Modelpublic

mradermacher/miiqu-f16-i1-GGUF

sourceHugging Faceotherupdated 2y agoView on Hugging Face

0likes545downloads

Model Card

weighted/imatrix quants of https://huggingface.co/Infinimol/miiqu-f16

static quants are available at https://huggingface.co/mradermacher/miiqu-f16-GGUF

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.

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

Link	Type	Size/GB	Notes
GGUF	i1-IQ1_S	19.4	for the desperate
GGUF	i1-IQ2_XXS	24.4
GGUF	i1-IQ2_XS	27.0
GGUF	i1-IQ2_S	28.4
GGUF	i1-IQ2_M	30.8
GGUF	i1-Q2_K	33.6	IQ3_XXS probably better
GGUF	i1-IQ3_XXS	35.1	lower quality
GGUF	i1-IQ3_XS	37.3
GGUF	i1-Q3KS	39.3	IQ3_XS probably better
GGUF	i1-IQ3_S	39.5	beats Q3_K*
GGUF	i1-IQ3_M	40.8
GGUF	i1-Q3KM	43.9	IQ3_S probably better
GGUF	i1-Q3KL	47.8	IQ3_M probably better
GGUF	i1-IQ4_XS	48.6
PART 1 PART 2	i1-Q4KS	51.7	optimal size/speed/quality
PART 1 PART 2	i1-Q4KM	54.6	fast, recommended
PART 1 PART 2	i1-Q5KS	62.6
PART 1 PART 2	i1-Q5KM	64.3
PART 1 PART 2	i1-Q6_K	74.5	practically like static Q6_K

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

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

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

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.