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
An architecture that makes any doodle realistic, in any specified style, using VQGAN, CLIP and some basic embedding arithmetics.

Sketch Simulator An architecture that makes any doodle realistic, in any specified style, using VQGAN, CLIP and some basic embedding arithmetics. See

12 Dec 18, 2022
Code for "Multi-View Multi-Person 3D Pose Estimation with Plane Sweep Stereo"

Multi-View Multi-Person 3D Pose Estimation with Plane Sweep Stereo This repository includes the source code for our CVPR 2021 paper on multi-view mult

Jiahao Lin 66 Jan 04, 2023
Using deep learning model to detect breast cancer.

Breast-Cancer-Detection Breast cancer is the most frequent cancer among women, with around one in every 19 women at risk. The number of cases of breas

1 Feb 13, 2022
Locally Differentially Private Distributed Deep Learning via Knowledge Distillation (LDP-DL)

Locally Differentially Private Distributed Deep Learning via Knowledge Distillation (LDP-DL) A preprint version of our paper: Link here This is a samp

Di Zhuang 3 Jan 08, 2023
Offline Multi-Agent Reinforcement Learning Implementations: Solving Overcooked Game with Data-Driven Method

Overcooked-AI We suppose to apply traditional offline reinforcement learning technique to multi-agent algorithm. In this repository, we implemented be

Baek In-Chang 14 Sep 16, 2022
The code of NeurIPS 2021 paper "Scalable Rule-Based Representation Learning for Interpretable Classification".

Rule-based Representation Learner This is a PyTorch implementation of Rule-based Representation Learner (RRL) as described in NeurIPS 2021 paper: Scal

Zhuo Wang 53 Dec 17, 2022
Non-Homogeneous Poisson Process Intensity Modeling and Estimation using Measure Transport

Non-Homogeneous Poisson Process Intensity Modeling and Estimation using Measure Transport This GitHub page provides code for reproducing the results i

Andrew Zammit Mangion 1 Nov 08, 2021
Linescanning - Package for (pre)processing of anatomical and (linescanning) fMRI data

line scanning repository This repository contains all of the tools used during the acquisition and postprocessing of line scanning data at the Spinoza

Jurjen Heij 4 Sep 14, 2022
Deep Reinforcement Learning for mobile robot navigation in ROS Gazebo simulator

DRL-robot-navigation Deep Reinforcement Learning for mobile robot navigation in ROS Gazebo simulator. Using Twin Delayed Deep Deterministic Policy Gra

87 Jan 07, 2023
This project is based on RIFE and aims to make RIFE more practical for users by adding various features and design new models

CPM 项目描述 CPM(Chinese Pretrained Models)模型是北京智源人工智能研究院和清华大学发布的中文大规模预训练模型。官方发布了三种规模的模型,参数量分别为109M、334M、2.6B,用户需申请与通过审核,方可下载。 由于原项目需要考虑大模型的训练和使用,需要安装较为复杂

hzwer 190 Jan 08, 2023
Back to Basics: Efficient Network Compression via IMP

Back to Basics: Efficient Network Compression via IMP Authors: Max Zimmer, Christoph Spiegel, Sebastian Pokutta This repository contains the code to r

IOL Lab @ ZIB 1 Nov 19, 2021
PRIME: A Few Primitives Can Boost Robustness to Common Corruptions

PRIME: A Few Primitives Can Boost Robustness to Common Corruptions This is the official repository of PRIME, the data agumentation method introduced i

Apostolos Modas 34 Oct 30, 2022
Simple tools for logging and visualizing, loading and training

TNT TNT is a library providing powerful dataloading, logging and visualization utilities for Python. It is closely integrated with PyTorch and is desi

1.5k Jan 02, 2023
🌎 The Modern Declarative Data Flow Framework for the AI Empowered Generation.

🌎 JSONClasses JSONClasses is a declarative data flow pipeline and data graph framework. Official Website: https://www.jsonclasses.com Official Docume

Fillmula Inc. 53 Dec 09, 2022
PyTorch implementation of SimCLR: A Simple Framework for Contrastive Learning of Visual Representations

PyTorch implementation of SimCLR: A Simple Framework for Contrastive Learning of Visual Representations

Thalles Silva 1.7k Dec 28, 2022
Deep Learning as a Cloud API Service.

Deep API Deep Learning as Cloud APIs. This project provides pre-trained deep learning models as a cloud API service. A web interface is available as w

Wu Han 4 Jan 06, 2023
An Evaluation of Generative Adversarial Networks for Collaborative Filtering.

An Evaluation of Generative Adversarial Networks for Collaborative Filtering. This repository was developed by Fernando B. Pérez Maurera. Fernando is

Fernando Benjamín PÉREZ MAURERA 0 Jan 19, 2022
Computer vision - fun segmentation experience using classic and deep tools :)

Computer_Vision_Segmentation_Fun Segmentation of Images and Video. Tools: pytorch Models: Classic model - GrabCut Deep model - Deeplabv3_resnet101 Flo

Mor Ventura 1 Dec 18, 2021
A large-scale video dataset for the training and evaluation of 3D human pose estimation models

ASPset-510 ASPset-510 (Australian Sports Pose Dataset) is a large-scale video dataset for the training and evaluation of 3D human pose estimation mode

Aiden Nibali 36 Oct 30, 2022
code for "AttentiveNAS Improving Neural Architecture Search via Attentive Sampling"

code for "AttentiveNAS Improving Neural Architecture Search via Attentive Sampling"

Facebook Research 94 Oct 26, 2022