Dynamo Sweeper Integration

AI Simulate’s Sweeper core does not depend on Dynamo. Dynamo owns its optional Planner and Router sweep configuration providers and the transitional DynamoReplayRunnerFactory.

Install

The supported prebuilt environment is the dynamo-planner image. The image builds AI Simulate and Dynamo from the same source revision and installs both wheels. The AI Simulate wheel remains inside the image and is not published as a release artifact.

For source development, build the matching Dynamo bindings and install both source distributions:

$
python -m pip install pip "maturin[patchelf]"
$
cd lib/bindings/python
$
maturin develop --uv --release --features aic-forward-pass
$
cd ../../..
$
python -m pip install --no-deps -e .
$
python -m pip install -e ./aisimulate
$
python -m pip install -r container/deps/requirements.planner.txt

The repository does not yet provide an ai-dynamo[simulation] extra. Add that extra only after the full AI Simulate and Replay refactor has a supported distribution path.

Run a Dynamo Sweep

$
python aisimulate/examples/sweeper/tools/run_sweep.py \
>
  --config aisimulate/examples/sweeper/configs/smart_sweep.yaml

Or compose the runtime directly:

1
from aisimulate.sweeper import SmartSearchConfig, Sweeper
2
from dynamo.replay.simulation import DynamoReplayRunnerFactory
3
4
config = SmartSearchConfig.from_yaml("smart_sweep.yaml")
5
candidates = Sweeper(
6
    runner_factory=DynamoReplayRunnerFactory(),
7
).run(config)

There is no --enable-dynamo flag. Selecting a Dynamo adapter in configuration and injecting the Dynamo runner are the explicit composition points.

Dynamo Providers

The ai-dynamo distribution registers these entry points:

Each adapter value is a search space, not one concrete Planner or Router config:

1
adapters:
2
  dynamo.router:
3
    search_space:
4
      mode: [kv_router, round_robin]
5
      overlap_score_credit: [0.0, 0.5, 1.0]
6
  dynamo.planner:
7
    search_space:
8
      scaling_policy: [disabled, load_180_5]
9
      fpm_sampling: [default]
10
      load_sensitivity: [default]

The Planner provider derives load-predictor parameters from all configured scaling intervals during generate_search_space. Enabled candidates materialize a concrete PlannerConfig runtime hook. The Router provider materializes either round-robin behavior without a hook or a concrete KV-router hook.

Replay Refactor Coordination

Replay is being decoupled in parallel. Backend-only Sweeper will compose with Dynamo-free Replay; Sweeper with Dynamo hooks will compose with Dynamo Replay. This branch keeps DynamoReplayRunnerFactory as the transitional boundary so the two refactors can land independently.