alibaba env var fix

.envrc

@@ -0,0 +1,5 @@
+if ! has nix_direnv_version || ! nix_direnv_version 3.0.6; then
+  source_url "https://raw.githubusercontent.com/nix-community/nix-direnv/3.0.6/direnvrc" "sha256-RYcUJaRMf8oF5LznDrlCXbkOQrywm0HDv1VjYGaJGdM="
+fi
+use flake .
+dotenv

dimos/perception/detection2d/module2D.py

@@ -23,7 +23,6 @@
 from reactivex.subject import Subject
 
 from dimos.core import In, Module, Out, rpc
-from dimos.models.vl import QwenVlModel, VlModel
 from dimos.msgs.sensor_msgs import Image
 from dimos.msgs.sensor_msgs.Image import sharpness_barrier
 from dimos.msgs.vision_msgs import Detection2DArray
@@ -40,7 +39,6 @@
 class Config:
     max_freq: float = 5  # hz
     detector: Optional[Callable[[Any], Detector]] = lambda: Yolo2DDetector()
-    vlmodel: VlModel = QwenVlModel
 
 
 class Detection2DModule(Module):
@@ -52,6 +50,7 @@ class Detection2DModule(Module):
     detections: Out[Detection2DArray] = None  # type: ignore
     annotations: Out[ImageAnnotations] = None  # type: ignore
 
+    # just for visualization, emits latest top 3 detections in a frame
     detected_image_0: Out[Image] = None  # type: ignore
     detected_image_1: Out[Image] = None  # type: ignore
     detected_image_2: Out[Image] = None  # type: ignore
@@ -60,7 +59,6 @@ def __init__(self, *args, **kwargs):
         super().__init__(*args, **kwargs)
         self.config: Config = Config(**kwargs)
         self.detector = self.config.detector()
-        self.vlmodel = self.config.vlmodel()
         self.vlm_detections_subject = Subject()
 
     def process_image_frame(self, image: Image) -> ImageDetections2D:

dimos/types/test_timestamped.py

@@ -330,7 +330,7 @@ def process_video_frame(frame):
         )
         assert diff <= 0.05
 
-    assert len(aligned_frames) > 3
+    assert len(aligned_frames) > 2
 
 
 def test_timestamp_alignment_primary_first():