-
### 🐛 Describe the bug
Error with attention mask in torch.nn.TransformerEncoderLayer
Easy example:
```
if __name__=="__main__":
import torch
model = torch.nn.TransformerEncoderLaye…
-
### 🐛 Describe the bug
```python
import torch
import torch.nn as nn
import torch.nn.functional as F
device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
class Attention(nn.M…
-
Hi
Does the latest docker container update produce the following error? any idea?
outline | Please migrate your code to use AWS SDK for JavaScript (v3).
outline | For …
-
https://github.com/Echo0125/Memory-and-Anticipation-Transformer/blob/main/src/rekognition_online_action_detection/models/transformer/multihead_attention.py
DotProductAttention Class is different fr…
-
Hi! I'm [reviewing your paper](https://github.com/openjournals/joss-reviews/issues/6175) for JOSS. This issue summarizes my (minor) comments on the paper, which I find to be well written and informati…
-
Hello, I have problem to install FBGEMM-gpu for torchREC
Initially I tried to install using pip but not successful.
```
(torchrec) ➜ torchrec git:(main) ✗ pip install fbgemm-gpu --index-url http…
-
### 🐛 Describe the bug
I'm not sure if the title of this issue is right, so please change this to whatever you think is most appropriate... in short, I couldn't decipher the cause of the problem :s…
-
### 文档中英文链接
英文文档:https://www.paddlepaddle.org.cn/documentation/docs/en/develop/api/paddle/bmm_en.html#bmm
中文文档:https://www.paddlepaddle.org.cn/documentation/docs/zh/develop/api/paddle/bmm_cn.html#…
-
The optional `tilize`/`untilize` arguments input/output activations can be removed, and can directly infer whether to tilize/untilize based on the input/output activations tensor formats. This will si…
-
Changing the Op from...
```c++
struct MorehOp {
void validate(const std::vector &input_tensors) const;
std::vector compute_output_shapes(const std::vector &input_tensors) const;
std::…