-
[ ] Installed Ubuntu 18
- [ ] Installed openssh
- [ ] Allowing ssh access
- [ ] Installed vncserver, but could not get it to work. So I deactivated it.
- [ ] Created 16gb of SWAP space
- [ ] Formatted and mounted 1TB hdd with disk utility fstab, gparted and mount
-
[ ] Hardware: Installed the RTX 2070.
- [ ] Needed to plug monitor in first to the onboard graphics card,
- [ ] Then turn off machine, and install RTX card
- [ ] Then turn machine back on (but still plugged into onboard graphics) and then install RTX 2070 NVidia drivers (RTX 2070 hardware needs to be installed for drivers to recognise them, even though monitor is plugged into onboard graphics)
- [ ] Then turn machine back off, plug monitor into display port of the RTX 2070
-
[ ] Installed ML frameworks (Followed instructions here https://github.com/fastai/fastai and here https://forums.fast.ai/t/fastai-v1-install-issues-thread/24111).
- [ ] Quick tip: Might need to run conda install -c pytorch -c fastai fastai and conda install pytorch torchvision -c pytorch multiple times
- [ ] conda environment for
- [ ] Pytorch
- [ ] and Fastai
- [ ]
-
[ ] Ngrok
- [ ] Installed ngrok,
- [ ] have setup a remote reserved address,
- [ ] runs as a background process,
- [ ] and starts on reboot (run
sudo crontab e
to see the crontab file for root)
-
[ ] Jupyter
- [ ] Installed jupyter notebook,
- [ ] run in background
-
[ ] Installing NVidia CUDA 10 drivers for the RTX 2070 (follwoing instructions here: https://www.pugetsystems.com/labs/hpc/How-To-Install-CUDA-10-together-with-9-2-on-Ubuntu-18-04-with-support-for-NVIDIA-20XX-Turing-GPUs-1236/#you-need-to-install-cuda-10-if)