Ecosyste.ms: Issues
An open API service for providing issue and pull request metadata for open source projects.
GitHub / speechbrain/speechbrain issues and pull requests
#2717 - Incorrect check on concat original start index
Pull Request -
State: closed - Opened by gfdb 4 months ago
- 1 comment
Labels: bug
#2716 - Ensure adapted models save their parameters
Pull Request -
State: closed - Opened by pplantinga 4 months ago
- 1 comment
#2715 - `AdaptedModel` does not save weights
Issue -
State: closed - Opened by pplantinga 4 months ago
Labels: bug
#2714 - Gammatone learnable filterbanks
Pull Request -
State: open - Opened by naspert 4 months ago
- 6 comments
#2713 - High Training Loss and WER in Transducer Recipe on LibriSpeech 960hrs Dataset
Issue -
State: open - Opened by nutan235 4 months ago
- 12 comments
Labels: bug
#2712 - Fetch take two
Pull Request -
State: closed - Opened by pplantinga 4 months ago
- 2 comments
#2711 - Fetch Semantics
Issue -
State: closed - Opened by pplantinga 5 months ago
- 1 comment
Labels: bug
#2710 - Harmonize Adapters and MultiWOZ w/ QLoRA
Issue -
State: open - Opened by pplantinga 5 months ago
- 1 comment
Labels: bug
#2708 - Fix: Update broken link to Colab preprocessing tutorial
Pull Request -
State: closed - Opened by shivamtawari 5 months ago
#2707 - Add adapter tutorial
Pull Request -
State: closed - Opened by pplantinga 5 months ago
- 2 comments
Labels: documentation, enhancement, ready to review
#2706 - Consistently fails after running for a while. Tried WAV MP3 formats.
Issue -
State: open - Opened by phalexo 5 months ago
- 2 comments
Labels: bug
#2705 - Regarding the vocoder, after we finish training it, how can we perform inference?
Issue -
State: open - Opened by a897456 5 months ago
Labels: bug
#2702 - Fix template inference links
Pull Request -
State: closed - Opened by pplantinga 5 months ago
- 3 comments
#2701 - Template inference script broken
Issue -
State: closed - Opened by pplantinga 5 months ago
- 6 comments
Labels: bug
#2700 - Streaming ASR guide/tutorial
Pull Request -
State: closed - Opened by asumagic 5 months ago
- 4 comments
Labels: documentation
#2699 - how to load everything from local path?
Issue -
State: closed - Opened by littlebowlnju 5 months ago
- 2 comments
Labels: bug
#2698 - Write tutorials for Adapters
Issue -
State: closed - Opened by asumagic 5 months ago
Labels: bug
#2696 - Tokotron: Tokenized TTS
Pull Request -
State: open - Opened by flexthink 5 months ago
Labels: enhancement
#2695 - Inference audio normalizer changes and use `load_audio` in more places
Pull Request -
State: open - Opened by asumagic 5 months ago
#2693 - Documentation updates and reorganization
Pull Request -
State: closed - Opened by asumagic 5 months ago
- 2 comments
Labels: documentation
#2692 - URL fixes and URL check fixes
Pull Request -
State: closed - Opened by asumagic 5 months ago
- 3 comments
#2689 - Voice analysis functions
Pull Request -
State: open - Opened by pplantinga 5 months ago
- 4 comments
Labels: enhancement
#2683 - Fixes slow iteration over tensor in GPU
Pull Request -
State: closed - Opened by nie3e 5 months ago
- 7 comments
#2677 - Positional Rotary Embeddings for transformers
Issue -
State: open - Opened by TParcollet 5 months ago
- 2 comments
Labels: enhancement
#2675 - Multi-Window Multi-Head Attention implementation for ASR transformer
Pull Request -
State: open - Opened by Kyhne 6 months ago
- 3 comments
#2669 - large CER
Issue -
State: open - Opened by ZCzzzzzz 6 months ago
- 6 comments
Labels: bug
#2669 - large CER
Issue -
State: open - Opened by ZCzzzzzz 6 months ago
