from llama_index.core import Settings
Settings.embed_model = "my embedding model"

I fixed it through above. Still, is there any other way. Please let me know