Spaces:
Sleeping
Sleeping
Use env vars MASTER_API_KEY, OLLAMA_KEY_1/2/3
Browse files
app.py
CHANGED
|
@@ -1,34 +1,40 @@
|
|
| 1 |
import os
|
| 2 |
-
import random
|
| 3 |
import httpx
|
| 4 |
from fastapi import FastAPI, Request, HTTPException
|
| 5 |
from fastapi.responses import JSONResponse, Response
|
| 6 |
|
| 7 |
app = FastAPI()
|
| 8 |
|
| 9 |
-
|
| 10 |
-
|
|
|
|
| 11 |
|
| 12 |
-
|
| 13 |
-
|
| 14 |
-
|
| 15 |
-
|
| 16 |
-
|
|
|
|
| 17 |
|
| 18 |
@app.get("/")
|
| 19 |
def root():
|
| 20 |
-
return {
|
|
|
|
|
|
|
|
|
|
|
|
|
| 21 |
|
| 22 |
@app.post("/v1/chat/completions")
|
| 23 |
async def chat(req: Request):
|
| 24 |
# Auth
|
| 25 |
auth_key = req.headers.get("Authorization", "").replace("Bearer ", "")
|
| 26 |
-
if auth_key !=
|
| 27 |
-
|
| 28 |
|
| 29 |
body = await req.json()
|
| 30 |
|
| 31 |
-
|
|
|
|
| 32 |
try:
|
| 33 |
async with httpx.AsyncClient(timeout=60) as client:
|
| 34 |
resp = await client.post(
|
|
@@ -36,17 +42,20 @@ async def chat(req: Request):
|
|
| 36 |
json=body,
|
| 37 |
headers={"Authorization": f"Bearer {key}"}
|
| 38 |
)
|
| 39 |
-
|
| 40 |
if resp.status_code == 200:
|
| 41 |
return Response(resp.content, status_code=200)
|
| 42 |
except Exception as e:
|
| 43 |
-
|
| 44 |
|
| 45 |
return JSONResponse({"error": "all keys failed"}, status_code=500)
|
| 46 |
|
| 47 |
@app.get("/v1/models")
|
| 48 |
def models():
|
| 49 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
| 50 |
try:
|
| 51 |
resp = httpx.get(
|
| 52 |
f"{BASE_URL}/v1/models",
|
|
|
|
| 1 |
import os
|
|
|
|
| 2 |
import httpx
|
| 3 |
from fastapi import FastAPI, Request, HTTPException
|
| 4 |
from fastapi.responses import JSONResponse, Response
|
| 5 |
|
| 6 |
app = FastAPI()
|
| 7 |
|
| 8 |
+
# Environment variables
|
| 9 |
+
BASE_URL = os.getenv("BASE_URL", "https://ollama.com")
|
| 10 |
+
MASTER_API_KEY = os.getenv("MASTER_API_KEY", "ollama-proxy-free")
|
| 11 |
|
| 12 |
+
# Load Ollama keys from env
|
| 13 |
+
OLLAMA_KEYS = []
|
| 14 |
+
for i in range(1, 10):
|
| 15 |
+
key = os.getenv(f"OLLAMA_KEY_{i}")
|
| 16 |
+
if key:
|
| 17 |
+
OLLAMA_KEYS.append(key)
|
| 18 |
|
| 19 |
@app.get("/")
|
| 20 |
def root():
|
| 21 |
+
return {
|
| 22 |
+
"status": "ok",
|
| 23 |
+
"master_key": MASTER_API_KEY[:10] + "...",
|
| 24 |
+
"ollama_keys": len(OLLAMA_KEYS)
|
| 25 |
+
}
|
| 26 |
|
| 27 |
@app.post("/v1/chat/completions")
|
| 28 |
async def chat(req: Request):
|
| 29 |
# Auth
|
| 30 |
auth_key = req.headers.get("Authorization", "").replace("Bearer ", "")
|
| 31 |
+
if auth_key != MASTER_API_KEY:
|
| 32 |
+
return JSONResponse({"error": "Unauthorized"}, status_code=401)
|
| 33 |
|
| 34 |
body = await req.json()
|
| 35 |
|
| 36 |
+
# Try each key
|
| 37 |
+
for key in OLLAMA_KEYS:
|
| 38 |
try:
|
| 39 |
async with httpx.AsyncClient(timeout=60) as client:
|
| 40 |
resp = await client.post(
|
|
|
|
| 42 |
json=body,
|
| 43 |
headers={"Authorization": f"Bearer {key}"}
|
| 44 |
)
|
|
|
|
| 45 |
if resp.status_code == 200:
|
| 46 |
return Response(resp.content, status_code=200)
|
| 47 |
except Exception as e:
|
| 48 |
+
continue
|
| 49 |
|
| 50 |
return JSONResponse({"error": "all keys failed"}, status_code=500)
|
| 51 |
|
| 52 |
@app.get("/v1/models")
|
| 53 |
def models():
|
| 54 |
+
auth_key = req.headers.get("Authorization", "").replace("Bearer ", "")
|
| 55 |
+
if auth_key != MASTER_API_KEY:
|
| 56 |
+
return JSONResponse({"error": "Unauthorized"}, status_code=401)
|
| 57 |
+
|
| 58 |
+
for key in OLLAMA_KEYS:
|
| 59 |
try:
|
| 60 |
resp = httpx.get(
|
| 61 |
f"{BASE_URL}/v1/models",
|