background-repeat: repeat-x;\n}\n@mixin gradient-x-three-colors($start-color: $blue, $mid-color: $purple, $color-stop: 50%, $end-color: $red) {\n  background-image: linear-gradient(to right, $start-color, $mid-color $color-stop, $end-color);\n  background-repeat: no-repeat;\n}\n@mixin gradient-y-three-colors($start-color: $blue, $mid-color: $purple, $color-stop: 50%, $end-color: $red) {\n  background-image: linear-gradient($start-color, $mid-color $color-stop, $end-color);\n  background-repeat: no-repeat;\n}\n@mixin...

    * Add `compute_sparse_core_stats` for sparse core users to profile the
      data with this API to get the `max_ids` and `max_unique_ids`. These
      numbers will be needed to configure the sparse core embedding mid level
      api.
    * Remove the `preprocess_features` method since that's no longer needed.

## Thanks to our Contributors

This release contains contributions from many people at Google, as well as:

becomes still more plain and convincing. Let the second Prism DH [in
_Fig._ 16.] be placed not immediately after the first, but at some
distance from it; suppose in the mid-way between it and the Wall on
which the oblong Spectrum PT is cast, so that the Light from the first
Prism may fall upon it in the form of an oblong Spectrum [Greek: pt]