Praxis 1.2.0 release#32

Open

chandrasekhard2 wants to merge 5 commits intomainfrom

Commits on Oct 16, 2023

Add activation quantization options for `quantization.for_transformer`

The praxis Authors
authored and
chandrasekhard2
committed
Allows custom implementations of StackedTransformer to be used with praxis.layers.transformer_models.TransformerLM class.

The praxis Authors
authored and
chandrasekhard2
committed
Praxis 1.2.0 release
chandrasekhard2
committed
Fix the bug in asymmetric activation quantization inference logic.

The praxis Authors
authored and
chandrasekhard2
committed
Update praxis 1.2.0 requirements
chandrasekhard2
committed