legraphista commited on
Commit
1c354b3
1 Parent(s): b438516

Upload README.md with huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +258 -0
README.md ADDED
@@ -0,0 +1,258 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model: facebook/llm-compiler-7b-ftd
3
+ extra_gated_button_content: I Accept Meta LLM Compiler License and AUP
4
+ extra_gated_description: The information you provide will be collected, stored, processed
5
+ and shared in accordance with the [Meta Privacy Policy](https://www.facebook.com/privacy/policy/).
6
+ extra_gated_fields:
7
+ Affiliation: text
8
+ ? By clicking Submit below I accept the terms of the license and acknowledge that
9
+ the information I provide will be collected stored processed and shared in accordance
10
+ with the Meta Privacy Policy
11
+ : checkbox
12
+ Country: country
13
+ Date of birth: date_picker
14
+ First Name: text
15
+ I accept the terms and conditions: checkbox
16
+ Last Name: text
17
+ geo: ip_location
18
+ extra_gated_prompt: "**Meta Large Language Model Compiler (LLM Compiler) LICENSE AGREEMENT**\n\
19
+ Version Release Date: 27th June 2024\n\u201C**Agreement**\u201D means the terms\
20
+ \ and conditions for use, reproduction, distribution and modification of the LLM\
21
+ \ Compiler Materials set forth herein.\n\u201C**Documentation**\u201D means the\
22
+ \ specifications, manuals and documentation accompanying the LLM Compiler distributed\
23
+ \ by Meta at:\n* [https://huggingface.co/facebook/llm-compiler-7b](https://huggingface.co/facebook/llm-compiler-7b)\
24
+ \ * [https://huggingface.co/facebook/llm-compiler-7b-ftd](https://huggingface.co/facebook/llm-compiler-7b-ftd)\
25
+ \ * [https://huggingface.co/facebook/llm-compiler-13b](https://huggingface.co/facebook/llm-compiler-13b)\
26
+ \ * [https://huggingface.co/facebook/llm-compiler-13b-ftd](https://huggingface.co/facebook/llm-compiler-13b-ftd)\n\
27
+ \u201C**Licensee**\u201D or \u201C**you**\u201D means you, or your employer or any\
28
+ \ other person or entity (if you are entering into this Agreement on such person\
29
+ \ or entity\u2019s behalf), of the age required under applicable laws, rules or\
30
+ \ regulations to provide legal consent and that has legal authority to bind your\
31
+ \ employer or such other person or entity if you are entering in this Agreement\
32
+ \ on their behalf.\n\u201C**Meta Large Language Model Compiler\u201D and \u201C\
33
+ LLM Compiler**\u201D mean the foundational large language models and software and\
34
+ \ algorithms, including machine-learning model code, trained model weights, inference-enabling\
35
+ \ code, training-enabling code, fine-tuning enabling code and other elements of\
36
+ \ the foregoing distributed by Meta at:\n* [https://huggingface.co/facebook/llm-compiler-7b](https://huggingface.co/facebook/llm-compiler-7b)\
37
+ \ * [https://huggingface.co/facebook/llm-compiler-7b-ftd](https://huggingface.co/facebook/llm-compiler-7b-ftd)\
38
+ \ * [https://huggingface.co/facebook/llm-compiler-13b](https://huggingface.co/facebook/llm-compiler-13b)\
39
+ \ * [https://huggingface.co/facebook/llm-compiler-13b-ftd](https://huggingface.co/facebook/llm-compiler-13b-ftd)\n\
40
+ \u201C**LLM Compiler Materials**\u201D means, collectively, Meta\u2019s proprietary\
41
+ \ LLM Compiler and Documentation (and any portion thereof) made available under\
42
+ \ this Agreement.\n\u201C**Meta**\u201D or \u201C**we**\u201D means Meta Platforms\
43
+ \ Ireland Limited (if you are located in or, if you are an entity, your principal\
44
+ \ place of business is in the EEA or Switzerland) and Meta Platforms, Inc. (if you\
45
+ \ are located outside of the EEA or Switzerland). \nBy clicking \u201CI Accept\u201D\
46
+ \ below or by using or distributing any portion or element of the LLM Compiler Materials,\
47
+ \ you agree to be bound by this Agreement.\n1. **License Rights and Redistribution**.\
48
+ \ \\\n\n a. <span style=\"text-decoration:underline;\">Grant of Rights</span>.\
49
+ \ You are granted a non-exclusive, worldwide, non-transferable and royalty-free\
50
+ \ limited license under Meta\u2019s intellectual property or other rights owned\
51
+ \ by Meta embodied in the LLM Compiler Materials to use, reproduce, distribute,\
52
+ \ copy, create derivative works of, and make modifications to the LLM Compiler Materials.\
53
