Pinned Repositories
adadim-llm
[ICLR 2024] Rethinking Channel Dimensions to Isolate Outliers for Low-bit Weight Quantization of Large Language Models
craft
Compression Aware Fine-Tuning
cuda-accel-vit
[EE451] Using CUDA to Accelerate Multi-head Attention in Vision Transformers
dalle-pictionary
Don't want to draw but still want to play pictionary? Play with DALL-E!
DeepPiCar
Deep Learning Autonomous Car based on Raspberry Pi, SunFounder PiCar-V Kit, TensorFlow, and Google's EdgeTPU Co-Processor
fast-compress-vit
Official Pytorch Implementation of "A Fast Training-free Compression Framework for Vision Transformers"
MLTherapy
[2021 USC IEEE Hack IoT] Machine Learning-guided Music & Light Therapy
neurologue.ai
[2021 USC Undergraduate SCymposium] Dementia Detection with Multimodal (Speech & Text) ML
semdedup
[EE499] Semantic Deduplication for Data Efficient Learning
iccad-tinyml-open
[ICCAD'22 TinyML Contest] Efficient Heart Stroke Detection on Low-cost Microcontrollers
johnheo's Repositories
johnheo/cuda-accel-vit
[EE451] Using CUDA to Accelerate Multi-head Attention in Vision Transformers
johnheo/fast-compress-vit
Official Pytorch Implementation of "A Fast Training-free Compression Framework for Vision Transformers"
johnheo/semdedup
[EE499] Semantic Deduplication for Data Efficient Learning
johnheo/adadim-llm
[ICLR 2024] Rethinking Channel Dimensions to Isolate Outliers for Low-bit Weight Quantization of Large Language Models
johnheo/craft
Compression Aware Fine-Tuning
johnheo/dalle-pictionary
Don't want to draw but still want to play pictionary? Play with DALL-E!
johnheo/DeepPiCar
Deep Learning Autonomous Car based on Raspberry Pi, SunFounder PiCar-V Kit, TensorFlow, and Google's EdgeTPU Co-Processor
johnheo/ee499_ml_spring23
johnheo/MLTherapy
[2021 USC IEEE Hack IoT] Machine Learning-guided Music & Light Therapy
johnheo/neurologue.ai
[2021 USC Undergraduate SCymposium] Dementia Detection with Multimodal (Speech & Text) ML
johnheo/johnheo
Config files for my GitHub profile.
johnheo/numlock-ee354
johnheo/ToMe
A method to increase the speed and lower the memory footprint of existing vision transformers.