Kev-4B Core AI: record the bundle's Hub revision ee978c7 in the card; SHA256SUMS for the card

README.md

@@ -278,8 +278,8 @@ phys_footprint of at most 1.15 GB.
 | `head/kev_head.json` | head size, scale, temperature, delimiter ids, provenance | 2,038 | `5ff071ff…cfba7416` |
 
 The repository root carries the base model's `config.json`, the adapter's `adapter_config.json` and the
-Apache-2.0 `LICENSE`, verbatim. `SHA256SUMS` lists every file. The Hub revision is recorded here after the
-upload.
+Apache-2.0 `LICENSE`, verbatim. `SHA256SUMS` lists every file. Its Hub revision is
+[`ee978c7`](https://huggingface.co/mlboydaisuke/Kev-4B-CoreAI/tree/ee978c72ad20f98d8548b5ef5a28e7e7e1102389) (2026-10-04).
 
 No AOT asset ships: the Swift runtime's specialization of the `.aimodel` equals the AOT asset bit for bit
 (above). To compile one for the Mac anyway (15.55 GB, the flag is required):

SHA256SUMS

@@ -1,5 +1,5 @@
 50cbab8a892c5f2993b8c7351a99182507472def3b1374558308605d99b86b32  LICENSE
-49f9ef81dab003ae1275678400962136811b0de05d2132fb9c8527a22d89c1fb  README.md
+27cbc6bbab46975319243be3ddecf878958329e2163d52c80372136f1de0114b  README.md
 8a05dfd6c5e7f61a62e6db5d8de8093a8dbbd105aa7ee7b5e7b73f6249f45617  adapter_config.json
 ddc63e1c717afa86c865bb5e01313d89d72bb53b97ad4a8a03ba8510c0621670  config.json
 36c392f75fa5eff2243b07664e1edb1dc4e771dfc9d145469ba175db54157106  gpu-pipelined/kev_4b_decode_fp16_metal_pf128/head/head.safetensors