Delete main.py

main.py

@@ -1,44 +0,0 @@
-import gradio as gr
-from transformers import AutoModelForCausalLM, AutoTokenizer
-import torch
-
-# Small public model, reliable
-MODEL_NAME = "facebook/opt-125m"
-
-tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)
-model = AutoModelForCausalLM.from_pretrained(MODEL_NAME).to("cpu")
-
-history = []
-
-def chat_with_airi(user_msg):
-    global history
-    prompt = ""
-    for u, a in history[-5:]:
-        prompt += f"User: {u}\nAiri: {a}\n"
-    prompt += f"User: {user_msg}\nAiri:"
-
-    inputs = tokenizer(prompt, return_tensors="pt").to("cpu")
-    with torch.no_grad():
-        output = model.generate(
-            **inputs,
-            max_new_tokens=60,
-            do_sample=True,
-            top_p=0.9,
-            temperature=0.8,
-            pad_token_id=tokenizer.eos_token_id
-        )
-    reply = tokenizer.decode(output[0], skip_special_tokens=True)
-    reply = reply.split("Airi:", 1)[-1].strip()
-
-    history.append([user_msg, reply])
-    return history, ""
-
-with gr.Blocks() as demo:
-    gr.HTML("<h2 style='text-align:center'>Airi — Mini Chat AI</h2>")
-    gr.HTML("<p style='text-align:center;color:#666;'>Small, Fast & Public Model</p>")
-
-    chat = gr.Chatbot()
-    msg = gr.Textbox(label="Talk to Airi…", placeholder="Write here…")
-    msg.submit(chat_with_airi, msg, [chat, msg])
-
-demo.launch()