Charles Weill

@marsggbo: For multiple GPUs on the same machine depends on https://github.com/tensorflow/adanet/issues/87 being fixed first. However, you can distribute training across multiple machines each with one GPU using TF_CONFIG: https://cloud.google.com/ml-engine/docs/tensorflow/distributed-training-details You...

@maxmarketit: We just released adanet=0.9.0 which includes better TF 2 support. Please try installing it, and let us know if it resolves your issue.

@le-dawg: There are two ways to visualize the models in TensorBoard: 1. Via the `Text` tab which will display the architecture per step. 1. Via the `Graph` tab where you...

@jankrynauw Thank you for the feature request. This looks like it could be implemented in a way similar to `adanet.Estimator(metric_fn=...)` which we added in https://github.com/tensorflow/adanet/commit/53f5f5b54f91a49da766f72b373a5e23cfc2db16. Unfortunately, we don't have much...

@picarus: This is a known issue when using the `adanet.Evaluator` in distributed training. One way you can make the evaluation much faster is to pass the [`steps`](https://adanet.readthedocs.io/en/latest/adanet.html#adanet.Evaluator) argument to its...

@picarus: Unfortunately nothing is very straightforward in TF. :) The challenges I see are: * Making sure this works for any number of workers and candidate subnetworks. * How to...

@chandramoulirajagopalan: The best way to get started will be to first [extend](url) [estimator_distributed_test_runner.py](https://github.com/tensorflow/adanet/blob/master/adanet/core/estimator_distributed_test_runner.py) to test your implementation. You can pass then pass the `tf.distribute.Strategy` you want to test to the...

@chandramoulirajagopalan: Just a heads up: `tf.distribute.MirroredStrategy` I believe is designed for multi-GPU, so may be difficult to test. But if you get it to run inside `estimator_distributed_test_runner.py`, then good work....

Good work getting that inside the runner. I'm surprised that the error is coming from deep down in TensorFlow Estimator. If you create a PR, I can have a look...

looks like the problem is your numeric_column has dtype `tf.float64`. Does it work if you change it to `tf.float32`?