Codec kernels¶
Restrict kernel selection to discovered neural-audio codec operations.
Use¶
result = model.apply_optimization_plan(
"codec-kernels",
mode="inference",
)
print(result.manifest())
Support¶
| Property | Value |
|---|---|
| Availability | Registered public pass: codec-kernels |
| Fidelity | Operator-equivalent intent; validate reconstructed audio |
| Runtime | CPU, CUDA, or MPS; Triton, CuTe, and CUDA extensions require CUDA |
| Registry name | codec-kernels |
| Pass ID | codec-kernels |
| Pass version | 1 |
| Restore | model.restore_optimization_plan(mode="inference") |
Unsupported explicit configurations must fail before mutation. A pass that
does not match a model reports not-applicable; it is not an acceleration.
Paper and GitHub¶
- Paper: Triton: an intermediate language and compiler for tiled neural network computations
- Upstream GitHub: Triton; NVIDIA CUTLASS / CuTe
- VoiceHub source: VoiceHub implementation
Verify¶
Compare the eager and optimized paths with the same checkpoint, input, seed, warm-up, device, and dtype. Record latency, memory, output quality, the exact source revision, and the optimization manifest.
See the related workflow and optimization API.