pymllm.orchestrator.detokenizer_process
=======================================

.. py:module:: pymllm.orchestrator.detokenizer_process

.. autoapi-nested-parse::

   DetokenizerProcess -- subprocess that converts token IDs back to text.

   Receives ``BatchTokenIDOut``-style dicts from the SchedulerProcess,
   detokenizes them, and forwards the decoded strings to the
   RequestResponseProcess.



Attributes
----------

.. autoapisummary::

   pymllm.orchestrator.detokenizer_process.logger


Classes
-------

.. autoapisummary::

   pymllm.orchestrator.detokenizer_process.DetokenizerProcess


Functions
---------

.. autoapisummary::

   pymllm.orchestrator.detokenizer_process.run_detokenizer_process


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

.. py:data:: logger

.. py:class:: DetokenizerProcess(recv_from_scheduler_addr, send_to_rr_addr, tokenizer_cfg = None)

   Runs inside a subprocess.  Detokenizes finished outputs.


   .. py:method:: init_sockets()


   .. py:method:: init_tokenizer()

      Load the tokenizer from the configured path.



   .. py:method:: event_loop()

      Infinite loop: recv token IDs -> detokenize -> send text to RR.



   .. py:method:: shutdown()


.. py:function:: run_detokenizer_process(recv_from_scheduler_addr, send_to_rr_addr, pipe_writer, tokenizer_cfg = None)

   Entry point for ``torch.multiprocessing.Process(target=...)``.