Memory Efficient Attention (O(sqrt(n)) for Jax and PyTorch

Overview

Memory Efficient Attention

arXiv PyPI version

This is unofficial implementation of Self-attention Does Not Need O(n^2) Memory for Jax and PyTorch.

Implementation is almost same as the one proposed in the paper, with additional masking and adding bias compatibility, batch dimensions support and PyTorch implementation. For computing attention, the proposed method requires only O(sqrt(n)) memory, and the provided functions can be used as a drop-in replacement for attention calculation.

Important Note: This implementation is a trade-off between memory requirements and runtime, so you should adjust key_chunk_size and query_chunk_size parameters to achieve the best configuration for your usecase. Here is a note from the paper's authors:

While a constant chunk size for the queries and a chunk size of sqrt(n) for the keys and values is optimal for memory consumption, the runtime is also affected by the choice of chunk size in practice, which is heavily affected by the choice of hardware. Ultimately, we have to leave this trade-off to the programmer, and expose the chunk sizes as arguments query_chunk_size and key_chunk_size. In Figure 1 we provide default values for the chunk sizes that lead to minimal runtime impact (on TPUv2), while still providing significant memory savings.

Quick Start

  1. Install the library
# for Jax
pip install memory-efficient-attention[jax]
# for PyTorch
pip install memory-efficient-attention[torch]
# for Running Tests
pip install memory-efficient-attention[testing]
  1. Compute attention with the proper function
0.5 bias = np.random.rand(1, b, 16, 128, 128).astype("float32") / 100 # Adjust chunk sizes efficient_dot_product_attention_jax(query, key, value, mask, bias, key_chunk_size=..., query_chunk_size=...)">
import numpy as np
# for PyTorch
from memory_efficient_attention import efficient_dot_product_attention_pt
# or for Jax
from memory_efficient_attention import efficient_dot_product_attention_jax

# Random Data (batch dimensions are not necessary)
b = 8
query = np.random.rand(1, b, 128, 16, 8).astype("float32")
key = np.random.rand(1, b, 128, 16, 8).astype("float32")
value = np.random.rand(1, b, 128, 16, 8).astype("float32")
# optional, for casual tasks, ...
mask = np.random.rand(1, b, 16, 128, 128) > 0.5
bias = np.random.rand(1, b, 16, 128, 128).astype("float32") / 100

# Adjust chunk sizes        
efficient_dot_product_attention_jax(query, key, value, mask, bias, key_chunk_size=..., query_chunk_size=...)

Citation

Please cite if this implementation helps your research. You can use the following BibTeX entry:

@misc{memory_efficient_attention,
	title = {Memory Efficient Attention},
	author = {Rezaei, Amin},
	howpublished = {\url{github.com/AminRezaei0x443/memory-efficient-attention}},
	year = {2021}
}

Also, for the paper:

@misc{rabe2021selfattention,
      title={Self-attention Does Not Need $O(n^2)$ Memory}, 
      author={Markus N. Rabe and Charles Staats},
      year={2021},
      eprint={2112.05682},
      archivePrefix={arXiv},
      primaryClass={cs.LG}
}
You might also like...
Memory-efficient optimum einsum using opt_einsum planning and PyTorch kernels.

opt-einsum-torch There have been many implementations of Einstein's summation. numpy's numpy.einsum is the least efficient one as it only runs in sing

A human-readable PyTorch implementation of "Self-attention Does Not Need O(n^2) Memory"

memory_efficient_attention.pytorch A human-readable PyTorch implementation of "Self-attention Does Not Need O(n^2) Memory" (Rabe&Staats'21). def effic

 Attention for PyTorch with Linear Memory Footprint
Attention for PyTorch with Linear Memory Footprint

Attention for PyTorch with Linear Memory Footprint Unofficially implements https://arxiv.org/abs/2112.05682 to get Linear Memory Cost on Attention (+

Local Attention - Flax module for Jax

Local Attention - Flax Autoregressive Local Attention - Flax module for Jax Install $ pip install local-attention-flax Usage from jax import random fr

Reimplementation of the paper "Attention, Learn to Solve Routing Problems!" in jax/flax.

