tensorflow errors and Error from chokidar (/webapp/frontend/node_modules...) errors

[myin4]

Hello @JeffreyCA ,
Hope you are doing well.

I am trying to build your spleeter-web on my local Mac machine and also on my Ubuntu server. But I always get the tensorflow error and Error from chokidar. You can see the error log below.

P/S: I did follow your instruction too. I am starting your spleeter-web with docker.

tensorflow errors log (both happened on local mac machine and Ubuntu server):
| 2021-11-23 18:03:22.755058: W tensorflow/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory
api_1 | 2021-11-23 18:03:22.755269: I tensorflow/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.

Error from chokidar log (This happended on Ubuntu server):

frontend_1 | Error from chokidar (/webapp/frontend/node_modules/@react-icons/all-files/ri): Error: ENOSPC: System limit for number of file watchers reached, watch '/webapp/frontend/node_modules/@react-icons/all-files/ri/RiFileList2Fill.esm.js'
frontend_1 | Error from chokidar (/webapp/frontend/node_modules/@react-icons/all-files/ri): Error: ENOSPC: System limit for number of file watchers reached, watch '/webapp/frontend/node_modules/@react-icons/all-files/ri/RiFileList2Fill.js'
frontend_1 | Error from chokidar (/webapp/frontend/node_modules/@react-icons/all-files/ri): Error: ENOSPC: System limit for number of file watchers reached, watch '/webapp/frontend/node_modules/@react-icons/all-files/ri/RiFileList2Line.d.ts'
frontend_1 | Error from chokidar (/webapp/frontend/node_modules/@react-icons/all-files/ri): Error: ENOSPC: System limit for

I am looking forward to your response!

Regards,
Peter

[JeffreyCA]

Hi, thanks for reporting this. This seems to be the same issue as #73.
Can you try the suggestion in that issue?

[myin4]

Hi, thanks for reporting this. This seems to be the same issue as #73.
Can you try the suggestion in that issue?

Hi @JeffreyCA ,

Thank you for your response!.

I did try your suggestion and The Error from chokidar log is gone.

I did build and run the spleeter-web on my Ubuntu server, But the spleeter model does not work. you can see the screenshot below

Regards,
Peter

[JeffreyCA]

Can you share the Docker logs? From celery-slow service

[myin4]

hi @JeffreyCA ,

Thank you for your response!

Here is the log file from my Ubuntu server:
ubuntu_server_logs.txt

Regards,
Peter

[JeffreyCA]

It looks like your Docker instances are running out of memory. How much RAM do you have on your machine? And can you try increasing the amount of memory available to Docker?

[myin4]

It looks like your Docker instances are running out of memory. How much RAM do you have on your machine? And can you try increasing the amount of memory available to Docker?

Hi @JeffreyCA ,

Thank you for your response!

I also tried to run spleeter-web on my MacBook Pro (RAM: 16 GB 2133 MHz LPDDR3, Processor: 2.9 GHz Quad-Core Intel Core i7) too, but spleeter-web did not work.

You can see the Docker Resources preferences:

Here is the screenshot of spleeter-web on my macbook pro:

And here the console log:
spleeter-web-celery-slow-1 | [2021-11-25 08:12:18,431: INFO/MainProcess] Task api.tasks.create_static_mix[27a79464-aed4-4495-a7d0-cc117d97e93f] received
spleeter-web-api-1 | [25/Nov/2021 08:12:18] "POST /api/mix/static/ HTTP/1.1" 201 429
spleeter-web-api-1 | [25/Nov/2021 08:12:18] "GET /api/source-track/ HTTP/1.1" 200 697
spleeter-web-celery-slow-1 | INFO:tensorflow:Apply unet for vocals_spectrogram
spleeter-web-celery-slow-1 | [2021-11-25 08:12:19,231: INFO/ForkPoolWorker-1] Apply unet for vocals_spectrogram
spleeter-web-celery-slow-1 | INFO:tensorflow:Apply unet for drums_spectrogram
spleeter-web-celery-slow-1 | [2021-11-25 08:12:20,066: INFO/ForkPoolWorker-1] Apply unet for drums_spectrogram
spleeter-web-celery-slow-1 | INFO:tensorflow:Apply unet for bass_spectrogram
spleeter-web-celery-slow-1 | [2021-11-25 08:12:21,032: INFO/ForkPoolWorker-1] Apply unet for bass_spectrogram
spleeter-web-api-1 | [25/Nov/2021 08:12:21] "GET /api/source-track/ HTTP/1.1" 200 697
spleeter-web-celery-slow-1 | INFO:tensorflow:Apply unet for other_spectrogram
spleeter-web-celery-slow-1 | [2021-11-25 08:12:22,032: INFO/ForkPoolWorker-1] Apply unet for other_spectrogram
spleeter-web-celery-slow-1 | [2021-11-25 08:12:24,701: WARNING/ForkPoolWorker-1] Can't load save_path when it is None.
spleeter-web-celery-slow-1 | [2021-11-25 08:12:24,761: INFO/ForkPoolWorker-1] Task api.tasks.create_static_mix[27a79464-aed4-4495-a7d0-cc117d97e93f] succeeded in 6.3273712420000265s: None

Regards,
Peter

[JeffreyCA]

Thanks, according to this: deezer/spleeter#306 (comment) it may happen when you stop the process mid-way through downloading Spleeter models.

Can you delete the pretrained-models Docker volume and rerun it again?