t2) // Z3 := b * t2 x3.Sub(y3, z3) // X3 := Y3 - Z3 z3.Add(x3, x3) // Z3 := X3 + X3 x3.Add(x3, z3) // X3 := X3 + Z3 z3.Sub(t1, x3) // Z3 := t1 - X3 x3.Add(t1, x3) // X3 := t1 + X3 y3.Mul({{.p}}B(), y3) // Y3 := b * Y3 t1.Add(t2, t2) // t1 := t2 + t2 t2.Add(t1, t2) // t2 := t1 + t2 y3.Sub(y3, t2) // Y3 := Y3 - t2 y3.Sub(y3, t0) // Y3 := Y3 - t0 t1.Add(y3, y3) // t1 := Y3 + Y3 y3.Add(t1, y3) // Y3 := t1 + Y3 t1.Add(t0, t0) // t1 := t0 + t0 t0.Add(t1, t0) // t0 := t1 + t0 t0.Sub(t0, t2) // t0 := t0 -...

t2) // Z3 := b * t2 x3.Sub(y3, z3) // X3 := Y3 - Z3 z3.Add(x3, x3) // Z3 := X3 + X3 x3.Add(x3, z3) // X3 := X3 + Z3 z3.Sub(t1, x3) // Z3 := t1 - X3 x3.Add(t1, x3) // X3 := t1 + X3 y3.Mul({{.p}}B(), y3) // Y3 := b * Y3 t1.Add(t2, t2) // t1 := t2 + t2 t2.Add(t1, t2) // t2 := t1 + t2 y3.Sub(y3, t2) // Y3 := Y3 - t2 y3.Sub(y3, t0) // Y3 := Y3 - t0 t1.Add(y3, y3) // t1 := Y3 + Y3 y3.Add(t1, y3) // Y3 := t1 + Y3 t1.Add(t0, t0) // t1 := t0 + t0 t0.Add(t1, t0) // t0 := t1 + t0 t0.Sub(t0, t2) // t0 := t0 -...

## 0.4.0

* Add `openapi_prefix`, support for reverse proxy and mounting sub-applications. See the docs at [https://fastapi.tiangolo.com/advanced/sub-applications-proxy/](https://fastapi.tiangolo.com/advanced/sub-applications-proxy/): [#26](https://github.com/tiangolo/fastapi/pull/26) by [@kabirkhan](https://github.com/kabirkhan).

*   Enable JIT-compiled i64-indexed kernels on GPU for large tensors with more than 2**32 elements.
    *   Unary GPU kernels: Abs, Atanh, Acos, Acosh, Asin, Asinh, Atan, Cos, Cosh, Sin, Sinh, Tan, Tanh.
    *   Binary GPU kernels: AddV2, Sub, Div, DivNoNan, Mul, MulNoNan, FloorDiv, Equal, NotEqual, Greater, GreaterEqual, LessEqual, Less.

* `tf.lite`
    * Add experimental supports conversion of models that may be larger than 2GB before buffer deduplication