How to use from
llama.cpp
Install (macOS, Linux)
curl -LsSf https://llama.app/install.sh | sh
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf rkstgr/typer-1.5b-base-gguf:Q8_0
# Run inference directly in the terminal:
llama cli -hf rkstgr/typer-1.5b-base-gguf:Q8_0
Install from WinGet (Windows)
winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf rkstgr/typer-1.5b-base-gguf:Q8_0
# Run inference directly in the terminal:
llama cli -hf rkstgr/typer-1.5b-base-gguf:Q8_0
Build from source code
git clone https://github.com/ggml-org/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-app
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama serve -hf rkstgr/typer-1.5b-base-gguf:Q8_0
# Run inference directly in the terminal:
./build/bin/llama cli -hf rkstgr/typer-1.5b-base-gguf:Q8_0
Use Docker
docker model run hf.co/rkstgr/typer-1.5b-base-gguf:Q8_0
Quick Links

Typer: LLM for Typst

Typer is the first LLM trained specifically for Typst, a performant modern typesetting system for generating PDF documents.

Model

The model is based on Qwen2.5-Coder and adapted using LoRA training. The base model is trained using continued pre-training (CPT), without any post-training techniques like instruction-tuning.

Dataset

The dataset consists of the official Typst documentation and public Typst packages.

Acknowledgements

Computational resources were provided by the German AI Service Center WestAI.

Downloads last month
56
GGUF
Model size
2B params
Architecture
qwen2
Hardware compatibility
Log In to add your hardware

8-bit

Inference Providers NEW
This model isn't deployed by any Inference Provider. ๐Ÿ™‹ Ask for provider support

Model tree for rkstgr/typer-1.5b-base-gguf

Quantized
(61)
this model
Quantizations
1 model