Commit Graph

  • 4e7cdb5d7e update RESULTS.md yaozengwei 2024-07-05 20:13:52 +08:00
  • bad44a7aed
    Update decoder.py Yifan Yang 2024-07-05 14:52:35 +08:00
  • 62b87cefa4 update results root 2024-07-05 05:14:05 +00:00
  • a04e70f1ce fix decoding issue root 2024-07-05 03:17:01 +00:00
  • 1d6e80195f update datasets for decoding root 2024-07-05 02:36:44 +00:00
  • 2502987dc4 support for cascades yifanyeung 2024-07-04 16:09:20 +08:00
  • 91488ce972 suport for cascades yfyeung 2024-07-04 15:53:15 +08:00
  • 7235b8561b fix black yfyeung 2024-07-04 14:55:00 +08:00
  • 53da22ecc7 update yfyeung 2024-07-04 14:53:49 +08:00
  • bacbffac11
    Merge cfb0cab3ebf719b393960ea3565361353f74eb28 into cbcac23d2617ccfdc8f1ecc14a00ba96413c3bf9 Wei Kang 2024-07-04 14:49:49 +08:00
  • 18da3e8975
    Merge branch 'k2-fsa:master' into dev/zipformer_lstm Yifan Yang 2024-07-04 14:20:43 +08:00
  • cbcac23d26
    Fix typos, remove unused packages, normalize comments (#1678) Yifan Yang 2024-07-04 14:19:45 +08:00
  • eb36aee973 Fix typos, remove unused packages, normalize comments yfyeung 2024-07-04 13:39:28 +08:00
  • 200f53d712 add zipformer lstm yfyeung 2024-07-03 22:04:23 +08:00
  • 63d501e6d6
    Merge 3f4e39d94f4fa43f79007752790fa08230fbffee into ebbd396c2bbe8f2bf626fef4e3778c32d28dc301 Wei Kang 2024-07-03 22:47:29 +09:00
  • ebbd396c2b
    update multi-hans-zh whisper-qwen-7b results (#1677) Yuekai Zhang 2024-07-03 19:55:12 +08:00
  • 1edcf21da0 fix typo root 2024-07-03 07:12:49 +00:00
  • 32191527b6 update qwen-7b whisper encoder results root 2024-07-03 07:11:53 +00:00
  • 28fbbd3d28 update qwen-7b whisper encoder results root 2024-07-03 07:09:49 +00:00
  • 36808b8940 fix the implementation of CoPE Fangjun Kuang 2024-07-03 13:43:03 +08:00
  • eaab2c819f
    Zipformer Onnx FP16 (#1671) Manix 2024-06-27 13:38:24 +05:30
  • 683ae6c2cc extending to export-onnx.py manickavela29 2024-06-27 06:01:27 +00:00
  • fa235adba2 minor refactor manickavela29 2024-06-27 05:32:53 +00:00
  • f657c364af lint fix manickavela29 2024-06-27 01:09:07 +00:00
  • f054f929a6 updating requirement file manickavela29 2024-06-26 18:46:20 +00:00
  • 5934b37e3f Zipformer Onnx fp16 manickavela29 2024-06-26 18:35:28 +00:00
  • cfb0cab3eb Initial tcpgen biasing pkufool 2024-06-26 11:42:17 +08:00
  • 9a5fc2ab71 add alimeeting norm Yuekai Zhang 2024-06-26 10:41:49 +08:00
  • b594a3875b
    Add CI for non-streaming zipformer about ksponspeech (#1667) Fangjun Kuang 2024-06-24 16:20:46 +08:00
  • b2644d4b0a Add CI for non-streaming zipformer about ksponspeech Fangjun Kuang 2024-06-24 14:17:46 +08:00
  • 031f892796
    Reformat by black non-streaming zipformer recipe for ksponspeech (#1665) Seung Hyun Lee 2024-06-24 16:28:09 +09:00
  • 0101720fb9 Reformat by black whsqkaak 2024-06-24 16:13:34 +09:00
  • 94bba4e8de
    Merge branch 'k2-fsa:master' into master Seung Hyun Lee 2024-06-24 16:06:34 +09:00
