bartowski commited on
Commit
03e5843
1 Parent(s): f4978fb

Main branch

Browse files
Files changed (2) hide show
  1. README.md +48 -0
  2. measurement.json +0 -0
README.md ADDED
@@ -0,0 +1,48 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ pipeline_tag: text-generation
3
+ tags:
4
+ - orca
5
+ - orca2
6
+ - microsoft
7
+ quantized_by: bartowski
8
+ ---
9
+
10
+ ## Exllama v2 Quantizations of Orca-2-13b
11
+
12
+ Using <a href="https://github.com/turboderp/exllamav2/releases/tag/v0.0.8">turboderp's ExLlamaV2 v0.0.8</a> for quantization.
13
+
14
+ Each branch contains an individual bits per weight, with the main one containing only the meaurement.json for further conversions.
15
+
16
+ Conversion was done using wikitext-103-raw-v1-test.parquet as calibration dataset.
17
+
18
+ Default arguments used except when the bits per weight is above 6.0, at that point the lm_head layer is quantized at 8 bits per weight instead of the default 6.
19
+
20
+ Original model: https://huggingface.co/microsoft/Orca-2-13b
21
+
22
+ ## Download instructions
23
+
24
+ With git:
25
+
26
+ ```shell
27
+ git clone --single-branch --branch 4_0 https://huggingface.co/bartowski/Orca-2-13b-exl2
28
+ ```
29
+
30
+ With huggingface hub (credit to TheBloke for instructions):
31
+
32
+ ```shell
33
+ pip3 install huggingface-hub
34
+ ```
35
+
36
+ To download the `main` (only useful if you only care about measurement.json) branch to a folder called `Orca-2-13b-exl2`:
37
+
38
+ ```shell
39
+ mkdir Orca-2-13b-exl2
40
+ huggingface-cli download bartowski/Orca-2-13b-exl2 --local-dir Orca-2-13b-exl2 --local-dir-use-symlinks False
41
+ ```
42
+
43
+ To download from a different branch, add the `--revision` parameter:
44
+
45
+ ```shell
46
+ mkdir Orca-2-13b-exl2
47
+ huggingface-cli download bartowski/Orca-2-13b-exl2 --revision 4_0 --local-dir Orca-2-13b-exl2 --local-dir-use-symlinks False
48
+ ```
measurement.json ADDED
The diff for this file is too large to render. See raw diff