Add Mistral-7B-OpenOrca conversation_temmplate (#2585)

docs/model_support.md

@@ -30,6 +30,7 @@
 - [NousResearch/Nous-Hermes-13b](https://huggingface.co/NousResearch/Nous-Hermes-13b)
 - [openaccess-ai-collective/manticore-13b-chat-pyg](https://huggingface.co/openaccess-ai-collective/manticore-13b-chat-pyg)
 - [OpenAssistant/oasst-sft-4-pythia-12b-epoch-3.5](https://huggingface.co/OpenAssistant/oasst-sft-4-pythia-12b-epoch-3.5)
+- [Open-Orca/Mistral-7B-OpenOrca](https://huggingface.co/Open-Orca/Mistral-7B-OpenOrca)
 - [VMware/open-llama-7b-v2-open-instruct](https://huggingface.co/VMware/open-llama-7b-v2-open-instruct)
 - [Phind/Phind-CodeLlama-34B-v2](https://huggingface.co/Phind/Phind-CodeLlama-34B-v2)
 - [project-baize/baize-v2-7b](https://huggingface.co/project-baize/baize-v2-7b)

fastchat/conversation.py

@@ -912,6 +912,20 @@ def get_conv_template(name: str) -> Conversation:
     )
 )
 
+# Open-Orca/Mistral-7B-OpenOrca template
+# source: https://huggingface.co/Open-Orca/Mistral-7B-OpenOrca
+# reference: https://huggingface.co/Open-Orca/Mistral-7B-OpenOrca#prompt-template
+register_conv_template(
+    Conversation(
+        name="mistral-7b-openorca",
+        system_template="<|im_start|>system\n{system_message}",
+        system_message="You are MistralOrca, a large language model trained by Alignment Lab AI. Write out your reasoning step-by-step to be sure you get the right answers!",
+        roles=("<|im_start|>user", "<|im_start|>assistant"),
+        sep_style=SeparatorStyle.CHATML,
+        sep="<|im_end|>",
+        stop_token_ids=[32000, 32001],
+    )
+)
 
 # Qwen-chat default template
 # source: https://huggingface.co/Qwen/Qwen-7B-Chat/blob/main/qwen_generation_utils.py#L130

fastchat/model/model_adapter.py

@@ -1345,12 +1345,24 @@ def get_default_conv_template(self, model_path: str) -> Conversation:
 
 
 class OpenOrcaAdapter(BaseModelAdapter):
-    "Model adapater for Open-Orca models (e.g., Open-Orca/OpenOrcaxOpenChat-Preview2-13B)" ""
+    """Model adapater for Open-Orca models wich may use different prompt templates
+    - (e.g. Open-Orca/OpenOrcaxOpenChat-Preview2-13B, Open-Orca/Mistral-7B-OpenOrca)
+    - `OpenOrcaxOpenChat-Preview2-13B` uses their "OpenChat Llama2 V1" prompt template.
+        - [Open-Orca/OpenOrcaxOpenChat-Preview2-13B #Prompt Template](https://huggingface.co/Open-Orca/OpenOrcaxOpenChat-Preview2-13B#prompt-template)
+    - `Mistral-7B-OpenOrca` uses the [OpenAI's Chat Markup Language (ChatML)](https://github.com/openai/openai-python/blob/main/chatml.md)
+        format, with <|im_start|> and <|im_end|> tokens added to support this.
+        - [Open-Orca/Mistral-7B-OpenOrca #Prompt Template](https://huggingface.co/Open-Orca/Mistral-7B-OpenOrca#prompt-template)
+    """
 
     use_fast_tokenizer = False
 
     def match(self, model_path: str):
-        return "openorca" in model_path.lower()
+        if "mistral-7b-openorca" in model_path.lower():
+            return get_conv_template("mistral-7b-openorca")
+        elif "openorca" in model_path.lower():
+            return get_conv_template("open-orca")
+        else:
+            return False
 
     def load_model(self, model_path: str, from_pretrained_kwargs: dict):
         revision = from_pretrained_kwargs.get("revision", "main")

fastchat/model/model_registry.py

@@ -325,6 +325,7 @@ def get_model_info(name: str) -> ModelInfo:
     "https://huggingface.co/HuggingFaceH4/zephyr-7b-alpha",
     "a chatbot fine-tuned from Mistral by Hugging Face",
 )
+
 register_model_info(
     [
         "Xwin-LM-7B-V0.1",
@@ -337,3 +338,10 @@ def get_model_info(name: str) -> ModelInfo:
     "https://github.com/Xwin-LM/Xwin-LM",
     "Chat models developed by Xwin-LM team",
 )
+
+register_model_info(
+    ["Mistral-7B-OpenOrca"],
+    "Open-Orca",
+    "https://huggingface.co/Open-Orca/Mistral-7B-OpenOrca",
+    "A fine-tune of [Mistral 7B](https://huggingface.co/mistralai/Mistral-7B-v0.1) using [OpenOrca dataset](https://huggingface.co/datasets/Open-Orca/OpenOrca)",
+)