Microsoft and Hugging Face Release ThinkingBox Workflow for Training Agents With Database Checks

Microsoft and Hugging Face released ThinkingBox-Training on GitHub, a workflow that turns executable checks of final database states into rewards for agent post-training. Hugging Face reported that after 50 GRPO steps on 187 tasks outside ThinkingBox-Bench, Qwen3.8-27B’s average benchmark success rose from 51.70% to 60.78%. Tasks passed on all 20 attempts increased from 38 to 107, although tasks solved at least once fell from 453 to 446. The release includes task-selection, rollout, reward, and Verl integration code. Its public sample task list exercises the interface; the private 187-task training pool used for the reported results is not included. Running the workflow requires NVIDIA GPUs with a compatible CUDA/NCCL stack, a local Qwen3.8-27B checkpoint, and OpenAI-compatible endpoints for the simulated user and judge. The validated reference configuration uses three nodes with eight H100 GPUs each.

Generated by AI from reports · Updated 1 hours ago

There is not enough continuous observation data to plot a trend yet.