About
weighted/imatrix quants of https://huggingface.co/CMSManhattan/JiRackUltra_32b
For a convenient overview and download list, visit our model page for this model.
static quants are available at https://huggingface.co/mradermacher/JiRackUltra_32b-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)
| Link | Type | Size/GB | Notes |
|---|---|---|---|
| GGUF | imatrix | 0.1 | imatrix file (for creating your own quants) |
| GGUF | i1-IQ1_S | 7.4 | for the desperate |
| GGUF | i1-IQ1_M | 8.0 | mostly desperate |
| GGUF | i1-IQ2_XXS | 9.1 | |
| GGUF | i1-IQ2_XS | 10.1 | |
| GGUF | i1-IQ2_S | 10.5 | |
| GGUF | i1-IQ2_M | 11.4 | |
| GGUF | i1-Q2_K_S | 11.6 | very low quality |
| GGUF | i1-Q2_K | 12.4 | IQ3_XXS probably better |
| GGUF | i1-IQ3_XXS | 12.9 | lower quality |
| GGUF | i1-IQ3_XS | 13.8 | |
| GGUF | i1-Q3_K_S | 14.5 | IQ3_XS probably better |
| GGUF | i1-IQ3_S | 14.5 | beats Q3_K* |
| GGUF | i1-IQ3_M | 14.9 | |
| GGUF | i1-Q3_K_M | 16.0 | IQ3_S probably better |
| GGUF | i1-Q3_K_L | 17.3 | IQ3_M probably better |
| GGUF | i1-IQ4_XS | 17.8 | |
| GGUF | i1-Q4_0 | 18.8 | fast, low quality |
| GGUF | i1-Q4_K_S | 18.9 | optimal size/speed/quality |
| GGUF | i1-Q4_K_M | 20.0 | fast, recommended |
| GGUF | i1-Q4_1 | 20.7 | |
| GGUF | i1-Q5_K_S | 22.7 | |
| GGUF | i1-Q5_K_M | 23.4 | |
| GGUF | i1-Q6_K | 27.0 | 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
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.