Fangjun Kuang
8905ef5d59
Fix gigaspeech dataset iterator.
...
Previously, it was reset after every epoch, which may cause it to
always use the first part of the gigaspeech dataset if you choose
a small --giga-prob.
2025-11-28 10:51:21 +08:00
Karel Vesely
693f069de7
zipformer/ctc_align.py ( #2020 )
...
* zipformer/ctc_align.py
- tool for forced-alignment with CTC model
- provides timeline, computes per-token and per-utterance acoustic confidences
- based on torchaudio `forced_align()`
- confidences are computed in several ways
other modifications:
- LibriSpeechAsrDataModel extended with `::load_manifest()` to allow
passing-in cutset from CLI.
- update @custom_fwd @custom_bwd in scaling.py
- streaming_decode.py update errs/recogs/log filenames '-' <-> '_'
* putting back `custom_bwd`, `custom_fwd`
* integrating remarks from PR
* update of argparse help strings
* ctc_align.py, avoid shadowing a variable
* Finalizing the code:
- adding some coderabbit suggestions.
- removing `word_table`, `decoding_graph` from aligner API (unused)
- improved consistency of variable names (confidences)
- updated docstrings
2025-10-06 07:49:37 +08:00
Amir Hussein
729a5ba3ec
IWSLT-Ta ASR/ST ( #1362 )
...
This is a pull request for Dialectal IWSLT-Tunisian 2022 shared task https://iwslt.org/2022/dialect ASR and ST recipes.
2025-09-22 09:58:00 +08:00
Amir Hussein
855536d355
HENT-SRT ( #2026 )
...
HENT-SRT: Hierarchical Efficient Neural Transducer with Self-Distillation for Joint Speech Recognition and Translation
Paper: https://arxiv.org/abs/2506.02157
2025-09-20 00:17:53 +08:00
Fangjun Kuang
63563d16d3
Fix setting joiner dim ( #2027 )
...
Fixes incorrect computation of encoder_dim when encoder_dim is a comma-separated list of integers by ensuring numeric (not lexicographic) max is used.
Fixes #2018
- Replace int(max(params.encoder_dim.split(","))) (lexicographic max on strings) with max(_to_int_tuple(params.encoder_dim)) (numeric max).
- Apply the fix consistently across all affected training scripts.
2025-09-19 09:42:41 +08:00
qweasdzxcvde
0c7ce5256f
add tot_score inf mask to make training stable ( #2019 )
...
I find there are some inf in tot_score, it makes model cannot converge, add inf mask can make training more stable.
2025-09-08 14:36:12 +08:00
Fangjun Kuang
34fc1fdf0d
Fix transformer decoder layer ( #1995 )
2025-07-18 20:12:29 +08:00
Bailey Machiko Hirota
5fe13078cc
Musan implementation for ReazonSpeech ( #1988 )
2025-07-18 17:16:19 +08:00
Yifan Yang
9fd0f2dc1d
support left pad for make_pad_mask ( #1990 )
2025-07-16 23:59:04 +08:00
Fangjun Kuang
e22bc78f98
Export streaming zipformer2 to RKNN ( #1977 )
2025-07-11 13:24:01 +08:00
Teo Wen Shen
da87e7fc99
add weights_only=False to torch.load ( #1984 )
2025-07-10 15:27:08 +08:00
Yifan Yang
89728dd4f8
Refactor data preparation for GigaSpeech recipe ( #1986 )
2025-07-10 11:17:37 +08:00
Mistmoon
9293edc62f
Add cr-ctc loss and ctc-decode in aishell ( #1980 )
2025-07-08 14:47:24 +08:00
Fangjun Kuang
fba5e67d5e
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
Fangjun Kuang
71377d21cd
Export streaming zipformer models with whisper feature to onnx ( #1973 )
2025-06-30 19:01:15 +08:00
Fangjun Kuang
abd9437e6d
Add more wheels for piper-phonemize ( #1969 )
2025-06-24 14:49:16 +08:00
Wei Kang
e1cf4dbace
rm zipvoice ( #1967 )
2025-06-23 19:22:35 +08:00
Wei Kang
343b8fa2dc
Using non strict match in context graph for contextual words ( #1952 )
2025-06-19 12:27:15 +08:00
Wei Kang
f80a2ee110
Decrease num_buckets & remove shuffle_buffer_size ( #1955 )
2025-06-19 12:26:37 +08:00
Wei Kang
3587c4b3b7
Fix decoding byte bpes tokens to words. ( #1966 )
2025-06-19 12:26:01 +08:00
Wei Kang
762f965cf7
[zipvoice] Add requirements.txt and pinyin.txt, remove k2 from pretrained model inference. ( #1965 )
...
