aitorrent/dolphin-2.9.3-qwen2-1.5b-GGUF

aitorrent/dolphin-2.9.3-qwen2-1.5b-GGUF
Photo by Rudney Uezu / Unsplash

About

This repository provides static quants for the dolphin-2.9.3-qwen2-1.5b model, originally from https://huggingface.co/cognitivecomputations/dolphin-2.9.3-qwen2-1.5b. Additionally, weighted/imatrix quants are available at https://huggingface.co/mradermacher/dolphin-2.9.3-qwen2-1.5b-i1-GGUF.

Usage

For guidance on using GGUF files, please refer to one of TheBloke's READMEs, which cover topics such as concatenating multi-part files.

Provided Quants

Below is a list of available quants, sorted by size (not necessarily quality). IQ-quants are often preferred over similar-sized non-IQ quants.

Link Type Size (GB) Notes
GGUF Q2_K 0.8
GGUF IQ3_XS 0.8
GGUF Q3_K_S 0.9
GGUF IQ3_S 0.9 Beats Q3_K*
GGUF IQ3_M 0.9
GGUF Q3_K_M 0.9 Lower quality
GGUF Q3_K_L 1.0
GGUF IQ4_XS 1.0
GGUF Q4_K_S 1.0 Fast, recommended
GGUF Q4_K_M 1.1 Fast, recommended
GGUF Q5_K_S 1.2
GGUF Q5_K_M 1.2
GGUF Q6_K 1.4 Very good quality
GGUF Q8_0 1.7 Fast, best quality
GGUF f16 3.2 16 bpw, overkill

Comparison Graph

ikawrakow has provided a handy graph comparing some lower-quality quant types (lower is better):

[image.png]

Additional Insights

For further discussion on the matter, see Artefact2's thoughts: https://gist.github.com/Artefact2/b5f810600771265fc1e39442288e8ec9

Credit: Original description from https://huggingface.co/cognitivecomputations/dolphin-2.9.3-qwen2-1.5b.