作者 | SHA1 メッセージ | 日付 |
---|---|---|
|
141672a0d4 kernels: disambiguate quantized types via a new ScalarType | 5 ヶ月 前 |
|
e3f07b22c3 feat: support for QQQ W4A8 quantization (#612) | 5 ヶ月 前 |
|
598afb63dd chore: add ignored layers for fp8 quant | 5 ヶ月 前 |
|
ba371fbbbd feat: AWQ marlin kernels (#603) | 5 ヶ月 前 |
|
98cb1c4cd1 feat: support fp8 via `llm-compressor` | 5 ヶ月 前 |