“2024-07-01”版本间的差异

来自cslt Wiki
跳转至： 导航、 搜索
 
（18位用户的32个中间修订版本未显示）
第6行： 第6行：
 
|Dong Wang
 
|Dong Wang
 
||
 
||
−
*
+
* A trial talk in Retired Professor Association of THU.
 +
* Check Aigraph slides
 
||
 
||
 
*
 
*
第17行： 第18行：
 
|Lantian Li
 
|Lantian Li
 
||
 
||
−
*
+
* GPU status [https://z1et6d3xtb.feishu.cn/wiki/XGcGwRK5viJmpRkjH9AczIhynCh]
 +
* Projects
 +
** AED -> model miniaturization/streaming
 +
** TSE -> 1st phase delivery (this week)
 +
** VSR -> no progress
 +
** Finance -> no progress
 +
* Papers
 +
** NeuralScoring
 +
* AI graph
 +
** Slides checking (10/50)
 +
** High school handbook (2/40)
 
||
 
||
 
*
 
*
第23行： 第34行：
 
*
 
*
 
|-
 
|-
 +
  
  
第28行： 第40行：
 
|Ying Shi
 
|Ying Shi
 
||
 
||
−
*  
+
* Text enroll Keyword Spotting
 +
** Training with sufficient data augmentation [failed]
 +
*  Cohort Conditional Chain Overlap ASR
 +
** Fix the bug about the positional embedding of Condition, the performance is still not good
 +
* Review several NC paper
 
||
 
||
−
*
+
* Text enroll Keyword Spotting
 +
** Employ Homo loss
 +
*  Cohort Conditional Chain Overlap ASR
 +
** Check the  SID performance of the current speaker embedding model on the training dataset
 +
** Reproduce the previous Cohort-SOT methods on the training dataset
 +
* Finish the  SPL response
 
||
 
||
 
*
 
*
第39行： 第60行：
 
|Zhenghai You
 
|Zhenghai You
 
||
 
||
−
*
+
* Complete the project deliverables
 
||
 
||
 
*
 
*
第49行： 第70行：
 
|Junming Yuan
 
|Junming Yuan
 
||
 
||
−
* find the Bug in SSL model finetuning experiment with multi-lingual  
+
* find the bug in SSL model finetuning experiment with multi-lingual  
 
** double check result in [https://z1et6d3xtb.feishu.cn/docx/B2Etd5sKwo9jiyx30Vwc9tbXnue]
 
** double check result in [https://z1et6d3xtb.feishu.cn/docx/B2Etd5sKwo9jiyx30Vwc9tbXnue]
−
** The results demonstrate the strong robustness of MT in the few-shot mix scenario.
+
** The results need check again.
 
||
 
||
 
*
 
*
第62行： 第83行：
 
|Chen Chen
 
|Chen Chen
 
||
 
||
−
*
+
* graduate
 +
* entropy (with avhubert/hubert for LRS3/GRID)
 
||
 
||
−
*
+
* finish ISCSLP paper
 
||
 
||
 
*
 
*
第73行： 第95行：
 
|Xiaolou Li
 
|Xiaolou Li
 
||
 
||
−
*
+
* Some exps on Bimamba (parameters search) [https://z1et6d3xtb.feishu.cn/docx/VTe2dvLpwoQX4DxNfTccNyZ3nsh?from=from_copylink]
 +
* Paper reading
 +
** MLLM survey, pre-train, fine-tune and modality alignment
 
||
 
||
 
*
 
*
第84行： 第108行：
 
|Zehua Liu
 
|Zehua Liu
 
||
 
||
−
*
+
*VSP-LLM Reproduce（LRS3(30h) wer:36.32 > wer: 29.2）[https://z1et6d3xtb.feishu.cn/docx/A10Ed7SceozES1xwK49cJGpHned?from=from_copylink]
 +
*still need work on the code
 
