(Terry) Tang, Yuxin Wu, Ziyue(Louis) Lu

# Release 1.7.0

## Major Features And Improvements

*   Eager mode is moving out of contrib, try `tf.enable_eager_execution()`.
*   Graph rewrites emulating fixed-point quantization compatible with TensorFlow
    Lite, supported by new `tf.contrib.quantize` package.
*   Easily customize gradient computation with `tf.custom_gradient`.

:= priv.pub.a[:k*l] computeMatrixA(A, ρ[:], p) for r := range l { priv.s1[r] = ntt(s1[r]) } for r := range k { priv.s2[r] = ntt(s2[r]) } for r := range k { priv.t0[r] = ntt(t0[r]) } // We need to put something in priv.seed, and putting random bytes feels // safer than putting anything predictable. drbg.Read(priv.seed[:]) // Making this format *even more* annoying, we need to recompute t1 from ρ, // s1, and s2 if we want to generate the public key. This is essentially as // much work as regenerating...