- 6 comments
Labels: bug
#2667 - ASR Model Trained with Orange/SSA-HuBERT-base-60k Returns Empty String During Inference
Issue -
State: open - Opened by ajesujoba 6 months ago
- 12 comments
Labels: bug
#2663 - SpeechLLM and Whisper
Issue -
State: open - Opened by TParcollet 6 months ago
- 6 comments
Labels: enhancement
#2662 - DDP backend enhancement
Issue -
State: closed - Opened by TParcollet 6 months ago
- 1 comment
Labels: refactor
#2661 - SpeechBrain and ONNX
Issue -
State: open - Opened by TParcollet 6 months ago
- 2 comments
Labels: bug, enhancement
#2660 - SpeechBrain pipelines
Issue -
State: open - Opened by TParcollet 6 months ago
Labels: enhancement
#2658 - Github release actions failing
Issue -
State: closed - Opened by pplantinga 6 months ago
- 1 comment
Labels: bug
#2656 - soundchoice-g2p on HF emits `<eos>` symbol (which HF tests don't expect)
Issue -
State: closed - Opened by asumagic 6 months ago
Labels: bug, regression, confirmed
#2655 - URL checks: false positives, filtering and fixes to do
Issue -
State: closed - Opened by asumagic 6 months ago
Labels: bug
#2651 - How do you add a yaml file?
Issue -
State: closed - Opened by StevePGLS 6 months ago
- 2 comments
Labels: bug
#2650 - Declaration: torchaudio_sox::load_audio_file(str _0, int? _1, int? _2, bool? _3, bool? _4, str? _5) -> (Tensor _0, int _1) Cast error details: Unable to cast Python instance of type <class 'pathlib.PosixPath'> to C++ type '?' (#define PYBIND11_DETAILED_ERROR_MESSAGES or compile in debug mode for details)
Issue -
State: open - Opened by bgenchel 6 months ago
- 3 comments
Labels: bug
#2647 - Fix to support CV 18
Pull Request -
State: closed - Opened by cvl01 6 months ago
- 2 comments
#2642 - Kmeans fix
Pull Request -
State: closed - Opened by poonehmousavi 6 months ago
- 5 comments
#2638 - VAD - very slow "apply_threshold" when running on cuda device + possible solution
Issue -
State: closed - Opened by nie3e 7 months ago
- 1 comment
Labels: bug
#2609 - Pin transient dep conllu to 4.x to workaround broken upstream
Pull Request -
State: closed - Opened by asumagic 7 months ago
- 1 comment
#2602 - Kmeans .fit() should be changed to .partial_fit()
Issue -
State: closed - Opened by Beilong-Tang 8 months ago
- 1 comment
Labels: bug
#2602 - Kmeans .fit() should be changed to .partial_fit()
Issue -
State: closed - Opened by Beilong-Tang 8 months ago
- 1 comment
Labels: bug
#2588 - Multi-GPU issue when pre-training Wav2Vec2
Issue -
State: closed - Opened by GasserElbanna 8 months ago
- 5 comments
Labels: bug
#2588 - Multi-GPU issue when pre-training Wav2Vec2
Issue -
State: closed - Opened by GasserElbanna 8 months ago
- 5 comments
Labels: bug
#2574 - Recipe for SEP-28k dataset
Pull Request -
State: closed - Opened by IliasMAOUDJ 8 months ago
- 6 comments
Labels: recipes
#2573 - Cannot reproduce the result in speech translation
Issue -
State: open - Opened by chentuochao 8 months ago
- 7 comments
Labels: bug, regression, confirmed
#2568 - ✨ Add SNAC
Pull Request -
State: open - Opened by julien-blanchon 9 months ago
- 8 comments
Labels: enhancement
#2566 - Modified conformer warmup
Pull Request -
State: open - Opened by asumagic 9 months ago
- 3 comments
Labels: enhancement, on hold
#2566 - Modified conformer warmup
Pull Request -
State: open - Opened by asumagic 9 months ago
- 3 comments
Labels: enhancement, on hold
#2558 - Add `quirks.py` for GPU/platform-specific quirks; disable CuDNN benchmarking on ROCm
Pull Request -
State: closed - Opened by asumagic 9 months ago
- 4 comments
Labels: enhancement, performance
