feng wang
feng wang
It will be a good idea to deploy this project with a docker container.
std::span is introduced in C++20, and std::mdspan is in C++23. It would be good to have them enabled in cereal.
It might be good to provide a dockerfile for easy deployment.
Baidu webdisk is unbelievably slow outside China. Is it possible to host the training set somewhere else?
Synchronized with upstream. Sorry for the previous pull request.
Hello, I went through your post 'Implementing Spatial Batch / Instance / Layer Normalization in Tensorflow'. In that post, it has been pointed out that > Batch Norm → Take...
### Trilium Version 0.60.4 ### What operating system are you using? Other Linux ### What is your setup? Server access only ### Operating System Version Archlinux ### Description 1. Create...
Steps to reproduce 1. Install actions-runner-linux-x64-2.299.2 in Archlinux/Ubuntu1804, setup token and service 2. Create an action script in folder `.github/workflows` including ```yaml name: CI on: push: branches: [ "main", "workflow"...
### Contact Details _No response_ ### What happened? I tested localscore on my home server, the test result is at https://www.localscore.ai/result/417. This server has two E5-2699 v3 CPUs installed, but...
I serialized my LLM model with cereal, the serialized model size starting from 3GB, and the weight loading time does matter to the performance. With this test program (https://raw.githubusercontent.com/lemire/Code-used-on-Daniel-Lemire-s-blog/refs/heads/master/2012/06/26/ioaccess.cpp), I...