# Multiple time fine-tuning VideoMAE model adding n class each time

**URL:** <https://discuss.huggingface.co/t/multiple-time-fine-tuning-videomae-model-adding-n-class-each-time/83589>\
**Category:** 🤗Transformers\
**Created:** [April 24, 2024, 6:55pm UTC](https://discuss.huggingface.co/t/multiple-time-fine-tuning-videomae-model-adding-n-class-each-time/83589 "2024-04-24T18:55:11Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![latif98](https://avatars.discourse-cdn.com/v4/letter/l/cdc98d/32.png) [@latif98](https://discuss.huggingface.co/u/latif98)\
**Post date:** [April 24, 2024, 6:55pm UTC](https://discuss.huggingface.co/t/multiple-time-fine-tuning-videomae-model-adding-n-class-each-time/83589/1 "2024-04-24T18:55:11Z")

</div>

Hello Everyone, I have got a video dataset of sign language, which has got multiple sub folders of numbers, alphabets, nouns, verbs and adjectives. I have fine-tuned the VideoMAE model on numbers videos only, which gave me a good accuracy. Moreover, my task is to fine-tune the VideoMAE model on all dataset which is relatively huge with respect to the computing facilities I have. How should I proceed to fine-tune the VideoMAE model on all dataset ?  
fine-tuning the model on classes of videos and having only one model at the end to work as classifier for my dataset.

Any help would be appreciated.