Commit Graph

  • f99f4d7c92 add decode seamlessm4t Yuekai Zhang 2023-09-05 11:12:19 +08:00
  • 77178c6311 comment out params related to the chunk size Triplecq 2024-01-14 17:35:20 -05:00
  • 7b6a89749d customize decoding script Triplecq 2024-01-14 17:29:22 -05:00
  • 04fa9e3e8c traning script completed Triplecq 2024-01-15 07:06:14 +09:00
  • 42c152f5cb decrease learning-rate to solve the error: RuntimeError: grad_scale is too small, exiting: 5.820766091346741e-11 Triplecq 2024-01-14 12:12:15 -05:00
  • ced8a53cdc Merge branch 'master' into rs Triplecq 2024-01-14 23:05:00 +09:00
  • 819db8fcad Merge branch 'master' of github.com:Triplecq/icefall Triplecq 2024-01-14 23:00:19 +09:00
  • dc2d531540 customized recipes for rs Triplecq 2024-01-14 22:28:53 +09:00
  • b1de6f266c customized recipes for reazonspeech Triplecq 2024-01-14 22:28:32 +09:00
  • 1e6fe2eae1 restore Triplecq 2024-01-14 08:05:49 -05:00
  • 5e9a171b20 customize tranning script for rs Triplecq 2024-01-14 07:45:33 -05:00
  • 8eae6ec7d1 Add pruned_transducer_stateless2 from reazonspeech branch Triplecq 2024-01-14 05:23:26 -05:00
  • af87726bf2 init zipformer recipe Triplecq 2024-01-14 19:13:21 +09:00
  • 05f093bd66 deploy: 398401ed277d4f895f624a95919c57edbbde4cba csukuangfj 2024-01-14 06:39:10 +00:00
  • 398401ed27
    Update kaldifeat installation doc (#1460) Fangjun Kuang 2024-01-14 14:38:41 +08:00
  • dca7e6b23e Update kaldifeat installation doc Fangjun Kuang 2024-01-14 14:33:14 +08:00
  • bce81394a4 Merge branch 'dev/aishell-zipformer-bbpe' of https://github.com/jinzr/icefall into dev/aishell-zipformer-bbpe JinZr 2024-01-11 09:56:01 +08:00
  • 9c9b23447b minor fixes JinZr 2024-01-11 09:55:54 +08:00
  • 5689a4cfc1 Update decode.py jinzr 2024-01-11 09:55:28 +08:00
  • 07899ee85a Merge branch 'dev/aishell-zipformer-bbpe' of https://github.com/jinzr/icefall into dev/aishell-zipformer-bbpe JinZr 2024-01-11 09:51:06 +08:00
  • fb9e52ae0e Update decode.py jinzr 2024-01-11 09:49:10 +08:00
  • 585ae128df Merge branch 'dev/aishell-zipformer-bbpe' of https://github.com/jinzr/icefall into dev/aishell-zipformer-bbpe JinZr 2024-01-11 09:41:40 +08:00
  • 5253941551 Update decode.py jinzr 2024-01-11 09:41:33 +08:00
  • d56d88b85e minor fix JinZr 2024-01-11 09:38:11 +08:00
  • 638cb5cc35 Update train.py jinzr 2024-01-10 22:07:27 +08:00
  • 5d630a8b36 Create decode.py jinzr 2024-01-10 21:47:46 +08:00
  • 0e7ec8ac56 Update train.py jinzr 2024-01-10 21:36:04 +08:00
  • 41413967d1 Update train.py jinzr 2024-01-10 21:34:21 +08:00
  • a7a4837425 init commit jinzr 2024-01-10 21:29:20 +08:00
  • 9975dd66f9 Small fix to make disambig file optional Xinyuan Li 2024-01-09 18:27:09 -05:00
  • ff89ef6643 Small fix to dependency Xinyuan Li 2024-01-09 14:06:00 -05:00
  • 9180211fdb Small fix to conformer dependency Xinyuan Li 2024-01-09 13:58:42 -05:00
  • bfd341d395 Small fix to relative path Xinyuan Li 2024-01-09 12:41:05 -05:00
  • a1d1f2e434 deploy: e2fcb42f5f176d9e39eb38506ab99d0a3adaf202 marcoyang1998 2024-01-09 07:44:23 +00:00
