⚡ WebGPU Benchmark Results (41.47x speedup)
#93
by
silait
- opened
Batch Size | WASM (fp32) | WebGPU (fp32) |
1 | 879.50 | 81.00 |
2 | 1846.70 | 297.60 |
4 | 3610.40 | 142.60 |
8 | 6935.90 | 380.10 |
16 | 13926.60 | 417.60 |
32 | 28082.20 | 677.20 |
- Model: Xenova/all-MiniLM-L6-v2
- Tests run: WASM (fp32), WebGPU (fp32)
- Sequence length: 512
- Browser: Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/123.0.0.0 Safari/537.36
- GPU: vendor=nvidia, architecture=lovelace, device=, description=