Siyuan Feng

icon indicating a website link https://syfeng.net

icon company SJTU icon location Shanghai icon location ML System & Compiler | ASF Member | PMC Member of Apache TVM

Results 99 comments of Siyuan Feng

[Question] Do we compare the batch performance with vllm?

cc @MasterJH5574 @yzh119

Comments for "How to Guides"

我仔细看了一下，应该是我review不仔细的问题。原本说的`?`是`Simple: nn.conv2d Composite: nn.pad?, nn.conv2d, nn.bias_add?, nn.relu? `其中的问号，应该是合理的。

Comments for "How to Guides"

@antonia0912 我本意是加上英文reduce，原因是reduce目前在并没有一个官方或者合理的翻译，一般平时交流还是一reduce/reduction居多，所以加上英文可能会更清楚一点

[Feature Request] HUAWEI Mate60 Pro Support

We have no plan to support HUAWEI officially shortly, but happy to see if there is community contribution for it :)

[BugFix][MSC] split name_string with index by colon from the right

cc @Archermmt

[Codegen, CUDA] Add FP8 Tensor Core Codegen

cc @vinx13

[BugFix] Fix Crash Cases Caused by "__tvm_meta__ = None"

cc @tqchen

[BugFix][TOPI] fix resize accuracy issue

Thanks for the fix. Could you please also add a topi test?

[Array] Add deduction guide for Array(IterType,IterType) constructor

@jikechao Could you please try to upgrade your cxx compiler?

[Doc] Refactor How-To

preview at http://8.130.134.126:58000/how_to/