* Add requirements.txt and pinyin.txt needed by zipvoice
* simplify the requirements for pretrained model inference
2025-06-18 18:38:46 +08:00
Wei Kang
06539d2b9d
Add Zipvoice ( #1964 )
...
* Add ZipVoice - a flow-matching based zero-shot TTS model.
2025-06-17 20:17:12 +08:00
Zengwei Yao
ffb7d05635
refactor branch exchange in cr-ctc ( #1954 )
2025-05-27 12:09:59 +08:00
Mahsa Yarmohammadi
021e1a8846
Add acknowledgment to README ( #1950 )
2025-05-22 22:06:35 +08:00
Tianxiang Zhao
30e7ea4b5a
Fix a bug in finetune.py --use-mux ( #1949 )
2025-05-22 12:05:01 +08:00
Fangjun Kuang
fd8f8780fa
Fix logging torch.dtype. ( #1947 )
2025-05-21 12:04:57 +08:00
Yifan Yang
e79833aad2
ensure SwooshL/SwooshR output dtype matches input dtype ( #1940 )
2025-05-12 19:28:48 +08:00
Yifan Yang
4627969ccd
fix bug: undefined name 'partial' ( #1941 )
2025-05-12 14:19:53 +08:00
Yifan Yang
cd7caf12df
Fix speech_llm recipe ( #1936 )
...
* fix training/decoding scripts, cleanup unused code, and ensure compliance with style checks
---------
Co-authored-by: Your Name <you@example.com>
Co-authored-by: Fangjun Kuang <csukuangfj@gmail.com>
2025-04-30 11:41:00 +08:00
Fangjun Kuang
cc2e64a6aa
Fix convert_texts_into_ids() in the tedlium3 recipe. ( #1929 )
2025-04-24 17:04:46 +08:00
Yifan Yang
5ec95e5482
Fix SpeechLLM recipe ( #1926 )
2025-04-23 16:18:38 +08:00
math345
64c5364085
Fix bug: When resuming training from a checkpoint, model_avg was not assigned, resulting in a None error. ( #1914 )
2025-04-10 11:37:28 +08:00
Fangjun Kuang
300a821f58
Fix aishell training ( #1916 )
2025-04-10 10:30:37 +08:00
Fangjun Kuang
171cf8c9fe
Avoid redundant computation in PiecewiseLinear. ( #1915 )
2025-04-09 11:52:37 +08:00
Wei Kang
86bd16d496
[KWS]Remove graph compiler ( #1905 )
2025-04-02 22:10:06 +08:00
Fangjun Kuang
db9fb8ad31
Add scripts to export streaming zipformer(v1) to RKNN ( #1882 )
2025-02-27 17:10:58 +08:00
Yuekai Zhang
2ba665abca
Add F5-TTS with semantic token training results ( #1880 )
...
* add cosy token
* update inference code
* add extract cosy token
* update results
* add requirements.txt
* update readme
---------
Co-authored-by: yuekaiz <yuekaiz@h20-7.cm.cluster>
Co-authored-by: yuekaiz <yuekaiz@mgmt1-login.cm.cluster>
2025-02-24 13:58:47 +08:00
Machiko Bailey
da597ad782
Update RESULTS.md ( #1873 )
2025-02-04 09:04:25 +08:00
Machiko Bailey
0855b0338a
Merge japanese-to-english multilingual branch ( #1860 )
...
