-
是不是在moe上运行效果会好点呢，另外请问moe...config文件在哪
-
### Describe the issue
MoE unit tests fail on older architecture.
The tests have a particular requirement. If that requirement is not met it is pointless to run the tests.
### Urgency
_No response…
-
### Description
If a message is edited before Discord embeds the contents of the link, when it finally embeds, it will replace the entire message contents with the message contents of when the link w…
-
软件版本：1.5.6
地址：https://www.bilibili.com/video/BV1iW411c7gL
恳请大佬研究研究，万分感谢
https://cdn-us.imgs.moe/2023/02/09/63e461fd1055b.png
[https://cdn-us.imgs.moe/2023/02/09/63e461fd1055b.png](url)
-
I have successfully used run_eval.py on a few of the long sequence forecasting data set, and the results look promising. From previous posts, I can see how a univariate sequence can be applied using t…
-
RUZAC updated
9 years ago
-
```
What steps will reproduce the problem?
I just tried running a 'Moe magic' migration from a git repository which
had 126 commits to migrate. However, due to branching history, Moe
apparently neede…
-
Thanks for the excellent work. Following the comment in #59, I am trying to train `dmoe_760m` using 16 GPUs (2 nodes) by changing distributed arguments to set up for two nodes but it is very slow in t…
-
```
[rank0]: Traceback (most recent call last):
[rank0]: File "Pai-Megatron-Patch-0925/toolkits/model_checkpoints_convertor/qwen/hf2mcore_qwen2_dense_and_moe_gqa.py", line 924, in
[rank0]: m…
-
我看到历史issue中有提到如果使用了moe模块的话，无法导出onnx。但是我使用wenet/bin/export_onnx_gpu.py是可以正常导出encoder.onnx 和 decoder.onnx的，而且cer跟torch checkpoint是对齐的。想请教下是有什么坑我没注意到么？还是目前已经支持moe的onnx导出了？