Pervasive Attention: 2D Convolutional Networks for Sequence-to-Sequence Prediction

Overview

This is a fork of Fairseq(-py) with implementations of the following models:

Pervasive Attention - 2D Convolutional Neural Networks for Sequence-to-Sequence Prediction

An NMT models with two-dimensional convolutions to jointly encode the source and the target sequences.

Pervasive Attention also provides an extensive decoding grid that we leverage to efficiently train wait-k models.

See README.

Efficient Wait-k Models for Simultaneous Machine Translation

Transformer Wait-k models (Ma et al., 2019) with unidirectional encoders and with joint training of multiple wait-k paths.

See README.

Fairseq Requirements and Installation

  • PyTorch version >= 1.4.0
  • Python version >= 3.6
  • For training new models, you'll also need an NVIDIA GPU and NCCL

Installing Fairseq

git clone https://github.com/elbayadm/attn2d
cd attn2d
pip install --editable .

License

fairseq(-py) is MIT-licensed. The license applies to the pre-trained models as well.

Citation

For Pervasive Attention, please cite:

@InProceedings{elbayad18conll,
    author ="Elbayad, Maha and Besacier, Laurent and Verbeek, Jakob",
    title = "Pervasive Attention: 2D Convolutional Neural Networks for Sequence-to-Sequence Prediction",
    booktitle = "Proceedings of the 22nd Conference on Computational Natural Language Learning",
    year = "2018",
 }

For our wait-k models, please cite:

@article{elbayad20waitk,
    title={Efficient Wait-k Models for Simultaneous Machine Translation},
    author={Elbayad, Maha and Besacier, Laurent and Verbeek, Jakob},
    journal={arXiv preprint arXiv:2005.08595},
    year={2020}
}

For Fairseq, please cite:

@inproceedings{ott2019fairseq,
  title = {fairseq: A Fast, Extensible Toolkit for Sequence Modeling},
  author = {Myle Ott and Sergey Edunov and Alexei Baevski and Angela Fan and Sam Gross and Nathan Ng and David Grangier and Michael Auli},
  booktitle = {Proceedings of NAACL-HLT 2019: Demonstrations},
  year = {2019},
}
Comments
  • The training speed of pervasive attention model

    The training speed of pervasive attention model

    Hello,

    I am trying to run the pervasive attention model recipes as recommended in the README: https://github.com/elbayadm/attn2d/blob/master/examples/pervasive/README.md

    Based on my observation, the model is too slow with almost 30 words per second on a single GPU 1080! May you please give me a rough idea of what should be the speed expectations based on your experiments?

    Thanks for your insights Parnia

    opened by papar22 1
  • AttributeError: 'float' object has no attribute 'data' on pytorch 0.4.1

    AttributeError: 'float' object has no attribute 'data' on pytorch 0.4.1

    I get the following error at trainer.backward_step() on running the demo script: File "attn2d/nmt/trainer.py", line 230, in backward_step self.clip_norm).data.item() AttributeError: 'float' object has no attribute 'data'

    which I fixed by: grad_norm = torch.nn.utils.clip_grad_norm_(self.model.parameters(), self.clip_norm)

    Which then runs OK for torch==0.4.1

    Which version of pytorch was the code written for?

    opened by adriangrepo 1
  • ModuleNotFoundError: No module named 'nmt.loader.pair_loader'

    ModuleNotFoundError: No module named 'nmt.loader.pair_loader'

    Hi, as title says, I got this error:

    Traceback (most recent call last): File "/home//repos/attn2d/train.py", line 84, in train(params) File "/home//repos/attn2d/train.py", line 20, in train from nmt.trainer import Trainer File "/home//repos/attn2d/nmt/trainer.py", line 17, in from nmt.loader.pair_loader import DataPair ModuleNotFoundError: No module named 'nmt.loader.pair_loader'

    opened by koszilard 1
  • 16 undefined names

    16 undefined names

    • Replace ‘false’ with ‘False’
    • Missing import tensorflow as tf
    • Missing import numpy as np
    • See #2

