dotnet / dotnet/machinelearning-modelbuilder
Senario Object Detection Json file (Vott) limit of 5MB
- Dominant language
- Dockerfile
- Stars
- 285
- Forks
- 66
- PR merge metrics
- No merged PRs in 30d
Description
Windows 10, Visual Studio 2022
- Model Builder Version 17.18.2.2415501
- Microsoft Visual Studio Enterprise 2022 (64-bit) - Current Version 17.10.2
**Describe the bug**
There is a file limit of 5 mb that prevents the process moving on. This limit should be much grater in supporting Vott files as it creates lots of metadata and json files. For me a tiny set cannot be read using this as it is limit on the json file and not running the ML.
**To Reproduce**
- On which step of the process did you run into an issue:
1. Add File -> Machine Learning model
2. Choose Senario "Object Detection"
3. Choose Locally(GPU) -> Next Step
4. Choose "vott file"
5. Press "..." button and choose a Json file (Exported from MS Vott) that is bigger then 5Mb
6. Look in the output section, error message:
"Unable to open XXXX.json in the JSON editor because it exceeds the 5 MB limit. Opening as plain text instead."
Stuck in process as next button is grayed out!
**Expected behavior**
A filesize limit that makes sense for object recognition. T.ex using CIFAR-10 or CIFAR-100. Currently cannot run our own generated object set....
**Screenshots**


**Additional context**
Limitations like this should be warnings one can bypass
Contributor guide
No contributing guide indexed for this repository
Assessment
This issue has not been assessed yet.