debug

app.py

@@ -54,7 +54,7 @@ sys.path.append('{}/third_party/Matcha-TTS'.format(ROOT_DIR))
 from huggingface_hub import snapshot_download
 snapshot_download('FunAudioLLM/CosyVoice2-0.5B', local_dir='pretrained_models/CosyVoice2-0.5B')
 snapshot_download('FunAudioLLM/CosyVoice-ttsfrd', local_dir='pretrained_models/CosyVoice-ttsfrd')
-os.system('cd pretrained_models/CosyVoice-ttsfrd/ && pip install ttsfrd_dependency-0.1-py3-none-any.whl && pip install ttsfrd-0.4.2-cp310-cp310-linux_x86_64.whl && tar -xvf resource.tar')
+os.system('cd pretrained_models/CosyVoice-ttsfrd/ && unzip resource.zip && pip install ttsfrd_dependency-0.1-py3-none-any.whl && pip install ttsfrd-0.4.2-cp310-cp310-linux_x86_64.whl && tar -xvf resource.tar')
 
 from cosyvoice.cli.cosyvoice import CosyVoice2
 from cosyvoice.utils.file_utils import load_wav, logging

cosyvoice/cli/cosyvoice.py

@@ -140,6 +140,7 @@ class CosyVoice2(CosyVoice):
             model_dir = snapshot_download(model_dir)
         with open('{}/cosyvoice.yaml'.format(model_dir), 'r') as f:
             configs = load_hyperpyyaml(f, overrides={'qwen_pretrain_path': os.path.join(model_dir, 'CosyVoice-BlankEN')})
+        print(f"Loading configs:{configs}")
         self.frontend = CosyVoiceFrontEnd(configs['get_tokenizer'],
                                           configs['feat_extractor'],
                                           '{}/campplus.onnx'.format(model_dir),
@@ -147,14 +148,18 @@ class CosyVoice2(CosyVoice):
                                           '{}/spk2info.pt'.format(model_dir),
                                           instruct,
                                           configs['allowed_special'])
+        print("Loaded frontend")
         self.sample_rate = configs['sample_rate']
         if torch.cuda.is_available() is False and load_jit is True:
             load_jit = False
             logging.warning('cpu do not support jit, force set to False')
         self.model = CosyVoice2Model(configs['llm'], configs['flow'], configs['hift'])
+        print("Loaded model")
+
         self.model.load('{}/llm.pt'.format(model_dir),
                         '{}/flow.pt'.format(model_dir),
                         '{}/hift.pt'.format(model_dir))
+        print(f"Model llm from {model_dir}/llm.pt, flow from {model_dir}/flow.pt, hift from {model_dir}/hift.pt")
         if load_jit:
             self.model.load_jit('{}/flow.encoder.fp32.zip'.format(model_dir))
         if load_trt is True and load_onnx is True: