File size: 1,186 Bytes
5aeff64 4b7a8dd 5aeff64 4b7a8dd 5aeff64 cb66b9a 4b7a8dd 5aeff64 4b7a8dd 5aeff64 670bf11 5aeff64 cb66b9a 5aeff64 670bf11 5aeff64 |
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 |
import gradio as gr
import torch
from transformers import T5Tokenizer, AutoModelForCausalLM, pipeline
from utils import translate_from_jp_to_en
tokenizer = T5Tokenizer.from_pretrained("rinna/japanese-gpt-1b")
model = AutoModelForCausalLM.from_pretrained("rinna/japanese-gpt-1b")
generator = pipeline("text-generation", tokenizer=tokenizer, model=model)
def generate(text, min_length=512):
out = generator(text, do_sample=True, min_length=min_length, max_length=1024, num_return_sequences=1)
text = out[0]['generated_text']
return text, translate_from_jp_to_en(text)
title = "JP GPT Demo"
description = "Demo for generating text in Japanase using a GPT model"
article = "Built by Narrativa"
examples = [['日本のeスポーツ障害者がステレオタイプを撃ち落とす', 128]]
gr.Interface(fn=generate, inputs=[gr.inputs.Textbox(lines=4, label="Prompt"),
gr.inputs.Slider(minimum=8, maximum=1024, step=8, default=128, label="Number of tokens")],
outputs=["text", "text"],
title=title, description=description,
article= article,
examples=examples).launch(enable_queue=True)
|