import gradio as gr
import torch
import torchaudio
import librosa
from modules.commons import build_model, load_checkpoint, recursive_munch
import yaml
from hf_utils import load_custom_model_from_hf
import numpy as np
from pydub import AudioSegment
# edge tts
import asyncio
import traceback
import edge_tts
# edge tts需要外网访问，这里设置你的网络代理
edge_proxy = "http://192.168.31.69:7890"
# Load model and configuration
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
dit_checkpoint_path, dit_config_path = load_custom_model_from_hf("Plachta/Seed-VC",
"DiT_seed_v2_uvit_whisper_small_wavenet_bigvgan_pruned.pth",
"config_dit_mel_seed_uvit_whisper_small_wavenet.yml")
config = yaml.safe_load(open(dit_config_path, 'r'))
model_params = recursive_munch(config['model_params'])
model = build_model(model_params, stage='DiT')
hop_length = config['preprocess_params']['spect_params']['hop_length']
sr = config['preprocess_params']['sr']
# Load checkpoints
model, _, _, _ = load_checkpoint(model, None, dit_checkpoint_path,
load_only_params=True, ignore_modules=[], is_distributed=False)
for key in model:
model[key].eval()
model[key].to(device)
model.cfm.estimator.setup_caches(max_batch_size=1, max_seq_length=8192)
# Load additional modules
from modules.campplus.DTDNN import CAMPPlus
campplus_ckpt_path = load_custom_model_from_hf("funasr/campplus", "campplus_cn_common.bin", config_filename=None)
campplus_model = CAMPPlus(feat_dim=80, embedding_size=192)
campplus_model.load_state_dict(torch.load(campplus_ckpt_path, map_location="cpu"))
campplus_model.eval()
campplus_model.to(device)
from modules.bigvgan import bigvgan
bigvgan_model = bigvgan.BigVGAN.from_pretrained('nvidia/bigvgan_v2_22khz_80band_256x', use_cuda_kernel=False)
# remove weight norm in the model and set to eval mode
bigvgan_model.remove_weight_norm()
bigvgan_model = bigvgan_model.eval().to(device)
# whisper
from transformers import AutoFeatureExtractor, WhisperModel
whisper_name = model_params.speech_tokenizer.whisper_name if hasattr(model_params.speech_tokenizer,
'whisper_name') else "openai/whisper-small"
whisper_model = WhisperModel.from_pretrained(whisper_name, torch_dtype=torch.float16).to(device)
del whisper_model.decoder
whisper_feature_extractor = AutoFeatureExtractor.from_pretrained(whisper_name)
# Generate mel spectrograms
mel_fn_args = {
"n_fft": config['preprocess_params']['spect_params']['n_fft'],
"win_size": config['preprocess_params']['spect_params']['win_length'],
"hop_size": config['preprocess_params']['spect_params']['hop_length'],
"num_mels": config['preprocess_params']['spect_params']['n_mels'],
"sampling_rate": sr,
"fmin": 0,
"fmax": None,
"center": False
}
from modules.audio import mel_spectrogram
to_mel = lambda x: mel_spectrogram(x, **mel_fn_args)
# f0 conditioned model
dit_checkpoint_path, dit_config_path = load_custom_model_from_hf("Plachta/Seed-VC",
"DiT_seed_v2_uvit_whisper_base_f0_44k_bigvgan_pruned_ft_ema.pth",
"config_dit_mel_seed_uvit_whisper_base_f0_44k.yml")
config = yaml.safe_load(open(dit_config_path, 'r'))
