-
# Failing Tests
> Please see the failing tests divided into sections below. Click on each section to expand. Feel free to get assigned to an issue by following the instructions [here](https://unify.ai…
-
Currently the brms model returns a summary that include the internal terms such as constant mu, kappa2, etc
![image](https://github.com/venpopov/bmm/assets/12783882/401950a0-ca13-4e2d-9ad8-8dc28f88…
-
Hi, Ali, I have been looking forward to implementing Attenion like Convolution. NATTEN is a very nice implementation.
[Naive PyTorch implementation](https://github.com/qwopqwop200/Neighborhood-Atte…
-
right now there are two copies of the website, one based on master (the default domain) and one based on develop (https://venpopov.github.io/bmm/dev). the developer notes are also different then, but…
-
### System Info
transformers 4.41.0
torch 2.3.0
GPU: NVIDIA GeForce RTX 4090, CUDA version 12.3
### Who can help?
_No response_
### Information
- [ ] The official example scripts
- [X] My own m…
-
Hi,
It mentioned in the blog that you were able to improve performance by 2x with bitdelta with punica.
Are you planning to release that implementation too?
https://www.together.ai/blog/bitde…
-
How should I fix this?
```
(phosa) [jalal@goku phosa]$ python demo.py --filename input/dark_bat.jpg --class_name bat
2021-03-26 16:55:48,497 INFO Calling with args: Namespace(class_name='bat', …
-
### 🐛 Describe the bug
Exporting a model with AOTInductor works when I don't specify `torch.set_float32_matmul_precision('high')`. But when I do set this setting to make full use of tensor cores, I…
-
### 🐛 Describe the bug
Internal xref: https://fb.workplace.com/groups/6829516587176185/posts/6917011115093398
Repro:
```
import torch
from torch.nn.attention import sdpa_kernel, SDPBackend
f…
-
Dear @i404788 ,
do you maybe have an S5 model that works with irregularly sampled data, I mean it can be trained on it, like in pendulum task?
If yes, could you share this branch?
Thanks.