Created all required files

app.py

@@ -1,64 +1,65 @@
-import gradio as gr
-from huggingface_hub import InferenceClient
-
-"""
-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")
+# app.py
 
+import gradio as gr
+import torch
+from transformers import AutoModelForCausalLM, AutoTokenizer
 
-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]})
+# Load the tokenizer and model
+# Using a specific revision to ensure compatibility
+tokenizer = AutoTokenizer.from_pretrained("microsoft/DialoGPT-medium")
+model = AutoModelForCausalLM.from_pretrained("microsoft/DialoGPT-medium")
 
-    messages.append({"role": "user", "content": message})
+# Define the prediction function
+def predict(message, history):
+    # 'history' is a list of lists, where each inner list has a user and a bot message.
+    # We need to format it for DialoGPT.
+    history_transformer_format = []
+    for user, bot in history:
+        history_transformer_format.append(user)
+        history_transformer_format.append(bot)
 
-    response = ""
+    # Join the history and the new message, separated by the EOS token
+    history_string = "".join(history_transformer_format)
+    input_text = history_string + message + tokenizer.eos_token
 
-    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
+    # Tokenize the input
+    new_user_input_ids = tokenizer.encode(input_text, return_tensors='pt')
 
-        response += token
-        yield response
+    # Generate a response
+    # The max_length is set to 1250 to allow for a decent conversation history.
+    bot_output_ids = model.generate(
+        new_user_input_ids,
+        max_length=1250,
+        pad_token_id=tokenizer.eos_token_id,
+        no_repeat_ngram_size=3,
+        do_sample=True,
+        top_k=100,
+        top_p=0.7,
+        temperature=0.8
+    )
 
+    # Decode the response, skipping the input part
+    response = tokenizer.decode(bot_output_ids[:, new_user_input_ids.shape[-1]:][0], skip_special_tokens=True)
+    
+    # Return an empty string to clear the textbox and the updated history
+    return "", history + [[message, response]]
 
-"""
-For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
-"""
-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)",
-        ),
-    ],
-)
+# Build the Gradio interface
+with gr.Blocks() as demo:
+    gr.Markdown("## DialoGPT-medium Chatbot")
+    gr.Markdown("This chatbot uses the microsoft/DialoGPT-medium model. Start typing to chat!")
+    
+    chatbot = gr.Chatbot()
+    textbox = gr.Textbox(placeholder="Type your message here and press Enter")
 
+    # When the user submits the textbox, call the 'predict' function
+    textbox.submit(
+        predict,
+        inputs=[textbox, chatbot],
+        outputs=[textbox, chatbot]
+    )
 
-if __name__ == "__main__":
-    demo.launch()
+# Enable the queue for better handling of multiple users and to enable API usage
+demo.queue()
+# Launch the app
+demo.launch()

requirements.txt

@@ -1 +1,3 @@
-huggingface_hub==0.25.2
+torch
+transformers
+gradio