  • 6f102d3470
    Add non-streaming Zipformer recipe for KsponSpeech (#1664) Seung Hyun Lee 2024-06-24 15:07:37 +09:00
  • 8108f57285 Add Zipformer recipe for KsponSpeech whsqkaak 2024-06-24 14:53:36 +09:00
  • c14fd3872d fix cast in zipformer_lora Teo 2024-06-23 00:22:43 +09:00
  • d1dfde2c9d cast grad_scale in whiten to float Teo 2024-06-23 00:20:11 +09:00
  • 3f4e39d94f add decode pinyin pkufool 2024-06-21 18:53:47 +08:00
  • 99ea92953f Add train pinyin pkufool 2024-06-21 18:11:30 +08:00
  • 19f88482be merge with master pkufool 2024-06-21 18:05:52 +08:00
  • 3059eb4511
    Fix doc URLs (#1660) Fangjun Kuang 2024-06-21 11:10:14 +08:00
  • 41af7cc144 Fix doc URLs Fangjun Kuang 2024-06-21 11:06:54 +08:00
  • ad4615eba3
    Merge 3364d9863c7816da59b9df30f58784617ab23b84 into ff2bef9e501a4b5ebfec04cbfe8afa2e8bea4b40 Zengwei Yao 2024-06-20 03:39:40 +00:00
  • ff2bef9e50
    update multi-hans whisper-qwen-1.5b results (#1657) Yuekai Zhang 2024-06-19 11:10:31 +08:00
  • 491266cb68 update multi-hans whisper-qwen-1.5b results root 2024-06-19 02:54:40 +00:00
  • 34c038b1bc
    Merge branch 'k2-fsa:master' into master Seung Hyun Lee 2024-06-19 10:27:30 +09:00
  • 2e05663fbb
    Add prepare.sh for KsponSpeech recipe. (#1656) Seung Hyun Lee 2024-06-18 17:54:39 +09:00
  • 7dda45c9bd Add prepare.sh for KsponSpeech recipe. whsqkaak 2024-06-18 16:40:30 +09:00
  • 1f5c0a87b9
    Add CI for ksponspeech (#1655) Fangjun Kuang 2024-06-16 19:15:09 +08:00
  • f04a3c3f08 Add CI for ksponspeech Fangjun Kuang 2024-06-16 17:04:39 +08:00
  • c13c7aa30b
    Add Streaming Zipformer-Transducer recipe for KsponSpeech (#1651) Seung Hyun Lee 2024-06-16 17:20:44 +09:00
  • 890eeec82c
    Add qwen-audio style model training: using whisper + qwen2 (#1652) Yuekai Zhang 2024-06-16 12:14:44 +08:00
  • 8b5aa993a4 update results Yuekai Zhang 2024-06-14 12:21:38 +08:00
  • c6e7344cc9 remove run.sh Yuekai Zhang 2024-06-14 12:00:28 +08:00
  • 6437fbf7d3 remove debug scripts Yuekai Zhang 2024-06-14 11:59:40 +08:00
  • 9ed428d7b1 add readme root 2024-06-14 03:55:57 +00:00
  • d1e31c7ac7 fix max_duration Yuekai Zhang 2024-06-14 09:43:52 +08:00
  • 618b686166 add sampler state_dict Yuekai Zhang 2024-06-14 09:41:08 +08:00
  • 7db5445d1e add unfreeze llm option root 2024-06-13 09:27:07 +00:00
  • dbe85c1f12 add speech io dataset root 2024-06-13 09:08:59 +00:00
  • 8226b628f4 add lora for second stage training root 2024-06-13 07:00:19 +00:00
  • 3195a55ac7 update test set with wenetspeech test meeting root 2024-06-12 01:50:57 +00:00
  • b26d3fa596 add logging root 2024-06-11 09:20:59 +00:00
  • 4ebccebcc0 removing debug log root 2024-06-11 09:17:31 +00:00
  • 271536248f fix decoding issue and padding to longest root 2024-06-11 09:04:29 +00:00
  • eb2c255e1e remove position ids root 2024-06-07 09:39:28 +00:00