JAX + Attention Learn To Solve Routing Problems Reinplementation of the paper Attention, Learn to Solve Routing Problems! using Jax and Flax. Fully su

Official and maintained implementation of the paper
Official and maintained implementation of the paper "OSS-Net: Memory Efficient High Resolution Semantic Segmentation of 3D Medical Data" [BMVC 2021].

OSS-Net: Memory Efficient High Resolution Semantic Segmentation of 3D Medical Data Christoph Reich, Tim Prangemeier, Özdemir Cetin & Heinz Koeppl | Pr

Lowest memory consumption and second shortest runtime in NTIRE 2022 challenge on Efficient Super-Resolution

FMEN Lowest memory consumption and second shortest runtime in NTIRE 2022 on Efficient Super-Resolution. Our paper: Fast and Memory-Efficient Network T

Implementation of
Implementation of "Efficient Regional Memory Network for Video Object Segmentation" (Xie et al., CVPR 2021).

RMNet This repository contains the source code for the paper Efficient Regional Memory Network for Video Object Segmentation. Cite this work @inprocee

Rethinking Space-Time Networks with Improved Memory Coverage for Efficient Video Object Segmentation
Rethinking Space-Time Networks with Improved Memory Coverage for Efficient Video Object Segmentation

STCN Rethinking Space-Time Networks with Improved Memory Coverage for Efficient Video Object Segmentation Ho Kei Cheng, Yu-Wing Tai, Chi-Keung Tang [a

Comments
  • feat: output_attentions

    feat: output_attentions

    I'm looking into hacking some of the models in the transformers library to use this library for attention, and I don't see a way to support output_attentions yet. This is a flag passed in transformers, where the attention weights are preserved and returned to the user, if it is set.

    I looked a little at implementing this in the torch backend, and I note the scan() function provides for only a single tensor return value. It seems to me that scan() function would be most clearly replaced by a for loop, but it could also be modified to handle tuples, or return_weights could be handled via accessing nonlocal data in some way instead of returning them through the chunk scanner. I'm also not sure how the output would best be passed to the user.

    Edit: Draft implementation 01/28 at https://github.com/AminRezaei0x443/memory-efficient-attention/compare/main...xloem:faba6371ac7faaa2040a2c26e15ae7ab87f94ce4 . I ended up extending the scan function for parity between implementations. Edit 2: Turns out it's the postsoftmax attention weights, not the presoftmax attention weights. I've updated this post and the draft implementation for this output: https://github.com/AminRezaei0x443/memory-efficient-attention/compare/main...xloem:return_weights

    opened by xloem 4
  • Provide a flag for the user to receive attention weights

    Provide a flag for the user to receive attention weights

    This is my draft code for #1. I saw this feature in the transformers library and wanted to implement it here.

    I'm curious what you think about this feature and implementation.

    The code is simply slightly instrumented so that the final attention weights can be returned to the user. Tests are augmented to test this use. In utils, the scan function is expanded to handle tuples.

    A change to dynamic_slice crept in from dev, to use slices rather than index_slice. I've retained this change because it looks like it would execute faster to me, but it can be removed.

    Rebased and squashed from 84724e1de4721ea0333d6bdbb91e8bce74fbeac .

    opened by xloem 2
  • Improve performance via batched-matmul and fused multiplies

    Improve performance via batched-matmul and fused multiplies

    Many thanks for providing this reference implementation.

    I tried integrating this into stable-diffusion / diffusers. A fix was required to make it work on Mac (PyTorch MPS backend):
    https://github.com/Birch-san/diffusers/pull/1/commits/04372140a25d7f53549175f1f196599c3e9bf3a5

    Knowing that computing attention via baddbmm()+bmm() can outperform einsum by 18%: I tried to rewrite the algorithm to use those.

