use multiprocess

src/depiction/parallel_ops/parallel_config.py

@@ -9,7 +9,7 @@ class ParallelConfig:
     """Encapsulates the configuration for parallel processing, and some common functionality to use this information.
     :param n_jobs: the number of parallel jobs to run
     :param task_size: the size of each task, if None, the number of tasks will be divided by the number of jobs
-    :param verbose: the verbosity level of the parallel processing (passed to joblib currently)
+    :param verbose: the verbosity level of the parallel processing (TODO currently unused)
     """
 
     n_jobs: int

src/depiction/parallel_ops/parallel_map.py

@@ -2,11 +2,10 @@
 
 import functools
 import operator
-
-import joblib
-
 from typing import TypeVar, TYPE_CHECKING, Callable, Any
 
+import multiprocess.pool
+
 if TYPE_CHECKING:
     from depiction.parallel_ops import ParallelConfig
 
@@ -41,9 +40,8 @@ def __call__(
         reduce_fn: Callable[[list[T]], U] | None = None,
     ) -> list[T] | U:
         reduce_fn = reduce_fn if reduce_fn is not None else list
-        joblib_parallel = joblib.Parallel(n_jobs=self.config.n_jobs, verbose=self.config.verbose)
-        operation = self._bind(operation=operation, bind_kwargs=bind_kwargs)
-        return reduce_fn(joblib_parallel(joblib.delayed(operation)(task) for task in tasks))
+        with multiprocess.pool.Pool(self.config.n_jobs) as pool:
+            return reduce_fn(pool.map(self._bind(operation=operation, bind_kwargs=bind_kwargs), tasks))
 
     def _bind(
         self,