measurement.json
Browse files- README.md +86 -0
- measurement.json +0 -0
README.md
ADDED
@@ -0,0 +1,86 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
library_name: transformers
|
3 |
+
extra_gated_heading: Access CodeGemma on Hugging Face
|
4 |
+
extra_gated_prompt: >-
|
5 |
+
To access CodeGemma on Hugging Face, you’re required to review and agree to
|
6 |
+
Google’s usage license. To do this, please ensure you’re logged-in to Hugging
|
7 |
+
Face and click below. Requests are processed immediately.
|
8 |
+
extra_gated_button_content: Acknowledge license
|
9 |
+
pipeline_tag: text-generation
|
10 |
+
widget:
|
11 |
+
- text: >
|
12 |
+
<start_of_turn>user
|
13 |
+
Write a Python function to calculate the nth fibonacci number.<end_of_turn>
|
14 |
+
<start_of_turn>model
|
15 |
+
inference:
|
16 |
+
parameters:
|
17 |
+
max_new_tokens: 200
|
18 |
+
license: gemma
|
19 |
+
license_link: https://ai.google.dev/gemma/terms
|
20 |
+
quantized_by: bartowski
|
21 |
+
---
|
22 |
+
|
23 |
+
## Exllama v2 Quantizations of codegemma-1.1-7b-it
|
24 |
+
|
25 |
+
Using <a href="https://github.com/turboderp/exllamav2/releases/tag/v0.0.20">turboderp's ExLlamaV2 v0.0.20</a> for quantization.
|
26 |
+
|
27 |
+
<b>The "main" branch only contains the measurement.json, download one of the other branches for the model (see below)</b>
|
28 |
+
|
29 |
+
Each branch contains an individual bits per weight, with the main one containing only the meaurement.json for further conversions.
|
30 |
+
|
31 |
+
Original model: https://huggingface.co/google/codegemma-1.1-7b-it
|
32 |
+
|
33 |
+
## Prompt format
|
34 |
+
|
35 |
+
No chat template specified so default is used. This may be incorrect, check original model card for details.
|
36 |
+
|
37 |
+
```
|
38 |
+
<|im_start|>system
|
39 |
+
{system_prompt}<|im_end|>
|
40 |
+
<|im_start|>user
|
41 |
+
{prompt}<|im_end|>
|
42 |
+
<|im_start|>assistant
|
43 |
+
|
44 |
+
```
|
45 |
+
|
46 |
+
## Available sizes
|
47 |
+
|
48 |
+
No GQA - VRAM requirements will be higher
|
49 |
+
|
50 |
+
| Branch | Bits | lm_head bits | Size (4k) | Size (16k) | Description |
|
51 |
+
| -------------------------------------------------------------- | ---- | ------------ | --------- | ---------- | ----------- |
|
52 |
+
| [8_0](https://huggingface.co/bartowski/codegemma-1.1-7b-it-exl2/tree/8_0) | 8.0 | 8.0 | 14.0 GB | 19.4 GB | Maximum quality that ExLlamaV2 can produce, near unquantized performance. |
|
53 |
+
| [6_5](https://huggingface.co/bartowski/codegemma-1.1-7b-it-exl2/tree/6_5) | 6.5 | 8.0 | 12.5 GB | 17.9 GB | Near unquantized performance at vastly reduced size, **recommended**. |
|
54 |
+
| [5_0](https://huggingface.co/bartowski/codegemma-1.1-7b-it-exl2/tree/5_0) | 5.0 | 6.0 | 10.9 GB | 16.3 GB | Slightly lower quality vs 6.5, but usable on 8GB cards with 4k context. |
|
55 |
+
| [4_25](https://huggingface.co/bartowski/codegemma-1.1-7b-it-exl2/tree/4_25) | 4.25 | 6.0 | 10.2 GB | 15.7 GB | GPTQ equivalent bits per weight. |
|
56 |
+
| [3_5](https://huggingface.co/bartowski/codegemma-1.1-7b-it-exl2/tree/3_5) | 3.5 | 6.0 | 9.5 GB | 14.9 GB | Lower quality, not recommended. |
|
57 |
+
|
58 |
+
## Download instructions
|
59 |
+
|
60 |
+
With git:
|
61 |
+
|
62 |
+
```shell
|
63 |
+
git clone --single-branch --branch 6_5 https://huggingface.co/bartowski/codegemma-1.1-7b-it-exl2 codegemma-1.1-7b-it-exl2-6_5
|
64 |
+
```
|
65 |
+
|
66 |
+
With huggingface hub (credit to TheBloke for instructions):
|
67 |
+
|
68 |
+
```shell
|
69 |
+
pip3 install huggingface-hub
|
70 |
+
```
|
71 |
+
|
72 |
+
To download a specific branch, use the `--revision` parameter. For example, to download the 6.5 bpw branch:
|
73 |
+
|
74 |
+
Linux:
|
75 |
+
|
76 |
+
```shell
|
77 |
+
huggingface-cli download bartowski/codegemma-1.1-7b-it-exl2 --revision 6_5 --local-dir codegemma-1.1-7b-it-exl2-6_5 --local-dir-use-symlinks False
|
78 |
+
```
|
79 |
+
|
80 |
+
Windows (which apparently doesn't like _ in folders sometimes?):
|
81 |
+
|
82 |
+
```shell
|
83 |
+
huggingface-cli download bartowski/codegemma-1.1-7b-it-exl2 --revision 6_5 --local-dir codegemma-1.1-7b-it-exl2-6.5 --local-dir-use-symlinks False
|
84 |
+
```
|
85 |
+
|
86 |
+
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
|
|