|
Download README.md from GameIntelligenceLab/Script2Play-4B: direct link, hf CLI and curl.
- Browser
- Download file 1.8 kB
-
https://huggingface.co/GameIntelligenceLab/Script2Play-4B/resolve/main/README.md
- Command line
-
hf download hf://GameIntelligenceLab/Script2Play-4B/README.md
-
curl -L -o README.md https://huggingface.co/GameIntelligenceLab/Script2Play-4B/resolve/main/README.md
1.8 kB
| base_model: Qwen/Qwen3.5-4B-Base | |
| pipeline_tag: image-text-to-text | |
| tags: | |
| - vision-language-action | |
| - game-playing | |
| - behavior-cloning | |
| # Script2Play-4B | |
| The 4B vision-language-action policy of | |
| [From Game Generation to Game Playing: Fully Automated Ability Training for Vision-Language-Action Models](https://arxiv.org/abs/2611.11136) | |
| ([project page](https://assettool-labs.github.io/script2play/)). It plays games from the screen: | |
| one frame and a task in, the next 200 ms of keyboard and mouse out. | |
| Qwen3.5-4B-Base post-trained by behavior cloning on 150.3 h of verified scripted-teacher | |
| demonstrations in generated games: 26.9 h of Minecraft and 123.4 h of Roblox. No ViZDoom or UE5 | |
| gameplay is in the training data, and no human gameplay, RL or DAgger is used. | |
| ## Usage | |
| Serve it with vLLM 0.26 and query its OpenAI-compatible endpoint; the inference client, the exact | |
| request the scores were measured with, and the benchmark harnesses are in | |
| [AssetTool-Labs/script2play-inference](https://github.com/AssetTool-Labs/script2play-inference). | |
| ```bash | |
| vllm serve PATH_TO_THIS_REPO --served-model-name script2play --trust-remote-code \ | |
| --max-model-len 8192 --max-num-batched-tokens 8192 --limit-mm-per-prompt '{"image":1,"video":0}' \ | |
| --attention-backend FLASHINFER --mm-encoder-attn-backend TORCH_SDPA | |
| ``` | |
| Each call sends the brief system prompt with the task, and one frame (832 px wide for Doom-like | |
| and Roblox games, 640x360 as is for Minecraft) as a JPEG. The reply is one line, | |
| `dx dy dz ; K1 ; K2 ; K3 ; K4`: the mouse displacement in px over the 200 ms (0.15 degrees of view | |
| turn per px) and the scroll steps, then the keys held in each 50 ms step. Sample at temperature | |
| 0.8, top-p 0.99, at most 48 tokens, stopping at `<|action_end|>` or `<|im_end|>`, with thinking | |
| disabled. | |