Spaces:
Paused
Paused
Commit ·
20de2d7
1
Parent(s): 8d78fb6
remove extra arguments
Browse files- mcp_server.py +12 -31
mcp_server.py
CHANGED
|
@@ -748,8 +748,6 @@ def align_songs_by_bpm_mcp(
|
|
| 748 |
def create_medley_mcp(
|
| 749 |
vocals_path: str,
|
| 750 |
instrumental_path: str,
|
| 751 |
-
vocals_gain: float = 0.6,
|
| 752 |
-
instrumental_gain: float = 1.2,
|
| 753 |
output_format: str = "wav",
|
| 754 |
) -> str:
|
| 755 |
"""
|
|
@@ -762,8 +760,6 @@ def create_medley_mcp(
|
|
| 762 |
|
| 763 |
vocals_path: Path to the vocals audio file or URL (supports common formats: WAV, MP3, FLAC, M4A)
|
| 764 |
instrumental_path: Path to the instrumental audio file or URL (supports common formats: WAV, MP3, FLAC, M4A)
|
| 765 |
-
vocals_gain: Gain multiplier for vocals (default: 0.6, typical range: 0.5-1.0)
|
| 766 |
-
instrumental_gain: Gain multiplier for instrumental (default: 1.2, typical range: 0.5-2.0)
|
| 767 |
output_format: Output format for medley ('wav' or 'mp3', default: 'wav')
|
| 768 |
|
| 769 |
Returns:
|
|
@@ -782,6 +778,9 @@ def create_medley_mcp(
|
|
| 782 |
- Handles different track lengths by extending shorter with silence
|
| 783 |
- Uses high-quality audio processing algorithms
|
| 784 |
"""
|
|
|
|
|
|
|
|
|
|
| 785 |
try:
|
| 786 |
result = create_medley(
|
| 787 |
vocals_path=vocals_path,
|
|
@@ -1132,12 +1131,6 @@ def replace_section_mcp(
|
|
| 1132 |
def replace_voice_mcp(
|
| 1133 |
source_audio_path: str,
|
| 1134 |
target_audio_path: str,
|
| 1135 |
-
diffusion_steps: int = 10,
|
| 1136 |
-
length_adjust: float = 1.0,
|
| 1137 |
-
inference_cfg_rate: float = 0.7,
|
| 1138 |
-
f0_condition: bool = False,
|
| 1139 |
-
auto_f0_adjust: bool = True,
|
| 1140 |
-
pitch_shift: int = 0,
|
| 1141 |
) -> str:
|
| 1142 |
"""
|
| 1143 |
Replace voice in source audio with voice from target audio using Seed-VC.
|
|
@@ -1149,12 +1142,6 @@ def replace_voice_mcp(
|
|
| 1149 |
Args:
|
| 1150 |
source_audio_path: Path to the source audio file or URL (voice to be replaced)
|
| 1151 |
target_audio_path: Path to the target audio file or URL (voice to use)
|
| 1152 |
-
diffusion_steps: Number of diffusion steps for inference (default: 10)
|
| 1153 |
-
length_adjust: Length adjustment factor (default: 1.0)
|
| 1154 |
-
inference_cfg_rate: Classifier-free guidance rate (default: 0.7)
|
| 1155 |
-
f0_condition: Whether to use F0 conditioning (default: False)
|
| 1156 |
-
auto_f0_adjust: Whether to auto-adjust F0 (default: True)
|
| 1157 |
-
pitch_shift: Pitch shift in semitones (default: 0)
|
| 1158 |
|
| 1159 |
Returns:
|
| 1160 |
Path to the generated voice-replaced audio file
|
|
@@ -1178,6 +1165,14 @@ def replace_voice_mcp(
|
|
| 1178 |
- Applies voice characteristics from target audio
|
| 1179 |
- Processing time depends on diffusion steps and audio length
|
| 1180 |
"""
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1181 |
return replace_voice_wrapper(
|
| 1182 |
source_audio_path=source_audio_path,
|
| 1183 |
target_audio_path=target_audio_path,
|
|
@@ -1452,12 +1447,6 @@ def create_interface() -> gr.TabbedInterface:
|
|
| 1452 |
inputs=[
|
| 1453 |
gr.Audio(type="filepath", label="Vocals Stem", sources=["upload"]),
|
| 1454 |
gr.Audio(type="filepath", label="Instrumental Stem", sources=["upload"]),
|
| 1455 |
-
gr.Number(
|
| 1456 |
-
value=0.6, label="Vocals Gain", minimum=0.1, maximum=3.0, step=0.1
|
| 1457 |
-
),
|
| 1458 |
-
gr.Number(
|
| 1459 |
-
value=1.2 , label="Instrumental Gain", minimum=0.1, maximum=3.0, step=0.1
|
| 1460 |
-
),
|
| 1461 |
gr.Dropdown(
|
| 1462 |
choices=["wav", "mp3"],
|
