⚡ WebGPU Benchmark Results (33.26x speedup)
#73
by
a414166402
- opened
| Batch Size | WASM (fp32) | WebGPU (fp32) |
| 1 | 473.20 | 23.30 |
| 2 | 957.00 | 59.60 |
| 4 | 1883.20 | 88.90 |
| 8 | 3943.20 | 181.00 |
| 16 | 8269.70 | 286.30 |
| 32 | 16568.80 | 498.10 |
- 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/124.0.0.0 Safari/537.36
- GPU: vendor=nvidia, architecture=turing, device=, description=