    I compared the speed of my optimized version, against the implementation in this repository. this result is for "everything fits in one chunk" perf (i.e. chunk size = max token length). I was unable to compare chunked perf, because although I got chunking working in my version: I wasn't able to get it working in the version in this repository (got some unexpected-shape tensors returned).

    compared to the implementation in this repository:
    my optimized version achieves a 2.78x speedup in the time it took to generate a 512x512 image with stable-diffusion v2.1-base (i.e. 4096 vision tokens, 5 attention heads, batch size of 2 due to CFG).

    here's my optimized implementation:
    https://github.com/Birch-san/diffusers/pull/1

    batched matmuls require a 3D tensor, i.e. [batch * num_heads, tokens, channels_per_head].

    code that currently integrates agains this repository's [batch, q_length, num_heads, qk_depth_per_head] format can migrate those tensors to the [batch * num_heads, q_length, channels_per_head] format favoured by my implementation like so:

    query = query.transpose(1,2).flatten(end_dim=1)
    key = key.transpose(1,2).flatten(end_dim=1)
    value = value.transpose(1,2).flatten(end_dim=1)
    

    the result that's returned, remains in [batch * num_heads, q_length, qk_depth_per_head] format, and can be restored to [batch, q_length, num_heads, qk_depth_per_head] format like so:

    result.unflatten(0, (-1, attn.heads)).transpose(1,2)
    

    I think a further speedup is possible too: by working out when chunking is not needed: we can compute whether unchunked attention would fit into memory, and prefer unchunked attention as a fast-path where possible. this will be useful in a Unet, which runs attention at various resolutions.

    EDIT:
    I have now added fast-paths for:

    • skipping kv-chunking when kv_chunk_size >= k_tokens
      • this turns the algorithm into "attention slicing"
    • skipping q-chunking when q_chunk_size >= q_tokens
    • skipping all chunking when the kv_chunk_size >= k_tokens and q_chunk_size >= q_tokens
    • skipping all chunking when the [email protected] matmul requires fewer bytes than a user-provided threshold
    opened by Birch-san 1
Releases(0.1.3)
  • 0.1.2(Mar 7, 2022)

    What's Changed

    This update fixes torch device handling issues in code. GPU and other kinds of tensors can be used safely.

    • Update utils.py by @yhgon in https://github.com/AminRezaei0x443/memory-efficient-attention/pull/5
    • Update attention_torch.py by @yhgon in https://github.com/AminRezaei0x443/memory-efficient-attention/pull/6

    New Contributors

    • @yhgon made their first contribution in https://github.com/AminRezaei0x443/memory-efficient-attention/pull/5

    Full Changelog: https://github.com/AminRezaei0x443/memory-efficient-attention/compare/0.1.1.0...0.1.2

    Source code(tar.gz)
    Source code(zip)
  • 0.1.1.0(Feb 3, 2022)

    Added mask, bias calculation functions for custom and memory efficient chunks computation. So now sublinear memory computation mask, bias are possible.

    Full Changelog: https://github.com/AminRezaei0x443/memory-efficient-attention/compare/0.1.1...0.1.1.0

    Source code(tar.gz)
    Source code(zip)
Owner
Amin Rezaei
Computer Science BSc, Neural Networks Enthusiast
Amin Rezaei
Source code for "Interactive All-Hex Meshing via Cuboid Decomposition [SIGGRAPH Asia 2021]".

Interactive All-Hex Meshing via Cuboid Decomposition Video demonstration This repository contains an interactive software to the PolyCube-based hex-me

Lingxiao Li 131 Dec 05, 2022
A PyTorch implementation of " EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks."

EfficientNet A PyTorch implementation of EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks. [arxiv] [Official TF Repo] Implemen

AhnDW 298 Dec 10, 2022
A PyTorch library for Vision Transformers

VFormer A PyTorch library for Vision Transformers Getting Started Read the contributing guidelines in CONTRIBUTING.rst to learn how to start contribut

Society for Artificial Intelligence and Deep Learning 142 Nov 28, 2022
This repository contains the data and code for the paper "Diverse Text Generation via Variational Encoder-Decoder Models with Gaussian Process Priors" ([email protected])

GP-VAE This repository provides datasets and code for preprocessing, training and testing models for the paper: Diverse Text Generation via Variationa

