Added llm api basic auth and disabled docs

.env.example

@@ -32,3 +32,5 @@ POSTGRES_PORT=5432
 INFERENCE_PROTOCOL=http
 INFERENCE_HOST=fyp-inference-dev
 INFERENCE_PORT=8001
+INFERENCE_USERNAME=admin
+INFERENCE_PASSWORD=changeme

.env.template

@@ -34,6 +34,8 @@ POSTGRES_PORT=5432
 INFERENCE_PROTOCOL=http
 INFERENCE_HOST=localhost
 INFERENCE_PORT=8001
+INFERENCE_USERNAME=admin
+INFERENCE_PASSWORD=change_this_to_a_secure_password
 
 # Production YAML (Ignore if you're setting up locally)
 FYP_DJANGO_IMAGE=dynavera-django:prod

config/settings.py

@@ -28,7 +28,9 @@ DJANGO_CELERY_BROKER_URL = os.getenv('DJANGO_CELERY_BROKER_URL', 'redis://localh
 INFERENCE_HOST = os.getenv('INFERENCE_HOST', 'localhost')
 INFERENCE_PORT = os.getenv('INFERENCE_PORT', '8001')
 INFERENCE_PROTOCOL = os.getenv('INFERENCE_PROTOCOL', 'http')
-INFERENCE_URL = f"{INFERENCE_PROTOCOL}://{INFERENCE_HOST}:{INFERENCE_PORT}"
+INFERENCE_USERNAME = os.getenv('INFERENCE_USERNAME', 'admin')
+INFERENCE_PASSWORD = os.getenv('INFERENCE_PASSWORD', 'changeme')
+INFERENCE_URL = f"{INFERENCE_PROTOCOL}://{INFERENCE_USERNAME}:{INFERENCE_PASSWORD}@{INFERENCE_HOST}:{INFERENCE_PORT}"
 INFERENCE_SEMANTIC_CHUNK_ENDPOINT = f"{INFERENCE_URL}/v1/semantic-chunk"
 INFERENCE_EMBEDDINGS_ENDPOINT = f"{INFERENCE_URL}/v1/embeddings"
 INFERENCE_CHAT_COMPLETIONS_ENDPOINT = f"{INFERENCE_URL}/v1/chat/completions"

gpu_server.py

@@ -8,8 +8,10 @@ from typing import Dict, Any
 import numpy as np
 from torch import cuda, no_grad, Tensor
 import torch.nn.functional as F
-from fastapi import FastAPI, Request, HTTPException
+import secrets
+from fastapi import FastAPI, Request, HTTPException, Depends
 from fastapi.responses import StreamingResponse
+from fastapi.security import HTTPBasic, HTTPBasicCredentials
 from llama_cpp import Llama
 from sentence_transformers import SentenceTransformer
 
@@ -65,10 +67,20 @@ async def lifespan(app: FastAPI):
     if cuda.is_available():
         cuda.empty_cache()
 
-app = FastAPI(title="Agentic GPU Node", lifespan=lifespan)
+app = FastAPI(title="Agentic GPU Node", lifespan=lifespan, docs_url=None, redoc_url=None, openapi_url=None)
+
+_security = HTTPBasic()
+_API_USER = os.getenv("INFERENCE_USERNAME", "admin")
+_API_PASS = os.getenv("INFERENCE_PASSWORD", "changeme")
+
+def require_auth(credentials: HTTPBasicCredentials = Depends(_security)):
+    valid_user = secrets.compare_digest(credentials.username.encode(), _API_USER.encode())
+    valid_pass = secrets.compare_digest(credentials.password.encode(), _API_PASS.encode())
+    if not (valid_user and valid_pass):
+        raise HTTPException(status_code=401, detail="Unauthorized", headers={"WWW-Authenticate": "Basic"})
 
 
-@app.get("/health")
+@app.get("/health", dependencies=[Depends(require_auth)])
 async def health():
     return {
         "status": "ok",
@@ -85,7 +97,7 @@ def pad_and_normalize(embeddings: Tensor, target_dimensions: int) -> Tensor:
     return F.normalize(embeddings, p=2, dim=1)
 
 
-@app.post("/v1/embeddings")
+@app.post("/v1/embeddings", dependencies=[Depends(require_auth)])
 async def embeddings(request: Request):
     data = await request.json()
     input_data = data.get("input", "")
@@ -148,7 +160,7 @@ async def embeddings(request: Request):
         },
     }
 
-@app.post("/v1/semantic-chunk")
+@app.post("/v1/semantic-chunk", dependencies=[Depends(require_auth)])
 async def semantic_chunk(request: Request):
     data = await request.json()
     raw_text = data.get("text", "")
@@ -208,7 +220,7 @@ async def semantic_chunk(request: Request):
         result = await loop.run_in_executor(None, _chunk_and_embed)
     return result
 
-@app.post("/v1/chat/completions")
+@app.post("/v1/chat/completions", dependencies=[Depends(require_auth)])
 async def chat_completions(request: Request):
     try:
         data = await request.json()