Repository to host and maintain scale-sim-v2 code

An integrated cache and memory access time, cycle time, area, leakage, and dynamic power model

Reorder-based post-training quantization for large language model

Using ideas from product quantization for state-of-the-art neural network compression.

Fast and accurate DRAM power and energy estimation tool

New generation entropy codecs : Finite State Entropy and Huff0

The official implementation of the EMNLP 2023 paper LLM-FP4

🤗 Transformers: State-of-the-art Machine Learning for Pytorch, TensorFlow, and JAX.

Code accompanying the paper "Massive Activations in Large Language Models"

This is a collection of our zero-cost NAS and efficient vision applications.

Post-Training Quantization for Vision transformers.

[ICML'21 Oral] I-BERT: Integer-only BERT Quantization

[ICML 2024 Oral] Any-Precision LLM: Low-Cost Deployment of Multiple, Different-Sized LLMs

A paper list of some recent Transformer-based CV works.

Code for "Atalanta: A Bit is Worth a ``Thousand'' Tensor Values"

[IJCAI 2022] FQ-ViT: Post-Training Quantization for Fully Quantized Vision Transformer

The largest collection of PyTorch image encoders / backbones. Including train, eval, inference, export scripts, and pretrained weights -- ResNet, ResNeXT, EfficientNet, NFNet, Vision Transformer (V…

[ICCV 2023] RepQ-ViT: Scale Reparameterization for Post-Training Quantization of Vision Transformers

A script for US visa appointments in Canada

Simulator for BitFusion

[MLSys 2024 Best Paper Award] AWQ: Activation-aware Weight Quantization for LLM Compression and Acceleration

A bit-level sparsity-awared multiply-accumulate process element.

A visualization and transformation of pytorch model

PDPU: An Open-Source Posit Dot-Product Unit for Deep Learning Applications