Spaces:
Paused
Paused
File size: 8,896 Bytes
78822f8 | 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 | import os
import subprocess
import logging
import json
import requests
import uvicorn
from fastapi import FastAPI, Depends, HTTPException, Request
from fastapi.middleware.cors import CORSMiddleware
from fastapi.security import HTTPBearer, HTTPAuthorizationCredentials
from fastapi.responses import StreamingResponse
from huggingface_hub import HfApi
logging.basicConfig(level=logging.INFO)
logger = logging.getLogger(__name__)
app = FastAPI(title="o87Dev Cloud LLM API")
security = HTTPBearer()
app.add_middleware(
CORSMiddleware,
allow_origins=["*"],
allow_methods=["*"],
allow_headers=["*"],
)
OLLAMA_BASE = "http://localhost:11434"
MODEL = os.environ.get("DEFAULT_MODEL", "qwen2.5-coder:7b-instruct-q4_K_M")
API_TOKEN = os.environ.get("API_TOKEN") # Set as Space secret
# ── Auth ──────────────────────────────────────────────────────────────────────
def verify_token(creds: HTTPAuthorizationCredentials = Depends(security)):
token = creds.credentials
# If API_TOKEN secret is set, validate against it directly (faster)
if API_TOKEN:
if token != API_TOKEN:
raise HTTPException(401, "Invalid token")
return token
# Fallback: validate as HF token
try:
HfApi().whoami(token=token)
except Exception:
raise HTTPException(401, "Invalid Hugging Face token")
return token
# ── Health ────────────────────────────────────────────────────────────────────
@app.get("/health")
async def health():
try:
r = requests.get(f"{OLLAMA_BASE}/api/tags", timeout=5)
models = [m["name"] for m in r.json().get("models", [])]
return {"status": "ok", "model": MODEL, "available_models": models}
except Exception as e:
return {"status": "starting", "error": str(e)}
# ── OpenAI-compatible /v1/chat/completions ────────────────────────────────────
@app.post("/v1/chat/completions")
async def chat_completions(request: Request, token: str = Depends(verify_token)):
body = await request.json()
model = body.get("model", MODEL)
stream = body.get("stream", False)
ollama_payload = {
"model": model,
"messages": body.get("messages", []),
"stream": stream,
"options": {
"num_ctx": body.get("max_tokens", 32768),
"temperature": body.get("temperature", 0.7),
}
}
if stream:
def generate():
try:
with requests.post(
f"{OLLAMA_BASE}/v1/chat/completions",
json=ollama_payload,
stream=True,
timeout=300
) as r:
for chunk in r.iter_content(chunk_size=None):
if chunk:
yield chunk
except Exception as e:
yield f"data: {{\"error\": \"{str(e)}\"}}\n\n"
return StreamingResponse(generate(), media_type="text/event-stream")
else:
try:
r = requests.post(
f"{OLLAMA_BASE}/v1/chat/completions",
json=ollama_payload,
timeout=300
)
return r.json()
except Exception as e:
raise HTTPException(500, str(e))
# ── Anthropic-compatible /v1/messages ─────────────────────────────────────────
@app.post("/v1/messages")
async def messages(request: Request, token: str = Depends(verify_token)):
body = await request.json()
model = body.get("model", MODEL)
stream = body.get("stream", False)
ollama_payload = {
"model": model,
"messages": body.get("messages", []),
"stream": stream,
"options": {
"num_ctx": body.get("max_tokens", 32768),
"temperature": body.get("temperature", 0.7),
}
}
if stream:
import time
def generate_anthropic():
msg_id = f"msg_{int(time.time())}"
yield f"event: message_start\ndata: {json.dumps({'type':'message_start','message':{'id':msg_id,'type':'message','role':'assistant','content':[],'model':model,'stop_reason':None,'usage':{'input_tokens':0,'output_tokens':0}}})}\n\n"
yield f"event: content_block_start\ndata: {json.dumps({'type':'content_block_start','index':0,'content_block':{'type':'text','text':''}})}\n\n"
yield f"event: ping\ndata: {{\"type\":\"ping\"}}\n\n"
output_tokens = 0
try:
with requests.post(
f"{OLLAMA_BASE}/v1/chat/completions",
json=ollama_payload,
stream=True,
timeout=300
) as r:
buffer = ""
for chunk in r.iter_content(chunk_size=None):
if not chunk:
continue
buffer += chunk.decode("utf-8", errors="ignore")
lines = buffer.split("\n")
buffer = lines.pop()
for line in lines:
line = line.strip()
if not line or not line.startswith("data: "):
continue
js = line[6:]
if js == "[DONE]":
break
try:
data = json.loads(js)
if data.get("usage"):
output_tokens = data["usage"].get("completion_tokens", 0)
delta = data.get("choices", [{}])[0].get("delta", {})
text = delta.get("content") or delta.get("reasoning") or ""
if text:
yield f"event: content_block_delta\ndata: {json.dumps({'type':'content_block_delta','index':0,'delta':{'type':'text_delta','text':text}})}\n\n"
if data.get("choices", [{}])[0].get("finish_reason"):
break
except Exception:
pass
except Exception as e:
yield f"event: content_block_delta\ndata: {json.dumps({'type':'content_block_delta','index':0,'delta':{'type':'text_delta','text':f'Error: {e}'}})}\n\n"
yield f"event: content_block_stop\ndata: {{\"type\":\"content_block_stop\",\"index\":0}}\n\n"
yield f"event: message_delta\ndata: {json.dumps({'type':'message_delta','delta':{'stop_reason':'end_turn','stop_sequence':None},'usage':{'output_tokens':output_tokens}})}\n\n"
yield f"event: message_stop\ndata: {{\"type\":\"message_stop\"}}\n\n"
return StreamingResponse(generate_anthropic(), media_type="text/event-stream")
else:
try:
r = requests.post(
f"{OLLAMA_BASE}/v1/chat/completions",
json=ollama_payload,
timeout=300
)
data = r.json()
content = data.get("choices", [{}])[0].get("message", {}).get("content", "")
return {
"id": data.get("id", f"msg_{int(__import__('time').time())}"),
"type": "message",
"role": "assistant",
"content": [{"type": "text", "text": content}],
"model": model,
"stop_reason": "end_turn",
"usage": {
"input_tokens": data.get("usage", {}).get("prompt_tokens", 0),
"output_tokens": data.get("usage", {}).get("completion_tokens", 0)
}
}
except Exception as e:
raise HTTPException(500, str(e))
# ── Models list ───────────────────────────────────────────────────────────────
@app.get("/v1/models")
async def list_models(token: str = Depends(verify_token)):
try:
r = requests.get(f"{OLLAMA_BASE}/api/tags", timeout=5)
models = [{"id": m["name"], "object": "model"} for m in r.json().get("models", [])]
return {"object": "list", "data": models}
except Exception:
return {"object": "list", "data": [{"id": MODEL, "object": "model"}]}
if __name__ == "__main__":
uvicorn.run(app, host="0.0.0.0", port=7860)
|