https://github.com/trrahul/llama2.cs

Inference Llama 2 in one file of pure C#
https://github.com/trrahul/llama2.cs

csharp llama llama2

Last synced: 29 days ago
JSON representation

Inference Llama 2 in one file of pure C#

• Host: GitHub
• URL: https://github.com/trrahul/llama2.cs
• Owner: trrahul
• License: mit
• Created: 2023-08-07T12:40:12.000Z (over 1 year ago)
• Default Branch: main
• Last Pushed: 2024-02-06T07:04:17.000Z (11 months ago)
• Last Synced: 2024-10-18T21:59:20.977Z (2 months ago)
• Topics: csharp, llama, llama2
• Language: C#
• Homepage:
• Size: 23.4 KB
• Stars: 92
• Watchers: 10
• Forks: 15
• Open Issues: 1
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# Andrej Karpathy's llama2.c in one file of pure C#.

[llama2.c](https://github.com/karpathy/llama2.c) is a very simple implementation

to run inference of models with a [Llama2](https://arxiv.org/pdf/2302.13971.pdf)-like transformer-based LLM architecture.  

This is a pure C# implementation of the same thing. It is optimized for speed and very simple to understand and modify.

## Usage

Requires .net7 or higher.

1. First put the stories15M.bin file in the same directory as the executable. You can download it from [here](https://huggingface.co/karpathy/tinyllamas/resolve/main/stories15M.bin)

2. Get tokenizer from [here](https://github.com/karpathy/llama2.c/blob/master/tokenizer.bin) and put it in the same directory as the executable.

```

dotnet build -c Release

```

### Generate a random story

```

.\bin\Release\net7.0\llama2.cs.exe stories15M.bin

```

![WindowsTerminal_LrRTW3joph](https://github.com/trrahul/llama2.cs/assets/7353840/3b469a99-b83a-43f1-b07d-227da7b9ebe0)

### Generate a random story with a given prompt

```

.\bin\Release\net7.0\llama2.cs.exe stories15M.bin  -i "A long time ago a"

```

### TODO

 - [ ] Inference with Llama2 checkpoints

 - [ ] Use high performance C# types from .net8?

 - [ ] Add training functionality