[DOCS] Update how to use package in IDE (#1682)

* added build artifacts dirs to gitignore
initial commit - 1 GPU, 1- show_in_window

* added show_in_window to all of the check tutorials.
MAJOR BUGS:
1. yolo related error in examples calling yolo AttributeError: 'Upsample' object has no attribute 'recompute_scale_factor'
2. duplicate display in all of the checks

* limiting torchvision version to 0.11.2 solves the YOLO bug
some displays are still duplicated but not all of them

* GPU comment added to the documentation for all of the relevant checks
Duplicate display bug not solved

Co-authored-by: Itay Gabbay <itay@deepchecks.com>

.gitignore

@@ -112,6 +112,10 @@ docs/source/user-guide/vision/auto_tutorials
 docs/source/tutorials/vision/*.html
 docs/source/tutorials/vision/deepchecks_formatted_image*.jpg
 docs/source/tutorials/vision/tomato-detection/
+docs/source/user-guide/vision/tutorials/tomato-detection
+docs/source/user-guide/vision/tutorials/hymenoptera_data
 docs/source/tutorials/vision/hymenoptera_data/
+docs/source/user-guide/vision/tutorials/EuroSAT
 docs/source/user-guide/general/exporting_results/wandb
 deepchecks/vision/datasets/classification/models/
+

docs/requirements.txt

@@ -13,5 +13,5 @@ readthedocs-sphinx-search
 numpydoc>=1.1.0
 pypandoc>=1.7.2
 sphinx-reredirects>=0.0.1
-ipywidgets @ git+https://github.com/deepchecks/ipywidgets.git@8ac487b64ffd48dbeaa0a47d1997be27f6052dbc
+#ipywidgets @ git+https://github.com/deepchecks/ipywidgets.git@8ac487b64ffd48dbeaa0a47d1997be27f6052dbc
 docutils

docs/source/checks/vision/data_integrity/plot_image_property_outliers.py

@@ -55,6 +55,13 @@
 result = check.run(train_data)
 result
 
+#%%
+# To display the results in an IDE like PyCharm, you can use the following code:
+
+#  result.show_in_window()
+#%%
+# The result will be displayed in a new window.
+
 #%%
 # Observe Graphic Result
 # ^^^^^^^^^^^^^^^^^^^^^^
@@ -72,5 +79,3 @@
 # in the check's run.
 
 result.value
-
-

docs/source/checks/vision/data_integrity/plot_label_property_outliers.py

@@ -54,6 +54,13 @@
 result = check.run(train_data)
 result
 
+#%%
+# To display the results in an IDE like PyCharm, you can use the following code:
+
+#  result.show_in_window()
+#%%
+# The result will be displayed in a new window.
+
 #%%
 # Observe Graphic Result
 # ^^^^^^^^^^^^^^^^^^^^^^

docs/source/checks/vision/model_evaluation/plot_class_performance.py

@@ -66,6 +66,18 @@
 check = ClassPerformance()
 check.run(train_ds, test_ds, mnist_model)
 
+#%%
+# If you have a GPU, you can speed up this check by calling:
+
+# check.run(train_ds, test_ds, mnist_model, device=<your GPU>)
+
+#%%
+# To display the results in an IDE like PyCharm, you can use the following code:
+
+#  check.run(train_ds, test_ds, mnist_model).show_in_window()
+#%%
+# The result will be displayed in a new window.
+
 #%%
 # Object Detection Class Performance
 # ==================================
@@ -94,6 +106,18 @@
 check = ClassPerformance(show_only='best')
 check.run(train_ds, test_ds, yolo)
 
+#%%
+# If you have a GPU, you can speed up this check by calling:
+
+# check.run(train_ds, test_ds, yolo, device=<your GPU>)
+
+#%%
+# To display the results in an IDE like PyCharm, you can use the following code:
+
+# check.run(train_ds, test_ds, yolo).show_in_window()
+#%%
+# The result will be displayed in a new window.
+
 #%%
 # Define a Condition
 # ==================
@@ -106,5 +130,6 @@
 result = check.run(train_ds, test_ds, yolo)
 result
 
+
 #%%
 # We detected that for several classes our model performance is below the threshold.

docs/source/checks/vision/model_evaluation/plot_confusion_matrix.py

@@ -37,3 +37,15 @@
 
 check = ConfusionMatrixReport(categories_to_display=10)
 check.run(train_ds, yolo)
+
+#%%
+# If you have a GPU, you can speed up this check by calling:
+
+# check.run(train_ds, yolo, device=<your GPU>)
+
+#%%
+# To display the results in an IDE like PyCharm, you can use the following code:
+
+#  check.run(train_ds, yolo).show_in_window()
+#%%
+# The result will be displayed in a new window.

