card: the wide variant now ships, point at it and state what it does not buy

README.md

@@ -178,4 +178,4 @@ they are theirs to correct.
 
 ## What this build does not change
 
-It does not touch the int4 tensors, so the dominant quantization error in the model is unchanged. Measured on this model's own shapes, the int4 group-16 grid carries about 9 percent weight error against 0.6 percent for the new int8 grid, which is where the remaining 0.007 gap to the base lives. That gap is documented, not hidden: a wider variant exists only as a measurement, and this repo stays behavior compatible with the FP8 card.
+It does not touch the int4 tensors, so the dominant quantization error in the model is unchanged. Measured on this model's own shapes, the int4 group-16 grid carries about 9 percent weight error against 0.6 percent for the new int8 grid, which is where the remaining 0.007 gap to the base lives. That gap is documented, not hidden. A wider variant that converts those same int4 tensors to int8 now exists as [WeMM-Embedding-2B-INT8-Wide](https://huggingface.co/ewin-reg/WeMM-Embedding-2B-INT8-Wide), and it measured 0.993015 against the base on the text protocol, which is this build's own 0.992999 to within the noise. Buy it for one fewer dtype family, not for better embeddings. This repo stays behavior compatible with the FP8 card.