Is n8n capable of handling the entire workflow for a SOTA text-to-image model, from ingesting a 1,000-pair dataset for fine tuning to running subsequent inference for image generation?
Yeah, totally. You can fine-tune the Flux model using Replicate and an HTTP Node in your workflow. But honestly, if you’re only doing it one time, it’s way faster and less hassle to just use the Replicate website form directly. If you plan on doing this over and over, that’s when you should definitely automate it with the HTTP Node.
Hope this helps
3 Likes
This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.