About finetuning on pretrained LLM models.

FoundationVision / LlamaGen

Autoregressive Model Beats Diffusion: 🦙 Llama for Scalable Image Generation

https://arxiv.org/abs/2406.06525

MIT License

1.36k stars 56 forks source link

About finetuning on pretrained LLM models. #73

Open ZeyuLing opened 3 weeks ago

ZeyuLing commented 3 weeks ago

Thanks for your pioneering work! We are interested about whether you tried to finetune a pretrained model like LLaMA-3.1/3.2 rather than train from scratch? If so, how's the results like? Thanks a lot!