Id recommend switching to the 2018 version of the course, which uses PY3 and is based on Pytorch. It’s really different from the previous version.
Thanks for the reply, I will look into it. I have had issues getting a GPU up and running at Google Cloud for ULMFiT, but am now trying to use an image with CUDA pre-installed, and update my condo environment according to instructions. Hopefully this will work out!
I followed the Ultimate guide to setting up a Google Cloud machine for fast.ai version 2 blog post to setup my GCP environment, but I’m facing some issues afterwards, like graphviz not being available and failing when trying to install it.
I raised an issue on howkhang/fastai-v2-setup, but seeing no update on the repo in 8 months and no answer to a PR made in January, I probably won’t receive an answer for the author.
Can anyone can have a look at this ? Thanks in advance