---
exported_from: Weyaxi/Bagel-Hermes-2x34B
language:
- en
library_name: transformers
license: other
license_link: https://huggingface.co/01-ai/Yi-34B-200K/blob/main/LICENSE
license_name: yi-license
quantized_by: mradermacher
tags:
- yi
- moe
---
## About

<!-- ### quantize_version: 1 -->
<!-- ### output_tensor_quantised: 1 -->
<!-- ### convert_type:  -->
<!-- ### vocab_type:  -->
weighted/imatrix quants of https://huggingface.co/Weyaxi/Bagel-Hermes-2x34B

<!-- provided-files -->
static quants are available at https://huggingface.co/mradermacher/Bagel-Hermes-2x34B-GGUF
## Usage

If you are unsure how to use GGUF files, refer to one of [TheBloke's
READMEs](https://huggingface.co/TheBloke/KafkaLM-70B-German-V0.1-GGUF) 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](https://huggingface.co/mradermacher/Bagel-Hermes-2x34B-i1-GGUF/resolve/main/Bagel-Hermes-2x34B.i1-IQ2_M.gguf) | i1-IQ2_M | 20.5 |  |
| [GGUF](https://huggingface.co/mradermacher/Bagel-Hermes-2x34B-i1-GGUF/resolve/main/Bagel-Hermes-2x34B.i1-Q2_K.gguf) | i1-Q2_K | 22.5 | IQ3_XXS probably better |
| [GGUF](https://huggingface.co/mradermacher/Bagel-Hermes-2x34B-i1-GGUF/resolve/main/Bagel-Hermes-2x34B.i1-IQ3_XXS.gguf) | i1-IQ3_XXS | 23.6 | lower quality |
| [GGUF](https://huggingface.co/mradermacher/Bagel-Hermes-2x34B-i1-GGUF/resolve/main/Bagel-Hermes-2x34B.i1-Q3_K_S.gguf) | i1-Q3_K_S | 26.4 | IQ3_XS probably better |
| [GGUF](https://huggingface.co/mradermacher/Bagel-Hermes-2x34B-i1-GGUF/resolve/main/Bagel-Hermes-2x34B.i1-Q3_K_M.gguf) | i1-Q3_K_M | 29.3 | IQ3_S probably better |
| [GGUF](https://huggingface.co/mradermacher/Bagel-Hermes-2x34B-i1-GGUF/resolve/main/Bagel-Hermes-2x34B.i1-Q3_K_L.gguf) | i1-Q3_K_L | 31.9 | IQ3_M probably better |
| [GGUF](https://huggingface.co/mradermacher/Bagel-Hermes-2x34B-i1-GGUF/resolve/main/Bagel-Hermes-2x34B.i1-IQ4_XS.gguf) | i1-IQ4_XS | 32.6 |  |
| [GGUF](https://huggingface.co/mradermacher/Bagel-Hermes-2x34B-i1-GGUF/resolve/main/Bagel-Hermes-2x34B.i1-Q4_K_S.gguf) | i1-Q4_K_S | 34.7 | optimal size/speed/quality |
| [GGUF](https://huggingface.co/mradermacher/Bagel-Hermes-2x34B-i1-GGUF/resolve/main/Bagel-Hermes-2x34B.i1-Q4_K_M.gguf) | i1-Q4_K_M | 36.8 | fast, recommended |
| [GGUF](https://huggingface.co/mradermacher/Bagel-Hermes-2x34B-i1-GGUF/resolve/main/Bagel-Hermes-2x34B.i1-Q5_K_S.gguf) | i1-Q5_K_S | 42.0 |  |
| [GGUF](https://huggingface.co/mradermacher/Bagel-Hermes-2x34B-i1-GGUF/resolve/main/Bagel-Hermes-2x34B.i1-Q6_K.gguf) | i1-Q6_K | 50.0 | practically like static Q6_K |


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

![image.png](https://www.nethype.de/huggingface_embed/quantpplgraph.png)

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

## Thanks

I thank my company, [nethype GmbH](https://www.nethype.de/), for letting
me use its servers and providing upgrades to my workstation to enable
this work in my free time.

<!-- end -->