mradermacher's picture
auto-patch README.md
3fe6ae5 verified
metadata
base_model: GOAT-AI/GOAT-70B-Storytelling
language:
  - en
library_name: transformers
license: llama2
model_type: llama
quantized_by: mradermacher
tags:
  - facebook
  - meta
  - pytorch
  - llama
  - llama-2
  - Storywriter

About

weighted/imatrix quants of https://huggingface.co/GOAT-AI/GOAT-70B-Storytelling

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)

Link Type Size/GB Notes
GGUF i1-IQ1_S 15.0 for the desperate
GGUF i1-IQ1_M 16.0 mostly desperate
GGUF i1-IQ2_XXS 18.7
GGUF i1-IQ2_XS 20.8
GGUF i1-IQ2_S 21.8
GGUF i1-IQ2_M 23.7
GGUF i1-Q2_K 25.9 IQ3_XXS probably better
GGUF i1-IQ3_XXS 27.4 lower quality
GGUF i1-IQ3_XS 28.6
GGUF i1-Q3_K_XS 28.7
GGUF i1-IQ3_S 30.3 beats Q3_K*
GGUF i1-Q3_K_S 30.3 IQ3_XS probably better
GGUF i1-IQ3_M 31.4
GGUF i1-Q3_K_M 33.7 IQ3_S probably better
GGUF i1-Q3_K_L 36.6 IQ3_M probably better
GGUF i1-IQ4_XS 37.2
GGUF i1-IQ4_NL 39.4 prefer IQ4_XS
GGUF i1-Q4_0 39.4 fast, low quality
GGUF i1-Q4_K_S 39.7 optimal size/speed/quality
GGUF i1-Q4_K_M 41.8 fast, recommended
GGUF i1-Q5_K_S 47.9
GGUF i1-Q5_K_M 49.2
PART 1 PART 2 i1-Q6_K 57.0 practically 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.