Rylan Schaeffer
Rylan Schaeffer
I have four questions towards creating custom distributional RL algorithms using Acme. 1. The only distributional agent I can find (in TF and jax) is D4PG. Is this correct? 2....
Hi! Is there a way to not use a `TemporaryDirectory()` when downloading DALL-E Mini, and instead specify a desired non-temporary directory to write the parameters to?
Hi! Could you recommend which model(s) perform best at small/tiny object segmentation?
If I can make one small quick suggestion, perhaps the two following plots in your README should have the same y axis limits. 
In Seaborn, if someone wants to plot a _smooth_ [survival function](https://en.wikipedia.org/wiki/Survival_function) (defined as 1-the CDF), there are two suboptimal solutions: - seaborn.kdeplot can plot a smooth CDF plot - seaborn.ecdfplot...
The documentation suggests that the Shampoo algorithm comes from the 2018 paper: https://arxiv.org/abs/1802.09568 Is this correct? If so, are there any plan to upgrade to the authors' newer 2021 paper:...
We're trying to fine-tune a linearized Vision Transformer by adapting code from https://github.com/google-research/vision_transformer/blob/main/vit_jax.ipynb. We're running into a really puzzling problem: when we load a model, we can train it, and...
I'm having fun playing with the Neural Tangents Cookbook.ipynb and I'd like to try extending it to multivariate regression. However, when I changed the output dimension of last layer in...
The Colab notebook showing how to train a custom agent (https://colab.research.google.com/github/google/dopamine/blob/master/dopamine/colab/agents.ipynb) doesn't actually do any learning. Could we have an example of modifying the DQN architecture or its loss function?
Two common performance metrics used in the RL literature are mean return and median return because the median is less influenced by outliers. However, `compute_metrics()` does not compute the median....