Add production runtime standard runner

[jarcherNV]

Promote the test-only inference loop into flashdreams.runtime as run_inference_session.

The runner validates mapping compatibility before runtime creation, maps global conditioning, drives a synchronous sequential session, writes step outputs, records timing metrics, and reliably closes output/session/runtime/metrics.

Add focused production runner tests for success, validation ordering, cleanup, and declared mapping compatibility failures. Update runtime docs to mark T4 complete and point at the new runner entry point.

[greptile-apps]

Greptile Summary

The PR promotes the inference loop into a production runtime runner and introduces shared replay, MP4-output, WebRTC-demo, and OmniDreams adapter surfaces.

• Validates mapping compatibility before creating runtime resources.
• Maps global and per-step inputs while driving a synchronous sequential session.
• Records timing metrics and closes output, session, runtime, and metrics resources.
• Adds shared demo specifications, replay and WebRTC entry points, MP4 output support, and focused tests.

Confidence Score: 5/5

The PR appears safe to merge.

The previously reported input-loss issue is fixed: the runner now constructs its per-step base from initial_inputs.step and metadata, and the regression coverage verifies that a pass-through mapping delivers those values to every session step. No blocking failure remains.

Important Files Changed

Filename	Overview
flashdreams/flashdreams/runtime/runner.py	Adds the production synchronous runner, including pre-runtime validation, preserved pass-through step inputs, timing collection, and ordered cleanup.
flashdreams/tests/test_runtime_runner.py	Covers successful multi-step execution, preservation of initial step inputs, validation ordering, mapping compatibility, metrics, and cleanup paths.
flashdreams/flashdreams/runtime/video_output.py	Adds an MP4 output target that collects runtime video chunks, validates layouts, and emits an output artifact.
flashdreams/flashdreams/runtime/demo/replay.py	Adds shared replay orchestration that prepares adapter scenarios and delegates execution to the production runner.
flashdreams/flashdreams/runtime/demo/webrtc.py	Adds shared WebRTC runtime, session-manager, application, and server construction.
integrations/omnidreams/omnidreams/demo/adapter.py	Adapts OmniDreams scenarios to the shared demo/runtime interfaces using global scenario conditioning.

Sequence Diagram

sequenceDiagram
  participant Demo as Demo/Caller
  participant Runner as run_inference_session
  participant Mapping as InputMapping
  participant Runtime as InferenceRuntime
  participant Session as InferenceSession
  participant Output as OutputTarget
  participant Metrics as MetricsRecorder

  Demo->>Runner: config, initial inputs, mapping, output
  Runner->>Mapping: validate compatibility
  Runner->>Mapping: map global conditioning
  Runner->>Runtime: create_runtime(config)
  Runtime-->>Runner: runtime
  Runner->>Runtime: start_session(mapped initial inputs)
  Runtime-->>Runner: session
  Runner->>Output: open()
  loop while next_step_request exists
    Runner->>Session: next_step_request()
    Runner->>Mapping: map_step_inputs(canonical inputs, preserved step base, request)
    Mapping-->>Runner: step inputs
    Runner->>Session: step(step inputs)
    Session-->>Runner: StepResult
    Runner->>Output: write(result)
    Runner->>Metrics: record timing samples
  end
  Runner->>Output: close()
  Runner->>Session: close()
  Runner->>Runtime: close()
  Runner->>Metrics: close()

Loading

_{Reviews (8): Last reviewed commit: "omnidreams: add shared demo API adapter" | Re-trigger Greptile}