tests/bnb/test_mixed_int8.py · 2230d149f0dc4de05ea7a9a1c5295b6d80e495cd · chenpangpang / transformers · GitLab

Find file Blame History Permalink

fix get_keys_to_not_convert() to return correct modules for full precision inference (#25105) · 2230d149
YQ authored Aug 02, 2023
```
* add test for `get_keys_to_not_convert`

* add minimum patch to keep mpt lm_head from 8bit quantization

* add reivsion to
```
2230d149

test_mixed_int8.py 32.9 KB