reporting

HTML fragments for reportable post-MIP node types.

Functions

build_post_mip_report_payloads	Collect current node observations and lineage once for all node renderers.
render_aiperf_report	Render AIPerf throughput/latency observations and timeout evidence.
render_downstream_evaluation_report	Render lmms-eval task metrics for downstream-evaluation nodes.
render_evaluation_report	Render one candidate-evaluation node.
render_global_kd_report	Render several candidate KD histories on shared, lineage-colored plots.

build_post_mip_report_payloads(root, nodes)

Collect current node observations and lineage once for all node renderers.

Parameters:

root (Path)

Return type:

dict[str, dict[str, Any]]

render_aiperf_report(section_id, payload)

Render AIPerf throughput/latency observations and timeout evidence.

Parameters:

• section_id (str)

• payload (Mapping[str, Any])

Return type:

str

render_downstream_evaluation_report(section_id, payload)

Render lmms-eval task metrics for downstream-evaluation nodes.

Parameters:

• section_id (str)

• payload (Mapping[str, Any])

Return type:

str

render_evaluation_report(section_id, payload)

Render one candidate-evaluation node.

Parameters:

• section_id (str)

• payload (Mapping[str, Any])

Return type:

str

render_global_kd_report(section_id, payload)

Render several candidate KD histories on shared, lineage-colored plots.

Parameters:

• section_id (str)

• payload (Mapping[str, Any])

Return type:

str