First upload
Browse files- Dockerfile +13 -0
- app.py +91 -0
- requirements.txt +4 -0
Dockerfile
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
FROM python:3.11.3
|
| 2 |
+
|
| 3 |
+
RUN useradd -m -u 1000 user
|
| 4 |
+
USER user
|
| 5 |
+
ENV PATH="/home/user/.local/bin:$PATH"
|
| 6 |
+
|
| 7 |
+
WORKDIR /app
|
| 8 |
+
|
| 9 |
+
COPY --chown=user ./requirements.txt requirements.txt
|
| 10 |
+
RUN pip install --trusted-host pypi.org --trusted-host pypi.python.org --trusted-host files.pythonhosted.org --no-cache-dir --upgrade -r requirements.txt
|
| 11 |
+
|
| 12 |
+
COPY --chown=user . /app
|
| 13 |
+
CMD ["uvicorn", "app:app", "--host", "0.0.0.0", "--port", "7860"]
|
app.py
ADDED
|
@@ -0,0 +1,91 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
import json
|
| 2 |
+
from fastapi import FastAPI, HTTPException
|
| 3 |
+
from dotenv import load_dotenv
|
| 4 |
+
import os
|
| 5 |
+
import re
|
| 6 |
+
import requests
|
| 7 |
+
from fastapi.middleware.cors import CORSMiddleware
|
| 8 |
+
from pydantic import BaseModel
|
| 9 |
+
from typing import List, Dict, Any, Optional, Tuple
|
| 10 |
+
|
| 11 |
+
load_dotenv()
|
| 12 |
+
|
| 13 |
+
app = FastAPI()
|
| 14 |
+
ranked_models = [
|
| 15 |
+
"llama-3.3-70b-versatile",
|
| 16 |
+
"llama3-70b-8192",
|
| 17 |
+
"meta-llama/llama-4-maverick-17b-128e-instruct",
|
| 18 |
+
"meta-llama/llama-4-scout-17b-16e-instruct",
|
| 19 |
+
"mistral-saba-24b",
|
| 20 |
+
"gemma2-9b-it",
|
| 21 |
+
"llama-3.1-8b-instant",
|
| 22 |
+
"llama3-8b-8192"
|
| 23 |
+
]
|
| 24 |
+
|
| 25 |
+
api_keys = []
|
| 26 |
+
|
| 27 |
+
for k,v in os.environ.items():
|
| 28 |
+
if re.match(r'^GROQ_\d+$', k):
|
| 29 |
+
api_keys.append(v)
|
| 30 |
+
|
| 31 |
+
app.add_middleware(
|
| 32 |
+
CORSMiddleware,
|
| 33 |
+
allow_credentials=True,
|
| 34 |
+
allow_headers=["*"],
|
| 35 |
+
allow_methods=["GET", "POST"],
|
| 36 |
+
allow_origins=["*"]
|
| 37 |
+
)
|
| 38 |
+
|
| 39 |
+
class ChatRequest(BaseModel):
|
| 40 |
+
models: Optional[List[Any]] = []
|
| 41 |
+
query: str
|
| 42 |
+
|
| 43 |
+
class ChatResponse(BaseModel):
|
| 44 |
+
output: str
|
| 45 |
+
|
| 46 |
+
@app.get("/")
|
| 47 |
+
def main_page():
|
| 48 |
+
return {"status": "ok"}
|
| 49 |
+
|
| 50 |
+
@app.post("/chat", response_model=ChatResponse)
|
| 51 |
+
def ask_groq_llm(req: ChatRequest):
|
| 52 |
+
models = req.models
|
| 53 |
+
query = req.query
|
| 54 |
+
looping = True
|
| 55 |
+
if models == []:
|
| 56 |
+
while looping:
|
| 57 |
+
for model in ranked_models:
|
| 58 |
+
for key in api_keys:
|
| 59 |
+
resp = requests.post("https://api.groq.com/openai/v1/chat/completions", verify=False, headers={"Content-Type": "application/json", "Authorization": f"Bearer {key}"}, data=json.dumps({"model": model, "messages": [{"role": "user", "content": query}]}))
|
| 60 |
+
if resp.status_code == 200:
|
| 61 |
+
respJson = resp.json()
|
| 62 |
+
print("Asked to", model, "with the key ID", str(api_keys.index(key)), ":", query)
|
| 63 |
+
return ChatResponse(output=respJson["choices"][0]["message"]["content"])
|
| 64 |
+
print(resp.status_code, resp.text)
|
| 65 |
+
looping = False
|
| 66 |
+
return ChatResponse(output="ERROR !")
|
| 67 |
+
elif len(models) == 1:
|
| 68 |
+
while looping:
|
| 69 |
+
for key in api_keys:
|
| 70 |
+
resp = requests.post("https://api.groq.com/openai/v1/chat/completions", verify=False, headers={"Content-Type": "application/json", "Authorization": f"Bearer {key}"}, data=json.dumps({"model": models[0], "messages": [{"role": "user", "content": query}]}))
|
| 71 |
+
if resp.status_code == 200:
|
| 72 |
+
respJson = resp.json()
|
| 73 |
+
print("Asked to", model, "with the key ID", str(api_keys.index(key)), ":", query)
|
| 74 |
+
return ChatResponse(output=respJson["choices"][0]["message"]["content"])
|
| 75 |
+
print(resp.status_code, resp.text)
|
| 76 |
+
looping = False
|
| 77 |
+
return ChatResponse(output="ERROR !")
|
| 78 |
+
else:
|
| 79 |
+
while looping:
|
| 80 |
+
order = {val: ind for ind, val in enumerate(ranked_models)}
|
| 81 |
+
sorted_models = sorted(models, key=lambda x: order.get(x, float('inf')))
|
| 82 |
+
for model in sorted_models:
|
| 83 |
+
for key in api_keys:
|
| 84 |
+
resp = requests.post("https://api.groq.com/openai/v1/chat/completions", verify=False, headers={"Content-Type": "application/json", "Authorization": f"Bearer {key}"}, data=json.dumps({"model": model, "messages": [{"role": "user", "content": query}]}))
|
| 85 |
+
if resp.status_code == 200:
|
| 86 |
+
respJson = resp.json()
|
| 87 |
+
print("Asked to", model, "with the key ID", str(api_keys.index(key)), ":", query)
|
| 88 |
+
return ChatResponse(output=respJson["choices"][0]["message"]["content"])
|
| 89 |
+
print(resp.status_code, resp.text)
|
| 90 |
+
looping = False
|
| 91 |
+
return ChatResponse(output="ERROR !")
|
requirements.txt
ADDED
|
@@ -0,0 +1,4 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
fastapi
|
| 2 |
+
uvicorn[standard]
|
| 3 |
+
python-dotenv
|
| 4 |
+
requests
|