    flake8 testing of https://github.com/elbayadm/attn2d on Python 3.7.0

    $ flake8 . --count --select=E901,E999,F821,F822,F823 --show-source --statistics

    ./nmt/scheduler.py:53:44: F821 undefined name 'exp'
                                               exp(self._iter / self.speed))
                                               ^
    ./nmt/optimizer.py:202:16: F821 undefined name 'nn'
            return nn.utils.clip_grad_norm_(params, self.grad_norm_max, self.grad_norm_type)
                   ^
    ./nmt/models/pooling.py:145:15: F821 undefined name 'PositionalPooling4'
            super(PositionalPooling4, self).__init__()
                  ^
    ./nmt/models/convs2s2D.py:523:20: F821 undefined name 'trg_emb'
                    if trg_emb.size(1) > max_h:
                       ^
    ./nmt/models/convs2s2D.py:524:31: F821 undefined name 'trg_emb'
                        trg_emb = trg_emb[:, -max_h:, :, :]
                                  ^
    ./nmt/models/convs2s2D.py:624:41: F821 undefined name 'trg_labels'
                    trg_labels = torch.cat((trg_labels, trg_labels_t),
                                            ^
    ./nmt/utils/logging.py:17:23: F821 undefined name 'DETOK'
        source = " ".join(DETOK.detokenize(source.split())).encode('utf-8')
                          ^
    ./nmt/utils/logging.py:18:19: F821 undefined name 'DETOK'
        gt = " ".join(DETOK.detokenize(gt.split())).encode('utf-8')
                      ^
    ./nmt/utils/logging.py:19:21: F821 undefined name 'DETOK'
        pred = " ".join(DETOK.detokenize(pred.split())).encode('utf-8')
                        ^
    ./nmt/utils/logging.py:129:16: F821 undefined name 'tf'
        _summary = tf.summary.scalar(name=key,
                   ^
    ./nmt/utils/logging.py:130:41: F821 undefined name 'tf'
                                     tensor=tf.Variable(value),
                                            ^
    ./nmt/utils/logging.py:132:15: F821 undefined name 'tf'
        summary = tf.Summary(value=[tf.Summary.Value(tag=key, simple_value=value)])
                  ^
    ./nmt/utils/logging.py:132:33: F821 undefined name 'tf'
        summary = tf.Summary(value=[tf.Summary.Value(tag=key, simple_value=value)])
                                    ^
    ./nmt/loss/working_loss.py:565:31: F821 undefined name 'false'
                p.requires_grad = false
                                  ^
    ./nmt/loss/_loss.py:402:31: F821 undefined name 'false'
                p.requires_grad = false
                                  ^
    ./nmt/loss/samplers/ngram.py:72:52: F821 undefined name 'score'
            return preds_matrix, np.ones(batch_size) * score, stats
                                                       ^
    16    F821 undefined name 'false'
    16
    
    opened by cclauss 1
  • Is this a typo?

    Is this a typo?

    at attn2d/nmt/models/pooling.py line 145.

    class PositionalPooling(nn.Module):
        def __init__(self, max_length, emb_size):
            super(PositionalPooling4, self).__init__()
            self.src_embedding = nn.Embedding(max_length, emb_size)
            self.trg_embedding = nn.Embedding(max_length, emb_size)
            self.src_embedding.weight.data.fill_(1)
            self.trg_embedding.weight.data.fill_(1)
            self.src_embedding.bias.data.fill_(0)
            self.trg_embedding.bias.data.fill_(0)
    

    Is PositionalPooling4 PositionalPooling?

    opened by elect000 1
  • Training/Eval Error for waitk model

    Training/Eval Error for waitk model

    🐛 Bug

    I'am trying to run the trainning code follow the waitk guide file , and fixed some bug just as @ereday this issue mentioned , but still got error when i ran the train code :

    RuntimeError: Output 0 of SplitBackward0 is a view and is being modified inplace. This view is the output of a function that returns multiple views. Such functions do not allow the output views to be modified inplace. You should replace the inplace operation by an out-of-place one.

