T5 Translate En Ru Zh Small 1024

🚀 T5 English, Russian and Chinese Multilingual Machine Translation

This T5-based model is designed for multitasking machine translation, enabling seamless translation between Russian, Chinese, and English.

🚀 Quick Start

This model is a conventional T5 transformer operating in multitasking mode for machine translation. It's precisely configured for the following language pairs: ru-zh, zh-ru, en-zh, zh-en, en-ru, and ru-en.

The model can directly translate between any two of Russian, Chinese, or English. To translate into the target language, prefix the text with 'translate to :'. The source language doesn't need to be specified, and the source text can be multilingual.

💻 Usage Examples

Basic Usage

Example: Translate Russian to Chinese

from transformers import T5ForConditionalGeneration, T5Tokenizer

device = 'cuda' #or 'cpu' for translate on cpu

model_name = 'utrobinmv/t5_translate_en_ru_zh_small_1024'
model = T5ForConditionalGeneration.from_pretrained(model_name)
model.to(device)
tokenizer = T5Tokenizer.from_pretrained(model_name)

prefix = 'translate to zh: '
src_text = prefix + "Цель разработки — предоставить пользователям личного синхронного переводчика."

# translate Russian to Chinese
input_ids = tokenizer(src_text, return_tensors="pt")

generated_tokens = model.generate(**input_ids.to(device))

result = tokenizer.batch_decode(generated_tokens, skip_special_tokens=True)
print(result)
#开发的目的就是向用户提供个性化的同步翻译。

Example: Translate Chinese to Russian

from transformers import T5ForConditionalGeneration, T5Tokenizer

device = 'cuda' #or 'cpu' for translate on cpu

model_name = 'utrobinmv/t5_translate_en_ru_zh_small_1024'
model = T5ForConditionalGeneration.from_pretrained(model_name)
model.to(device)
tokenizer = T5Tokenizer.from_pretrained(model_name)

prefix = 'translate to ru: '
src_text = prefix + "开发的目的就是向用户提供个性化的同步翻译。"

# translate Russian to Chinese
input_ids = tokenizer(src_text, return_tensors="pt")

generated_tokens = model.generate(**input_ids.to(device))

result = tokenizer.batch_decode(generated_tokens, skip_special_tokens=True)
print(result)
#Цель разработки - предоставить персонализированный синхронный перевод для пользователей.

📚 Documentation

Languages covered

Russian (ru_RU), Chinese (zh_CN), English (en_US)

📄 License

This project is licensed under the Apache-2.0 license.

📦 Additional Information