According to the website (https://vectordash.com/hosting/) they use a highly isolated Ubuntu image, so the person hosting the service shouldn't have access to the VM with your model or data on it. It would be nice if there was some third party audit of the software though, the models, the code, and even the training data can be pretty sensitive for researchers.
Visual Studio Code with the LaTeX Workshop addon is definitely my favorite LaTeX editor. The integration with the Chktex linter, latexmk, git, and all that jazz just makes it so much easier to focus on writing.
For research management I had been using Mendeley for a while and got a bit frustrated with the way it handled bibtex. Like it got really annoying when I had papers which fell into multiple categories and/or were used in multiple papers. My new setup is to use JabRef to manage individual bibtex files for specific projects and to use Mendeley just for document management and notes.
Hot damn this has got me all giddy. How will this work on single node multi-GPU systems? For example, with PyTorch you have to either use threading, multiprocessing, or even MPI. Can you think of a not-too-scary way to use eager execution with multiple GPUs?
For installing it, yeah pip is great too, but for building conda includes third party tools and libraries and stuff. e.g. in order to use the MPI backend for PyTorch's distributed processing you need to build it yourself and conda just makes it a bit easier. That and I had a real bad experience with trying to build Tensorflow (and Bazel) to run on an HPC cluster.
Not yet! I'm not using convnets or backprop or anything so I don't think it would be beneficial that way, but you could get something similar to what I'm doing by looking at Fritzke's Growing Neural Gas[1]
Hell, being able to effortlessly switch between PyTorch and Numpy/SciPy/sklearn/skimage has been so helpful for the project I'm working on. That and I have tensors in later layers whose shapes depend on the training of the previous layers.