June 2025 monthly summary for tensorflow/tensorflow focusing on feature delivery and stability improvements. Implemented granular control to disable StableHLO->XLA replacement via the _no_xla_call_module attribute, updated the function replacement pass to honor the attribute, and added tests to validate behavior. Fixed forward for the broken tests related to this feature and stabilized the test suite for this transformation workflow.

May 2025 monthly summary for tensorflow/tensorflow focusing on feature delivery and impact. The period delivered key stability and integration improvements in StableHLO with selective XlaCallModule replacement, enhancing compatibility with quantization workflows and internal compiler usage. No explicit bug fixes were reported in the provided data; the emphasis was on feature delivery and internal tooling visibility.

Implemented a precise, minimal-impact fix to preserve parameter dtype in normalization layers across flax, improving numerical stability and consistency in mixed-precision training.