Mirrors/AudioGPT
1
0
Fork 0
mirror of https://github.com/AIGC-Audio/AudioGPT.git synced 2026-05-18 05:04:58 +02:00
Code Issues Packages Projects Releases Wiki Activity

update

Browse Source
This commit is contained in:
lmz
2023-04-30 23:29:53 +08:00
parent 526a05a268
commit f61a97c296
1 changed files with 3 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
audio-chatgpt.py
View File

@@ -184,7 +184,7 @@ class T2A:
def select_best_audio(self, prompt, wav_list): def select_best_audio(self, prompt, wav_list):
from wav_evaluation.models.CLAPWrapper import CLAPWrapper from wav_evaluation.models.CLAPWrapper import CLAPWrapper
clap_model = CLAPWrapper('useful_ckpts/CLAP/CLAP_weights_2022.pth', 'useful_ckpts/CLAP/config.yml', clap_model = CLAPWrapper('text_to_audio/Make_An_Audio/useful_ckpts/CLAP/CLAP_weights_2022.pth', 'text_to_audio/Make_An_Audio/useful_ckpts/CLAP/config.yml',
use_cuda=torch.cuda.is_available()) use_cuda=torch.cuda.is_available())
text_embeddings = clap_model.get_text_embeddings([prompt]) text_embeddings = clap_model.get_text_embeddings([prompt])
score_list = [] score_list = []
@@ -216,7 +216,7 @@ class I2A:
print("Initializing Make-An-Audio-Image to %s" % device) print("Initializing Make-An-Audio-Image to %s" % device)
self.device = device self.device = device
self.sampler = self._initialize_model('text_to_audio/Make_An_Audio/configs/img_to_audio/img2audio_args.yaml', 'text_to_audio/Make_An_Audio/useful_ckpts/ta54_epoch=000216.ckpt', device=device) self.sampler = self._initialize_model('text_to_audio/Make_An_Audio/configs/img_to_audio/img2audio_args.yaml', 'text_to_audio/Make_An_Audio/useful_ckpts/ta54_epoch=000216.ckpt', device=device)
self.vocoder = VocoderBigVGAN('text_to_audio/Make_An_Audio_img/vocoder/logs/bigv16k53w',device=device) self.vocoder = VocoderBigVGAN('text_to_audio/Make_An_Audio/vocoder/logs/bigv16k53w',device=device)
def _initialize_model(self, config, ckpt, device): def _initialize_model(self, config, ckpt, device):
config = OmegaConf.load(config) config = OmegaConf.load(config)
@@ -420,7 +420,7 @@ class Inpaint:
print("Initializing Make-An-Audio-inpaint to %s" % device) print("Initializing Make-An-Audio-inpaint to %s" % device)
self.device = device self.device = device
self.sampler = self._initialize_model_inpaint('text_to_audio/Make_An_Audio/configs/inpaint/txt2audio_args.yaml', 'text_to_audio/Make_An_Audio/useful_ckpts/inpaint7_epoch00047.ckpt') self.sampler = self._initialize_model_inpaint('text_to_audio/Make_An_Audio/configs/inpaint/txt2audio_args.yaml', 'text_to_audio/Make_An_Audio/useful_ckpts/inpaint7_epoch00047.ckpt')
self.vocoder = VocoderBigVGAN('text_to_audio/Make_An_Audio/vocoder/logs/bigv16k53',device=device) self.vocoder = VocoderBigVGAN('text_to_audio/Make_An_Audio/vocoder/logs/bigv16k53w',device=device)
self.cmap_transform = matplotlib.cm.viridis self.cmap_transform = matplotlib.cm.viridis
def _initialize_model_inpaint(self, config, ckpt): def _initialize_model_inpaint(self, config, ckpt):