+ \ \n\n b. <span style=\"text-decoration:underline;\">Redistribution and Use</span>.\
54
+ \ \n\n i. If you distribute or make available the LLM Compiler Materials (or\
55
+ \ any derivative works thereof), or a product or service that uses any of them,\
56
+ \ including another AI model, you shall (A) provide a copy of this Agreement with\
57
+ \ any such LLM Compiler Materials; and (B) prominently display \u201CBuilt with\
58
+ \ LLM Compiler\u201D on a related website, user interface, blogpost, about page,\
59
+ \ or product documentation. If you use the LLM Compiler Materials to create, train,\
60
+ \ fine tune, or otherwise improve an AI model, which is distributed or made available,\
61
+ \ you shall also include \u201CLLM Compiler\u201D at the beginning of any such AI\
62
+ \ model name.\n\n ii. If you receive LLM Compiler Materials, or any derivative\
63
+ \ works thereof, from a Licensee as part of an integrated end user product, then\
64
+ \ Section 2 of this Agreement will not apply to you. \n\n iii. You must retain\
65
+ \ in all copies of the LLM Compiler Materials that you distribute the following\
66
+ \ attribution notice within a \u201CNotice\u201D text file distributed as a part\
67
+ \ of such copies: \u201CLLM Compiler is licensed under the LLM Compiler License,\
68
+ \ Copyright \xA9 Meta Platforms, Inc. All Rights Reserved.\u201D\n\n iv. Your\
69
+ \ use of the LLM Compiler Materials must comply with applicable laws and regulations\
70
+ \ (including trade compliance laws and regulations) and adhere to the Acceptable\
71
+ \ Use Policy for Llama Materials (available at https://llama.meta.com/llama3/use-policy),\
72
+ \ which is hereby incorporated by reference into this Agreement.\n\n v. You will\
73
+ \ not use the LLM Compiler Materials or any output or results of the LLM Compiler\
74
+ \ Materials to improve any other large language model. \n\n2. **Additional Commercial\
75
+ \ Terms**. If, on the LLM Compiler release date, the monthly active users of the\
76
+ \ products or services made available by or for Licensee, or Licensee\u2019s affiliates,\
77
+ \ is greater than 700 million monthly active users in the preceding calendar month,\
78
+ \ you must request a license from Meta, which Meta may grant to you in its sole\
79
+ \ discretion, and you are not authorized to exercise any of the rights under this\
80
+ \ Agreement unless or until Meta otherwise expressly grants you such rights. \n\
81
+ 3**. Disclaimer of Warranty**. UNLESS REQUIRED BY APPLICABLE LAW, THE LLM COMPILER\
82
+ \ MATERIALS AND ANY OUTPUT AND RESULTS THEREFROM ARE PROVIDED ON AN \u201CAS IS\u201D\
83
+ \ BASIS, WITHOUT WARRANTIES OF ANY KIND, AND META DISCLAIMS ALL WARRANTIES OF ANY\
84
+ \ KIND, BOTH EXPRESS AND IMPLIED, INCLUDING, WITHOUT LIMITATION, ANY WARRANTIES\
85
+ \ OF TITLE, NON-INFRINGEMENT, MERCHANTABILITY, OR FITNESS FOR A PARTICULAR PURPOSE.\
86
+ \ YOU ARE SOLELY RESPONSIBLE FOR DETERMINING THE APPROPRIATENESS OF USING OR REDISTRIBUTING\
87
+ \ THE LLM COMPILER MATERIALS AND ASSUME ANY RISKS ASSOCIATED WITH YOUR USE OF THE\
88
+ \ LLM COMPILER MATERIALS AND ANY OUTPUT AND RESULTS.\n4. **Limitation of Liability**.\
89
+ \ IN NO EVENT WILL META OR ITS AFFILIATES BE LIABLE UNDER ANY THEORY OF LIABILITY,\
90
+ \ WHETHER IN CONTRACT, TORT, NEGLIGENCE, PRODUCTS LIABILITY, OR OTHERWISE, ARISING\
91
+ \ OUT OF THIS AGREEMENT, FOR ANY LOST PROFITS OR ANY INDIRECT, SPECIAL, CONSEQUENTIAL,\
92
+ \ INCIDENTAL, EXEMPLARY OR PUNITIVE DAMAGES, EVEN IF META OR ITS AFFILIATES HAVE\
93
+ \ BEEN ADVISED OF THE POSSIBILITY OF ANY OF THE FOREGOING.\n5. **Intellectual Property**.\n\
94
+ \n a. No trademark licenses are granted under this Agreement, and in connection\
95
+ \ with the LLM Compiler Materials, neither Meta nor Licensee may use any name or\
96
+ \ mark owned by or associated with the other or any of its affiliates, except as\
97
+ \ required for reasonable and customary use in describing and redistributing the\
98
+ \ LLM Compiler Materials or as set forth in this Section 5(a). Meta hereby grants\
99
+ \ you a license to use LLM Compiler (the \u201CMark\u201D) solely as required to\
100
+ \ comply with the last sentence of Section 1.b.i. You will comply with Meta\u2019\
101
+ s brand guidelines (currently accessible at[ https://about.meta.com/brand/resources/meta/company-brand/)](https://about.meta.com/brand/resources/meta/company-brand/).\
102
