bartowski commited on
Commit
f49558f
1 Parent(s): c51ea0b

measurement.json

Browse files
Files changed (2) hide show
  1. README.md +62 -0
  2. measurement.json +0 -0
README.md ADDED
@@ -0,0 +1,62 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ quantized_by: bartowski
4
+ pipeline_tag: text-generation
5
+ ---
6
+
7
+ ## Exllama v2 Quantizations of Mistral-Plus-7B
8
+
9
+ Using <a href="https://github.com/turboderp/exllamav2/releases/tag/v0.0.15">turboderp's ExLlamaV2 v0.0.15</a> for quantization.
10
+
11
+ <b>The "main" branch only contains the measurement.json, download one of the other branches for the model (see below)</b>
12
+
13
+ Each branch contains an individual bits per weight, with the main one containing only the meaurement.json for further conversions.
14
+
15
+ Original model: https://huggingface.co/zhengchenphd/Mistral-Plus-7B
16
+
17
+ | Branch | Bits | lm_head bits | VRAM (4k) | VRAM (16k) | VRAM (32k) | Description |
18
+ | ----- | ---- | ------- | ------ | ------ | ------ | ------------ |
19
+ | [8_0](https://huggingface.co/bartowski/Mistral-Plus-7B-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. |
20
+ | [6_5](https://huggingface.co/bartowski/Mistral-Plus-7B-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**. |
21
+ | [5_0](https://huggingface.co/bartowski/Mistral-Plus-7B-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. |
22
+ | [4_25](https://huggingface.co/bartowski/Mistral-Plus-7B-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. |
23
+ | [3_5](https://huggingface.co/bartowski/Mistral-Plus-7B-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. |
24
+
25
+ ## Download instructions
26
+
27
+ With git:
28
+
29
+ ```shell
30
+ git clone --single-branch --branch 6_5 https://huggingface.co/bartowski/Mistral-Plus-7B-exl2 Mistral-Plus-7B-exl2-6_5
31
+ ```
32
+
33
+ With huggingface hub (credit to TheBloke for instructions):
34
+
35
+ ```shell
36
+ pip3 install huggingface-hub
37
+ ```
38
+
39
+ To download the `main` (only useful if you only care about measurement.json) branch to a folder called `Mistral-Plus-7B-exl2`:
40
+
41
+ ```shell
42
+ mkdir Mistral-Plus-7B-exl2
43
+ huggingface-cli download bartowski/Mistral-Plus-7B-exl2 --local-dir Mistral-Plus-7B-exl2 --local-dir-use-symlinks False
44
+ ```
45
+
46
+ To download from a different branch, add the `--revision` parameter:
47
+
48
+ Linux:
49
+
50
+ ```shell
51
+ mkdir Mistral-Plus-7B-exl2-6_5
52
+ huggingface-cli download bartowski/Mistral-Plus-7B-exl2 --revision 6_5 --local-dir Mistral-Plus-7B-exl2-6_5 --local-dir-use-symlinks False
53
+ ```
54
+
55
+ Windows (which apparently doesn't like _ in folders sometimes?):
56
+
57
+ ```shell
58
+ mkdir Mistral-Plus-7B-exl2-6.5
59
+ huggingface-cli download bartowski/Mistral-Plus-7B-exl2 --revision 6_5 --local-dir Mistral-Plus-7B-exl2-6.5 --local-dir-use-symlinks False
60
+ ```
61
+
62
+ Want to support my work? Visit my ko-fi page here: https://ko-fi.com/bartowski
measurement.json ADDED
The diff for this file is too large to render. See raw diff