add llm endpoint
Browse files
main.py
CHANGED
|
@@ -1,13 +1,15 @@
|
|
| 1 |
-
from fastapi import FastAPI, HTTPException, Query, Path
|
| 2 |
from fastapi.middleware.cors import CORSMiddleware
|
| 3 |
from pydantic import BaseModel, Field
|
| 4 |
-
from typing import List
|
| 5 |
import json
|
| 6 |
import os
|
| 7 |
import logging
|
| 8 |
from txtai.embeddings import Embeddings
|
| 9 |
import pandas as pd
|
| 10 |
import glob
|
|
|
|
|
|
|
| 11 |
|
| 12 |
# Set up logging
|
| 13 |
logging.basicConfig(level=logging.INFO)
|
|
@@ -19,6 +21,8 @@ app = FastAPI(
|
|
| 19 |
version="1.0.0"
|
| 20 |
)
|
| 21 |
|
|
|
|
|
|
|
| 22 |
# Enable CORS
|
| 23 |
app.add_middleware(
|
| 24 |
CORSMiddleware,
|
|
@@ -138,6 +142,92 @@ def check_and_index_csv_files():
|
|
| 138 |
else:
|
| 139 |
logger.info(f"Index already exists for: {csv_file}")
|
| 140 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 141 |
@app.on_event("startup")
|
| 142 |
async def startup_event():
|
| 143 |
check_and_index_csv_files()
|
|
|
|
| 1 |
+
from fastapi import FastAPI, HTTPException, Query, Path, Header, Depends
|
| 2 |
from fastapi.middleware.cors import CORSMiddleware
|
| 3 |
from pydantic import BaseModel, Field
|
| 4 |
+
from typing import List, Optional, Dict
|
| 5 |
import json
|
| 6 |
import os
|
| 7 |
import logging
|
| 8 |
from txtai.embeddings import Embeddings
|
| 9 |
import pandas as pd
|
| 10 |
import glob
|
| 11 |
+
import uuid
|
| 12 |
+
import httpx
|
| 13 |
|
| 14 |
# Set up logging
|
| 15 |
logging.basicConfig(level=logging.INFO)
|
|
|
|
| 21 |
version="1.0.0"
|
| 22 |
)
|
| 23 |
|
| 24 |
+
CHAT_AUTH_KEY = os.environ.get("CHAT_AUTH_KEY", "default_secret_key")
|
| 25 |
+
|
| 26 |
# Enable CORS
|
| 27 |
app.add_middleware(
|
| 28 |
CORSMiddleware,
|
|
|
|
| 142 |
else:
|
| 143 |
logger.info(f"Index already exists for: {csv_file}")
|
| 144 |
|
| 145 |
+
|
| 146 |
+
# ... [Previous code for DocumentRequest, QueryRequest, save_embeddings, load_embeddings, create_index, query_index, process_csv_file, check_and_index_csv_files remains the same]
|
| 147 |
+
|
| 148 |
+
class ChatRequest(BaseModel):
|
| 149 |
+
query: str = Field(..., description="The user's query")
|
| 150 |
+
index_id: str = Field(..., description="Unique identifier for the index to query")
|
| 151 |
+
conversation_id: Optional[str] = Field(None, description="Unique identifier for the conversation")
|
| 152 |
+
model_id: str = Field(..., description="Identifier for the LLM model to use")
|
| 153 |
+
user_id: str = Field(..., description="Unique identifier for the user")
|
| 154 |
+
|
| 155 |
+
async def get_api_key(x_api_key: str = Header(...)) -> str:
|
| 156 |
+
if x_api_key != CHAT_AUTH_KEY:
|
| 157 |
+
raise HTTPException(status_code=403, detail="Invalid API key")
|
| 158 |
+
return x_api_key
|
| 159 |
+
|
| 160 |
+
async def make_llm_request(api_key: str, llm_request: Dict[str, str]) -> Dict:
|
| 161 |
+
"""
|
| 162 |
+
Make a request to the LLM service.
