Jesper Nielsen
Jesper Nielsen
**Describe the bug** There appears to be several bugs when trying to use `einops` with TensorFlow tensors that have a dynamic shape. **Reproduction steps** Steps to reproduce the behavior: ```python...
I'm trying to fit a `NegativeBinomial`, but its `cdf` seems to be missing some gradients: ```python count = tf.compat.v1.get_variable("count", shape=()) logit = tf.compat.v1.get_variable("logit", shape=()) value = tf.compat.v1.get_variable("value", shape=()) cdf =...
DRAFT, in progress work on rewriting GPflow landing page / getting started documentation. A rendered version of this is available at https://gpflow.github.io/GPflow/rework/index.html Please use this to give feedback on in-progress...
DeepMind has this library: https://github.com/deepmind/einshape Maybe this could replace some of our leading_reshape etc. I know this is similar to einops, but einops don't seem to work for TensorFlow, see:...
Use TensorFlow stateless_* random functions for more deterministic randomness. This will allow our users to get truly deterministic randomness from GPflow. This should be optional, default should be similar to...
Need to dynamically update our documentation with the latest CITATIONS.cff Need to figure out exactly who will go on the list? Can we even find everybody's real names?
**PR type:** bugfix / enhancement / new feature / doc improvement **Related issue(s)/PRs:** ## Summary **Proposed changes** * ... * ... * ... **What alternatives have you considered?** ### Minimal...
Add a benchmark suite to monitor regressions. 1. Have some "minimal" set of benchmarks for continuously monitoring regression. 2. Should be extensible to make it easy to test/benchmark new features,...
- [x] Disable the obsolete CircleCI build for the `docs` repo. - [x] Replace readthedocs pages with a link to the new pages. - [x] Remove references to GPflow 1.x....