    To Reproduce

    Steps to reproduce the behavior (always include the command you ran):

    k=7
    MODEL=tf_wait${k}_wmt14Ende
    CUDA_VISIBLE_DEVICES=0 python train.py $DATA_BIN -s en -t de --left-pad-source False \
        --user-dir examples/waitk --arch waitk_transformer_small \
        --save-dir $Workdir/checkpoints/$MODEL --tensorboard-logdir $Workdir/logs/$MODEL \
        --seed 1 --no-epoch-checkpoints --no-progress-bar --log-interval 10  \
        --optimizer adam --adam-betas '(0.9, 0.98)' --weight-decay 0.0001 \
        --max-tokens 4000 --update-freq 2 --max-update 50000 \
        --lr-scheduler inverse_sqrt --warmup-updates 4000 --warmup-init-lr '1e-07' --lr 0.002 \
        --min-lr '1e-9' --criterion label_smoothed_cross_entropy --label-smoothing 0.1 \
        --share-decoder-input-output-embed --waitk  $k
    
    1. See error

    Expected behavior

    Environment

    • fairseq Version (e.g., 1.0 or master):
    • PyTorch Version (e.g., 1.0)
    • OS (e.g., Linux):
    • How you installed fairseq (pip, source):
    • Build command you used (if compiling from source):
    • Python version:
    • CUDA/cuDNN version:
    • GPU models and configuration:
    • Any other relevant information:

    Additional context

    this repo seems out-of-date and the issue raised half years ago is still no replied.

    bug 
    opened by EricLina 0
  • ModuleNotFoundError: No module named 'examples.simultaneous'

    ModuleNotFoundError: No module named 'examples.simultaneous'

    🐛 Bug

    Hi,

    I was trying to evaluate the pre-trained models under "Efficient Wait-k Models for Simultaneous Machine Translation". For this, I followed the instructions given in the readme. Specifically, I did followings:

    After downloading model and data and placing them underpre_saved:

    cd ~/attn2d/pre_saved
    tar xzf iwslt14_de_en.tar.gz
    tar xzf tf_waitk_model.tar.gz
    
    
    k=5 # Evaluation time k
    output=wait$k.log
    CUDA_VISIBLE_DEVICES=0 python generate.py pre_saved/iwslt14_deen_bpe10k_binaries/ -s de -t en --gen-subset test --path pre_saved/tf_waitk_model.tar.gz --task waitk_translation --eval-waitk $k --model-overrides "{'max_source_positions': 1024, 'max_target_positions': 1024}" --left-pad-source False --user-dir examples/waitk --no-progress-bar --max-tokens 8000 --remove-bpe --beam 1 2>&1 | tee -a $output
    

    It generates following error message:

    Traceback (most recent call last):
      File "generate.py", line 11, in <module>
        cli_main()
      File "/home/attn2d/fairseq_cli/generate.py", line 276, in cli_main
        parser = options.get_generation_parser()
      File "/home/attn2d/fairseq/options.py", line 33, in get_generation_parser
        parser = get_parser("Generation", default_task)
      File "/home/attn2d/fairseq/options.py", line 197, in get_parser
        utils.import_user_module(usr_args)
      File "/home/attn2d/fairseq/utils.py", line 350, in import_user_module
        importlib.import_module(module_name)
      File "/home/anaconda3/envs/py37/lib/python3.7/importlib/__init__.py", line 127, in import_module
        return _bootstrap._gcd_import(name[level:], package, level)
      File "<frozen importlib._bootstrap>", line 1006, in _gcd_import
      File "<frozen importlib._bootstrap>", line 983, in _find_and_load
      File "<frozen importlib._bootstrap>", line 967, in _find_and_load_unlocked
      File "<frozen importlib._bootstrap>", line 677, in _load_unlocked
      File "<frozen importlib._bootstrap_external>", line 728, in exec_module
      File "<frozen importlib._bootstrap>", line 219, in _call_with_frames_removed
      File "/home/attn2d/examples/waitk/__init__.py", line 1, in <module>
        from . import models, tasks
      File "/home/attn2d/examples/waitk/models/__init__.py", line 7, in <module>
        importlib.import_module('examples.simultaneous.models.' + model_name)
      File "/home/anaconda3/envs/py37/lib/python3.7/importlib/__init__.py", line 127, in import_module
        return _bootstrap._gcd_import(name[level:], package, level)
    ModuleNotFoundError: No module named 'examples.simultaneous'
    

    EDIT

    Okay, here is more detail about this:

    I believe thisline is responsible for the error message shared above.

