Add files using upload-large-folder tool

.gitattributes

@@ -57,3 +57,6 @@ 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
+meta/tasks.jsonl filter=lfs diff=lfs merge=lfs -text
+meta/sharded_index.json filter=lfs diff=lfs merge=lfs -text
+meta/episodes.jsonl filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,134 @@
+Language Table (LeRobot) — Embedding-Only Release
+(DINOv3 + SigLIP2 image features; EmbeddingGemma task-text features)
+
+This repository packages a re-encoded variant of IPEC-COMMUNITY/language_table_lerobot where raw videos are replaced by fixed-length image embeddings, and task strings are augmented with text embeddings. All indices, splits, and semantics remain consistent with the source dataset while storage and I/O are substantially lighter. To make the dataset practical to upload/download and stream from the Hub, we also consolidated ~0.5M tiny per-episode Parquet files into N large Parquet shards under a single data/ folder. The file meta/sharded_index.json preserves a precise mapping from each original episode (referenced by a normalized identifier of the form data/chunk-{episode_chunk:03d}/episode_{episode_index:06d}.parquet) to its shard path and row range, so you keep original addressing without paying the small-file tax.
+
+- Robot: xArm
+- Modalities kept: states, actions, timestamps, frame/episode indices, image embeddings, task-text embeddings
+- Removed: raw video tensors (column observation.images.rgb)
+- License: apache-2.0 (inherits from source)
+
+----------------------------------------------------------------
+
+Quick Stats
+
+From meta/info.json and meta/task_text_embeddings_info.json:
+
+- Episodes: 442,226
+- Frames: 7,045,476
+- Tasks (unique): 127,605
+- Chunks (original layout): 443 (chunks_size=1000)
+- Shards (this release): N Parquet files under data/ (see meta/sharded_index.json)
+- FPS: 10
+- Image embeddings (per frame):
+  - observation.images.rgb_dinov3 → float32 [1024] (DINOv3 ViT-L/16 CLS)
+  - observation.images.rgb_siglip2 → float32 [768] (SigLIP2-base)
+- Task-text embeddings (per unique task):
+  - embedding → float32 [768] from google/embeddinggemma-300m
+  - Count: 127,605 rows (one per task)
+
+Note: This is an embedding-only package. video_path is omitted and the original observation.images.rgb pixels are dropped.
+
+----------------------------------------------------------------
+
+Contents
+
+.
+├─ meta/
+│  ├─ info.json                       # dataset overview & feature schema (updated to shards)
+│  ├─ sharded_index.json              # mapping: original-episode-id → shard path + row range
+│  ├─ tasks.jsonl                     # {"task_index": int, "task": str}
+│  ├─ episodes.jsonl                     # {"task_index": int, "task": str, "length": int}
+│  └─ task_text_embeddings_info.json  # model/dim/normalized/count/file for task embeddings
+├─ data/
+│  ├─ shard-00000-of-000NN.parquet
+│  ├─ shard-00001-of-000NN.parquet
+│  ├─ …                               # N large Parquet shards for fast HF upload/streaming
+│  └─ task_text_embeddings.parquet    # task_index, task, 768-D EmbeddingGemma vector
+└─ README.md
+
+----------------------------------------------------------------
+
+How This Was Generated (Reproducible Pipeline)
+
+1) Episode → Image Embeddings (drop pixels)
+convert_lerobot_to_embeddings_mono.py (GPU-accelerated preprocessing).
+Adds:
+- observation.images.rgb_dinov3 (float32[1024])
+- observation.images.rgb_siglip2 (float32[768])
+Removes:
+- observation.images.rgb (raw frames)
+
+2) Task-Text Embeddings (one row per unique task)
+build_task_text_embeddings.py with SentenceTransformer("google/embeddinggemma-300m") → data/task_text_embeddings.parquet + meta/task_text_embeddings_info.json.
+
+3) Data Consolidation (this release)
+All per-episode Parquets were consolidated into N large Parquet shards in one data/ folder.
+- The index meta/sharded_index.json records, for each episode, its normalized source identifier data/chunk-{episode_chunk:03d}/episode_{episode_index:06d}.parquet, the destination shard path, and the (row_offset, num_rows) range inside that shard.
+- This preserves original addressing while making Hub sync/clone/stream far faster and more reliable.
+
+----------------------------------------------------------------
+
+Metadata (Excerpts)
+
+meta/task_text_embeddings_info.json
+{
+  "model": "google/embeddinggemma-300m",
+  "dimension": 768,
+  "normalized": false,
+  "count": 127605,
+  "file": "task_text_embeddings.parquet"
+}
+
+meta/info.json (embedding-only + shards)
+{
+  "codebase_version": "v2.0-embeddings-sharded",
+  "robot_type": "xarm",
+  "total_episodes": 442226,
+  "total_frames": 7045476,
+  "total_tasks": 127605,
+  "total_videos": 442226,
+  "total_chunks": 443,
+  "chunks_size": 1000,
+  "fps": 10,
+  "splits": { "train": "0:442226" },
+  "data_path": "data/shard-{shard_id:05d}-of-{num_shards:05d}.parquet",
+  "index_path": "meta/sharded_index.json",
+  "num_shards": "NN",
+  "features": {
+    "observation.state": { "dtype": "float32", "shape": [8],
+      "names": { "motors": ["x","y","z","roll","pitch","yaw","pad","gripper"] } },
+    "action": { "dtype": "float32", "shape": [7],
+      "names": { "motors": ["x","y","z","roll","pitch","yaw","gripper"] } },
+    "timestamp": { "dtype": "float32", "shape": [1], "names": null },
+    "frame_index": { "dtype": "int64", "shape": [1], "names": null },
+    "episode_index": { "dtype": "int64", "shape": [1], "names": null },
+    "index": { "dtype": "int64", "shape": [1], "names": null },
+    "task_index": { "dtype": "int64", "shape": [1], "names": null },
+    "observation.images.rgb_dinov3": { "dtype": "float32", "shape": [1024], "names": null },
+    "observation.images.rgb_siglip2": { "dtype": "float32", "shape": [768], "names": null }
+  }
+}
+
+----------------------------------------------------------------
+
+Environment & Dependencies
+
+Python ≥ 3.9 • PyTorch ≥ 2.1 • transformers • sentence-transformers • pyarrow • tqdm • decord (and optionally av)
+
+----------------------------------------------------------------
+
+Provenance, License, and Citation
+
+- Source dataset: IPEC-COMMUNITY/language_table_lerobot
+- License: apache-2.0 (inherits from the source)
+- Encoders to cite:
+  - facebook/dinov3-vitl16-pretrain-lvd1689m
+  - google/siglip2-base-patch16-384
+  - google/embeddinggemma-300m
+
+----------------------------------------------------------------
+
+Changelog
+
+- v2.0-embeddings-sharded — Replaced video tensors with DINOv3 + SigLIP2 features; added EmbeddingGemma task-text embeddings; consolidated per-episode Parquets into N shards with a repo-local index; preserved original indexing/splits via normalized episode identifiers.