docs/source/checks/vision/model_evaluation/plot_image_segment_performance.py

@@ -33,6 +33,16 @@
 result = ImageSegmentPerformance().run(coco_data, model)
 result
 
+#%%
+# If you have a GPU, you can speed up this check by passing it as an argument to .run() as device=<your GPU>
+#
+# To display the results in an IDE like PyCharm, you can use the following code:
+
+#  result.show_in_window()
+#%%
+# The result will be displayed in a new window.
+
+
 #%%
 # Observe the check’s output
 # --------------------------

docs/source/checks/vision/model_evaluation/plot_mean_average_precision_report.py

@@ -44,6 +44,16 @@
 result = check.run(test_ds, yolo)
 result
 
+#%%
+# If you have a GPU, you can speed up this check by passing it as an argument to .run() as device=<your GPU>
+#
+# To display the results in an IDE like PyCharm, you can use the following code:
+
+#  result.show_in_window()
+#%%
+# The result will be displayed in a new window.
+
+
 # %%
 # Observe the check’s output
 # --------------------------

docs/source/checks/vision/model_evaluation/plot_mean_average_recall_report.py

@@ -50,6 +50,15 @@
 result = check.run(test_ds, yolo)
 result
 
+#%%
+# If you have a GPU, you can speed up this check by passing it as an argument to .run() as device=<your GPU>
+#
+# To display the results in an IDE like PyCharm, you can use the following code:
+
+#  result.show_in_window()
+#%%
+# The result will be displayed in a new window.
+
 #%%
 # Observe the check’s output
 # --------------------------

docs/source/checks/vision/model_evaluation/plot_model_error_analysis.py

@@ -51,6 +51,16 @@
 check = ModelErrorAnalysis(min_error_model_score=-0.1)
 check.run(train_ds, test_ds, mnist_model)
 
+#%%
+# If you have a GPU, you can speed up this check by passing it as an argument to .run() as device=<your GPU>
+#
+# To display the results in an IDE like PyCharm, you can use the following code:
+
+#  check.run(train_ds, test_ds, mnist_model).show_in_window()
+#%%
+# The result will be displayed in a new window.
+
+
 #%%
 # Object Detection Class Performance
 # ----------------------------------

docs/source/checks/vision/model_evaluation/plot_robustness_report.py

@@ -91,6 +91,15 @@
 result = RobustnessReport().run(mnist_dataloader_test, model)
 result
 
+#%%
+# If you have a GPU, you can speed up this check by passing it as an argument to .run() as device=<your GPU>
+#
+# To display the results in an IDE like PyCharm, you can use the following code:
+
+#  result.show_in_window()
+#%%
+# The result will be displayed in a new window.
+
 #%%
 # Observe the check’s output
 # --------------------------

docs/source/checks/vision/model_evaluation/plot_simple_model_comparison.py

@@ -69,6 +69,15 @@ class has equal probability.
 #%%
 result
 
+#%%
+# If you have a GPU, you can speed up this check by passing it as an argument to .run() as device=<your GPU>
+#
+# To display the results in an IDE like PyCharm, you can use the following code:
+
+#  result.show_in_window()
+#%%
+# The result will be displayed in a new window.
+
 #%%
 # Observe the check's output
 # --------------------------

docs/source/checks/vision/model_evaluation/plot_single_dataset_scalar_performance.py

@@ -50,6 +50,15 @@
 result = check.run(train_ds, mnist_model)
 result
 
+#%%
+# If you have a GPU, you can speed up this check by passing it as an argument to .run() as device=<your GPU>
+#
+# To display the results in an IDE like PyCharm, you can use the following code:
+
+#  result.show_in_window()
+#%%
+# The result will be displayed in a new window.
+
 #%%
 # The result value is a dictionary with the following fields:
 # score - the actual result,

docs/source/checks/vision/model_evaluation/plot_train_test_prediction_drift.py

@@ -84,6 +84,15 @@
 check = TrainTestPredictionDrift()
 check.run(train_ds, test_ds, model)
 
+#%%
+# If you have a GPU, you can speed up this check by passing it as an argument to .run() as device=<your GPU>
+#
+# To display the results in an IDE like PyCharm, you can use the following code:
+
+#  check.run(train_ds, test_ds, model).show_in_window()
+#%%
+# The result will be displayed in a new window.
+
 #%%
 # Understanding the results
 # -------------------------
@@ -146,12 +155,14 @@ def collate_test(batch):
 mod_train_ds = MNISTData(mod_train_loader)
 mod_test_ds = MNISTData(mod_test_loader)
 