    I changed this importlib.import_module('examples.simultaneous.models.' + model_name) to importlib.import_module('examples.waitk.models.' + model_name)

    Then, I got another error:

      File "generate.py", line 11, in <module>
        cli_main()
      File "/home/attn2d/fairseq_cli/generate.py", line 276, in cli_main
        parser = options.get_generation_parser()
      File "/home/attn2d/fairseq/options.py", line 33, in get_generation_parser
        parser = get_parser("Generation", default_task)
      File "/home/attn2d/fairseq/options.py", line 197, in get_parser
        utils.import_user_module(usr_args)
      File "/home/attn2d/fairseq/utils.py", line 350, in import_user_module
        importlib.import_module(module_name)
      File "/home/anaconda3/lib/python3.7/importlib/__init__.py", line 127, in import_module
        return _bootstrap._gcd_import(name[level:], package, level)
      File "<frozen importlib._bootstrap>", line 1006, in _gcd_import
      File "<frozen importlib._bootstrap>", line 983, in _find_and_load
      File "<frozen importlib._bootstrap>", line 967, in _find_and_load_unlocked
      File "<frozen importlib._bootstrap>", line 677, in _load_unlocked
      File "<frozen importlib._bootstrap_external>", line 728, in exec_module
      File "<frozen importlib._bootstrap>", line 219, in _call_with_frames_removed
      File "/home/attn2d/examples/waitk/__init__.py", line 1, in <module>
        from . import models, tasks
      File "/home/attn2d/examples/waitk/models/__init__.py", line 8, in <module>
        importlib.import_module('examples.waitk.models.' + model_name)
      File "/home/anaconda3/lib/python3.7/importlib/__init__.py", line 127, in import_module
        return _bootstrap._gcd_import(name[level:], package, level)
      File "/home/attn2d/examples/waitk/__init__.py", line 1, in <module>
        from . import models, tasks
      File "/home/attn2d/examples/waitk/models/__init__.py", line 8, in <module>
        importlib.import_module('examples.waitk.models.' + model_name)
      File "/home/anaconda3/lib/python3.7/importlib/__init__.py", line 127, in import_module
        return _bootstrap._gcd_import(name[level:], package, level)
      File "/home/attn2d/examples/waitk/models/waitk_transformer.py", line 24, in <module>
        from examples.simultaneous.modules import TransformerEncoderLayer, TransformerDecoderLayer
    
    

    So, I changed this line here to ```from examples.waitk.modules import TransformerEncoderLayer, ```` too. Then when I tried once more, I got the following error:

      File "generate.py", line 11, in <module>
        cli_main()
      File "/home/attn2d/fairseq_cli/generate.py", line 276, in cli_main
        parser = options.get_generation_parser()
      File "/home/attn2d/fairseq/options.py", line 33, in get_generation_parser
        parser = get_parser("Generation", default_task)
      File "/home/attn2d/fairseq/options.py", line 197, in get_parser
        utils.import_user_module(usr_args)
      File "/home/attn2d/fairseq/utils.py", line 350, in import_user_module
        importlib.import_module(module_name)
      File "/home/anaconda3/lib/python3.7/importlib/__init__.py", line 127, in import_module
        return _bootstrap._gcd_import(name[level:], package, level)
      File "<frozen importlib._bootstrap>", line 1006, in _gcd_import
      File "<frozen importlib._bootstrap>", line 983, in _find_and_load
      File "<frozen importlib._bootstrap>", line 967, in _find_and_load_unlocked
      File "<frozen importlib._bootstrap>", line 677, in _load_unlocked
      File "<frozen importlib._bootstrap_external>", line 728, in exec_module
      File "<frozen importlib._bootstrap>", line 219, in _call_with_frames_removed
      File "/home/attn2d/examples/waitk/__init__.py", line 1, in <module>
        from . import models, tasks
      File "/home/attn2d/examples/waitk/models/__init__.py", line 8, in <module>
        importlib.import_module('examples.waitk.models.' + model_name)
      File "/home/anaconda3/lib/python3.7/importlib/__init__.py", line 127, in import_module
        return _bootstrap._gcd_import(name[level:], package, level)
      File "/home/attn2d/examples/waitk/__init__.py", line 1, in <module>
        from . import models, tasks
      File "/home/attn2d/examples/waitk/models/__init__.py", line 8, in <module>
        importlib.import_module('examples.waitk.models.' + model_name)
      File "/home/anaconda3/lib/python3.7/importlib/__init__.py", line 127, in import_module
        return _bootstrap._gcd_import(name[level:], package, level)
      File "/home/attn2d/examples/waitk/models/waitk_transformer.py", line 25, in <module>
        from examples.waitk.modules import TransformerEncoderLayer, TransformerDecoderLayer
      File "/home/attn2d/examples/waitk/modules/__init__.py", line 2, in <module>
        from .controller import Controller
    

