Hugging Face's logo

legraphista
/
Meta-Llama-3-8B-Instruct-abliterated-v3-IMat-GGUF

Text Generation
GGUF
quantized
GGUF
imatrix
quantization
imat
static
16bit
8bit
6bit
5bit
4bit
3bit
2bit
1bit
Model card Files Community
legraphista commited on
Commit
0f1edc8
•
1 Parent(s): 22ef95e

Upload README.md with huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +161 -0
README.md ADDED
@@ -0,0 +1,161 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model: failspy/Meta-Llama-3-8B-Instruct-abliterated-v3
3
+ inference: false
4
+ library_name: gguf
5
+ license: llama3
6
+ pipeline_tag: text-generation
7
+ quantized_by: legraphista
8
+ tags:
9
+ - quantized
10
+ - GGUF
11
+ - imatrix
12
+ - quantization
13
+ - imat
14
+ - imatrix
15
+ - static
16
+ - 16bit
17
+ - 8bit
18
+ - 6bit
19
+ - 5bit
20
+ - 4bit
21
+ - 3bit
22
+ - 2bit
23
+ - 1bit
24
+ ---
25
+
26
+ # Meta-Llama-3-8B-Instruct-abliterated-v3-IMat-GGUF
27
+ _Llama.cpp imatrix quantization of failspy/Meta-Llama-3-8B-Instruct-abliterated-v3_
28
+
29
+ Original Model: [failspy/Meta-Llama-3-8B-Instruct-abliterated-v3](https://huggingface.co/failspy/Meta-Llama-3-8B-Instruct-abliterated-v3)
30
+ Original dtype: `BF16` (`bfloat16`)
31
+ Quantized by: llama.cpp [b3058](https://github.com/ggerganov/llama.cpp/releases/tag/b3058)
32
+ IMatrix dataset: [here](https://gist.githubusercontent.com/bartowski1182/eb213dccb3571f863da82e99418f81e8/raw/b2869d80f5c16fd7082594248e80144677736635/calibration_datav3.txt)
33
+
34
+ - [Files](#files)
35
+ - [IMatrix](#imatrix)
36
+ - [Common Quants](#common-quants)
37
+ - [All Quants](#all-quants)
38
+ - [Downloading using huggingface-cli](#downloading-using-huggingface-cli)
39
+ - [Inference](#inference)
40
+ - [Simple chat template](#simple-chat-template)
41
+ - [Chat template with system prompt](#chat-template-with-system-prompt)
42
+ - [Llama.cpp](#llama-cpp)
43
+ - [FAQ](#faq)
44
+ - [Why is the IMatrix not applied everywhere?](#why-is-the-imatrix-not-applied-everywhere)
45
+ - [How do I merge a split GGUF?](#how-do-i-merge-a-split-gguf)
46
+
47
+ ---
48
+
49
+ ## Files
50
+
51
+ ### IMatrix
52
+ Status: ⏳ Processing
53
+ Link: [here](https://huggingface.co/legraphista/Meta-Llama-3-8B-Instruct-abliterated-v3-IMat-GGUF/blob/main/imatrix.dat)
54
+
55
+ ### Common Quants
56
+ | Filename | Quant type | File Size | Status | Uses IMatrix | Is Split |
57
+ | -------- | ---------- | --------- | ------ | ------------ | -------- |
58
+ | Meta-Llama-3-8B-Instruct-abliterated-v3.Q8_0 | Q8_0 | - | ⏳ Processing | ⚪ Static | -
59
+ | Meta-Llama-3-8B-Instruct-abliterated-v3.Q6_K | Q6_K | - | ⏳ Processing | ⚪ Static | -
60
+ | Meta-Llama-3-8B-Instruct-abliterated-v3.Q4_K | Q4_K | - | ⏳ Processing | 🟢 IMatrix | -
61
+ | Meta-Llama-3-8B-Instruct-abliterated-v3.Q3_K | Q3_K | - | ⏳ Processing | 🟢 IMatrix | -
