File size: 3,060 Bytes
6011819
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
66
67
68
69
70
71
---
tags:
- merge
- mergekit
- lazymergekit
- mistralai/Mistral-7B-Instruct-v0.2
- cognitivecomputations/dolphin-2.8-mistral-7b-v02
base_model:
- mistralai/Mistral-7B-Instruct-v0.2
- cognitivecomputations/dolphin-2.8-mistral-7b-v02
license: apache-2.0
quantized_by: bartowski
pipeline_tag: text-generation
---

## Exllama v2 Quantizations of pandafish-2-7b-32k

Using <a href="https://github.com/turboderp/exllamav2/releases/tag/v0.0.17">turboderp's ExLlamaV2 v0.0.17</a> for quantization.

<b>The "main" branch only contains the measurement.json, download one of the other branches for the model (see below)</b>

Each branch contains an individual bits per weight, with the main one containing only the meaurement.json for further conversions.

Original model: https://huggingface.co/ichigoberry/pandafish-2-7b-32k

| Branch | Bits | lm_head bits | VRAM (4k) | VRAM (16k) | VRAM (32k) | Description |
| ----- | ---- | ------- | ------ | ------ | ------ | ------------ |
| [8_0](https://huggingface.co/bartowski/pandafish-2-7b-32k-exl2/tree/8_0) | 8.0 | 8.0 | 8.4 GB | 9.8 GB | 11.8 GB | Maximum quality that ExLlamaV2 can produce, near unquantized performance. |
| [6_5](https://huggingface.co/bartowski/pandafish-2-7b-32k-exl2/tree/6_5) | 6.5 | 8.0 | 7.2 GB | 8.6 GB | 10.6 GB | Very similar to 8.0, good tradeoff of size vs performance, **recommended**. |
| [5_0](https://huggingface.co/bartowski/pandafish-2-7b-32k-exl2/tree/5_0) | 5.0 | 6.0 | 6.0 GB | 7.4 GB |  9.4 GB | Slightly lower quality vs 6.5, but usable on 8GB cards. |
| [4_25](https://huggingface.co/bartowski/pandafish-2-7b-32k-exl2/tree/4_25) | 4.25 | 6.0 | 5.3 GB | 6.7 GB | 8.7 GB | GPTQ equivalent bits per weight, slightly higher quality. |
| [3_5](https://huggingface.co/bartowski/pandafish-2-7b-32k-exl2/tree/3_5) | 3.5 | 6.0 | 4.7 GB | 6.1 GB | 8.1 GB | Lower quality, only use if you have to. |

## Download instructions

With git:

```shell
git clone --single-branch --branch 6_5 https://huggingface.co/bartowski/pandafish-2-7b-32k-exl2 pandafish-2-7b-32k-exl2-6_5
```

With huggingface hub (credit to TheBloke for instructions):

```shell
pip3 install huggingface-hub
```

To download the `main` (only useful if you only care about measurement.json) branch to a folder called `pandafish-2-7b-32k-exl2`:

```shell
mkdir pandafish-2-7b-32k-exl2
huggingface-cli download bartowski/pandafish-2-7b-32k-exl2 --local-dir pandafish-2-7b-32k-exl2 --local-dir-use-symlinks False
```

To download from a different branch, add the `--revision` parameter:

Linux:

```shell
mkdir pandafish-2-7b-32k-exl2-6_5
huggingface-cli download bartowski/pandafish-2-7b-32k-exl2 --revision 6_5 --local-dir pandafish-2-7b-32k-exl2-6_5 --local-dir-use-symlinks False
```

Windows (which apparently doesn't like _ in folders sometimes?):

```shell
mkdir pandafish-2-7b-32k-exl2-6.5
huggingface-cli download bartowski/pandafish-2-7b-32k-exl2 --revision 6_5 --local-dir pandafish-2-7b-32k-exl2-6.5 --local-dir-use-symlinks False
```

Want to support my work? Visit my ko-fi page here: https://ko-fi.com/bartowski