|
import gradio as gr |
|
import torch |
|
from transformers import AutoTokenizer, AutoModelForCausalLM |
|
import spaces |
|
|
|
title = "Fakypedia" |
|
|
|
DESCRIPTION = """\ |
|
# Generate a silly article |
|
A bilingual (English and Hebrew) [nonsense generation model](https://huggingface.co/Norod78/SmolLM-135M-FakyPedia-EngHeb) which produces silly Wikipedia-like abstract text. |
|
Tap on the \"Submit\" button to generate a silly and/or fake \"Wikipedia-Like\" article based on the input title |
|
""" |
|
|
|
article = "<p>This model extended the tokenizer of and is a fine-tuned of [SmolLM-135M-Instruct](https://huggingface.co/HuggingFaceTB/SmolLM-135M-Instruct)</p>" |
|
|
|
CUDA_AVAILABLE = torch.cuda.is_available() |
|
device = torch.device("cuda" if CUDA_AVAILABLE else "cpu") |
|
|
|
model_id = "Norod78/SmolLM-135M-FakyPedia-EngHeb" |
|
tokenizer = AutoTokenizer.from_pretrained(model_id) |
|
tokenizer.pad_token_id = tokenizer.eos_token_id |
|
bos_token = tokenizer.bos_token |
|
model = AutoModelForCausalLM.from_pretrained(model_id).to(device) |
|
model.generation_config.pad_token_id = tokenizer.pad_token_id |
|
|
|
torch.manual_seed(1234) |
|
|
|
@spaces.GPU |
|
def generate_fakypedia(article_title: str): |
|
with torch.no_grad(): |
|
result = "" |
|
string_to_tokenize= f"{bos_token}\\%{article_title}" |
|
input_ids = tokenizer( string_to_tokenize, return_tensors="pt").input_ids.to(device) |
|
sample_outputs = model.generate(input_ids, do_sample=True,repetition_penalty=1.05, top_k = 40, top_p = 0.950, temperature=0.80, max_length=192, num_return_sequences=3) |
|
|
|
if article_title == None or len(article_title) == 0: |
|
result += f"# Fakypedia results with random titles \n" |
|
article_title = "" |
|
else: |
|
result += f"# Fakypedia results for \"{article_title}\" \n" |
|
for i, sample_output in enumerate(sample_outputs): |
|
decoded_output = tokenizer.decode(sample_output, skip_special_tokens=True) |
|
decoded_output = decoded_output.replace(f"\%{article_title}", f"## {i+1}. {article_title}").replace("\%", " ").replace("\\n", " \n") |
|
decoded_output = decoded_output.replace("## \n", "\n") |
|
result += "{}\n".format(decoded_output) |
|
return result |
|
|
|
demo = gr.Interface( |
|
generate_fakypedia, |
|
inputs=gr.Textbox(lines=1, label="Enter a title for the article (or leave blank for a random one)"), |
|
outputs=gr.Markdown(label="Generated fakypedia article"), |
|
title=title, |
|
description=DESCRIPTION, |
|
article=article, |
|
examples=["Hugging face", "A socially awkward potato", "讚讜专讜谉 讗讚诇专", "Elon Musk"], |
|
allow_flagging="never", |
|
) |
|
|
|
demo.queue() |
|
demo.launch() |
