README.md

metadata

license: other
tags:
  - cua-lite
  - gui
  - sft
task_categories:
  - image-text-to-text
configs:
  - config_name: default
    data_files:
      - split: train
        path:
          - '*/*/train*parquet'
          - '*/*/train/*.parquet'
          - '*/*/train/*/*.parquet'
  - config_name: desktop
    data_files:
      - split: train
        path:
          - desktop/*/train*parquet
          - desktop/*/train/*.parquet
          - desktop/*/train/*/*.parquet
  - config_name: web
    data_files:
      - split: train
        path:
          - web/*/train*parquet
          - web/*/train/*.parquet
          - web/*/train/*/*.parquet
  - config_name: desktop.use
    data_files:
      - split: train
        path:
          - desktop/use/train*parquet
          - desktop/use/train/*.parquet
          - desktop/use/train/*/*.parquet
  - config_name: web.use
    data_files:
      - split: train
        path:
          - web/use/train*parquet
          - web/use/train/*.parquet
          - web/use/train/*/*.parquet

cua-lite/Lite.CUAGym

Lite.CUAGym GPT-5.5 teacher trajectories from the July 2026 V1 and current-image collections. All annotated trajectories are published except /opt/env tool leaks; quality gates live in metadata.others.exclude_reason. For SFT, filter with not exclude_reason and episode_return > 0.5.

Origin

Load via datasets

from datasets import load_dataset

# entire dataset
ds = load_dataset("cua-lite/Lite.CUAGym")

# just one platform
ds = load_dataset("cua-lite/Lite.CUAGym", "desktop")

# just one (platform, task_type) cohort
ds = load_dataset("cua-lite/Lite.CUAGym", "desktop.use")

You can also filter by metadata.platform / metadata.task_type / metadata.others.* after loading; every row carries a rich metadata struct (see schema below).

Schema

Each row has these columns:

column	type	notes
images	list[Image]	embedded PNG/JPEG bytes; HF viewer renders thumbnails
messages	list[struct]	OpenAI-style turns with role + structured content
metadata	struct	{platform, task_type, extra_tool_schemas, valid_actions, others{...}}

Coordinate values in messages are normalized to [0, 1000] integers.

Image-dedup (grounding.* / understanding cohorts). These cohorts are single-image-per-row and many rows share the same screenshot, so to avoid re-embedding identical image bytes once per instruction they are stored folded: one row per unique screenshot (image embedded once), carrying an extra _folded column — a JSON string with the authoritative list of {messages, metadata} members for that screenshot. The row's top-level messages is the members concatenated for viewer convenience. use cohorts are not folded. Use lite.data.hf.download to consume this repo — it unfolds automatically back to one row per instruction; reading the parquet directly yields the folded form.

Layout

<platform>/<task_type>/<split>/shard-NNNNN-of-NNNNN.parquet                  # single-variant cohort
<platform>/<task_type>/<split>/<variant>/shard-NNNNN-of-NNNNN.parquet        # multi-variant cohort

• platform ∈ {desktop, mobile, web}
• task_type ∈ {understanding, grounding.action, grounding.point, grounding.bbox, use} — used verbatim as the dir component
• HF config names are <platform>.<task_type> by default (e.g. mobile.grounding.action) — UNLESS the dataset was staged with --config-names, which sets verbatim, explicitly-chosen config names (see the configs: block above for the authoritative list). The agent registry lookup key in code is <agent>@<platform>@<task_type> (e.g. qwen3_vl@mobile@grounding.action); only this user-facing token uses . between platform and task_type, because @ triggers a 403 on the dataset-viewer's signed image URLs.
• HF split names stay train / validation (the datasets library blacklists <>:/\|?* in split names; everything else is fine in config_name)
• validation is an in-distribution held-out slice (never used in training); test is reserved for out-of-distribution benchmark datasets

Stats

platform	task_type	variant	train	validation
desktop	use	use	8,721	0
web	use	use	1,330	0

Local mirror & SFT export

For local workflows (SFT export, dedup, mixing across datasets), use lite.data.hf.download to mirror this repo back to the canonical local layout:

$CUA_LITE_DATASETS_ROOT/cua-lite/Lite.CUAGym/
  images/<hash[:2]>/<hash>.<ext>                          # content-addressed image store
  <platform>/<task_type>/<split>[/<variant>].parquet      # rows reference images by relative path

Rows in the local parquet have images: list[str]; bytes are extracted to the image store. lite.train.export.export_sft consumes the local form directly with --image-root=$CUA_LITE_DATASETS_ROOT.

• Total unique images: 189,285
• Image store size: 100.32 GB

Notes

Staged via lite.data.hf.stage from rollout log-roots: .data/rollout/lite.cuagym/gpt/teacher_v3_audited_medium_20260714/train_annotated_v3_20260730, .data/rollout/lite.cuagym/gpt/current_remaining_20260729/train_annotated_v3_20260730 (row filter: none).

License & citation

other