-
**What would you like to be added/modified**:
Based on existing datasets, the issue aims to build a benchmark for domain-specific large models on KubeEdge-Ianvs. Namely, it aims to help all Edge AI a…
-
Just today, [Intel's CEO said](https://www.theregister.com/2024/06/05/ai_pc_confusion/):
> "AI PC benchmarks are needed. We don't have proper comparisons yet."
Just marking my interest in helpin…
-
The draft benchmark version looks good, which is a nice start point. Thanks for the great efforts!
Issues: There are too many metrics, which lead to two problems:
> - Difficult for readers…
-
In a future version, I suggest to add “leaderboard/forum” for academic and industrial users to publish their latest progresses and discuss feedbacks/questions/concerns about data, metrics, benchma…
-
Would it be possible to make the maps site a tarbell for running locally instead of having to use the AMI image?
-
The map site seems to be very fragile. I host it in AWS. However, the site becomes non-responsive after a few examples.
-
# Interesting papers
- NASA JPL의 ROSA - Robot Operating System Agent [링크](https://github.com/nasa-jpl/rosa)
- LLM을 이용해 ROS1/2 기반의 로봇들과 소통 가능
https://github.com/user-attachments/assets/961db…
-
Hello
is the Yolov10 of size **m** for object detection also officially supported?
I've seen from the benchmark/evaluation that the Yolov10 of size **n** and **s** are supported.
https://github.com…
-
Hi! When running an old AWQ quant (for example deepseek 1.3B coder), it works fine. However, when trying to run `casperhansen/mistral-instruct-v0.2-gemvfast-awq`:
```sh
❯ python examples/benchmark…
-
Hi,
I'm using "selfie_multiclass_256x256" and "deeplab_v3" models for background removal.
The "selfie multiclass 256x256" model works very well. But the "deeplabv3" model does not work well.
D…