add support for LocalAI (#1608)

### What problem does this PR solve?

#762 

### Type of change
- [x] New Feature (non-breaking change which adds functionality)

---------

Co-authored-by: Zhedong Cen <cenzhedong2@126.com>

rag/llm/chat_model.py

@@ -348,6 +348,82 @@ class OllamaChat(Base):
         yield 0
 
 
+class LocalAIChat(Base):
+    def __init__(self, key, model_name, base_url):
+        if base_url[-1] == "/":
+            base_url = base_url[:-1]
+        self.base_url = base_url + "/v1/chat/completions"
+        self.model_name = model_name.split("___")[0]
+
+    def chat(self, system, history, gen_conf):
+        if system:
+            history.insert(0, {"role": "system", "content": system})
+        for k in list(gen_conf.keys()):
+            if k not in ["temperature", "top_p", "max_tokens"]:
+                del gen_conf[k]
+        headers = {
+            "Content-Type": "application/json",
+        }
+        payload = json.dumps(
+            {"model": self.model_name, "messages": history, **gen_conf}
+        )
+        try:
+            response = requests.request(
+                "POST", url=self.base_url, headers=headers, data=payload
+            )
+            response = response.json()
+            ans = response["choices"][0]["message"]["content"].strip()
+            if response["choices"][0]["finish_reason"] == "length":
+                ans += (
+                    "...\nFor the content length reason, it stopped, continue?"
+                    if is_english([ans])
+                    else "······\n由于长度的原因，回答被截断了，要继续吗？"
+                )
+            return ans, response["usage"]["total_tokens"]
+        except Exception as e:
+            return "**ERROR**: " + str(e), 0
+
+    def chat_streamly(self, system, history, gen_conf):
+        if system:
+            history.insert(0, {"role": "system", "content": system})
+        ans = ""
+        total_tokens = 0
+        try:
+            headers = {
+                "Content-Type": "application/json",
+            }
+            payload = json.dumps(
+                {
+                    "model": self.model_name,
+                    "messages": history,
+                    "stream": True,
+                    **gen_conf,
+                }
+            )
+            response = requests.request(
+                "POST",
+                url=self.base_url,
+                headers=headers,
+                data=payload,
+            )
+            for resp in response.content.decode("utf-8").split("\n\n"):
+                if "choices" not in resp:
+                    continue
+                resp = json.loads(resp[6:])
+                if "delta" in resp["choices"][0]:
+                    text = resp["choices"][0]["delta"]["content"]
+                else:
+                    continue
+                ans += text
+                total_tokens += 1
+            yield ans
+
+        except Exception as e:
+            yield ans + "\n**ERROR**: " + str(e)
+
+        yield total_tokens
+
+
 class LocalLLM(Base):
     class RPCProxy:
         def __init__(self, host, port):