  • 639feab4df update dataset with aishell 2 root 2024-06-07 07:49:38 +00:00
  • 8afb0d647f fix template root 2024-06-07 07:23:13 +00:00
  • 16f18080be update prompt for decoding Yuekai Zhang 2024-06-07 10:57:35 +08:00
  • 40e4ac480c change prompt Yuekai Zhang 2024-06-07 10:14:28 +08:00
  • 68b99f456f fix debug Yuekai Zhang 2024-06-07 09:53:40 +08:00
  • 8bbd06112a add decode log Yuekai Zhang 2024-06-06 22:01:17 +08:00
  • 412e926941 fix down sample method root 2024-06-06 13:25:29 +00:00
  • 796663066f mask unrelated labels root 2024-06-06 08:57:26 +00:00
  • 3ac27d5ad4 fix requirements Yuekai Zhang 2024-06-06 16:24:27 +08:00
  • 09ec0d6553 add requirements.txt Yuekai Zhang 2024-06-06 01:04:19 -07:00
  • 19b5b86f9b fix decoding issues Yuekai Zhang 2024-06-06 12:25:10 +08:00
  • 3dbbc29429 add decode file Yuekai Zhang 2024-06-04 22:56:57 -07:00
  • b5a906cbbd fix bugs Yuekai Zhang 2024-06-04 18:46:16 +08:00
  • e495c9d732 add whisper llm Yuekai Zhang 2024-06-03 23:10:58 -07:00
  • 3c970e7fad Add pretrained model link in RESULTS.md whsqkaak 2024-06-13 17:50:57 +09:00
  • ed9cc836ce Replace codes copied from librispeech recipe with symlink whsqkaak 2024-06-13 17:13:17 +09:00
  • 3b40d9bbb1
    Zipformer recipe for ReazonSpeech (#1611) Triplecq 2024-06-13 02:19:03 -04:00
  • db5c61d371 Reformat by black whsqkaak 2024-06-13 15:02:59 +09:00
  • ee21954c15 Add Streaming Zipformer-Transducer recipe for KsponSpeech whsqkaak 2024-06-13 14:41:55 +09:00
  • d5be739639
    add distill whisper results (#1648) Yuekai Zhang 2024-06-13 00:20:04 +08:00
  • 7b03f27183 deploy: 13f55d073513b3beaefdf0b7e16237b35199ca04 csukuangfj 2024-06-12 16:02:46 +00:00
  • 13f55d0735
    Add merge_tokens for ctc forced alignment (#1649) Fangjun Kuang 2024-06-12 17:45:13 +08:00
  • ec0389a3c1
    Add doc about FST-based CTC forced alignment. (#1482) Fangjun Kuang 2024-06-12 17:36:57 +08:00
  • cb21b878c0 Merge remote-tracking branch 'dan/master' into doc-force-alignment-kaldi Fangjun Kuang 2024-06-12 17:34:26 +08:00
  • f515aed0cc Finish kaldi-based approach Fangjun Kuang 2024-06-12 17:32:49 +08:00
  • 0936c80be7 add distill whisper results root 2024-06-12 08:43:05 +00:00
  • 68e202c52a minor fixes Fangjun Kuang 2024-06-12 15:38:23 +08:00
  • 4d5c1f2e60
    Remove inf from stored stats (#1647) Daniel Povey 2024-06-10 22:41:54 +08:00
  • 29bb04a51c Remove inf from stored stats Daniel Povey 2024-06-10 22:40:51 +08:00
  • 130a18cc10
    support torch 2.3.1 in docker (#1646) Fangjun Kuang 2024-06-06 22:27:29 +08:00
  • 428b5d2a74 support torch 2.3.1 in docker Fangjun Kuang 2024-06-06 14:19:04 +08:00
  • 06232dce2e WIP: Begin to add Contextual positional encoding Fangjun Kuang 2024-06-05 14:54:42 +08:00
  • b88062292b
    Typo fixes (#1643) Fangjun Kuang 2024-06-03 16:49:21 +08:00
  • 8460b632e0 Typo fixes Fangjun Kuang 2024-06-03 16:18:58 +08:00