    So, to fix it I commented out following lines inexamples/waitk/modules/init.py:

    from .controller import Controller
    from .branch_controller import BranchController
    from .oracle import SimulTransOracleDP, SimulTransOracleDP1
    

    Next, I've tried to use the generation command given in the readme once more..

    CUDA_VISIBLE_DEVICES=0 python generate.py pretrained-sources/iwslt14_deen_bpe10k_binaries/ -s de -t en --gen-subset test --path pretrained-sources/model.pt --task waitk_translation --eval-waitk $k --model-overrides "{'max_source_positions': 1024, 'max_target_positions': 1024}" --left-pad-source False --user-dir examples/waitk --no-progress-bar --max-tokens 8000 --remove-bpe --beam 1 2>&1 | tee -a $output

    I got this error:

    
    2021-09-20 20:29:46 | INFO | fairseq_cli.generate | Namespace(all_gather_list_size=16384, beam=1, bpe=None, checkpoint_suffix='', cpu=False, criterion='cross_entropy', data='pretrained-sources/iwslt14_deen_bpe10k_binaries/', data_buffer_size=0, dataset_impl=None, decoding_format=None, diverse_beam_groups=-1, diverse_beam_strength=0.5, diversity_rate=-1.0, empty_cache_freq=0, eval_bleu=False, eval_bleu_args=None, eval_bleu_detok='space', eval_bleu_detok_args=None, eval_bleu_print_samples=False, eval_bleu_remove_bpe=None, eval_tokenized_bleu=False, eval_waitk=5, force_anneal=None, fp16=False, fp16_init_scale=128, fp16_no_flatten_grads=False, fp16_scale_tolerance=0.0, fp16_scale_window=None, gen_subset='test', iter_decode_eos_penalty=0.0, iter_decode_force_max_iter=False, iter_decode_max_iter=10, iter_decode_with_beam=1, iter_decode_with_external_reranker=False, left_pad_source='False', left_pad_target='False', lenpen=1, load_alignments=False, log_format=None, log_interval=100, lr_scheduler='fixed', lr_shrink=0.1, match_source_len=False, max_len_a=0, max_len_b=200, max_sentences=None, max_source_positions=1024, max_target_positions=1024, max_tokens=8000, memory_efficient_fp16=False, min_len=1, min_loss_scale=0.0001, model_overrides="{'max_source_positions': 1024, 'max_target_positions': 1024}", model_parallel_size=1, momentum=0.99, nbest=1, no_beamable_mm=False, no_early_stop=False, no_progress_bar=True, no_repeat_ngram_size=0, num_shards=1, num_workers=1, optimizer='nag', path='pretrained-sources/model.pt', prefix_size=0, print_alignment=False, print_step=False, quantization_config_path=None, quiet=False, remove_bpe='@@ ', replace_unk=None, required_batch_size_multiple=8, results_path=None, retain_iter_history=False, sacrebleu=False, sampling=False, sampling_topk=-1, sampling_topp=-1.0, score_reference=False, seed=1, shard_id=0, skip_invalid_size_inputs_valid_test=False, source_lang='de', target_lang='en', task='waitk_translation', temperature=1.0, tensorboard_logdir='', threshold_loss_scale=None, tokenizer=None, truncate_source=False, unkpen=0, unnormalized=False, upsample_primary=1, user_dir='examples/waitk', warmup_updates=0, weight_decay=0.0)
    2021-09-20 20:29:46 | INFO | fairseq.tasks.translation | [de] dictionary: 8848 types
    2021-09-20 20:29:46 | INFO | fairseq.tasks.translation | [en] dictionary: 6632 types
    2021-09-20 20:29:46 | INFO | fairseq.data.data_utils | loaded 6750 examples from: pretrained-sources/iwslt14_deen_bpe10k_binaries/test.de-en.de
    2021-09-20 20:29:46 | INFO | fairseq.data.data_utils | loaded 6750 examples from: pretrained-sources/iwslt14_deen_bpe10k_binaries/test.de-en.en
    2021-09-20 20:29:46 | INFO | fairseq.tasks.translation | pretrained-sources/iwslt14_deen_bpe10k_binaries/ test de-en 6750 examples
    2021-09-20 20:29:46 | INFO | fairseq_cli.generate | loading model(s) from pretrained-sources/model.pt
    Traceback (most recent call last):
      File "generate.py", line 11, in <module>
        cli_main()
      File "/home/attn2d/fairseq_cli/generate.py", line 278, in cli_main
        main(args)
      File "/home/attn2d/fairseq_cli/generate.py", line 36, in main
        return _main(args, sys.stdout)
      File "/home/attn2d/fairseq_cli/generate.py", line 103, in _main
        num_workers=args.num_workers,
      File "/home/attn2d/fairseq/tasks/fairseq_task.py", line 181, in get_batch_iterator
        required_batch_size_multiple=required_batch_size_multiple,
      File "/home/attn2d/fairseq/data/data_utils.py", line 220, in batch_by_size
        from fairseq.data.data_utils_fast import batch_by_size_fast
      File "fairseq/data/data_utils_fast.pyx", line 1, in init fairseq.data.data_utils_fast
        # cython: language_level=3
    ValueError: numpy.ndarray size changed, may indicate binary incompatibility. Expected 88 from C header, got 80 from PyObject
    ##########
    