  • e2fcb42f5f
    fix typo (#1455) Xiaoyu Yang 2024-01-09 15:41:37 +08:00
  • 083744d894 fix typo marcoyang 2024-01-09 15:39:43 +08:00
  • 42863b1f8e change starting index; support different subsets marcoyang 2024-01-09 10:14:23 +08:00
  • 5445ea6df6
    Use shuffled LibriSpeech cuts instead (#1450) zr_jin 2024-01-08 15:09:21 +08:00
  • b9b56eb879
    Minor fixes to the VCTK data prep scripts (#1441) zr_jin 2024-01-08 14:28:07 +08:00
  • b5dfe244be leave the old code in comments for reference jinzr 2024-01-08 14:24:55 +08:00
  • 9c29aaa61c use shuffled LibriSpeech cuts instead jinzr 2024-01-05 19:44:10 +08:00
  • 44b9730e82 update RESULTS.md marcoyang 2024-01-05 17:39:16 +08:00
  • 4f849985c4 add file marcoyang 2024-01-05 16:17:40 +08:00
  • e31f7d5abf add decode.py marcoyang 2024-01-05 16:17:29 +08:00
  • 000e54a319 minor fix marcoyang 2024-01-05 16:14:44 +08:00
  • 6665469a4e add args in asr datamodule marcoyang 2024-01-05 16:12:24 +08:00
  • 716b82cc3a
    streaming_decode.py, relax the audio range from [-1,+1] to [-10,+10] (#1448) Karel Vesely 2024-01-05 03:21:27 +01:00
  • bd5b9b0c05 streaming_decode.py, relax the audio range from [-1,+1] to [-10,+10] Karel Vesely 2024-01-04 15:40:46 +01:00
  • 8136ad775b
    Use high_freq -400 in computing fbank features. (#1447) Fangjun Kuang 2024-01-04 13:59:32 +08:00
  • 830206dfcd Use high_freq -400 in computing fbank features. Fangjun Kuang 2024-01-04 13:42:13 +08:00
  • a232cebbc8
    small fix for decode.py Yifan Yang 2024-01-03 17:00:13 +08:00
  • 8023493029
    update Yifan Yang 2024-01-01 20:09:22 +08:00
  • a01a4231c4
    Merge branch 'k2-fsa:master' into dev/ssl Yifan Yang 2024-01-01 20:08:48 +08:00
  • f42258caf8
    Update compute_fbank_commonvoice_splits.py (#1437) zr_jin 2023-12-30 13:03:26 +08:00
  • e06e128d73
    Fix finalize error in contextgraph Huahuan Zheng 2023-12-29 13:59:18 +08:00
  • 4ace5fac5d
    Merge branch 'k2-fsa:master' into fix/vctk zr_jin 2023-12-28 11:33:36 +08:00
  • e2209f93c7 Update prepare.sh jinzr 2023-12-28 11:23:38 +08:00
  • f38aec6f85 Update prepare.sh jinzr 2023-12-28 11:18:36 +08:00
  • 2436597f7f Zipformer recipe Chen 2023-12-28 05:37:40 +09:00
  • e1de397dbc Support finetuning the multi_zh model with icmc data wd929 2023-12-27 17:11:11 +08:00
  • 7310489dc9 Support training for aec-iva and gss data wd929 2023-12-27 17:00:45 +08:00
  • 70f22900b9 Update compute_fbank_commonvoice_splits.py jinzr 2023-12-27 16:26:20 +08:00
  • 140e6381ad
    Refactor CI tests for librispeech (#1436) Fangjun Kuang 2023-12-27 13:21:14 +08:00
  • b4c5db5a84 minor fixes Fangjun Kuang 2023-12-27 13:19:48 +08:00
  • 7567ce040f disable conformer_ctc export Fangjun Kuang 2023-12-27 12:48:19 +08:00
  • af20319f17 Refactor the CI tests for librispeech Fangjun Kuang 2023-12-27 11:36:39 +08:00
  • db52fe2349
    Refactor CI test for aishell (#1435) Fangjun Kuang 2023-12-26 20:29:43 +08:00
