File size: 7,011 Bytes
0bc2945 735f3aa 0bc2945 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 |
from fastapi import FastAPI, HTTPException
from pydantic import BaseModel
import httpx
import requests
import os
from dotenv import load_dotenv
from typing import List, Optional, Dict, Any
import json
load_dotenv()
OPENROUTER_API_KEY = os.getenv("OPEN_ROUTER_API_KEY")
PERSONALITY_URL = os.getenv("PERSONALITY_URL")
app = FastAPI(title="ER5 v1.0.0")
class ChatMessage(BaseModel):
role: str
content: str
class ChatRequest(BaseModel):
messages: List[ChatMessage]=[]
temperature: Optional[float] = 0.7
max_tokens: Optional[int] = 1000
user_preferences: Optional[Dict[str, Any]] = None
context: Optional[str] = ""
personality: Optional[str] = None
class ChatResponse(BaseModel):
response: str
context: Optional[str] = None
def build_system_message(preferences: Dict[str, Any]) -> dict:
"""Create a system message based on schemaless user preferences"""
if not preferences:
return {"role": "system", "content": "You are a helpful assistant."}
preferences_str = "\n".join(
f"- {key}: {value if not isinstance(value, (list, dict)) else json.dumps(value)}"
for key, value in preferences.items()
)
system_content = f"""You are a helpful assistant. Please consider these user preferences in your responses:
{preferences_str}
Tailor your responses to align with these user preferences and characteristics."""
return {"role": "system", "content": system_content}
def parse_context(context_str: str) -> List[dict]:
"""Parse context string into messages list"""
if not context_str:
return []
try:
return json.loads(context_str)
except json.JSONDecodeError:
return []
async def call_openrouter(messages: List[dict], temperature: float, max_tokens: int):
url = "https://openrouter.ai/api/v1/chat/completions"
headers = {
"Authorization": f"Bearer {OPENROUTER_API_KEY}",
"Content-Type": "application/json"
}
payload = {
"model": "qwen/qwen-2.5-7b-instruct",
"messages": messages,
"temperature": temperature,
"max_tokens": max_tokens
}
async with httpx.AsyncClient() as client:
response = await client.post(url, json=payload, headers=headers)
if response.status_code != 200:
raise HTTPException(status_code=response.status_code,
detail=f"OpenRouter API error: {response.text}")
return response.json()['choices'][0]['message']['content']
# @app.post("/chat", response_model=ChatResponse)
async def chat_endpoint(request: ChatRequest):
try:
include_preferences = await should_include_preferences_llm(request.messages[-1].content)
messages = []
if include_preferences and request.user_preferences:
messages.append(build_system_message(request.user_preferences))
context_messages = parse_context(request.context)
messages.extend(context_messages)
messages.extend([{"role": msg.role, "content": msg.content}
for msg in request.messages])
messages = messages[-10:]
print(messages)
response = await decide_chat_api(
messages=messages,
temperature=request.temperature,
max_tokens=request.max_tokens,
personality=request.personality
)
assistant_message = response
context_messages = messages[1:] if include_preferences else messages
context_messages.append({"role": "assistant", "content": assistant_message})
updated_context = json.dumps(context_messages[-10:])
return ChatResponse(response=assistant_message, context=updated_context)
except Exception as e:
raise HTTPException(status_code=500, detail=str(e))
async def should_include_preferences_llm(query: str) -> bool:
"""
Use the LLM to decide if user preferences should be included based on the query.
"""
url = "https://openrouter.ai/api/v1/chat/completions"
headers = {
"Authorization": f"Bearer {OPENROUTER_API_KEY}",
"Content-Type": "application/json"
}
prompt = (
"You are an assistant that classifies user queries. "
"Given the query below, decide if user-specific context or preferences are relevant. "
"Respond with 'true' if any kind of personal context or preferences should be included in the response, "
"and 'false' otherwise.\n\n"
f"Query: {query}\n\n"
"Is this query related to personal context such as user preferences, lifestyle, habits, or any other factors "
"that could affect the response?"
)
payload = {
"model": "qwen/qwen-2.5-7b-instruct",
"messages": [{"role": "user", "content": prompt}],
"temperature": 0.0,
"max_tokens": 5
}
async with httpx.AsyncClient() as client:
response = await client.post(url, json=payload, headers=headers)
if response.status_code != 200:
raise HTTPException(status_code=response.status_code,
detail=f"OpenRouter API error: {response.text}")
llm_response = response.json()['choices'][0]['message']['content'].strip().lower()
return llm_response == "true"
async def call_personality_api(messages: List[dict], personality: str):
"""
Call the external Personality API to handle chat completions.
"""
url = PERSONALITY_URL + "/chat"
message_str = "\n".join([msg['content'] for msg in messages]) if messages else ""
payload = {
"message": message_str,
"personality": personality
}
payload = {
"message": message_str,
"personality": 'humanish'
}
response = requests.post(url, json=payload)
if response.status_code != 200:
raise HTTPException(status_code=response.status_code,
detail=f"Personality API error: {response.text}")
response_json = response.json()
if "response" in response_json:
return response_json["response"]
async def decide_chat_api(messages: List[dict], temperature: float, max_tokens: int,personality:str):
"""
This function decides which API to call based on the presence of the PERSONALITY_URL environment variable.
"""
if PERSONALITY_URL and PERSONALITY_URL.strip() and personality:
print("yes")
return await call_personality_api(messages,personality)
else:
return await call_openrouter(messages, temperature, max_tokens)
# if __name__ == "__main__":
# import uvicorn
# uvicorn.run(app, host="0.0.0.0", port=8000)
|