app.py

import gradio as gr
from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline

MODEL_NAME = "TinyLlama/TinyLlama-1.1B-Chat-v1.0"

# Load tokenizer & model
#tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)
tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)
model = AutoModelForCausalLM.from_pretrained(
    MODEL_NAME,
    device_map="auto",   # Automatically use GPU if available
    torch_dtype="auto"
)

# Create generation pipeline
story_generator = pipeline(
    "text-generation",
    model=model,
    tokenizer=tokenizer
)

# Function to generate stories
def generate_story(prompt, max_length=300, temperature=0.8):
    outputs = story_generator(
        prompt,
        max_length=max_length,
        temperature=temperature,
        do_sample=True,
        top_p=0.95,
        top_k=50
    )
    return outputs[0]["generated_text"]

# Gradio UI
with gr.Blocks() as demo:
    gr.Markdown("# 📖 Interactive Story Generator (open-gpt-oss-20b)")
    gr.Markdown("Type a prompt and let the AI continue your story with a powerful 20B model.")
    prompt = gr.Textbox(
        label="Your Story Prompt",
        placeholder="e.g., In the far future, humanity discovered a hidden planet...",
        lines=3
    )

    max_length = gr.Slider(50, 1000, value=300, step=50, label="Story Length")
    temperature = gr.Slider(0.1, 1.5, value=0.8, step=0.1, label="Creativity")

    generate_btn = gr.Button("✨ Generate Story")
    output = gr.Textbox(label="Generated Story", lines=20)

    generate_btn.click(
        fn=generate_story,
        inputs=[prompt, max_length, temperature],
        outputs=output
    )

demo.launch()