You are in phase 0 of LORA creation (Planning), which requires the least effort and is accessible to everyone. However, there are a few things to consider before requesting a model.
In the training category under "Request your LORA.”
Are you sure that current AI models are unable to do what you have proposed?
Sometimes, a concept/character that is trained in a LORA was already understood by the previous AI model and did not require a LORA. You can do a small test by generating an image and adding the relevant tags if you're not sure.
Are you sure that what you want to train doesn't already exist in Civitai, Hugging Face, or another platform where people upload their LORAs?
It's quite obvious, but it's advisable not to create a model with an existing concept in Civitai as it contributes to creating redundant models, unless you want to contribute something new or improve an existing one.
Is that LORA useful or necessary for the community in general, or is it just a whim for yourself?
There's nothing wrong with requesting strange or less desired LORAs, but it's useful to prioritize those that you know will have high demand.
Does the request comply with the limits of our training lab?
In the documentation section, you will find the limits. Make sure to read them to avoid future moderation issues.
<aside> <img src="/icons/send_blue.svg" alt="/icons/send_blue.svg" width="40px" /> Remember that to contribute datasets to the community, you must have requested the "Dataset Creator" role in the chat. Otherwise, you won't be able to contribute datasets. To get the role, simply follow the steps in the
General Guide to Creating Datasets and Tagging LORAs in Spanish
</aside>