Update app.py

app.py

@@ -1,37 +1,71 @@
-import gradio as gr
-from transformers import AutoTokenizer, AutoModelForCausalLM
+from transformers import GPT2LMHeadModel, GPT2Tokenizer
 import torch
+import gradio as gr
 
+# Load tokenizer dan model IndoGPT (pastikan support GPT2-style)
 model_name = "cahya/gpt2-small-indonesian-522M"
+tokenizer = GPT2Tokenizer.from_pretrained(model_name)
+model = GPT2LMHeadModel.from_pretrained(model_name)
 
-tokenizer = AutoTokenizer.from_pretrained(model_name)
-model = AutoModelForCausalLM.from_pretrained(model_name)
-
+# Load prompt dasar
 with open("prompt.txt", "r", encoding="utf-8") as f:
-    base_prompt = f.read().strip()
-
-def respond(messages):
-    history = [f"Arya: {m[0]}\nCici: {m[1]}" for m in messages if m[0] and m[1]]
-    last_user = messages[-1][0] if messages else ""
-    full_prompt = base_prompt + "\n" + "\n".join(history) + f"\nArya: {last_user}\nCici:"
-
-    inputs = tokenizer(full_prompt, return_tensors="pt", truncation=True, max_length=1024)
-    with torch.no_grad():
-        outputs = model.generate(
-            **inputs,
-            max_new_tokens=100,
-            pad_token_id=tokenizer.eos_token_id,
-            do_sample=True,
-            top_k=50,
-            top_p=0.95,
-            temperature=0.8
-        )
-
-    generated = tokenizer.decode(outputs[0], skip_special_tokens=True)
-    response = generated.split("Cici:")[-1].strip().split("Arya:")[0].strip()
-    return response
-
-chatbot_ui = gr.Chatbot(label="Cici 🤭💕").style(height=450)
-demo = gr.ChatInterface(fn=respond, chatbot=chatbot_ui, title="Cici AI Roleplay", theme="soft")
+    base_prompt = f.read()
+
+# Simpan riwayat chat
+history = []
+
+def chat(user_input):
+    global history
+    # Tambahkan input user ke riwayat
+    history.append(f"Arya: {user_input}")
+    prompt = base_prompt + "\n" + "\n".join(history) + "\nCici:"
+
+    # Tokenisasi prompt
+    inputs = tokenizer(prompt, return_tensors="pt", truncation=True, max_length=1024)
+    
+    # Generate teks
+    outputs = model.generate(
+        **inputs,
+        max_new_tokens=100,
+        do_sample=True,
+        top_k=50,
+        top_p=0.95,
+        temperature=0.8,
+        pad_token_id=tokenizer.eos_token_id
+    )
+
+    result = tokenizer.decode(outputs[0], skip_special_tokens=True)
+    
+    # Ambil jawaban terakhir dari "Cici:"
+    cici_reply = result.split("Cici:")[-1].strip().split("Arya:")[0].strip()
+    
+    # Tambahkan ke riwayat
+    history.append(f"Cici: {cici_reply}")
+    
+    # Format untuk gr.Chatbot (message history)
+    chat_messages = []
+    for msg in history:
+        if msg.startswith("Arya:"):
+            chat_messages.append(("Arya", msg[6:]))
+        elif msg.startswith("Cici:"):
+            chat_messages.append(("Cici 🤭💕", msg[6:]))
+    
+    return chat_messages
+
+def reset():
+    global history
+    history = []
+    return []
+
+with gr.Blocks() as demo:
+    gr.Markdown("## 🤖 Chatbot Roleplay Cici 🤭💕")
+    chatbot_ui = gr.Chatbot(label="Cici 🤭💕", height=450)
+    txt = gr.Textbox(label="Ketik pesan kamu...")
+
+    btn_send = gr.Button("Kirim")
+    btn_clear = gr.Button("Reset Chat")
+
+    btn_send.click(chat, inputs=txt, outputs=chatbot_ui)
+    btn_clear.click(reset, outputs=chatbot_ui)
 
 demo.launch()