Document verified Real-World One-Shot training release

README.md

@@ -23,7 +23,7 @@ Expert demonstrations for **robotic value models**, covering real-world and simu
 
 The common release format is **LeRobot Dataset v3.0**: per-step Parquet data, camera videos, and task/episode metadata. Robot platforms, cameras, and state/action schemas are specified separately for each subset.
 
-**Available now: Real-World Full-Shot — 2,000 episodes across 20 tasks; Simulation Full-Shot — 1,500 episodes across 15 tasks.** Simulation One-Shot is available: 15 episodes / 8,080 rows. Real-World One-Shot remains reserved.
+**Available now: Real-World Full-Shot — 2,000 episodes across 20 tasks; Simulation Full-Shot — 1,500 episodes across 15 tasks.** Both One-Shot subsets are available: Real-World — 20 episodes / 15,667 rows; Simulation — 15 episodes / 8,080 rows.
 
 [Real-world data](#2-real-world) · [Simulation](#3-simulation) · [Quick start](#5-download-and-reading) · [Repository files](#4-repository-layout)
 
@@ -31,10 +31,10 @@ The common release format is **LeRobot Dataset v3.0**: per-step Parquet data, ca
 
 | Domain | Tasks | Full-Shot | One-Shot |
 | --- | ---: | ---: | ---: |
-| [Real-World](#2-real-world) | 20 | **2,000** | 20<sup>†</sup> |
+| [Real-World](#2-real-world) | 20 | **2,000** | **20** |
 | [Simulation](#3-simulation) | 15 | **1,500** | **15** |
 
-Counts are **episodes**, not video files or individual camera views. Full-Shot provides **100 episodes per task**; One-Shot selects **one episode per task from the same Full-Shot pool**, rather than adding independently collected demonstrations. † denotes the defined size of a reserved subset.
+Counts are **episodes**, not video files or individual camera views. Full-Shot provides **100 episodes per task**; One-Shot selects **one episode per task from the same Full-Shot pool**, rather than adding independently collected demonstrations. All four subsets are now available.
 
 **Which setting should I use?**
 
@@ -52,7 +52,7 @@ Real robot demonstrations collected with **G2A**, covering **20 manipulation tas
 | Setting | Per task | Episodes | Data |
 | --- | ---: | ---: | --- |
 | [**Full-Shot**](#22-full-shot) | 100 | **2,000** | [Files](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/tree/main/data/real-world/full-shot) |
-| [One-Shot](#23-one-shot) | 1 | 20 | [Reserved](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/tree/main/data/real-world/one-shot) |
+| [**One-Shot**](#23-one-shot) | 1 | **20** | [Files](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/tree/main/data/real-world/one-shot) |
 
 <details>
 <summary><strong>Task catalog: all 20 tasks and short descriptions</strong></summary>
@@ -161,9 +161,9 @@ This is a key-field summary, not the entire vector. Additional pose, mode, opera
 
 ### 2.3 One-Shot
 
-**Use for one-reference conditioning or task-specific adaptation.** The setting selects one episode for each of the same **20 real-world tasks**, for **20 episodes** in total. These are selected from Full-Shot, not an additional task collection.
+**20 tasks × 1 demonstration = 20 episodes, with 15,667 synchronized rows.** The supplied One-Shot selection is available in the [directory](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/tree/main/data/real-world/one-shot). Original episode 000 is retained for each source task, with no trimming or padding. All 60 original HEVC (H.265) videos are preserved byte-for-byte. The release uses LeRobot v3.0 paths, 30 FPS, `float32[290]` state and `float32[52]` action. Image and numeric statistics are recomputed for this subset. The [task–episode–row mapping](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/data/real-world/one-shot/provenance/episode_mapping.json) preserves source task identifiers. Camera aliases match the Full-Shot naming: `hand_left → cam_left_wrist`, `hand_right → cam_right_wrist`, and `top_head → cam_high`.
 
-The [One-Shot directory](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/tree/main/data/real-world/one-shot) is reserved; its trajectory files and episode selection are not yet released. Until then, the [Full-Shot quick start](#5-download-and-reading) is the available entry point. Choosing your own single episode does not reproduce the official One-Shot selection.
+Use `prefix = "data/real-world/one-shot"` in the download example; expect 20 episodes and 15,667 rows.
 
