Josh Rosen

Properly handle opaque whiteout entries in layer flattener

6

This PR fixes a bug in the flattener's handling of layers with opaque whiteout entries (`.wh..wh..opq` files). This bug could lead to deleted files re-appearing in flattened output. ## Bug...

java_tls_test suite fails with test error then hangs indefinitely

1

The `src/py4j/tests/java_tls_test.py` test errors out and then hangs indefinitely, both in local tests and in GitHub Actions. This problem appears to be specific to that suite, since all other test...

help() causes TypeError

3

When I call `help(amazon)` I get the following error ``` Python 2.7.2 |EPD 7.1-2 (32-bit)| (default, Jul 27 2011, 13:29:32) [GCC 4.0.1 (Apple Inc. build 5493)] on darwin Type "help",...

Add tests for Python APIs

1

This library should test its Python APIs. Note that Codecov has Python integration, so we can also measure coverage there.

Option for non-smooth scrolling

1

I appreciate the option to use the j/k keys to scroll through webpages, but I don't like the smooth scrolling animation. Would it be possible to add an option to...

Arguments are incorrectly double-quoted when nesting remote commands

13

Consider the following example: ``` python In [2]: from plumbum import SshMachine In [3]: remote = SshMachine('localhost', ssh_opts=['-tt']) In [4]: ls = remote['ls'] In [5]: sudo = remote['sudo'] In [6]:...

I ran across a confusing `JavaException` when trying to call a Java method stored in a class variable: ``` python from jnius import autoclass System = autoclass('java.lang.System') # This works:...

Display additional information on the layer graph

1

Could we encode additional information on the layer graph? Maybe we could use edge labels to convey information about the sizes of the outputs at each layer.

Images should be normalized according to the same scale across timesteps

1

Images should be normalized according to the same scale across timesteps.

Show how training error decreases over time

4

We should add a view that shows how the training error decreases over time. Some ideas: - Add a sparkline showing the training loss versus time. This could be drawn...