*   Add Conv3DTranspose as a builtin op.
*   `tf.summary`:
    *   Fix `tf.summary.should_record_summaries()` so it correctly reflects when
        summaries will be written, even when `tf.summary.record_if()` is not n
        effect, by returning True tensor if default writer is present.
*   Grappler:
    *   Disable default Grappler optimization timeout to make the optimization

### Breaking Changes

* PR [#2434](https://github.com/tiangolo/fastapi/pull/2434) includes several improvements that shouldn't affect normal use cases, but could affect in advanced scenarios:

overflowed). // // The paper suggests in Section 4 that we can do an "Almost Montgomery // Multiplication" by subtracting only in the overflow case, but the // cost is very similar since the constant time subtraction tells us if // x >= m as a side effect, and taking care of the broken invariant is // highly undesirable (see https://go.dev/issue/13907). copy(x.reset(n).limbs, T[n:]) x.maybeSubtractModulus(choice(c), m) // The following specialized cases follow the exact same algorithm, but // optimized...

overflowed). // // The paper suggests in Section 4 that we can do an "Almost Montgomery // Multiplication" by subtracting only in the overflow case, but the // cost is very similar since the constant time subtraction tells us if // x >= m as a side effect, and taking care of the broken invariant is // highly undesirable (see https://go.dev/issue/13907). copy(x.reset(n).limbs, T[n:]) x.maybeSubtractModulus(choice(c), m) // The following specialized cases follow the exact same algorithm, but // optimized...