⚡ WebGPU Benchmark Results (97.38x speedup)
#58
by
Xenova
HF staff
- opened
| Batch Size | WASM (fp16) | WebGPU (fp16) |
| 1 | 1074.60 | 81.20 |
| 2 | 2150.80 | 57.50 |
| 4 | 4336.50 | 189.30 |
| 8 | 8737.80 | 327.40 |
| 16 | 17437.00 | 179.30 |
| 32 | 34789.40 | 449.80 |
| 64 | 69838.60 | 717.20 |
- Model: Xenova/all-MiniLM-L6-v2
- Tests run: WASM (fp16), WebGPU (fp16)
- Sequence length: 512
- Browser: Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/122.0.0.0 Safari/537.36
- GPU: vendor=nvidia, architecture=turing, device=, description=