End of training

README.md

@@ -1,9 +1,15 @@
 ---
 base_model: meta-llama/Llama-2-7b-hf
 tags:
+- alignment-handbook
 - trl
 - dpo
 - generated_from_trainer
+- trl
+- dpo
+- generated_from_trainer
+datasets:
+- HuggingFaceH4/ultrafeedback_binarized
 model-index:
 - name: llama2-7b-dpo-full-wo-medication_qa-ep3
   results: []
@@ -14,17 +20,17 @@ should probably proofread and complete it, then remove this comment. -->
 
 # llama2-7b-dpo-full-wo-medication_qa-ep3
 
-This model is a fine-tuned version of [meta-llama/Llama-2-7b-hf](https://huggingface.co/meta-llama/Llama-2-7b-hf) on an unknown dataset.
+This model is a fine-tuned version of [meta-llama/Llama-2-7b-hf](https://huggingface.co/meta-llama/Llama-2-7b-hf) on the HuggingFaceH4/ultrafeedback_binarized dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.6540
-- Rewards/chosen: 0.0513
-- Rewards/rejected: -0.0539
-- Rewards/accuracies: 0.7778
-- Rewards/margins: 0.1052
-- Logps/rejected: -1095.5646
-- Logps/chosen: -485.5474
-- Logits/rejected: -1.1199
-- Logits/chosen: -0.0017
+- Loss: 0.6486
+- Rewards/chosen: 0.0485
+- Rewards/rejected: -0.0716
+- Rewards/accuracies: 0.7847
+- Rewards/margins: 0.1201
+- Logps/rejected: -1097.3336
+- Logps/chosen: -485.8272
+- Logits/rejected: -1.1127
+- Logits/chosen: -0.0114
 
 ## Model description
 

all_results.json

@@ -1,5 +1,18 @@
 {
     "epoch": 1.0,
+    "eval_logits/chosen": -0.011370762251317501,
+    "eval_logits/rejected": -1.1126513481140137,
+    "eval_logps/chosen": -485.8271789550781,
+    "eval_logps/rejected": -1097.3336181640625,
+    "eval_loss": 0.6485693454742432,
+    "eval_rewards/accuracies": 0.7847222089767456,
+    "eval_rewards/chosen": 0.04850021377205849,
+    "eval_rewards/margins": 0.12006408721208572,
+    "eval_rewards/rejected": -0.07156387716531754,
+    "eval_runtime": 191.3892,
+    "eval_samples": 1728,
+    "eval_samples_per_second": 9.029,
+    "eval_steps_per_second": 0.282,
     "train_loss": 0.6277184326883772,
     "train_runtime": 2301.3874,
     "train_samples": 9071,

config.json

@@ -23,6 +23,6 @@
   "tie_word_embeddings": false,
   "torch_dtype": "bfloat16",
   "transformers_version": "4.39.0.dev0",
-  "use_cache": false,
+  "use_cache": true,
   "vocab_size": 32000
 }

eval_results.json

@@ -0,0 +1,16 @@
+{
+    "epoch": 1.0,
+    "eval_logits/chosen": -0.011370762251317501,
+    "eval_logits/rejected": -1.1126513481140137,
+    "eval_logps/chosen": -485.8271789550781,
+    "eval_logps/rejected": -1097.3336181640625,
+    "eval_loss": 0.6485693454742432,
+    "eval_rewards/accuracies": 0.7847222089767456,
+    "eval_rewards/chosen": 0.04850021377205849,
+    "eval_rewards/margins": 0.12006408721208572,
+    "eval_rewards/rejected": -0.07156387716531754,
+    "eval_runtime": 191.3892,
+    "eval_samples": 1728,
+    "eval_samples_per_second": 9.029,
+    "eval_steps_per_second": 0.282
+}