Apppublic

hispa-nico/gemma4-e2b

sourceHugging Faceupdated 9d agoView on Hugging Face

0likes

Dockerfile21 linesDownload Raw Back to root

1FROM ghcr.io/ggml-org/llama.cpp:server2CMD ["-hf", "huihui-ai/Huihui-gemma-4-E2B-it-qat-q4_0-unquantized-abliterated-GGUF:Q4_0", \3	 "-t", "2", \4     "-tb", "2", \5     "-np", "1", \6     "--cpu-strict", "1", \7     "--device", "none", \8     "-fa", "on", \9     "-c", "4096", \10     "--poll", "100", \11     "-lm", "none", \12     "--jinja", \13     "--temp", "1.0", \14     "--top-k", "64", \15     "--min-p", "0.95", \16     "--spec-type", "draft-mtp", \17	 "--spec-draft-n-max", "3", \18	 "--spec-draft-n-min", "1", \19     "--no-mmproj", \20     "--rope-scaling", "none", \21	 "--port", "7860"]