SparseML v0.12.1 Patch Release
jeanniefinks
released this
05 May 18:39
·
2 commits
to release/0.12
since this release
This is a patch release for 0.12.0 that contains the following changes:
- Disabling of distillation modifiers no longer crashes Hugging Face Transformers integrations
--distillation_teacher disable
- Numeric stability is provided for distillation modifiers using log_softmax instead of softmax.
- Accuracy and performance issues were addressed for quantized graphs in image classification and NLP.
- When using mixed precision for a quantized recipe with image classification, crashes no longer occur.