Update README.md

README.md

@@ -2,6 +2,8 @@
 datasets:
 - ehartford/dolphin
 - jondurbin/airoboros-2.2.1
+- ehartford/samantha-data
+- WizardLM/WizardLM_evol_instruct_V2_196k
 language:
 - en
 license: llama2
@@ -10,6 +12,8 @@ license: llama2
 Dolphin 2.2 🐬
 https://erichartford.com/dolphin
 
+New in this release:  The EOS token works now, and I have added multi-turn conversational data so it has learned to integrate the history with its response when appropriate.
+
 <img src="https://cdn-uploads.huggingface.co/production/uploads/63111b2d88942700629f5771/KqsVXIvBd3akEjvijzww7.png" width="600" />
 
 Dolphin-2.2-70b's training was sponsored by [a16z](https://a16z.com/supporting-the-open-source-ai-community/).
@@ -29,7 +33,7 @@ I modified the dataset for uncensoring, deduping, cleaning, and quality.
 
 I added Jon Durbin's excellent Airoboros dataset to increase creativity.
 
-I added a curated subset of Samantha and WizardLM data to train it for multi-turn conversation.
+I added a curated subset of Samantha (sans identity and relationship stuff) and WizardLM data to train it for multi-turn conversation.
 
 ## Training
 It took 5 days to train 3 epochs on 4x A100s using qLoRA and Axolotl