Spaces:
Sleeping
Sleeping
| import os | |
| import torch | |
| import gradio as gr | |
| import time | |
| from transformers import AutoTokenizer, pipeline | |
| # Modelo de Meta capaz de traducir a más de 200 idiomas | |
| model = 'facebook/nllb-200-distilled-600M' | |
| #model = 'JustFrederik/nllb-200-distilled-600M-ct2-int8' | |
| tokenizer = 'facebook/nllb-200-distilled-600M' # El modelo Quantizado no tiene tokenizador | |
| # Pero como traduce mal al asturiano, añadimos este otro fine-tuned | |
| # Es del proyecto AINA: https://huggingface.co/projecte-aina | |
| model_ast = "projecte-aina/aina-translator-es-ast" | |
| flores_codes = {} | |
| flores_codes["Asturianu"] = "ast_Latn" | |
| flores_codes["Castellano"] = "spa_Latn" | |
| flores_codes["Català"] = "cat_Latn" | |
| flores_codes["English"] = "eng_Latn" | |
| flores_codes["Euskera"] = "eus_Latn" | |
| flores_codes["Galego"] = "glg_Latn" | |
| def translation(source, target, text): | |
| #start_time = time.time() | |
| source = flores_codes[source] | |
| target = flores_codes[target] | |
| if target == flores_codes["Asturianu"]: | |
| texto_castellano = source | |
| if source != flores_codes["Castellano"]: | |
| translator = pipeline('translation', model=model, tokenizer=tokenizer, src_lang=source, tgt_lang=flores_codes["Castellano"]) | |
| texto_castellano = translator(text, max_length=400) | |
| translator_ast = pipeline('translation', model=model_ast, tokenizer=tokenizer, src_lang=flores_codes["Castellano"], tgt_lang=flores_codes["Asturianu"]) | |
| output = translator_ast(text, max_length=400) | |
| else: | |
| translator = pipeline('translation', model=model, tokenizer=tokenizer, src_lang=source, tgt_lang=target) | |
| output = translator(text, max_length=400) | |
| #end_time = time.time() | |
| output = output[0]['translation_text'] | |
| #result = {'inference_time': end_time - start_time, | |
| # 'source': source, | |
| # 'target': target, | |
| # 'result': output} | |
| #return result | |
| return output; | |
| if __name__ == '__main__': | |
| print('\tIniciando...') | |
| # define gradio demo | |
| lang_codes = list(flores_codes.keys()) | |
| inputs = [gr.Dropdown(lang_codes, value='Castellano', label='Idioma original'), | |
| gr.Dropdown(lang_codes, value='Asturianu', label='Traducir al...'), | |
| gr.Textbox(label="Texto a traducir"), | |
| ] | |
| outputs = [gr.Textbox(label="Texto traducido"),] | |
| title = "Traductor Multilingüe" | |
| description = """ | |
| Este traductor utiliza el siguiente modelo de lenguaje de Meta: https://github.com/facebookresearch/fairseq/tree/nllb\n | |
| Excepto para traducir al asturiano que usa el modelo del proyecto AINA: https://huggingface.co/projecte-aina/aina-translator-es-ast\n | |
| Adaptado de: https://huggingface.co/spaces/Azwaw/Text_Translation_Multi-languages | |
| """ | |
| gr.Interface(translation, | |
| inputs, | |
| outputs, | |
| title=title, | |
| description=description, | |
| submit_btn="Traducir" | |
| ).launch() | |