domains / families / science / nvflare-convert-huggingface
nvflare-convert-huggingface
separate
Convert existing Hugging Face Transformers Trainer or TRL SFTTrainer training code into an NVFLARE federated job using flare.patch(trainer), local validation, and job export; use when the user names Hugging Face or preliminary source inspection identifies one Hugging Face owner, and not for manual PyTorch loops, Lightning, inference-only pipelines, deployment, or experiment workflows.
When to use
- Converting transformers.Trainer code
- Converting Seq2SeqTrainer code
- Converting TRL SFTTrainer code
- Converting other Trainer subclasses
- Full-model fine-tuning conversion
- PEFT/LoRA fine-tuning conversion
- Replicated torch.distributed training conversion
When not to use
- Do not use for an `AutoModel` driven by a manual PyTorch loop without a Hugging Face Trainer (route to `nvflare-convert-pytorch`), PyTorch Lightning (route to `nvflare-convert-lightning`, including Lightning modules that contain Transforme…
Skill page
Source
| Repository | Path |
|---|
| NVIDIA/skills | skills/nvflare-convert-huggingface/SKILL.md |