model_params = recursive_munch(config['model_params'])
model_f0 = build_model(model_params, stage='DiT')
hop_length = config['preprocess_params']['spect_params']['hop_length']
sr = config['preprocess_params']['sr']
# Load checkpoints
model_f0, _, _, _ = load_checkpoint(model_f0, None, dit_checkpoint_path,
load_only_params=True, ignore_modules=[], is_distributed=False)
for key in model_f0:
model_f0[key].eval()
model_f0[key].to(device)
model_f0.cfm.estimator.setup_caches(max_batch_size=1, max_seq_length=8192)
# f0 extractor
from modules.rmvpe import RMVPE
model_path = load_custom_model_from_hf("lj1995/VoiceConversionWebUI", "rmvpe.pt", None)
rmvpe = RMVPE(model_path, is_half=False, device=device)
mel_fn_args_f0 = {
"n_fft": config['preprocess_params']['spect_params']['n_fft'],
"win_size": config['preprocess_params']['spect_params']['win_length'],
"hop_size": config['preprocess_params']['spect_params']['hop_length'],
"num_mels": config['preprocess_params']['spect_params']['n_mels'],
"sampling_rate": sr,
"fmin": 0,
"fmax": None,
"center": False
}
to_mel_f0 = lambda x: mel_spectrogram(x, **mel_fn_args_f0)
bigvgan_44k_model = bigvgan.BigVGAN.from_pretrained('nvidia/bigvgan_v2_44khz_128band_512x', use_cuda_kernel=False)
# remove weight norm in the model and set to eval mode
bigvgan_44k_model.remove_weight_norm()
bigvgan_44k_model = bigvgan_44k_model.eval().to(device)
def adjust_f0_semitones(f0_sequence, n_semitones):
factor = 2 ** (n_semitones / 12)
return f0_sequence * factor
def crossfade(chunk1, chunk2, overlap):
fade_out = np.cos(np.linspace(0, np.pi / 2, overlap)) ** 2
fade_in = np.cos(np.linspace(np.pi / 2, 0, overlap)) ** 2
if len(chunk2) < overlap:
chunk2[:overlap] = chunk2[:overlap] * fade_in[:len(chunk2)] + (chunk1[-overlap:] * fade_out)[:len(chunk2)]
else:
chunk2[:overlap] = chunk2[:overlap] * fade_in + chunk1[-overlap:] * fade_out
return chunk2
# streaming and chunk processing related params
overlap_frame_len = 16
bitrate = "320k"
@torch.no_grad()
@torch.inference_mode()
def voice_conversion(tts_text, tts_choice, speed, pitch, target, diffusion_steps, length_adjust, inference_cfg_rate,
f0_condition,
auto_f0_adjust,
pitch_shift):
print(length_adjust)
edge_audio = None
speed_str = f"{speed:+d}%"
pitch_str = f"{pitch:+d}Hz"
try:
print(tts_choice)
asyncio.run(
edge_tts.Communicate(
tts_text, "-".join(tts_choice.split("-")[:-1]), rate=speed_str, pitch=pitch_str, proxy=edge_proxy
).save(edge_output_filename)
)
edge_audio = gr.Audio(value=edge_output_filename)
except EOFError:
yield None, None, None
except:
info = traceback.format_exc()
print(info)
yield None, None, None
inference_module = model if not f0_condition else model_f0
mel_fn = to_mel if not f0_condition else to_mel_f0
bigvgan_fn = bigvgan_model if not f0_condition else bigvgan_44k_model
sr = 22050 if not f0_condition else 44100
hop_length = 256 if not f0_condition else 512
max_context_window = sr // hop_length * 30
overlap_wave_len = overlap_frame_len * hop_length
# Load audio
source_audio = librosa.load(edge_output_filename, sr=sr)[0]