#2552 - AMD ROCm: `torch.backends.cudnn.benchmark` should be set to `False` by default on ROCm
Issue -
State: closed - Opened by asumagic 9 months ago
Labels: bug, confirmed
#2548 - Less aggressive SpecAug to reduce(?) divergence
Pull Request -
State: closed - Opened by asumagic 9 months ago
- 1 comment
#2533 - Training regression for Conformer-Transducer models
Issue -
State: open - Opened by asumagic 10 months ago
- 4 comments
Labels: bug
#2441 - SpeechT5 integration
Pull Request -
State: open - Opened by helleuch 12 months ago
- 3 comments
Labels: enhancement
#2424 - Adding spoken dialogue state tracking recipes for Spoken MultiWoz and SpokenWoz datasets
Pull Request -
State: open - Opened by LucasDruart about 1 year ago
- 8 comments
#2405 - add GigaSpeech dataset in SpeechBrain
Pull Request -
State: closed - Opened by Adel-Moumen about 1 year ago
- 3 comments
Labels: enhancement, ready to review, recipes
#2301 - Beam search too slow on the transducer-conformer recipe
Issue -
State: open - Opened by asumagic about 1 year ago
- 1 comment
Labels: bug, performance, confirmed
#2221 - Issue while training our own data "All tensors must have same number of dimensions"
Issue -
State: open - Opened by Manusha5 over 1 year ago
- 7 comments
Labels: bug
#1565 - Wav2vec2-FR representations for prediction of arousal and valence dimensions of emotion for RECOLA dataset
Pull Request -
State: open - Opened by SinaAlisamir over 2 years ago
- 6 comments
#1462 - Loss is NaN for LibriSpeech recipe
Issue -
State: closed - Opened by Asma-droid over 2 years ago
- 13 comments
#1347 - Fix tensorboard logger
Pull Request -
State: closed - Opened by TParcollet almost 3 years ago
- 6 comments
Labels: on hold
#1347 - Fix tensorboard logger
Pull Request -
State: closed - Opened by TParcollet almost 3 years ago
- 6 comments
Labels: on hold
#1311 - Fix libresspech lm dataset
Pull Request -
State: closed - Opened by Moumeneb1 almost 3 years ago
- 5 comments
Labels: on hold
#1311 - Fix libresspech lm dataset
Pull Request -
State: closed - Opened by Moumeneb1 almost 3 years ago
- 5 comments
Labels: on hold
#1303 - bug fix Pretrained.load_audio
Pull Request -
State: closed - Opened by goexle about 3 years ago
- 28 comments
#1303 - bug fix Pretrained.load_audio
Pull Request -
State: closed - Opened by goexle about 3 years ago
- 28 comments
#1283 - add HifiGAN vocoder for LJSpeech TTS
Pull Request -
State: closed - Opened by BenoitWang about 3 years ago
- 7 comments
Labels: ready to review
#1192 - [Feature] Add the must-c v1 ST recipe
Pull Request -
State: open - Opened by freddy5566 about 3 years ago
- 16 comments
Labels: enhancement
#1192 - [Feature] Add the must-c v1 ST recipe
Pull Request -
State: open - Opened by freddy5566 about 3 years ago
- 16 comments
Labels: enhancement
#1155 - Pretrained.from_hparams() failing on windows when non admin
Issue -
State: closed - Opened by gfournier about 3 years ago
- 8 comments
Labels: bug
#512 - PyPI installation fails with a conda environment on Windows OS
Issue -
State: closed - Opened by KacperKubara about 4 years ago
- 12 comments
Labels: good first issue
#100 - Integrated the ISTFT
Pull Request -
State: closed - Opened by williamaris almost 5 years ago
- 3 comments
#100 - Integrated the ISTFT
Pull Request -
State: closed - Opened by williamaris almost 5 years ago
- 3 comments
#99 - Add transducer loss
Pull Request -
State: closed - Opened by aheba almost 5 years ago
- 11 comments
Labels: enhancement, ready to review
#99 - Add transducer loss
Pull Request -
State: closed - Opened by aheba almost 5 years ago
- 11 comments
Labels: enhancement, ready to review
