Top AI Repos — open-source AI, indexed and scored
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Go with your own intelligence - Write Go applications that directly integrate llama.cpp for local inference using hardware acceleration on Linux, macOS, Windows, & WebAssembly.
| Date | Stars |
|---|---|
| 2026-07-30 | 533 |
| 2026-08-06 | 533 |
| 2026-08-07 | 534 |
| 2026-08-15 | 552 |
| 2026-08-18 | 559 |
| 2026-08-20 | 561 |
| 2026-08-22 | 564 |
| 2026-08-24 | 566 |
| 2026-08-25 | 566 |
| 2026-08-26 | 567 |
| 2026-08-27 | 569 |
| 2026-08-28 | 570 |
| 2026-08-29 | 574 |
| 2026-08-30 | 574 |
| 2026-09-01 | 575 |
| 2026-09-02 | 581 |
| 2026-09-03 | 582 |
| 2026-09-04 | 585 |
| 2026-09-05 | 586 |
| 2026-09-06 | 587 |
| 2026-09-07 | 590 |
| 2026-09-08 | 594 |
| 2026-09-09 | 597 |
| 2026-09-10 | 599 |
| 2026-09-12 | 604 |
| 2026-09-13 | 609 |
| 2026-09-14 | 623 |
| 2026-09-15 | 627 |
| 2026-09-16 | 629 |
| 2026-09-18 | 633 |
| 2026-09-19 | 633 |
| 2026-09-21 | 635 |
| 2026-09-23 | 638 |
| 2026-09-24 | 639 |
| 2026-09-26 | 642 |
| 2026-09-30 | 643 |
| 2026-10-01 | 647 |
| 2026-10-02 | 648 |
| 2026-10-04 | 649 |
| 2026-10-05 | 648 |
| 2026-10-06 | 648 |
| 2026-10-07 | 647 |
| 2026-10-08 | 650 |
| 2026-10-09 | 651 |
| 2026-10-10 | 651 |
| 2026-10-12 | 652 |
| 2026-10-14 | 653 |
| 2026-10-16 | 653 |
| 2026-10-17 | 654 |
| 2026-10-18 | 654 |
| 2026-10-19 | 655 |
| 2026-10-20 | 657 |
| 2026-10-21 | 657 |
| 2026-10-22 | 657 |
| 2026-10-23 | 658 |
| 2026-10-25 | 660 |
| 2026-10-26 | 661 |
| 2026-10-27 | 661 |
| 2026-10-28 | 661 |
| 2026-10-29 | 662 |
| 2026-10-30 | 663 |
| 2026-11-01 | 664 |
| 2026-11-03 | 666 |
| 2026-11-04 | 667 |
Today
+1 stars today
This week
+6 stars this week
This month
+19 stars this month
Momentum
45.0
growth rate 0.91%/day
# yzma - Go with your own intelligence
[](https://pkg.go.dev/github.com/hybridgroup/yzma) [](https://github.com/hybridgroup/yzma/actions/workflows/linux.yml) [](https://github.com/hybridgroup/yzma/actions/workflows/macos.yml) [](https://github.com/hybridgroup/yzma/actions/workflows/windows.yml) [](https://github.com/hybridgroup/yzma/actions/workflows/wasm.yml) [](https://github.com/hybridgroup/llama-cpp-builder/releases) [](https://bsky.app/profile/yzma.ai) [](https://mastodon.social/@yzma_ai)
`yzma` lets you write Go applications that directly integrate [`llama.cpp`](https://github.com/ggml-org/llama.cpp) for fully local inference using hardware acceleration.
- Run the latest Vision Language Models (VLM) and Large/Small/Tiny Language Models (LLM) on Linux, macOS, or Windows.
- Use any available hardware acceleration such as [CUDA](https://en.wikipedia.org/wiki/CUDA), [Metal](https://en.wikipedia.org/wiki/Metal_(API)), [OpenVINO](https://en.wikipedia.org/wiki/OpenVINO), or [Vulkan](https://en.wikipedia.org/wiki/Vulkan) for maximum performance.
- Run in a browser as well, with [TinyGo](https://tinygo.org) and WebAssembly, on the CPU or on the GPU with [WebGPU](https://en.wikipedia.org/wiki/WebGPU).
- `yzma` uses the [`purego`](https://github.com/ebitengine/purego) and [`ffi`](https://github.com/JupiterRider/ffi) packages so CGo is not needed.
- Works with the newest `llama.cpp` releases so you can use the latest features, performance improvements, and bugfixes.
The documentation is at **[yzma.ai](https://yzma.ai)**.
This example uses the [SmolLM2-135M-GGUF](https://huggingface.co/QuantFactory/SmolLM2-135M-GGUF) model:
```go
package main
import (
"fmt"
"os"
"path/filepath"
"github.com/hybridgroup/yzma/pkg/download"
"github.com/hybridgroup/yzma/pkg/llama"
)
var (
modelFile = "SmolLM2-135M.Q4_K_M.gguf"
prompt = "Are you ready to go?"
libPath = os.Getenv("YZMA_LIB")
responseLength int32 = 12
)
func main() {
llama.Load(libPath)
llama.LogSet(llama.LogSilent())
llama.Init()
model, _ := llama.ModelLoadFromFile(filepath.Join(download.DefaultModelsDir(), modelFile), llama.ModelDefaultParams())
ctx, _ := llama.InitFromModel(model, llama.ContextDefaultParams())
vocab := llama.ModelGetVocab(model)
tokens := llama.Tokenize(vocab, prompt, true, false)
batch := llama.BatchGetOne(tokens)
sampler := llama.SamplerChainInit(llama.SamplerChainDefaultParams())
llama.SamplerChainAdd(sampler, llama.SamplerInitGreedy())
for pos := int32(0); pos < responseLength; pos += batch.NTokens {
llama.Decode(ctx, batch)
token := llama.SamplerSample(sampler, ctx, -1)
if llama.VocabIsEOG(vocab, token) {
fmt.Println()
break
}
buf := make([]byte, 36)
len := llama.TokenToPiece(vocab, token, buf, 0, true)
fmt.Print(string(buf[:len]))
batch = llama.BatchGetOne([]llama.Token{token})
}
fmt.Println()
}
```
[Install `yzma`](https://yzma.ai/getting-started/install/), then download the model using the `yzma` command line tool:
```shell
$ yzma model get -u https://huggingface.co/QuantFactory/SmolLM2-135M-GGUF/resolve/main/SmolLM2-135M.Q4_K_M.gguf
```
And run the Go program:
```Excerpt of 14,536 characters
Read on GitHub
Ron Evans · The Hybrid Group
440
17
Samy Fodil · Taubyte
3
3
1
1
Would you bet a product on this? Bounded 0–100 and slow moving.
matched fp:fc5db135d7fce425, topic:llm, topic:llama
matched fp:fc5db135d7fce425, topic:llamacpp
matched fp:fc5db135d7fce425, topic:cuda
matched fp:fc5db135d7fce425, topic:vlm, readme:vision-language