![]() ![]() Usernames that are against the rules, NSFW, hard to mention, or containing special characters will be changed without your concern, no fonts, no uni-code.ĭon't ask for any roles, you can take self assigned roles. Profiles with inappropriate avatars and usernames are not allowed. These are in some cases against the discord TOS and if you need a bot for something use one of the bots already in the server. No advertising/linking other Discord servers. Contribute to BIRD-UP73/nsfw-bot development by creating an account on GitHub. No unsolicited DMs or ads to other members. You may help other members, but do not falsely imply that you have a role or other position of authority. Do not attempt to impersonate/roleplay.ĭon't try to fake your identity, gender, or impersonate other users by changing usernames or by any other possible way. Please don't forget that it's a public server and many of our members are younger! No pornographic/adult/other NSFW material.Ĥ. ![]() ĭon't be toxic & racist, No hate speech regarding others, No cursing, vulgar, abusive language and adult conversation are strictly prohibited. Keep all your conversations in English/Hindi and only English characters are permitted. ![]() Put the right content into the right channel. Spam does include excessive emoji use, excessive caps spam/all-caps, excessive Discord formatting, messages spanning multiple lines that are designed to interrupt or derail conversation, posting the same message across multiple channels, and inciting spam. No sexual comments/harassment, pornography, gore/death/extreme violence, content related to self-harm or harming others, threats/bullying or animal cruelty. Hate speech of any kind will not be tolerated, regardless of religion, sexuality, gender etc. Thanks to Stable Cascade’s modular approach, the expected VRAM requirements for inference can be kept to approximately 20gb but can be further lowered by using the smaller variants (as mentioned before, this may also decrease the final output quality).ĭuring our evaluations, we found Stable Cascade performs best in both prompt alignment and aesthetic quality in almost all model comparisons.Treat everyone with respect. For Stage B, both achieve great results, however, the 1.5 billion excels at reconstructing fine details. However, the 1B parameter version can be used for those who want to focus on the lowest hardware requirements. It is recommended to use the 3.6B model for Stage C as this model has the highest quality outputs. Stages C & B will be released with two different models: 1B & 3.6B parameters for Stage C and 700M & 1.5B parameters for Stage B. For most uses, it will provide minimal additional benefit & we suggest simply training Stage C and using Stages A and B in their original state. Stages A and B can optionally be finetuned for additional control, but this would be comparable to finetuning the VAE in a Stable Diffusion model. If you want to use nudefrom command you have to register an app on reddit then replace the redditappclientid with your apps clientid. Add a prefix for the commands for example. This comes with a 16x cost reduction compared to training a similar-sized Stable Diffusion model (as shown in the original paper). First you have to create an app at the discord developer portal then you have to create a bot and paste its token in the configs.json file. The Latent Generator phase, Stage C, transforms the user inputs into compact 24x24 latents that are passed along to the Latent Decoder phase (Stages A & B), which is used to compress images, similar to what the job of the VAE is in Stable Diffusion, but achieving much higher compression.īy decoupling the text-conditional generation (Stage C) from the decoding to the high-resolution pixel space (Stage A & B), we can allow additional training or finetunes, including ControlNets and LoRAs to be completed singularly on Stage C. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |