GitHub ↗
site index
Engine

What the engine is, what runs where, and who it is built on.

engine
Docs

Every device, every component, and how to run them.

docs
Blog

The build log: what moved over, what got faster, what broke.

blog
Benchmarks

Measured, not projected. Every number with how it was taken.

wiki
1bit serve

The engine runs inside Lemonade: Lemonade runs 1bit serve as a backend, the way it runs llama-server. 1bit serve…

serve
Lemonade

The 1bit engine runs inside Lemonade. Lemonade is the server users talk to, with its own catalog, downloads, router and…

lemonade
NPU

The fast lane runs one whole-layer kernel per decoder layer, plus an lm-head kernel. For each token, the host submits…

npu
HRX + Vulkan

One llama.cpp build with two GPU backends, ggml-hrx (AMD's HRX runtime) and ggml-vulkan. Its single llama-server…

hrx
Vulkan (upstream)

1bit serve --device vulkan runs a llama-server built from upstream llama.cpp (MIT), pinned to its latest release in…

vulkan
Lean (ROCmFP4, ROCmI4)

1bit serve --lean trades accuracy for speed. It runs models in the AMD-focused formats of ROCmFPX (MIT), a llama.cpp…

lean
ZINC

ZINC (MIT) is a single-binary GGUF engine written in Zig. It has its own kernels for four GPU backends, and serves…

zinc
Apple Silicon

On macOS, 1bit serve --device mlx serves MLX models. The executor is the server binary of lemon-mlx-engine (fork…

apple
Laya router

Step 4 of the port (docs/PORTING.md): Laya picks where each request runs. Laya (Apache-2.0) is a non-autoregressive…

laya
Tokenizers

The engine reads any model's tokenizer.json through Hugging Face tokenizers, the Rust library behind the tokenizers npm…

tokenizers
Linux kernel

The engine's NPU driver (amdxdna) and GPU driver (amdgpu) come from the kernel, so the kernel is pinned like every…

kernel
Porting map

This repository is the working 1bit-MONSTER engine, ported without its history. Each step below is one PR (or a short…

PORTING