-
ERROR: Cannot install -r D:\Comfyui\ComfyUI-aki-v1.2\ComfyUI-aki-v1.2\custom_nodes\ComfyUI_LayerStyle\requirements.txt (line 36), fairscale>=0.4.4 and torchscale==0.3.0 because these package versions …
-
# ComfyUI Error Report
## Error Details
- **Node Type:** DeepFakeDefender_Loader
- **Exception Type:** RuntimeError
- **Exception Message:** Unknown model (convnextv2_tiny.fcmae_ft_in1k)
## Stack…
-
**Describe the bug**
Trying to install Kosmos-G, there's a disclaimer that says:
> Make sure you have Pytorch 1.13.0 installed.
However on
```shell
bash vl_setup.sh
```
We got
```she…
-
Hello,
I have followed the training configuration introduced here (https://github.com/microsoft/torchscale/issues/52) with retnet_medium architecture. I have some questions that I would appreciate …
-
Thanks for a great work!
I am just wondering what would happen if we had higher D. This is because the RetNet configs that you can obtain from the torchscale (and also mine) have typically `D=128` …
-
The fix of normalization Rnm is totally wrong.
The added max value in clam needed because of wrong placement of abs() operation.
More thorough explanation I put here:
https://github.com/microsoft/t…
-
Hi! Congrats on the clean RetNet code. I'm experimenting with the torchscale codebase and happened to find your repo with a link to a checkpoint in HF.
I noticed it is now gone, do you have plans t…
-
![image](https://github.com/syncdoth/RetNet/assets/902005/8eef7829-88ae-49e1-a65f-cd882268e688)
Trying to compare with other transformer architectures. But as soon as the training starts, the gradi…
-
Hi,
I found that the model used in BIET-3 based on torchscale is not as what the paper described.
In the multiway transformer, the self-attention layer should be shared across different modality. …
-
```
image-reward 1.5 requires fairscale==0.4.13, but you have fairscale 0.4.0 which is incompatible.
torchscale 0.3.0 requires fairscale==0.4.0, but you have fairscale 0.4.13 which is incompatible.
…