Update app.py

app.py

@@ -1,70 +1,152 @@
 import gradio as gr
-from huggingface_hub import InferenceClient
+from langchain_community.vectorstores import Chroma
+from langchain_core.documents import Document
+from langchain_openai import ChatOpenAI
+from langchain_community.embeddings import HuggingFaceEmbeddings
+from langchain_core.prompts import ChatPromptTemplate
+from langchain_core.output_parsers import StrOutputParser
+from langchain_core.runnables import RunnablePassthrough
+import json
+import os
 
+# -------------------------------
+# CONFIGURATION
+# -------------------------------
+# For Hugging Face Spaces, set this as a Secret in your Space settings
+os.environ["OPENAI_API_KEY"] = os.getenv("OPENAI_API_KEY", "sk-your-key-here")
 
-def respond(
-    message,
-    history: list[dict[str, str]],
-    system_message,
-    max_tokens,
-    temperature,
-    top_p,
-    hf_token: gr.OAuthToken,
-):
-    """
-    For more information on `huggingface_hub` Inference API support, please check the docs: https://huggingface.co/docs/huggingface_hub/v0.22.2/en/guides/inference
-    """
-    client = InferenceClient(token=hf_token.token, model="openai/gpt-oss-20b")
+# -------------------------------
+# LOAD DATA AND BUILD RAG CHAIN
+# -------------------------------
+def clean_metadata(metadata):
+    """Convert list values to comma-separated strings for ChromaDB compatibility"""
+    cleaned = {}
+    for key, value in metadata.items():
+        if isinstance(value, list):
+            cleaned[key] = ", ".join(str(v) for v in value)
+        elif isinstance(value, (str, int, float, bool)) or value is None:
+            cleaned[key] = value
+        else:
+            cleaned[key] = str(value)
+    return cleaned
 
-    messages = [{"role": "system", "content": system_message}]
+print("Loading documents...")
+docs = []
+with open("helpwildlife_rag.jsonl", "r", encoding="utf-8") as f:
+    for line in f:
+        entry = json.loads(line)
+        metadata = entry.get("metadata", {})
+        docs.append(Document(
+            page_content=entry["text"],
+            metadata=clean_metadata(metadata)
+        ))
+print(f"✓ Loaded {len(docs)} documents")
 
-    messages.extend(history)
+print("Loading embedding model...")
+embeddings = HuggingFaceEmbeddings(
+    model_name="sentence-transformers/all-MiniLM-L6-v2"
+)
 
-    messages.append({"role": "user", "content": message})
+print("Building vector store...")
+try:
+    vectorstore = Chroma.from_documents(docs, embedding=embeddings)
+except TypeError:
+    vectorstore = Chroma.from_documents(docs, embedding_function=embeddings)
 
-    response = ""
+retriever = vectorstore.as_retriever(search_kwargs={"k": 3})
+print("✓ Vector store created")
 
-    for message in client.chat_completion(
-        messages,
-        max_tokens=max_tokens,
-        stream=True,
-        temperature=temperature,
-        top_p=top_p,
-    ):
-        choices = message.choices
-        token = ""
-        if len(choices) and choices[0].delta.content:
-            token = choices[0].delta.content
+prompt_template = """
+You are a compassionate wildlife advice assistant.
+Your only source of information is the text provided in the CONTEXT section below.
+Do not use outside knowledge, guesses, or general reasoning.
+If the context does not contain enough information to answer fully,
+say "I'm not sure based on the available guidance."
 
-        response += token
-        yield response
+Never suggest killing or harming any animal.
+Always emphasise that taking an injured or distressed animal to a local wildlife rescue
+or 24/7 animal hospital is the safest and most humane course of action.
+Discourage people from trying to handle or treat the animal themselves,
+and note that general veterinary clinics may euthanise wild animals unnecessarily.
 
+Adopt a voice of empathy and respect for all life, consistent with vegan principles:
+there are no animals that are pests, vermin, or unworthy of care.
+Respond clearly and calmly, with brief, practical, step-by-step guidance suitable for the public.
 
+---------------------
+CONTEXT (from HelpWildlife data file):
+{context}
+---------------------
+QUESTION:
+{question}
+
+YOUR ANSWER:
 """
-For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
-"""
-chatbot = gr.ChatInterface(
-    respond,
-    type="messages",
-    additional_inputs=[
-        gr.Textbox(value="You are a friendly Chatbot.", 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 (nucleus sampling)",
-        ),
-    ],
+
+prompt = ChatPromptTemplate.from_template(prompt_template)
+llm = ChatOpenAI(model_name="gpt-4o-mini", temperature=0)
+
+def format_docs(docs):
+    return "\n\n".join(doc.page_content for doc in docs)
+
+rag_chain = (
+    {
+        "context": retriever | format_docs,
+        "question": RunnablePassthrough()
+    }
+    | prompt
+    | llm
+    | StrOutputParser()
 )
+print("✓ RAG chain ready")
 
-with gr.Blocks() as demo:
-    with gr.Sidebar():
-        gr.LoginButton()
-    chatbot.render()
+# -------------------------------
+# GRADIO INTERFACE
+# -------------------------------
+def ask_wildlife_question(question):
+    """Process a wildlife question and return an answer"""
+    if not question.strip():
+        return "Please enter a question about wildlife."
+    
+    try:
+        answer = rag_chain.invoke(question)
+        return answer
+    except Exception as e:
+        return f"Error: {str(e)}\n\nPlease check your OpenAI API key is set correctly."
 
+# Example questions
+examples = [
+    "I found a baby hedgehog out during the day. What should I do?",
+    "There's a bird that seems injured in my garden. How can I help?",
+    "I found a baby bird on the ground. Should I put it back in the nest?",
+    "A fox is limping in my backyard. What should I do?",
+    "How do I know if a wild animal needs help?"
+]
+
+# Create Gradio interface
+demo = gr.Interface(
+    fn=ask_wildlife_question,
+    inputs=gr.Textbox(
+        label="Ask a Wildlife Question",
+        placeholder="e.g., I found a baby bird on the ground...",
+        lines=3
+    ),
+    outputs=gr.Textbox(
+        label="Compassionate Advice",
+        lines=10
+    ),
+    title="🦔 Wildlife Rescue Assistant",
+    description="""
+    Ask questions about helping wildlife in distress. This assistant provides compassionate, 
+    evidence-based advice prioritizing the wellbeing of all animals.
+    
+    ⚠️ **Important**: This tool provides general guidance only. For urgent situations, 
+    contact your local wildlife rescue or 24/7 animal hospital immediately.
+    """,
+    examples=examples,
+    theme=gr.themes.Soft(),
+    allow_flagging="never"
+)
 
 if __name__ == "__main__":
-    demo.launch()
+    demo.launch()