62
+ | Meta-Llama-3-8B-Instruct-abliterated-v3.Q2_K | Q2_K | - | ⏳ Processing | 🟢 IMatrix | -
63
+
64
+
65
+ ### All Quants
66
+ | Filename | Quant type | File Size | Status | Uses IMatrix | Is Split |
67
+ | -------- | ---------- | --------- | ------ | ------------ | -------- |
68
+ | Meta-Llama-3-8B-Instruct-abliterated-v3.BF16 | BF16 | - | ⏳ Processing | ⚪ Static | -
69
+ | Meta-Llama-3-8B-Instruct-abliterated-v3.FP16 | F16 | - | ⏳ Processing | ⚪ Static | -
70
+ | Meta-Llama-3-8B-Instruct-abliterated-v3.Q8_0 | Q8_0 | - | ⏳ Processing | ⚪ Static | -
71
+ | Meta-Llama-3-8B-Instruct-abliterated-v3.Q6_K | Q6_K | - | ⏳ Processing | ⚪ Static | -
72
+ | Meta-Llama-3-8B-Instruct-abliterated-v3.Q5_K | Q5_K | - | ⏳ Processing | ⚪ Static | -
73
+ | Meta-Llama-3-8B-Instruct-abliterated-v3.Q5_K_S | Q5_K_S | - | ⏳ Processing | ⚪ Static | -
74
+ | Meta-Llama-3-8B-Instruct-abliterated-v3.Q4_K | Q4_K | - | ⏳ Processing | 🟢 IMatrix | -
75
+ | Meta-Llama-3-8B-Instruct-abliterated-v3.Q4_K_S | Q4_K_S | - | ⏳ Processing | 🟢 IMatrix | -
76
+ | Meta-Llama-3-8B-Instruct-abliterated-v3.IQ4_NL | IQ4_NL | - | ⏳ Processing | 🟢 IMatrix | -
77
+ | Meta-Llama-3-8B-Instruct-abliterated-v3.IQ4_XS | IQ4_XS | - | ⏳ Processing | 🟢 IMatrix | -
78
+ | Meta-Llama-3-8B-Instruct-abliterated-v3.Q3_K | Q3_K | - | ⏳ Processing | 🟢 IMatrix | -
79
+ | Meta-Llama-3-8B-Instruct-abliterated-v3.Q3_K_L | Q3_K_L | - | ⏳ Processing | 🟢 IMatrix | -
80
+ | Meta-Llama-3-8B-Instruct-abliterated-v3.Q3_K_S | Q3_K_S | - | ⏳ Processing | 🟢 IMatrix | -
81
+ | Meta-Llama-3-8B-Instruct-abliterated-v3.IQ3_M | IQ3_M | - | ⏳ Processing | 🟢 IMatrix | -
82
+ | Meta-Llama-3-8B-Instruct-abliterated-v3.IQ3_S | IQ3_S | - | ⏳ Processing | 🟢 IMatrix | -
83
+ | Meta-Llama-3-8B-Instruct-abliterated-v3.IQ3_XS | IQ3_XS | - | ⏳ Processing | 🟢 IMatrix | -
84
+ | Meta-Llama-3-8B-Instruct-abliterated-v3.IQ3_XXS | IQ3_XXS | - | ⏳ Processing | 🟢 IMatrix | -
85
+ | Meta-Llama-3-8B-Instruct-abliterated-v3.Q2_K | Q2_K | - | ⏳ Processing | 🟢 IMatrix | -
86
+ | Meta-Llama-3-8B-Instruct-abliterated-v3.Q2_K_S | Q2_K_S | - | ⏳ Processing | 🟢 IMatrix | -
87
+ | Meta-Llama-3-8B-Instruct-abliterated-v3.IQ2_M | IQ2_M | - | ⏳ Processing | 🟢 IMatrix | -
88
+ | Meta-Llama-3-8B-Instruct-abliterated-v3.IQ2_S | IQ2_S | - | ⏳ Processing | 🟢 IMatrix | -
89
+ | Meta-Llama-3-8B-Instruct-abliterated-v3.IQ2_XS | IQ2_XS | - | ⏳ Processing | 🟢 IMatrix | -
90
+ | Meta-Llama-3-8B-Instruct-abliterated-v3.IQ2_XXS | IQ2_XXS | - | ⏳ Processing | 🟢 IMatrix | -
91
+ | Meta-Llama-3-8B-Instruct-abliterated-v3.IQ1_M | IQ1_M | - | ⏳ Processing | 🟢 IMatrix | -
92
+ | Meta-Llama-3-8B-Instruct-abliterated-v3.IQ1_S | IQ1_S | - | ⏳ Processing | 🟢 IMatrix | -
93
+
94
+
95
+ ## Downloading using huggingface-cli
96
