Ecosyste.ms: Issues

An open API service for providing issue and pull request metadata for open source projects.

GitHub / seannaren/deepspeech.pytorch issues and pull requests

#100 - Use bucketing sampler while continue_from

Pull Request - State: closed - Opened by EgorLakomkin over 7 years ago - 2 comments

#99 - Training deepspeech.pytorch with variable label size

Issue - State: closed - Opened by Deepayan137 over 7 years ago - 1 comment

#98 - fixed super() initialization error

Pull Request - State: closed - Opened by aaronzira over 7 years ago - 1 comment

#98 - fixed super() initialization error

Pull Request - State: closed - Opened by aaronzira over 7 years ago - 1 comment

#97 - Removed encoding since it happens internally

Pull Request - State: closed - Opened by SeanNaren over 7 years ago

#96 - Error while compiling training script...

Issue - State: closed - Opened by anindyasarkarIITH over 7 years ago - 5 comments

#96 - Error while compiling training script...

Issue - State: closed - Opened by anindyasarkarIITH over 7 years ago - 5 comments

#95 - Add beam search

Pull Request - State: closed - Opened by ryanleary over 7 years ago - 9 comments

#95 - Add beam search

Pull Request - State: closed - Opened by ryanleary over 7 years ago - 9 comments

#94 - Infinity loss warning (received an inf loss, setting loss value to 0)

Issue - State: closed - Opened by aaronzira over 7 years ago - 2 comments

#93 - Python 3: DeepSpeech.PyTorch X PyTorch Audio

Issue - State: closed - Opened by dlmacedo over 7 years ago - 4 comments

#92 - Fixed a bug and cleaned up the code in continue_from section of train.py

Pull Request - State: closed - Opened by SiddGururani over 7 years ago - 1 comment

#92 - Fixed a bug and cleaned up the code in continue_from section of train.py

Pull Request - State: closed - Opened by SiddGururani over 7 years ago - 1 comment

#91 - save wer/cer and avg_loss during training

Pull Request - State: closed - Opened by EgorLakomkin over 7 years ago - 1 comment

#90 - Segfault: warp-ctc pytorch bindings

Issue - State: closed - Opened by jerinphilip over 7 years ago - 10 comments

#89 - Added parameter size count, various refactor of main training class

Pull Request - State: closed - Opened by SeanNaren over 7 years ago

#88 - Add parameter count on Training/Benchmarking

Issue - State: closed - Opened by SeanNaren over 7 years ago
Labels: enhancement

#88 - Add parameter count on Training/Benchmarking

Issue - State: closed - Opened by SeanNaren over 7 years ago
Labels: enhancement

#87 - Updated Test variable and README information

Pull Request - State: closed - Opened by SeanNaren over 7 years ago - 2 comments

#87 - Updated Test variable and README information

Pull Request - State: closed - Opened by SeanNaren over 7 years ago - 2 comments

#86 - Language model to predict.py

Issue - State: closed - Opened by mit456 over 7 years ago - 27 comments

#86 - Language model to predict.py

Issue - State: closed - Opened by mit456 over 7 years ago - 27 comments

#85 - Pre-trained models tracker

Issue - State: closed - Opened by SeanNaren over 7 years ago - 24 comments

#85 - Pre-trained models tracker

Issue - State: closed - Opened by SeanNaren over 7 years ago - 24 comments

#84 - Fix typo in train.py

Pull Request - State: closed - Opened by ryanleary over 7 years ago

#84 - Fix typo in train.py

Pull Request - State: closed - Opened by ryanleary over 7 years ago

#83 - Sampler that splits utterances into buckets

Pull Request - State: closed - Opened by EgorLakomkin over 7 years ago - 11 comments

#83 - Sampler that splits utterances into buckets

Pull Request - State: closed - Opened by EgorLakomkin over 7 years ago - 11 comments

#82 - Error when import warpctc_pytorch

Issue - State: closed - Opened by philokey over 7 years ago - 2 comments

#81 - Added tensorboard logging

Pull Request - State: closed - Opened by SiddGururani over 7 years ago - 1 comment