data/shard-00000-of-00064.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ccf8b9ec861c95e985607cd7ee5b18bb219003a1ecae00119644de2fdc259e0f
+size 580852220

data/shard-00002-of-00064.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e28d580a26e7ef867d5be776c5acf03f29124377f0926e9113a4c7016c1023ca
+size 592295490

data/shard-00003-of-00064.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:97ac76676022b776ffc3f9d0d1f71cde9a25a70bb91286e394ffae60a705f5bd
+size 592767349

data/shard-00004-of-00064.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1f4b0d746818458491ca7bf552cdc6a50bd4e9709a2a7c0d1e7f1f722b8e274c
+size 576418975

data/shard-00005-of-00064.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a050d1c98c88a6c32ffb89f4370d92771ad29031bb2c0ecfab78f285f5283271
+size 581436552

data/shard-00006-of-00064.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a343e5dc38ac55d8ea6c9b004523a6a02806c3f9363cb90f35b3e258738d0cd3
+size 578627566

data/shard-00007-of-00064.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:54e7202447de6a7e0598577b18e61283b066524ac56d1ab9133b859f85eb34c4
+size 583892168

data/shard-00009-of-00064.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c5574183ff80cb7c24bf7f00d5918dd981315e450ed42aea95488a7e2ba6c904
+size 580744077

