swanlab convert

bash

swanlab convert [OPTIONS]

Option	Description
`-t`, `--type`	Select the conversion type, options include `tensorboard`, `wandb`, `mlflow`, `wandb-local`, default is `tensorboard`.
`-p`, `--project`	Set the SwanLab project name for the conversion, default is None.
`-w`, `--workspace`	Set the workspace where the SwanLab project is located, default is None.
`--mode`	Set the SwanLab logging mode, options include `online`, `local`, `offline`, `disabled`, default is `online`.
`-l`, `--logdir`	Set the log file save path for the SwanLab project, default is None.
`--tb-log-dir`	Path to the TensorBoard log files (tfevent) to be converted.
`--tb-types`	The types of TensorBoard logs to convert, options include `scalar`, `image`, `audio`, `text`, default is `scalar`; separate multiple types with commas.
`--wb-project`	Name of the W&B project to be converted.
`--wb-entity`	Entity where the W&B project to be converted is located.
`--wb-runid`	ID of the W&B Run to be converted.
`--wb-dir`	Directory where the W&B local log files are stored, default is `./wandb`.
`--wb-run-dir`	The specific W&B local run directory name; if omitted, all runs under `--wb-dir` will be converted.
`--mlflow-url`	The tracking URL of the MLflow server, default is `http://127.0.0.1:5000`.
`--mlflow-exp`	The name or ID of the MLflow experiment to be converted (required).
`--mlflow-runid`	ID of a specific MLflow Run to be converted.
`--resume`	Resume mode: uses the source Run ID as the SwanLab Run ID for resuming; must be used together with `--wb-runid`.

Introduction

Convert content from other logging tools into SwanLab projects.
Supported tools for conversion include: TensorBoard, Weights & Biases, MLflow.