Visualization block copy

inference/core/workflows/core_steps/visualizations/bounding_box/v1.py

@@ -41,13 +41,13 @@ class BoundingBoxManifest(ColorableVisualizationManifest):
     )
 
     thickness: Union[int, Selector(kind=[INTEGER_KIND])] = Field(  # type: ignore
-        description="Thickness of the bounding box in pixels.",
+        description="Set the thickness of the bounding box edges.",
         default=2,
         examples=[2, "$inputs.thickness"],
     )
 
     roundness: Union[FloatZeroToOne, Selector(kind=[FLOAT_ZERO_TO_ONE_KIND])] = Field(  # type: ignore
-        description="Roundness of the corners of the bounding box.",
+        description="Define the roundness of the bounding box corners.",
         default=0.0,
         examples=[0.0, "$inputs.roundness"],
     )

inference/core/workflows/core_steps/visualizations/common/base.py

@@ -34,12 +34,12 @@ class VisualizationManifest(WorkflowBlockManifest, ABC):
     )
     image: Selector(kind=[IMAGE_KIND]) = Field(
         title="Input Image",
-        description="The input image for this step.",
+        description="Select the input image to visualize on.",
         examples=["$inputs.image", "$steps.cropping.crops"],
         validation_alias=AliasChoices("image", "images"),
     )
     copy_image: Union[bool, Selector(kind=[BOOLEAN_KIND])] = Field(  # type: ignore
-        description="Duplicate the image contents (vs overwriting the image in place). Deselect for chained visualizations that should stack on previous ones where the intermediate state is not needed.",
+        description="Enable this option to create a copy of the input image for visualization, preserving the original. Use this when stacking multiple visualizations.",
         default=True,
         examples=[True, False],
     )

inference/core/workflows/core_steps/visualizations/common/base_colorable.py

@@ -77,7 +77,7 @@ class ColorableVisualizationManifest(PredictionsVisualizationManifest, ABC):
         Selector(kind=[STRING_KIND]),
     ] = Field(  # type: ignore
         default="DEFAULT",
-        description="Color palette to use for annotations.",
+        description="Select a color palette for the visualised elements.",
         examples=["DEFAULT", "$inputs.color_palette"],
     )
 
@@ -86,14 +86,14 @@ class ColorableVisualizationManifest(PredictionsVisualizationManifest, ABC):
         Selector(kind=[INTEGER_KIND]),
     ] = Field(  # type: ignore
         default=10,
-        description="Number of colors in the color palette. Applies when using a matplotlib `color_palette`.",
+        description="Specify the number of colors in the palette. This applies when using custom or Matplotlib palettes.",
         examples=[10, "$inputs.palette_size"],
     )
 
     custom_colors: Union[List[str], Selector(kind=[LIST_OF_VALUES_KIND])] = (
         Field(  # type: ignore
             default=[],
-            description='List of colors to use for annotations when `color_palette` is set to "CUSTOM".',
+            description="Define a list of custom colors for bounding boxes in HEX format.",
             examples=[["#FF0000", "#00FF00", "#0000FF"], "$inputs.custom_colors"],
         )
     )
@@ -103,7 +103,7 @@ class ColorableVisualizationManifest(PredictionsVisualizationManifest, ABC):
         Selector(kind=[STRING_KIND]),
     ] = Field(  # type: ignore
         default="CLASS",
-        description="Strategy to use for mapping colors to annotations.",
+        description="Choose how bounding box colors are assigned.",
         examples=["CLASS", "$inputs.color_axis"],
     )
 

inference/core/workflows/prototypes/block.py

@@ -24,7 +24,9 @@ class WorkflowBlockManifest(BaseModel, ABC):
     )
 
     type: str
-    name: str = Field(title="Step Name", description="The unique name of this step.")
+    name: str = Field(
+        title="Step Name", description="Enter a unique identifier for this step."
+    )
 
     @classmethod
     @abstractmethod

tests/inference/models_predictions_tests/test_owlv2.py

@@ -1,13 +1,20 @@
-import pytest
 import gc
-from unittest.mock import MagicMock
 import os
+from unittest.mock import MagicMock
+
+import pytest
 import torch
+
+from inference.core.cache.model_artifacts import get_cache_file_path
 from inference.core.entities.requests.inference import ObjectDetectionInferenceRequest
 from inference.core.entities.requests.owlv2 import OwlV2InferenceRequest
-from inference.models.owlv2.owlv2 import OwlV2, SerializedOwlV2, Owlv2Singleton, LazyImageRetrievalWrapper
 from inference.core.env import OWLV2_VERSION_ID
-from inference.core.cache.model_artifacts import get_cache_file_path
+from inference.models.owlv2.owlv2 import (
+    LazyImageRetrievalWrapper,
+    OwlV2,
+    Owlv2Singleton,
+    SerializedOwlV2,
+)
 
 
 @pytest.mark.slow
@@ -58,13 +65,14 @@ def test_owlv2():
     assert abs(532 - posts[3].x) < 1.5
     assert abs(572 - posts[4].x) < 1.5
 
+
 def test_owlv2_serialized():
     image = {
         "type": "url",
         "value": "https://media.roboflow.com/inference/seawithdock.jpeg",
     }
 
-    training_data=[
+    training_data = [
         {
             "image": image,
             "boxes": [
@@ -93,19 +101,20 @@ def test_owlv2_serialized():
         hf_id=f"google/{OWLV2_VERSION_ID}",
     )
     assert os.path.exists(serialized_pt)
-    pt_path = get_cache_file_path(file=SerializedOwlV2.weights_file_path, model_id=model_id)
+    pt_path = get_cache_file_path(
+        file=SerializedOwlV2.weights_file_path, model_id=model_id
+    )
     os.makedirs(os.path.dirname(pt_path), exist_ok=True)
     os.rename(serialized_pt, pt_path)
     serialized_owlv2 = SerializedOwlV2(model_id=model_id)
-    
+
     # Get the image hash before inference
     image_wrapper = LazyImageRetrievalWrapper(request.image)
     image_hash = image_wrapper.image_hash
     assert image_hash in serialized_owlv2.owlv2.cpu_image_embed_cache
-    
+
     response = serialized_owlv2.infer_from_request(request)
-
-
+
     assert len(response.predictions) == 5
     posts = [p for p in response.predictions if p.class_name == "post"]
     posts.sort(key=lambda x: x.x)
@@ -114,15 +123,13 @@ def test_owlv2_serialized():
     assert abs(264 - posts[2].x) < 1.5
     assert abs(532 - posts[3].x) < 1.5
     assert abs(572 - posts[4].x) < 1.5
-    
+
     pt_path = serialized_owlv2.save_small_model_without_image_embeds()
     assert os.path.exists(pt_path)
     pt_dict = torch.load(pt_path)
     assert len(pt_dict["image_embeds"]) == 0
 
 
-
-
 @pytest.mark.slow
 def test_owlv2_multiple_prompts():
     image = {
@@ -411,12 +418,14 @@ def test_owlv2_multiple_training_images_repeated_inference():
         assert p1.height == p2.height
         assert p1.confidence == p2.confidence
 
+
 @pytest.mark.slow
 def test_owlv2_model_unloaded_when_garbage_collected():
     model = OwlV2()
     del model
     gc.collect()
     assert len(Owlv2Singleton._instances) == 0
 
+
 if __name__ == "__main__":
     test_owlv2()

tests/inference/models_predictions_tests/test_sam2.py

@@ -17,6 +17,7 @@
 from inference.core.workflows.core_steps.models.foundation.segment_anything2.v1 import (
     convert_sam2_segmentation_response_to_inference_instances_seg_response,
 )
+
 try:
     from inference.models.sam2 import SegmentAnything2
     from inference.models.sam2.segment_anything2 import (

tests/inference_cli/unit_tests/lib/workflows/test_common.py

@@ -86,25 +86,22 @@ def test_aggregate_batch_processing_results_when_json_output_is_expected_and_res
                 continue
             decoded_results.append(json.loads(line))
     print(decoded_results)
-    assert (
-        decoded_results
-        == [
-            {
-                "some": "value",
-                "image": "other.jpg",
-                "other": 3.0,
-                "list_field": [1, 2, 3],
-                "object_field": {"nested": "value"},
-            },
-            {
-                "some": "value",
-                "image": "some.jpg",
-                "other": 3.0,
-                "list_field": [1, 2, 3],
-                "object_field": {"nested": "value"},
-            }
-        ]
-    )
+    assert decoded_results == [
+        {
+            "some": "value",
+            "image": "other.jpg",
+            "other": 3.0,
+            "list_field": [1, 2, 3],
+            "object_field": {"nested": "value"},
+        },
+        {
+            "some": "value",
+            "image": "some.jpg",
+            "other": 3.0,
+            "list_field": [1, 2, 3],
+            "object_field": {"nested": "value"},
+        },
+    ]
 
