WIA Datasets
Collection
World-model datasets for the WIA project. • 4 items • Updated • 1
messages listlengths 3 127 | domain stringclasses 7
values | dataset stringclasses 11
values |
|---|---|---|
[
{
"role": "system",
"content": "Role: Operate as a language world model for a mobile accessibility-control environment. You model how the environment changes after an issued action and report only the resulting observation.\n\nInput: Interpret the goal or task metadata, latest observation, action block, and... | android | AndroidControl |
[{"role":"system","content":"Role: You are an observation generator for a language world model of a (...TRUNCATED) | search | OpenResearcher |
[{"role":"system","content":"Role: Serve as a language world model for a web search and page-browsin(...TRUNCATED) | search | OpenResearcher |
[{"role":"system","content":"Role: You are a language world model for a mobile accessibility-control(...TRUNCATED) | android | AndroidControl |
[{"role":"system","content":"Role: Operate as a language world model for a web search and page-brows(...TRUNCATED) | search | OpenResearcher |
[{"role":"system","content":"Role: You are the simulator side of a language world model for a softwa(...TRUNCATED) | swe | SWE-Gym |
[{"role":"system","content":"Role: You are a state-transition world model for a terminal command-exe(...TRUNCATED) | terminal | TMax |
[{"role":"system","content":"Role: You are an observation generator for a language world model of an(...TRUNCATED) | mcp | AWM |
[{"role":"system","content":"Role: Operate as a language world model for a terminal command-executio(...TRUNCATED) | terminal | TMax |
[{"role":"system","content":"Role: Function as the environment model for a mobile accessibility-cont(...TRUNCATED) | android | AndroidControl |
Each line is one sample, and each sample contains an ordered set of messages.
| File | Purpose | Samples | Top-level fields |
|---|---|---|---|
train.jsonl |
World-model supervised fine-tuning data | 9,181 | 3 |
| Field | Type | Purpose |
|---|---|---|
messages |
array | Ordered conversation messages in the sample. |
domain |
string | Task domain, such as swe, terminal, or web. |
dataset |
string | Source dataset name. |
Each message in messages[] contains three fields:
| Field | Type | Purpose |
|---|---|---|
role |
string | Message role: system, user, or assistant. |
content |
string | Message content. |
loss |
boolean | Whether the message contributes to the training loss. |
pip install -U huggingface_hub
from huggingface_hub import snapshot_download
snapshot_download(
repo_id="LightningCreeper/WIA-Corpus-SFT",
repo_type="dataset",
allow_patterns=["train.jsonl"],
local_dir="WIA-Corpus-SFT",
)