Delete app.py

app.py

@@ -1,344 +0,0 @@
-# import os
-# import gradio as gr
-# from pathlib import Path
-# from pypdf import PdfReader
-# from sentence_transformers import SentenceTransformer
-# import faiss
-# import numpy as np
-# from groq import Groq
-
-# # ── Config ─────────────────────────────────────────────────────────────────────
-# EMBEDDING_MODEL = "all-MiniLM-L6-v2"
-# CHUNK_SIZE      = 500
-# CHUNK_OVERLAP   = 100
-# TOP_K           = 5
-# LLM_MODEL       = "llama3-8b-8192"
-
-# # ── Load embedder once at startup ──────────────────────────────────────────────
-# embedder = SentenceTransformer(EMBEDDING_MODEL)
-
-# # ── RAG helpers ────────────────────────────────────────────────────────────────
-
-# def extract_text(pdf_path: str) -> str:
-#     reader = PdfReader(pdf_path)
-#     return "".join(page.extract_text() or "" for page in reader.pages)
-
-# def chunk_text(text: str) -> list:
-#     chunks, start = [], 0
-#     while start < len(text):
-#         chunks.append(text[start : start + CHUNK_SIZE])
-#         start += CHUNK_SIZE - CHUNK_OVERLAP
-#     return chunks
-
-# def build_index(chunks: list):
-#     vecs = embedder.encode(chunks, show_progress_bar=False).astype("float32")
-#     idx  = faiss.IndexFlatL2(vecs.shape[1])
-#     idx.add(vecs)
-#     return idx
-
-# def retrieve(query: str, index, chunks: list) -> list:
-#     q_vec = embedder.encode([query], show_progress_bar=False).astype("float32")
-#     _, ids = index.search(q_vec, TOP_K)
-#     return [chunks[i] for i in ids[0] if i < len(chunks)]
-
-# # ── Process PDF ────────────────────────────────────────────────────────────────
-
-# def process_pdf(pdf_file, index_state, chunks_state):
-#     if pdf_file is None:
-#         return "⚠️ Please upload a PDF.", index_state, chunks_state
-#     try:
-#         text = extract_text(pdf_file)
-#     except Exception as e:
-#         return f"❌ Failed to read PDF: {e}", index_state, chunks_state
-#     if not text.strip():
-#         return "❌ No text found. Use a text-based PDF.", index_state, chunks_state
-#     chunks = chunk_text(text)
-#     index  = build_index(chunks)
-#     name   = Path(pdf_file).stem
-#     return f"✅ **{name}** — {len(chunks)} chunks indexed. Start chatting!", index, chunks
-
-# # ── Chat ───────────────────────────────────────────────────────────────────────
-
-# def respond(
-#     message,
-#     history: list[dict],
-#     system_message,
-#     max_tokens,
-#     temperature,
-#     top_p,
-#     groq_api_key,
-#     index_state,
-#     chunks_state,
-# ):
-#     if not groq_api_key.strip():
-#         yield "⚠️ Please enter your Groq API key in the sidebar."
-#         return
-#     if index_state is None:
-#         yield "⚠️ Please upload and process a PDF first."
-#         return
-
-#     context = retrieve(message, index_state, chunks_state)
-#     rag_system = (
-#         system_message + "\n\n"
-#         "Answer ONLY from the document context below. "
-#         "If the answer is not there, say you don't know.\n\n"
-#         "DOCUMENT CONTEXT:\n" + "\n\n---\n\n".join(context)
-#     )
-
-#     messages = [{"role": "system", "content": rag_system}]
-#     messages.extend(history)
-#     messages.append({"role": "user", "content": message})
-
-#     client   = Groq(api_key=groq_api_key.strip())
-#     response = ""
-
-#     try:
-#         stream = client.chat.completions.create(
-#             model=LLM_MODEL,
-#             messages=messages,
-#             max_tokens=max_tokens,
-#             temperature=temperature,
-#             top_p=top_p,
-#             stream=True,
-#         )
-#         for chunk in stream:
-#             token = chunk.choices[0].delta.content or ""
-#             response += token
-#             yield response
-#     except Exception as e:
-#         yield f"❌ Groq error: {e}"
-
-# # ── UI ─────────────────────────────────────────────────────────────────────────
-
-# with gr.Blocks(title="RAG PDF Chatbot") as demo:
-
-#     # State lives here — one place only, shared with everything below
-#     index_state  = gr.State(None)
-#     chunks_state = gr.State(None)
-
-#     with gr.Sidebar():
-#         gr.Markdown("## 🔑 Groq API Key")
-#         groq_key = gr.Textbox(
-#             label="API Key",
-#             placeholder="gsk_...",
-#             type="password",
-#             info="Free at console.groq.com",
-#         )
-
-#         gr.Markdown("---")
-#         gr.Markdown("## 📄 Upload PDF")
-#         pdf_upload  = gr.File(label="PDF file", file_types=[".pdf"], type="filepath")
-#         process_btn = gr.Button("⚡ Process PDF", variant="primary")
-#         status_box  = gr.Markdown("Upload a PDF to begin.")
