configs/config.yaml

# TRN-10: Hydra-composable FATHOM config tree.
# Invoke: python -m train.grpo model=qwen_1_5b train=grpo data=v1 reward=v1

defaults:
  - model: qwen_1_5b        # OVERRIDE: model=qwen_0_5b_smoke for laptop
  - train: grpo             # OVERRIDE: train=sft
  - data: v1
  - reward: v1
  - _self_

seed: 42
output_dir: outputs