Merge branch 'main' of https://huggingface.co/datasets/1x-technologies/worldmodel

README.md

@@ -12,7 +12,51 @@ Download with:
 huggingface-cli download 1x-technologies/worldmodel --repo-type dataset --local-dir data
 ```
 
-Current version: v1.1
+Changes from v1.1:
+- New train and val dataset of 100 hours, replacing the v1.1 datasets
+- Blur applied to faces
+
+Contents of train/val_v2.0:
+
+The training dataset is shareded into 100 independent shards. The definitions are as follows:
+
+- **video_{shard}.bin**: 8x8x8 image patches at 30hz, with 17 frame temporal window, encoded using [NVIDIA Cosmos Tokenizer](https://github.com/NVIDIA/Cosmos-Tokenizer) "Cosmos-Tokenizer-DV8x8x8".
+- **segment_idx_{shard}.bin** - Maps each frame `i` to its corresponding segment index. You may want to use this to separate non-contiguous frames from different videos (transitions).
+- **states_{shard}.bin** - States arrays (defined below in `Index-to-State Mapping`) stored in `np.float32` format. For frame `i`, the corresponding state is represented by `states_{shard}[i]`.
+- **metadata** - The `metadata.json` file provides high-level information about the entire dataset, while `metadata_{shard}.json` files contain specific details for each shard. 
+
+  #### Index-to-State Mapping (NEW)
+  ```
+   {
+        0: HIP_YAW
+        1: HIP_ROLL
+        2: HIP_PITCH
+        3: KNEE_PITCH
+        4: ANKLE_ROLL
+        5: ANKLE_PITCH
+        6: LEFT_SHOULDER_PITCH
+        7: LEFT_SHOULDER_ROLL
+        8: LEFT_SHOULDER_YAW
+        9: LEFT_ELBOW_PITCH
+        10: LEFT_ELBOW_YAW
+        11: LEFT_WRIST_PITCH
+        12: LEFT_WRIST_ROLL
+        13: RIGHT_SHOULDER_PITCH
+        14: RIGHT_SHOULDER_ROLL
+        15: RIGHT_SHOULDER_YAW
+        16: RIGHT_ELBOW_PITCH
+        17: RIGHT_ELBOW_YAW
+        18: RIGHT_WRIST_PITCH
+        19: RIGHT_WRIST_ROLL
+        20: NECK_PITCH
+        21: Left hand closure state (0 = open, 1 = closed)
+        22: Right hand closure state (0 = open, 1 = closed)
+        23: Linear Velocity
+        24: Angular Velocity
+    }
+
+
+Previous version: v1.1
 
 - **magvit2.ckpt** - weights for [MAGVIT2](https://github.com/TencentARC/Open-MAGVIT2) image tokenizer we used. We provide the encoder (tokenizer) and decoder (de-tokenizer) weights.
 
@@ -25,7 +69,7 @@ Contents of train/val_v1.1:
   - **neck_desired** `(N, 1)`: Desired neck pitch.
   - **l_hand_closure** `(N, 1)`: Left hand closure state (0 = open, 1 = closed).
   - **r_hand_closure** `(N, 1)`: Right hand closure state (0 = open, 1 = closed).
-  #### Index-to-Joint Mapping
+  #### Index-to-Joint Mapping (OLD)
   ```
    {
         0: HIP_YAW

unpack_data.py

@@ -0,0 +1,32 @@
+"""Example script to unpack one shard of the 1xGPT v2.0 video dataset."""
+
+import json
+import pathlib
+import subprocess
+
+import numpy as np
+
+dir_path = pathlib.Path("val_v2.0")
+rank = 0
+
+# load metadata.json
+metadata = json.load(open(dir_path / "metadata.json"))
+metadata_shard = json.load(open(dir_path / f"metadata_{rank}.json"))
+
+total_frames = metadata_shard["shard_num_frames"]
+
+
+maps = [
+    ("segment_idx", np.int32, []),
+    ("states", np.float32, [25]),
+]
+
+video_path = dir_path / "video_0.mp4"
+
+for m, dtype, shape in maps:
+    filename = dir_path / f"{m}_{rank}.bin"
+    print("Reading", filename, [total_frames] + shape)
+    m_out = np.memmap(filename, dtype=dtype, mode="r", shape=tuple([total_frames] + shape))
+    assert m_out.shape[0] == total_frames
+    print(m, m_out[:100])
+