-
-#         gr.Markdown("---")
-#         gr.Markdown(
-#             "**Model:** Llama 3 8B (Groq)\n\n"
-#             "**Embeddings:** all-MiniLM-L6-v2\n\n"
-#             "**Vector store:** FAISS top-5"
-#         )
-
-#     # PDF processing updates the shared state
-#     process_btn.click(
-#         fn=process_pdf,
-#         inputs=[pdf_upload, index_state, chunks_state],
-#         outputs=[status_box, index_state, chunks_state],
-#     )
-
-#     # ChatInterface reuses the SAME groq_key, index_state, chunks_state objects
-#     # defined above — NOT new gr.State() instances
-#     gr.ChatInterface(
-#         fn=respond,
-#         additional_inputs=[
-#             gr.Textbox(
-#                 value="You are a helpful assistant that answers questions about documents.",
-#                 label="System message",
-#             ),
-#             gr.Slider(minimum=1,   maximum=2048, value=512,  step=1,    label="Max new tokens"),
-#             gr.Slider(minimum=0.1, maximum=4.0,  value=0.7,  step=0.1,  label="Temperature"),
-#             gr.Slider(minimum=0.1, maximum=1.0,  value=0.95, step=0.05, label="Top-p"),
-#             groq_key,      # same textbox object from sidebar
-#             index_state,   # same State object — not a new one
-#             chunks_state,  # same State object — not a new one
-#         ],
-#     )
-
-# if __name__ == "__main__":
-#     demo.queue().launch()
-
-import os
-import gradio as gr
-from pathlib import Path
-from pypdf import PdfReader
-from sentence_transformers import SentenceTransformer
-import faiss
-import numpy as np
-from groq import Groq
-
-# ── Config ─────────────────────────────────────────────────────────────────────
-EMBEDDING_MODEL = "all-MiniLM-L6-v2"
-CHUNK_SIZE      = 500
-CHUNK_OVERLAP   = 100
-TOP_K           = 5
-LLM_MODEL       = "llama3-8b-8192"
-
-# ── Load embedder once at startup ──────────────────────────────────────────────
-embedder = SentenceTransformer(EMBEDDING_MODEL)
-
-# ── RAG helpers ────────────────────────────────────────────────────────────────
-
-def extract_text(pdf_path: str) -> str:
-    reader = PdfReader(pdf_path)
-    return "".join(page.extract_text() or "" for page in reader.pages)
-
-def chunk_text(text: str) -> list:
-    chunks, start = [], 0
-    while start < len(text):
-        chunks.append(text[start : start + CHUNK_SIZE])
-        start += CHUNK_SIZE - CHUNK_OVERLAP
-    return chunks
-
-def build_index(chunks: list):
-    vecs = embedder.encode(chunks, show_progress_bar=False).astype("float32")
-    idx  = faiss.IndexFlatL2(vecs.shape[1])
-    idx.add(vecs)
-    return idx
-
-def retrieve(query: str, index, chunks: list) -> list:
-    q_vec = embedder.encode([query], show_progress_bar=False).astype("float32")
-    _, ids = index.search(q_vec, TOP_K)
-    return [chunks[i] for i in ids[0] if i < len(chunks)]
-
-# ── Process PDF ────────────────────────────────────────────────────────────────
-
-def process_pdf(pdf_file, index_state, chunks_state):
-    if pdf_file is None:
-        return "⚠️ Please upload a PDF first.", index_state, chunks_state
-    try:
-        text = extract_text(pdf_file)
-    except Exception as e:
-        return f"❌ Failed to read PDF: {e}", index_state, chunks_state
-    if not text.strip():
-        return "❌ No text found. Use a text-based PDF.", index_state, chunks_state
-    chunks = chunk_text(text)
-    index  = build_index(chunks)
-    name   = Path(pdf_file).stem
-    return f"✅ **{name}** — {len(chunks)} chunks indexed. Ready to chat!", index, chunks
-
-# ── Chat ───────────────────────────────────────────────────────────────────────
-
-def respond(
-    message,
-    history,
-    groq_api_key,
-    pdf_file,
-    system_message,
-    max_tokens,
-    temperature,
-    top_p,
-    index_state,
-    chunks_state,
-):
-    # Auto-process PDF if not yet indexed
-    if index_state is None:
-        if pdf_file is None:
-            yield "⚠️ Please upload a PDF using the panel on the left."
