asteroid-team/asteroid issues and pull requests

#707 - Fix a missing super-init in the GEVD beamformer class

Pull Request - State: closed - Opened by ovistetom 25 days ago - 1 comment

#706 - Training and modification

Issue - State: open - Opened by 2624428587 about 1 month ago
Labels: question

#705 - New comes for separation support

Issue - State: open - Opened by MonolithFoundation about 2 months ago
Labels: enhancement, help wanted

#704 - PIT loss returns wrong loss

Issue - State: closed - Opened by RahulDasari 2 months ago - 1 comment
Labels: question

#703 - Use Pretrained Model for multi-channel input

Issue - State: open - Opened by Raphael310 3 months ago - 1 comment
Labels: question

#702 - As We Speak: Real-Time Visually Guided Speaker Separation and Localization

Issue - State: open - Opened by SutirthaChakraborty 4 months ago
Labels: question

#700 - Fix dns_challenge recipe issue forcing INTERSPEECH 2020 branch

Pull Request - State: closed - Opened by Rodolfo-S 7 months ago - 1 comment

#699 - Separation with sound event time stamps

Issue - State: open - Opened by chnk58hoang 7 months ago - 1 comment
Labels: question

#698 - How to use GPU rather than CPU

Issue - State: open - Opened by SoftologyPro 9 months ago - 1 comment
Labels: question

#697 - ConvTasNet pretrained huggingface model inference setup

Issue - State: open - Opened by Rodolfo-S 9 months ago - 1 comment
Labels: question

#696 - Fixed a bug caused by a type error

Pull Request - State: closed - Opened by Xinyi-Cheng 10 months ago - 2 comments

#695 - Bump black from 22.3.0 to 24.3.0 in /requirements

Pull Request - State: open - Opened by dependabot[bot] 11 months ago
Labels: dependencies

#694 - KeyError: 'source_2_path' when training ConvTasNet using enh_single mode

Issue - State: open - Opened by lingzhic 11 months ago - 3 comments
Labels: question

#693 - mossFormer2 & sepTDA models

Issue - State: closed - Opened by jeromew 11 months ago - 2 comments
Labels: enhancement, help wanted

#692 - test model

Issue - State: open - Opened by zhang-chongxiang 11 months ago
Labels: question

#691 - PIT Loss for multichannel audio for speech separation

Issue - State: open - Opened by SutirthaChakraborty 11 months ago - 1 comment
Labels: question

#690 - Support for mobile development (Android)?

Issue - State: closed - Opened by lxaw 12 months ago - 3 comments
Labels: question

#689 - Inference script to hear samples from AVSpeech Model ?

Issue - State: open - Opened by SutirthaChakraborty about 1 year ago
Labels: question

#688 - _reorder_sources() in OLA compares scaled/windowed frame with un-scaled/un-windowed frame

Pull Request - State: closed - Opened by hendriks73 about 1 year ago - 3 comments

#687 - Transfer learning of pretrained model on custom dataset

Issue - State: closed - Opened by thedragoncoder about 1 year ago - 2 comments
Labels: question

#686 - how to train the dataset

Issue - State: closed - Opened by zhang-chongxiang about 1 year ago - 3 comments
Labels: question

#685 - Input and output mean scales updated during training (X-UMX)

Issue - State: closed - Opened by DavidDiazGuerra about 1 year ago - 5 comments

#684 - Quick fix for XUMX in torch 2.0

Pull Request - State: closed - Opened by DavidDiazGuerra about 1 year ago - 2 comments

#683 - RunTimeError related with NCCL when training librimix recipe

Issue - State: open - Opened by garcesote about 1 year ago - 3 comments
Labels: question

#682 - Upgrade dependencies to torch 2.x and lightning 2.x

Pull Request - State: closed - Opened by mpariente over 1 year ago

#679 - KeyError while evaluating musdb18 (X-UMX)

Issue - State: closed - Opened by jerryuhoo over 1 year ago - 2 comments
Labels: question

#678 - LambdaOverlapAdd Param replace window="hanning" with window="hann"

Pull Request - State: closed - Opened by blackhu over 1 year ago - 1 comment

#677 - LambdaOverlapAdd Param replace `window="hanning"` with `window="hann"`

Pull Request - State: closed - Opened by blackhu over 1 year ago

#676 - refer to asteroid/dsp/overlap_add.py:31 LambdaOverlapAdd model use file_separate output audio file is All noise

Issue - State: closed - Opened by blackhu over 1 year ago - 2 comments
Labels: bug, help wanted

#675 - Update overlap_add.py, auto casting device for window

Pull Request - State: closed - Opened by popcornell over 1 year ago

#674 - val loss in distribute training

Issue - State: open - Opened by LiuSiQi-TJ over 1 year ago - 2 comments
Labels: bug, help wanted