    I just gave up after that.. @elbayadm I hope you can help me on this.

    Code sample

    Environment

    I have followed the instructions in the README to install my environment. :

    git clone https://github.com/elbayadm/attn2d
    cd attn2d
    pip install --editable .
    
    

    As a result, I have the following libraries in my environment:

    Python 3.7.10 | packaged by conda-forge | (default, Feb 19 2021, 16:07:37) 
    [GCC 9.3.0] on linux
    Type "help", "copyright", "credits" or "license" for more information.
    >>> import torch
    torc>>> torch.__version__
    '1.9.0+cu102'
    >>> import fairseq
    >>> fairseq.__version__
    '0.9.0'
    >>> 
    $ python --version
    Python 3.7.10
    

    Operating system: Linux

    bug 
    opened by ereday 2
Owner
Maha
PhD student (LIG & INRIA)
Maha
Sequence modeling benchmarks and temporal convolutional networks

Sequence Modeling Benchmarks and Temporal Convolutional Networks (TCN) This repository contains the experiments done in the work An Empirical Evaluati

CMU Locus Lab 3.5k Jan 1, 2023
Unofficial TensorFlow implementation of Protein Interface Prediction using Graph Convolutional Networks.

[TensorFlow] Protein Interface Prediction using Graph Convolutional Networks Unofficial TensorFlow implementation of Protein Interface Prediction usin

YeongHyeon Park 9 Oct 25, 2022
MVGCN: a novel multi-view graph convolutional network (MVGCN) framework for link prediction in biomedical bipartite networks.

MVGCN MVGCN: a novel multi-view graph convolutional network (MVGCN) framework for link prediction in biomedical bipartite networks. Developer: Fu Hait

null 13 Dec 1, 2022
Code for our ICASSP 2021 paper: SA-Net: Shuffle Attention for Deep Convolutional Neural Networks

SA-Net: Shuffle Attention for Deep Convolutional Neural Networks (paper) By Qing-Long Zhang and Yu-Bin Yang [State Key Laboratory for Novel Software T

Qing-Long Zhang 199 Jan 8, 2023
Improving Convolutional Networks via Attention Transfer (ICLR 2017)

Attention Transfer PyTorch code for "Paying More Attention to Attention: Improving the Performance of Convolutional Neural Networks via Attention Tran

Sergey Zagoruyko 1.4k Dec 23, 2022
Codes for TIM2021 paper "Anchor-Based Spatio-Temporal Attention 3-D Convolutional Networks for Dynamic 3-D Point Cloud Sequences"

