dataset: v1.2.1
#1
by AmineHA - opened
- README.md +53 -3
- full.parquet +3 -0
- hard.parquet +3 -0
- version.json +6 -0
README.md
CHANGED
|
@@ -1,3 +1,53 @@
|
|
| 1 |
-
---
|
| 2 |
-
|
| 3 |
-
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
configs:
|
| 3 |
+
- config_name: default
|
| 4 |
+
data_files:
|
| 5 |
+
- split: full
|
| 6 |
+
path: full.parquet
|
| 7 |
+
- split: hard
|
| 8 |
+
path: hard.parquet
|
| 9 |
+
---
|
| 10 |
+
|
| 11 |
+
# WebArena-Verified
|
| 12 |
+
|
| 13 |
+
## Dataset description
|
| 14 |
+
|
| 15 |
+
WebArena-Verified is a curated benchmark dataset of web tasks designed for reproducible
|
| 16 |
+
evaluation of web agents across multiple realistic websites.
|
| 17 |
+
|
| 18 |
+
## Splits
|
| 19 |
+
|
| 20 |
+
- `full`: 812 rows
|
| 21 |
+
- `hard`: 258 rows
|
| 22 |
+
|
| 23 |
+
## Schema notes
|
| 24 |
+
|
| 25 |
+
The table below reflects inferred column types from the `full` split during artifact generation.
|
| 26 |
+
|
| 27 |
+
| Column | Type |
|
| 28 |
+
| --- | --- |
|
| 29 |
+
| `sites` | `Value('string')` |
|
| 30 |
+
| `task_id` | `Value('int64')` |
|
| 31 |
+
| `intent_template_id` | `Value('int64')` |
|
| 32 |
+
| `start_urls` | `Value('string')` |
|
| 33 |
+
| `intent` | `Value('string')` |
|
| 34 |
+
| `intent_template` | `Value('string')` |
|
| 35 |
+
| `instantiation_dict` | `Value('string')` |
|
| 36 |
+
| `eval` | `Value('string')` |
|
| 37 |
+
| `revision` | `Value('int64')` |
|
| 38 |
+
|
| 39 |
+
## Metadata
|
| 40 |
+
|
| 41 |
+
- Version: `v1.2.1`
|
| 42 |
+
- Git commit: `0d2691719ba0f0c411ac49b11b75c51f4f750d93`
|
| 43 |
+
- Generated at (UTC): `2026-02-07T21:07:25Z`
|
| 44 |
+
- Dataset hash: `0e90f3feabe68cb9c7285e6989e36862a47622e7ac81b562f3126152350cb5ac`
|
| 45 |
+
- License: `Apache-2.0`
|
| 46 |
+
- Language: `en`
|
| 47 |
+
- Task categories: `web-navigation`, `information-retrieval`
|
| 48 |
+
|
| 49 |
+
## Notes
|
| 50 |
+
|
| 51 |
+
- Expected split counts: `full=812`, `hard=258`
|
| 52 |
+
- If custom split names are not supported by the Dataset Viewer, use one config per split
|
| 53 |
+
mapped to `train` as a compatibility fallback.
|
full.parquet
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ae70cbfba57d8edd509e44ed7ed31536446df80e455593a00048bee29b461d19
|
| 3 |
+
size 142939
|
hard.parquet
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:aee9ebb53935b95e4dd03d294b32648cc08665f9953d5e7cecc1ad551e02e122
|
| 3 |
+
size 77570
|
version.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"dataset_hash": "0e90f3feabe68cb9c7285e6989e36862a47622e7ac81b562f3126152350cb5ac",
|
| 3 |
+
"generated_at": "2026-02-07T21:07:25Z",
|
| 4 |
+
"git_commit": "0d2691719ba0f0c411ac49b11b75c51f4f750d93",
|
| 5 |
+
"version": "v1.2.1"
|
| 6 |
+
}
|