Skip to content

test_e2e_pipeline

ipw.tests.integration.test_e2e_pipeline

Tier 3 end-to-end integration tests for the full profiling pipeline.

These tests run the full pipeline including: - Energy monitor subprocess - gRPC telemetry stream - Inference client (Ollama or vLLM) - Dataset iteration - Profiling record construction - Analysis and export

They require a running inference server and energy monitor binary.

test_profile_pipeline_ollama()

Placeholder: full profile run with Ollama backend.

Source code in intelligence-per-watt/src/ipw/tests/integration/test_e2e_pipeline.py 
30
31
32
def test_profile_pipeline_ollama() -> None:
    """Placeholder: full profile run with Ollama backend."""
    pytest.skip("Full e2e pipeline test not yet implemented")

test_agentic_run_pipeline()

Placeholder: full agentic run with agent + dataset + telemetry.

Source code in intelligence-per-watt/src/ipw/tests/integration/test_e2e_pipeline.py 
35
36
37
def test_agentic_run_pipeline() -> None:
    """Placeholder: full agentic run with agent + dataset + telemetry."""
    pytest.skip("Full e2e agentic pipeline test not yet implemented")

test_export_roundtrip()

Placeholder: export JSONL + HF dataset, reload and verify.

Source code in intelligence-per-watt/src/ipw/tests/integration/test_e2e_pipeline.py 
40
41
42
def test_export_roundtrip() -> None:
    """Placeholder: export JSONL + HF dataset, reload and verify."""
    pytest.skip("Full e2e export roundtrip test not yet implemented")

test_analysis_from_run()

Placeholder: run analysis (accuracy, regression) on profiling output.

Source code in intelligence-per-watt/src/ipw/tests/integration/test_e2e_pipeline.py 
45
46
47
def test_analysis_from_run() -> None:
    """Placeholder: run analysis (accuracy, regression) on profiling output."""
    pytest.skip("Full e2e analysis test not yet implemented")