François Chollet
François Chollet
This is great, thanks for the proposal. LGTM. Note that this is currently doable in TF in a performant way via a XLA op: ```python from tensorflow.compiler.tf2xla.python import xla output...
Since Keras is in stable state after the release of 1.0, the changes to the documentation should be rare and limited. But there will be changes. I can post an...
Another solution: - create a new github repo - in the keras source, `cd` to `docs/` and run `python autogen.py` - `cp` the resulting `sources` dir to the new repo...
All of the steps above are bash commands. Automation is one script away. On 14 June 2016 at 19:12, Tsukasa ŌMOTO [email protected] wrote: > I hope we could automate these...
Thanks for the PR! - Can you provide the git diff? - Did you run it on all 3 backends?
Sounds good -- thank you!
1. Why not just call `flatten` on your structure before calling `build()`? 2. If you pass a nested structure to `build()`, do you expect to also pass the same nested...
The format test is failing (you have a line that's too long). You can run `sh shell/format.sh` to fix it.
The issue here is that you are attempting to `model = TorchModuleWrapper(dino)` as if it were a Keras `Model`. In reality, it is a `Layer`, so it doesn't have the...
@nkovela1 this is fixed, right?