ref_audio = librosa.load(target, sr=sr)[0]
# Process audio
source_audio = torch.tensor(source_audio).unsqueeze(0).float().to(device)
ref_audio = torch.tensor(ref_audio[:sr * 25]).unsqueeze(0).float().to(device)
# Resample
ref_waves_16k = torchaudio.functional.resample(ref_audio, sr, 16000)
converted_waves_16k = torchaudio.functional.resample(source_audio, sr, 16000)
# if source audio less than 30 seconds, whisper can handle in one forward
if converted_waves_16k.size(-1) <= 16000 * 30:
alt_inputs = whisper_feature_extractor([converted_waves_16k.squeeze(0).cpu().numpy()],
return_tensors="pt",
return_attention_mask=True,
sampling_rate=16000)
alt_input_features = whisper_model._mask_input_features(
alt_inputs.input_features, attention_mask=alt_inputs.attention_mask).to(device)
alt_outputs = whisper_model.encoder(
alt_input_features.to(whisper_model.encoder.dtype),
head_mask=None,
output_attentions=False,
output_hidden_states=False,
return_dict=True,
)
S_alt = alt_outputs.last_hidden_state.to(torch.float32)
S_alt = S_alt[:, :converted_waves_16k.size(-1) // 320 + 1]
else:
overlapping_time = 5 # 5 seconds
S_alt_list = []
buffer = None
traversed_time = 0
while traversed_time < converted_waves_16k.size(-1):
if buffer is None: # first chunk
chunk = converted_waves_16k[:, traversed_time:traversed_time + 16000 * 30]
else:
chunk = torch.cat(
[buffer, converted_waves_16k[:, traversed_time:traversed_time + 16000 * (30 - overlapping_time)]],
dim=-1)
alt_inputs = whisper_feature_extractor([chunk.squeeze(0).cpu().numpy()],
return_tensors="pt",
return_attention_mask=True,
sampling_rate=16000)
alt_input_features = whisper_model._mask_input_features(
alt_inputs.input_features, attention_mask=alt_inputs.attention_mask).to(device)
alt_outputs = whisper_model.encoder(
alt_input_features.to(whisper_model.encoder.dtype),
head_mask=None,
output_attentions=False,
output_hidden_states=False,
return_dict=True,
)
S_alt = alt_outputs.last_hidden_state.to(torch.float32)
S_alt = S_alt[:, :chunk.size(-1) // 320 + 1]
if traversed_time == 0:
S_alt_list.append(S_alt)
else:
S_alt_list.append(S_alt[:, 50 * overlapping_time:])
buffer = chunk[:, -16000 * overlapping_time:]
traversed_time += 30 * 16000 if traversed_time == 0 else chunk.size(-1) - 16000 * overlapping_time
S_alt = torch.cat(S_alt_list, dim=1)
ori_waves_16k = torchaudio.functional.resample(ref_audio, sr, 16000)
ori_inputs = whisper_feature_extractor([ori_waves_16k.squeeze(0).cpu().numpy()],
return_tensors="pt",
return_attention_mask=True)
ori_input_features = whisper_model._mask_input_features(
ori_inputs.input_features, attention_mask=ori_inputs.attention_mask).to(device)
with torch.no_grad():
ori_outputs = whisper_model.encoder(
ori_input_features.to(whisper_model.encoder.dtype),
head_mask=None,
output_attentions=False,
output_hidden_states=False,
return_dict=True,
)
S_ori = ori_outputs.last_hidden_state.to(torch.float32)
S_ori = S_ori[:, :ori_waves_16k.size(-1) // 320 + 1]
mel = mel_fn(source_audio.to(device).float())
