BELLE-7B-2M — самая продвинутая итерация в линейке, обученная на массивном наборе из 2 миллионов инструкций. Эта языковая модель демонстрирует впечатляющие результаты в задачах генерации контента и глубокого понимания контекста.
BELLE is based on Bloomz-7b1-mt and finetuned with 2M Chinese data combined with 50,000 pieces of English data from the open source Stanford-Alpaca, resulting in good Chinese instruction understanding and response generation capabilities. The code of Chinese data generation and other detailed information can be found in our Github project repository: https://github.com/LianjiaTech/BELLE. We trained models using datasets of different sizes (200,000, 600,000, 1,000,000, and 2,000,000 samples) for instruction learning, and we obtained different model versions as shown below: