Skip to content

Conversation

@Noah-Flame
Copy link
Contributor

No description provided.

@ftshijt
Copy link
Member

ftshijt commented Feb 25, 2021

Please fix the CI bugs

@ftshijt ftshijt added the enhancement New feature or request label Mar 4, 2021
@ftshijt
Copy link
Member

ftshijt commented Mar 12, 2021

Please fix the conflict as soon as possible, thanks!

SVS/bin/train.py Outdated
help="how many learning_steps use lr_decay",
)

parser.add_argument("--vocoder_category", type=str, help="The vocoder to be used (e.g., pyworld).")
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

please uniform the interface with current vocoder

self.d_model = d_model
self.feat_dim_pw = feat_dim_pw
self.vocoder_category = vocoder_category
self.output_fc_pw = nn.Linear(d_output, self.feat_dim_pw*2+1)
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

this depends on the vocoder category right? add a if here

align_root_path=args.train_align,
pitch_beat_root_path=args.train_pitch,
wav_root_path=args.train_wav,
pw_f0_root_path=args.train_pw_f0,
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think we can uniform this right? Don't need to check vocoder category here, but check it inside the initialization of SVSDataset (init)

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

enhancement New feature or request

Projects

None yet

Development

Successfully merging this pull request may close these issues.

7 participants