tokenizer_kernel / README.md
Allanatrix's picture
Publish tokenizer_kernel kernel and performance card
df4a67f verified
|
Raw
History Blame Contribute Delete
1.15 kB
---
license: apache-2.0
tags:
- cuda
- kernel
- gpu-optimization
- hpc
---
# tokenizer_kernel
Experimental tokenizer/matrix CUDA kernel file.
This repository contains the standalone CUDA source for the `tokenizer_kernel` lane from
the PyC kernel lab. It is a source artifact for inspection and benchmarking;
it is not a precompiled binary and the result below is not a universal ranking.
## Performance
| Kernel | GPU / architecture | Shape | Best recorded result | Evidence |
|---|---|---|---|---|
| `tokenizer_kernel` | not recorded | not recorded | Not measured in the published campaign | No published performance receipt was found for this lane. |
![Performance plot](performance.svg)
The result is reported with the original campaign's timing and correctness
context. Compare kernels only when GPU, CUDA version, matrix shape, warmup,
repeats, and reference/correctness mode match.
## Source
- `kernel.cu` — copied from `kernels/prototypes/experimental/tokenizer_matmul/kernel.cu`.
- Original lane tags: `cuda, tokenizer, experimental`.
## Build/run contract
```text
{nvcc} -O2 -c {source} -o {build_dir}/{name}.o
(compile-only)
```