Upload README.md with huggingface_hub

README.md

@@ -25,9 +25,13 @@ decoder once per click.
 
 The encoder ships in fp32 only, and that is not an omission. Its backbone is RepViT,
 which is convolutional, and XNNPACK serializes convolution weights as fp32 whatever
-dtype the graph carries — fp16 came out at 19.8 MB (100.5%) and dynamic int8 at
-19.7 MB, so neither buys anything. At 19.7 MB the fp32 encoder is already smaller
-than SAM 2.1 hiera-tiny's *fp16* encoder (55.6 MB).
+dtype the graph carries, so fp16 came out at 19.8 MB (100.5%) and buys nothing.
+Dynamic int8 is a different matter: it annotates `aten.linear` and this graph has
+none — all 50 nn.Linear in EdgeTAM sit in the mask decoder, not on the
+image-embedding path — so it leaves the encoder at the same 19.7 MB. Static int8 is
+the recipe that would reach convolution weights, and it has not been run here. At
+19.7 MB the fp32 encoder is already smaller than SAM 2.1 hiera-tiny's *fp16* encoder
+(55.6 MB).
 
 EdgeTAM is Meta's on-device SAM 2 (CVPR 2025). Its encoder is **5.5× smaller than
 SAM 2.1 hiera-tiny's** (19.7 MB vs 109.2 MB) for the same output contract, so an
@@ -80,13 +84,3 @@ the decoder are dropped, since their lowered form mis-sizes on a single-point ex
 The GPU-specific rewrites in the LiteRT build of this model (splitting the
 squeeze-excite mean, replacing ConvTranspose2d) are ML Drift workarounds and are not
 needed here — XNNPACK runs the stock graph.
-
-<!-- funnel:v1 -->
-
----
-
-**More models in this format:** [ExecuTorch Model Zoo](https://huggingface.co/collections/mlboydaisuke/executorch-model-zoo-6a7ff328390b63075ffeae5e) — 31 models, each with the recipe that produced it.
-
-**Want a different model on-device?** [Open a request](https://github.com/john-rocky/on-device-requests) — free, open weights only; the export and its measured numbers get published publicly.
-
-<!-- /funnel:v1 -->