Expose static llama in OSS (#16184)

Summary:

This exposes a static llama model for CoreML.

We want to unify development behind one static model so that we do not need to apply fixes in multiple places (e.g., iOS 26 fixes).

Reviewed By: billmguo

Differential Revision: D88875745

Author: metascroy

.ci/scripts/test_ane_static_llama.sh

@@ -28,6 +28,11 @@ pushd $EXECUTORCH_ROOT/examples/apple/coreml/llama
 # Download stories llama110m artifacts
 download_stories_model_artifacts
 
+# Test static ANE llama model
+python export_static_llm_coreml.py --checkpoint stories110M.pt --params params.json --output model.pte
+python run_static_llm.py --model static_llm_coreml_model.pte --params $HOME/models/llama1b/params.json --tokenizer tokenizer.model --prompt "Once upon a time," --lookahead
+
+# Test export of deprecated model
 python export.py -n model.pte -p params.json -c stories110M.pt --seq_length 32 --max_seq_length 64 --dtype fp16 --coreml-quantize c4w --embedding-quantize 4,32
 
 popd