Team, if you would like to help us in our AI efforts, you can contribute to our data testing by testing! Purpose of this AI Effort and Image Intents Model? This is a custom built model, built off of the renowned Roberta model, and is designed and trained against a dataset of thousands of text messages (publicly available, non SoMee users, consent granted) to help identify if a request is an image intent or not an image intent. What we'll use this model for? Ultimately, we'll use it to extend our chat capabilities for Chat Sprinkle and dedicated AI bots (others can use it as well). We'll be able to integrate this model into our AI flow, so that when you are chatting with the bots, if you generate an image intent, the bot will know this, and generate an image for you :) How to participate 1. Visit: https://huggingface.co/spaces/smeintadmin/image_intents 2. Enter in a message into the field and click submit; it'll let you know if it is, or is not an image intent. 3. If it did it right, just add more messages, keep testing :) 3a. If it did it WRONG, though, and it SHOULD have been an image intent given the context of your message, then go to our support channel: https://t.me/+s7Q9W_gX9G45MDEx, enter in to the #train channel and paste your message and result there, so that we can add it to our fine tuning dataset. Once we have enough messages to fine-tine more, we'll fine tune the model again. Now's your time to actually help out and train AI models with SoMee and Chat Sprinkle :)