#80 - Manually set autograd params for reduced memory

Pull Request - State: closed - Opened by ryanleary over 7 years ago - 1 comment

#80 - Manually set autograd params for reduced memory

Pull Request - State: closed - Opened by ryanleary over 7 years ago - 1 comment

#79 - flags --visdom and --continue_from break the code

Issue - State: closed - Opened by SiddGururani over 7 years ago - 4 comments

#78 - Validation loss increasing while WER decreases

Issue - State: closed - Opened by SiddGururani over 7 years ago - 25 comments

#77 - Error while executing training script

Issue - State: closed - Opened by anindyasarkarIITH over 7 years ago - 3 comments

#76 - KenLM integration (Beam search)

Issue - State: closed - Opened by SeanNaren over 7 years ago - 23 comments
Labels: enhancement, help wanted

#75 - Batch_norm W, b gradients for RNN0 layer are None

Issue - State: closed - Opened by SiddGururani over 7 years ago - 4 comments

#74 - deepspeech.Pytorch vs deepspeech.Torch Memory Usage

Issue - State: closed - Opened by dlmacedo over 7 years ago - 2 comments

#73 - Fix augment loading audio

Pull Request - State: closed - Opened by SeanNaren over 7 years ago - 2 comments

#73 - Fix augment loading audio

Pull Request - State: closed - Opened by SeanNaren over 7 years ago - 2 comments

#72 - function augment_audio_with_sox check

Issue - State: closed - Opened by jeon30c over 7 years ago - 2 comments

#70 - Added noise injection script

Pull Request - State: closed - Opened by SeanNaren over 7 years ago

#70 - Added noise injection script

Pull Request - State: closed - Opened by SeanNaren over 7 years ago

#69 - Noise injection - files used

Issue - State: closed - Opened by aaronzira over 7 years ago - 2 comments

#67 - Loss is 0.000 When using --cuda parameter.

Issue - State: closed - Opened by lintangsutawika over 7 years ago - 6 comments

#67 - Loss is 0.000 When using --cuda parameter.

Issue - State: closed - Opened by lintangsutawika over 7 years ago - 6 comments

#66 - Error when running train.py

Issue - State: closed - Opened by lintangsutawika over 7 years ago - 2 comments

#66 - Error when running train.py

Issue - State: closed - Opened by lintangsutawika over 7 years ago - 2 comments

#65 - Output Layer Type

Issue - State: closed - Opened by ryanleary over 7 years ago - 9 comments

#64 - Fix train continue bug on cuda models

Pull Request - State: closed - Opened by ryanleary over 7 years ago - 1 comment

#64 - Fix train continue bug on cuda models

Pull Request - State: closed - Opened by ryanleary over 7 years ago - 1 comment

#63 - Continue_from checkpoint fails with --cuda

Issue - State: closed - Opened by aaronzira over 7 years ago - 1 comment

#62 - Fix benchmark script

Pull Request - State: closed - Opened by SeanNaren over 7 years ago

#61 - Convergence speed difference between pytorch and torch

Issue - State: closed - Opened by Minju-Jung over 7 years ago - 8 comments

#61 - Convergence speed difference between pytorch and torch

Issue - State: closed - Opened by Minju-Jung over 7 years ago - 8 comments

#60 - Filtering by minimum and maximum duration individually

Pull Request - State: closed - Opened by EgorLakomkin over 7 years ago - 1 comment

#59 - Pre-trained models

Issue - State: closed - Opened by danielhauagge over 7 years ago - 13 comments
Labels: help wanted, Blocked

#58 - Batch sizes

Issue - State: closed - Opened by ryanleary over 7 years ago - 8 comments

#57 - Return the current validation error/loss

Pull Request - State: closed - Opened by ryanleary over 7 years ago

#56 - Fix noise arg default in train script

Pull Request - State: closed - Opened by ryanleary over 7 years ago - 4 comments

#55 - FP16 support

Issue - State: closed - Opened by SeanNaren over 7 years ago - 10 comments
Labels: enhancement, Blocked

#55 - FP16 support

Issue - State: closed - Opened by SeanNaren over 7 years ago - 10 comments
Labels: enhancement, Blocked

