babysor / babysor/MockingBird

tag 0.0.1 和最新版, 选择不同合成器模型,总会在不同的参数上出现 size mismatch, 导致语音合成失败

Open
#708 2 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

错误信息:
size mismatch for encoder.embedding.weight: copying a param with shape torch.Size([70, 512]) from checkpoint, the shape in current model is torch.Size([75, 512]).

请问
1)这个是什么原因导致的呢?
是编码器和合成器不匹配?

2),这个有办法从代码层面调整么?

3)这个问题怎样解决呢?

Contributor guide

No contributing guide indexed for this repository

Research direction

Start by reproducing the checkpoint-loading failure with the 0.0.1 tag and the reported synthesizer-model combinations. Compare the encoder embedding shapes in each checkpoint and current model, then determine whether compatible model pairing or a code-level adjustment is required. Done means the selected models load without size-mismatch errors and speech synthesis completes.

Written by the indexing model from the issue text.

Assessment

Tech stack
python, pytorch
Domain
machine-learning
Issue type
Bug
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.