Wanyu Du 18 Dec 29, 2022
Deep motion generator collections

GenMotion GenMotion (/gen’motion/) is a Python library for making skeletal animations. It enables easy dataset loading and experiment sharing for synt

23 May 24, 2022
A Python toolbox to create adversarial examples that fool neural networks in PyTorch, TensorFlow, and JAX

Foolbox Native: Fast adversarial attacks to benchmark the robustness of machine learning models in PyTorch, TensorFlow, and JAX Foolbox is a Python li

Bethge Lab 2.4k Dec 25, 2022
A stock generator that assess a list of stocks and returns the best stocks for investing and money allocations based on users choices of volatility, duration and number of stocks

Stock-Generator Please visit "Stock Generator.ipynb" for a clearer view and "Stock Generator.py" for scripts. The stock generator is designed to allow

jmengnyay 1 Aug 02, 2022
YolactEdge: Real-time Instance Segmentation on the Edge

YolactEdge, the first competitive instance segmentation approach that runs on small edge devices at real-time speeds. Specifically, YolactEdge runs at up to 30.8 FPS on a Jetson AGX Xavier (and 172.7

Haotian Liu 1.1k Jan 06, 2023
[ICCV21] Code for RetrievalFuse: Neural 3D Scene Reconstruction with a Database

RetrievalFuse Paper | Project Page | Video RetrievalFuse: Neural 3D Scene Reconstruction with a Database Yawar Siddiqui, Justus Thies, Fangchang Ma, Q

Yawar Nihal Siddiqui 75 Dec 22, 2022
Vikrant Deshpande 1 Nov 17, 2022
Visualizing Yolov5's layers using GradCam

YOLO-V5 GRADCAM I constantly desired to know to which part of an object the object-detection models pay more attention. So I searched for it, but I di

Pooya Mohammadi Kazaj 200 Jan 01, 2023
Training Certifiably Robust Neural Networks with Efficient Local Lipschitz Bounds (Local-Lip)

Training Certifiably Robust Neural Networks with Efficient Local Lipschitz Bounds (Local-Lip) Introduction TL;DR: We propose an efficient and trainabl

17 Dec 01, 2022
CUP-DNN is a deep neural network model used to predict tissues of origin for cancers of unknown of primary.

CUP-DNN CUP-DNN is a deep neural network model used to predict tissues of origin for cancers of unknown of primary. The model was trained on the expre

1 Oct 27, 2021
Awesome Transformers in Medical Imaging

This repo supplements our Survey on Transformers in Medical Imaging Fahad Shamshad, Salman Khan, Syed Waqas Zamir, Muhammad Haris Khan, Munawar Hayat,

Fahad Shamshad 666 Jan 06, 2023
ExCon: Explanation-driven Supervised Contrastive Learning

ExCon: Explanation-driven Supervised Contrastive Learning Contributors of this repo: Zhibo Zhang ( Zhibo (Darren) Zhang 18 Nov 01, 2022

🔎 Super-scale your images and run experiments with Residual Dense and Adversarial Networks.

Image Super-Resolution (ISR) The goal of this project is to upscale and improve the quality of low resolution images. This project contains Keras impl

idealo 4k Jan 08, 2023
CVPR 2021 Official Pytorch Code for UC2: Universal Cross-lingual Cross-modal Vision-and-Language Pre-training

UC2 UC2: Universal Cross-lingual Cross-modal Vision-and-Language Pre-training Mingyang Zhou, Luowei Zhou, Shuohang Wang, Yu Cheng, Linjie Li, Zhou Yu,

Mingyang Zhou 28 Dec 30, 2022
Source codes of CenterTrack++ in 2021 ICME Workshop on Big Surveillance Data Processing and Analysis

MOT Tracked object bounding box association (CenterTrack++) New association method based on CenterTrack. Two new branches (Tracked Size and IOU) are a

36 Oct 04, 2022
Package to compute Mauve, a similarity score between neural text and human text. Install with `pip install mauve-text`.

MAUVE MAUVE is a library built on PyTorch and HuggingFace Transformers to measure the gap between neural text and human text with the eponymous MAUVE

Krishna Pillutla 182 Jan 02, 2023