* add streaming support to reazonresearch
* update README for streaming
* Update RESULTS.md
* add onnx decode
---------
Co-authored-by: root <root@KDA03.cm.cluster>
Co-authored-by: Fangjun Kuang <csukuangfj@gmail.com>
Co-authored-by: root <root@KDA01.cm.cluster>
Co-authored-by: zr_jin <peter.jin.cn@gmail.com>
2025-02-04 01:33:09 +08:00
Yuekai Zhang
dd5d7e358b
F5-TTS Training Recipe for WenetSpeech4TTS ( #1846 )
...
* add f5
* add infer
* add dit
* add README
* update pretrained checkpoint usage
---------
Co-authored-by: yuekaiz <yuekaiz@h20-5.cm.cluster>
Co-authored-by: yuekaiz <yuekaiz@l20-3.cm.cluster>
Co-authored-by: yuekaiz <yuekaiz@h20-6.cm.cluster>
Co-authored-by: zr_jin <peter.jin.cn@gmail.com>
2025-01-27 16:33:02 +08:00
zr_jin
39c466e802
Update shared ( #1868 )
2025-01-21 11:04:11 +08:00
zr_jin
79074ef0d4
removed the erroneous ‘’continual'' implementation ( #1865 )
2025-01-16 20:51:28 +08:00
zr_jin
8ab0352e60
Update style_check.yml ( #1866 )
2025-01-16 17:36:09 +08:00
Han Zhu
ab91112909
Improve infinity-check ( #1862 )
...
1. Attach the inf-check hooks if the grad scale is getting too small.
2. Add try-catch to avoid OOM in the inf-check hooks.
3. Set warmup_start=0.1 to reduce chances of divergence
2025-01-09 15:05:38 +08:00
Seonuk Kim
8d602806c3
Update conformer.py ( #1859 )
...
* Update conformer.py
feedforward dimention -> feedforward dimension
* Update conformer.py
feedforward dimention -> feedforward dimension
* Update conformer.py
feedforward dimention -> feedforward dimension
* Update conformer.py
feedforward dimention -> feedforward dimension
* Update conformer.py
feedforward dimention -> feedforward dimension
* Update conformer.py
feedforward dimention -> feedforward dimension
* Update conformer.py
feedforward dimention -> feedforward dimension
* Update conformer.py
feedforward dimention -> feedforward dimension
* Update conformer.py
feedforward dimention -> feedforward dimension
* Update conformer.py
Swich -? Swish
2025-01-06 17:31:13 +08:00
Seonuk Kim
3b6d54007b
Update conformer.py ( #1857 )
...
* Update conformer.py
feedforward dimention -> feedforward dimension
* Update conformer.py
feedforward dimention -> feedforward dimension
* Update conformer.py
feedforward dimention -> feedforward dimension
* Update conformer.py
feedforward dimention -> feedforward dimension
* Update conformer.py
feedforward dimention -> feedforward dimension
* Update conformer.py
feedforward dimention -> feedforward dimension
* Update conformer.py
feedforward dimention -> feedforward dimension
* Update conformer.py
feedforward dimention -> feedforward dimension
* Update conformer.py
feedforward dimention -> feedforward dimension
2025-01-06 13:17:02 +08:00
Fangjun Kuang
3b263539cd
Publish MatchaTTS onnx models trained with LJSpeech to huggingface ( #1854 )
2025-01-02 15:54:34 +08:00
Fangjun Kuang
bfffda5afb
Add MatchaTTS for the Chinese dataset Baker ( #1849 )
2024-12-31 17:17:05 +08:00
Han Zhu
df46a3eaf9
Warn instead of raising exceptions in inf-check ( #1852 )
2024-12-31 16:52:06 +08:00
Yifan Yang
a2b0f6057c
Small fix ( #1853 )
2024-12-31 07:41:44 +08:00
