# VIAME training crash

**URL:** https://discourse.kitware.com/t/viame-training-crash/1674
**Category:** VIAME & DIVE
**Created:** [July 2, 2026, 8:26pm UTC](https://discourse.kitware.com/t/viame-training-crash/1674 "2026-07-02T20:26:06Z")
**Posts on this page:** 2
**Page:** 1

<div class="post-metadata">

### Author: ![samloutet](https://discourse.kitware.com/letter_avatar_proxy/v4/letter/s/e9c0ed/32.png) [@samloutet](https://discourse.kitware.com/u/samloutet)
#### Post date: [July 2, 2026, 8:26pm UTC](https://discourse.kitware.com/t/viame-training-crash/1674/1 "2026-07-02T20:26:07Z")

</div>

Hello,

I am trying to run a training on DIVE Desktop and I think that it is crashing before finishing correctly. I am getting some error messages in the run log but I don’t really know how to interpret them or if they are just warnings. The training does produce a detector.pipe but resulting model does not perform well even on the same data that it was trained on, and I cannot open the training folders on DIVE Desktop afterwards without the program crashing.

Here is the [runlog.txt](https://unbcloud-my.sharepoint.com/:t:/g/personal/n74sp_unb_ca/IQAoerMPoi_WQIzs3_4XxJT7AcFsPJ-7xY5KiXNmZctcBjs?e=bwr6At) from the training (let me know if this is accessible).

I’m just wondering what is going on and if there is any way to stop these errors from happening. I am running VIAME v0.22.4 and DIVE Desktop Release 1.9.10. The computer is a Dell Precision 3580 running Windows 11.

Thanks!

---

<div class="post-metadata">

### Author: ![matt\_dawkins](https://discourse.kitware.com/user_avatar/discourse.kitware.com/matt_dawkins/32/523_2.png) [@matt\_dawkins](https://discourse.kitware.com/u/matt_dawkins)
#### Post date: [August 4, 2026, 9:40pm UTC](https://discourse.kitware.com/t/viame-training-crash/1674/2 "2026-08-04T21:40:19Z")

</div>

Sorry for the delayed reply, I don’t check this list as often as the email help list.

This is a Windows MAX\_PATH (260-character) failure, due to trying to make a path that’s too big causing training to fail. Potential workaround:

1. Move the training project to a shorter path (e.g. C:\data\proj) — the simplest fix
2. Or enable Windows long paths: set HKLM\SYSTEM\CurrentControlSet\Control\FileSystem\LongPathsEnabled = 1 (Python3.10 respects this automatically).
