Posts
-
Helen district heat generation dataset 2015-2020
I was notified yesterday that Helsinki district heating company Helen published their district heat generation data from 2015-2020. This is very nice indeed, since generation data has been notoriously hard to get as a researcher. I’m leaving here some notes on how to read the data in a Jupyter notebook.
-
fast.ai model development on DataCrunch.io
For developing and training deep learning models, a GPU-enabled computing resource is a must. I did not even bother trying on my laptop, but went straight ahead to using cloud resources as recommended by fast.ai tutorial. I started first using Paperspace Gradient as per fast.ai instructions, but it soon occurred to me that the free notebook servers were most of the time loaded full. Paying a bit extra would not have been a problem, but in order to access the paid GPU servers, a monthly subscription would have been a must. I could not justify that cost yet, so I arrived at another suggested solution, DataCrunch.io
-
Hello, World!
Hello world, this is my first Jekyll blog post.
subscribe via RSS