Implementation of H-Transformer-1D, Hierarchical Attention for Sequence Learning

Overview

H-Transformer-1D

Implementation of H-Transformer-1D, Transformer using hierarchical Attention for sequence learning with subquadratic costs.

For now, the H-Transformer will only act as a long-context encoder

Install

$ pip install h-transformer-1d

Usage

import torch
from h_transformer_1d import HTransformer1D

model = HTransformer1D(
    num_tokens = 256,          # number of tokens
    dim = 512,                 # dimension
    depth = 2,                 # depth
    max_seq_len = 8192,        # maximum sequence length
    heads = 8,                 # heads
    dim_head = 64,             # dimension per head
    block_size = 128           # block size
)

x = torch.randint(0, 256, (1, 8000))   # variable sequence length
mask = torch.ones((1, 8000)).bool()    # variable mask length

# network will automatically pad to power of 2, do hierarchical attention, etc

logits = model(x, mask = mask) # (1, 8000, 256)

Citations

@misc{zhu2021htransformer1d,
    title   = {H-Transformer-1D: Fast One-Dimensional Hierarchical Attention for Sequences}, 
    author  = {Zhenhai Zhu and Radu Soricut},
    year    = {2021},
    eprint  = {2107.11906},
    archivePrefix = {arXiv},
    primaryClass = {cs.LG}
}
Comments
  • Masking not working in training, thanks

    Masking not working in training, thanks

    Hi, I have tried to train the model on GPU with masking enabled. The line 94 t = torch.flip(t, dims = (2,)) reports an error: RuntimeError: "flip_cuda" not implemented for 'Bool', even though I have tried to move mask to CPU.

    Any ideas to solve the problem? Thanks a lot.

    opened by junyongyou 6
  • Application to sequence classification?

    Application to sequence classification?

    Hi,

    Forgive the naive question, I am trying to make sense of this paper but it's tough going. If I understand correctly, this attention mechanism focuses mainly on nearby tokens and only attends to distant tokens via a hierarchical, low-rank approximation. In that case, can the usual sequence classification approach of having a global [CLS] token that can attend to all other tokens (and vice versa) still work? If not, how can this attention mechanism handle the text classification tasks in the long range arena benchmark?

    Cheers for whatever insights you can share, and thanks for the great work!

    opened by trpstra 4
  • eos token does not work in batch mode generation

    eos token does not work in batch mode generation

    When generating the sequence with current code it seems the eos_token will work when generating one sequence at a time https://github.com/lucidrains/h-transformer-1d/blob/main/h_transformer_1d/autoregressive_wrapper.py#L59

    opened by tmphex 4
  • RuntimeError: Tensor type unknown to einops <class 'torch.return_types.max'>

    RuntimeError: Tensor type unknown to einops

    lib/python3.7/site-packages/einops/_backends.py", line 52, in get_backend raise RuntimeError('Tensor type unknown to einops {}'.format(type(tensor)))

    RuntimeError: Tensor type unknown to einops <class 'torch.return_types.max'>

    I understand why this gets raised. Could it be a pytorch version problem? Mine is 1.6.0

    opened by wajihullahbaig 4
  • Algorithm Mismatch

    Algorithm Mismatch

    Paper Implementation

    In the implementation, we get blocked Q, K, V tensors by level with the code below.

    https://github.com/lucidrains/h-transformer-1d/blob/110cab0038898560d72d460bfef8ca8b7f17f0a5/h_transformer_1d/h_transformer_1d.py#L164-L179

    And return the final result of matrix-matrix product with Equation 29 or 69 with the for loop below.

    https://github.com/lucidrains/h-transformer-1d/blob/110cab0038898560d72d460bfef8ca8b7f17f0a5/h_transformer_1d/h_transformer_1d.py#L234-L247

    What is problem?

    However, according to the current code, it is not possible to include information about the level 0 white blocks in the figure below. (Equation 70 of the paper includes the corresponding attention matrix entries.)

    fig2

    I think you should also add an off-diagonal term of near-interaction (level 0) to match Equation 69!

    opened by jinmang2 3
  • I have some questions about implementation details

    I have some questions about implementation details

    Thanks for making your implementation public. I have three questions about your h-transformer 1d implementation.

