-
### 1. Issue or feature description
只要带nvidia.com/gpu相关配置启动pod,
然后进入bash, 输入nvidia-smi ,就会报 Segmentation fault
### 2. Steps to reproduce the issue
1. 使用 下面命令安装hami
helm install hami hami-charts/h…
hxh71 updated
1 month ago
-
### Short description
One of our clients reported this issue.
### Steps to reproduce
1. Open GPU Instance(g2.2xlarge) on AWS Marketplace AMI
2. Add broadcast object, add 2 adaptive options
3. …
-
Make sure we add GPU instance support for AWS deployments. This is a tracker issue for various pieces of this problem, and based on experiences with astroML demo prep.
Todo:
[ ] Start the GPU nod…
-
The job fails with this error :
```
tensorflow.python.framework.errors_impl.InvalidArgumentError: No OpKernel was registered to support Op 'NcclAllReduce' used by node allreduce/allreduce/NcclAllRe…
-
**Problem**
Larger (especially ML/AL) workloads require access to GPUs for parallel processing. During summer school 2023, GPU instances were made available in the platform backed by Anvil and JetStr…
fbaig updated
2 months ago
-
### Checklist
- [X] 1. If the issue you raised is not a feature but a question, please raise a discussion at https://github.com/sgl-project/sglang/discussions/new/choose Otherwise, it will be closed.…
-
### Already reported ? *
- [X] I have searched the existing open and closed issues.
### Regression?
No
### System Info and Version
irrelevant - not a software bug - documentation bug
### Descrip…
-
This is an issue discovered in fluidstack implementation #3086. In fluidstack, the instances are heterogeneous across different regions, for example:
| region | acc | acc_count | vCPUs | instance_t…
-
**Description**
Triton does not clear or release GPU memory when there is a pause in inference. In the attached diagrams the same model is being used. It is served via ONNX.
![image (1)](https:…
-
I setup a brand new camera on Scrypted, added it to to Home Assistant to use the go2rtc camera integration with the custom instance (for a GPU) and for this camera only I get a cryptic error which doe…