feat(model): Support Yi-1.5 models (#1516)

README.md

@@ -158,6 +158,9 @@ At present, we have introduced several key features to showcase our current capa
   We offer extensive model support, including dozens of large language models (LLMs) from both open-source and API agents, such as LLaMA/LLaMA2, Baichuan, ChatGLM, Wenxin, Tongyi, Zhipu, and many more. 
 
   - News
+    - 🔥🔥🔥  [Yi-1.5-34B-Chat](https://huggingface.co/01-ai/Yi-1.5-34B-Chat)
+    - 🔥🔥🔥  [Yi-1.5-9B-Chat](https://huggingface.co/01-ai/Yi-1.5-9B-Chat)
+    - 🔥🔥🔥  [Yi-1.5-6B-Chat](https://huggingface.co/01-ai/Yi-1.5-6B-Chat)
     - 🔥🔥🔥  [Qwen1.5-110B-Chat](https://huggingface.co/Qwen/Qwen1.5-110B-Chat)
     - 🔥🔥🔥  [Qwen1.5-MoE-A2.7B-Chat](https://huggingface.co/Qwen/Qwen1.5-MoE-A2.7B-Chat)
     - 🔥🔥🔥  [Meta-Llama-3-70B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-70B-Instruct)

README.zh.md

@@ -152,6 +152,9 @@
   海量模型支持，包括开源、API代理等几十种大语言模型。如LLaMA/LLaMA2、Baichuan、ChatGLM、文心、通义、智谱等。当前已支持如下模型: 
 
   - 新增支持模型
+    - 🔥🔥🔥  [Yi-1.5-34B-Chat](https://huggingface.co/01-ai/Yi-1.5-34B-Chat)
+    - 🔥🔥🔥  [Yi-1.5-9B-Chat](https://huggingface.co/01-ai/Yi-1.5-9B-Chat)
+    - 🔥🔥🔥  [Yi-1.5-6B-Chat](https://huggingface.co/01-ai/Yi-1.5-6B-Chat)
     - 🔥🔥🔥  [Qwen1.5-110B-Chat](https://huggingface.co/Qwen/Qwen1.5-110B-Chat)
     - 🔥🔥🔥  [Qwen1.5-MoE-A2.7B-Chat](https://huggingface.co/Qwen/Qwen1.5-MoE-A2.7B-Chat)
     - 🔥🔥🔥  [Meta-Llama-3-70B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-70B-Instruct)

dbgpt/configs/model_config.py

@@ -174,6 +174,10 @@ LLM_MODEL_CONFIG = {
     # https://huggingface.co/01-ai/Yi-34B-Chat-4bits
     "yi-34b-chat-4bits": os.path.join(MODEL_PATH, "Yi-34B-Chat-4bits"),
     "yi-6b-chat": os.path.join(MODEL_PATH, "Yi-6B-Chat"),
+    # https://huggingface.co/01-ai/Yi-1.5-6B-Chat
+    "yi-1.5-6b-chat": os.path.join(MODEL_PATH, "Yi-1.5-6B-Chat"),
+    "yi-1.5-9b-chat": os.path.join(MODEL_PATH, "Yi-1.5-9B-Chat"),
+    "yi-1.5-34b-chat": os.path.join(MODEL_PATH, "Yi-1.5-34B-Chat"),
     # https://huggingface.co/google/gemma-7b-it
     "gemma-7b-it": os.path.join(MODEL_PATH, "gemma-7b-it"),
     # https://huggingface.co/google/gemma-2b-it

dbgpt/model/adapter/base.py

@@ -524,11 +524,11 @@ def get_model_adapter(
     """
     adapter = None
     # First find adapter by model_name
-    for adapter_entry in model_adapters:
+    for adapter_entry in model_adapters[::-1]:
         if adapter_entry.model_adapter.match(model_type, model_name, None):
             adapter = adapter_entry.model_adapter
             break
-    for adapter_entry in model_adapters:
+    for adapter_entry in model_adapters[::-1]:
         if adapter_entry.model_adapter.match(model_type, None, model_path):
             adapter = adapter_entry.model_adapter
             break

dbgpt/model/adapter/hf_adapter.py

@@ -135,6 +135,41 @@ class YiAdapter(NewHFChatModelAdapter):
         )
 
 
+class Yi15Adapter(YiAdapter):
+    """Yi 1.5 model adapter."""
+
+    def do_match(self, lower_model_name_or_path: Optional[str] = None):
+        return (
+            lower_model_name_or_path
+            and "yi-" in lower_model_name_or_path
+            and "1.5" in lower_model_name_or_path
+            and "chat" in lower_model_name_or_path
+        )
+
+    def get_str_prompt(
+        self,
+        params: Dict,
+        messages: List[ModelMessage],
+        tokenizer: Any,
+        prompt_template: str = None,
+        convert_to_compatible_format: bool = False,
+    ) -> Optional[str]:
+        str_prompt = super().get_str_prompt(
+            params,
+            messages,
+            tokenizer,
+            prompt_template,
+            convert_to_compatible_format,
+        )
+        terminators = [
+            tokenizer.eos_token_id,
+        ]
+        exist_token_ids = params.get("stop_token_ids", [])
+        terminators.extend(exist_token_ids)
+        params["stop_token_ids"] = terminators
+        return str_prompt
+
+
 class Mixtral8x7BAdapter(NewHFChatModelAdapter):
     """
     https://huggingface.co/mistralai/Mixtral-8x7B-Instruct-v0.1
@@ -335,7 +370,10 @@ class Llama3Adapter(NewHFChatModelAdapter):
         return str_prompt
 
 
+# The following code is used to register the model adapter
+# The last registered model adapter is matched first
 register_model_adapter(YiAdapter)
+register_model_adapter(Yi15Adapter)
 register_model_adapter(Mixtral8x7BAdapter)
 register_model_adapter(SOLARAdapter)
 register_model_adapter(GemmaAdapter)