-            return
-        try:
-            text = extract_text(pdf_file)
-            chunks = chunk_text(text)
-            index_state = build_index(chunks)
-            chunks_state = chunks
-        except Exception as e:
-            yield f"❌ Failed to process PDF: {e}"
-            return
-
-    if not groq_api_key or not groq_api_key.strip():
-        yield "⚠️ Please enter your Groq API key in the panel on the left."
-        return
-
-    context = retrieve(message, index_state, chunks_state)
-    rag_system = (
-        system_message + "\n\n"
-        "Answer ONLY from the document context below. "
-        "If the answer is not there, say you don't know.\n\n"
-        "DOCUMENT CONTEXT:\n" + "\n\n---\n\n".join(context)
-    )
-
-    messages = [{"role": "system", "content": rag_system}]
-    messages.extend(history)
-    messages.append({"role": "user", "content": message})
-
-    response = ""
-    try:
-        client = Groq(api_key=groq_api_key.strip())
-        stream = client.chat.completions.create(
-            model=LLM_MODEL,
-            messages=messages,
-            max_tokens=max_tokens,
-            temperature=temperature,
-            top_p=top_p,
-            stream=True,
-        )
-        for chunk in stream:
-            token = chunk.choices[0].delta.content or ""
-            response += token
-            yield response
-    except Exception as e:
-        yield f"❌ Groq error: {e}"
-
-# ── UI ─────────────────────────────────────────────────────────────────────────
-
-with gr.Blocks(title="RAG PDF Chatbot") as demo:
-
-    index_state  = gr.State(None)
-    chunks_state = gr.State(None)
-
-    with gr.Sidebar():
-        gr.Markdown("## 🔑 Groq API Key")
-        groq_key = gr.Textbox(
-            label="API Key",
-            placeholder="gsk_...",
-            type="password",
-            info="Free at console.groq.com",
-        )
-
-        gr.Markdown("---")
-        gr.Markdown("## 📄 PDF Document")
-        pdf_upload  = gr.File(label="Upload PDF", file_types=[".pdf"], type="filepath")
-        process_btn = gr.Button("⚡ Process PDF", variant="primary")
-        status_box  = gr.Markdown("Upload a PDF to begin.")
-
-        gr.Markdown("---")
-        gr.Markdown(
-            "**Model:** Llama 3 8B (Groq)\n\n"
-            "**Embeddings:** all-MiniLM-L6-v2\n\n"
-            "**Vector store:** FAISS top-5"
-        )
-
-    process_btn.click(
-        fn=process_pdf,
-        inputs=[pdf_upload, index_state, chunks_state],
-        outputs=[status_box, index_state, chunks_state],
-    )
-
-    gr.ChatInterface(
-        fn=respond,
-        additional_inputs=[
-            groq_key,      # ← exact same object as sidebar
-            pdf_upload,    # ← exact same object as sidebar
-            gr.Textbox(
-                value="You are a helpful assistant that answers questions about documents.",
-                label="System message",
-            ),
-            gr.Slider(minimum=1,   maximum=2048, value=512,  step=1,    label="Max new tokens"),
-            gr.Slider(minimum=0.1, maximum=4.0,  value=0.7,  step=0.1,  label="Temperature"),
-            gr.Slider(minimum=0.1, maximum=1.0,  value=0.95, step=0.05, label="Top-p"),
-            index_state,   # ← exact same State object
-            chunks_state,  # ← exact same State object
-        ],
-        additional_inputs_accordion=gr.Accordion("⚙️ Settings", open=False),
-    )
-
-if __name__ == "__main__":
-    demo.queue().launch()