RandomGitUser321

I've actually rethought it a little and it should likely have some extra separate tags like "deblur(promax), inpainting(promax), outpainting(promax)" and so on. Not everyone will use the promax version of...

@yiyixuxu I think it's about done for real this time, let me know if there's anything else. Check my original post for more details on the changes. Added in cfg...

> thanks! I think this PR should focus on adding the new callback and cfg_rescale, I think we should not include these changes introduced for `encode_prompt` in this PR ~~I...

@yiyixuxu Alright, I reworked the legacy callback/callback_steps back in. ~~The legacy `callback` will require `(self, step_idx, t, latents)`, but has 1:1 parity with the newer `callback_on_step_end` method~~. I also included...

@yiyixuxu @sayakpaul I reverted the negative_prompt changes and fully updated my original post to be more clear about the changes.

Since the legacy implementation of callback doesn't appear to have the ability to interrupt, I'm just going to roll this back. If people want to interrupt, they should be using...

It's already there for when it's ksampling. You get current step and total steps through `message['type'] == 'progress':` and the entries will be called `value` and `max` In the websocket...

within the `@routes.get("/queue")` API call in server.py: ```python @routes.get("/queue") async def get_queue(request): queue_info = {} current_queue = self.prompt_queue.get_current_queue() queue_info['queue_running'] = current_queue[0] queue_info['queue_pending'] = current_queue[1] return web.json_response(queue_info) ``` If you add...