ceos/unsloth
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

fix: always force-uncheck trainOnCompletions for pure audio models in dataset check

Browse source 
Separate pure-audio from audio-VLM logic in runDatasetCheck so pure
audio models are always forced to trainOnCompletions=false regardless
of dataset type, while audio VLMs (gemma3n) only uncheck when the
dataset is audio.
This commit is contained in:
Roland Tannous 2026-03-10 19:02:49 +00:00
commit 846cc2cf2a
1 changed files with 6 additions and 1 deletions
Download patch file Download diff file Expand all files Collapse all files

7
studio/frontend/src/features/training/stores/training-config-store.ts
View file

@ -213,7 +213,12 @@ export const useTrainingConfigStore = create<TrainingConfigStore>()(
if (isVisionModel && isImage) {
updates.trainOnCompletions = false;
}
if (isAudioModel && isAudio) {
// Pure audio model → always uncheck regardless of dataset.
if (isAudioModel && !isVisionModel) {
updates.trainOnCompletions = false;
}
// Audio-capable vision model (e.g. gemma3n) + audio dataset → uncheck.
if (isAudioModel && isVisionModel && isAudio) {
updates.trainOnCompletions = false;
}
}