#54 - Refactor model serialization & deserialization for CPU/GPU portability

Pull Request - State: closed - Opened by ryanleary over 7 years ago - 3 comments

#54 - Refactor model serialization & deserialization for CPU/GPU portability

Pull Request - State: closed - Opened by ryanleary over 7 years ago - 3 comments

#53 - Noise inject #35

Pull Request - State: closed - Opened by SeanNaren over 7 years ago - 1 comment

#52 - Test args

Pull Request - State: closed - Opened by ryanleary over 7 years ago

#52 - Test args

Pull Request - State: closed - Opened by ryanleary over 7 years ago

#51 - Add .gitignore

Pull Request - State: closed - Opened by ryanleary over 7 years ago

#51 - Add .gitignore

Pull Request - State: closed - Opened by ryanleary over 7 years ago

#50 - Create release script for model

Issue - State: closed - Opened by SeanNaren over 7 years ago - 1 comment
Labels: enhancement

#49 - Persist metadata, use in prediction, various fixes

Pull Request - State: closed - Opened by SeanNaren over 7 years ago

#47 - Fix model saving for different RNN types

Pull Request - State: closed - Opened by ryanleary over 7 years ago - 4 comments

#47 - Fix model saving for different RNN types

Pull Request - State: closed - Opened by ryanleary over 7 years ago - 4 comments

#46 - Implement Row Convolution in PyTorch

Issue - State: closed - Opened by SeanNaren over 7 years ago - 18 comments
Labels: enhancement

#46 - Implement Row Convolution in PyTorch

Issue - State: closed - Opened by SeanNaren over 7 years ago - 18 comments
Labels: enhancement

#45 - What kind of RTF should one expect?

Issue - State: closed - Opened by migueljette over 7 years ago - 14 comments

#45 - What kind of RTF should one expect?

Issue - State: closed - Opened by migueljette over 7 years ago - 14 comments

#44 - Fix for #43

Pull Request - State: closed - Opened by SeanNaren over 7 years ago - 1 comment

#43 - 'iteration' argument is not set when continuing from checkpoint

Issue - State: closed - Opened by EgorLakomkin over 7 years ago - 1 comment

#42 - Added merging manifests script

Pull Request - State: closed - Opened by SeanNaren over 7 years ago

#41 - Added benchmark script, updated README with instructions

Pull Request - State: closed - Opened by SeanNaren over 7 years ago

#40 - librosa.core.load is significantly slower than scipy.io.wavfile.read

Issue - State: closed - Opened by EgorLakomkin over 7 years ago - 3 comments

#39 - LibriSpeech dataset

Pull Request - State: closed - Opened by EgorLakomkin over 7 years ago - 2 comments

#38 - Fixed epoch count

Pull Request - State: closed - Opened by SeanNaren over 7 years ago

#37 - visdom assert 'AssertionError: Y should be same size as X'

Issue - State: closed - Opened by demobin8 over 7 years ago - 3 comments

#36 - Batch benchmarking script

Issue - State: closed - Opened by SeanNaren over 7 years ago - 1 comment
Labels: enhancement

#36 - Batch benchmarking script

Issue - State: closed - Opened by SeanNaren over 7 years ago - 1 comment
Labels: enhancement

#35 - Noise injection

Issue - State: closed - Opened by SeanNaren over 7 years ago - 5 comments
Labels: enhancement

#35 - Noise injection

Issue - State: closed - Opened by SeanNaren over 7 years ago - 5 comments
Labels: enhancement

#34 - added support of simple_rnn, gru and lstm for the deepspeech module

Pull Request - State: closed - Opened by EgorLakomkin over 7 years ago - 4 comments

#33 - Issue w/ TED Dataloader

Issue - State: closed - Opened by ryanleary over 7 years ago - 4 comments

#33 - Issue w/ TED Dataloader

Issue - State: closed - Opened by ryanleary over 7 years ago - 4 comments

#32 - RuntimeError: CUDNN_STATUS_INTERNAL_ERROR

Issue - State: closed - Opened by demobin8 over 7 years ago - 16 comments

#31 - train complete but TypeError: checkpoint() takes at least 5 arguments (4 given)

Issue - State: closed - Opened by demobin8 over 7 years ago - 1 comment

#30 - During training utterances are always processed sorted by duration

Issue - State: closed - Opened by EgorLakomkin over 7 years ago - 16 comments
Labels: enhancement

#30 - During training utterances are always processed sorted by duration

Issue - State: closed - Opened by EgorLakomkin over 7 years ago - 16 comments
Labels: enhancement