diff options
| author | Laurent Mazare <laurent.mazare@gmail.com> | 2023-09-30 19:25:47 +0200 |
|---|---|---|
| committer | GitHub <noreply@github.com> | 2023-09-30 18:25:47 +0100 |
| commit | deee7612da7dcda1aa1cfd4237f4858d9f5ed8c7 (patch) | |
| tree | 4c67080d778fb502ee665eb036b35b8c72a10103 /candle-core/src/quantized/simd128.rs | |
| parent | 06207332bc58e20680dd1925b7d90bac51f4f21c (diff) | |
| download | candle-deee7612da7dcda1aa1cfd4237f4858d9f5ed8c7.tar.gz candle-deee7612da7dcda1aa1cfd4237f4858d9f5ed8c7.tar.bz2 candle-deee7612da7dcda1aa1cfd4237f4858d9f5ed8c7.zip | |
Quantized version of mistral. (#1009)
* Quantized version of mistral.
* Integrate the quantized mistral variant.
* Use the quantized weight files.
* Tweak the quantization command.
* Fix the dtype when computing the rotary embeddings.
* Update the readme with the quantized version.
* Fix the decoding of the remaining tokens.
Diffstat (limited to 'candle-core/src/quantized/simd128.rs')
0 files changed, 0 insertions, 0 deletions