`tf.distribute.experimental.MultiWorkerMirroredStrategy`
    *   Update NVIDIA `NCCL` to `2.5.7-1` for better performance and performance
        tuning. Please see
        [nccl developer guide](https://docs.nvidia.com/deeplearning/sdk/nccl-developer-guide/docs/env.html)
        for more information on this.
    *   Support gradient `allreduce` in `float16`. See this

receiving programmatic focus was a keyboard interaction,\n// or the browser has somehow determined that the user is primarily a keyboard user and/or\n// wants focus outlines to always be presented.\n//\n// See https://developer.mozilla.org/en-US/docs/Web/CSS/:focus-visible\n// and https://developer.paciellogroup.com/blog/2018/03/focus-visible-and-backwards-compatibility/\n[tabindex=\"-1\"]:focus:not(:focus-visible) {\n  outline: 0 !important;\n}\n\n\n// Content grouping\n//\n// 1. Add the correct box sizing...

receiving programmatic focus was a keyboard interaction,\n// or the browser has somehow determined that the user is primarily a keyboard user and/or\n// wants focus outlines to always be presented.\n//\n// See https://developer.mozilla.org/en-US/docs/Web/CSS/:focus-visible\n// and https://developer.paciellogroup.com/blog/2018/03/focus-visible-and-backwards-compatibility/\n[tabindex=\"-1\"]:focus:not(:focus-visible) {\n  outline: 0 !important;\n}\n\n\n// Content grouping\n//\n// 1. Add the correct box sizing...

- `EventRecorder()` is exposed to `FrameworkHandle` interface so that scheduler plugin developers can choose to log cluster-level events. ([#92010](https://github.com/kubernetes/kubernetes/pull/92010), [@Huang-Wei](https://github.com/Huang-Wei)) [SIG Scheduling]