|
| 163 |
+
"""
|
| 164 |
+
try:
|
| 165 |
+
async with httpx.AsyncClient() as client:
|
| 166 |
+
response = await client.post(
|
| 167 |
+
"https://pvanand-audio-chat.hf.space/llm-agent",
|
| 168 |
+
headers={
|
| 169 |
+
"accept": "application/json",
|
| 170 |
+
"X-API-Key": api_key,
|
| 171 |
+
"Content-Type": "application/json"
|
| 172 |
+
},
|
| 173 |
+
json=llm_request
|
| 174 |
+
)
|
| 175 |
+
|
| 176 |
+
if response.status_code != 200:
|
| 177 |
+
raise HTTPException(status_code=response.status_code, detail="Error from LLM service")
|
| 178 |
+
|
| 179 |
+
return response.json()
|
| 180 |
+
except httpx.HTTPError as e:
|
| 181 |
+
logger.error(f"HTTP error occurred while making LLM request: {str(e)}")
|
| 182 |
+
raise HTTPException(status_code=500, detail=f"HTTP error occurred while making LLM request: {str(e)}")
|
| 183 |
+
except Exception as e:
|
| 184 |
+
logger.error(f"Unexpected error occurred while making LLM request: {str(e)}")
|
| 185 |
+
raise HTTPException(status_code=500, detail=f"Unexpected error occurred while making LLM request: {str(e)}")
|
| 186 |
+
|
| 187 |
+
|
| 188 |
+
@app.post("/rag-chat/", response_model=dict, tags=["Chat"])
|
| 189 |
+
async def chat(request: ChatRequest, api_key: str = Depends(get_api_key)):
|
| 190 |
+
"""
|
| 191 |
+
Chat endpoint that uses embeddings search and LLM for response generation.
|
| 192 |
+
"""
|
| 193 |
+
try:
|
| 194 |
+
# Load embeddings for the specified index
|
| 195 |
+
document_list = load_embeddings(request.index_id)
|
| 196 |
+
|
| 197 |
+
# Perform embeddings search
|
| 198 |
+
search_results = embeddings.search(request.query, 5) # Get top 5 relevant results
|
| 199 |
+
context = "\n".join([document_list[idx[0]] for idx in search_results])
|
| 200 |
+
|
| 201 |
+
# Create RAG prompt
|
| 202 |
+
rag_prompt = f"please answer the user's question:\n\nUser's question:{request.query} Based on the following context, \n\nContext:\n{context} \n\nAnswer:"
|
| 203 |
+
|
| 204 |
+
rag_system_prompt = "You are a helpful assistant tasked with providing answers from the given context"
|
| 205 |
+
# Generate conversation_id if not provided
|
| 206 |
+
conversation_id = request.conversation_id or str(uuid.uuid4())
|
| 207 |
+
|
| 208 |
+
# Prepare the request for the LLM service
|
| 209 |
+
llm_request = {
|
| 210 |
+
"prompt": rag_prompt,
|
| 211 |
+
"system_message": rag_system_prompt,
|
| 212 |
+
"model_id": request.model_id,
|
| 213 |
+
"conversation_id": conversation_id,
|
| 214 |
+
"user_id": request.user_id
|
| 215 |
+
}
|
| 216 |
+
|
| 217 |
+
# Make request to LLM service
|
| 218 |
+
llm_response = await make_llm_request(api_key, llm_request)
|
| 219 |
+
|
| 220 |
+
logger.info(f"Chat response generated successfully for user: {request.user_id}")
|
| 221 |
+
return {
|
| 222 |
+
"response": llm_response,
|
| 223 |
+
"conversation_id": conversation_id
|
| 224 |
+
}
|
| 225 |
+
|
| 226 |
+
except Exception as e:
|
| 227 |
+
logger.error(f"Error in chat endpoint: {str(e)}")
|
| 228 |
+
raise HTTPException(status_code=500, detail=f"Error in chat endpoint: {str(e)}")
|
| 229 |
+
|
| 230 |
+
|
| 231 |
@app.on_event("startup")
|
| 232 |
async def startup_event():
|
| 233 |
check_and_index_csv_files()
|