+#%%
 # Run the check
 # -------------
 
 check = TrainTestPredictionDrift()
 check.run(mod_train_ds, mod_test_ds, model)
 
+#%%
 # Add a condition
 # ---------------
 # We could also add a condition to the check to alert us to changes in the prediction

docs/source/checks/vision/train_test_validation/plot_feature_label_correlation_change.py

@@ -160,6 +160,13 @@ def mnist_batch_to_images_with_bias(batch):
 check = FeatureLabelCorrelationChange()
 check.run(train_ds, test_ds)
 
+#%%
+# To display the results in an IDE like PyCharm, you can use the following code:
+
+#  check.run(train_ds, test_ds).show_in_window()
+#%%
+# The result will be displayed in a new window.
+
 #%%
 # We can see that the check detected the bias we inserted, and that the
 # brightness property of the image has a high PPS, meaning it can be used to

docs/source/checks/vision/train_test_validation/plot_heatmap_comparison.py

@@ -86,6 +86,13 @@
 check = HeatmapComparison()
 check.run(train_ds, test_ds)
 
+#%%
+# To display the results in an IDE like PyCharm, you can use the following code:
+
+#  check.run(train_ds, test_ds).show_in_window()
+#%%
+# The result will be displayed in a new window.
+
 #%%
 # Limit to Specific Classes
 # =========================

docs/source/checks/vision/train_test_validation/plot_image_dataset_drift.py

@@ -118,6 +118,12 @@ def batch_to_images(self, batch):
 check = ImageDatasetDrift()
 check.run(train_dataset=drifted_train_ds, test_dataset=test_ds_coco)
 
+#%%
+# To display the results in an IDE like PyCharm, you can use the following code:
+
+#  check.run(drifted_train_ds, test_ds_coco).show_in_window()
+#%%
+# The result will be displayed in a new window.
 
 #%%
 # Define a Condition

docs/source/checks/vision/train_test_validation/plot_image_property_drift.py

@@ -80,6 +80,13 @@
 check_result = ImagePropertyDrift().run(train_dataset, test_dataset)
 check_result
 
+#%%
+# To display the results in an IDE like PyCharm, you can use the following code:
+
+#  check_result.show_in_window()
+#%%
+# The result will be displayed in a new window.
+
 #%%
 # Observe the check’s output 
 # --------------------------

docs/source/checks/vision/train_test_validation/plot_new_labels.py

@@ -31,6 +31,13 @@
 result = NewLabels().run(coco_train, coco_test)
 result
 
+#%%
+# To display the results in an IDE like PyCharm, you can use the following code:
+
+#  result.show_in_window()
+#%%
+# The result will be displayed in a new window.
+
 # %%
 # Observe the check’s output
 # ~~~~~~~~~~~~~~~~~~

docs/source/checks/vision/train_test_validation/plot_similar_image_leakage.py

@@ -64,6 +64,13 @@
 check = SimilarImageLeakage()
 check.run(train_ds, test_ds)
 
+#%%
+# To display the results in an IDE like PyCharm, you can use the following code:
+
+#  check.run(train_ds, test_ds).show_in_window()
+#%%
+# The result will be displayed in a new window.
+
 #%%
 # As we can see, no similar images were found.
 

docs/source/checks/vision/train_test_validation/plot_train_test_label_drift.py

@@ -85,6 +85,13 @@
 mnist_model = load_mnist_model(pretrained=True)
 ClassPerformance().run(train_ds, test_ds, mnist_model)
 
+#%%
+# To display the results in an IDE like PyCharm, you can use the following code:
+
+#  ClassPerformance().run(train_ds, test_ds, mnist_model).show_in_window()
+#%%
+# The result will be displayed in a new window.
+
 #%%
 # MNIST with label drift
 # ======================

makefile

@@ -164,7 +164,7 @@ requirements: $(ENV)
 			"torch==1.10.2+cpu" "torchvision==0.11.3+cpu" "torchaudio==0.10.2+cpu" \
 			-f https://s3.amazonaws.com/pytorch/whl/torch_stable.html; \
 	else \
-		$(PIP) install -q torch torchvision torchaudio; \
+		$(PIP) install -q torch "torchvision==0.11.2" torchaudio; \
 	fi;
 
 	@$(PIP) install -U pip