added auto-detection for the output type on the client side to be able to download binary files such as videos

source/manager/ModelManager.py

@@ -1,3 +1,4 @@
+import asyncio
 import json
 import os
 import typing
@@ -10,6 +11,11 @@ from source import model, api
 
 
 class ModelManager:
+    """
+    The model manager
+    Load the list of models available, ensure that only one model is loaded at the same time.
+    """
+
     def __init__(self, application: api.Application, model_library: os.PathLike | str):
         self.application: api.Application = application
         self.model_library: Path = Path(model_library)
@@ -20,7 +26,9 @@ class ModelManager:
         self.models: dict[str, model.base.BaseModel] = {}
 
         # the currently loaded model
-        # TODO(Faraphel): load more than one model at a time ? require a way more complex manager to handle memory issue
+        # TODO(Faraphel): load more than one model at a time ?
+        #  would require a way more complex manager to handle memory issue
+        #  having two calculations at the same time might not be worth it either
         self.current_loaded_model: typing.Optional[model.base.BaseModel] = None
 
         # lock to avoid concurrent inference and concurrent model loading and unloading