 ## 3. Simulation
 
@@ -231,13 +231,13 @@ subset_catalog.json               Machine-readable subset index
 data/
 ├── real-world/
 │   ├── full-shot/                Available LeRobot dataset root
-│   └── one-shot/                 Reserved
+│   └── one-shot/                 Available LeRobot dataset root
 └── simulation/
     ├── full-shot/                Available LeRobot dataset root
     └── one-shot/                 Available LeRobot dataset root
 ```
 
-**A released subset directory is the LeRobot `root`; the repository root is not.** The reserved directories currently contain only a README.
+**A released subset directory is the LeRobot `root`; the repository root is not.** All four subset directories contain released training data.
 
 <details>
 <summary><strong>Format reference: files inside either released Full-Shot subset</strong></summary>
@@ -298,6 +298,6 @@ print(sample["action"].shape, sample["observation.state"].shape)
 
 Expected: **2,000 episodes**, **1,530,987 time steps**, `action` shape **[52]**, and `observation.state` shape **[290]**. `dataset[0]` reads one time step, not an entire episode. LeRobot resolves the corresponding camera frames from video metadata.
 
-The example pins one repository revision for both download and reading. To load simulation instead, set `prefix = "data/simulation/full-shot"`; expect 1,500 episodes, 789,612 rows, and `float32[14]` actions and states. For Simulation One-Shot, use `prefix = "data/simulation/one-shot"` and expect 15 episodes / 8,080 rows. Real-World One-Shot remains reserved.
+The example pins one repository revision for both download and reading. To load simulation instead, set `prefix = "data/simulation/full-shot"`; expect 1,500 episodes, 789,612 rows, and `float32[14]` actions and states. For Simulation One-Shot, use `prefix = "data/simulation/one-shot"` and expect 15 episodes / 8,080 rows. Real-World One-Shot uses `prefix = "data/real-world/one-shot"` with 20 episodes / 15,667 rows.
 
 **License:** real-world and supplemental simulation licensing remains unspecified. The RoboDojo-derived portion follows its upstream terms (dataset card: Apache-2.0). No blanket license is assigned to the combined repository.

README_zh.md

@@ -6,7 +6,7 @@
 
 各子集统一以 **LeRobot Dataset v3.0** 作为发布格式：逐时间步 Parquet 数据、相机视频，以及任务和轨迹元数据。不同子集的机器人、相机、状态与动作字段以各自说明为准，不要求跨平台完全相同。
 
-**当前可用：真机 Full-Shot，20 个任务、2,000 条轨迹；仿真 Full-Shot，15 个任务、1,500 条轨迹。** 仿真 One-Shot 已发布，15 条轨迹、8,080 行；真机 One-Shot 仍为预留目录。
+**当前可用：真机 Full-Shot，20 个任务、2,000 条轨迹；仿真 Full-Shot，15 个任务、1,500 条轨迹。** 两类 One-Shot 均已发布：真机 20 条轨迹、15,667 行；仿真 15 条轨迹、8,080 行。
 
 [真实世界](#2-真实世界) · [仿真世界](#3-仿真世界) · [快速开始](#5-下载与读取) · [仓库目录](#4-仓库目录)
 
@@ -14,10 +14,10 @@
 
 | 领域 | 任务数 | Full-Shot | One-Shot |
 | --- | ---: | ---: | ---: |
-| [真实世界](#2-真实世界) | 20 | **2,000** | 20<sup>†</sup> |
+| [真实世界](#2-真实世界) | 20 | **2,000** | **20** |
 | [仿真世界](#3-仿真世界) | 15 | **1,500** | **15** |
 
-表中数量均为**轨迹条数**，不是视频文件数或相机视角数。Full-Shot **每任务 100 条**；One-Shot **从对应 Full-Shot 集合中每任务选择 1 条**，不是额外采集的数据。† 表示预留子集的定义规模。
+表中数量均为**轨迹条数**，不是视频文件数或相机视角数。Full-Shot **每任务 100 条**；One-Shot **从对应 Full-Shot 集合中每任务选择 1 条**，不是额外采集的数据。四个子集现均已发布。
 
 **什么时候使用哪一种设置？**
 
@@ -35,7 +35,7 @@
 | 设置 | 每任务 | 总轨迹数 | 数据 |
 | --- | ---: | ---: | --- |
 | [**Full-Shot**](#22-full-shot) | 100 | **2,000** | [文件](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/tree/main/data/real-world/full-shot) |
-| [One-Shot](#23-one-shot) | 1 | 20 | [预留](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/tree/main/data/real-world/one-shot) |
+| [**One-Shot**](#23-one-shot) | 1 | **20** | [文件](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/tree/main/data/real-world/one-shot) |
 
 <details>
 <summary><strong>任务清单：20 个任务及简要说明</strong></summary>
@@ -144,9 +144,9 @@
 
 ### 2.3 One-Shot
 
-**用于单条示例参考或任务适配。** 从相同的 **20 个真机任务**中，每任务选择一条 Full-Shot 轨迹，共 **20 条**，不是另一套任务或额外采集的数据。
+**20 个任务 × 1 条 = 20 条轨迹，共 15,667 行同步数据。** 提供的 One-Shot 选择已在[目录](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/tree/main/data/real-world/one-shot)发布。保留每个源任务的原始 episode 000，不裁剪、不补帧；60 个原始 HEVC（H.265）视频保持字节不变。采用 LeRobot v3.0 路径、30 FPS，状态为 `float32[290]`，动作为 `float32[52]`。图像与数值统计量均按本子集重新计算。[任务—episode—行范围映射](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/data/real-world/one-shot/provenance/episode_mapping.json)保留源任务标识。相机别名与 Full-Shot 一致：`hand_left → cam_left_wrist`、`hand_right → cam_right_wrist`、`top_head → cam_high`。
 
-[One-Shot 目录](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/tree/main/data/real-world/one-shot)目前为预留位置，轨迹文件与正式示例选择尚未发布。当前可用入口为 [Full-Shot 快速开始](#5-下载与读取)；自行挑选一条轨迹不等于复现正式 One-Shot 的示例选择。
+读取时在下载示例中设置 `prefix = "data/real-world/one-shot"`，预期为 20 条轨迹、15,667 行。
 
 ## 3. 仿真世界
 
@@ -214,13 +214,13 @@ subset_catalog.json               机器可读子集索引
 data/
 ├── real-world/
 │   ├── full-shot/                当前可用的 LeRobot 数据根目录
-│   └── one-shot/                 预留
+│   └── one-shot/                 当前可用的 LeRobot 数据根目录
 └── simulation/
     ├── full-shot/                当前可用的 LeRobot 数据根目录
     └── one-shot/                 当前可用的 LeRobot 数据根目录
 ```
 
