-
The vLLM [fused moe kernel](https://github.com/vllm-project/vllm/blob/main/vllm/model_executor/layers/fused_moe.py) used for Mixtral uses the standard data parallel parallelization which works well wi…
-
### 🐛 Describe the bug
(I'll add actual benchmarking details and logs and output_code.py in a bit)
I'm doing min_sum and mul_sum in two setups:
1. (D, ) x (D, ) -> scalar
2. (B, N, 1, D) x (B,…
-
Thank you for your nice work.
Since the code is not yet open, I write down my question for your kernel design.
In the paper, given the image feature information, you set this feature as the convol…
-
Thanks for sharing! Just found out `Attention.get_att_weight` is calculating attention in a for-loop? this looks rather slow isn't it?
`4-2.Seq2Seq(Attention)/Seq2Seq(Attention).ipynb`
```pyth…
-
**Describe the bug**
Hello, After clicking on "Outpaint" in the screenshot below I get the following error:
![image](https://user-images.githubusercontent.com/4301170/197384390-e9e8672e-db9e-48b…
-
开发计划可参考以下节点:
1. 方案撰写,xx.xx~xx.xx
2. 开发自测,xx.xx~xx.xx
3. 提出 PR/MR,xx.xx~xx.xx
4. review( 3个赞),xx.xx~xx.xx
6. maintainer 合入
-
您代码里面的解码的前向传播没怎么看懂
`< def forward(self, inputs, init_state, contexts):
if not self.config.global_emb:
embs = self.embedding(inputs)
outputs, state, attns = [], i…
-
hi @xcmyz
after successful run of preprocess.py
when i run train.py it gives following error
```
Use FastSpeech
Model Has Been Defined
Number of TTS Parameters: 25367169
Load data to buffer
…
-
Hello,
I am using a U-Net augmentation (specifically: https://github.com/juntang-zhuang/LadderNet) to perform segmentation of hands. To be specific, I am classifying each pixel of an image to one o…
-
### 🐛 Describe the bug
From https://discuss.pytorch.org/t/torch-function-runtime-dependent-on-scipy-call/139483:
I've noticed that certain PyTorch functions run slower when I make calls to `scipy.…