Enable loading model from hub that has already been converted #19

Job	Run time
build (3.10, macos-15, test_modeling_gemma2.py)	4m 17s
build (3.10, macos-15, test_modeling_gemma.py)	4m 17s
build (3.10, macos-15, test_modeling_llama.py)	4m 19s
build (3.10, macos-15, test_modeling_olmo.py)	4m 19s
build (3.10, macos-15, test_modeling.py)	1m 12s
build (3.10, macos-15, test_modeling_qwen2.py)	3m 56s
build (3.11, macos-15, test_modeling_gemma2.py)	4m 16s
build (3.11, macos-15, test_modeling_gemma.py)	4m 17s
build (3.11, macos-15, test_modeling_llama.py)	4m 26s
build (3.11, macos-15, test_modeling_olmo.py)	4m 18s
build (3.11, macos-15, test_modeling.py)	49s
build (3.11, macos-15, test_modeling_qwen2.py)	4m 22s
build (3.12, macos-15, test_modeling_gemma2.py)	4m 14s
build (3.12, macos-15, test_modeling_gemma.py)	4m 15s
build (3.12, macos-15, test_modeling_llama.py)	4m 19s
build (3.12, macos-15, test_modeling_olmo.py)	4m 25s
build (3.12, macos-15, test_modeling.py)	1m 17s
build (3.12, macos-15, test_modeling_qwen2.py)	4m 22s
	1h 7m 40s

Provide feedback