mirror of https://github.com/k2-fsa/icefall.git synced 2025-08-09 18:12:19 +00:00

History

Fix CI tests. (#1974 )

- Introduce unified AMP helpers (create_grad_scaler, torch_autocast) to handle 
  deprecations in PyTorch ≥2.3.0

- Replace direct uses of torch.cuda.amp.GradScaler and torch.cuda.amp.autocast 
  with the new utilities across all training and inference scripts

- Update all torch.load calls to include weights_only=False for compatibility with 
  newer PyTorch versions

2025-07-01 13:47:55 +08:00

conformer_ctc

Fix buffer size of DynamicBucketingSampler (#1468 )

2024-01-21 02:10:42 +08:00

local

Strengthened style constraints (#1527 )

2024-03-04 23:28:04 +08:00

pruned_transducer_stateless2

Fix CI tests. (#1974 )

2025-07-01 13:47:55 +08:00

zipformer

Fix CI tests. (#1974 )

2025-07-01 13:47:55 +08:00

.gitignore

Add Blankskip to Zipformer+CTC (#730 )

2022-12-21 17:41:31 +08:00

prepare.sh

fix gigaspeech_prepare.sh (#1734 )

2024-08-28 12:15:01 +08:00

README.md

Add Zipformer recipe for GigaSpeech (#1254 )

2023-10-21 15:36:59 +08:00

RESULTS.md

Add Zipformer recipe for GigaSpeech (#1254 )

2023-10-21 15:36:59 +08:00

shared

GigaSpeech recipe (#120 )

2022-04-14 16:07:22 +08:00

README.md

GigaSpeech

GigaSpeech, an evolving, multi-domain English speech recognition corpus with 10,000 hours of high quality labeled audio, collected from audiobooks, podcasts and YouTube, covering both read and spontaneous speaking styles, and a variety of topics, such as arts, science, sports, etc. More details can be found: https://github.com/SpeechColab/GigaSpeech

Download

Apply for the download credentials and download the dataset by following https://github.com/SpeechColab/GigaSpeech#download. Then create a symlink

ln -sfv /path/to/GigaSpeech download/GigaSpeech

Performance Record

	Dev	Test
`zipformer`	10.25	10.38
`conformer_ctc`	10.47	10.58
`pruned_transducer_stateless2`	10.40	10.51

See RESULTS for details.