stormlog.infer.cli

CLI for Stormlog inference profiling.

Functions

build_parser()	Build the stormlog infer parser.
cmd_analyze(args)	Analyze an inference JSONL artifact.
cmd_collect_server(args)	Collect telemetry on the server while a profile uses the same run ID.
cmd_import_execution(args)	Append a vLLM execution log's final steps to an existing artifact.
cmd_import_trace(args)	Append profiler-trace GPU activity to an existing inference artifact.
cmd_profile(args)	Run active inference profiling.
main([argv])	Run the inference CLI and return a code from stormlog.exit_codes.

stormlog.infer.cli.main(argv=None)

Run the inference CLI and return a code from stormlog.exit_codes.

Parameters:

argv (Sequence[str] | None)

Return type:

int

stormlog.infer.cli.build_parser()

Build the stormlog infer parser.

Return type:

ArgumentParser

stormlog.infer.cli.cmd_profile(args)

Run active inference profiling.

Parameters:

args (Namespace)

Return type:

int

stormlog.infer.cli.cmd_analyze(args)

Analyze an inference JSONL artifact.

Parameters:

args (Namespace)

Return type:

int

stormlog.infer.cli.cmd_collect_server(args)

Collect telemetry on the server while a profile uses the same run ID.

Parameters:

args (Namespace)

Return type:

int

stormlog.infer.cli.cmd_import_trace(args)

Append profiler-trace GPU activity to an existing inference artifact.

Parameters:

args (Namespace)

Return type:

int

stormlog.infer.cli.cmd_import_execution(args)

Append a vLLM execution log’s final steps to an existing artifact.

Parameters:

args (Namespace)

Return type:

int