Reproducing evals
Commands to run and compare eval baselines.
pip install -e "packages/techtide-swarm[dev]"
export OPENROUTER_API_KEY=...
export SWARM_EVAL_BUDGET_USD=5.0
swarm eval --save-baseline --compareCommands to run and compare eval baselines.
pip install -e "packages/techtide-swarm[dev]"
export OPENROUTER_API_KEY=...
export SWARM_EVAL_BUDGET_USD=5.0
swarm eval --save-baseline --compare