babysor / babysor/MockingBird

请问如何对多个数据集进行合并,并进行训练以降低Loss?

Open
#135 3 comments 0 reactions 0 assignees View on GitHub
Dominant language
Python
Stars
36.9k
Forks
5.2k
PR merge metrics
No merged PRs in 30d

Description

请问如何对多个数据集进行合并,并进行训练以降低Loss?
我已经下载了三个数据集

├──aidatatang_200zh
│ ├──corpus
│ │ ├──dev
│ │ ├──test
│ │ └──train
│ └──transcript
├──data_aishell3
│ ├──test
│ │ └──wav
│ └──train
│ └──wav
├──MAGICDATA
│ └──train
│ ├──14_3466
│ ├──14_3664
│ ........
│ └──5_970
└──SV2TTS
└──synthesizer
├──audio
├──embeds
└──mels
现在我应该如何将这三个混合起来?

Contributor guide

No contributing guide indexed for this repository

Research direction

The issue names the aidatatang_200zh, data_aishell3, MAGICDATA, and SV2TTS dataset directories but no repository file or test. Start by locating the dataset-loading and training entry points, then establish the expected data format and document a reproducible multi-dataset training workflow with a measurable loss outcome.

Written by the indexing model from the issue text.

Assessment

Tech stack
python, pytorch
Domain
data, machine-learning
Issue type
Documentation
Difficulty
5/5
Estimated time
Over a week
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
18/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.