perf: cuBLAS matmul fwd/bwd

Route Tensor::matmul and matmul_backward through cuBLAS Sgemm instead of
the hand-written tiled kernel. fp32 → same GEMM up to rounding order, so
the T3 cuBLAS tolerance and downstream grad-checks are preserved.

- cublas.rs: thread-local persistent handle + row-major sgemm helper with
  transpose flags (col-major⟺row-major as the T3 oracle does).
- matmul_backward: dA/dB via cuBLAS OP_T, dropping the two transpose
  kernels + their allocations the T3 version ran.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

crates/xtrain-cuda/src/lib.rs

@@ -1,3 +1,5 @@
+#[cfg(not(no_cuda))]
+pub mod cublas;
 pub mod device;
 pub mod error;
 pub mod ffi;