||
 
||
 
*
 
*
第95行： 第120行：
 
|Pengqi Li
 
|Pengqi Li
 
||
 
||
−
*
+
* Modify NC-paper
 +
* Expand experiments for supervise(ASP)(finished code)
 
||
 
||
 
*
 
*
第106行： 第132行：
 
|Wan Lin
 
|Wan Lin
 
||
 
||
−
*
+
* Neural Scoring
 +
** trials test: vox1-e, vox1-h
 +
** cn: ns(all-genres and 3-genres fine-tuning)
 +
** variable-chunk training(2-10s): <br /> training, looks similar to the results of 4/6s
 +
** weekly report
 
||
 
||
 
*
 
*
第117行： 第147行：
 
|Tianhao Wang
 
|Tianhao Wang
 
||
 
||
−
*
+
* Neural Scoring [https://z1et6d3xtb.feishu.cn/docx/BywjdkGvNou12sxQ4dAcxYa9noh]:
 +
** vox: vox1-e, vox1-h test [https://z1et6d3xtb.feishu.cn/docx/KqXyd7Ok1ofyFIxMvdJcrLcxnDb]
 +
** cn: three genres fine-tuning: resnet and ns, 2s and 4s
 +
** weekly report
 
||
 
||
 
*
 
*
第128行： 第161行：
 
|Zhenyu Zhou
 
|Zhenyu Zhou
 
||
 
||
−
*
+
*Huawei Project Submission
 
||
 
||
 
*
 
*
第139行： 第172行：
 
|Junhui Chen
 
|Junhui Chen
 
||
 
||
−
*
+
* Neural Scoring [https://z1et6d3xtb.feishu.cn/docx/BywjdkGvNou12sxQ4dAcxYa9noh]:
 +
** vox: vox1-e, vox1-h test [https://z1et6d3xtb.feishu.cn/docx/KqXyd7Ok1ofyFIxMvdJcrLcxnDb]
 +
** vox: one transformer encoder layer training
 +
** cn: three genres fine-tuning: resnet and ns, 2s and 4s
 
||
 
||
 
*
 
*
第150行： 第186行：
 
|Jiaying Wang
 
|Jiaying Wang
 
||
 
||
−
*
+
* Preliminary validation: cohort works[https://z1et6d3xtb.feishu.cn/docx/IeIydyjzJozfUWxvmNfcWASBngg]
 
||
 
||
 
*
 
*
第161行： 第197行：
 
|Yu Zhang
 
|Yu Zhang
 
||
 
||
−
*
+
* Finance
 +
** Data Collection (2015 - 2019 HS300 stocks)
 +
* AED
 +
** 8k 2s CNN model training and Window inference code
 
||
 
||
 
*
 
*
第172行： 第211行：
 
|Wenqiang Du
 
|Wenqiang Du
 
||
 
||
−
*
+
* Quantified the kws model of Aibabel
 +
* Training dialect models for AIbabel（ Uyghur language）
 
||
 
||
−
*
+
* Train a joint model for Chinese, Uyghur, and Kazakh languages
 
||
 
||
 
*
 
*
第183行： 第223行：
 
|Yang Wei
 
|Yang Wei
 
||
 
||
−
*
+
* AIbabel
 +
** Learn to train and test KWS model
 
||
 
||
 
*
 
*
第193行： 第234行：
 
|Lily
 
|Lily
 
||
 
||
−
*
+
* Assisted with design of AI courses for primary, high school
 +
* Some chores about 'AIradiance'
 +
** Prepare application for Beijing Science and Technology Progress Award
 +
** Live broadcast
 +
** Prepare the content for the daily sign poster for July, August
 
||
 
||
 
*
 
*
第203行： 第248行：
 
|Turi
 
|Turi
 
||
 
||
−
*
+
* Thesis Proposal Defense
 +
* Data Collection
 +
** 31K utterences so far
 
||
 
||
 
*
 
*
第221行： 第268行：
 
|Qi Qu
 
|Qi Qu
 
||
 
||
−
*  
+
* AED:
 +
** CED + Linear: c/jni/python lib development and test.
 
