im stuck between text-to-image & text-generation
text-generations using huggingface llms is hard AF!
way harder than images, with images i could just pass the entire prompt to the llm while on text prompts there are weights & T & P values that have some sort of effect… meanwhile im unsure if im even passing the arguments of the prompts properly as each text-gen llm is a different story!
its literally trial & error
Leave a Reply
You must be logged in to post a comment.