— llama.cpp — ramalama — LM Studio — koboldcpp — Jan AI — Text Generation Web UI — LoLLMs Kimi-K2.6 is uploaded as int4, and with a small patch by jukofyork, Q4_0 maps 1:1 from the original FFN tensors, making it lossless quality For that reason, Q40 is provided as the highest quality, with any non-FFN tensors set to Q80 (which is why the metadata shows Q80`, I should fix that but would need to re-upload and it changes nothing) Some of these quants (Q3KXL, Q4KL etc) are the s If the model is bigger than 50GB, it will have been split into multiple files. In order to download them all to a local folder, run: You can either specify a new local-dir (moonshotaiKimi-K2.6-Q80) or download them all in place (./) Previously, you would download Q4044/48/8_8, and these would have their weights interleaved in memory in order to improve performance on ARM and AVX machines by loading up more data in one pass. Now, however, there is something called «online repacking» for weights. details in this PR. If you use Q4_0 and your hardware would benefit from repacking…
Модальности:
Генерация текста
Области применения:
Диалог / чат
Задача: Генерация текста
Автор: bartowski
Теги: gguf, compressed-tensors, endpoints_compatible, imatrix, conversational
Лайков: 4 | Загрузок: 1,169
Описание основано на материалах HuggingFace. Перевод выполнен автоматически.
