-
We need some way to keep historical (kind of like a lab-book), JMH results and documentation "did X does not work".
ktoso updated
8 years ago
-
I'm seeing some strange performance issues with the GROMACS test on our system. I.e. occasionally, it just runs 10 times slower. Looking at `htop`, I see individual cores not being used - even though …
-
It would be nice to mark the benchmark logs as artifacts.
We could then write a script that would use GitHub CLI tool to download the artifacts of main branch and plot the benchmarks.
-
Hi, @haesleinhuepf recommended I post this here.
Because you guys are making this evaluation benchmark public, couldn't LLMs use this as training data and therefore overfit to it?
So, it will be h…
-
Hi!
For testing purposes, I would like to try C/R xHPL with cuda-checkpoint. However, in 99% it fails on the first step when `cuda-checkpoint --pid ... --action lock` is called.
For testing I u…
-
Via:
```java
static {
Json.mapper.registerModule(new AfterburnerModule());
Json.prettyMapper.registerModule(new AfterburnerModule());
}
```
and:
```java
.end(Json.encodeToBuff…
-
官方demo是通过对一个数据集文件按照一定划分出训练、验证、测试集。但是我的训练,验证和测试集是划分好的。已经单独保存了三个`.inter`文件。这种情况下如何生成训练、验证和测试的dataset以及dataloader呢?实在是找不到相关的demo或者api
-
During the process of executing the command, we will call the dictFind function at least twice to find the current key, once for lookupkey and once for dbset. If the key is relatively long, calling di…
kukey updated
2 months ago
-
Let's use this issue to gather instructions on how to profile one's CPUNVMe setup.
(@tjruwase and I have been editing this post)
You need to do this on every new CPU/NVMe setup in order to confi…
-
### Run Information
Name | Value
-- | --
Architecture | x64
OS | ubuntu 22.04
Queue | ViperUbuntu
Baseline | [1b5c48dc5958e20b4aa0f4cbfc21fddb8f81052c](https://github.com/dotnet/runtime/co…