use a smaller model to fix API test run on CPU node

Author: papa99do

tests/api_tests/v1/tests/application_tests/test_env_var_changes.py

@@ -79,7 +79,7 @@ def test_inference_cache(self):
         """
 
         # Restart marqo with new max values
-        new_models = ["open_clip/ViT-H-14/laion2b_s32b_b79k"]
+        new_models = ["open_clip/ViT-B-32/laion2B-s34B-b79K"]
         index_name = "test_multiple_env_vars"
         utilities.rerun_marqo_with_env_vars(
             env_vars=[