Upload README.md with huggingface_hub

README.md

@@ -1,21 +1,9 @@
 ---
-dataset_info:
-  features:
-  - name: prompt
-    dtype: string
-  - name: response
-    dtype: string
-  - name: score
-    dtype: float64
-  splits:
-  - name: train
-    num_bytes: 9946
-    num_examples: 10
-  download_size: 15323
-  dataset_size: 9946
-configs:
-- config_name: default
-  data_files:
-  - split: train
-    path: data/train-*
+{}
 ---
+
+# Phi-Chat: Best answers from microsoft/phi-2
+
+This is a dataset produced by microsoft/phi-2 under the supervision of OpenAssistant/reward-model-deberta-v3-large-v2.
+
+8 responses are produced for each prompt and the best are saved for future fine-tuning or improvement.