* Disabling TensorFloat-32 execution now causes TPUs to use float32 precision for float32 matmuls and other ops. TPUs have always used bfloat16 precision for certain ops, like matmul, when such ops had float32 inputs. Now, disabling TensorFloat-32 by calling `tf.config.experimental.enable_tensor_float_32_execution(False)` will cause TPUs to use float32 precision for such ops instead of bfloat16.

*  `tf.experimental.dtensor`

- Fixes an issue with the max-in-flight API server filter where the filter could take a long time to process an incoming request if it had been a long time since the last request. ([#96282](https://github.com/kubernetes/kubernetes/pull/96282), [@staebler](https://github.com/staebler)) [SIG API Machinery]

    * Read the new docs: [Dependencies with `yield` and `except`](https://fastapi.tiangolo.com/tutorial/dependencies/dependencies-with-yield/#dependencies-with-yield-and-except).

In short, if you had dependencies that looked like:

```Python
def my_dep():