Model save

README.md

@@ -2,13 +2,9 @@
 license: mit
 library_name: peft
 tags:
-- alignment-handbook
-- generated_from_trainer
 - trl
 - dpo
 - generated_from_trainer
-datasets:
-- HuggingFaceH4/ultrafeedback_binarized
 base_model: microsoft/phi-2
 model-index:
 - name: phi-2-gpo-ultrafeedback-lora
@@ -20,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # phi-2-gpo-ultrafeedback-lora
 
-This model is a fine-tuned version of [lole25/phi-2-sft-ultrachat-lora](https://huggingface.co/lole25/phi-2-sft-ultrachat-lora) 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.
 It achieves the following results on the evaluation set:
 - Loss: 0.0004
 - Rewards/chosen: -0.0084

adapter_config.json

@@ -20,9 +20,9 @@
   "revision": null,
   "target_modules": [
     "q_proj",
-    "dense",
+    "v_proj",
     "k_proj",
-    "v_proj"
+    "dense"
   ],
   "task_type": "CAUSAL_LM"
 }

all_results.json

@@ -9,13 +9,13 @@
     "eval_rewards/chosen": -0.008376287296414375,
     "eval_rewards/margins": 0.009309147484600544,
     "eval_rewards/rejected": -0.017685433849692345,
-    "eval_runtime": 324.63,
+    "eval_runtime": 323.1582,
     "eval_samples": 2000,
-    "eval_samples_per_second": 6.161,
-    "eval_steps_per_second": 0.385,
+    "eval_samples_per_second": 6.189,
+    "eval_steps_per_second": 0.387,
     "train_loss": 2.2877212975025803e-05,
-    "train_runtime": 862.5384,
+    "train_runtime": 855.8096,
     "train_samples": 30567,
-    "train_samples_per_second": 70.877,
-    "train_steps_per_second": 1.106
+    "train_samples_per_second": 71.434,
+    "train_steps_per_second": 1.115
 }

eval_results.json

@@ -9,8 +9,8 @@
     "eval_rewards/chosen": -0.008376287296414375,
     "eval_rewards/margins": 0.009309147484600544,
     "eval_rewards/rejected": -0.017685433849692345,
-    "eval_runtime": 324.63,
+    "eval_runtime": 323.1582,
     "eval_samples": 2000,
-    "eval_samples_per_second": 6.161,
-    "eval_steps_per_second": 0.385
+    "eval_samples_per_second": 6.189,
+    "eval_steps_per_second": 0.387
 }

runs/Mar07_15-08-43_gpu4-119-4/events.out.tfevents.1709784607.gpu4-119-4.1178546.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:967414148e00d6529964d711be469779c24daac955c98f3b07374fdaf5593d95
+size 8366

runs/Mar07_15-08-43_gpu4-119-4/events.out.tfevents.1709785786.gpu4-119-4.1178546.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5dc823ac15042bf88ec88508f727f3eeac7c4f7f786de01bb96441f34a84208a
+size 828

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 2.0,
     "train_loss": 2.2877212975025803e-05,
-    "train_runtime": 862.5384,
+    "train_runtime": 855.8096,
     "train_samples": 30567,
-    "train_samples_per_second": 70.877,
-    "train_steps_per_second": 1.106
+    "train_samples_per_second": 71.434,
+    "train_steps_per_second": 1.115
 }

trainer_state.json

@@ -1501,9 +1501,9 @@
       "step": 954,
       "total_flos": 0.0,
       "train_loss": 2.2877212975025803e-05,
-      "train_runtime": 862.5384,
-      "train_samples_per_second": 70.877,
-      "train_steps_per_second": 1.106
+      "train_runtime": 855.8096,
+      "train_samples_per_second": 71.434,
+      "train_steps_per_second": 1.115
     }
   ],
   "logging_steps": 10,

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a0dd2b02ab585a34a9bf39314ee5dc3e74f98b4692efe90bf2da48b933a6a62a
+oid sha256:5bffb371a56113cbe802831c08a188d09b0ca6cbb3c7dac180dc167c8b87b2b5
 size 5816