Use Hyperism tar shards instead of expanded unzip tree

README.md

@@ -58,7 +58,10 @@ Install the Hugging Face CLI if needed:
 pip install -U "huggingface_hub[cli]"
 ```
 
-Download the dataset while preserving repository paths. The expanded ADT frame tree is not stored directly in the repository; download `aria_digital_twin/ADT_tars/*.tar` and restore it locally.
+Download the dataset while preserving repository paths. The expanded ADT and Hyperism frame trees are not stored directly in the repository; download the archive shards and restore them locally.
+
+- ADT archives: `aria_digital_twin/ADT_tars/*.tar`
+- Hyperism archives: `hyperism_required_shards/*.tar`
 
 ```bash
 DATA_ROOT=/mnt/nvme6/jieneng/data/spatialencoder_full
@@ -67,10 +70,11 @@ mkdir -p "$DATA_ROOT"
 hf download qicq1c/spatialencoder_full \
   --type dataset \
   --local-dir "$DATA_ROOT" \
-  --exclude "aria_digital_twin/ADT/**"
+  --exclude "aria_digital_twin/ADT/**" \
+  --exclude "hyperism/hyperism/unzip/**"
 ```
 
-The ADT tar archives are about 405 GiB. After extraction, `aria_digital_twin/ADT/` is about 406 GiB. Keep both the archives and expanded ADT tree if you want the download to be resumable and reproducible.
+The ADT tar archives are about 405 GiB. After extraction, `aria_digital_twin/ADT/` is about 406 GiB. The Hyperism tar shards are about 35.6 GiB. Keep the archives and expanded trees if you want the download to be resumable and reproducible.
 
 ## 2. Restore ADT From Tar Archives
 
@@ -92,35 +96,43 @@ aria_digital_twin/ADT/Apartment_release_clean_seq131_M1292/depth_frames/...
 aria_digital_twin/ADT/Apartment_release_clean_seq131_M1292/rgb_frames/...
 ```
 
-## 3. Expand Packed Add-Ons If Present
+## 3. Restore Hyperism From Tar Shards
 
-If the download contains archive files such as `pickle.zip`, `hyperism-train-json.zip`, or `hyperism-val-json.zip`, unzip them at the data root:
+Restore the Hyperism frame tree from the downloaded tar shards:
 
 ```bash
 cd "$DATA_ROOT"
+mkdir -p hyperism/hyperism/unzip
 
-for z in pickle.zip hyperism-train-json.zip hyperism-val-json.zip; do
-  if [ -f "$z" ]; then
-    unzip -o "$z" -d "$DATA_ROOT"
-  fi
+for shard in hyperism_required_shards/*.tar; do
+  tar --skip-old-files -xf "$shard" -C hyperism/hyperism/unzip
 done
 ```
 
-If the download contains `hyperism_required_shards/*.tar`, expand those shards into the Hyperism frame directory:
+This recreates paths such as:
+
+```text
+hyperism/hyperism/unzip/ai_001_001/...
+hyperism/hyperism/unzip/ai_001_002/...
+```
+
+## 4. Expand Packed Add-Ons If Present
+
+If the download contains archive files such as `pickle.zip`, `hyperism-train-json.zip`, or `hyperism-val-json.zip`, unzip them at the data root:
 
 ```bash
-mkdir -p "$DATA_ROOT/hyperism/hyperism/unzip"
+cd "$DATA_ROOT"
 
-if [ -d "$DATA_ROOT/hyperism_required_shards" ]; then
-  for shard in "$DATA_ROOT"/hyperism_required_shards/*.tar; do
-    tar -xf "$shard" -C "$DATA_ROOT/hyperism/hyperism/unzip"
-  done
-fi
+for z in pickle.zip hyperism-train-json.zip hyperism-val-json.zip; do
+  if [ -f "$z" ]; then
+    unzip -o "$z" -d "$DATA_ROOT"
+  fi
+done
 ```
 
 Skip this step for files that are already expanded in the final tree.
 
-## 4. Check Required Files
+## 5. Check Required Files
 
 Run these checks before training:
 
@@ -149,7 +161,7 @@ Expected minimum result:
 - `BoxFromMotion/dataset/{CA-1M,hyperism,ADT}.json` exist.
 - Hyperism and ADT frame paths referenced by the json files exist under `BOX_DATA_PATH`.
 
-## 5. Set Training Environment
+## 6. Set Training Environment
 
 From the SpatialEncoder code checkout:
 
@@ -179,7 +191,7 @@ wget -P "$BOX_WEIGHTS_PATH" \
   https://huggingface.co/facebook/sam3.1/resolve/main/sam3.1_multiplex.pt
 ```
 
-## 6. Dataset Smoke Tests
+## 7. Dataset Smoke Tests
 
 The merged training config samples datasets according to:
 
@@ -225,7 +237,7 @@ Train Epoch: [0][   0/...] ... Losses/train_all_loss: ...
 
 The first batch can be slow because workers are filling caches. Later steps should have near-zero `Data Time`.
 
-## 7. Mixed Training
+## 8. Mixed Training
 
 Once all three single-dataset smoke tests print loss, launch the mixed run: