⚡ WebGPU Benchmark Results (88.85x speedup)
#105
by
Xenova
HF staff
- opened
Batch Size | WASM (fp32) | WebGPU (fp32) |
1 | 2116.80 | 247.60 |
2 | 4068.10 | 307.50 |
4 | 8315.50 | 338.80 |
8 | 16611.50 | 369.50 |
16 | 32932.30 | 491.10 |
32 | 65857.60 | 741.20 |
- Model: sentence-transformers/all-MiniLM-L12-v1
- 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/129.0.0.0 Safari/537.36
- GPU: vendor=nvidia, architecture=turing, device=, description=