Learning Self-Orchestration in Decentralized Agent Swarms
Paper • 2610.22322 • Published • 7
Qwen3-VL-4B-Instruct — Minecraft Three-Agent SFT
Paper: Learning Self-Orchestration in Decentralized Agent Swarms
Code: https://github.com/Embodied-Minds-Lab/MPT
For the post-trained model initialized from this checkpoint, see Xkev/mc_society.
For the single-agent counterpart, see Xkev/mc_single_sft.
Training
- Base model:
Qwen/Qwen3-VL-4B-Instruct - Dataset: Minecraft iron-pickaxe tasks
- Framework: transformers
Trainer - Hyperparameters:
lr=1e-5,lr_warmup_ratio=0.03, cosine schedule,epochs=2
Intended use
Research on multi-agent post-training and embodied agents. Not intended for general dialog or production.
License
Apache-2.0.
Model tree for Xkev/mc_society_sft
Base model
Qwen/Qwen3-VL-4B-Instruct