Model save

README.md

@@ -2,14 +2,10 @@
 license: mit
 library_name: peft
 tags:
-- alignment-handbook
-- generated_from_trainer
 - trl
 - dpo
 - generated_from_trainer
 base_model: microsoft/phi-2
-datasets:
-- HuggingFaceH4/ultrafeedback_binarized
 model-index:
 - name: phi-2-gpo-test-iter-2
   results: []
@@ -20,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # phi-2-gpo-test-iter-2
 
-This model is a fine-tuned version of [BraylonDash/phi-2-gpo-test-iter-2](https://huggingface.co/BraylonDash/phi-2-gpo-test-iter-2) on the HuggingFaceH4/ultrafeedback_binarized dataset.
+This model is a fine-tuned version of [microsoft/phi-2](https://huggingface.co/microsoft/phi-2) on the None dataset.
 
 ## Model description
 

adapter_config.json

@@ -19,9 +19,9 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
+    "v_proj",
     "q_proj",
     "k_proj",
-    "v_proj",
     "dense"
   ],
   "task_type": "CAUSAL_LM"

all_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 2.0,
-    "train_loss": 0.0007018463802523911,
-    "train_runtime": 3.9331,
+    "train_loss": 0.00032374734291806817,
+    "train_runtime": 3.5844,
     "train_samples": 30567,
-    "train_samples_per_second": 1.017,
-    "train_steps_per_second": 0.509
+    "train_samples_per_second": 2.232,
+    "train_steps_per_second": 0.558
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 2.0,
-    "train_loss": 0.0007018463802523911,
-    "train_runtime": 3.9331,
+    "train_loss": 0.00032374734291806817,
+    "train_runtime": 3.5844,
     "train_samples": 30567,
-    "train_samples_per_second": 1.017,
-    "train_steps_per_second": 0.509
+    "train_samples_per_second": 2.232,
+    "train_steps_per_second": 0.558
 }

trainer_state.json

@@ -11,11 +11,11 @@
     {
       "epoch": 1.0,
       "learning_rate": 5e-06,
-      "logits/chosen": -0.04497978091239929,
-      "logits/rejected": 0.10493635386228561,
-      "logps/chosen": -188.7808074951172,
-      "logps/rejected": -71.98552703857422,
-      "loss": 0.0007,
+      "logits/chosen": 0.09949145466089249,
+      "logits/rejected": -0.044053155928850174,
+      "logps/chosen": -259.9786682128906,
+      "logps/rejected": -270.3120422363281,
+      "loss": 0.0003,
       "rewards/accuracies": 0.0,
       "rewards/chosen": 0.0,
       "rewards/margins": 0.0,
@@ -26,10 +26,10 @@
       "epoch": 2.0,
       "step": 2,
       "total_flos": 0.0,
-      "train_loss": 0.0007018463802523911,
-      "train_runtime": 3.9331,
-      "train_samples_per_second": 1.017,
-      "train_steps_per_second": 0.509
+      "train_loss": 0.00032374734291806817,
+      "train_runtime": 3.5844,
+      "train_samples_per_second": 2.232,
+      "train_steps_per_second": 0.558
     }
   ],
   "logging_steps": 10,

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7e6cbd28a4b67d20357865520500264c0fb976059c91628b12c95ac60d8a86df
-size 5880
+oid sha256:4b25fdb9fd76c7e1484f459e5f5fc184f530c6ad3b0aaf217f3e4e5e0e10ad14
+size 5816