| 1463 |
value="wav",
|
|
@@ -1703,15 +1692,7 @@ def create_interface() -> gr.TabbedInterface:
|
|
| 1703 |
type="filepath",
|
| 1704 |
label="Target Audio (voice to use) - Local file or URL",
|
| 1705 |
sources=["upload"],
|
| 1706 |
-
)
|
| 1707 |
-
gr.Number(value=10, label="Diffusion Steps", minimum=1, maximum=50),
|
| 1708 |
-
gr.Number(value=1.0, label="Length Adjust", minimum=0.1, maximum=3.0),
|
| 1709 |
-
gr.Number(value=0.7, label="Inference CFG Rate", minimum=0.0, maximum=1.0),
|
| 1710 |
-
gr.Checkbox(value=False, label="F0 Condition"),
|
| 1711 |
-
gr.Checkbox(value=True, label="Auto F0 Adjust"),
|
| 1712 |
-
gr.Number(
|
| 1713 |
-
value=0, label="Pitch Shift (semitones)", minimum=-12, maximum=12
|
| 1714 |
-
),
|
| 1715 |
],
|
| 1716 |
outputs=gr.Audio(label="Voice-Replaced Audio", type="filepath"),
|
| 1717 |
title="Voice Replacement with Seed-VC",
|
|
|
|
| 748 |
def create_medley_mcp(
|
| 749 |
vocals_path: str,
|
| 750 |
instrumental_path: str,
|
|
|
|
|
|
|
| 751 |
output_format: str = "wav",
|
| 752 |
) -> str:
|
| 753 |
"""
|
|
|
|
| 760 |
|
| 761 |
vocals_path: Path to the vocals audio file or URL (supports common formats: WAV, MP3, FLAC, M4A)
|
| 762 |
instrumental_path: Path to the instrumental audio file or URL (supports common formats: WAV, MP3, FLAC, M4A)
|
|
|
|
|
|
|
| 763 |
output_format: Output format for medley ('wav' or 'mp3', default: 'wav')
|
| 764 |
|
| 765 |
Returns:
|
|
|
|
| 778 |
- Handles different track lengths by extending shorter with silence
|
| 779 |
- Uses high-quality audio processing algorithms
|
| 780 |
"""
|
| 781 |
+
vocals_gain: float = 0.6
|
| 782 |
+
instrumental_gain: float = 1.2
|
| 783 |
+
|
| 784 |
try:
|
| 785 |
result = create_medley(
|
| 786 |
vocals_path=vocals_path,
|
|
|
|
| 1131 |
def replace_voice_mcp(
|
| 1132 |
source_audio_path: str,
|
| 1133 |
target_audio_path: str,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1134 |
) -> str:
|
| 1135 |
"""
|
| 1136 |
Replace voice in source audio with voice from target audio using Seed-VC.
|
|
|
|
| 1142 |
Args:
|
| 1143 |
source_audio_path: Path to the source audio file or URL (voice to be replaced)
|
| 1144 |
target_audio_path: Path to the target audio file or URL (voice to use)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1145 |
|
| 1146 |
Returns:
|
| 1147 |
Path to the generated voice-replaced audio file
|
|
|
|
| 1165 |
- Applies voice characteristics from target audio
|
| 1166 |
- Processing time depends on diffusion steps and audio length
|
| 1167 |
"""
|
| 1168 |
+
|
| 1169 |
+
diffusion_steps: int = 35
|
| 1170 |
+
length_adjust: float = 1.0
|
| 1171 |
+
inference_cfg_rate: float = 0.5
|
| 1172 |
+
f0_condition: bool = False
|
| 1173 |
+
auto_f0_adjust: bool = False
|
| 1174 |
+
pitch_shift: int = 0
|
| 1175 |
+
|
| 1176 |
return replace_voice_wrapper(
|
| 1177 |
source_audio_path=source_audio_path,
|
| 1178 |
target_audio_path=target_audio_path,
|
|
|
|
| 1447 |
inputs=[
|
| 1448 |
gr.Audio(type="filepath", label="Vocals Stem", sources=["upload"]),
|
| 1449 |
gr.Audio(type="filepath", label="Instrumental Stem", sources=["upload"]),
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1450 |
gr.Dropdown(
|
| 1451 |
choices=["wav", "mp3"],
|
| 1452 |
value="wav",
|
|
|
|
| 1692 |
type="filepath",
|
| 1693 |
label="Target Audio (voice to use) - Local file or URL",
|
| 1694 |
sources=["upload"],
|
| 1695 |
+
)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1696 |
],
|
| 1697 |
outputs=gr.Audio(label="Voice-Replaced Audio", type="filepath"),
|
| 1698 |
title="Voice Replacement with Seed-VC",
|