Upload MobileGym batch 1

README.md

@@ -0,0 +1,119 @@
+---
+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: mobile.use
+  data_files:
+  - split: train
+    path:
+    - "mobile/use/train*parquet"
+    - "mobile/use/train/*.parquet"
+    - "mobile/use/train/*/*.parquet"
+---
+
+# cua-lite/MobileGym
+
+MobileGym GPT-5.5 grounded teacher trajectories; every trajectory kept, quality gates tagged in metadata.others.exclude_reason (filter with not exclude_reason and episode_return>0.5)
+
+## Origin
+
+
+
+## Load via `datasets`
+
+```python
+from datasets import load_dataset
+
+# entire dataset
+ds = load_dataset("cua-lite/MobileGym")
+
+# just one (platform, task_type) cohort
+ds = load_dataset("cua-lite/MobileGym", "mobile.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 |
+|---|---|---|---:|---:|
+| mobile | use | use | 320 | 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:
+
+```
+$CUA_LITE_DATASETS_ROOT/cua-lite/MobileGym/
+  images/<hash[:2]>/<hash>.<ext>                          # content-addressed image store
+  <platform>/<task_type>/<split>[/<variant>].parquet      # rows reference images by relative path
+```
+
+Rows in the local parquet have `images: list[str]`; bytes are extracted to
+the image store. `lite.train.export.export_sft` consumes the local
+form directly with `--image-root=$CUA_LITE_DATASETS_ROOT`.
+
+- Total unique images: **2,054**
+- Image store size: **1.07 GB**
+
+## Notes
+
+Staged via `lite.data.hf.stage` from rollout log-roots: .data/rollout/mobilegym/gpt/398f9e5ab/train_annotated (row filter: none).
+
+## License & citation
+
+other
+
+

mobile/use/train/shard-00000-of-00019.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a329aafbdfa5566b23688ae6a5d1e7ea2992ceb68e3539d480c6dcdb8b8d3767
+size 112440121

mobile/use/train/shard-00001-of-00019.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dc1452d1bda2d8179102af8dbd2cc75eb6c59753523643b42eabbaa2d95e8503
+size 150753075

mobile/use/train/shard-00002-of-00019.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ffa465560a40a994471b222776e613258c30acbd94b301a85ee562f401294524
+size 105389245

mobile/use/train/shard-00003-of-00019.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a80d734bd41aeb04b861438af2fc0545b7bb0af0e56a04b2505c07c2f0153c36
+size 73454687

mobile/use/train/shard-00004-of-00019.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4f534f1ea4e82a8eb9cda18d4d0c59cbdd65c88814ee963bbe467a8b3061040c
+size 93957850

mobile/use/train/shard-00005-of-00019.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b6d21e7969d79010787fa36e4be800c28547c6fa8d0ff6b72c73ac81cd33a383
+size 117670748

mobile/use/train/shard-00006-of-00019.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:24ce2fc422b2dd3ccbd27125843b52ddc790e8eb648bcde9f6a47429565173ba
+size 86531686

mobile/use/train/shard-00007-of-00019.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b4fca783dfd8c452307977b55383e8c8295db393c79585b6dadc7b49fba8376a
+size 75723793

mobile/use/train/shard-00008-of-00019.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:88d2654b2bbfc4f2d01475c1a9d4d5fc04d1e86afc1825dbd236be35a88fd52b
+size 169501507

mobile/use/train/shard-00009-of-00019.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8938acf7fc1544d2610bff4cb78acd799f40414b75e4c8ddb10acda42e472a3b
+size 186623285

mobile/use/train/shard-00010-of-00019.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:961e13ab7f6fea20c870b55a91a63d93da1564a07bae6124838a6f9c810bb76c
+size 111165412

mobile/use/train/shard-00011-of-00019.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:428b50ee95db308f005b84d6c17e3004687240d62fe2edca850181135e40246e
+size 92464962

mobile/use/train/shard-00012-of-00019.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ce463a6697039d2d9d5061192a8cea187ff3a0756096b23025a7d1bc609f16fd
+size 93763218

mobile/use/train/shard-00013-of-00019.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9716fa1f1f89c7194d546a02e4b35659b1f8ebccf00499e5ddeaa1384ab8b83e
+size 59660712

mobile/use/train/shard-00014-of-00019.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6402c29d9b2a1f01a84c16f9f49b09ae5be3b7c11d1ba60baaeb3a777a4ddb8c
+size 85479773

mobile/use/train/shard-00015-of-00019.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2c00590c2d660fdae4c6f6b253ebaa28cdbee209e82b0c36862a4d532b0ad518
+size 69915168

mobile/use/train/shard-00016-of-00019.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4829c39fc70e3a820a6539fa65c804aad3430ed1db0537f3cce4091ada927316
+size 61349868

mobile/use/train/shard-00017-of-00019.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:031e589af232734efd03720d948dd591cdcfc736f0c21a94004207aae3e8d8ec
+size 85163696

mobile/use/train/shard-00018-of-00019.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dd86f508f8a769f73ba777e8b7afd57d04c530719e38289de52ae767a211f4e5
+size 32685907

stats.json

@@ -0,0 +1,10 @@
+{
+  "rows_in": 0,
+  "rows_out": 320,
+  "rows_dropped": 0,
+  "unique_images": 2054,
+  "image_store_bytes": 1065253722,
+  "by_partition": {
+    "mobile::use::train::use": 320
+  }
+}