Žiga Avsec

Hey, you can access the data, but you have to set up your Google Cloud account to pay for the download. I actually don't know how to set this up...

Hey! Enformer pre-trained model wasn't trained on other species than human and mouse, nor was it evaluated on them. The Enformer model architecture specified in enformer.py could be used for...

By 256 do you mean 256 bins at 128 bp resolutoin or 256 bp? I would recommend lowering the number of channels by 4x to say 384 instead of lowering...

Hi, thanks for reaching out. Currently, there is no pipeline/script to run the Enformer. I suggest implementing the pipeline yourself by extracting the relevant code from the [enformer-usage colab](https://colab.research.google.com/github/deepmind/deepmind_research/blob/master/enformer/enformer-usage.ipynb). The...

oh. thanks. need to see what happened to the blog

ok. my domain has expired (forgot to update it). Will try to get it back and running

You can read the blog-post also here for now: https://github.com/Avsecz/avsecz.github.io/blob/master/_posts/2016-06-18-aws-tensorflow-setup.md

Check out also [Paperspace](https://www.paperspace.com/), [google ml engine](https://cloud.google.com/ml-engine/) and [Floydhub](https://www.floydhub.com/) if you want to train model in the cloud - https://medium.com/@rupak.thakur/aws-vs-paperspace-vs-floydhub-choosing-your-cloud-gpu-partner-350150606b39. I find their setup to be easier, especially floydhub.

Hey, at the moment not. You would need to change the lines: [39-47](https://github.com/Avsecz/aws-tensorflow-setup/blob/master/setup-aws-tensorflow.bash#L39-L47). Feel free to make a pull request. Best, Žiga

Awesome! I didn't know about the shuffle='batch' trick - very neat! > successfully read directly from hdf5 files with multiple workers concurrently. In the current PR, the data are not...