Text Generation
Transformers
English
orpo
qlora
trl
bartowski commited on
Commit
6c0a6f3
1 Parent(s): 84574b8

measurement.json

Browse files
Files changed (2) hide show
  1. README.md +74 -0
  2. measurement.json +0 -0
README.md ADDED
@@ -0,0 +1,74 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ language:
3
+ - en
4
+ license: apache-2.0
5
+ library_name: transformers
6
+ tags:
7
+ - orpo
8
+ - qlora
9
+ - trl
10
+ datasets:
11
+ - alvarobartt/dpo-mix-7k-simplified
12
+ - argilla/dpo-mix-7k
13
+ base_model: mistralai/Mistral-7B-v0.1
14
+ pipeline_tag: text-generation
15
+ inference: false
16
+ quantized_by: bartowski
17
+ ---
18
+
19
+ ## Exllama v2 Quantizations of Mistral-7B-v0.1-ORPO
20
+
21
+ Using <a href="https://github.com/turboderp/exllamav2/releases/tag/v0.0.16">turboderp's ExLlamaV2 v0.0.16</a> for quantization.
22
+
23
+ <b>The "main" branch only contains the measurement.json, download one of the other branches for the model (see below)</b>
24
+
25
+ Each branch contains an individual bits per weight, with the main one containing only the meaurement.json for further conversions.
26
+
27
+ Original model: https://huggingface.co/alvarobartt/Mistral-7B-v0.1-ORPO
28
+
29
+ | Branch | Bits | lm_head bits | VRAM (4k) | VRAM (16k) | VRAM (32k) | Description |
30
+ | ----- | ---- | ------- | ------ | ------ | ------ | ------------ |
31
+ | [8_0](https://huggingface.co/bartowski/Mistral-7B-v0.1-ORPO-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. |
32
+ | [6_5](https://huggingface.co/bartowski/Mistral-7B-v0.1-ORPO-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**. |
33
+ | [5_0](https://huggingface.co/bartowski/Mistral-7B-v0.1-ORPO-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. |
34
+ | [4_25](https://huggingface.co/bartowski/Mistral-7B-v0.1-ORPO-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. |
35
+ | [3_5](https://huggingface.co/bartowski/Mistral-7B-v0.1-ORPO-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. |
36
+
37
+ ## Download instructions
38
+
39
+ With git:
40
+
41
+ ```shell
42
+ git clone --single-branch --branch 6_5 https://huggingface.co/bartowski/Mistral-7B-v0.1-ORPO-exl2 Mistral-7B-v0.1-ORPO-exl2-6_5
43
+ ```
44
+
45
+ With huggingface hub (credit to TheBloke for instructions):
46
+
47
+ ```shell
48
+ pip3 install huggingface-hub
49
+ ```
50
+
51
+ To download the `main` (only useful if you only care about measurement.json) branch to a folder called `Mistral-7B-v0.1-ORPO-exl2`:
52
+
53
+ ```shell
54
+ mkdir Mistral-7B-v0.1-ORPO-exl2
55
+ huggingface-cli download bartowski/Mistral-7B-v0.1-ORPO-exl2 --local-dir Mistral-7B-v0.1-ORPO-exl2 --local-dir-use-symlinks False
56
+ ```
57
+
58
+ To download from a different branch, add the `--revision` parameter:
59
+
60
+ Linux:
61
+
62
+ ```shell
63
+ mkdir Mistral-7B-v0.1-ORPO-exl2-6_5
64
+ huggingface-cli download bartowski/Mistral-7B-v0.1-ORPO-exl2 --revision 6_5 --local-dir Mistral-7B-v0.1-ORPO-exl2-6_5 --local-dir-use-symlinks False
65
+ ```
66
+
67
+ Windows (which apparently doesn't like _ in folders sometimes?):
68
+
69
+ ```shell
70
+ mkdir Mistral-7B-v0.1-ORPO-exl2-6.5
71
+ huggingface-cli download bartowski/Mistral-7B-v0.1-ORPO-exl2 --revision 6_5 --local-dir Mistral-7B-v0.1-ORPO-exl2-6.5 --local-dir-use-symlinks False
72
+ ```
73
+
74
+ 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