Alexander Dmitriev

Another interesting case is attempting to train a model with the same params twice without restarting `python` in between. In this case it throws some internal error the first time,...

Nope, not at all. Though interestingly enough this time when i ran the same code it worked for a 112K documents dataset. May be because this time the GPU-support failed...

I have my ebcli inside a folder in the repo. The ebcli folder is ignored within the `.ebignore` as ``` ebcli ebcli/ ebcli/* ``` With this patch the `ebcli/lib64` symling...

I second merging the PR. I *haven't reviewed the PR*, but this bug makes me want to self-host my stuff somewhere far away from AWS.

Is there any *sensible* progress on this issue?

If this is still relevant - technically yes, but no. My conflict here is between: 1. Using just one json query for all the data that i need: Temp /...