mel2 = mel_fn(ref_audio.to(device).float())
target_lengths = torch.LongTensor([int(mel.size(2) * length_adjust)]).to(mel.device)
target2_lengths = torch.LongTensor([mel2.size(2)]).to(mel2.device)
feat2 = torchaudio.compliance.kaldi.fbank(ref_waves_16k,
num_mel_bins=80,
dither=0,
sample_frequency=16000)
feat2 = feat2 - feat2.mean(dim=0, keepdim=True)
style2 = campplus_model(feat2.unsqueeze(0))
if f0_condition:
F0_ori = rmvpe.infer_from_audio(ref_waves_16k[0], thred=0.03)
F0_alt = rmvpe.infer_from_audio(converted_waves_16k[0], thred=0.03)
F0_ori = torch.from_numpy(F0_ori).to(device)[None]
F0_alt = torch.from_numpy(F0_alt).to(device)[None]
voiced_F0_ori = F0_ori[F0_ori > 1]
voiced_F0_alt = F0_alt[F0_alt > 1]
log_f0_alt = torch.log(F0_alt + 1e-5)
voiced_log_f0_ori = torch.log(voiced_F0_ori + 1e-5)
voiced_log_f0_alt = torch.log(voiced_F0_alt + 1e-5)
median_log_f0_ori = torch.median(voiced_log_f0_ori)
median_log_f0_alt = torch.median(voiced_log_f0_alt)
# shift alt log f0 level to ori log f0 level
shifted_log_f0_alt = log_f0_alt.clone()
if auto_f0_adjust:
shifted_log_f0_alt[F0_alt > 1] = log_f0_alt[F0_alt > 1] - median_log_f0_alt + median_log_f0_ori
shifted_f0_alt = torch.exp(shifted_log_f0_alt)
if pitch_shift != 0:
shifted_f0_alt[F0_alt > 1] = adjust_f0_semitones(shifted_f0_alt[F0_alt > 1], pitch_shift)
else:
F0_ori = None
F0_alt = None
shifted_f0_alt = None
# Length regulation
cond, _, codes, commitment_loss, codebook_loss = inference_module.length_regulator(S_alt, ylens=target_lengths,
n_quantizers=3,
f0=shifted_f0_alt)
prompt_condition, _, codes, commitment_loss, codebook_loss = inference_module.length_regulator(S_ori,
ylens=target2_lengths,
n_quantizers=3,
f0=F0_ori)
max_source_window = max_context_window - mel2.size(2)
# split source condition (cond) into chunks
processed_frames = 0
generated_wave_chunks = []
# generate chunk by chunk and stream the output
while processed_frames < cond.size(1):
chunk_cond = cond[:, processed_frames:processed_frames + max_source_window]
is_last_chunk = processed_frames + max_source_window >= cond.size(1)
cat_condition = torch.cat([prompt_condition, chunk_cond], dim=1)
with torch.autocast(device_type=device.type, dtype=torch.float16):
# Voice Conversion
vc_target = inference_module.cfm.inference(cat_condition,
torch.LongTensor([cat_condition.size(1)]).to(mel2.device),
mel2, style2, None, diffusion_steps,
inference_cfg_rate=inference_cfg_rate)
vc_target = vc_target[:, :, mel2.size(-1):]
vc_wave = bigvgan_fn(vc_target.float())[0]
if processed_frames == 0:
if is_last_chunk:
output_wave = vc_wave[0].cpu().numpy()
generated_wave_chunks.append(output_wave)
output_wave = (output_wave * 32768.0).astype(np.int16)
mp3_bytes = AudioSegment(
output_wave.tobytes(), frame_rate=sr,
sample_width=output_wave.dtype.itemsize, channels=1
).export(format="mp3", bitrate=bitrate).read()
yield edge_audio, mp3_bytes, (sr, np.concatenate(generated_wave_chunks))
break
output_wave = vc_wave[0, :-overlap_wave_len].cpu().numpy()
generated_wave_chunks.append(output_wave)
