XNNPACK: new-flow export (dynamic seq, get_model_schema, int8 embedding table)

Apply the model card standard

Apply the model card standard

Stop restating quantized, and drop the unused default

Correct the published metadata

Document why Core ML is not shipped

MLX: quantize embeddings (4w/64) and emit get_model_schema

Update README.md

Add LFM2.5-ColBERT-350M (XNNPACK 8da4w + MLX int4) for react-native-executorch

initial commit