    1. The number of levels M

    https://github.com/lucidrains/h-transformer-1d/blob/63063d5bb036b56a7205aadc5c8198da02d698f6/h_transformer_1d/h_transformer_1d.py#L105-L114

    In papers, eq (32) gives a guide on how M is determined.

    img1

    In your code implementations,

    • n is sequence length which is not padded
    • bsz is block size (is same to N_r which is numerical rank of the off-diagonal blocks)
    • Because code line 111 already contains level 0, M is equal to int(log2(n // bsz)) - 1

    However, in the Section 6.1 Constructing Hierarchical Attention, I found that sequence length(L) must be a multiple of 2. In my opinion, eq (31)'s L is equal to 2^{M+1}. In implementation, n is not padded sequence. So one of M is missing.

    Since x is the sequence padded by processing below, https://github.com/lucidrains/h-transformer-1d/blob/63063d5bb036b56a7205aadc5c8198da02d698f6/h_transformer_1d/h_transformer_1d.py#L83-L91

    I think the above implementation should be modified as below

    107    num_levels = int(log2(x.size(1) // bsz)) - 1 
    

    2. Super- and Sub-diagonal blocks of the coarsened matrix \tilde{A} as level-l

    https://github.com/lucidrains/h-transformer-1d/blob/63063d5bb036b56a7205aadc5c8198da02d698f6/h_transformer_1d/h_transformer_1d.py#L190-L198

    Ys conatins y and A computed as calculate_Y_and_A. For examples,

    # Ys
    [
        (batch_size*n_heads, N_b(2), N_r), (batch_size*n_heads, N_b(2)),  # level 2, (Y(2), tilde_A(2))
        (batch_size*n_heads, N_b(1), N_r), (batch_size*n_heads, N_b(1)),  # level 1, (Y(1), tilde_A(1))
        (batch_size*n_heads, N_b(0), N_r), (batch_size*n_heads, N_b(0)),  # level 0, (Y(0), tilde_A(0))
    ]
    

    In eq (29), Y is calculated as Y = AV = Y(0) + P(0)( Y(1) + P(1)Y(2) ) However, in code line 190, Y is calculated using only level-0 and level-1 blocks, no matter how many M there are. Y = AV = Y(0) + P(0)Y(1)

    Does increasing the level cause performance degradation issues in implementation? I'm so curious!


    3. Comparison with Luna: Linear Unified Nested Attention

    h-transformer significantly exceeded the scores of BigBird and Luna in LRA. However, what I regretted while reading the paper was that there was no comparison of computation time with other sub-quadratic and Luna. Is this algorithm much faster than other sub-quadratic? And how about compared to Luna?


    Thanks again for the implementation release! The idea of ​​calculating off-diagonal with flip was amazing and I learned a lot. Thank you!! 😄

    opened by jinmang2 3
  • Add Norm Missing

    Add Norm Missing

    I am using code now, and i wonder is there implemented add norm? I only find layer norm, but no add operation. Here is code in h-transformer-1d.py line 489 ... Is this a bug or something ? Thanks @Lucidrains

    for ind in range(depth): attn = attn_class(dim, dim_head = dim_head, heads = heads, block_size = block_size, pos_emb = self.pos_emb, **attn_kwargs) ff = FeedForward(dim, mult = ff_mult)

            if shift_tokens:
                attn, ff = map(lambda t: PreShiftTokens(shift_token_ranges, t), (attn, ff))
    
            attn, ff = map(lambda t: PreNorm(dim, t), (attn, ff))
            layers.append(nn.ModuleList([attn ,ff]))_
    
    opened by wwx13 2
  • Mask not working

    Mask not working

    def forward(self, x, mask = None):
        b, n, device = *x.shape, x.device
        assert n <= self.max_seq_len, 'sequence length must be less than the maximum sequence length'
        x = self.token_emb(x)
        x = self.layers(x)
        return self.to_logits(x)
    

    I think... Masking does not work ???

    opened by wwx13 2
  • One simple question

    One simple question

    Hi, Phil!

    One simple question, (my math is not good) https://github.com/lucidrains/h-transformer-1d/blob/7c11d036d53926495ec0917a34a1aad7261892b5/train.py#L65

    why not be randint(0, self.data.size(0)-self.seq_len+1)? Since the high part should be excluded

    opened by CiaoHe 2
  • Mini-batching (b > 1) does not work with masking

    Mini-batching (b > 1) does not work with masking

    When using x and mask that have batch size larger than 1 following error is arises:

    import torch
    from h_transformer_1d import HTransformer1D
    
    model = HTransformer1D(
        num_tokens = 256,          # number of tokens
        dim = 512,                 # dimension
        depth = 2,                 # depth
        causal = False,            # autoregressive or not
        max_seq_len = 8192,        # maximum sequence length
        heads = 8,                 # heads
        dim_head = 64,             # dimension per head
        block_size = 128           # block size
    )
    
    batch_size = 2
    x = torch.randint(0, 256, (batch_size, 8000))   # variable sequence length
    mask = torch.ones((batch_size, 8000)).bool()    # variable mask length
    
