fsdl-text-recognizer-2022
fsdl-text-recognizer-2022 copied to clipboard
Source of the FSDL 2022 labs, which are at https://github.com/full-stack-deep-learning/fsdl-text-recognizer-2022-labs
command: ```bash python -m flake8 --select=ANN --ignore=ANN101,ANN102,ANN002,ANN003,ANN202 \ --mypy-init-return --suppress-none-returning ./text-recognizer ./training ``` currently 228 lints
it is no longer in the requirements, isn't in the linting, and doesn't run on 2022 branch at https://github.com/full-stack-deep-learning/fsdl-text-recognizer-2021/commit/0e875130ac429c68e1188a43ebe68045dafb4116 Process: add to pre-commit lint all files add to dev-lint environment...
right now, the "support" files for the model testing are written by hand, and so they need to be updated alternatively, we could draw from the validation or test sets...
I checked this issue has not been duplicated. Hi @charlesfrye , I'm not sure if this repo is accepting PRs, but I spotted a bug in the `data/emnist.py` file. It...
Account for `y` labels being offset by `NUM_SPECIAL_TOKENS` when calling `np.bincount` in emnist balance subsampling. The offsetting is found here: https://github.com/the-full-stack/fsdl-text-recognizer-2022/blob/ac59bfe43ea3e1ef1e03e4fb3b1bcf715a973063/text_recognizer/data/emnist.py#L104 and here: https://github.com/the-full-stack/fsdl-text-recognizer-2022/blob/ac59bfe43ea3e1ef1e03e4fb3b1bcf715a973063/text_recognizer/data/emnist.py#L106 `np.bincount` will prepend zeros for elements...
Hi, When launching the lab 2a on Pytorch Lightning in Colab (https://colab.research.google.com/github/full-stack-deep-learning/fsdl-text-recognizer-2022-labs/blob/main/lab02/notebooks/lab02a_lightning.ipynb#scrollTo=sVx7C7H0PIZC), the first cell runs fine: ``` lab_idx = 2 if "bootstrap" not in locals() or bootstrap.run: # path...
if we can't reach `ec2.gradio.app` while trying to create a shareable URL, we go down this isn't needed for the deployed version, so we should make the `share` behavior configurable...
will need to rebuild the frontend docker container
`conda env update` reuses the existing environment, so when you make a big change to the yaml, things can fail on machines that ran `conda-update` previously. something like ```bash no-environment:...
Will there be a dedicated forum for questions and tips related to FSDL 2022? I (as well as many others) find Discourse-like community forums tremendously helpful for other data science...