model_translation.py

"""
File: model_translation.py

Description: 
   Loading models for text translations (EN->FR, FR->EN)

Author: Didier Guillevic
Date: 2024-03-16
"""

import torch
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM

src_langs = set(["ar", "en", "fa", "fr", "he", "ja", "zh"])
model_names = {
    "ar": "Helsinki-NLP/opus-mt-ar-en",
    "en": "Helsinki-NLP/opus-mt-en-fr",
    "fa": "Helsinki-NLP/opus-mt-tc-big-fa-itc",
    "fr": "Helsinki-NLP/opus-mt-fr-en",
    "he": "Helsinki-NLP/opus-mt-tc-big-he-en",
    "ja": "Helsinki-NLP/opus-mt-jap-en",
    "zh": "Helsinki-NLP/opus-mt-zh-en",
}

# Registry for all loaded bilingual models
tokenizer_model_registry = {}

device = 'cuda' if torch.cuda.is_available() else 'cpu'

def get_tokenizer_model_for_src_lang(src_lang: str) -> (AutoTokenizer, AutoModelForSeq2SeqLM):
    """
    Return the (tokenizer, model) for a given source language.
    """
    src_lang = src_lang.lower()

    # Already loaded?
    if src_lang in tokenizer_model_registry:
        return tokenizer_model_registry.get(src_lang)

    # Load tokenizer and model
    model_name = model_names.get(src_lang)
    if not model_name:
        raise Exception(f"No model defined for language: {src_lang}")
    
    # We will leave the models on the CPU (for now)
    tokenizer = AutoTokenizer.from_pretrained(model_name)
    model = AutoModelForSeq2SeqLM.from_pretrained(model_name)
    if model.config.torch_dtype != torch.float16:
        model = model.half()
    model = model.to(device)
    tokenizer_model_registry[src_lang] = (tokenizer, model)

    return (tokenizer, model)

# Max number of words for given input text
# - Usually 512 tokens (max position encodings, as well as max length)
# - Let's set to some number of words somewhat lower than that threshold
# - e.g. 200 words
max_words_per_chunk = 200

#
# Multilingual translation model
#
model_MADLAD_name = "google/madlad400-3b-mt"
#model_MADLAD_name = "google/madlad400-7b-mt-bt"
tokenizer_multilingual = AutoTokenizer.from_pretrained(model_MADLAD_name, use_fast=True)
model_multilingual = AutoModelForSeq2SeqLM.from_pretrained(
    model_MADLAD_name, device_map="auto", torch_dtype=torch.float16)