#672 - Fix continuous integration

Pull Request - State: closed - Opened by LeonieBorne over 1 year ago - 2 comments

#671 - Limit torch and pytorch-lightning to 1.x

Pull Request - State: closed - Opened by LeonieBorne over 1 year ago - 2 comments

#670 - HOW TO USE OTHER DATASET TO TRAIN Multi-Decoder-DPRNN MODEL

Issue - State: open - Opened by xiaoyu942 over 1 year ago - 3 comments
Labels: question

#669 - torchmetrics>=0.11.0

Issue - State: open - Opened by AceHack over 1 year ago - 2 comments

#668 - Add a colab notebook demo, and a separate script. Support from_pretrained for easier use.

Pull Request - State: closed - Opened by JunzheJosephZhu over 1 year ago - 8 comments

#667 - About the usage

Issue - State: open - Opened by will46616 over 1 year ago - 1 comment

#665 - Add a new pre-trained model (X-UMXL)

Pull Request - State: closed - Opened by r-sawata over 1 year ago - 2 comments

#664 - X-UMX: lr_scheduler_step() missing 1 required positional argument: 'metric'

Issue - State: closed - Opened by DavidDiazGuerra over 1 year ago - 4 comments

#663 - Legacy Install Failure

Issue - State: open - Opened by Mareentier almost 2 years ago - 1 comment
Labels: bug, help wanted

#662 - RuntimeError: istft requires a complex-valued input tensor matching the output from stft with return_complex=True.

Issue - State: open - Opened by lucyanddarlin almost 2 years ago - 4 comments
Labels: question

#661 - why can't I run the ./run.sh ?

Issue - State: closed - Opened by lucyanddarlin almost 2 years ago - 14 comments
Labels: question

#657 - [src] Fix argparse key based on Python version (#628)

Pull Request - State: closed - Opened by jbartolewska almost 2 years ago - 1 comment

#656 - Wrong release for DNS Challenge 2020

Issue - State: open - Opened by jonpodtu almost 2 years ago - 3 comments
Labels: bug, help wanted

#653 - Update Multi-Decoder DPRNN with inference function on a single file out-of-the-box

Pull Request - State: closed - Opened by JunzheJosephZhu about 2 years ago - 2 comments

#651 - How to train with wsj0 dataset

Issue - State: open - Opened by tongyu1 about 2 years ago - 1 comment
Labels: question

#650 - Do you have the pre-trained models for AV-Speech?

Issue - State: closed - Opened by SutirthaChakraborty about 2 years ago - 1 comment
Labels: question

#649 - Fix cumulative normalization

Pull Request - State: closed - Opened by zmolikova about 2 years ago - 1 comment

#648 - Cumulative normalization, wrong variance computation

Issue - State: closed - Opened by zmolikova about 2 years ago - 1 comment

#647 - LibriVAD can not runing from stage 3

Issue - State: closed - Opened by QHenry1990 over 2 years ago - 1 comment
Labels: question

#646 - BaseEncoderMaskerDecoder instance cannot be converted to TorchScript?

Issue - State: open - Opened by urig over 2 years ago - 2 comments
Labels: bug, help wanted

#645 - Upgrade deprecated huggingface_hub.cached_download

Pull Request - State: closed - Opened by mattiadg over 2 years ago - 3 comments

#644 - Warning cached_download deprecated

Issue - State: closed - Opened by mattiadg over 2 years ago - 2 comments
Labels: bug, help wanted

#643 - Missing model.eval() in eval.py scripts

Issue - State: closed - Opened by stachu86 over 2 years ago - 1 comment
Labels: bug, help wanted

#640 - SinkPIT

Issue - State: closed - Opened by xixihahaggg over 2 years ago - 5 comments
Labels: question

#639 - NO asteroid_filterbanks

Issue - State: closed - Opened by nicoyang-21 over 2 years ago

#638 - Update prepare_data.sh

Pull Request - State: closed - Opened by mystlee over 2 years ago

#632 - Create README.md

Pull Request - State: closed - Opened by JunzheJosephZhu over 2 years ago - 1 comment

#631 - PMSQE ：In what range should pmsqe converge?

Issue - State: open - Opened by Fangbo0506 over 2 years ago
Labels: question

#630 - AttributeError: module 'asteroid.filterbanks.transforms' has no attribute 'take_mag'

Issue - State: closed - Opened by jingxuan9862 over 2 years ago - 1 comment
Labels: bug, help wanted

#628 - KeyError: Options in training stage for recipes

Issue - State: open - Opened by joeoct93 over 2 years ago - 8 comments
Labels: bug, help wanted