---
license: mit
task_categories:
  - text-generation
language:
  - en
tags:
  - textworld
  - world-model
  - w2r
  - tw-wm-tm
  - gpt-5
  - step170
pretty_name: TW WM-TM Step170 TextWorld WM/W2R Trajectories - GPT-5 Agent
---

# TW WM-TM Step170 TextWorld WM/W2R Trajectories - GPT-5 Agent

This dataset contains TextWorld task-success trajectories for the world model `Ricardo-H/tw-wm-tm-0501-step-170`, evaluated with `gpt-5-20250807` (CloudGPT) as the agent.

## Configuration

- World model: `Ricardo-H/tw-wm-tm-0501-step-170`
- World model serving: vLLM, tensor parallel size 1, `max_model_len=32768`
- Agent: `gpt-5-20250807` via CloudGPT
- Eval concurrency: 8
- Temperature: 1
- Task: TextWorld test split (200 items)

## Metrics

- WM total items: 200
- WM processed items: 200
- WM API errors: 0
- WM accuracy: 100.00%
- W2R total items: 200
- W2R processed items: 200
- W2R errors: 0
- W2R accuracy: 100.00%

## Files

- `textworld_*.json`: agent and world-model interaction trajectories.
- `_metrics.json`: WM metrics.
- `valid_on_real_env/textworld_*.json`: W2R replay results on the real TextWorld environment.
- `valid_on_real_env/_metrics.json`: W2R metrics.
- `metadata.json`: configuration and run notes.