File size: 2,672 Bytes
3aa4e07
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
ecb4cbc
 
 
 
 
3aa4e07
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
---
base_model:
- tokyotech-llm/Swallow-70b-instruct-hf
- nitky/Swallow-70b-NVE-RP
exported_from: nitky/Swallow-70b-RP
language:
- en
library_name: transformers
license: llama2
model_type: llama
quantized_by: mradermacher
tags:
- mergekit
- merge
---
## About

<!-- ### quantize_version: 1 -->
<!-- ### output_tensor_quantised: 1 -->
<!-- ### convert_type:  -->
<!-- ### vocab_type:  -->
weighted/imatrix quants of https://huggingface.co/nitky/Swallow-70b-RP

<!-- provided-files -->
static quants are available at https://huggingface.co/mradermacher/Swallow-70b-RP-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/Swallow-70b-RP-i1-GGUF/resolve/main/Swallow-70b-RP.i1-IQ2_M.gguf) | i1-IQ2_M | 23.4 |  |
| [GGUF](https://huggingface.co/mradermacher/Swallow-70b-RP-i1-GGUF/resolve/main/Swallow-70b-RP.i1-Q2_K.gguf) | i1-Q2_K | 25.7 | IQ3_XXS probably better |
| [GGUF](https://huggingface.co/mradermacher/Swallow-70b-RP-i1-GGUF/resolve/main/Swallow-70b-RP.i1-IQ3_XXS.gguf) | i1-IQ3_XXS | 26.8 | lower quality |
| [GGUF](https://huggingface.co/mradermacher/Swallow-70b-RP-i1-GGUF/resolve/main/Swallow-70b-RP.i1-Q3_K_M.gguf) | i1-Q3_K_M | 33.5 | IQ3_S probably better |
| [GGUF](https://huggingface.co/mradermacher/Swallow-70b-RP-i1-GGUF/resolve/main/Swallow-70b-RP.i1-Q4_K_S.gguf) | i1-Q4_K_S | 39.5 | optimal size/speed/quality |
| [GGUF](https://huggingface.co/mradermacher/Swallow-70b-RP-i1-GGUF/resolve/main/Swallow-70b-RP.i1-Q4_K_M.gguf) | i1-Q4_K_M | 41.6 | fast, recommended |
| [PART 1](https://huggingface.co/mradermacher/Swallow-70b-RP-i1-GGUF/resolve/main/Swallow-70b-RP.i1-Q6_K.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/Swallow-70b-RP-i1-GGUF/resolve/main/Swallow-70b-RP.i1-Q6_K.gguf.part2of2) | i1-Q6_K | 56.8 | 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 -->