||
 
||
−
*
+
* AED:
 +
** CED: Linear to be trained on data.
 +
** On-device demo.
 
||
 
||
 
*   
 
*   
 
|-
 
|-

2024年7月1日 (一) 10:58的最后版本

People This Week Next Week Task Tracking (DeadLine)
Dong Wang
  • A trial talk in Retired Professor Association of THU.
  • Check Aigraph slides
Lantian Li
  • GPU status [1]
  • Projects
    • AED -> model miniaturization/streaming
    • TSE -> 1st phase delivery (this week)
    • VSR -> no progress
    • Finance -> no progress
  • Papers
    • NeuralScoring
  • AI graph
    • Slides checking (10/50)
    • High school handbook (2/40)
Ying Shi
  • Text enroll Keyword Spotting
    • Training with sufficient data augmentation [failed]
  • Cohort Conditional Chain Overlap ASR
    • Fix the bug about the positional embedding of Condition, the performance is still not good
  • Review several NC paper
  • Text enroll Keyword Spotting
    • Employ Homo loss
  • Cohort Conditional Chain Overlap ASR
    • Check the SID performance of the current speaker embedding model on the training dataset
    • Reproduce the previous Cohort-SOT methods on the training dataset
  • Finish the SPL response
Zhenghai You
  • Complete the project deliverables
Junming Yuan
  • find the bug in SSL model finetuning experiment with multi-lingual
    • double check result in [2]
    • The results need check again.
Chen Chen
  • graduate
  • entropy (with avhubert/hubert for LRS3/GRID)
  • finish ISCSLP paper
Xiaolou Li
  • Some exps on Bimamba (parameters search) [3]
  • Paper reading
    • MLLM survey, pre-train, fine-tune and modality alignment
Zehua Liu
  • VSP-LLM Reproduce（LRS3(30h) wer:36.32 > wer: 29.2）[4]
  • still need work on the code
Pengqi Li
  • Modify NC-paper
  • Expand experiments for supervise(ASP)(finished code)
Wan Lin
  • Neural Scoring
    • trials test: vox1-e, vox1-h
    • cn: ns(all-genres and 3-genres fine-tuning)
    • variable-chunk training(2-10s):
      training, looks similar to the results of 4/6s
    • weekly report
Tianhao Wang
  • Neural Scoring [5]:
    • vox: vox1-e, vox1-h test [6]
    • cn: three genres fine-tuning: resnet and ns, 2s and 4s
    • weekly report
Zhenyu Zhou
  • Huawei Project Submission
Junhui Chen
  • Neural Scoring [7]:
    • vox: vox1-e, vox1-h test [8]
    • vox: one transformer encoder layer training
    • cn: three genres fine-tuning: resnet and ns, 2s and 4s
Jiaying Wang
  • Preliminary validation: cohort works[9]
Yu Zhang
  • Finance
    • Data Collection (2015 - 2019 HS300 stocks)
  • AED
    • 8k 2s CNN model training and Window inference code
Wenqiang Du
  • Quantified the kws model of Aibabel
  • Training dialect models for AIbabel（ Uyghur language）
  • Train a joint model for Chinese, Uyghur, and Kazakh languages
Yang Wei
  • AIbabel
    • Learn to train and test KWS model
Lily
  • Assisted with design of AI courses for primary, high school
  • Some chores about 'AIradiance'
    • Prepare application for Beijing Science and Technology Progress Award
    • Live broadcast
    • Prepare the content for the daily sign poster for July, August
Turi
  • Thesis Proposal Defense
  • Data Collection
    • 31K utterences so far
Yue Gu
  • Prepare the live content
  • writing paper
Qi Qu
  • AED:
    • CED + Linear: c/jni/python lib development and test.
  • AED:
    • CED: Linear to be trained on data.
    • On-device demo.
取自“http://index.cslt.org/mediawiki/index.php?title=2024-07-01&oldid=41973”