sanchit-gandhi seq2seq-speech issues

sanchit-gandhi / seq2seq-speech

Repository for fine-tuning Transformers 🤗 based seq2seq speech models in JAX/Flax.

34 stars 6 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

May we know the version of jax/flax/transformers?

#98 snoop2head opened 1 year ago
0
do not have file requirements.txt

#97 pphuc25 opened 1 year ago
0
error while finetuning wav2vec2 with bart.

#96 arvindmn01 opened 1 year ago
1
whisper

#95 surajpachouri opened 1 year ago
0
It worked demo in the morning 29 May, but now it doesn't. It gives Error when downloading audio files of different formats (m4a, mp3, occ)

#94 maxdovbysh opened 1 year ago
1
Analyse data script

#93 patrickvonplaten closed 2 years ago
0
[E22] Add missing junk token (<silence>)

#92 sanchit-gandhi closed 2 years ago
0
[Whisper] Training

#91 patrickvonplaten closed 2 years ago
0
[SWBD] Torchaudio resampler (12x faster than Librosa)

#90 sanchit-gandhi closed 2 years ago
0
[Earnings22] Black box systems

#89 sanchit-gandhi closed 2 years ago
0
[SWBD+Fisher] Black-box pre-processing

#88 sanchit-gandhi closed 2 years ago
0
More scripts for ngram

#87 patrickvonplaten closed 2 years ago
0
[black-box] Add scripts for ERSR black-box runs and code changes

#86 sanchit-gandhi closed 2 years ago
0
[Error Correction] Add SPGISpeech and tidy up comments

#85 sanchit-gandhi closed 2 years ago
0
[ASR Examples] Only filter training samples by audio length criterion

#84 sanchit-gandhi closed 2 years ago
0
[Ngram] compute ngrams & evaluate

#83 patrickvonplaten closed 2 years ago
3
[prepare_dataset] Improve error correction vs normalisation

#82 sanchit-gandhi closed 2 years ago
0
[CTC] SWBD Tokenizer

#81 sanchit-gandhi closed 2 years ago
0
Improve error correction vs. normalization

#80 patrickvonplaten closed 2 years ago
1
[n-gram] Add dummy script

#79 sanchit-gandhi closed 2 years ago
0
[n-gram] Update pre-processing to align with CTC + S2S

#78 sanchit-gandhi closed 2 years ago
0
Eval samples filtering

#77 mutiann closed 2 years ago
1
[Seq2Seq] Fix pad_shard_unpad kwarg bug

#76 sanchit-gandhi closed 2 years ago
0
Random `TypeError` with NumPy array with all values -100

#75 versae closed 2 years ago
3
Recovering from crashed run

#74 versae opened 2 years ago
3
[Seq2Seq] Fix pad_shard_unpad kwarg bug

#73 sanchit-gandhi closed 2 years ago
0
Add RNN-T Model & Training Script

#72 sanchit-gandhi closed 2 years ago
1
[Seq2Seq] Template scripts

#71 sanchit-gandhi closed 2 years ago
0
[Imports] Fix Flax model paths (for TPU) and omit RNN-T (for GPU)

#70 sanchit-gandhi closed 2 years ago
0
[Pre-processing] Make ds specific (for ERSR)

#69 sanchit-gandhi closed 2 years ago
0
[Clean] Tidy up imports

#68 sanchit-gandhi closed 2 years ago
0
[CTC] Add regularisation

#67 sanchit-gandhi closed 2 years ago
0
[Dataloader] Handle incomplete final batch

#66 sanchit-gandhi closed 2 years ago
0
[`_do_init`] Update modeling files to handle `_do_init` kwarg

#65 sanchit-gandhi closed 2 years ago
0
[Earnings] Final preprocessing

#64 sanchit-gandhi closed 2 years ago
0
[CTC] Fix missing backslash in dummy CTC run script

#63 sanchit-gandhi closed 2 years ago
0
[Run] Fix bugs in dummy run scripts

#62 sanchit-gandhi closed 2 years ago
0
[Datasets] fix hashing/caching bug

#61 sanchit-gandhi closed 2 years ago
0
[CTC] Filter by min target length

#60 sanchit-gandhi closed 2 years ago
0
[CTC] Fix eval/test split bug

#59 sanchit-gandhi closed 2 years ago
0
[GigaSpeech] Add preprocessing

#58 sanchit-gandhi closed 2 years ago
0
[Earnings 22] Add preprocessing

#57 sanchit-gandhi closed 2 years ago
0
[CTC] Update dummy run script

#56 sanchit-gandhi closed 2 years ago
0
Update dummy run scripts

#55 sanchit-gandhi closed 2 years ago
0
[Eval] Fix eval steps bug

#54 sanchit-gandhi closed 2 years ago
0
[SWB] Finalise preprocessing

#53 sanchit-gandhi closed 2 years ago
0
[AMI] Add Seq2Seq + CTC training scripts

#52 sanchit-gandhi closed 2 years ago
0
[Seq2Seq Train] Beam at each eval steps, greedy at end of epoch

#51 sanchit-gandhi closed 2 years ago
0
[wandb] Fix push to hub, log run id

#50 sanchit-gandhi closed 2 years ago
0
[Preprocess] Gigaspeech and SWB

#49 sanchit-gandhi closed 2 years ago
0