Spaces:
Running
Running
File size: 8,843 Bytes
4986fe4 d38c2eb ef215d3 9fa0f10 378f2c3 8834a20 fc764d5 a111cf9 1d32d66 a111cf9 31eab42 378f2c3 d38c2eb 4986fe4 378f2c3 b955cc1 4986fe4 b955cc1 a111cf9 4986fe4 a111cf9 b955cc1 378f2c3 b955cc1 006f05b 9422734 aeb51aa 9422734 a111cf9 9422734 a111cf9 9422734 a111cf9 9422734 a111cf9 9422734 4d88866 fa65dba a111cf9 4e5813e aeb51aa 1ba1f47 4d88866 8ce4a3a a111cf9 7bee578 a01be99 9422734 a111cf9 9422734 a111cf9 9422734 a111cf9 7bee578 9422734 a111cf9 2c1c62a 7ef5d89 c613f2b 7ef5d89 a68045e 614a889 7ef5d89 ac4bad0 3109050 9bfb2c6 d7b15ab 9bfb2c6 1d32d66 a0270ea 924ff1d 9bfb2c6 b955cc1 c5cedd6 9bfb2c6 ba11b8c 31eab42 ba11b8c 9bfb2c6 3109050 1d32d66 a0270ea 6a3f7b0 7ef5d89 a111cf9 6a84e5c 8e4491b |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 |
import os
from dotenv import load_dotenv
from fastapi import FastAPI, HTTPException, Request
from fastapi.responses import StreamingResponse, HTMLResponse, JSONResponse, FileResponse
from pydantic import BaseModel
import httpx
from pathlib import Path # Import Path from pathlib
import requests
import re
import cloudscraper
import json
from typing import Optional
import datetime
load_dotenv()
app = FastAPI()
# Get API keys and secret endpoint from environment variables
api_keys_str = os.getenv('API_KEYS')
valid_api_keys = api_keys_str.split(',') if api_keys_str else []
secret_api_endpoint = os.getenv('SECRET_API_ENDPOINT')
secret_api_endpoint_2 = os.getenv('SECRET_API_ENDPOINT_2')
secret_api_endpoint_3 = os.getenv('SECRET_API_ENDPOINT_3') # New endpoint for searchgpt
# Validate if the main secret API endpoints are set
if not secret_api_endpoint or not secret_api_endpoint_2 or not secret_api_endpoint_3:
raise HTTPException(status_code=500, detail="API endpoint(s) are not configured in environment variables.")
# Define models that should use the secondary endpoint
alternate_models = {"gpt-4o-mini", "claude-3-haiku", "llama-3.1-70b", "mixtral-8x7b"}
class Payload(BaseModel):
model: str
messages: list
stream: bool
@app.get("/favicon.ico")
async def favicon():
# The favicon.ico file is in the same directory as the app
favicon_path = Path(__file__).parent / "favicon.ico"
return FileResponse(favicon_path, media_type="image/x-icon")
def generate_search(query: str, systemprompt: Optional[str] = None, stream: bool = True) -> str:
headers = {"User-Agent": ""}
# Use the provided system prompt, or default to "Be Helpful and Friendly"
system_message = systemprompt or "Be Helpful and Friendly"
# Create the prompt history with the user query and system message
prompt = [
{"role": "user", "content": query},
]
prompt.insert(0, {"content": system_message, "role": "system"})
# Prepare the payload for the API request
payload = {
"is_vscode_extension": True,
"message_history": prompt,
"requested_model": "searchgpt",
"user_input": prompt[-1]["content"],
}
# Send the request to the chat endpoint
response = requests.post(secret_api_endpoint_3, headers=headers, json=payload, stream=True)
streaming_text = ""
# Process the streaming response
for value in response.iter_lines(decode_unicode=True):
if value.startswith("data: "):
try:
json_modified_value = json.loads(value[6:])
content = json_modified_value.get("choices", [{}])[0].get("delta", {}).get("content", "")
if content.strip(): # Only process non-empty content
cleaned_response = {
"created": json_modified_value.get("created"),
"id": json_modified_value.get("id"),
"model": "searchgpt",
"object": "chat.completion",
"choices": [
{
"message": {
"content": content
}
}
]
}
if stream:
yield f"data: {json.dumps(cleaned_response)}\n\n"
streaming_text += content
except json.JSONDecodeError:
continue
if not stream:
yield streaming_text
@app.get("/searchgpt")
async def search_gpt(q: str, stream: Optional[bool] = False, systemprompt: Optional[str] = None):
if not q:
raise HTTPException(status_code=400, detail="Query parameter 'q' is required")
if stream:
return StreamingResponse(
generate_search(q, systemprompt=systemprompt, stream=True),
media_type="text/event-stream"
)
else:
