Meet the Underdog Saluki 27B: A 2-bit Qwen3.8-27B That Beats the Original at Tool Calling
Saluki 27B, a 7.89 GB 2-bit Qwen3.8-27B GGUF, beats the full model on tool calling.
MarkTechPost profiles Underdog Saluki 27B, a 7.89 GB 2-bit GGUF quantization of Qwen3.8-27B released under Apache 2.0. The quantized model reportedly outperforms the roughly 54 GB original on tool calling while scoring lower on competition math and reasoning. It is a compressed release of an existing Qwen model rather than a new frontier training run.
- 7.89 GB 2-bit GGUF versus about a 54 GB original.
- Reportedly beats the base model on tool calling.
- Scores lower on competition math and reasoning.
- Released under the Apache 2.0 license.
Underdog Saluki 27B is a 7.89 GB, 2-bit GGUF of Qwen3.8-27B under Apache 2.0. It beats the 54 GB original on tool calling but gives up ground on competition math and reasoning. The post Meet the Underdog Saluki 27B: A 2-bit Qwen3.8-27B That Beats the Original at Tool Calling appeared first on MarkTechPost.
The full text could not be extracted from this site (paywall, bot protection or heavy scripting). Read it at marktechpost.com.