    # network will automatically pad to power of 2, do hierarchical attention, etc
    
    logits = model(x, mask = mask) # (1, 8000, 256)
    

    Gives following error:

    ~/git/h-transformer-1d/h_transformer_1d/h_transformer_1d.py in masked_aggregate(tensor, mask, dim, average)
         19     diff_len = len(tensor.shape) - len(mask.shape)
         20     mask = mask[(..., *((None,) * diff_len))]
    ---> 21     tensor = tensor.masked_fill(~mask, 0.)
         22 
         23     total_el = mask.sum(dim = dim)
    
    RuntimeError: The size of tensor a (2) must match the size of tensor b (16) at non-singleton dimension 0
    

    It seems the tensor has shape heads * batch in 0 dimension and not batch what mask has.

    opened by jaak-s 2
  • Example in README does not work

    Example in README does not work

    Executing the example:

    import torch
    from h_transformer_1d import HTransformer1D
    
    model = HTransformer1D(
        num_tokens = 256,          # number of tokens
        dim = 512,                 # dimension
        depth = 2,                 # depth
        causal = False,            # autoregressive or not
        max_seq_len = 8192,        # maximum sequence length
        heads = 8,                 # heads
        dim_head = 64,             # dimension per head
        block_size = 128           # block size
    )
    
    x = torch.randint(0, 256, (1, 8000))   # variable sequence length
    mask = torch.ones((1, 8000)).bool()    # variable mask length
    
    # network will automatically pad to power of 2, do hierarchical attention, etc
    
    logits = model(x, mask = mask) # (1, 8000, 256)
    

    Gives the following error:

    ~/miniconda3/lib/python3.7/site-packages/rotary_embedding_torch/rotary_embedding_torch.py in apply_rotary_emb(freqs, t, start_index)
         43     assert rot_dim <= t.shape[-1], f'feature dimension {t.shape[-1]} is not of sufficient size to rotate in all the positions {rot_dim}'
         44     t_left, t, t_right = t[..., :start_index], t[..., start_index:end_index], t[..., end_index:]
    ---> 45     t = (t * freqs.cos()) + (rotate_half(t) * freqs.sin())
         46     return torch.cat((t_left, t, t_right), dim = -1)
         47 
    
    RuntimeError: The size of tensor a (8192) must match the size of tensor b (8000) at non-singleton dimension 1
    
    opened by jaak-s 2
  • Fix indexing

    Fix indexing

    I am fixing a few apparent bugs in the code. The upshot is that the attention now supports a block size of the (next largest power of two) of the input length, and for this value of the block size it becomes exact. This allows one to look at the systematic error in the output as a function of decreased block size (and memory usage).

    I've found this module to reduce memory consumption by a factor of two, but the approximation quickly becomes too inaccurate with decreasing block size to use it as a drop-in replacement for an existing (full) attention layer.

    This repository shows how to compute the full attention with linear memory complexity: https://github.com/CHARM-Tx/linear_mem_attention_pytorch

    opened by jglaser 0
  • Approximated values are off

    Approximated values are off

    I wrote a simple test to check the output of the hierarchical transformer self attention against the BERT self attention from huggingface transformers.

    import torch
    import torch.nn as nn
    import math
    
    from h_transformer_1d.h_transformer_1d import HAttention1D
    
    def transpose_for_scores(x, num_attention_heads, attention_head_size):
        new_x_shape = x.size()[:-1] + (num_attention_heads, attention_head_size)
        x = x.view(*new_x_shape)
        return x.permute(0, 2, 1, 3)
    
    def bert_self_attention(query, key, value_layer, attention_mask=None, num_attention_heads=1):
            dim_head = query.size()[-1] // num_attention_heads
            all_head_size = dim_head*num_attention_heads
    
            query_layer = transpose_for_scores(query, num_attention_heads, dim_head)
            key_layer = transpose_for_scores(key, num_attention_heads, dim_head)
            value_layer = transpose_for_scores(value, num_attention_heads, dim_head)
    
            # Take the dot product between "query" and "key" to get the raw attention scores.
            attention_scores = torch.matmul(query_layer, key_layer.transpose(-1, -2))
    
            attention_scores = attention_scores / math.sqrt(dim_head)
    
            if attention_mask is not None:
                # Apply the attention mask is (precomputed for all layers in BertModel forward() function)
                attention_scores = attention_scores + attention_mask
    
