​

distributed-llm-pretraining-torchtitan

Provides PyTorch-native distributed LLM pretraining using torchtitan with 4D parallelism (FSDP2, TP, PP, CP). Use when pretraining Llama 3.1, DeepSeek V3, or custom models at scale from 8 to 512+ GPUs with Float8, torch.compile, and distributed checkpointing.

​

Install

npx research-agent-skills install distributed-llm-pretraining-torchtitan
npx skills add KalarisLabs/research-agent-skills --skill distributed-llm-pretraining-torchtitan

​

When to use it

​

Full playbook