bartowski commited on
Commit
2f5f063
1 Parent(s): 7ad7d1f

measurement.json

Browse files
Files changed (2) hide show
  1. README.md +83 -0
  2. measurement.json +0 -0
README.md ADDED
@@ -0,0 +1,83 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: other
3
+ base_model: meta-llama/Meta-Llama-3-8B
4
+ tags:
5
+ - generated_from_trainer
6
+ - axolotl
7
+ model-index:
8
+ - name: out
9
+ results: []
10
+ datasets:
11
+ - cognitivecomputations/Dolphin-2.9
12
+ - teknium/OpenHermes-2.5
13
+ - m-a-p/CodeFeedback-Filtered-Instruction
14
+ - cognitivecomputations/dolphin-coder
15
+ - cognitivecomputations/samantha-data
16
+ - microsoft/orca-math-word-problems-200k
17
+ - Locutusque/function-calling-chatml
18
+ - internlm/Agent-FLAN
19
+ quantized_by: bartowski
20
+ pipeline_tag: text-generation
21
+ ---
22
+
23
+ ## Exllama v2 Quantizations of dolphin-2.9.1-llama-3-8b
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/cognitivecomputations/dolphin-2.9.1-llama-3-8b
32
+
33
+ ## Prompt format
34
+
35
+ ```
36
+ <|im_start|>system
37
+ {system_prompt}<|im_end|>
38
+ <|im_start|>user
39
+ {prompt}<|im_end|>
40
+ <|im_start|>assistant
41
+
42
+ ```
43
+
44
+ ## Available sizes
45
+
46
+
47
+ | Branch | Bits | lm_head bits | VRAM (4k) | VRAM (8K) | VRAM (16k) | VRAM (32k) | Description |
48
+ | ----- | ---- | ------- | ------ | ------ | ------ | ------ | ------------ |
49
+ | [8_0](https://huggingface.co/bartowski/dolphin-2.9.1-llama-3-8b-exl2/tree/8_0) | 8.0 | 8.0 | 10.1 GB | 10.5 GB | 11.5 GB | 13.6 GB | Maximum quality that ExLlamaV2 can produce, near unquantized performance. |
50
+ | [6_5](https://huggingface.co/bartowski/dolphin-2.9.1-llama-3-8b-exl2/tree/6_5) | 6.5 | 8.0 | 8.9 GB | 9.3 GB | 10.3 GB | 12.4 GB | Very similar to 8.0, good tradeoff of size vs performance, **recommended**. |
51
+ | [5_0](https://huggingface.co/bartowski/dolphin-2.9.1-llama-3-8b-exl2/tree/5_0) | 5.0 | 6.0 | 7.7 GB | 8.1 GB | 9.1 GB | 11.2 GB | Slightly lower quality vs 6.5, but usable on 8GB cards. |
52
+ | [4_25](https://huggingface.co/bartowski/dolphin-2.9.1-llama-3-8b-exl2/tree/4_25) | 4.25 | 6.0 | 7.0 GB | 7.4 GB | 8.4 GB | 10.5 GB | GPTQ equivalent bits per weight, slightly higher quality. |
53
+ | [3_5](https://huggingface.co/bartowski/dolphin-2.9.1-llama-3-8b-exl2/tree/3_5) | 3.5 | 6.0 | 6.4 GB | 6.8 GB | 7.8 GB | 9.9 GB | Lower quality, only use if you have to. |
54
+
55
+ ## Download instructions
56
+
57
+ With git:
58
+
59
+ ```shell
60
+ git clone --single-branch --branch 6_5 https://huggingface.co/bartowski/dolphin-2.9.1-llama-3-8b-exl2 dolphin-2.9.1-llama-3-8b-exl2-6_5
61
+ ```
62
+
63
+ With huggingface hub (credit to TheBloke for instructions):
64
+
65
+ ```shell
66
+ pip3 install huggingface-hub
67
+ ```
68
+
69
+ To download a specific branch, use the `--revision` parameter. For example, to download the 6.5 bpw branch:
70
+
71
+ Linux:
72
+
73
+ ```shell
74
+ huggingface-cli download bartowski/dolphin-2.9.1-llama-3-8b-exl2 --revision 6_5 --local-dir dolphin-2.9.1-llama-3-8b-exl2-6_5 --local-dir-use-symlinks False
75
+ ```
76
+
77
+ Windows (which apparently doesn't like _ in folders sometimes?):
78
+
79
+ ```shell
80
+ huggingface-cli download bartowski/dolphin-2.9.1-llama-3-8b-exl2 --revision 6_5 --local-dir dolphin-2.9.1-llama-3-8b-exl2-6.5 --local-dir-use-symlinks False
81
+ ```
82
+
83
+ 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