microsoft / microsoft/TRELLIS

Overflow error in spconv

Open
#188 3 comments 4 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
Python
Stars
13.7k
Forks
1.3k
PR merge metrics
No merged PRs in 30d

Description

Hi, I got this error on some batches when training mesh decoder.

RuntimeError: /io/build/temp.linux-x86_64-cpython-310/spconv/build/core_cc/src/cumm/conv/main/ConvMainUnitTest/ConvMainUnitTest_matmul_split_Simt_f32f32f32_0.cu(222)
int64_t(N) * int64_t(C) * tv::bit_size(algo_desp.dtype_a) / 8 < int_max assert faild. your data exceed int32 range. this will be fixed in cumm + nvrtc (spconv 2.2/2.3).

Contributor guide

No contributing guide indexed for this repository

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start by reproducing mesh decoder training with the batches that trigger the reported spconv assertion. Inspect the referenced ConvMainUnitTest_matmul_split_Simt_f32f32f32_0.cu assertion and the spconv/cumm version behavior around the int32 range limit. Done means affected batches train without the overflow error, but the issue does not identify a source file or reproduction data.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
machine-learning
Issue type
Bug
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
28/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.