The normalization layer for robotics data. Convert, inspect, visualize, score, and discover datasets across every major format.

RLDS  ===\          /===> LeRobot
Zarr  ===| Episode/Frame |===> RoboDM
HDF5  ===/          \===> RLDS
View on GitHub Browse Datasets
$ pip install -e ".[all]"
Features

Everything you need for robotics data

One toolkit to convert, inspect, score, filter, segment, and browse robotics datasets.

Format Conversion
Convert between RLDS, LeRobot, Zarr, HDF5, Rosbag, and RoboDM with a single command. Hub-and-spoke architecture means O(n) not O(n²).
$ forge convert hf://lerobot/pusht ./output --format rlds
Dataset Inspection
Auto-detect format, list episodes, cameras, action/state dimensions, FPS, and schema. Works with local paths and HuggingFace URIs.
$ forge inspect hf://lerobot/aloha_sim_cube
Quality Scoring
Score every episode 0-10 with 8 research-backed metrics. Detect jerky demos, dead actions, gripper chatter, and idle periods from proprioception alone.
$ forge quality ./my_dataset --export report.json
Episode Filtering
Filter datasets by quality score, flags, or episode IDs. Supports dry-run previews and pre-computed quality reports.
$ forge filter ./dataset ./filtered --min-quality 6.0
Episode Segmentation
PELT changepoint detection on proprioception signals. Automatically split episodes into sub-skills, regime changes, and idle periods.
$ forge segment ./dataset --label --plot timeline.png
Dataset Registry
Curated catalog of 23+ prominent robotics datasets. Search, filter, and download by name. Use dataset IDs directly in any command.
$ forge inspect droid # resolves via registry
See it in action

Real output from real datasets

Every output below was generated by running Forge on the pusht dataset.

forge inspect
$ forge inspect hf://lerobot/pusht Dataset: lerobot/pusht Format: lerobot-v3 (v3.0) Episodes: 206 Total frames: 25,650 Observation Schema ┏━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━┳━━━━━━━┓ ┃ Field ┃ Type ┃ Shape ┃ ┡━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━╇━━━━━━━┩ │ observation.state │ float32 │ (2,) │ │ next.success │ bool │ (1,) │ │ next.reward │ float32 │ () │ │ next.done │ bool │ () │ └───────────────────┴─────────┴───────┘ Action: float32 (2,) Cameras: image: 96x96 (rgb) FPS: 10 Language: yes (100% coverage) Sample: "Push the T-shaped block onto the T-shaped target."
forge quality
$ forge quality hf://lerobot/pusht Analyzing episodes... ━━━━━━━━━━━━━━━━━━ 206/206 ╭────────── Quality Report: pusht (206 episodes) ──────────╮ │ │ │ Overall Quality Score: 8.5 / 10 │ │ │ │ Smoothness (LDLJ) ███████░░░ 0.75 OK │ │ Dead Actions █████████░ 0.99 OK │ │ Gripper Health ██████████ 1.00 OK │ │ Static Detection ██████████ 1.00 OK │ │ Timestamp Regularity ██████████ 1.00 OK │ │ Action Saturation ████████░░ 0.87 OK │ │ Action Diversity ███░░░░░░░ 0.30 OK │ │ │ ╰───────────────────────────────────────────────────────╯
forge segment
$ forge segment pusht --sample 8 --label --penalty aic Resolved from registry: PushT (lerobot) Format: lerobot-v3 | Signal: observation.state | Penalty: aic Segmentation Results ┏━━━━━━━━━━━━━━━━┳━━━━━━━━┳━━━━━━━━━━┳━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━┓ ┃ Episode ┃ Frames ┃ Segments ┃ Changepoints ┃ Labels ┃ ┡━━━━━━━━━━━━━━━━╇━━━━━━━━╇━━━━━━━━━━╇━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━┩ │ episode_000000 │ 161 │ 6 │ 20, 33, 63... │ moving -> fine_m... │ │ episode_000001 │ 118 │ 6 │ 13, 34, 49... │ fine_m -> fine_m... │ │ episode_000002 │ 141 │ 7 │ 12, 27, 56... │ moving -> fine_m... │ │ episode_000003 │ 159 │ 7 │ 28, 42, 60... │ fine_m -> fine_m... │ │ episode_000004 │ 159 │ 8 │ 12, 22, 45... │ moving -> fine_m... │ │ episode_000005 │ 157 │ 6 │ 30, 47, 83... │ moving -> moving... │ │ episode_000006 │ 69 │ 4 │ 14, 46, 57 │ fine_m -> moving... │ │ episode_000007 │ 169 │ 7 │ 12, 43, 59... │ moving -> fine_m... │ └────────────────┴────────┴──────────┴─────────────────┴───────────────────────┘ ╭────────────────── Summary ──────────────────╮ │ Episodes: 8 │ │ Mean segments/episode: 6.38 │ │ Range: 4 — 8 │ │ Total changepoints: 43 │ ╰───────────────────────────────────────────╯
forge filter
$ forge filter ./dataset ./filtered --min-quality 6.0 Filtering episodes... ━━━━━━━━━━━━━━━━━━ 206/206 ┏━━━━━━━━━━━━━━━━┳━━━━━━━┳━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━┓ ┃ Episode ┃ Score ┃ Flags ┃ Status ┃ ┡━━━━━━━━━━━━━━━━╇━━━━━━━╇━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━┩ │ episode_000000 │ 8.7 │ │ KEEP │ │ episode_000001 │ 9.1 │ │ KEEP │ │ episode_000002 │ 8.4 │ │ KEEP │ │ episode_000003 │ 5.2 │ jerky, hesitant │ EXCL │ │ episode_000004 │ 7.8 │ │ KEEP │ │ episode_000005 │ 3.1 │ mostly_static │ EXCL │ │ episode_000006 │ 8.9 │ │ KEEP │ │ ... 199 more episodes ... │ └────────────────┴───────┴────────────────────┴────────┘ ╭────────────── Filter Results ──────────────╮ │ │ │ Episodes kept: 189 / 206 │ │ Episodes excluded: 17 │ │ Written to: ./filtered/ │ │ │ ╰─────────────────────────────────────────────╯
forge segment pusht --label --plot timeline.png — semantic phase labels via proprioception
Segmentation timeline visualization
Architecture

