].btn-block {\n  width: 100%;\n}\n\n.fade {\n  transition: opacity 0.15s linear;\n}\n\n@media (prefers-reduced-motion: reduce) {\n  .fade {\n    transition: none;\n  }\n}\n\n.fade:not(.show) {\n  opacity: 0;\n}\n\n.collapse:not(.show) {\n  display: none;\n}\n\n.collapsing {\n  position: relative;\n  height: 0;\n  overflow: hidden;\n  transition: height 0.35s ease;\n}\n\n@media (prefers-reduced-motion: reduce) {\n  .collapsing {\n    transition: none;\n  }\n}\n\n.collapsing.width {\n  width: 0;\n...

].btn-block {\n  width: 100%;\n}\n\n.fade {\n  transition: opacity 0.15s linear;\n}\n\n@media (prefers-reduced-motion: reduce) {\n  .fade {\n    transition: none;\n  }\n}\n\n.fade:not(.show) {\n  opacity: 0;\n}\n\n.collapse:not(.show) {\n  display: none;\n}\n\n.collapsing {\n  position: relative;\n  height: 0;\n  overflow: hidden;\n  transition: height 0.35s ease;\n}\n\n@media (prefers-reduced-motion: reduce) {\n  .collapsing {\n    transition: none;\n  }\n}\n\n.dropup,\n.dropright,\n.dropdown,\n.dropleft...

        for NVIDIA Ampere based GPUs which causes certain float32 ops, such as
        matrix multiplications and convolutions, to run much faster on Ampere
        GPUs but with reduced precision. This reduced precision has not been
        found to effect convergence quality of deep learning models in practice.
        TensorFloat-32 is enabled by default, but can be disabled with

- Scores from PodTopologySpreading have reduced differentiation as maxSkew increases. ([#90820](https://github.com/kubernetes/kubernetes/pull/90820), [@alculquicondor](https://github.com/alculquicondor)) [SIG Scheduling]