Update app.py

app.py

@@ -1,64 +1,77 @@
 import gradio as gr
-from huggingface_hub import InferenceClient
+from fuzzywuzzy import process
+from transformers import pipeline
 
-"""
-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("HuggingFaceH4/zephyr-7b-beta")
+# Our dictionary of 20 dental terms
+dental_terms = {
+    "cavity": "A cavity is a hole in a tooth caused by decay.",
+    "gingivitis": "Gingivitis is the inflammation of the gums, often caused by plaque buildup.",
+    "implant": "A dental implant is a surgical component that interfaces with the jawbone to support a dental prosthesis.",
+    "orthodontics": "Orthodontics is a branch of dentistry that corrects teeth and jaw alignment issues.",
+    "plaque": "Plaque is a sticky, colorless film of bacteria that forms on teeth.",
+    "enamel": "Enamel is the hard, outer surface layer of your teeth that protects against decay.",
+    "braces": "Braces are orthodontic devices used to straighten teeth and correct bite issues.",
+    "root canal": "A root canal is a treatment to repair and save a badly damaged or infected tooth.",
+    "crown": "A crown is a dental cap placed over a tooth to restore its shape, size, and strength.",
+    "veneers": "Veneers are thin shells placed over the front of teeth to improve appearance.",
+    "halitosis": "Halitosis is chronic bad breath caused by bacteria or other factors.",
+    "periodontitis": "Periodontitis is a serious gum infection that damages gums and can destroy the jawbone.",
+    "denture": "Dentures are removable appliances that replace missing teeth and surrounding tissues.",
+    "bridge": "A dental bridge is a fixed prosthetic device that replaces missing teeth.",
+    "tartar": "Tartar is hardened plaque that forms on teeth and can only be removed by a dentist.",
+    "x-ray": "A dental x-ray is an imaging technique used to view the inside of teeth and surrounding tissues.",
+    "flossing": "Flossing is the process of cleaning between your teeth with dental floss.",
+    "sealant": "A sealant is a protective coating applied to teeth to prevent decay.",
+    "bitewing": "A bitewing is a type of dental x-ray that shows the upper and lower back teeth.",
+    "occlusion": "Occlusion refers to the alignment and contact between teeth when the jaws close."
+}
 
+# Set up a Flan-T5 pipeline (instruction-tuned T5)
+# We'll use the "text2text-generation" pipeline for T5-like models
+# Model: google/flan-t5-base
+generation_pipeline = pipeline(
+    "text2text-generation",
+    model="google/flan-t5-base"
+)
 
-def respond(
-    message,
-    history: list[tuple[str, str]],
-    system_message,
-    max_tokens,
-    temperature,
-    top_p,
-):
-    messages = [{"role": "system", "content": system_message}]
-
-    for val in history:
-        if val[0]:
-            messages.append({"role": "user", "content": val[0]})
-        if val[1]:
-            messages.append({"role": "assistant", "content": val[1]})
-
-    messages.append({"role": "user", "content": message})
+def chatbot_response(message, history):
+    """
+    Hybrid response logic:
+    1) Check if user input matches a known dental term (exactly or via fuzzy matching).
+    2) If found or close match, return the definition from our dictionary.
+    3) Otherwise, use Flan-T5 to generate an open-ended response.
+    """
+    print(f"User Input: {message}")
+    print(f"Chat History: {history}")
 
-    response = ""
+    # Normalize user input to lowercase for simpler matching
+    user_input_lower = message.lower()
 
-    for message in client.chat_completion(
-        messages,
-        max_tokens=max_tokens,
-        stream=True,
-        temperature=temperature,
-        top_p=top_p,
-    ):
-        token = message.choices[0].delta.content
+    # 1) Exact match check
+    if user_input_lower in dental_terms:
+        return dental_terms[user_input_lower]
 
-        response += token
-        yield response
+    # 2) Fuzzy match check
+    closest_match, score = process.extractOne(user_input_lower, dental_terms.keys())
+    if score >= 80:
+        return f"Did you mean '{closest_match}'? {dental_terms[closest_match]}"
 
+    # 3) If no match or fuzzy match is too low, use Flan-T5 for generation
+    # We'll prompt it directly with the user's message.
+    # Adjust parameters as desired for creativity, length, etc.
+    result = generation_pipeline(message, max_length=100, num_return_sequences=1)
+    generated_text = result[0]["generated_text"]
+    return generated_text
 
-"""
-For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
-"""
+# Gradio chat interface
 demo = gr.ChatInterface(
-    respond,
-    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)",
-        ),
-    ],
+    fn=chatbot_response,
+    title="Hybrid Dental Terminology Chatbot",
+    description=(
+        "Enter a dental term to get its definition (20 known terms). "
+        "If the term isn't recognized, Flan-T5 will respond."
+    )
 )
 
-
 if __name__ == "__main__":
     demo.launch()