base_model: 152334H/miqu-1-70b-sf
language:
- en
library_name: transformers
license: other
quantized_by: mradermacher
tags:
- merge
- moe
About
weighted/imatrix quants of https://huggingface.co/lodrick-the-lafted/Grafted-Wind-Elementals-2x70B
imatrix training data reduced to 130k tokens because llama otherwise corrupts the imatrix.
static quants are available at https://huggingface.co/mradermacher/Grafted-Wind-Elementals-2x70B-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 | i1-Q2_K | 46.3 | IQ3_XXS probably better |
| PART 1 PART 2 | i1-Q3_K_S | 54.6 | IQ3_XS probably better |
| PART 1 PART 2 | i1-Q3_K_M | 60.6 | IQ3_S probably better |
| PART 1 PART 2 | i1-Q3_K_L | 65.6 | IQ3_M probably better |
| PART 1 PART 2 | i1-Q4_K_S | 71.7 | optimal size/speed/quality |
| PART 1 PART 2 | i1-Q4_K_M | 76.0 | fast, recommended |
| PART 1 PART 2 | i1-Q5_K_S | 86.6 | |
| PART 1 PART 2 PART 3 | i1-Q6_K | 103.2 | 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
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.