-
I use `SD_METAL=ON` to build `sd.cpp` on m1. However, the generated image seems to be blured or tends to be in the cartoon-like style. This happens on all models (v2, SDXL, SDXL turbo). This problem o…
Ucag updated
6 months ago
-
https://github.com/openai/guided-diffusion/issues/106
-
-
Hi Eley,
In your RAL paper, you mentioned that "The training dataset consists of 376 human-human demonstrations (179,993 environment interactions) on the collaborative carrying task collected by 5 …
-
RX7900XTX on 23.1.2
I tried with Waifu Diffusion this time, i get the same error with OrangeMix Abyss2
```
vulkan devices are available.
cuda devices are not available.
Running on local URL: …
-
You might want to loosen this version dependency because AFAIK older fastai versions rely on older pytorch versions that did not support ARM/M1 chips at all, so fastai tries to install with CUDA suppo…
-
I'm always get black images. All Filters, img2img and txt2img
many warning outputs
System is Win11 and stock RX5700
![image](https://user-images.githubusercontent.com/127530420/232443179-878c2b69…
-
I did everything well except when I run lcd train_hulc
I can not find the right 12_all_trajectories.pt in https://github.com/ezhang7423/hulc-data.git too. The dir "hulc-trajectories" is empty.
…
-
Hi~, the description of realworld data shown in the paper and the data downloaded from [https://xskill.cs.columbia.edu/data/real_kitchen_data.zip](url) seems not match , could you help me, thank you…
-
## Enviornment
* CUDA: 11.6
* Python: 3.8.10
* PyTorch: 1.13.1
* DeepSpeed: 0.8.1
* Transformers: 4.26.1
* Diffusers: 0.13.0
* GPU: T4
## Code
```python
"""
Simple script to benchmar…