  • 83ac992569 more fixes Fangjun Kuang 2023-12-26 20:18:59 +08:00
  • 7a462e2663 Install tree for cpu docker Fangjun Kuang 2023-12-26 20:02:24 +08:00
  • c54af53813 Refactor CI for aishell Fangjun Kuang 2023-12-26 19:57:38 +08:00
  • e257b44763 Add wenetspeech kws recipe pkufool 2023-12-26 10:32:02 +08:00
  • 6dfba8b4a3 deploy: 835a92eba51a939c6b4a069a53cc1e3ddeabd9a5 csukuangfj 2023-12-25 12:33:16 +00:00
  • 835a92eba5
    Add doc about how to use the CPU-only docker images (#1432) Fangjun Kuang 2023-12-25 20:23:56 +08:00
  • 6c1b66bc2e Add doc about how to use the CPU-only docker images Fangjun Kuang 2023-12-25 20:20:50 +08:00
  • ddd7131317
    Update TTS export-onnx.py scripts for handling variable token counts (#1430) Ali Haznedaroğlu 2023-12-25 14:44:07 +03:00
  • c855a58cfd
    Generate the dependency matrix by code for GitHub Actions (#1431) Fangjun Kuang 2023-12-25 19:41:09 +08:00
  • 747dc2f0e5
    Update export-onnx.py Ali Haznedaroğlu 2023-12-25 13:51:18 +03:00
  • 48b945d1a3
    Update export-onnx.py Ali Haznedaroğlu 2023-12-25 13:50:48 +03:00
  • 278d637013
    Update export-onnx.py for variable token counts Ali Haznedaroğlu 2023-12-25 13:48:03 +03:00
  • 44bc60ff38 Add gigaspeech kws recipe pkufool 2023-12-25 18:04:24 +08:00
  • 68c1ab1e2d Install kaldi-decoder for docker Fangjun Kuang 2023-12-25 13:51:28 +08:00
  • 010c311fd6 Use cpu docker to run yesno Fangjun Kuang 2023-12-25 13:47:02 +08:00
  • 6dee6bda6d generate build matrix via scripts Fangjun Kuang 2023-12-25 12:34:54 +08:00
  • 17dab02dc9 various fixes to context graph to support kws system and bugs of hotwords pkufool 2023-12-25 12:18:32 +08:00
  • e5bb1ae86c
    Use the CPU docker in CI to simplify the test code (#1427) Fangjun Kuang 2023-12-24 13:40:33 +08:00
  • 5be11bb7b3 rename csukuangfj to k2-fsa Fangjun Kuang 2023-12-23 23:21:09 +08:00
  • 23351ea177 install matplotlib Fangjun Kuang 2023-12-23 22:42:53 +08:00
  • 2af46ef97a change k2-fsa to csukuangfj for testing Fangjun Kuang 2023-12-23 22:25:02 +08:00
  • 9aef34511f Install graphviz in docker Fangjun Kuang 2023-12-23 22:20:00 +08:00
  • f95fcd7d11 Use cpu docker to run the test Fangjun Kuang 2023-12-23 22:08:39 +08:00
  • 75c5389979 update yifanyeung 2023-12-23 15:27:34 +08:00
  • 79a42148db
    Add CI test to cover zipformer/train.py (#1424) Fangjun Kuang 2023-12-23 00:38:36 +08:00
  • 7c07f0c9b9 Merge branch 'ci-train-2' into ci-train Fangjun Kuang 2023-12-23 00:31:53 +08:00
  • fd09e5befe minor fixes Fangjun Kuang 2023-12-23 00:24:20 +08:00
  • 360c30bb59 Rename profile to my_profile Fangjun Kuang 2023-12-23 00:18:05 +08:00
  • eb34ed5560 fix torchaudio Fangjun Kuang 2023-12-22 23:41:36 +08:00
  • 0979a3eb0f install ffmpeg for torch >= 2.1.0 Fangjun Kuang 2023-12-22 23:36:54 +08:00
  • 77125064cb Add SSL yifanyeung 2023-12-22 23:01:47 +08:00
  • 5f5699c5dd first working version Fangjun Kuang 2023-12-22 22:27:53 +08:00
  • 9a707e07f7 first working version Fangjun Kuang 2023-12-22 22:27:53 +08:00