Link the published Medium write-up

README.md

@@ -32,6 +32,9 @@ are frozen, publicly available source models, and only two small projector
 heads (a few million parameters total) were trained from scratch on public
 datasets to align them into a shared space.
 
+📝 Full write-up (how it was built, and seven rounds of what did and didn't
+work): [I Built a Multimodal Embedding Model From Scratch on an RTX 4060](https://medium.com/@mithilmaske/i-built-a-multimodal-embedding-model-from-scratch-on-an-rtx-4060-text-image-audio-and-video-ab1fef04f1cd)
+
 ## Architecture
 
 Three frozen encoders, two trained projectors — everything large stays
@@ -81,7 +84,9 @@ raw training data volume was — R@1 only moved once the image-caption
 training set was scaled to ~172k pairs (round 7). The full round-by-round
 write-up, with what did and didn't move the needle, lives in the source
 repo's README and feasibility notes:
-[github.com/mithilai/QuadEmbed](https://github.com/mithilai/QuadEmbed).
+[github.com/mithilai/QuadEmbed](https://github.com/mithilai/QuadEmbed), and
+as a full narrative write-up on Medium:
+[I Built a Multimodal Embedding Model From Scratch on an RTX 4060](https://medium.com/@mithilmaske/i-built-a-multimodal-embedding-model-from-scratch-on-an-rtx-4060-text-image-audio-and-video-ab1fef04f1cd).
 
 ## Usage