Refactor loading_from_pretrained local cache for Llama and Gemma models#1220

Closed

brendanlong wants to merge 3 commits intoTransformerLensOrg:devfrom

brendanlong:brendanlong/load-from-pretrained-cache

Commits on Mar 28, 2026

Add regression tests for convert_hf_model_config covering all architecture branches

claude
authored and
brendanlong
committed

Commits on Mar 29, 2026

Replace hardcoded Llama config dicts with LlamaConfig cache

claude
authored and
brendanlong
committed
Replace hardcoded Gemma config dicts with HF config cache entries

claude
authored and
brendanlong
committed