Training in progress, step 6, checkpoint

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:45b69b38f30c9123191de0b8828377ed369b0f837f3773e9ff0d96384f02ce82
-size 6979130
+oid sha256:a9e53fe451017a13010594dadfb5397d4346a7e0c68a800e93e1755de161d87a
+size 7309882

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5a97cec4ccaf7b86834a2fdb06e10135db381a62fd0dc050dc49420427220909
+oid sha256:7a34833b1148ea746b3aaefd6499b7286814fc0502accf3a99066160145f0cf2
 size 1064

last-checkpoint/trainer_state.json

@@ -1,9 +1,9 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.050955414012738856,
+  "epoch": 0.061146496815286625,
   "eval_steps": 500,
-  "global_step": 5,
+  "global_step": 6,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -97,12 +97,30 @@
       "rewards/margins": 0.006132860202342272,
       "rewards/rejected": -0.2816582918167114,
       "step": 5
+    },
+    {
+      "epoch": 0.061146496815286625,
+      "grad_norm": 1.755154013633728,
+      "learning_rate": 9.79591836734694e-05,
+      "log_odds_chosen": 0.09281320869922638,
+      "log_odds_ratio": -0.6549628973007202,
+      "logits/chosen": -0.5153884291648865,
+      "logits/rejected": -0.5407605767250061,
+      "logps/chosen": -2.8005495071411133,
+      "logps/rejected": -2.8862051963806152,
+      "loss": 3.2371,
+      "nll_loss": 3.171649217605591,
+      "rewards/accuracies": 0.625,
+      "rewards/chosen": -0.2800549566745758,
+      "rewards/margins": 0.008565555326640606,
+      "rewards/rejected": -0.2886205315589905,
+      "step": 6
     }
   ],
   "logging_steps": 1,
-  "max_steps": 5,
+  "max_steps": 294,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 1,
+  "num_train_epochs": 3,
   "save_steps": 1,
   "stateful_callbacks": {
     "TrainerControl": {
@@ -111,7 +129,7 @@
         "should_evaluate": false,
         "should_log": false,
         "should_save": true,
-        "should_training_stop": true
+        "should_training_stop": false
       },
       "attributes": {}
     }

last-checkpoint/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:17dece46f2086166fcf5d15834c0696840efd4c8cdeff75c12fb5279c16e0afe
+oid sha256:8ee18ea0c5cd8ab4dc26bcfd2b5da8b145de03666713371dc25009f17b330bef
 size 5560