---
license: apache-2.0
base_model: unsloth/Qwen3.8-27B-GGUF
tags:
  - gguf
  - vocabulary-pruning
  - ascii-condensed
---

# Qwen3.8-27B ASCII-Condensed UD-Q3_K_XL GGUF

A vocabulary-condensed version of Unsloth's Qwen3.8-27B UD-Q3_K_XL GGUF, intended for ASCII-heavy text and code workloads. It was made using the ASCII pruning approach described by [bsaleh03](https://huggingface.co/bsaleh03/Qwen3.8-27B-ASCII-Condensed).

Use `Qwen3.8-27B-ASCII-Condensed-UD-Q3_K_XL-v3.gguf` as the target model in llama.cpp. Non-ASCII text remains representable through byte fallback but typically uses more tokens.

Source: [unsloth/Qwen3.8-27B-GGUF](https://huggingface.co/unsloth/Qwen3.8-27B-GGUF), based on [Qwen/Qwen3.8-27B](https://huggingface.co/Qwen/Qwen3.8-27B). Apache-2.0 license, inherited from the source model.