typo in label studio

src/label_studio.py

@@ -249,7 +249,7 @@ def import_image_tasks(label_studio_project, image_names, local_image_dir, predi
     for index, image_name in enumerate(image_names):
         print(f"Importing {image_name} into Label Studio")
         data_dict = {'image': os.path.join("/data/local-files/?d=input/", os.path.basename(image_name))}
-        if predictions:
+        if predictions is not None:
             prediction = predictions[index]
             # Skip predictions if there are none
             if prediction.empty:

src/pipeline.py

@@ -155,7 +155,8 @@ def run(self):
         # Align the predictions with the cropped images
         # Run the annotation pipeline
         if len(image_paths) > 0:
-            label_studio.upload_to_label_studio(images=image_paths, 
+            full_image_paths = [os.path.join(self.config.active_learning.image_dir, image) for image in image_paths]
+            label_studio.upload_to_label_studio(images=full_image_paths, 
                                                 sftp_client=self.sftp_client, 
                                                 label_studio_project=self.label_studio_project, 
                                                 images_to_annotate_dir=self.config.active_learning.image_dir, 

src/reporting.py

@@ -18,16 +18,25 @@ def __init__(self, report_dir, image_dir, pipeline_monitor=None):
         self.image_dir = image_dir
         self.pipeline_monitor = pipeline_monitor
 
+    def concat_predictions(self):
+        """Concatenate predictions
+        
+        Args:
+            predictions: List of dataframes containing predictions
+        """
+        self.all_predictions = pd.concat(self.pipeline_monitor.predictions, ignore_index=True)
+
     def generate_report(self):
         """Generate a report"""
-        self.all_predictions = pd.concat(self.pipeline_monitor.predictions)
 
+        self.concat_predictions()
         self.write_predictions()
         self.write_metrics()
         self.generate_video()
 
     def write_predictions(self):
         """Write predictions to a csv file"""
+        self.concat_predictions()
         self.all_predictions.to_csv(f"{self.report_dir}/predictions.csv", index=False)
 
         return f"{self.report_dir}/predictions.csv"
@@ -37,6 +46,7 @@ def get_coco_datasets(self):
 
     def generate_video(self):
         """Generate a video from the predictions"""
+        self.concat_predictions()
         visualizer = PredictionVisualizer(self.all_predictions, self.report_dir)
         output_path = f"{self.report_dir}/predictions.mp4"
         images = self.all_predictions['image_path'].unique()
@@ -52,6 +62,7 @@ def write_metrics(self):
         Args:
             pipeline_monitor: PipelineEvaluation instance containing model performance metrics
         """
+        self.concat_predictions()
         # Get current timestamp
         timestamp = datetime.now().strftime("%Y-%m-%d %H:%M:%S")
 

submit.sh

@@ -10,9 +10,9 @@
 #SBATCH --output=/home/b.weinstein/logs/BOEM%j.out   # Standard output and error log
 #SBATCH --error=/home/b.weinstein/logs/BOEM%j.err
 #SBATCH --partition=gpu
-#SBATCH --gpus=3
+#SBATCH --gpus=1
 
 source activate BOEM
 
 cd ~/BOEM/
-python main.py active_learning.gpus = 3
+python main.py active_learning.pool_limit=5000 active_learning.n_images=100

tests/test_reporting.py

@@ -29,7 +29,7 @@ def pipeline_monitor(config, random_model, random_crop_model):
 
 @pytest.fixture
 def reporter(config, pipeline_monitor, tmpdir_factory):
-    report_dir = tmpdir_factory.mktemp("report_dir").strpath
+    report_dir = tmpdir_factory.mktemp("report_dir").strpath  
     return reporting.Reporting(report_dir, config.reporting.image_dir, pipeline_monitor)
 
 def test_generate_video(reporter):