What is the GPU mem requirement for finetuning?

Alpha-VLLM / Lumina-T2X

Lumina-T2X is a unified framework for Text to Any Modality Generation

MIT License

1.82k stars 74 forks source link

Closed ninjasaid2k closed 1 week ago

zhuole1025 commented 1 week ago

We now support single gpu training now. However, using one GPU for full fine-tuning can be slow.

nunbuzor commented 1 week ago

We now support single gpu training now. However, using one GPU for full fine-tuning can be slow.

can dreambooth finetuning be done on a consumer GPU?