-**已发布子集的目录才是 LeRobot 的 `root`，不是整个仓库根目录。** 预留目录目前只包含 README。
+**已发布子集的目录才是 LeRobot 的 `root`，不是整个仓库根目录。** 四个子集目录均已包含发布的训练数据。
 
 <details>
 <summary><strong>格式参考：两个已发布 Full-Shot 子集内部的文件</strong></summary>
@@ -281,6 +281,6 @@ print(sample["action"].shape, sample["observation.state"].shape)
 
 预期为 **2,000 条轨迹**、**1,530,987 个时间步**，`action` 维度为 `float32[52]`，`observation.state` 维度为 `float32[290]`。`dataset[0]` 读��的是一个时间步，而不是整条轨迹；LeRobot 根据视频元数据读取对应的相机帧。
 
-示例固定同一个仓库版本用于下载和读取。读取仿真时设置 `prefix = "data/simulation/full-shot"`，预期为 1,500 条轨迹、789,612 行，状态与动作均为 `float32[14]`。读取仿真 One-Shot 时设置 `prefix = "data/simulation/one-shot"`，预期为 15 条轨迹、8,080 行；真机 One-Shot 仍为预留。
+示例固定同一个仓库版本用于下载和读取。读取仿真时设置 `prefix = "data/simulation/full-shot"`，预期为 1,500 条轨迹、789,612 行，状态与动作均为 `float32[14]`。读取仿真 One-Shot 时设置 `prefix = "data/simulation/one-shot"`，预期为 15 条轨迹、8,080 行；读取真机 One-Shot 时设置 `prefix = "data/real-world/one-shot"`，预期为 20 条轨迹、15,667 行。
 
 **许可：**真机与仿真补充示例的许可尚未指定。RoboDojo 来源部分遵循上游适用条款（其卡片声明 Apache-2.0）；不为整个混合仓库指定统一许可。

subset_catalog.json

@@ -115,18 +115,35 @@
       "task_name_correspondence": {
         "sweep_objects": "sweep_blocks"
       }
-    }
-  ],
-  "planned_subsets": [
+    },
     {
       "subset_id": "real-world-one-shot-data",
       "domain": "real-world",
       "shot_regime": "one-shot",
-      "availability": "reserved",
+      "availability": "available",
       "tasks": 20,
       "episodes_per_task": 1,
       "episodes": 20,
-      "path_in_repo": "data/real-world/one-shot"
+      "path_in_repo": "data/real-world/one-shot",
+      "name_en": "Real-World One-Shot Training Data",
+      "name_zh": "真机 One-Shot 训练集",
+      "format": "LeRobot v3.0",
+      "robot_type": "G2A",
+      "frames": 15667,
+      "fps": 30,
+      "camera_streams_per_episode": 3,
+      "state_dimensions": 290,
+      "action_dimensions": 52,
+      "splits": {
+        "train": "0:20"
+      },
+      "video_codecs": [
+        "hevc"
+      ],
+      "trimmed_unmatched_final_rows": 0,
+      "episode_mapping": "data/real-world/one-shot/provenance/episode_mapping.json",
+      "source_episode_selection": "original episode 000 per source task"
     }
-  ]
+  ],
+  "planned_subsets": []
 }