Update app.py

app.py

@@ -1,50 +1,83 @@
-import os
+# app.py  ──────────────────────────────────────────────────────────────────────
+import torch
 import gradio as gr
+from transformers import AutoTokenizer, AutoModelForCausalLM
+
+TITLE = "Talk To Me Morty"
 
-title = "Talk To Me Morty"
-description = """
-<p>
-<center>
-The bot was trained on Rick and Morty dialogues Kaggle Dataset using DialoGPT.
-<img src="https://huggingface.co/spaces/kingabzpro/Rick_and_Morty_Bot/resolve/main/img/rick.png" alt="rick" width="200"/>
-</center>
+DESCRIPTION = """
+<p style='text-align:center'>
+The bot was trained on a Rick & Morty dialogues dataset with DialoGPT.<br>
+<img src="https://huggingface.co/spaces/kingabzpro/Rick_and_Morty_Bot/resolve/main/img/rick.png"
+     alt="Rick"
+     width="200">
 </p>
 """
-article = "<p style='text-align: center'><a href='https://medium.com/geekculture/discord-bot-using-dailogpt-and-huggingface-api-c71983422701' target='_blank'>Complete Tutorial</a></p><p style='text-align: center'><a href='https://dagshub.com/kingabzpro/DailoGPT-RickBot' target='_blank'>Project is Available at DAGsHub</a></p></center><center><img src='https://visitor-badge.glitch.me/badge?page_id=kingabzpro/Rick_and_Morty_Bot' alt='visitor badge'></center></p>"
-examples = [["How are you Rick?"]]
-from transformers import AutoModelForCausalLM, AutoTokenizer
-import torch
 
+ARTICLE = """
+<p style='text-align:center'>
+<a href="https://medium.com/geekculture/discord-bot-using-dailogpt-and-huggingface-api-c71983422701"
+   target="_blank">Complete Tutorial</a> ·
+<a href="https://dagshub.com/kingabzpro/DailoGPT-RickBot"
+   target="_blank">Project on DAGsHub</a>
+</p>
+<p style='text-align:center'>
+<img src="https://visitor-badge.glitch.me/badge?page_id=kingabzpro/Rick_and_Morty_Bot"
+     alt="visitor badge">
+</p>
+"""
+
+# ─── Load model once at start ────────────────────────────────────────────────
 tokenizer = AutoTokenizer.from_pretrained("ericzhou/DialoGPT-Medium-Rick_v2")
-model = AutoModelForCausalLM.from_pretrained("ericzhou/DialoGPT-Medium-Rick_v2")
-
-def predict(input, history=[]):
-    # tokenize the new input sentence
-    new_user_input_ids = tokenizer.encode(input + tokenizer.eos_token, return_tensors='pt')
-
-    # append the new user input tokens to the chat history
-    bot_input_ids = torch.cat([torch.LongTensor(history), new_user_input_ids], dim=-1)
-
-    # generate a response 
-    history = model.generate(bot_input_ids, max_length=4000, pad_token_id=tokenizer.eos_token_id).tolist()
-
-    # convert the tokens to text, and then split the responses into lines
-    response = tokenizer.decode(history[0]).split("<|endoftext|>")
-    #print('decoded_response-->>'+str(response))
-    response = [(response[i], response[i+1]) for i in range(0, len(response)-1, 2)]  # convert to tuples of list
-    #print('response-->>'+str(response))
-    return response, history
-
-gr.Interface(fn=predict,
-             title=title,
-             description=description,
-             examples=examples,
-             inputs=["text", "state"],
-             outputs=["chatbot", "state"],
-             theme='gradio/seafoam').launch()
-
- #theme ="grass", 
- #title = title, 
- #flagging_callback=hf_writer,
- #description = description, 
- #article = article
+model     = AutoModelForCausalLM.from_pretrained("ericzhou/DialoGPT-Medium-Rick_v2")
+
+# ─── Chat handler ────────────────────────────────────────────────────────────
+def chat(user_msg: str, history_ids: list[int] | None):
+    if not user_msg:
+        return [], history_ids or []
+
+    new_ids = tokenizer.encode(user_msg + tokenizer.eos_token,
+                               return_tensors="pt")
+
+    bot_input = (
+        torch.cat([torch.LongTensor(history_ids), new_ids], dim=-1)
+        if history_ids else new_ids
+    )
+
+    history_ids = model.generate(
+        bot_input,
+        max_length=min(4096, bot_input.shape[-1] + 200),
+        pad_token_id=tokenizer.eos_token_id,
+        do_sample=True,
+        temperature=0.7,
+        top_p=0.92,
+        top_k=50,
+    ).tolist()
+
+    turns = tokenizer.decode(history_ids[0], skip_special_tokens=False) \
+                     .split("<|endoftext|>")
+    # pack into (user, bot) pairs for Chatbot component
+    pairs = [(turns[i], turns[i + 1]) for i in range(0, len(turns) - 1, 2)]
+    return pairs, history_ids
+
+# ─── Gradio UI ───────────────────────────────────────────────────────────────
+with gr.Blocks(theme=gr.themes.Seafoam()) as demo:
+    gr.Markdown(f"<h1 style='text-align:center'>{TITLE}</h1>")
+    gr.Markdown(DESCRIPTION)
+
+    chatbot = gr.Chatbot(height=450)
+    state   = gr.State([])
+
+    with gr.Row():
+        prompt = gr.Textbox(placeholder="Ask Rick anything…", scale=4, show_label=False)
+        send   = gr.Button("Send", variant="primary")
+
+    # send on click or ↵
+    send.click(chat, inputs=[prompt, state], outputs=[chatbot, state])
+    prompt.submit(chat, inputs=[prompt, state], outputs=[chatbot, state])
+
+    gr.Examples([["How are you, Rick?"], ["Tell me a joke!"]], inputs=prompt)
+    gr.Markdown(ARTICLE)
+
+if __name__ == "__main__":
+    demo.launch()