CoolFace
Modelpublic

mradermacher/LLaMA-7B-HF-i1-GGUF

sourceHugging Faceotherupdated 1y agoView on Hugging Face
0likes1.5kdownloads
Model Card

About

<!-- ### quantizeversion: 2 --> <!-- ### outputtensorquantised: 1 --> <!-- ### converttype: hf --> <!-- ### vocab_type: --> <!-- ### tags: nicoboss --> weighted/imatrix quants of https://huggingface.co/Neko-Institute-of-Science/LLaMA-7B-HF

<!-- provided-files --> static quants are available at https://huggingface.co/mradermacher/LLaMA-7B-HF-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
PART 1 PART 2i1-IQ1_S3.2for the desperate
PART 1 PART 2i1-IQ1_M3.4mostly desperate
PART 1 PART 2i1-IQ2_XXS3.8
GGUFi1-IQ4_NL3.9prefer IQ4_XS
GGUFi1-Q404_43.9fast on arm, low quality
GGUFi1-Q404_83.9fast on arm+i8mm, low quality
GGUFi1-Q408_83.9fast on arm+sve, low quality
PART 1 PART 2i1-IQ2_XS4.2
GGUFi1-Q4_14.3
PART 1 PART 2i1-IQ2_S4.5
PART 1 PART 2i1-Q2KS4.7very low quality
PART 1 PART 2i1-IQ2_M4.8
PART 1 PART 2i1-Q2_K5.2IQ3_XXS probably better
PART 1 PART 2i1-IQ3_XXS5.3lower quality
PART 1 PART 2i1-IQ3_XS5.7
PART 1 PART 2i1-IQ3_S6.0beats Q3_K*
PART 1 PART 2i1-Q3KS6.0IQ3_XS probably better
PART 1 PART 2i1-IQ3_M6.3
PART 1 PART 2i1-Q3KM6.7IQ3_S probably better
PART 1 PART 2i1-Q3KL7.3IQ3_M probably better
PART 1 PART 2i1-IQ4_XS7.3
PART 1 PART 2i1-Q4_07.8fast, low quality
PART 1 PART 2i1-Q4KS7.8optimal size/speed/quality
PART 1 PART 2i1-Q4KM8.3fast, recommended
PART 1 PART 2i1-Q5KS9.4
PART 1 PART 2i1-Q5KM9.7
PART 1 PART 2i1-Q6_K11.2practically 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 -->