Codes for TIM2021 paper "Anchor-Based Spatio-Temporal Attention 3-D Convolutional Networks for Dynamic 3-D Point Cloud Sequences"

Intelligent Robotics and Machine Vision Lab 4 Jul 19, 2022
CoSMA: Convolutional Semi-Regular Mesh Autoencoder. From Paper "Mesh Convolutional Autoencoder for Semi-Regular Meshes of Different Sizes"

Mesh Convolutional Autoencoder for Semi-Regular Meshes of Different Sizes Implementation of CoSMA: Convolutional Semi-Regular Mesh Autoencoder arXiv p

Fraunhofer SCAI 10 Oct 11, 2022
Rethinking Semantic Segmentation from a Sequence-to-Sequence Perspective with Transformers

Segmentation Transformer Implementation of Segmentation Transformer in PyTorch, a new model to achieve SOTA in semantic segmentation while using trans

Abhay Gupta 161 Dec 8, 2022
Implementation of SETR model, Original paper: Rethinking Semantic Segmentation from a Sequence-to-Sequence Perspective with Transformers.

SETR - Pytorch Since the original paper (Rethinking Semantic Segmentation from a Sequence-to-Sequence Perspective with Transformers.) has no official

zhaohu xing 112 Dec 16, 2022
Understanding and Improving Encoder Layer Fusion in Sequence-to-Sequence Learning (ICLR 2021)

Understanding and Improving Encoder Layer Fusion in Sequence-to-Sequence Learning (ICLR 2021) Citation Please cite as: @inproceedings{liu2020understan

Sunbow Liu 22 Nov 25, 2022
[CVPR 2021] Rethinking Semantic Segmentation from a Sequence-to-Sequence Perspective with Transformers

[CVPR 2021] Rethinking Semantic Segmentation from a Sequence-to-Sequence Perspective with Transformers

Fudan Zhang Vision Group 897 Jan 5, 2023
Sequence to Sequence Models with PyTorch

Sequence to Sequence models with PyTorch This repository contains implementations of Sequence to Sequence (Seq2Seq) models in PyTorch At present it ha

Sandeep Subramanian 708 Dec 19, 2022
Sequence-to-Sequence learning using PyTorch

Seq2Seq in PyTorch This is a complete suite for training sequence-to-sequence models in PyTorch. It consists of several models and code to both train

Elad Hoffer 514 Nov 17, 2022
An implementation of a sequence to sequence neural network using an encoder-decoder

Keras implementation of a sequence to sequence model for time series prediction using an encoder-decoder architecture. I created this post to share a

Luke Tonin 195 Dec 17, 2022
Sequence lineage information extracted from RKI sequence data repo

Pango lineage information for German SARS-CoV-2 sequences This repository contains a join of the metadata and pango lineage tables of all German SARS-

Cornelius Roemer 24 Oct 26, 2022
Official repository of OFA. Paper: Unifying Architectures, Tasks, and Modalities Through a Simple Sequence-to-Sequence Learning Framework

Paper | Blog OFA is a unified multimodal pretrained model that unifies modalities (i.e., cross-modality, vision, language) and tasks (e.g., image gene

OFA Sys 1.4k Jan 8, 2023
TalkNet 2: Non-Autoregressive Depth-Wise Separable Convolutional Model for Speech Synthesis with Explicit Pitch and Duration Prediction.

TalkNet 2 [WIP] TalkNet 2: Non-Autoregressive Depth-Wise Separable Convolutional Model for Speech Synthesis with Explicit Pitch and Duration Predictio

Rishikesh (ऋषिकेश) 69 Dec 17, 2022
Cancer Drug Response Prediction via a Hybrid Graph Convolutional Network

DeepCDR Cancer Drug Response Prediction via a Hybrid Graph Convolutional Network This work has been accepted to ECCB2020 and was also published in the

Qiao Liu 50 Dec 18, 2022
Price-Prediction-For-a-Dream-Home - A machine learning based linear regression trained model for house price prediction.

Price-Prediction-For-a-Dream-Home ROADMAP TO THIS LINEAR REGRESSION BASED HOUSE PRICE PREDICTION PREDICTION MODEL Import all the dependencies of the p

DIKSHA DESWAL 1 Dec 29, 2021