update(colab): add vol_aug

update(colab): add vol_aug
This commit is contained in:
YuriHead 2023-06-13 23:06:27 +08:00 committed by GitHub
commit a4d5a1b4cf
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 12 additions and 4 deletions

View File

@ -163,8 +163,8 @@
"#@markdown Although the pretrained model generally does not cause any copyright problems, please pay attention to it. For example, ask the author in advance, or the author has indicated the feasible use in the description clearly.\n",
"\n",
"download_pretrained_model = True #@param {type:\"boolean\"}\n",
"D_0_URL = \"https://huggingface.co/datasets/ms903/sovits4.0-768vec-layer12/resolve/main/sovits_768l12_pre_large_320k/clean_D_320000.pth\" #@param {type:\"string\"}\n",
"G_0_URL = \"https://huggingface.co/datasets/ms903/sovits4.0-768vec-layer12/resolve/main/sovits_768l12_pre_large_320k/clean_G_320000.pth\" #@param {type:\"string\"}\n",
"D_0_URL = \"https://huggingface.co/datasets/ms903/sovits4.0-768vec-layer12/resolve/main/sovits_768l12_pre_large_320k/clean_D_320000.pth\" #@param [\"https://huggingface.co/datasets/ms903/sovits4.0-768vec-layer12/resolve/main/sovits_768l12_pre_large_320k/clean_D_320000.pth\", \"https://huggingface.co/1asbgdh/sovits4.0-volemb-vec768/resolve/main/clean_D_320000.pth\"] {allow-input: true}\n",
"G_0_URL = \"https://huggingface.co/datasets/ms903/sovits4.0-768vec-layer12/resolve/main/sovits_768l12_pre_large_320k/clean_G_320000.pth\" #@param [\"https://huggingface.co/datasets/ms903/sovits4.0-768vec-layer12/resolve/main/sovits_768l12_pre_large_320k/clean_G_320000.pth\", \"https://huggingface.co/1asbgdh/sovits4.0-volemb-vec768/resolve/main/clean_G_320000.pth\"] {allow-input: true}\n",
"\n",
"download_pretrained_diffusion_model = True #@param {type:\"boolean\"}\n",
"diff_model_URL = \"https://huggingface.co/datasets/ms903/Diff-SVC-refactor-pre-trained-model/resolve/main/fix_pitch_add_vctk_600k/model_0.pt\" #@param {type:\"string\"}\n",
@ -173,10 +173,14 @@
"\n",
"if download_pretrained_model:\n",
" !curl -L {D_0_URL} -o logs/44k/D_0.pth\n",
" !md5sum logs/44k/D_0.pth\n",
" !curl -L {G_0_URL} -o logs/44k/G_0.pth\n",
" !md5sum logs/44k/G_0.pth\n",
"\n",
"if download_pretrained_diffusion_model:\n",
" !curl -L {diff_model_URL} -o logs/44k/diffusion/model_0.pt"
" !mkdir -p logs/44k/diffusion\n",
" !curl -L {diff_model_URL} -o logs/44k/diffusion/model_0.pt\n",
" !md5sum logs/44k/diffusion/model_0.pt"
]
},
{
@ -281,6 +285,8 @@
"%cd /content/so-vits-svc\n",
"\n",
"speech_encoder = \"vec768l12\" #@param [\"vec768l12\", \"vec256l9\", \"hubertsoft\", \"whisper-ppg\", \"whisper-ppg-large\"]\n",
"use_vol_aug = False #@param {type:\"boolean\"}\n",
"vol_aug = \"--vol_aug\" if use_vol_aug else \"\"\n",
"\n",
"from pretrain.meta import download_dict\n",
"download_dict = download_dict()\n",
@ -291,8 +297,9 @@
"import os\n",
"if not os.path.exists(output):\n",
" !curl -L {url} -o {output}\n",
" !md5sum {output}\n",
"\n",
"!python preprocess_flist_config.py --speech_encoder={speech_encoder}"
"!python preprocess_flist_config.py --speech_encoder={speech_encoder} {vol_aug}"
]
},
{
@ -320,6 +327,7 @@
" import os\n",
" if not os.path.exists(\"./pretrain/nsf_hifigan/model\"):\n",
" !curl -L https://github.com/openvpi/vocoders/releases/download/nsf-hifigan-v1/nsf_hifigan_20221211.zip -o nsf_hifigan_20221211.zip\n",
" !md5sum nsf_hifigan_20221211.zip\n",
" !unzip nsf_hifigan_20221211.zip\n",
" !rm -rf pretrain/nsf_hifigan\n",
" !mv -v nsf_hifigan pretrain\n",