Vincent Zhong
Vincent Zhong
See #6601 since we updated to the default to be `xgrammar` alreaadya in the meantime, but one docs mention doesn't use it
> [!NOTE] > Edit update: I just realized the original was 28 micros, so this result seems fine, and is comparable. I investigated this on latest sgl, which has upgraded...
Is it tested for MoE?
## Repro B200x8, TP 0 Here is my script how to use it: `head_dim = 128, num_heads = 64, typical decode batch ≈ 8 → jobs = 512.` ```python python3...
> I don't understand "it's not a good experimental setup", can you explain more? Based on my understanding of sgl kernel, in order to use it at runtime I can...