diff --git a/src/lighteval/main_accelerate.py b/src/lighteval/main_accelerate.py
index 001585b87..2e160501a 100644
--- a/src/lighteval/main_accelerate.py
+++ b/src/lighteval/main_accelerate.py
@@ -32,6 +32,8 @@
 
 TOKEN = os.getenv("HF_TOKEN")
 CACHE_DIR: str = os.getenv("HF_HOME", "/scratch")
+print(f"CACHE_DIR: {CACHE_DIR}")
+print(f"ENV: {os.environ}")
 
 HELP_PANNEL_NAME_1 = "Common Paramaters"
 HELP_PANNEL_NAME_2 = "Logging Parameters"
@@ -63,7 +65,7 @@ def accelerate(  # noqa C901
     ] = None,
     cache_dir: Annotated[
         str, Option(help="Cache directory for datasets and models.", rich_help_panel=HELP_PANNEL_NAME_1)
-    ] = CACHE_DIR,
+    ] = "cache/model",
     num_fewshot_seeds: Annotated[
         int, Option(help="Number of seeds to use for few-shot evaluation.", rich_help_panel=HELP_PANNEL_NAME_1)
     ] = 1,
@@ -113,6 +115,7 @@ def accelerate(  # noqa C901
     accelerator = Accelerator(kwargs_handlers=[InitProcessGroupKwargs(timeout=timedelta(seconds=3000))])
 
     env_config = EnvConfig(token=TOKEN, cache_dir=cache_dir)
+    print(f"ENV CONFIG: {env_config}")
     evaluation_tracker = EvaluationTracker(
         output_dir=output_dir,
         save_details=save_details,
diff --git a/tests/test_main.py b/tests/test_main.py
index ad98d238e..732f945a7 100644
--- a/tests/test_main.py
+++ b/tests/test_main.py
@@ -59,7 +59,7 @@ def run_model_predictions_full(model: str, tasks: tuple):
         output_dir="",
         dataset_loading_processes=1,
         save_details=True,
-        cache_dir="cache/models",
+        # cache_dir="cache/models",
     )
     return results
 
@@ -75,7 +75,7 @@ def run_model_predictions_lite(model: str, tasks: tuple):
         dataset_loading_processes=1,
         save_details=True,
         max_samples=10,
-        cache_dir="cache/models",
+        # cache_dir="cache/models",
     )
     return results