:py:mod:`IMLCV.configs.hpc_ugent`
=================================

.. py:module:: IMLCV.configs.hpc_ugent


Module Contents
---------------

Classes
~~~~~~~

.. autoapisummary::

   IMLCV.configs.hpc_ugent.SlurmProviderVSC



Functions
~~~~~~~~~

.. autoapisummary::

   IMLCV.configs.hpc_ugent.get_slurm_provider
   IMLCV.configs.hpc_ugent.config



Attributes
~~~~~~~~~~

.. autoapisummary::

   IMLCV.configs.hpc_ugent.ROOT_DIR
   IMLCV.configs.hpc_ugent.logger
   IMLCV.configs.hpc_ugent.translate_table


.. py:data:: ROOT_DIR

   

.. py:data:: logger

   

.. py:data:: translate_table

   

.. py:class:: SlurmProviderVSC(cluster=None, **kwargs)

   Bases: :py:obj:`parsl.providers.slurm.slurm.SlurmProvider`

   Specifies cluster and partition for sbatch, scancel, and squeue

   .. py:method:: submit(command, tasks_per_node, job_name='parsl.slurm')

      Submit the command as a slurm job.

      This function differs in its parent in the self.execute_wait()
      call, in which the slurm partition is explicitly passed as a command
      line argument as this is necessary for some SLURM-configered systems
      (notably, Belgium's HPC infrastructure).
      In addition, the way in which the job_id is extracted from the returned
      log after submission is slightly modified, again to account for
      the specific cluster configuration of HPCs in Belgium.

      :param command: Command to be made on the remote side.
      :type command: str
      :param tasks_per_node: Command invocations to be launched per node
      :type tasks_per_node: int
      :param job_name: Name for the job
      :type job_name: str

      :returns: If at capacity, returns None; otherwise, a string identifier for the job
      :rtype: None or str


   .. py:method:: _status()

      Returns the status list for a list of job_ids
      :param self:

      :returns: Status list of all jobs
      :rtype: [status...]


   .. py:method:: cancel(job_ids)

      Cancels the jobs specified by a list of job ids
      Args:
      job_ids : [<job_id> ...]
      Returns :
      [True/False...] : If the cancel operation fails the entire list will be False.



.. py:function:: get_slurm_provider(env, label, path_internal: Path | str, cpu_cluster, gpu_cluster=None, account=None, channel=LocalChannel(), gpu=False, cores=None, open_mp_threads_per_core: int | None = None, parsl_cores=False, mem=None, memory_per_core=None, walltime='48:00:00', init_blocks=1, min_blocks=1, max_blocks=1, parallelism=1, use_work_queue: bool = True, wq_timeout: int = 120, gpu_part='gpu_rome_a100', cpu_part='cpu_rome', py_env=None)


.. py:function:: config(env=None, singlepoint_nodes=16, walltime='48:00:00', bootstrap=False, memory_per_core=None, min_memery_per_node=None, path_internal: Path | None = None, cpu_cluster=None, gpu_cluster=None, py_env=None)