GrumblePuss

This is already a feature as listed in the notes from the dev. When using the plugin, you can specify 768x768 as the resolution.

This may already be available in some generators. If it's not available in your favorite one, you can simply fork the generator and modify the code to pass along the higher resolution request.

P.S. There is a feedback form at the bottom of the notes page. What you post there goes straight to the dev.

About our participation in the matter: only the dev can apply new training data. However, I'm sure feedback is greatly appreciated.

It is possible (and quite standard) to add additional training data to a model after deployment. It takes a long time, and a considerable amount of GPU power. It is also the only way we are likely to get good celebrity support again, because the vast majority of models are now shipped without training on celebrity data sets anymore.

Because AI text gen is not just spitting out text from some lookup table: it's looking through billions of chunks of text, looking for patterns, then combining them together to form a response. There is a an absolutely massive amount of number crunching that goes into the process. In addition, your want for it to "remember" things is an extremely tall order because, AI doesn't remember anything.

Every query you send (this is true for all AI models) includes:

• role instructions
• previous messages
• "memories" (blocks of text with summarized context)
• your query

This is always sent as one big blob of text that the AI has to read, parse, process, group, compare, quantize, then finally generate a response. The more data you send, the more it has to process, the more resources are required.

First, you will need a powerful PC and GPU. AI is very resource hungry and takes a lot of processing power to be very good. This is especially true if you want it to "remember" things.

I wasn't insulting, I didn't call names, and I didn't cuss. I was blunt and called attention to some flaws in their argument.

You immediately lost your argument when you stated that you are using the exact same prompts. The notes about the new model clearly state that you will need to adjust your prompts for the new model. There have also been numerous examples, both here and in the Discord server of people getting extremely consistent results by simply prompting correctly. The old model isn't coming back, get over it. Adapt and learn to use the new model. Or, at least put as much effort into learning the new model as you have put into complaining.

What do mean? Which part is the bad omen?

Perhaps it's two celebrities in the same prompt? The assumption may be that if they are both there, then you may be depicting those famous people as doing something together, which could be a big red flag.

The training is happening on the server side, by the dev. Many model creators refuse to use celebrities and famous people as part of the training data (with a common exception for public figures) for their models. However, there is no rule about creating databases of these images and then using them to train models that you host. That's what the training is for, to fill in the gaps that are in the model.

Refer to these notes if you want some more info.

What you are referring to is known as "training". Training is the process of showing the AI model thousands upon thousands of images. These images come from curated datasets. Each image is accompanied by some text that describes in clear terms exactly what is in each image. The AI model analyzes each image, compares the image to the text, then stores the data for later use. The more images and text, the more the model will understand. The more the model understands, the better it will be able to use the data it has.

Currently, the model is in training. As training continues you will see the image generator steadily improve until it understands what the old model did.