How hard is it to get a llama fine-tuning pipeline (with [qlora](https://github.com/artidoro/qlora)) on EC2 using reasonable resources? Take script from #2 and implement on bare EC2.
How hard is it to get a llama fine-tuning pipeline (with qlora) on EC2 using reasonable resources?
Take script from #2 and implement on bare EC2.