diff --git a/micro_sam/evaluation/inference.py b/micro_sam/evaluation/inference.py
index b033055f..5d86067a 100644
--- a/micro_sam/evaluation/inference.py
+++ b/micro_sam/evaluation/inference.py
@@ -550,9 +550,13 @@ def run_amg(
     iou_thresh_values: Optional[List[float]] = None,
     stability_score_values: Optional[List[float]] = None,
     peft_kwargs: Optional[Dict] = None,
+    cache_embeddings: bool = False
 ) -> str:
-    embedding_folder = os.path.join(experiment_folder, "embeddings")  # where the precomputed embeddings are saved
-    os.makedirs(embedding_folder, exist_ok=True)
+    if cache_embeddings:
+        embedding_folder = os.path.join(experiment_folder, "embeddings")  # where the precomputed embeddings are saved
+        os.makedirs(embedding_folder, exist_ok=True)
+    else:
+        embedding_folder = None
 
     predictor = util.get_sam_model(model_type=model_type, checkpoint_path=checkpoint, peft_kwargs=peft_kwargs)
     amg = AutomaticMaskGenerator(predictor)
@@ -572,9 +576,15 @@ def run_amg(
     )
 
     instance_segmentation.run_instance_segmentation_grid_search_and_inference(
-        amg, grid_search_values,
-        val_image_paths, val_gt_paths, test_image_paths,
-        embedding_folder, prediction_folder, gs_result_folder,
+        segmenter=amg,
+        grid_search_values=grid_search_values,
+        val_image_paths=val_image_paths,
+        val_gt_paths=val_gt_paths,
+        test_image_paths=test_image_paths,
+        embedding_dir=embedding_folder,
+        prediction_dir=prediction_folder,
+        result_dir=gs_result_folder,
+        experiment_folder=experiment_folder,
     )
     return prediction_folder
 
@@ -592,9 +602,13 @@ def run_instance_segmentation_with_decoder(
     val_gt_paths: List[Union[str, os.PathLike]],
     test_image_paths: List[Union[str, os.PathLike]],
     peft_kwargs: Optional[Dict] = None,
+    cache_embeddings: bool = False,
 ) -> str:
-    embedding_folder = os.path.join(experiment_folder, "embeddings")  # where the precomputed embeddings are saved
-    os.makedirs(embedding_folder, exist_ok=True)
+    if cache_embeddings:
+        embedding_folder = os.path.join(experiment_folder, "embeddings")  # where the precomputed embeddings are saved
+        os.makedirs(embedding_folder, exist_ok=True)
+    else:
+        embedding_folder = None
 
     predictor, decoder = get_predictor_and_decoder(
         model_type=model_type, checkpoint_path=checkpoint, peft_kwargs=peft_kwargs,
@@ -616,6 +630,6 @@ def run_instance_segmentation_with_decoder(
         segmenter, grid_search_values,
         val_image_paths, val_gt_paths, test_image_paths,
         embedding_dir=embedding_folder, prediction_dir=prediction_folder,
-        result_dir=gs_result_folder,
+        result_dir=gs_result_folder, experiment_folder=experiment_folder,
     )
     return prediction_folder
diff --git a/micro_sam/evaluation/instance_segmentation.py b/micro_sam/evaluation/instance_segmentation.py
index 5e657190..dbee0959 100644
--- a/micro_sam/evaluation/instance_segmentation.py
+++ b/micro_sam/evaluation/instance_segmentation.py
@@ -276,12 +276,15 @@ def run_instance_segmentation_inference(
         assert os.path.exists(image_path), image_path
         image = imageio.imread(image_path)
 
-        embedding_path = os.path.join(embedding_dir, f"{os.path.splitext(image_name)[0]}.zarr")
-        image_embeddings = util.precompute_image_embeddings(
-            predictor, image, embedding_path, ndim=2, verbose=verbose_embeddings
-        )
+        if embedding_dir is None:
+            segmenter.initialize(image)
+        else:
+            embedding_path = os.path.join(embedding_dir, f"{os.path.splitext(image_name)[0]}.zarr")
+            image_embeddings = util.precompute_image_embeddings(
+                predictor, image, embedding_path, ndim=2, verbose=verbose_embeddings
+            )
+            segmenter.initialize(image, image_embeddings)
 
-        segmenter.initialize(image, image_embeddings)
         masks = segmenter.generate(**generate_kwargs)
 
         if len(masks) == 0:  # the instance segmentation can have no masks, hence we just save empty labels
@@ -362,6 +365,7 @@ def run_instance_segmentation_grid_search_and_inference(
     test_image_paths: List[Union[str, os.PathLike]],
     embedding_dir: Union[str, os.PathLike],
     prediction_dir: Union[str, os.PathLike],
+    experiment_folder: Union[str, os.PathLike],
     result_dir: Union[str, os.PathLike],
     fixed_generate_kwargs: Optional[Dict[str, Any]] = None,
     verbose_gs: bool = True,
@@ -379,6 +383,7 @@ def run_instance_segmentation_grid_search_and_inference(
         test_image_paths: The input images for inference.
         embedding_dir: Folder to cache the image embeddings.
         prediction_dir: Folder to save the predictions.
+        experiment_dir: Folder for caching best 
         result_dir: Folder to cache the evaluation results per image.
         fixed_generate_kwargs: Fixed keyword arguments for the `generate` method of the segmenter.
         verbose_gs: Whether to run the gridsearch for individual images in a verbose mode.
@@ -394,7 +399,7 @@ def run_instance_segmentation_grid_search_and_inference(
     print("Best grid-search result:", best_msa, "with parmeters:\n", best_param_str)
     print()
 
-    save_grid_search_best_params(best_kwargs, best_msa, Path(embedding_dir).parent)
+    save_grid_search_best_params(best_kwargs, best_msa, experiment_folder)
 
     generate_kwargs = {} if fixed_generate_kwargs is None else fixed_generate_kwargs
     generate_kwargs.update(best_kwargs)