“2024-08-26”版本间的差异

来自cslt Wiki
跳转至: 导航搜索
第49行: 第49行:
 
|Junming Yuan
 
|Junming Yuan
 
||
 
||
* Confirmed that the performance gap of the 10% is determined by the impact of GPUs.
+
* Confirmed that the performance gap of the 10% is determined by the impact of GPUs[https://z1et6d3xtb.feishu.cn/docx/PaATdHi26oEc0Pxovd4cSyp0nQ2].
 
** To fully reproduce the official model, it would take approximately 32 days.
 
** To fully reproduce the official model, it would take approximately 32 days.
 
* Investigate how to train Hubert with Mix-speech (in progress)
 
* Investigate how to train Hubert with Mix-speech (in progress)

2024年8月26日 (一) 09:39的版本

People This Week Next Week Task Tracking (DeadLine)
Dong Wang
Lantian Li
Ying Shi
Zhenghai You
Junming Yuan
  • Confirmed that the performance gap of the 10% is determined by the impact of GPUs[1].
    • To fully reproduce the official model, it would take approximately 32 days.
  • Investigate how to train Hubert with Mix-speech (in progress)
Chen Chen
Xiaolou Li
Zehua Liu
Pengqi Li
  • Extend Proposal for 'HOW PHONEMES CONTRIBUTE TO DEEP SPEAKER MODELS?'[2]
    • Reviewing code, paper.
    • Analyzing di-phones in Audio-Mnist.
    • Start exp with TIMIT dataset.
Wan Lin
Tianhao Wang
  • AudioSep reproducing
  • IS24 poster
Zhenyu Zhou
Junhui Chen
Jiaying Wang
Yu Zhang
Wenqiang Du
Yang Wei
Lily
Turi
  • Added more sections to the draft paper
    • Need to refine and do more experiments
Yue Gu
Qi Qu