skills.rest
Ascend
AscendOFFICIAL
Explore all skills by this author →
WebsiteGitHub

mindspeed-fsdp2-config-migration

Translate legacy MindSpeed MM FSDP2 training configurations into valid YAML with mapping tables and migration checklists.

31|1|Updated Mar 28, 2026
Search Tags:#mlops#yaml#mapping#distributed-training#config-migration#fsdp2#mindspeed-mm
One-click install
npx skills add https://github.com/Ascend/agent-skills --skill mindspeed-fsdp2-config-migration
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: mindspeed-fsdp2-config-migration
Source: https://github.com/Ascend/agent-skills/tree/main/skills/mindspeed-mm-fsdp2-migration/mindspeed-fsdp2-config-migration
Command: npx skills add https://github.com/Ascend/agent-skills --skill mindspeed-fsdp2-config-migration

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill helps teams translate legacy MindSpeed MM FSDP2 training configurations into valid YAML, ensuring proper model/dataset registration alignment and safe field layering.

Core Features & Use Cases

  • Config mapping ensures model_id, dataset_type, and training.plugin are consistently reflected in MindSpeed MM FSDP2 YAML.
  • Validation enforces strict/extra field layering, and checks for required fields before deployment.
  • Use Case: When migrating an existing training setup to FSDP2, generate a ready-to-run YAML draft with a mapping table and a checklist for acceptance.

Quick Start

Provide the source config and target registrations to produce a compliant MindSpeed FSDP2 YAML config.

Frequently Asked Questions about mindspeed-fsdp2-config-migration

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I migrate legacy MindSpeed MM training configurations to FSDP2 YAML?▼

To migrate legacy MindSpeed MM configurations to FSDP2 YAML, provide the source config and target registrations to generate a compliant YAML draft, complete with a mapping table and validation checklist for safe field layering.

What needs to be mapped when converting configs for MindSpeed MM FSDP2?▼

Config mapping for MindSpeed MM FSDP2 requires aligning model_id, dataset_type, and training.plugin fields to ensure legacy training configurations are correctly reflected in the generated YAML.

How does config validation handle strict and extra fields during FSDP2 migration?▼

Config validation enforces strict and extra field layering by checking for required fields before deployment, guarding the migrated FSDP2 YAML against missing values and invalid module references.

Can I use this config migration process for multi-asset training setups?▼

Yes, the FSDP2 config migration process validates multi-asset options and establishes accurate model and dataset registration mappings to ensure multi-asset training setups deploy correctly.

What do I need to prepare before starting an FSDP2 config migration?▼

Before starting an FSDP2 config migration, you need the legacy MindSpeed MM source configuration and the target model and dataset registrations to produce a valid, ready-to-run YAML output.

Why are missing fields and invalid module references blocked during YAML generation?▼

Missing fields and invalid module references are blocked during YAML generation because the migration checklist validates required fields and strict layering, ensuring deployment safety and proper registration alignment.

Related Skills

Explore More Software Engineering Skills→
AscendAscend
31

mindspeed-fsdp2-model-migration

Migrates models to MindSpeed-MM FSDP2 preserving source semantics via @model_register and signature checks.

Official
Advanced
vikrant-projectvikrant-project
2

pytorch-fsdp

Implement and optimize PyTorch FSDP for distributed training.

Community
Advanced
distribucionesfayosmoreno-sysdistribucionesfayosmoreno-sys

ml-pipeline-workflow

Orchestrate end-to-end MLOps pipelines from data preparation to deployment.

Community
Advanced
NVIDIA-NeMoNVIDIA-NeMo
852

mlm-bridge-training

Run, compare, and translate Megatron-LM and Bridge training configurations.

Official
Intermediate