data/shard-00011-of-00064.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4691c98d1f8a66372d39c89a1aaeedc8b5611cc36cf437051ce74bcc697f37e7
+size 585345800

data/shard-00014-of-00064.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2edb83742907d66c6dac65e1bd3e8d1cfa0f55d813434c588ffe99c70cd4b0d8
+size 582627118

data/shard-00015-of-00064.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c3a0ed457038c0e8539408a672d5a4cc07059d6d2cdf5aaa2aa602fa9a0a9067
+size 586080680

data/shard-00016-of-00064.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4a5c96ef90a6de33570b11f40375020a1a01cef0d89a173e108ff94239a06426
+size 597316914

data/shard-00018-of-00064.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3047b2de3da780c9d5c4c90f2f1d6aff9110df8d54f324d50ceedd91ec9f8f38
+size 588221309

data/shard-00019-of-00064.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:55623aee0ce5e884314dda0d765ef484e6b920871f3b4dd3c42dd1be1b5c25bc
+size 586150399

data/shard-00020-of-00064.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:76025b656cca058411e73e7ddaedc3714033fba97196348bc96e278ee1ac4dc6
+size 569948180

data/shard-00021-of-00064.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c01899f745acc8cb6b9bb1c8713aff7c6fd866aeb9e613864af85e527a482fa6
+size 580789235

data/shard-00022-of-00064.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d5d8cce813ad3c41a67131d4db34198397cf21e974962ba052b06a19c844d592
+size 578191137

data/shard-00023-of-00064.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f2bb90aa459bb8fa54e8b79235533082381cea6854b97b72703c9c47834f5b94
+size 579833479

data/shard-00024-of-00064.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7090d3d925d15f6319048da284753ee235956535892bc5526d7adb4daf8e9bb0
+size 581710958

data/shard-00026-of-00064.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fe866dd1500cbf3d8b09ea0396b4bbec33b09e48e75d2b8a853fed52acc29411
+size 587988022

data/shard-00027-of-00064.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:80446c61aae7097954a91ba0ad2cf3238ec26e288c90bb604732be8fab7d480a
+size 585253202

data/shard-00029-of-00064.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b449b018b9069011f2b86314662c207badc1986edcf83d856ba38881a38bb552
+size 578967954

data/shard-00030-of-00064.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:31f732e072d08411f047f74ff310ac23378a2fd81325da37f1e64fc6d0080ab0
+size 580630730

data/shard-00032-of-00064.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fce71b8067c9b8131f5b8bbd76599fa6313f2cd6a40bde86decffb646b9b1b22
+size 581070876

data/shard-00033-of-00064.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c66531f222a7405ee5f919147b0be19f5fe510eec2a0e3f0ac93291a8dfaf128
+size 585217278

data/shard-00036-of-00064.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:67717a13dd14fa76392b57fc00695eed991ad0ba776be9da395d4c2b1bd26cf6
+size 586601101

data/shard-00038-of-00064.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9648728bc3e44d90c236e34dd9cfe651d787640bc96ce7a61facc537598d4d47
+size 581174605

data/shard-00040-of-00064.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c42c80c21fe32100dcc3e61e06e255a681c1dbf3c7f653ff123bc66ec51097a8
+size 585700214

data/shard-00041-of-00064.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:66249c4e8fb9b5781a9d38c8050a712cb113ec4be347e6ca3e877a69148bfa7b
+size 584247528

data/shard-00042-of-00064.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:04a3c7805be65c7506d13e2600dd70b7476eef33fb86010d8a06c0f2ddb93114
+size 586361593

data/shard-00043-of-00064.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d645aa8b208658740e7529a195a5faf3b1b2b63c84e97ba81be329ed655b9308
+size 582058452

data/shard-00044-of-00064.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6520aca0d800644dd16ee144378425964eda3c486fe235766348f1f17f09bc2c
+size 590702048

data/shard-00045-of-00064.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c177831993da3e0d55c61acaae28ef3bc2fc03562a6eaf83c8c778555c224a91
+size 587549762

data/shard-00047-of-00064.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:76ffd3559829dae7c088c46cd70c22ed31f332136a7ca21a9b8290da39bc5b60
+size 585014424

data/shard-00048-of-00064.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8e42c041597f52784490c4bd8a2a1ec47c43859ff22c83ff08f77c64ababb205
+size 581356845