#98 - librispeech: change place of data_prepare
Pull Request -
State: closed - Opened by jjery2243542 almost 5 years ago
- 1 comment
Labels: ready to review
#98 - librispeech: change place of data_prepare
Pull Request -
State: closed - Opened by jjery2243542 almost 5 years ago
- 1 comment
Labels: ready to review
#97 - Pooling Doctest fails
Issue -
State: closed - Opened by pplantinga almost 5 years ago
- 2 comments
Labels: bug, documentation
#97 - Pooling Doctest fails
Issue -
State: closed - Opened by pplantinga almost 5 years ago
- 2 comments
Labels: bug, documentation
#96 - Move integration tests to minimal examples folder
Pull Request -
State: closed - Opened by pplantinga almost 5 years ago
- 1 comment
Labels: bug, ready to review
#96 - Move integration tests to minimal examples folder
Pull Request -
State: closed - Opened by pplantinga almost 5 years ago
- 1 comment
Labels: bug, ready to review
#95 - Yaml upgrades
Pull Request -
State: closed - Opened by pplantinga almost 5 years ago
- 3 comments
Labels: enhancement
#95 - Yaml upgrades
Pull Request -
State: closed - Opened by pplantinga almost 5 years ago
- 3 comments
Labels: enhancement
#94 - Move LibriSpeech and VoxCeleb preparation to recipes dir
Issue -
State: closed - Opened by pplantinga almost 5 years ago
Labels: refactor
#94 - Move LibriSpeech and VoxCeleb preparation to recipes dir
Issue -
State: closed - Opened by pplantinga almost 5 years ago
Labels: refactor
#93 - Data_preparation in two places?
Issue -
State: closed - Opened by TParcollet almost 5 years ago
- 2 comments
Labels: question
#93 - Data_preparation in two places?
Issue -
State: closed - Opened by TParcollet almost 5 years ago
- 2 comments
Labels: question
#92 - Librispeech preparation
Pull Request -
State: closed - Opened by jjery2243542 almost 5 years ago
- 6 comments
Labels: refactor
#91 - Remove all doctest code that creates "exp" directory
Pull Request -
State: closed - Opened by pplantinga almost 5 years ago
- 1 comment
Labels: bug, ready to review
#91 - Remove all doctest code that creates "exp" directory
Pull Request -
State: closed - Opened by pplantinga almost 5 years ago
- 1 comment
Labels: bug, ready to review
#90 - When running pytest, an new exp folder appears
Issue -
State: closed - Opened by mravanelli almost 5 years ago
- 2 comments
Labels: bug
#89 - Replicability Issue
Issue -
State: closed - Opened by mravanelli almost 5 years ago
Labels: bug
#88 - Cannot run minimal examples from recipe/minimal_example/*
Issue -
State: closed - Opened by mravanelli almost 5 years ago
- 2 comments
Labels: bug
#87 - ARPA LM Reading and query interface
Pull Request -
State: closed - Opened by Gastron almost 5 years ago
- 7 comments
Labels: enhancement
#86 - Remove extra output size
Pull Request -
State: closed - Opened by pplantinga almost 5 years ago
- 2 comments
Labels: bug
#85 - Fix RNN to support RNN-T and multi-gpu training
Pull Request -
State: closed - Opened by mravanelli almost 5 years ago
- 11 comments
Labels: work in progress, ready to review
#84 - separate last layer (head) & activation from CRDNN block and add it into forward fct to be more transparent
Pull Request -
State: closed - Opened by aheba almost 5 years ago
#83 - WIP: seq2seq model training and decoding
Pull Request -
State: closed - Opened by jjery2243542 almost 5 years ago
- 7 comments
Labels: enhancement, work in progress
#82 - Add convl_layer option. If True, reshape the input(dim3*dim4), otherwise use last dim for matmul
Pull Request -
State: closed - Opened by aheba almost 5 years ago
- 3 comments
Labels: enhancement