Databricks

rubenteixeira · ‎01-12-2023

I'm doing parameter tuning for a NeuralProphet model (you can see in the image the parameters and code for training)

When I try to parallelize the training, it gives me Permission Error.

Why can't I access the folder '/databricks/spark/work/*'? Do I need some aditional permissions to write on the cluster?

I also leave the error traceback. Thank you

Traceback (most recent call last):
  File "/databricks/spark/python/pyspark/worker.py", line 876, in main
    process()
  File "/databricks/spark/python/pyspark/worker.py", line 868, in process
    serializer.dump_stream(out_iter, outfile)
  File "/databricks/spark/python/pyspark/serializers.py", line 325, in dump_stream
    vs = list(itertools.islice(iterator, batch))
  File "/databricks/spark/python/pyspark/util.py", line 84, in wrapper
    return f(*args, **kwargs)
  File "<command-3900447791436819>", line 4, in <lambda>
  File "<command-3900447791436817>", line 36, in neural_prophet_cv
  File "/local_disk0/.ephemeral_nfs/cluster_libraries/python/lib/python3.9/site-packages/neuralprophet/forecaster.py", line 795, in fit
    metrics_df = self._train(
  File "/local_disk0/.ephemeral_nfs/cluster_libraries/python/lib/python3.9/site-packages/neuralprophet/forecaster.py", line 2657, in _train
    self.trainer.fit(
  File "/local_disk0/.ephemeral_nfs/cluster_libraries/python/lib/python3.9/site-packages/mlflow/utils/autologging_utils/safety.py", line 555, in safe_patch_function
    patch_function(call_original, *args, **kwargs)
  File "/local_disk0/.ephemeral_nfs/cluster_libraries/python/lib/python3.9/site-packages/mlflow/utils/autologging_utils/safety.py", line 254, in patch_with_managed_run
    result = patch_function(original, *args, **kwargs)
  File "/local_disk0/.ephemeral_nfs/cluster_libraries/python/lib/python3.9/site-packages/mlflow/pytorch/_pytorch_autolog.py", line 370, in patched_fit
    result = original(self, *args, **kwargs)
  File "/local_disk0/.ephemeral_nfs/cluster_libraries/python/lib/python3.9/site-packages/mlflow/utils/autologging_utils/safety.py", line 536, in call_original
    return call_original_fn_with_event_logging(_original_fn, og_args, og_kwargs)
  File "/local_disk0/.ephemeral_nfs/cluster_libraries/python/lib/python3.9/site-packages/mlflow/utils/autologging_utils/safety.py", line 471, in call_original_fn_with_event_logging
    original_fn_result = original_fn(*og_args, **og_kwargs)
  File "/local_disk0/.ephemeral_nfs/cluster_libraries/python/lib/python3.9/site-packages/mlflow/utils/autologging_utils/safety.py", line 533, in _original_fn
    original_result = original(*_og_args, **_og_kwargs)
  File "/local_disk0/.ephemeral_nfs/cluster_libraries/python/lib/python3.9/site-packages/pytorch_lightning/trainer/trainer.py", line 696, in fit
    self._call_and_handle_interrupt(
  File "/local_disk0/.ephemeral_nfs/cluster_libraries/python/lib/python3.9/site-packages/pytorch_lightning/trainer/trainer.py", line 650, in _call_and_handle_interrupt
    return trainer_fn(*args, **kwargs)
  File "/local_disk0/.ephemeral_nfs/cluster_libraries/python/lib/python3.9/site-packages/pytorch_lightning/trainer/trainer.py", line 735, in _fit_impl
    results = self._run(model, ckpt_path=self.ckpt_path)
  File "/local_disk0/.ephemeral_nfs/cluster_libraries/python/lib/python3.9/site-packages/pytorch_lightning/trainer/trainer.py", line 1154, in _run
    self._log_hyperparams()
  File "/local_disk0/.ephemeral_nfs/cluster_libraries/python/lib/python3.9/site-packages/pytorch_lightning/trainer/trainer.py", line 1222, in _log_hyperparams
    logger.log_hyperparams(hparams_initial)
  File "/local_disk0/.ephemeral_nfs/cluster_libraries/python/lib/python3.9/site-packages/pytorch_lightning/utilities/rank_zero.py", line 32, in wrapped_fn
    return fn(*args, **kwargs)
  File "/local_disk0/.ephemeral_nfs/cluster_libraries/python/lib/python3.9/site-packages/pytorch_lightning/loggers/tensorboard.py", line 211, in log_hyperparams
    self.log_metrics(metrics, 0)
  File "/local_disk0/.ephemeral_nfs/cluster_libraries/python/lib/python3.9/site-packages/pytorch_lightning/utilities/rank_zero.py", line 32, in wrapped_fn
    return fn(*args, **kwargs)
  File "/local_disk0/.ephemeral_nfs/cluster_libraries/python/lib/python3.9/site-packages/neuralprophet/logger.py", line 29, in log_metrics
    super(MetricsLogger, self).log_metrics(metrics, step)
  File "/local_disk0/.ephemeral_nfs/cluster_libraries/python/lib/python3.9/site-packages/pytorch_lightning/utilities/rank_zero.py", line 32, in wrapped_fn
    return fn(*args, **kwargs)
  File "/local_disk0/.ephemeral_nfs/cluster_libraries/python/lib/python3.9/site-packages/pytorch_lightning/loggers/tensorboard.py", line 236, in log_metrics
    raise ValueError(m) from ex
ValueError: 
 you tried to log -1 which is currently not supported. Try a dict or a scalar/tensor.

Debayan · ‎01-12-2023

Hi, Could you please check on cluster-level permissions and let us know if it helps?

Please refer: https://docs.databricks.com/security/access-control/cluster-acl.html#cluster-level-permissions

rubenteixeira · ‎01-13-2023

Hello. Thank you for your answer. I have 'Can manage' permissions.

P.S. I'm using Azure Databricks.

Kaniz · ‎01-16-2023

Hi @Rúben Teixeira, Do you still need help?

rubenteixeira · ‎01-20-2023

Hello! Yes, I do. I requested the permission of Contributor, on the Databricks service, on the Azure Portal, and now this Tip appeared. How can I write to files in a repository?