d

Dockerfile

@@ -6,7 +6,7 @@ RUN pip install --no-cache-dir -r requirements.txt
 
 WORKDIR /app
 COPY app.py .
-COPY dataset ./dataset
+COPY dataset /tmp
 
 # Hugging Face cache fix
 ENV TRANSFORMERS_CACHE=/app/models/.cache

app.py

@@ -11,7 +11,7 @@ import gradio as gr
 from gradio import mount_gradio_app
 
 # ------------------- Config ------------------- #
-DATA_PATH = "./dataset/pentagon_core.json"
+DATA_PATH = "/tmp/pentagon_core.json"  # Use /tmp for temporary storage
 EMBEDDING_MODEL = "./models/all-MiniLM-L6-v2"
 QA_MODEL = "./models/bart-large-cnn"
 DEVICE = "cuda" if os.environ.get("USE_CUDA") == "1" else "cpu"
@@ -60,7 +60,7 @@ def upload_knowledge(data: UploadData):
     return {"message": "Data uploaded and indexed."}
 
 # --------- Ask Endpoint --------- #
-@app.get("/ask/")
+@app.get("/ask/") 
 def ask(question: str, top_k: int = 3):
     question_embedding = embedder.encode([question], convert_to_numpy=True)
     distances, indices = index.search(question_embedding, top_k)