Error with guild.yml but not guild run train.py

I am using tensorflow 2.0a and guild train.py works fine for me. However, when i specify a guild.yml file and run guild run train it fails with the following error:

You are about to run clamp-detector:train
  batch_size: 2
  epochs: 2
  evaluation_batch_size: 32
  gradient_clipvalue: 1.0
  learning_rate: 0.001
  model_dir: runs/model
  oversample_factor: 15
  seed: 1000
  steps_per_epoch: 100
  validation_split_seed: 71751
  validation_steps: 50
Continue? (Y/n) Y
Traceback (most recent call last):
  File "/home/richard/anaconda3/envs/clamp/lib/python3.7/runpy.py", line 193, in _run_module_as_main
    "__main__", mod_spec)
  File "/home/richard/anaconda3/envs/clamp/lib/python3.7/runpy.py", line 85, in _run_code
    exec(code, run_globals)
  File "/home/richard/.local/lib/python3.7/site-packages/guild/op_main.py", line 272, in <module>
    main()
  File "/home/richard/.local/lib/python3.7/site-packages/guild/op_main.py", line 53, in main
    _main()
  File "/home/richard/.local/lib/python3.7/site-packages/guild/op_main.py", line 79, in _main
    _apply_plugins()
  File "/home/richard/.local/lib/python3.7/site-packages/guild/op_main.py", line 107, in _apply_plugins
    _apply_plugin(name)
  File "/home/richard/.local/lib/python3.7/site-packages/guild/op_main.py", line 111, in _apply_plugin
    plugin.patch_env()
  File "/home/richard/.local/lib/python3.7/site-packages/guild/plugins/tensorflow_util.py", line 39, in patch_env
    self._try_patch_tensorflow()
  File "/home/richard/.local/lib/python3.7/site-packages/guild/plugins/tensorflow_util.py", line 52, in _try_patch_tensorflow
    tensorflow.summary.FileWriter, "add_summary",
AttributeError: module 'tensorboard.summary._tf.summary' has no attribute 'FileWriter'

Tensorboard is version 1.13.1, tensorboardX is version 1.7

This is not a show stopper for me since I can run with the other command but I thought it might be helpful to report

[gar1t]

@samedii The fix for this will be out this week. In the meantime, you can work around the problem using the pattern used in this example:

https://github.com/guildai/examples/tree/master/tensorflow2

The Guild file (guild.yml) disables all plugins, which is required to skip the logic triggering the exception you reported. The operation also explicitly defines output-scalars, which capture the output generated by the Keras training loop.

If your operation generates any additional (or alternative) output, update output-scalars in your Guild file accordingly.

If you need any help with regular expressions - just ping us on Slack. We’re working on docs with examples for all of this but I wanted to get you an update ASAP.

Thank again for the report!

[samedii]

Working nicely!