BTW did you happen to try this non-async lib? Seems like it would be nice to avoid async if possible to simplify the explanation/exploration https://
websocket-client.readthedocs.io/en/latest/exam
ples.html
…
@jeremyphoward
-
Non-async WebSocket library for simplified implementation
By
–
-
IPython Async REPL Implementation Issues and Solutions
By
–
What was the issue with async? Did this stuff not work? : https://
blog.jupyter.org/ipython-7-0-as
ync-repl-a35ce050f7f7
… -
Batch vs Epoch Level Logging Best Practices Training
By
–
AFAIK you pretty much always want batch-level logging for training and epoch-level for validation (and end-of-epoch training)?
-
Library Feature Request for Standardized Implementation
By
–
I hope you do add it at some point – it's a pretty slim library, and it provides a feature pretty much everyone needs, and it's something that needs to be done in a consistent standard way across projects.
-
Diffusion Models Learning Lesson from John O Whitaker
By
–
Also, be sure to check out this great lesson from @johnowhitaker which introduces some cool diffusion ideas: https://
youtube.com/watch?v=844LY0
vYQhc
… -
Mid-U Guidance: Fast Classifier Guidance for Latent Diffusion
By
–
Check out this powerful new diffusion inference technique from @johnowhitaker – 'Mid-U Guidance': Fast Classifier Guidance for Latent Diffusion Models. It lets you very flexible push your generated images towards almost any concept you can imagine.
-
Improving devcontainer.json setup for Codespaces development
By
–
BTW we do have a .devcontainer.json which we set up in the early days of Codespaces. I'm sure it could be much better now!
-
Exploring nbterm and jpterm development tools for AI workflows
By
–
Wow exciting news! You might find nbterm and jpterm from @davidbrochart useful to study.
-
94.5% Accuracy Fashion MNIST Achievement in Five Epochs
By
–
94.5% on Fashion MNIST in 5 epochs is pretty amazing!
-
Bypassing AI Detection: Natural Text and Watermarking Constraints
By
–
I'm not sure about that. E.g. tricking the ZeroGPT algo means actually writing more natural text (so it'll be better, not worse), and tricking the watermarking means not being limited by the constraints of the watermarking algo.