Enable loading model from hub that has already been converted #36

Job	Run time
build (3.10, macos-15, test_modeling_gemma2.py)	2m 53s
build (3.10, macos-15, test_modeling_gemma.py)	2m 54s
build (3.10, macos-15, test_modeling_llama.py)	2m 52s
build (3.10, macos-15, test_modeling_olmo.py)	2m 52s
build (3.10, macos-15, test_modeling.py)	55s
build (3.10, macos-15, test_modeling_qwen2.py)	2m 51s
build (3.11, macos-15, test_modeling_gemma2.py)	2m 54s
build (3.11, macos-15, test_modeling_gemma.py)	2m 53s
build (3.11, macos-15, test_modeling_llama.py)	2m 53s
build (3.11, macos-15, test_modeling_olmo.py)	2m 52s
build (3.11, macos-15, test_modeling.py)	39s
build (3.11, macos-15, test_modeling_qwen2.py)	2m 53s
build (3.12, macos-15, test_modeling_gemma2.py)	2m 52s
build (3.12, macos-15, test_modeling_gemma.py)	2m 52s
build (3.12, macos-15, test_modeling_llama.py)	2m 52s
build (3.12, macos-15, test_modeling_olmo.py)	2m 51s
build (3.12, macos-15, test_modeling.py)	57s
build (3.12, macos-15, test_modeling_qwen2.py)	2m 51s
	45m 36s

Provide feedback