sensorium
sensorium copied to clipboard
Code base for the SENSORIUM competition.
@MaxFBurg After all the issues were fixed with independent_bn_bias, we have a new problem: The code works as intended, but it is not backwards compatible. If Sensorium and neuralpredictors with...
Hi @KonstantinWilleke @mohammadbashiri @kklurz , I think (some) of you specified the package dependencies. Do you know why `numpy>=1.22.0` is required and if a lower version would work as well...
https://github.com/sinzlab/sensorium/blob/e5017df2ff89c60a4d0a7687c4bde67774de346b/sensorium/training/trainers.py#L85
Was talking to @KonstantinWilleke just now and we had some ideas for improving the participation experience: 1. Provide the demo notebooks in Google colab 2. Provide an extra notebook to...
I would suggest to add **kwargs [here](https://github.com/sinzlab/sensorium/blob/main/sensorium/models/models.py#L54) and [here](https://github.com/sinzlab/sensorium/blob/main/sensorium/models/models.py#L124) as otherwise it does not allow to change some of the parameters, which are not stated in the function while they...
1. The standard trainer will not stop and will not give any warning if the loss when to 'nan', it will just keep training. 2. It would be nice to...
When running `submission.generate_submission_file(trained_model=model, dataloaders=dataloaders, data_key=dataset_name, path="./submission_files/", device="cpu")` I still get a `RuntimeError: No CUDA GPUs are available` even when using device='cpu'
Hi, I was using your submission example notebook to generate my submission files and I came across a couple of errors that I tried resolving manually and just wanted to...