logits/chosen": -0.092443086206913,
-      "logits/rejected": -0.12563319504261017,
-      "logps/chosen": -0.608110249042511,
-      "logps/rejected": -0.7250441312789917,
-      "loss": 0.751,
-      "nll_loss": 0.6914761066436768,
-      "rewards/accuracies": 0.75,
-      "rewards/chosen": -0.06081102788448334,
-      "rewards/margins": 0.011693387292325497,
-      "rewards/rejected": -0.07250441610813141,
-      "step": 89
-    },
-    {
-      "epoch": 0.9171974522292994,
-      "grad_norm": 0.5988894701004028,
-      "learning_rate": 6.938775510204082e-05,
-      "log_odds_chosen": 0.2530748248100281,
-      "log_odds_ratio": -0.5891579389572144,
-      "logits/chosen": -0.08331869542598724,
-      "logits/rejected": -0.09794420003890991,
-      "logps/chosen": -0.567914605140686,
-      "logps/rejected": -0.6916981339454651,
-      "loss": 0.7225,
-      "nll_loss": 0.6635410785675049,
-      "rewards/accuracies": 0.78125,
-      "rewards/chosen": -0.05679146572947502,
-      "rewards/margins": 0.012378348037600517,
-      "rewards/rejected": -0.06916981935501099,
-      "step": 90
-    },
-    {
-      "epoch": 0.9273885350318471,
-      "grad_norm": 0.6276490092277527,
-      "learning_rate": 6.904761904761905e-05,
-      "log_odds_chosen": 0.1880638748407364,
-      "log_odds_ratio": -0.6190271973609924,
-      "logits/chosen": 0.024591676890850067,
-      "logits/rejected": 0.03547324985265732,
-      "logps/chosen": -0.6322571635246277,
-      "logps/rejected": -0.7290378212928772,
-      "loss": 0.7772,
-      "nll_loss": 0.7153192758560181,
-      "rewards/accuracies": 0.65625,
-      "rewards/chosen": -0.06322571635246277,
-      "rewards/margins": 0.009678068570792675,
-      "rewards/rejected": -0.07290378957986832,
-      "step": 91
-    },
-    {
-      "epoch": 0.9375796178343949,
-      "grad_norm": 0.43707507848739624,
-      "learning_rate": 6.870748299319728e-05,
-      "log_odds_chosen": 0.2878055274486542,
-      "log_odds_ratio": -0.5840734839439392,
-      "logits/chosen": -0.0694851279258728,
-      "logits/rejected": -0.07452072948217392,
-      "logps/chosen": -0.5883352160453796,
-      "logps/rejected": -0.7242968678474426,
-      "loss": 0.7213,
-      "nll_loss": 0.6628736853599548,
-      "rewards/accuracies": 0.78125,
-      "rewards/chosen": -0.0588335283100605,
-      "rewards/margins": 0.013596156612038612,
-      "rewards/rejected": -0.07242967933416367,
-      "step": 92
-    },
-    {
-      "epoch": 0.9477707006369427,
-      "grad_norm": 0.4694580137729645,
-      "learning_rate": 6.836734693877551e-05,
-      "log_odds_chosen": 0.2617676556110382,
-      "log_odds_ratio": -0.5865938663482666,
-      "logits/chosen": -0.031561583280563354,
-      "logits/rejected": -0.025262409821152687,
-      "logps/chosen": -0.5720298290252686,
-      "logps/rejected": -0.695450484752655,
-      "loss": 0.7592,
-      "nll_loss": 0.7005623579025269,
-      "rewards/accuracies": 0.75,
-      "rewards/chosen": -0.05720299109816551,
-      "rewards/margins": 0.012342060916125774,
-      "rewards/rejected": -0.06954504549503326,
-      "step": 93
     }
   ],
   "logging_steps": 1,

last-checkpoint/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5ace7eb9613605d1a80effde8f5e673e03dc8c398af3367a0d9d550ae1639482
+oid sha256:b6e9bd93f66e1ed0f17812a85321161533aa48011f619aebcabb9399556f35a5
 size 5560

runs/Aug18_16-36-13_e21432fd5565/events.out.tfevents.1723998982.e21432fd5565.2815.0

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:bee74032e5aa0b8e956c410adc7ba6f3d8da2599ca7ad1c5f89fceffb5bd7dcb
-size 32200
+oid sha256:14bc2e3b52c429ccd4ee023d6616fde9d944ae2eb21ad5d03011e5a4d591ea9b
+size 33045