+ If you do not have hugginface-cli installed:
97
+ ```
98
+ pip install -U "huggingface_hub[cli]"
99
+ ```
100
+ Download the specific file you want:
101
+ ```
102
+ huggingface-cli download legraphista/Meta-Llama-3-8B-Instruct-abliterated-v3-IMat-GGUF --include "Meta-Llama-3-8B-Instruct-abliterated-v3.Q8_0.gguf" --local-dir ./
103
+ ```
104
+ If the model file is big, it has been split into multiple files. In order to download them all to a local folder, run:
105
+ ```
106
+ huggingface-cli download legraphista/Meta-Llama-3-8B-Instruct-abliterated-v3-IMat-GGUF --include "Meta-Llama-3-8B-Instruct-abliterated-v3.Q8_0/*" --local-dir ./
107
+ # see FAQ for merging GGUF's
108
+ ```
109
+
110
+ ---
111
+
112
+ ## Inference
113
+
114
+ ### Simple chat template
115
+ ```
116
+ <|begin_of_text|><|start_header_id|>user<|end_header_id|>
117
+
118
+ {user_prompt}<|eot_id|><|start_header_id|>assistant<|end_header_id|>
119
+
120
+ {assistant_response}<|eot_id|><|start_header_id|>user<|end_header_id|>
121
+
122
+ {next_user_prompt}<|eot_id|>
123
+ ```
124
+
125
+ ### Chat template with system prompt
126
+ ```
127
+ <|begin_of_text|><|start_header_id|>system<|end_header_id|>
128
+
129
+ {system_prompt}<|eot_id|><|start_header_id|>user<|end_header_id|>
130
+
131
+ {user_prompt}<|eot_id|><|start_header_id|>assistant<|end_header_id|>
132
+
133
+ {assistant_response}<|eot_id|><|start_header_id|>user<|end_header_id|>
134
+
135
+ {next_user_prompt}<|eot_id|>
136
+ ```
137
+
138
+ ### Llama.cpp
139
+ ```
140
+ llama.cpp/main -m Meta-Llama-3-8B-Instruct-abliterated-v3.Q8_0.gguf --color -i -p "prompt here (according to the chat template)"
141
+ ```
142
+
143
+ ---
144
+
145
+ ## FAQ
146
+
147
+ ### Why is the IMatrix not applied everywhere?
148
+ According to [this investigation](https://www.reddit.com/r/LocalLLaMA/comments/1993iro/ggufs_quants_can_punch_above_their_weights_now/), it appears that lower quantizations are the only ones that benefit from the imatrix input (as per hellaswag results).
149
+
150
+ ### How do I merge a split GGUF?
151
+ 1. Make sure you have `gguf-split` available
152
+ - To get hold of `gguf-split`, navigate to https://github.com/ggerganov/llama.cpp/releases
153
+ - Download the appropriate zip for your system from the latest release
154
+ - Unzip the archive and you should be able to find `gguf-split`
155
+ 2. Locate your GGUF chunks folder (ex: `Meta-Llama-3-8B-Instruct-abliterated-v3.Q8_0`)
156
+ 3. Run `gguf-split --merge Meta-Llama-3-8B-Instruct-abliterated-v3.Q8_0/Meta-Llama-3-8B-Instruct-abliterated-v3.Q8_0-00001-of-XXXXX.gguf Meta-Llama-3-8B-Instruct-abliterated-v3.Q8_0.gguf`
157
+ - Make sure to point `gguf-split` to the first chunk of the split.
158
+
159
+ ---
160
+
161
+ Got a suggestion? Ping me [@legraphista](https://x.com/legraphista)!