List your company

Inference Latency on ViT encoder prefill 729 tokens

9.7Latency (ms)W4A8
9.6410.04510.4510.855Dec 27, 2024

Evaluation results

MethodLinks
W4A8Model=LLaVA-onevision-7B, Hardware=RTX 4090, Quantization=W4A82024.129.7
FP16Model=LLaVA-onevision-7B, Hardware=RTX 4090, Quantization=FP162024.1211.2