nuscenes-trainval: index and manifest

.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
+nuscenes-trainval/index.json filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,26 @@
+---
+license: cc-by-nc-sa-4.0
+pretty_name: confhead cache
+tags:
+- depth-estimation
+- nuscenes
+---
+
+# confhead cache
+
+Frozen Depth Anything V2 outputs on nuScenes, cached for the training of a
+per-pixel confidence head: decoder features, disparity, temporal spread, and
+the lidar and radar points of each camera image. Each folder is one cache,
+built by [confhead](https://github.com/kgqiao/Vision-Project-CSCI-566/tree/main/confhead).
+
+The cache derives from [nuScenes](https://www.nuscenes.org/), which is under
+CC BY-NC-SA 4.0 and the nuScenes terms of use. The same terms apply here.
+
+## Restore a cache
+
+    confhead cache download <cache> <folder>
+
+Or by hand: download `<cache>/` and extract each shard into one folder.
+
+    hf download nadayusc/confhead-cache --repo-type dataset --include "<cache>/*" --local-dir .
+    for f in <cache>/shards/*.tar.zst; do tar --zstd -xf "$f" -C <cache>; done

nuscenes-trainval/index.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5e3d1371f8a8a9ef718dccd8dac8b00f079d122277534b275d821fc60df75132
+size 39350814

nuscenes-trainval/manifest.json

@@ -4,7 +4,7 @@
  "fingerprint": "99d44ff8c71c5355",
  "scenes_per_shard": 25,
  "compression": "tar, zstd level 3",
- "complete": false,
+ "complete": true,
  "shards": [
   {
    "name": "00000",