run-pipeline

To run a Snakemake pipeline on a SLURM cluster, this Skill orchestrates end-to-end workflows by parsing dataset arguments, validating configurations, and submitting jobs with a dedicated SLURM profile to execute ML experiments.

Yes, you can preview expected Snakemake outputs before SLURM submission by performing a dry run. This Skill verifies dataset existence and validates the workflow to determine expected outputs without launching actual compute jobs.

The best way to track experiment provenance for Snakemake workflows is through MLflow tracking. This Skill captures SLURM logs and integrates MLflow tracking automatically during pipeline submission to maintain complete experiment records.

To execute knowledge distillation targets like student_fusion_kd using a Snakemake pipeline, you parse the dataset and optional target from arguments. This Skill applies the appropriate KD-GAT pipeline configuration for targets including teacher_fusion and student_fusion.

Yes, this pipeline orchestration tool supports automotive KD-GAT experiments across multiple datasets. It applies appropriate pipeline configurations for common targets such as teachers, students_nokd, and students across all specified datasets.

Snakemake dataset validation fails before SLURM submission when the specified dataset does not exist. This Skill parses and validates the dataset from arguments first, ensuring dataset existence before proceeding to dry-run validation and job submission.