Using TensorBoard with dstack to run ML experiments at scale

TensorBoard is probably the simplest tool to track experiment metrics. You don’t need to host a tracking server. Just save logs locally and use the TensorBoard CLI to visualize it.

I’m excited to share our new tutorial on using TensorBoard with dstack. When combined with dstack, TensorBoard becomes even better because dstack helps to conveniently store logs in your storage.

Check out the tutorial and let us know what you think! TensorBoard - dstack