cellmap_flow.finetune.markers
=============================

.. py:module:: cellmap_flow.finetune.markers

.. autoapi-nested-parse::

   The lines the trainer prints for the job manager to find in its log.

   The training job's loop (session_loop) and lora_trainer print these to
   stdout, which tee copies into training_log.txt; the job manager reads them
   back from there to follow the job, and the finetune tab reads the log it
   streams. The two sides can be different versions -- a training job
   outlives a dashboard upgrade -- so these only ever change compatibly.

   The patterns are the ones the job manager has always used, unchanged.



Attributes
----------

.. autoapisummary::

   cellmap_flow.finetune.markers.FINETUNED_MODEL_YAML
   cellmap_flow.finetune.markers.TRAINING_ITERATION_COMPLETE
   cellmap_flow.finetune.markers.RESTART_FAILED
   cellmap_flow.finetune.markers.INFERENCE_SERVER_FAILED
   cellmap_flow.finetune.markers.RESTART_STATUS
   cellmap_flow.finetune.markers.TRAINING_DIVERGED
   cellmap_flow.finetune.markers.RESTARTING_TRAINING
   cellmap_flow.finetune.markers.WAITING_FOR_RESTART
   cellmap_flow.finetune.markers.EPOCH_START_RE
   cellmap_flow.finetune.markers.EPOCH_SUMMARY_RE
   cellmap_flow.finetune.markers.ITERATION_COMPLETE_RE
   cellmap_flow.finetune.markers.MODEL_YAML_RE
   cellmap_flow.finetune.markers.STATUS_MARKER_RE
   cellmap_flow.finetune.markers.SERVER_URL_RE


Functions
---------

.. autoapisummary::

   cellmap_flow.finetune.markers.emit


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

.. py:data:: FINETUNED_MODEL_YAML
   :value: 'FINETUNED_MODEL_YAML:'


.. py:data:: TRAINING_ITERATION_COMPLETE
   :value: 'TRAINING_ITERATION_COMPLETE:'


.. py:data:: RESTART_FAILED
   :value: 'RESTART_FAILED:'


.. py:data:: INFERENCE_SERVER_FAILED
   :value: 'INFERENCE_SERVER_FAILED:'


.. py:data:: RESTART_STATUS
   :value: 'RESTART_STATUS:'


.. py:data:: TRAINING_DIVERGED
   :value: 'TRAINING_DIVERGED'


.. py:data:: RESTARTING_TRAINING
   :value: 'RESTARTING_TRAINING'


.. py:data:: WAITING_FOR_RESTART
   :value: 'WAITING_FOR_RESTART'


.. py:data:: EPOCH_START_RE

.. py:data:: EPOCH_SUMMARY_RE

.. py:data:: ITERATION_COMPLETE_RE

.. py:data:: MODEL_YAML_RE

.. py:data:: STATUS_MARKER_RE

.. py:data:: SERVER_URL_RE

.. py:function:: emit(marker: str, value=None) -> None

   Print ``marker`` (and its value) as a line of its own, at once.

   Flushed, because the job manager is reading the log while the job runs.