nodes/wildminder-ComfyUI-VoxCPM
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

fix bug with stereo files

Browse Source
This commit is contained in:
WildAi
2025-12-12 20:50:50 +03:00
parent 45a209c489
commit 09479557cd
1 changed files with 3 additions and 15 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/voxcpm/model/voxcpm.py
+3 -15
View File
@@ -126,7 +126,6 @@ class VoxCPMModel(nn.Module):
self.device = "mps"
else:
self.device = "cpu"
# print(f"Running on device: {self.device}, dtype: {self.config.dtype}")
# Text-Semantic LM
self.base_lm = MiniCPMModel(config.lm_config)
@@ -337,14 +336,8 @@ class VoxCPMModel(nn.Module):
# Mix down to mono if needed
if audio.dim() > 1 and audio.size(0) > 1:
audio = audio.mean(dim=0, keepdim=True)
elif audio.dim() > 1 and audio.size(1) > 1 and prompt_waveform is not None:
# Handle ComfyUI tensor shape [batch, channels, samples] or [channels, samples]
if audio.size(0) == 1 and audio.size(1) > 1:
# e.g. [1, 2, N] -> [1, 1, N]
audio = audio.mean(dim=1, keepdim=True)
elif audio.size(0) > 1:
# [2, N] -> [1, N]
audio = audio.mean(dim=0, keepdim=True)
# Note: If Mono [1, Samples], the above check fails (1 is not > 1), so it remains [1, Samples]. Correct.
if sr != self.sample_rate:
audio = torchaudio.functional.resample(audio, sr, self.sample_rate)
@@ -449,14 +442,9 @@ class VoxCPMModel(nn.Module):
else:
audio, sr = torchaudio.load(prompt_wav_path)
# Simplified standard mix logic here as well
if audio.dim() > 1 and audio.size(0) > 1:
audio = audio.mean(dim=0, keepdim=True)
elif audio.dim() > 1 and audio.size(1) > 1 and prompt_waveform is not None:
# Handle ComfyUI tensor shape
if audio.size(0) == 1 and audio.size(1) > 1:
audio = audio.mean(dim=1, keepdim=True)
elif audio.size(0) > 1:
audio = audio.mean(dim=0, keepdim=True)
if sr != self.sample_rate:
audio = torchaudio.functional.resample(audio, sr, self.sample_rate)