Add training instruction pairs

.gitattributes

@@ -58,3 +58,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 # Video files - compressed
 *.mp4 filter=lfs diff=lfs merge=lfs -text
 *.webm filter=lfs diff=lfs merge=lfs -text
+egoexo4d-exoviews-with-egocaptions.json filter=lfs diff=lfs merge=lfs -text

README.md

@@ -6,7 +6,10 @@ tags:
 
 Exocentric training videos for **Ego2ExoVLM**, from [From My View to Yours: Learning Egocentric Cues from Exocentric Video using Privileged Egocentric Supervision](https://arxiv.org/abs/2501.05711) (ECCV 2026).
 
-`exocentric_videos.zip` contains 9,392 exocentric keystep segments from the [EgoExo4D](https://ego-exo4d-data.org/) keystep recognition training set (one exocentric view per segment). Please see the [GitHub repository](https://github.com/dominickrei/EgoExo4ADL) for training instructions.
+* `exocentric_videos.zip`: 9,392 exocentric keystep segments from the [EgoExo4D](https://ego-exo4d-data.org/) keystep recognition training set (one exocentric view per segment)
+* `egoexo4d-exoviews-with-egocaptions.json`: 28,176 instruction pairs (three per video) containing the responses of the egocentric teacher (VideoLLaMA3-7B), paired with the exocentric videos
+
+Please see the [GitHub repository](https://github.com/dominickrei/EgoExo4ADL) for training instructions.
 
 ## Citation
 ```bibtex

egoexo4d-exoviews-with-egocaptions.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1d426837d0c1432237c1f00d0e536b509075884b12f58428423a70eb7c5498cb
+size 18181135