-
Moe workload generated by AICB using the following command cannot be parsed:
```bash
sh scripts/megatron_gpt.sh \
--nnodes 1 --node_rank 0 --nproc_per_node 8 --master_addr localhost --master_port 2…
-
I converted my model to NCNN using the following command:
```bash
!yolo export model=jameslahm/yolov10n format=ncnn
```
The output was:
```
...
############# pass_level1
############# pass_l…
-
### Which OpenObserve functionalities are the source of the bug?
metrics
### Is this a regression?
Yes
### Description
hello, while trying to run the simple query
`topk(5, count by (__name__)(…
-
A newly released paper, [SparQ Attention: Bandwidth-Efficient LLM Inference](https://arxiv.org/pdf/2312.04985.pdf), suggests a method for increasing the inference throughput of LLMs up to 8x by reduci…
-
Hi everyone
I was able to successfully run the code on Visium HD data. The output has block#, X, Y,K and P.
1. How do I merge this into the corresponding 2um bin in the original adata object, s…
-
**描述这个 bug**
对 bug 作一个清晰简明的描述。
DCN在mvlen100k下uni100与full的AUC差距过大, uni100时auc仅0.58左右，而full时auc为0.78左右，请问原因？谢谢
**如何复现**
复现这个 bug 的步骤：
1. 您引入的额外 yaml 文件
```
load_col:
inter: ['user_id', '…
-
https://github.com/llm-attacks/llm-attacks/blob/355b2ad245f8723b0f78e04d421eb8d13c90def8/llm_attacks/base/attack_manager.py#L622-L638
This means that if `control_cand.shape[0]` is `1` and not `de…
-
We should add what is called elsewhere `topk` or `maxk`, e.g. https://pytorch.org/docs/stable/generated/torch.topk.html
In Base we have `partialsort` and `partialsortperm` doing something similar …
-
Hi,
I attempted to add `topk` support, but for some reason, I am getting the following notorious error:
```
AttributeError: 'torch.return_types.topk' object has no attribute '_trt'
```
Al…
-
感谢您参与 X2Paddle 社区！ 问题模版为了 X2Paddle 能更好的迭代，例如新功能发布、 RoadMaps 和错误跟踪. :smile_cat:
## 问题描述
- 错误信息
- 错误截图
## 具体信息
- 转换模型后用处
- [ ] 使用 Paddle 框架/ PaddleInference 推理预测
- [ ] 使用 Paddle-L…