            # Normalize the attention scores to probabilities.
            attention_probs = nn.functional.softmax(attention_scores, dim=-1)
    
            # This is actually dropping out entire tokens to attend to, which might
            # seem a bit unusual, but is taken from the original Transformer paper.
            #attention_probs = self.dropout(attention_probs)
    
            context_layer = torch.matmul(attention_probs, value_layer)
    
            context_layer = context_layer.permute(0, 2, 1, 3).contiguous()
            new_context_layer_shape = context_layer.size()[:-2] + (all_head_size,)
            context_layer = context_layer.view(*new_context_layer_shape)
    
            return context_layer, attention_probs
    
    if __name__ == "__main__":
        query = torch.tensor([[[0.1,0.2],[-0.5,0.7],[-0.5,-0.75],[.123,.456]]])
    #    query = torch.tensor([[[0.1,0.2],[-0.5,0.7]]])
        key = value = query
    
        n_heads = 1
        attn, probs = bert_self_attention(query, key, value, num_attention_heads=n_heads)
        print('bert_self_attention out: ', attn)
    
        block_size = 1
        for _ in range(0,2):
            dim_head = query.size()[-1]//n_heads
            h_attn = HAttention1D(
                dim=query.size()[-1],
                heads=n_heads,
                dim_head=dim_head,
                block_size=block_size
            )
    
            h_attn.to_qkv = torch.nn.Identity()
            h_attn.to_out = torch.nn.Identity()
    
            qkv = torch.stack([query, key, value], dim=2)
            qkv = torch.flatten(qkv, start_dim=2)
    
            attn_scores = h_attn(qkv)
            print('hattention_1d: (block_size = {})'.format(block_size), attn_scores)
    
            block_size *= 2
    

    This is the output I get

    bert_self_attention:  tensor([[[-0.1807,  0.1959],
             [-0.2096,  0.2772],
             [-0.2656, -0.0568],
             [-0.1725,  0.2442]]])
    hattention_1d: (block_size = 1) tensor([[[-0.2000,  0.4500],
             [-0.2000,  0.4500],
             [-0.1885, -0.1470],
             [-0.1885, -0.1470]]])
    

    before it errors out with

    assert num_levels >= 0, 'number of levels must be at least greater than 0'
    

    Some of the values are off in absolute magnitude by more than a factor of two.

    Looking at the code, this line seems problematic: https://github.com/lucidrains/h-transformer-1d/blob/8afd75cc6bc41754620bb6ab3737176cb69bdf93/h_transformer_1d/h_transformer_1d.py#L172

    I believe it should read

    num_levels = int(log2(pad_to_len // bsz)) - 1
    

    If I make that change, the approximated attention output is much closer to the exact one:

    bert_self_attention out:  tensor([[[-0.1807,  0.1959],
             [-0.2096,  0.2772],
             [-0.2656, -0.0568],
             [-0.1725,  0.2442]]])
    hattention_1d: (block_size = 1) tensor([[[-0.2590,  0.2020],
             [-0.2590,  0.2020],
             [-0.2590,  0.2020],
             [-0.2590,  0.2020]]])
    hattention_1d: (block_size = 2) tensor([[[-0.1808,  0.1972],
             [-0.1980,  0.2314],
             [-0.2438,  0.0910],
             [-0.1719,  0.2413]]])
    
    opened by jglaser 1
Owner
Phil Wang
Working with Attention. It's all we need
Phil Wang
Pytorch-Swin-Unet-V2 - a modified version of Swin Unet based on Swin Transfomer V2

Swin Unet V2 Swin Unet V2 is a modified version of Swin Unet arxiv based on Swin

Chenxu Peng 26 Dec 03, 2022
This repo contains the code for the paper "Efficient hierarchical Bayesian inference for spatio-temporal regression models in neuroimaging" that has been accepted to NeurIPS 2021.

Dugh-NeurIPS-2021 This repo contains the code for the paper "Efficient hierarchical Bayesian inference for spatio-temporal regression models in neuroi

Ali Hashemi 5 Jul 12, 2022
Official pytorch implement for “Transformer-Based Source-Free Domain Adaptation”

Official implementation for TransDA Official pytorch implement for “Transformer-Based Source-Free Domain Adaptation”. Overview: Result: Prerequisites:

stanley 54 Dec 22, 2022
Implementation of Basic Machine Learning Algorithms on small datasets using Scikit Learn.

Basic Machine Learning Algorithms All the basic Machine Learning Algorithms are implemented in Python using libraries Acknowledgements Machine Learnin

Piyal Banik 47 Oct 16, 2022
Implementation of the paper Scalable Intervention Target Estimation in Linear Models (NeurIPS 2021), and the code to generate simulation results.

