Adam Murphy
I use VS Code and there is an [extension](https://marketplace.visualstudio.com/items?itemName=njpwerner.autodocstring) which automatically generates docstrings in readthedocs format. [train_transformer_style](https://flow-forecast.readthedocs.io/en/latest/pytorch_training.html) looks really good, so I assume you'd want them all in that style?
Agreed! It was my first time working with FF yesterday and it was hard to follow along. My first thought is just to split the massive functions up into smaller...
Hey, @joelrorseth here is the link to my file [flow_forecast_utils.py](https://github.com/theadammurphy/pytorch_learn_by_converting/blob/e199caeadd451451b368c33a7a73a69027a00801/electrochem/flow-forecast/flow_forecast_utils.py) that I've tried to use to manage experiments. The most important functions are: 1. `make_final_config` 2. `get_sweep_config` 3. `train` 4....
Wonderful! I'm very happy to hear that @joelrorseth! Glad to be of assistance :)
Update: just installed `premailer` with `conda install -c conda-forge premailer` for python 3.10 and it works fine. Still an issue that premailer is not listed as a dependency though
Do conda installs work differently somehow?
True :) I just wanted to highlight that the experience with conda is not as seamless as it could be On Tue, 1 Nov 2022 at 10:59, Pascal van Kooten...
One reason must be that a SequentialChain encapsulates the logic into a single class you can move around and use however you like. If you are using multiple individual chains,...
@prosa-trebu Some ideas - Have the first chain be separate so that if it's False it breaks and if it's True it moves into the actual larger chain - Use...
Just want to say to you both @Delgan and @DeflateAwning that this was a lovely, civil disagreement to read. I understand your frustration @DeflateAwning but also appreciate @Delgan's commitment to...