data/shard-00049-of-00064.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4bd133fc82831b5e1aebaa14ac763aff4abaec0b9b545c206db4026a714380b6
+size 588388922

data/shard-00050-of-00064.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d038126d8bf6e86138ae95ff15f8560d6e91379283bf5c1d2197c71e0324cc5f
+size 583872160

data/shard-00054-of-00064.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:64c68eeb21fd0091d2621e00c4166286fa3ed6be7887d67bab2afa73b3d862f6
+size 579455336

data/shard-00055-of-00064.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e1b9dd909f279369525f1175cb45cca74e8158efb272b2bfd3b28639904dc7c5
+size 584316769

data/shard-00058-of-00064.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a248234fe241cbf21137609aeeef54158af769fe4aa50884b1b58401ab4f732c
+size 580714856

data/shard-00059-of-00064.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4cbf160083516562f23e897c8e1bf83d7909e1c735d60d2fe10a9e1f969416d1
+size 582300781

data/shard-00061-of-00064.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:de608b2eed1eaa27c208ee383fecc7babd81a014e65a515232d60f6f5631ba98
+size 586658758

data/shard-00063-of-00064.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4b98cfdf28d781f1c3ba5b8d8fe24a6d1d77ffe8e2c6da610f1ff5fe7474b1ff
+size 576188672

data/task_text_embeddings.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9ee2fb531e60b20611130627a7c931cd3bdd32905d15789aed1cd27147f5bca5
+size 365401699

meta/episodes.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ab183d7292a2e6daa63edf30e9154c5bdc261ca132c364b33d5d6f71a3580ee1
+size 43671939

meta/info.json

@@ -0,0 +1,103 @@
+{
+  "codebase_version": "v2.0-embeddings-sharded",
+  "robot_type": "xarm",
+  "total_episodes": 442226,
+  "total_frames": 7045476,
+  "total_tasks": 127605,
+  "total_videos": 442226,
+  "total_chunks": 443,
+  "chunks_size": 1000,
+  "fps": 10,
+  "splits": {
+    "train": "0:442226"
+  },
+  "data_path": "data/shard-{shard_id:05d}-of-{num_shards:05d}.parquet",
+  "features": {
+    "observation.state": {
+      "dtype": "float32",
+      "shape": [
+        8
+      ],
+      "names": {
+        "motors": [
+          "x",
+          "y",
+          "z",
+          "roll",
+          "pitch",
+          "yaw",
+          "pad",
+          "gripper"
+        ]
+      }
+    },
+    "action": {
+      "dtype": "float32",
+      "shape": [
+        7
+      ],
+      "names": {
+        "motors": [
+          "x",
+          "y",
+          "z",
+          "roll",
+          "pitch",
+          "yaw",
+          "gripper"
+        ]
+      }
+    },
+    "timestamp": {
+      "dtype": "float32",
+      "shape": [
+        1
+      ],
+      "names": null
+    },
+    "frame_index": {
+      "dtype": "int64",
+      "shape": [
+        1
+      ],
+      "names": null
+    },
+    "episode_index": {
+      "dtype": "int64",
+      "shape": [
+        1
+      ],
+      "names": null
+    },
+    "index": {
+      "dtype": "int64",
+      "shape": [
+        1
+      ],
+      "names": null
+    },
+    "task_index": {
+      "dtype": "int64",
+      "shape": [
+        1
+      ],
+      "names": null
+    },
+    "observation.images.rgb_dinov3": {
+      "dtype": "float32",
+      "shape": [
+        1024
+      ],
+      "names": null
+    },
+    "observation.images.rgb_siglip2": {
+      "dtype": "float32",
+      "shape": [
+        768
+      ],
+      "names": null
+    }
+  },
+  "num_shards": 64,
+  "index_path": "meta/sharded_index.json"
+}

meta/sharded_index.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d2d21f3f1c84332d211e2de3c8134dd8f4d114ecfb47dea71d37a63f6d435f75
+size 102754075

meta/task_text_embeddings_info.json

@@ -0,0 +1,7 @@
+{
+  "model": "google/embeddinggemma-300m",
+  "dimension": 768,
+  "normalized": false,
+  "count": 127605,
+  "file": "task_text_embeddings.parquet"
+}