Update on "[Executorch] Add quantized kv cache to oss ci"

Fixes to make sure quantized kv cache works in oss

Differential Revision: [D66269487](https://our.internmc.facebook.com/intern/diff/D66269487/)

[ghstack-poisoned]

Author: kimishpatel

examples/models/llama/source_transformation/quantized_kv_cache.py

@@ -227,6 +227,7 @@ def replace_kv_cache_with_quantized_kv_cache(module):
         assert op is not None
     except:
         import glob
+
         import executorch
         from executorch.extension.pybindings import portable_lib  # noqa # usort: skip