Scalable Intervention Target Estimation in Linear Models Implementation of the paper Scalable Intervention Target Estimation in Linear Models (NeurIPS

0 Oct 25, 2021
A PyTorch Implementation of "SINE: Scalable Incomplete Network Embedding" (ICDM 2018).

Scalable Incomplete Network Embedding ⠀⠀ A PyTorch implementation of Scalable Incomplete Network Embedding (ICDM 2018). Abstract Attributed network em

Benedek Rozemberczki 69 Sep 22, 2022
A list of Machine Learning Art Colabs

ML Visual Art Colabs A list of cool Colabs on Machine Learning Imagemaking or other artistic purposes 3D Ken Burns Effect Ken Burns Effect by Manuel R

Derrick Schultz (he/him) 789 Dec 12, 2022
Paper: Cross-View Kernel Similarity Metric Learning Using Pairwise Constraints for Person Re-identification

Cross-View Kernel Similarity Metric Learning Using Pairwise Constraints for Person Re-identification T M Feroz Ali, Subhasis Chaudhuri, ICVGIP-20-21

T M Feroz Ali 3 Jun 17, 2022
WSDM‘2022: Knowledge Enhanced Sports Game Summarization

Knowledge Enhanced Sports Game Summarization Cooming Soon! :) Data will be released after approval process. Code will be published once the author of

Jiaan Wang 14 Jul 13, 2022
Unofficial implementation of "TTNet: Real-time temporal and spatial video analysis of table tennis" (CVPR 2020)

TTNet-Pytorch The implementation for the paper "TTNet: Real-time temporal and spatial video analysis of table tennis" An introduction of the project c

Nguyen Mau Dung 438 Dec 29, 2022
TigerLily: Finding drug interactions in silico with the Graph.

Drug Interaction Prediction with Tigerlily Documentation | Example Notebook | Youtube Video | Project Report Tigerlily is a TigerGraph based system de

Benedek Rozemberczki 91 Dec 30, 2022
MERLOT: Multimodal Neural Script Knowledge Models

merlot MERLOT: Multimodal Neural Script Knowledge Models MERLOT is a model for learning what we are calling "neural script knowledge" -- representatio

Rowan Zellers 190 Dec 22, 2022
Neural Network Libraries

Neural Network Libraries Neural Network Libraries is a deep learning framework that is intended to be used for research, development and production. W

Sony 2.6k Dec 30, 2022
Code for Piggyback: Adapting a Single Network to Multiple Tasks by Learning to Mask Weights

Piggyback: https://arxiv.org/abs/1801.06519 Pretrained masks and backbones are available here: https://uofi.box.com/s/c5kixsvtrghu9yj51yb1oe853ltdfz4q

Arun Mallya 165 Nov 22, 2022
Old Photo Restoration (Official PyTorch Implementation)

Bringing Old Photo Back to Life (CVPR 2020 oral)

Microsoft 11.3k Dec 30, 2022
Data and codes for ACL 2021 paper: Towards Emotional Support Dialog Systems

Emotional-Support-Conversation Copyright © 2021 CoAI Group, Tsinghua University. All rights reserved. Data and codes are for academic research use onl

126 Dec 21, 2022
Automatic number plate recognition using tech: Yolo, OCR, Scene text detection, scene text recognation, flask, torch

Automatic Number Plate Recognition Automatic Number Plate Recognition (ANPR) is the process of reading the characters on the plate with various optica

Meftun AKARSU 52 Dec 22, 2022
An NLP library with Awesome pre-trained Transformer models and easy-to-use interface, supporting wide-range of NLP tasks from research to industrial applications.

简体中文 | English News [2021-10-12] PaddleNLP 2.1版本已发布!新增开箱即用的NLP任务能力、Prompt Tuning应用示例与生成任务的高性能推理! 🎉 更多详细升级信息请查看Release Note。 [2021-08-22]《千言:面向事实一致性的生

6.9k Jan 01, 2023
Code for Towards Streaming Perception (ECCV 2020) :car:

sAP — Code for Towards Streaming Perception ECCV Best Paper Honorable Mention Award Feb 2021: Announcing the Streaming Perception Challenge (CVPR 2021

Martin Li 85 Dec 22, 2022
[CVPR'21] Multi-Modal Fusion Transformer for End-to-End Autonomous Driving

TransFuser This repository contains the code for the CVPR 2021 paper Multi-Modal Fusion Transformer for End-to-End Autonomous Driving. If you find our

695 Jan 05, 2023