remove academic and change to iic/CosyVoice_ttsfrd

.gitmodules

@@ -1,6 +1,3 @@
-[submodule "third_party/AcademiCodec"]
-	path = third_party/AcademiCodec
-	url = https://github.com/yangdongchao/AcademiCodec.git
 [submodule "third_party/Matcha-TTS"]
 	path = third_party/Matcha-TTS
 	url = https://github.com/shivammehta25/Matcha-TTS.git

README.md

@@ -33,7 +33,7 @@ sudo yum install sox sox-devel
 
 **Model download**
 
-We strongly recommand that you download our pretrained `CosyVoice-300M` `CosyVoice-300M-SFT` `CosyVoice-300M-Instruct` model and `speech_kantts_ttsfrd` resource.
+We strongly recommand that you download our pretrained `CosyVoice-300M` `CosyVoice-300M-SFT` `CosyVoice-300M-Instruct` model and `CosyVoice-ttsfrd` resource.
 
 If you are expert in this field, and you are only interested in training your own CosyVoice model from scratch, you can skip this step.
 
@@ -43,7 +43,7 @@ from modelscope import snapshot_download
 snapshot_download('iic/CosyVoice-300M', local_dir='pretrained_models/CosyVoice-300M')
 snapshot_download('iic/CosyVoice-300M-SFT', local_dir='pretrained_models/CosyVoice-300M-SFT')
 snapshot_download('iic/CosyVoice-300M-Instruct', local_dir='pretrained_models/CosyVoice-300M-Instruct')
-snapshot_download('speech_tts/speech_kantts_ttsfrd', local_dir='pretrained_models/speech_kantts_ttsfrd')
+snapshot_download('iic/CosyVoice-ttsfrd', local_dir='pretrained_models/CosyVoice-ttsfrd')
 ```
 
 ``` sh
@@ -52,12 +52,12 @@ mkdir -p pretrained_models
 git clone https://www.modelscope.cn/iic/CosyVoice-300M.git pretrained_models/CosyVoice-300M
 git clone https://www.modelscope.cn/iic/CosyVoice-300M-SFT.git pretrained_models/CosyVoice-300M-SFT
 git clone https://www.modelscope.cn/iic/CosyVoice-300M-Instruct.git pretrained_models/CosyVoice-300M-Instruct
-git clone https://www.modelscope.cn/speech_tts/speech_kantts_ttsfrd.git pretrained_models/speech_kantts_ttsfrd
+git clone https://www.modelscope.cn/iic/CosyVoice-ttsfrd.git pretrained_models/CosyVoice-ttsfrd
 ```
 
 Unzip `ttsfrd` resouce and install `ttsfrd` package
 ``` sh
-cd pretrained_models/speech_kantts_ttsfrd/
+cd pretrained_models/CosyVoice-ttsfrd/
 unzip resource.zip -d .
 pip install ttsfrd-0.3.6-cp38-cp38-linux_x86_64.whl
 ```

cosyvoice/cli/frontend.py

@@ -50,7 +50,7 @@ class CosyVoiceFrontEnd:
         self.inflect_parser = inflect.engine()
         self.frd = ttsfrd.TtsFrontendEngine()
         ROOT_DIR = os.path.dirname(os.path.abspath(__file__))
-        assert self.frd.initialize('{}/../../pretrained_models/speech_kantts_ttsfrd/resource'.format(ROOT_DIR)) is True, 'failed to initialize ttsfrd resource'
+        assert self.frd.initialize('{}/../../pretrained_models/CosyVoice-ttsfrd/resource'.format(ROOT_DIR)) is True, 'failed to initialize ttsfrd resource'
         self.frd.set_lang_type('pinyin')
         self.frd.enable_pinyin_mix(True)
         self.frd.set_breakmodel_index(1)

cosyvoice/hifigan/generator.py

@@ -27,8 +27,8 @@ from torch.nn.utils import weight_norm
 from torch.distributions.uniform import Uniform
 
 from cosyvoice.transformer.activation import Snake
-from academicodec.utils import get_padding
+from cosyvoice.utils.common import get_padding
-from academicodec.utils import init_weights
+from cosyvoice.utils.common import init_weights
 
 
 """hifigan based generator implementation.

cosyvoice/utils/common.py

@@ -91,3 +91,13 @@ def th_accuracy(pad_outputs: torch.Tensor, pad_targets: torch.Tensor,
         pad_pred.masked_select(mask) == pad_targets.masked_select(mask))
     denominator = torch.sum(mask)
     return (numerator / denominator).detach()
+
+
+def get_padding(kernel_size, dilation=1):
+    return int((kernel_size * dilation - dilation) / 2)
+
+
+def init_weights(m, mean=0.0, std=0.01):
+    classname = m.__class__.__name__
+    if classname.find("Conv") != -1:
+        m.weight.data.normal_(mean, std)

runtime/python/Dockerfile

@@ -5,8 +5,11 @@ WORKDIR /opt/CosyVoice
 
 RUN sed -i s@/archive.ubuntu.com/@/mirrors.aliyun.com/@g /etc/apt/sources.list
 RUN apt-get update -y
-RUN apt-get -y install python3-dev cmake python3-pip git
+RUN apt-get -y install python3-dev cmake python3-pip git unzip
 RUN git clone --recursive https://github.com/FunAudioLLM/CosyVoice.git
-RUN cd CosyVoice && pip3 install -r requirements.txt -i https://mirrors.aliyun.com/pypi/simple/ --trusted-host=mirrors.aliyun.com
+RUN cd CosyVoice && pip3 install --default-timeout=3600 -r requirements.txt -i https://mirrors.aliyun.com/pypi/simple/ --trusted-host=mirrors.aliyun.com
+RUN apt install git-lfs && git lfs install
+RUN cd CosyVoice && git clone https://www.modelscope.cn/iic/CosyVoice-ttsfrd.git pretrained_models/CosyVoice-ttsfrd
+RUN cd CosyVoice/pretrained_models/CosyVoice-ttsfrd && unzip resource.zip -d . && pip3 install ttsfrd-0.3.6-cp38-cp38-linux_x86_64.whl
 RUN cd CosyVoice/runtime/python && python3 -m grpc_tools.protoc -I. --python_out=. --grpc_python_out=. cosyvoice.proto
 CMD ["/bin/bash", "-c", "cd /opt/CosyVoice/CosyVoice/runtime/python && . ./path/sh && python3 server.py --port 50000 --max_conc 4 --model_dir speech_tts/CosyVoice-300M && sleep infinity"]