# For non-streaming, collect the text and return as JSON response
response_text = "".join([chunk for chunk in generate_search(q, systemprompt=systemprompt, stream=False)])
return JSONResponse(content={"response": response_text})
@app.get("/", response_class=HTMLResponse)
async def root():
# Open and read the content of index.html (in the same folder as the app)
file_path = "index.html"
try:
with open(file_path, "r") as file:
html_content = file.read()
return HTMLResponse(content=html_content)
except FileNotFoundError:
return HTMLResponse(content="<h1>File not found</h1>", status_code=404)
async def get_models():
try:
# Load the models from models.json in the same folder
file_path = Path(__file__).parent / 'models.json'
with open(file_path, 'r') as f:
return json.load(f)
except FileNotFoundError:
raise HTTPException(status_code=404, detail="models.json not found")
except json.JSONDecodeError:
raise HTTPException(status_code=500, detail="Error decoding models.json")
@app.get("/models")
async def fetch_models():
return await get_models()
available_model_ids = [
"gpt-4o", "gpt-4o-mini", "claude-3-haiku", "llama-3.1-405b", "llama-3.1-70b",
"llama-3.1-8b", "gemini-1.5-flash", "mixtral-8x7b" , "command-r","gemini-pro",
"gpt-3.5-turbo", "command"
]
@app.post("/chat/completions")
@app.post("/v1/chat/completions")
async def get_completion(payload: Payload,request: Request):
model_to_use = payload.model if payload.model else "gpt-4o-mini"
# Validate model availability
if model_to_use not in available_model_ids:
raise HTTPException(
status_code=400,
detail=f"Model '{model_to_use}' is not available. Check /models for the available model list."
)
# Proceed with the request handling
payload_dict = payload.dict()
payload_dict["model"] = model_to_use
# Select the appropriate endpoint
endpoint = secret_api_endpoint_2 if model_to_use in alternate_models else secret_api_endpoint
current_time = datetime.datetime.now().strftime("%Y-%m-%d %H:%M:%S")
aaip = request.client.host
print(f"Time: {current_time}, {aaip}")
print(payload_dict)
async def stream_generator(payload_dict):
scraper = cloudscraper.create_scraper() # Create a CloudScraper session
try:
# Send POST request using CloudScraper
response = scraper.post(f"{endpoint}/v1/chat/completions", json=payload_dict, stream=True)
# Check response status
if response.status_code == 422:
raise HTTPException(status_code=422, detail="Unprocessable entity. Check your payload.")
elif response.status_code == 400:
raise HTTPException(status_code=400, detail="Bad request. Verify input data.")
elif response.status_code == 403:
raise HTTPException(status_code=403, detail="Forbidden. You do not have access to this resource.")
elif response.status_code == 404:
raise HTTPException(status_code=404, detail="The requested resource was not found.")
elif response.status_code >= 500:
raise HTTPException(status_code=500, detail="Server error. Try again later.")
# Stream response lines to the client
for line in response.iter_lines():
if line:
yield line.decode('utf-8') + "\n"
except requests.exceptions.RequestException as req_err:
# Handle request-specific errors
raise HTTPException(status_code=500, detail=f"Request failed: {req_err}")
except Exception as e:
# Handle unexpected errors
raise HTTPException(status_code=500, detail=f"An unexpected error occurred: {e}")
return StreamingResponse(stream_generator(payload_dict), media_type="application/json")
@app.get("/playground", response_class=HTMLResponse)
async def playground():
# Open and read the content of playground.html (in the same folder as the app)
file_path = "playground.html"
try:
with open(file_path, "r") as file:
html_content = file.read()
return HTMLResponse(content=html_content)
except FileNotFoundError:
return HTMLResponse(content="<h1>playground.html not found</h1>", status_code=404)
@app.on_event("startup")
async def startup_event():
print("API endpoints:")
print("GET /")
print("GET /models")
print("GET /searchgpt") # We now have the new search API
print("POST /chat/completions")
if __name__ == "__main__":
import uvicorn
uvicorn.run(app, host="0.0.0.0", port=8000)
|