Hub-and-spoke, not N×M

Add a reader, get all writers for free. Add a writer, get all readers for free.

RLDS / Open-X
LeRobot v2/v3
GR00T
Zarr
HDF5
Rosbag / MCAP
RoboDM
→
Episode / Frame
Intermediate Representation
→
RLDS
LeRobot v2/v3
RoboDM
Compatibility

Format support matrix

Read, write, and visualize across every major robotics data format.

Format Read Write Visualize Notes
RLDS ✓ ✓ ✓ Open-X, TensorFlow Datasets
LeRobot v2/v3 ✓ ✓ ✓ HuggingFace, Parquet + MP4
GR00T ✓ — ✓ NVIDIA Isaac, LeRobot v2 with embodiment metadata
RoboDM ✓ ✓ ✓ Berkeley's .vla format, up to 70x compression
Zarr ✓ — ✓ Diffusion Policy, UMI
HDF5 ✓ — ✓ robomimic, ACT/ALOHA
Rosbag ✓ — ✓ ROS1 .bag, ROS2 MCAP
Quality

8 research-backed metrics

Score every episode 0-10 from proprioception data alone. No video processing needed.

Smoothness (LDLJ)
Jerk-based smoothness
Dead Actions
Zero/constant detection
Gripper Chatter
Rapid open/close cycles
Static Detection
Idle period detection
Timestamp Regularity
Dropped frames & jitter
Action Saturation
Time at hardware limits
Action Entropy
Diversity vs repetition
Path Length
Wandering & hesitation
$ forge quality hf://lerobot/aloha_sim_cube --export report.json
$ forge filter ./dataset ./filtered --min-quality 6.0 # remove bad demos
Registry

23+ curated robotics datasets

Browse, search, and download by name. Use dataset IDs directly in any command.

DROID
~76,000 episodes
rlds
Bridge V2
~60,096 episodes
rlds
Open-X Embodiment
~2,200,000 episodes
rlds
AgiBot World
~1,000,000 episodes
lerobot-v2
RoboSet
~100,000 episodes
hdf5
ALOHA
Bi-manual demos
lerobot
Browse All Datasets
Get Started

Up and running in 60 seconds

1
Install
Clone the repo and install with pip.
$ git clone https://github.com/arpitg1304/forge.git
$ cd forge
$ pip install -e ".[all]"
2
Try a demo dataset
Download, inspect, and score a small dataset in one command.
$ forge demo
# Downloads pusht, runs inspect + quality
3
Convert and go
Convert to any format you need for training.
$ forge convert droid ./output --format lerobot-v3

Forge — MIT License — GitHub · Dataset Registry