Add app.py and requirements.txt for Hugging Face Space

app.py

@@ -0,0 +1,70 @@
+# app.py
+import gradio as gr
+from transformers import AutoTokenizer, AutoModelForCausalLM
+from langchain_community.vectorstores import FAISS
+from langchain_huggingface import HuggingFaceEmbeddings
+import torch
+import os
+
+# Load FAISS vectorstore
+embedding_model = HuggingFaceEmbeddings(model_name="sentence-transformers/all-MiniLM-L6-v2")
+vectorstore = FAISS.load_local("embeddings/am_index", embedding_model, allow_dangerous_deserialization=True)
+
+# Load model
+model_name = "HuggingFaceH4/zephyr-7b-beta"
+tokenizer = AutoTokenizer.from_pretrained(model_name, trust_remote_code=True)
+model = AutoModelForCausalLM.from_pretrained(model_name, trust_remote_code=True, torch_dtype=torch.float16 if torch.cuda.is_available() else torch.float32)
+
+# Set device
+device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
+model = model.to(device)
+
+# QA Function
+def generate_answer(query, api_key):
+    if api_key.strip() != "am123456":
+        return "❌ Invalid API Key."
+
+    docs = vectorstore.similarity_search(query, k=2)
+    context = "\n".join([doc.page_content for doc in docs])
+
+    prompt = f"""You are a domain expert in Additive Manufacturing.
+Based on the context, answer the following question in a short and precise paragraph.
+
+### Context:
+{context}
+
+### Question:
+{query}
+
+### Answer:
+"""
+    inputs = tokenizer(prompt, return_tensors="pt", truncation=True, max_length=1024).to(device)
+    with torch.no_grad():
+        outputs = model.generate(**inputs, max_new_tokens=300, temperature=0.5, top_p=0.95, do_sample=False)
+
+    decoded = tokenizer.decode(outputs[0], skip_special_tokens=True)
+    final_answer = decoded.split("### Answer:")[-1].strip()
+    return final_answer
+
+# Gradio UI
+with gr.Blocks(css="""
+    #main-title { width: 50%; }
+    .image-class { float: right; width: 50px; margin-left: auto; }
+""") as demo:
+    with gr.Row():
+        gr.Markdown("## 🧠 Additive Manufacturing LLM", elem_id="main-title")
+        gr.Image("logo.png", elem_id="logo", show_label=False, show_download_button=False, height=50, container=False)
+
+    gr.Markdown("Answer technical questions with a focused and clean response. No extra metadata.")
+
+    query = gr.Textbox(label="Ask your Additive Manufacturing question")
+    key = gr.Textbox(label="Enter API Key (e.g., am123456)")
+    output = gr.Textbox(label="Answer")
+
+    btn = gr.Button("Get Answer")
+    btn.click(fn=generate_answer, inputs=[query, key], outputs=output)
+
+    gr.Markdown("### 🎉 Hosted permanently on Hugging Face Spaces")
+
+if __name__ == "__main__":
+    demo.launch()

requirements.txt

@@ -0,0 +1,6 @@
+gradio
+transformers
+langchain-community
+torch
+faiss-cpu
+sentence-transformers