 
 def test_aggregate_batch_processing_results_when_json_output_is_expected_and_results_not_present(

tests/workflows/unit_tests/core_steps/cache/test_cache.py

@@ -2,20 +2,16 @@
 
 import numpy as np
 
+from inference.core.workflows.core_steps.cache.cache_get.v1 import CacheGetBlockV1
+from inference.core.workflows.core_steps.cache.cache_set.v1 import CacheSetBlockV1
 from inference.core.workflows.core_steps.common.entities import StepExecutionMode
-from inference.core.workflows.core_steps.cache.cache_get.v1 import (
-    CacheGetBlockV1,
-)
-from inference.core.workflows.core_steps.cache.cache_set.v1 import (
-    CacheSetBlockV1,
-)
-
 from inference.core.workflows.execution_engine.entities.base import (
     ImageParentMetadata,
     VideoMetadata,
     WorkflowImageData,
 )
 
+
 def test_cache_on_video() -> None:
     # given
     metadata = VideoMetadata(
@@ -34,28 +30,19 @@ def test_cache_on_video() -> None:
     cache_set_block = CacheSetBlockV1(step_execution_mode=StepExecutionMode.LOCAL)
 
     # empty result
-    get_empty = cache_get_block.run(
-        image=image,
-        key="foo"
-    )
+    get_empty = cache_get_block.run(image=image, key="foo")
     assert get_empty == {
         "output": False,
     }
 
     # set then get
-    cache_set_block.run(
-        image=image,
-        key="foo",
-        value="bar"
-    )
-    get_full = cache_get_block.run(
-        image=image,
-        key="foo"
-    )
+    cache_set_block.run(image=image, key="foo", value="bar")
+    get_full = cache_get_block.run(image=image, key="foo")
     assert get_full == {
         "output": "bar",
     }
 
+
 def test_cache_with_no_metadata() -> None:
     # given
     image = WorkflowImageData(
@@ -66,28 +53,19 @@ def test_cache_with_no_metadata() -> None:
     cache_set_block = CacheSetBlockV1(step_execution_mode=StepExecutionMode.LOCAL)
 
     # empty result
-    get_empty = cache_get_block.run(
-        image=image,
-        key="foo"
-    )
+    get_empty = cache_get_block.run(image=image, key="foo")
     assert get_empty == {
         "output": False,
     }
 
     # set then get
-    cache_set_block.run(
-        image=image,
-        key="foo",
-        value="bar"
-    )
-    get_full = cache_get_block.run(
-        image=image,
-        key="foo"
-    )
+    cache_set_block.run(image=image, key="foo", value="bar")
+    get_full = cache_get_block.run(image=image, key="foo")
     assert get_full == {
         "output": "bar",
     }
 
+
 def test_cache_on_multiple_videos() -> None:
     # given
     metadata_1 = VideoMetadata(
@@ -120,33 +98,20 @@ def test_cache_on_multiple_videos() -> None:
     cache_set_block = CacheSetBlockV1(step_execution_mode=StepExecutionMode.LOCAL)
 
     # empty result
-    get_empty = cache_get_block.run(
-        image=image_1,
-        key="foo"
-    )
+    get_empty = cache_get_block.run(image=image_1, key="foo")
     assert get_empty == {
         "output": False,
     }
 
     # set then get
-    cache_set_block.run(
-        image=image_1,
-        key="foo",
-        value="bar"
-    )
-    get_full = cache_get_block.run(
-        image=image_1,
-        key="foo"
-    )
+    cache_set_block.run(image=image_1, key="foo", value="bar")
+    get_full = cache_get_block.run(image=image_1, key="foo")
     assert get_full == {
         "output": "bar",
     }
 
     # make sure it doesn't bleed over
-    get_empty = cache_get_block.run(
-        image=image_2,
-        key="foo"
-    )
+    get_empty = cache_get_block.run(image=image_2, key="foo")
     assert get_empty == {
         "output": False,
-    }
+    }