fumiama/Retrieval-based-Voice-Conversion-WebUI
1
0
Fork 0
mirror of https://github.com/fumiama/Retrieval-based-Voice-Conversion-WebUI.git synced 2026-06-06 09:40:24 +08:00
Code Activity

optimize(infer.synthesizer): all modules inherit from one

Browse Source
This commit is contained in:
源文雨
2024-06-10 21:34:35 +09:00
parent b67050b2f7
commit e33ef19200
5 changed files with 127 additions and 231 deletions
Download Patch File Download Diff File Expand all files Collapse all files

24
infer/modules/train/train.py
View File

@@ -415,6 +415,7 @@ def train_and_evaluate(
for batch_idx, info in data_iterator:
# Data
## Unpack
pitch = pitchf = None
if hps.if_f0 == 1:
(
phone,
@@ -444,22 +445,13 @@ def train_and_evaluate(
# Calculate
with autocast(enabled=hps.train.fp16_run):
if hps.if_f0 == 1:
(
y_hat,
ids_slice,
x_mask,
z_mask,
(z, z_p, m_p, logs_p, m_q, logs_q),
) = net_g(phone, phone_lengths, pitch, pitchf, spec, spec_lengths, sid)
else:
(
y_hat,
ids_slice,
x_mask,
z_mask,
(z, z_p, m_p, logs_p, m_q, logs_q),
) = net_g(phone, phone_lengths, spec, spec_lengths, sid)
(
y_hat,
ids_slice,
x_mask,
z_mask,
(z, z_p, m_p, logs_p, m_q, logs_q),
) = net_g(phone, phone_lengths, spec, spec_lengths, sid, pitch, pitchf)
mel = spec_to_mel_torch(
spec,
hps.data.filter_length,

13
infer/modules/vc/pipeline.py
View File

@@ -290,10 +290,15 @@ class Pipeline(object):
feats = feats.to(feats0.dtype)
p_len = torch.tensor([p_len], device=self.device).long()
with torch.no_grad():
hasp = pitch is not None and pitchf is not None
arg = (feats, p_len, pitch, pitchf, sid) if hasp else (feats, p_len, sid)
audio1 = (net_g.infer(*arg)[0][0, 0]).data.cpu().float().numpy()
del arg
audio1 = (
net_g.infer(
feats,
p_len,
sid,
pitch=pitch,
pitchf=pitchf,
)[0, 0]
).data.cpu().float().numpy()
del feats, p_len, padding_mask
if torch.cuda.is_available():
torch.cuda.empty_cache()