app.py

from transformers import pipeline
import gradio as gr
import spaces
# Initialize the text generation pipeline with optimizations
pipe = pipeline("text-generation", model="SakanaAI/EvoLLM-JP-v1-7B")
# Define a function to generate text based on user input
def generate_text(prompt):
    result = pipe(prompt, max_length=50, num_return_sequences=1)
    return result[0]['generated_text']

# Define a function to generate text based on user input
@spaces.GPU
def generate_text(prompt):
    result = pipe(prompt, max_length=50, num_return_sequences=1)
    return result[0]['generated_text']

# Create a Gradio interface with batching enabled
iface = gr.Interface(
    fn=generate_text, 
    inputs=gr.inputs.Textbox(lines=2, placeholder="Enter your prompt here..."), 
    outputs="text",
    title="Text Generation with SakanaAI/EvoLLM-JP-v1-7B",
    description="Enter a prompt and the model will generate a continuation of the text.",
    batch=True,
    max_batch_size=4
)

# Launch the interface
if __name__ == "__main__":
    iface.launch()