previous_chunk = vc_wave[0, -overlap_wave_len:]
processed_frames += vc_target.size(2) - overlap_frame_len
output_wave = (output_wave * 32768.0).astype(np.int16)
mp3_bytes = AudioSegment(
output_wave.tobytes(), frame_rate=sr,
sample_width=output_wave.dtype.itemsize, channels=1
).export(format="mp3", bitrate=bitrate).read()
yield edge_audio, mp3_bytes, None
elif is_last_chunk:
output_wave = crossfade(previous_chunk.cpu().numpy(), vc_wave[0].cpu().numpy(), overlap_wave_len)
generated_wave_chunks.append(output_wave)
processed_frames += vc_target.size(2) - overlap_frame_len
output_wave = (output_wave * 32768.0).astype(np.int16)
mp3_bytes = AudioSegment(
output_wave.tobytes(), frame_rate=sr,
sample_width=output_wave.dtype.itemsize, channels=1
).export(format="mp3", bitrate=bitrate).read()
yield edge_audio, mp3_bytes, (sr, np.concatenate(generated_wave_chunks))
break
else:
output_wave = crossfade(previous_chunk.cpu().numpy(), vc_wave[0, :-overlap_wave_len].cpu().numpy(),
overlap_wave_len)
generated_wave_chunks.append(output_wave)
previous_chunk = vc_wave[0, -overlap_wave_len:]
processed_frames += vc_target.size(2) - overlap_frame_len
output_wave = (output_wave * 32768.0).astype(np.int16)
mp3_bytes = AudioSegment(
output_wave.tobytes(), frame_rate=sr,
sample_width=output_wave.dtype.itemsize, channels=1
).export(format="mp3", bitrate=bitrate).read()
yield edge_audio, mp3_bytes, None
# edge tts
edge_output_filename = "edge_output.mp3"
# 如果不使用下面的默认音色，可以通过以下代码获取全部edge tts音色，需要外网访问
# tts_voices_list = asyncio.get_event_loop().run_until_complete(edge_tts.list_voices(proxy=edge_proxy))
# tts_speakers = [f"{v['ShortName']}-{v['Gender']}" for v in tts_voices_list]
# 默认的edge tts音色
tts_speakers = ['zh-HK-HiuGaaiNeural-Female', 'zh-HK-HiuMaanNeural-Female', 'zh-HK-WanLungNeural-Male',
'zh-CN-XiaoxiaoNeural-Female', 'zh-CN-XiaoyiNeural-Female', 'zh-CN-YunjianNeural-Male',
'zh-CN-YunxiNeural-Male', 'zh-CN-YunxiaNeural-Male', 'zh-CN-YunyangNeural-Male',
'zh-CN-liaoning-XiaobeiNeural-Female', 'zh-TW-HsiaoChenNeural-Female', 'zh-TW-YunJheNeural-Male',
'zh-TW-HsiaoYuNeural-Female', 'zh-CN-shaanxi-XiaoniNeural-Female', 'en-AU-NatashaNeural-Female',
'en-AU-WilliamNeural-Male', 'en-CA-ClaraNeural-Female', 'en-CA-LiamNeural-Male', 'en-HK-SamNeural-Male',
'en-HK-YanNeural-Female', 'en-IN-NeerjaExpressiveNeural-Female', 'en-IN-NeerjaNeural-Female',
'en-IN-PrabhatNeural-Male', 'en-IE-ConnorNeural-Male', 'en-IE-EmilyNeural-Female',
'en-KE-AsiliaNeural-Female', 'en-KE-ChilembaNeural-Male', 'en-NZ-MitchellNeural-Male',
'en-NZ-MollyNeural-Female', 'en-NG-AbeoNeural-Male', 'en-NG-EzinneNeural-Female',
'en-PH-JamesNeural-Male', 'en-PH-RosaNeural-Female', 'en-SG-LunaNeural-Female',
'en-SG-WayneNeural-Male',
'en-US-AvaMultilingualNeural-Female', 'en-US-AndrewMultilingualNeural-Male',
'en-US-EmmaMultilingualNeural-Female', 'en-US-BrianMultilingualNeural-Male', 'en-US-AvaNeural-Female',
'en-US-AndrewNeural-Male', 'en-US-EmmaNeural-Female', 'en-US-BrianNeural-Male',
'en-ZA-LeahNeural-Female',
