File size: 2,886 Bytes
86d7d4c
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
f3c6a43
 
 
 
 
86d7d4c
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
63
64
65
---
datasets:
- JeanKaddour/minipile
exported_from: chargoddard/llama3-42b-v0
language:
- en
library_name: transformers
license: llama3
quantized_by: mradermacher
tags:
- axolotl
- mergekit
- llama
---
## About

<!-- ### quantize_version: 1 -->
<!-- ### output_tensor_quantised: 1 -->
<!-- ### convert_type:  -->
<!-- ### vocab_type:  -->
static quants of https://huggingface.co/chargoddard/llama3-42b-v0


<!-- provided-files -->
weighted/imatrix quants seem not to be available (by me) at this time. If they do not show up a week or so after the static ones, I have probably not planned for them. Feel free to request them by opening a Community Discussion.
## 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/llama3-42b-v0-GGUF/resolve/main/llama3-42b-v0.Q2_K.gguf) | Q2_K | 16.4 |  |
| [GGUF](https://huggingface.co/mradermacher/llama3-42b-v0-GGUF/resolve/main/llama3-42b-v0.Q3_K_S.gguf) | Q3_K_S | 19.1 |  |
| [GGUF](https://huggingface.co/mradermacher/llama3-42b-v0-GGUF/resolve/main/llama3-42b-v0.IQ3_S.gguf) | IQ3_S | 19.1 | beats Q3_K* |
| [GGUF](https://huggingface.co/mradermacher/llama3-42b-v0-GGUF/resolve/main/llama3-42b-v0.IQ3_M.gguf) | IQ3_M | 19.7 |  |
| [GGUF](https://huggingface.co/mradermacher/llama3-42b-v0-GGUF/resolve/main/llama3-42b-v0.Q3_K_M.gguf) | Q3_K_M | 21.1 | lower quality |
| [GGUF](https://huggingface.co/mradermacher/llama3-42b-v0-GGUF/resolve/main/llama3-42b-v0.Q3_K_L.gguf) | Q3_K_L | 22.9 |  |
| [GGUF](https://huggingface.co/mradermacher/llama3-42b-v0-GGUF/resolve/main/llama3-42b-v0.Q4_K_S.gguf) | Q4_K_S | 24.8 | fast, recommended |
| [GGUF](https://huggingface.co/mradermacher/llama3-42b-v0-GGUF/resolve/main/llama3-42b-v0.Q4_K_M.gguf) | Q4_K_M | 26.2 | fast, recommended |
| [GGUF](https://huggingface.co/mradermacher/llama3-42b-v0-GGUF/resolve/main/llama3-42b-v0.Q6_K.gguf) | Q6_K | 35.5 | very good quality |
| [GGUF](https://huggingface.co/mradermacher/llama3-42b-v0-GGUF/resolve/main/llama3-42b-v0.Q8_0.gguf) | Q8_0 | 46.0 | fast, best quality |


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 -->