caret-down() {\n  border-top: $caret-width solid;\n  border-right: $caret-width solid transparent;\n  border-bottom: 0;\n  border-left: $caret-width solid transparent;\n}\n\n@mixin caret-up() {\n  border-top: 0;\n  border-right: $caret-width solid transparent;\n  border-bottom: $caret-width solid;\n  border-left: $caret-width solid transparent;\n}\n\n@mixin caret-right() {\n  border-top: $caret-width solid transparent;\n  border-right: 0;\n  border-bottom: $caret-width solid transparent;\n  border-left:...

caret-down() {\n  border-top: $caret-width solid;\n  border-right: $caret-width solid transparent;\n  border-bottom: 0;\n  border-left: $caret-width solid transparent;\n}\n\n@mixin caret-up() {\n  border-top: 0;\n  border-right: $caret-width solid transparent;\n  border-bottom: $caret-width solid;\n  border-left: $caret-width solid transparent;\n}\n\n@mixin caret-right() {\n  border-top: $caret-width solid transparent;\n  border-right: 0;\n  border-bottom: $caret-width solid transparent;\n  border-left:...

        libraries and headers.
    *   TF code now resides in `tensorflow_core` and `tensorflow` is just a
        virtual pip package. No code changes are needed for projects using
        TensorFlow, the change is transparent
*   XLA
    *   XLA HLO graphs can be inspected with interactive_graphviz tool now.
*   Estimator
    *   Use tf.compat.v1.estimator.inputs instead of tf.estimator.inputs