11 31 56

youssef boulaouane

byoussef

Boulaouaney

AI & ML interests

None yet

Recent Activity

Reacted to rwightman's post with 🚀 10 days ago

Want to validate some hparams or figure out what `timm` model to use before commiting to download or training with a large dataset? Try mini-imagenet: https://huggingface.co/datasets/timm/mini-imagenet I had this sitting on my drive and forgot where I pulled it together from. It's 100 classes of imagenet, 50k train and 10k val images (from ImageNet-1k train set), and 5k test images (from ImageNet-1k val set). 7.4GB instead of > 100GB for the full ImageNet-1k. This ver is not reduced resolution like some other 'mini' versions. Super easy to use with timm train/val scripts, checkout the dataset card. I often check fine-tuning with even smaller datasets like: * https://huggingface.co/datasets/timm/resisc45 * https://huggingface.co/datasets/timm/oxford-iiit-pet But those are a bit small to train any modest size model w/o starting from pretrained weights.

upvoted a collection 15 days ago

Nov 15 Releases 🍂

upvoted a collection 20 days ago

OpenCoder

View all activity

Organizations

Posts 1

Post

924

On-Device deployment ready MobileNet-V4 models:
byoussef/mobilenetv4-conv-pretrained-tflite-models-66726c945040fc61f5a50b63

Models converted from timm's pretrained Pytorch weights to TFLite (fp32 & fp16).
All conv models fully compatible with TFLite Gpu Delegate on Android for fast on-device inference 🚀
Can also run super fast on Google-Edge-TPU on Pixel phones 🚀

Collections 1

spaces 1

Running

👀

MobileNetV4 TFLite

models 10

datasets

None public yet