'en-ZA-LukeNeural-Male', 'en-TZ-ElimuNeural-Male', 'en-TZ-ImaniNeural-Female',
'en-GB-LibbyNeural-Female',
'en-GB-MaisieNeural-Female', 'en-GB-RyanNeural-Male', 'en-GB-SoniaNeural-Female',
'en-GB-ThomasNeural-Male', 'en-US-AnaNeural-Female', 'en-US-AriaNeural-Female',
'en-US-ChristopherNeural-Male', 'en-US-EricNeural-Male', 'en-US-GuyNeural-Male',
'en-US-JennyNeural-Female', 'en-US-MichelleNeural-Female', 'en-US-RogerNeural-Male',
'en-US-SteffanNeural-Male', 'ko-KR-HyunsuMultilingualNeural-Male', 'ko-KR-InJoonNeural-Male',
'ko-KR-SunHiNeural-Female']
def app():
with gr.Blocks(title="Seed VC Edge TTS") as demo:
with gr.Row():
gr.Markdown(value="""Seed VC Edge TTS
输入文本，使用Edge TTS获取待处理音频，再进行推理
""")
with gr.Row():
with gr.Column():
tts_text = gr.Textbox(label="Input Text / 输入文本", lines=5, value="这是一个示例文本")
tts_choice = gr.Dropdown(
label="Edge TTS Speaker / Edge TTS 音色",
choices=tts_speakers,
allow_custom_value=False,
value="zh-CN-YunjianNeural-Male"
)
tts_speed = gr.Slider(
minimum=-100,
maximum=100,
label="Edge TTS Speed(%) / Edge TTS 语速(%)",
value=-10,
step=5,
interactive=True
)
tts_pitch = gr.Slider(
minimum=-20,
maximum=20,
value=0,
label="Pitch Adjustment / 音调调整",
step=1
)
diffusion_steps = gr.Slider(minimum=1, maximum=200, value=10, step=1,
label="Diffusion Steps / 扩散步数",
info="10 by default, 50~100 for best quality / 默认为 10，50~100 为最佳质量")
length_adjust = gr.Slider(minimum=0.5, maximum=2.0, step=0.1, value=1.0,
label="Length Adjust / 长度调整",
info="<1.0 for speed-up speech, >1.0 for slow-down speech / <1.0 加速语速，>1.0 减慢语速")
inference_cfg_rate = gr.Slider(minimum=0.0, maximum=1.0, step=0.1, value=0.7,
label="Inference CFG Rate",
info="has subtle influence / 有微小影响")
f0_condition = gr.Checkbox(label="Use F0 conditioned model / 启用F0输入", value=False,
info="Must set to true for singing voice conversion / 歌声转换时必须勾选")
auto_f0_adjust = gr.Checkbox(label="Auto F0 adjust / 自动F0调整", value=True,
info="Roughly adjust F0 to match target voice. Only works when F0 conditioned model is used. / 粗略调整 F0 以匹配目标音色，仅在勾选 '启用F0输入' 时生效")
pitch_shift = gr.Slider(label='Pitch shift / 音调变换', minimum=-24, maximum=24, step=1, value=0,
info="Pitch shift in semitones, only works when F0 conditioned model is used / 半音数的音高变换，仅在勾选 '启用F0输入' 时生效")
with gr.Column():
reference_audio = gr.Audio(type="filepath", label="Reference Audio / 参考音频")
edge_tts_output = gr.Audio(type="filepath", label="Edge TTS Audio / Edge TTS 音频")
stream_audio_output = gr.Audio(label="Stream Output Audio / 流式输出", streaming=True, format='mp3')
full_audio_output = gr.Audio(label="Full Output Audio / 完整输出", streaming=False, format='wav')
submit_btn = gr.Button(value="推理", variant='primary')
submit_btn.click(voice_conversion,
inputs=[tts_text, tts_choice, tts_speed, tts_pitch, reference_audio, diffusion_steps,
length_adjust,
inference_cfg_rate,
f0_condition, auto_f0_adjust,
pitch_shift], outputs=[edge_tts_output, stream_audio_output, full_audio_output])
demo.queue(api_open=True).launch(debug=True, show_error=True)
if __name__ == "__main__":
app()