CoolFace
Modelpublic

mradermacher/codegemma-7b-it-i1-GGUF

sourceHugging Facegemmaupdated 2y agoView on Hugging Face
0likes1.4kdownloads
Model Card

About

<!-- ### quantizeversion: 2 --> <!-- ### outputtensorquantised: 1 --> <!-- ### converttype: hf --> <!-- ### vocab_type: --> <!-- ### tags: nicoboss --> weighted/imatrix quants of https://huggingface.co/google/codegemma-7b-it

<!-- provided-files --> static quants are available at https://huggingface.co/mradermacher/codegemma-7b-it-GGUF

Usage

If you are unsure how to use GGUF files, refer to one of TheBloke's READMEs for more details, including on how to concatenate multi-part files.

Provided Quants

(sorted by size, not necessarily quality. IQ-quants are often preferable over similar sized non-IQ quants)

LinkTypeSize/GBNotes
GGUFi1-IQ1_S2.3for the desperate
GGUFi1-IQ1_M2.4mostly desperate
GGUFi1-IQ2_XXS2.7
GGUFi1-IQ2_XS2.9
GGUFi1-IQ2_S3.0
GGUFi1-IQ2_M3.2
GGUFi1-Q2_K3.6IQ3_XXS probably better
GGUFi1-IQ3_XXS3.6lower quality
GGUFi1-IQ3_XS3.9
GGUFi1-IQ3_S4.1beats Q3_K*
GGUFi1-Q3KS4.1IQ3_XS probably better
GGUFi1-IQ3_M4.2
GGUFi1-Q3KM4.5IQ3_S probably better
GGUFi1-Q3KL4.8IQ3_M probably better
GGUFi1-IQ4_XS4.9
GGUFi1-Q404_45.1fast on arm, low quality
GGUFi1-Q404_85.1fast on arm+i8mm, low quality
GGUFi1-Q408_85.1fast on arm+sve, low quality
GGUFi1-Q4_05.1fast, low quality
GGUFi1-Q4KS5.1optimal size/speed/quality
GGUFi1-Q4KM5.4fast, recommended
GGUFi1-Q5KS6.1
GGUFi1-Q5KM6.2
GGUFi1-Q6_K7.1practically like static Q6_K

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

image.png

And here are Artefact2's thoughts on the matter: https://gist.github.com/Artefact2/b5f810600771265fc1e39442288e8ec9

FAQ / Model Request

See https://huggingface.co/mradermacher/model_requests for some answers to questions you might have and/or if you want some other model quantized.

Thanks

I thank my company, nethype GmbH, for letting me use its servers and providing upgrades to my workstation to enable this work in my free time. Additional thanks to @nicoboss for giving me access to his private supercomputer, enabling me to provide many more imatrix quants, at much higher quality, than I would otherwise be able to.

<!-- end -->