Each branch contains an individual bits per weight, with the main one containing only the meaurement.json for further conversions. Conversion was done using wikitext-103-raw-v1-test.parquet as calibration dataset. Default arguments used except when the bits per weight is above 6.0, at that point the lm_head layer is quantized at 8 bits per weight instead of the default 6. Original model: https://huggingface.co/fblgit/una-cybertron-7b-v2-bf16 To download the main (only useful if you only care about measurement.json) branch to a folder called una-cybertron-7b-v2-bf16-exl2: To download from a different branch, add the —revision parameter:
Модальности:
Генерация текста
Задача: Генерация текста
Автор: bartowski
Теги: juanako, UNA, cybertron, fbl, endpoints_compatible
Лайков: 3 | Загрузок: 1
Описание основано на материалах HuggingFace. Перевод выполнен автоматически.