Jack

> > > A responsive free edit UI probably requires a pure front-end solution instead of gradio solution we have now. > > > > > > Hi guys! I...

> SDXL ipadapter uses clip-g as preprocessor which is slower than clip-h. You can try switch to `ip-adapter_sdxl_vit-h` which uses clip-h as preprocessor. Besides I found clip would init everytime....

> > > SDXL ipadapter uses clip-g as preprocessor which is slower than clip-h. You can try switch to `ip-adapter_sdxl_vit-h` which uses clip-h as preprocessor. > > > > >...

> @nhannguyen1003 maybe paddlepaddle2.5 or 2.6 请问paddleocr 支持batch 推理吗？

> Nice work! I have checked it. Thanks for your contribution. I feel this Euler based sampler can also be used in SD3 version. Wether Euler sampler is good enough...

> @JettHu Yes, our 1-step unified lora is able to inference on 1-8 steps with the help of TCDScheduler within single checkpoint! And we found previous implementation of it behaves...

> i think i have solved this issue. I found it strange that it had to be something with models in memory getting corrupt, and after reloading it was ok....

> > Found this while searching. How did this thread morph from talking about int8 to talking about fp8? > > Also, would it be possible leverage the gguf format...

> note, most of this wont be possible until the next paper; as only simulated quantization is implemented. no quantization; no speedup/optimizations, only proof of concept > > probably will...

The same error as blow. Is there any good way to solve this? > torch._dynamo.exc.Unsupported: Tensor.item > > from user code: > File "/home/admin/miniconda3/envs/diffusers/lib/python3.10/site-packages/DeepCache/extension/deepcache.py", line 41, in wrapped_forward > self.cur_timestep...