Hjalmar K. Turesson

I keep removing the model because it is big (5 GB) and my google drive is running out of space. I'll happily upload and keep it available if it could...

[Bert trained on Spider](http://www.mediafire.com/file/cans69jkpietag4/spider_bert.tar.xz/file). I think it gets around 60% accuracy. It'll stay up until I've trained a better model.

Dask uses the multiprocessing package (i.e. no extra dependencies). We can get the number of cups from that: `multiprocessing.cpu_count()`. In `audio.py`: adding to line 4: ```python import multiprocessing ``` and...

I tried by setting `partition_size` to 3% of the available memory too. Increasing the `partition_size` beyond 3% of available memory results in out of memory errors. ```python import psutil partition_size...

**System**: Ubuntu 20.04.5 LTS with 62.7 GM memory **Files**: 107 wav files sampled at 96kHz with sizes between 13.4 and 212.6 MB (most around 130 MB) **Traceback**: ```console $ vak...

Hi @NickleDave. I just sent you a link to a part of the data set.

Same problem on Ubuntu 22.04.

Hi @NickleDave Try my fix on many small files. It works fine on my large files. It should be good if it works on small too. In `audio.py`: ```python import...

Worked perfectly! Thanks

I'm using Studio on Ubuntu (22.04) too. I managed to connect to OpenBCI Daisy by running Studio as root (`$ sudo NMStudio`). Adding my user name to the groups `dialout`...