CoolFace
Modelpublic

AXONVERTEX-AI-RESEARCH/LFM2.5-8B-A1B-Q8_0-GGUF

sourceHugging Faceotherupdated 4mo agoView on Hugging Face
0likes67downloads
Model Card

LFM2.5-8B-A1B Q8_0 GGUF

This repository contains a Q8_0 GGUF quantized version of LiquidAI/LFM2.5-8B-A1B for llama.cpp-compatible runtimes.

Model relationship

  • —Base model: LiquidAI/LFM2.5-8B-A1B
  • —Relation: Quantized GGUF version
  • —Quantization: Q8_0
  • —Format: GGUF
  • —Runtime target: llama.cpp, Ollama, LM Studio, Jan, and other GGUF-compatible runtimes

Files

  • —LFM2.5-8B-A1B-Q8_0.gguf

License

Hugging Face metadata uses license: other because the upstream LiquidAI license identifier is not currently accepted as a valid Hugging Face YAML license enum.

Use of this model should follow the original LiquidAI model license terms from the upstream repository.

Run with llama.cpp

llama-cli -hf AXONVERTEX-AI-RESEARCH/LFM2.5-8B-A1B-Q80-GGUF:Q80 \ --jinja \ --ctx-size 4096 \ --temp 0.2 \ --top-p 0.8 \ --repeat-penalty 1.05

Run with Ollama

ollama run hf.co/AXONVERTEX-AI-RESEARCH/LFM2.5-8B-A1B-Q80-GGUF:Q80

Source

This GGUF was derived from the official LiquidAI GGUF release: LiquidAI/LFM2.5-8B-A1B-GGUF

AXONVERTEX-AI-RESEARCH/LFM2.5-8B-A1B-Q8_0-GGUF · CoolFace