File size: 446 Bytes
1a1dc3b
fa09834
bba0c2d
 
7bee3fa
fa09834
7faf1af
1
2
3
4
5
6
7
FROM ghcr.io/ggml-org/llama.cpp:full

RUN apt update && apt install wget -y

RUN wget "https://huggingface.co/unsloth/GLM-4.7-Flash-REAP-23B-A3B-GGUF/resolve/main/GLM-4.7-Flash-REAP-23B-A3B-Q2_K.gguf" -O /model.gguf

CMD ["--server", "-m", "/model.gguf", "--port", "7860", "--host", "0.0.0.0", "-n", "768", "--no-mmap", "--jinja", "--temp", "1.0", "--top-p", "0.95", "--top-k", "64", "--reasoning-budget", "0", "--reasoning", "off", "-c", "4096"]