diff --git a/supervisely/train/requirements.txt b/supervisely/train/requirements.txt
index a39f6deb9b05..073b87a59366 100644
--- a/supervisely/train/requirements.txt
+++ b/supervisely/train/requirements.txt
@@ -1 +1 @@
-supervisely==6.73.4
+supervisely==6.73.41
diff --git a/supervisely/train/src/sly_project_cached.py b/supervisely/train/src/sly_project_cached.py
new file mode 100644
index 000000000000..5fcce7d5b886
--- /dev/null
+++ b/supervisely/train/src/sly_project_cached.py
@@ -0,0 +1,74 @@
+import os
+
+import supervisely as sly
+from supervisely.project.download import (
+    download_to_cache,
+    copy_from_cache,
+    is_cached,
+    get_cache_size,
+)
+from sly_utils import get_progress_cb
+import sly_train_globals as g
+
+
+def download_project(
+    api: sly.Api,
+    project_info: sly.ProjectInfo,
+    project_dir: str,
+    use_cache: bool,
+):
+    if os.path.exists(project_dir):
+        sly.fs.clean_dir(project_dir)
+    if not use_cache:
+        total = project_info.items_count
+        download_progress = get_progress_cb("Downloading input data...", total * 2)
+        sly.download(
+            api=api,
+            project_id=project_info.id,
+            dest_dir=project_dir,
+            dataset_ids=None,
+            log_progress=True,
+            progress_cb=download_progress,
+            cache=g.my_app.cache
+        )
+        return
+
+    # get datasets to download and cached
+    dataset_infos = api.dataset.get_list(project_info.id)
+    to_download = [info for info in dataset_infos if not is_cached(project_info.id, info.name)]
+    cached = [info for info in dataset_infos if is_cached(project_info.id, info.name)]
+    if len(cached) == 0:
+        log_msg = "No cached datasets found"
+    else:
+        log_msg = "Using cached datasets: " + ", ".join(
+            f"{ds_info.name} ({ds_info.id})" for ds_info in cached
+        )
+    sly.logger.info(log_msg)
+    if len(to_download) == 0:
+        log_msg = "All datasets are cached. No datasets to download"
+    else:
+        log_msg = "Downloading datasets: " + ", ".join(
+            f"{ds_info.name} ({ds_info.id})" for ds_info in to_download
+        )
+    sly.logger.info(log_msg)
+    # get images count
+    total = sum([ds_info.images_count for ds_info in to_download])
+    # download
+    download_progress = get_progress_cb("Downloading input data...", total * 2)
+    download_to_cache(
+        api=api,
+        project_id=project_info.id,
+        dataset_infos=to_download,
+        log_progress=True,
+        progress_cb=download_progress,
+    )
+    # copy datasets from cache
+    total = sum([get_cache_size(project_info.id, ds.name) for ds in dataset_infos])
+    dataset_names = [ds_info.name for ds_info in dataset_infos]
+    download_progress = get_progress_cb("Retreiving data from cache...", total, is_size=True)
+    copy_from_cache(
+        project_id=project_info.id,
+        dest_dir=project_dir,
+        dataset_names=dataset_names,
+        progress_cb=download_progress,
+    )
diff --git a/supervisely/train/src/sly_train.py b/supervisely/train/src/sly_train.py
index 1aadd1d3137e..527d8a3db079 100644
--- a/supervisely/train/src/sly_train.py
+++ b/supervisely/train/src/sly_train.py
@@ -10,6 +10,7 @@
     root_source_dir, scratch_str, finetune_str
 
 import ui as ui
+from sly_project_cached import download_project
 from sly_train_utils import init_script_arguments
 from sly_utils import get_progress_cb, upload_artifacts
 from splits import get_train_val_sets, verify_train_val_sets
@@ -39,9 +40,13 @@ def train(api: sly.Api, task_id, context, state, app_logger):
         sly.fs.mkdir(project_dir, remove_content_if_exists=True)  # clean content for debug, has no effect in prod
 
         # download and preprocess Sypervisely project (using cache)
-        download_progress = get_progress_cb("Download data (using cache)", g.project_info.items_count * 2)
         try:
-            sly.download_project(api, project_id, project_dir, cache=my_app.cache, progress_cb=download_progress)
+            download_project(
+                api=api,
+                project_info=g.project_info,
+                project_dir=project_dir,
+                use_cache=True,
+            )
         except Exception as e:
             sly.logger.warn("Can not download project")
             raise Exception(