File size: 1,375 Bytes
d6f1344 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 |
from transformers import pipeline
from datasets import load_dataset
import gradio as gr
# --- Hugging Face Pipeline Setup ---
# Replace 'model_id' with the identifier of the model you wish to use
model_id = "declare-lab/flan-alpaca-gpt4-xl"
text_generator = pipeline(model=model_id)
# --- Dataset Loading (Optional) ---
# Replace with the dataset of your choice
dataset_name = "luisotorres/wikipedia-crypto-articles" # Example dataset
dataset_split = 'train' # Choose from 'train', 'test', 'validation', etc.
dataset = load_dataset(dataset_name, split=dataset_split)
# --- Text Generation Function ---
def generate_text(prompt, max_tokens=128, system_message=""):
# Generating text using the pipeline
generated_text = text_generator(prompt, max_length=max_tokens, do_sample=True)[0]["generated_text"]
# Incorporating the system message
full_message = system_message + "\n\nGenerated Text:\n" + generated_text
return full_message
# --- Gradio Interface Setup ---
iface = gr.Interface(
fn=generate_text,
inputs=[
gr.inputs.Textbox(lines=2, placeholder="Enter Prompt Here..."),
gr.inputs.Slider(minimum=1, maximum=256, default=128, label="Max Tokens"),
gr.inputs.Textbox(label="System Message", placeholder="Enter a system message here...")
],
outputs="text"
)
# --- Launch the Interface ---
iface.launch()
|