Paolo Galeone

Hm.. there should be something I can't see since those lines are correct (check if the location of that lines is correct too), the permissions are correct and lighttpd starts,...

It would be cool. Is the wildcard certs nowadays 100% supported by letsencrypt? If so, this addition is more than welcome

Why not, if it's something that can be automatized why not adding this feature

I guess we should go with this approach: 1. If the image `dtype` is `tf.uin8`, then we expect values on [0-255] (follow the tensorflow convention) 2. If the image `dytpe`...

The log method (in utils) is extremely useful when doing custom logging, it should be documented and public; perhaps moved in a better location. The `validate_objects` instead is too broad,...

Yup - we can start creating some notebook, first run it on a Binder, so people can try ashpy - and when colab starts supporting Python 3.7 we'll switch

> The parameter measure_performance_freq is responsible for the frequency of performance measurements. Performance measurements during training and validation. > In the current implementation the average loss is measured as a...

Point 3: refactor everything. It happens often that our loss function is the composition of executors (e.g. is a sum executor). Term A + Term B. With the actual way...

Hi! Do you mean changing the name in the saved model from `StatefulPartitionedCall:1` to `B`?

:thinking: it should be this way. Can you show me: 1. In **python** how do you load the saved model, and feed the model to get the values A,B,C,D (and...