作者 | SHA1 備註 | 提交日期 |
---|---|---|
|
00503b9fc1 feat: non-uniform quantization via `compressed-tensors` for llama | 5 月之前 |
|
6671e3a162 feat: add CPU offloading support (#598) | 6 月之前 |
|
497bf64942 chore: simplify pipeline parallel code in llama | 6 月之前 |
|
4599c98f99 feat: dynamic image size support for VLMs | 6 月之前 |