[BUG] Failed to load model with docker up

[hexsix]

Issue Type

Bug Report

vc client version number

tag v.1.5.3.5 & dannadori/vcclient:20230602_014557

OS

Docker

GPU

4080

Clear setting

yes

Sample model

no

Input chunk num

no

Voice Changer type

RVC

Model type

pyTorch

Situation

I've tried build image myself on tag v.1.5.3.5 with command npm run build:docker:vcclient. And start with USE_LOCAL=on bash start_docker.sh.

Also I've tried image from dockerhub dannadori/vcclient:20230602_014557. And start with bash start_docker.sh.

They are both the same issue as shown below.

I guess it would be some PATH issue. According to commit 99f93f9

Screenshot:

Log:

~/projects/voice-changer$ bash start_docker.sh
VC Client start...(with gpu)
exec with [UID : 1000, GID: 1000]
/voice-changer/server/MMVCServerSIO.py:4: DeprecationWarning: The distutils package is deprecated and slated for removal in Python 3.12. Use setuptools or check PEP 632 for potential alternatives
  from distutils.util import strtobool
    Booting PHASE :__main__
    Voice Changerを起動しています。
+[Errno 13] Permission denied: 'samples_0003_o.json'
[Voice Changer] loading sample info error: RVCModelSample.__init__() got an unexpected keyword argument 'icon'
    External_Port:18888 Internal_Port:18888
    External_IP:172.25.219.245
    protocol: HTTPS(self-signed), key:keys/20230616_165542.key, cert:keys/20230616_165542.cert
    -- ---- --
    ブラウザで次のURLを開いてください.
    https://<IP>:<PORT>/
    多くの場合は次のいずれかのURLにアクセスすると起動します。
    https://localhost:18888/
    https://172.25.219.245:18888/
    Booting PHASE :__mp_main__
    サーバプロセスを起動しています。
    Booting PHASE :MMVCServerSIO
VoiceChanger Initialized (GPU_NUM:1, mps_enabled:False)
[2023-06-16 16:55:59] connet sid : xleitP9SG610uuM7AAAC
[2023-06-16 16:55:59] connet sid : 4jenEcWyGX5eSLXrAAAD
2023-06-16 16:55:59 | INFO | faiss.loader | Loading faiss with AVX2 support.
2023-06-16 16:55:59 | INFO | faiss.loader | Successfully loaded faiss with AVX2 support.
[Voice Changer] RVC initialization:  VoiceChangerParams(model_dir='model_dir', content_vec_500='pretrain/checkpoint_best_legacy_500.pt', content_vec_500_onnx=None, content_vec_500_onnx_on=False, hubert_base='pretrain/hubert_base.pt', hubert_base_jp='pretrain/rinna_hubert_base_jp.pt', hubert_soft='pretrain/hubert/hubert-soft-0d54a1f4.pt', nsf_hifigan='pretrain/nsf_hifigan/model')
[Voice Changer] loading sample info error: RVCModelSample.__init__() got an unexpected keyword argument 'icon'
[2023-06-16 16:56:06] connet sid : u8t472tW3Gjg1zkiAAAG
[2023-06-16 16:56:06] connet sid : YfcqsquyKk2dYIaxAAAH
Exception ignored in: <function RVC.__del__ at 0x7fca554bb400>
Traceback (most recent call last):
  File "/voice-changer/server/voice_changer/RVC/RVC.py", line 375, in __del__
    del self.pipeline
AttributeError: pipeline
[Voice Changer] RVC initialization:  VoiceChangerParams(model_dir='model_dir', content_vec_500='pretrain/checkpoint_best_legacy_500.pt', content_vec_500_onnx=None, content_vec_500_onnx_on=False, hubert_base='pretrain/hubert_base.pt', hubert_base_jp='pretrain/rinna_hubert_base_jp.pt', hubert_soft='pretrain/hubert/hubert-soft-0d54a1f4.pt', nsf_hifigan='pretrain/nsf_hifigan/model')
[Voice Changer] loading sample info error: RVCModelSample.__init__() got an unexpected keyword argument 'icon'

P.S. I've also add -v ./docker_vcclient/pretrain:/voice-changer/server/pretrain to avoid redownload models. I thought it would be better remove the cp command in https://github.com/w-okada/voice-changer/blob/master/docker_vcclient/exec.sh#L5 And just map the volumn using docker command

-v `pwd`/docker_vcclient/weights:/voice-changer/server/pretrain

[w-okada]

I updated docker. please try.

[hexsix]

I updated docker. please try.

@w-okada It works for me. Thank you.