+ \ All goodwill arising out of your use of the Mark will inure to the benefit of\
103
+ \ Meta. \n\n b. Subject to Meta\u2019s ownership of LLM Compiler Materials and\
104
+ \ derivatives made by or for Meta, with respect to any derivative works and modifications\
105
+ \ of the LLM Compiler Materials that are made by you, as between you and Meta, you\
106
+ \ are and will be the owner of such derivative works and modifications.\n\n c.\
107
+ \ If you institute litigation or other proceedings against Meta or any entity (including\
108
+ \ a cross-claim or counterclaim in a lawsuit) alleging that the LLM Compiler Materials\
109
+ \ or LLM Compiler outputs or results, or any portion of any of the foregoing, constitutes\
110
+ \ infringement of intellectual property or other rights owned or licensable by you,\
111
+ \ then any licenses granted to you under this Agreement shall terminate as of the\
112
+ \ date such litigation or claim is filed or instituted. You will indemnify and hold\
113
+ \ harmless Meta from and against any claim by any third party arising out of or\
114
+ \ related to your use or distribution of the LLM Compiler Materials.\n\n6. **Term\
115
+ \ and Termination**. The term of this Agreement will commence upon your acceptance\
116
+ \ of this Agreement or access to the LLM Compiler Materials and will continue in\
117
+ \ full force and effect until terminated in accordance with the terms and conditions\
118
+ \ herein. Meta may terminate this Agreement if you are in breach of any term or\
119
+ \ condition of this Agreement. Upon termination of this Agreement, you shall delete\
120
+ \ and cease use of the LLM Compiler Materials. Sections 3, 4 and 7 shall survive\
121
+ \ the termination of this Agreement. \n7. **Governing Law and Jurisdiction**. This\
122
+ \ Agreement will be governed and construed under the laws of the State of California\
123
+ \ without regard to choice of law principles, and the UN Convention on Contracts\
124
+ \ for the International Sale of Goods does not apply to this Agreement. The courts\
125
+ \ of California shall have exclusive jurisdiction of any dispute arising out of\
126
+ \ this Agreement. "
127
+ inference: false
128
+ library_name: gguf
129
+ license: other
130
+ pipeline_tag: text-generation
131
+ quantized_by: legraphista
132
+ tags:
133
+ - quantized
134
+ - GGUF
135
+ - quantization
136
+ - imat
137
+ - imatrix
138
+ - static
139
+ - 16bit
140
+ - 8bit
141
+ - 6bit
142
+ - 5bit
143
+ - 4bit
144
+ - 3bit
145
+ - 2bit
146
+ - 1bit
147
+ ---
148
+
149
+ # llm-compiler-7b-ftd-IMat-GGUF
150
+ _Llama.cpp imatrix quantization of facebook/llm-compiler-7b-ftd_
151
+
152
+ Original Model: [facebook/llm-compiler-7b-ftd](https://huggingface.co/facebook/llm-compiler-7b-ftd)
153
+ Original dtype: `BF16` (`bfloat16`)
154
+ Quantized by: llama.cpp [b3256](https://github.com/ggerganov/llama.cpp/releases/tag/b3256)
155
+ IMatrix dataset: [here](https://gist.githubusercontent.com/bartowski1182/eb213dccb3571f863da82e99418f81e8/raw/b2869d80f5c16fd7082594248e80144677736635/calibration_datav3.txt)
156
+
157
+ - [Files](#files)
158
+ - [IMatrix](#imatrix)
159
+ - [Common Quants](#common-quants)
160
+ - [All Quants](#all-quants)
161
+ - [Downloading using huggingface-cli](#downloading-using-huggingface-cli)
162
+ - [Inference](#inference)
163
+ - [Llama.cpp](#llama-cpp)
164
+ - [FAQ](#faq)
165
+ - [Why is the IMatrix not applied everywhere?](#why-is-the-imatrix-not-applied-everywhere)
166
+ - [How do I merge a split GGUF?](#how-do-i-merge-a-split-gguf)
167
+
168
+ ---
169
+
170
+ ## Files
171
+
172
+ ### IMatrix
173
+ Status: ⏳ Processing
174
+ Link: [here](https://huggingface.co/legraphista/llm-compiler-7b-ftd-IMat-GGUF/blob/main/imatrix.dat)
175
+
176
+ ### Common Quants
177
+ | Filename | Quant type | File Size | Status | Uses IMatrix | Is Split |
178
+ | -------- | ---------- | --------- | ------ | ------------ | -------- |
179
+ | llm-compiler-7b-ftd.Q8_0 | Q8_0 | - | ⏳ Processing | ⚪ Static | -
180
+ | llm-compiler-7b-ftd.Q6_K | Q6_K | - | ⏳ Processing | ⚪ Static | -
181
+ | llm-compiler-7b-ftd.Q4_K | Q4_K | - | ⏳ Processing | 🟢 IMatrix | -
182
+ | llm-compiler-7b-ftd.Q3_K | Q3_K | - | ⏳ Processing | 🟢 IMatrix | -
183
+ | llm-compiler-7b-ftd.Q2_K | Q2_K | - | ⏳ Processing | 🟢 IMatrix | -
184
+
185
+
186
+ ### All Quants
187
+ | Filename | Quant type | File Size | Status | Uses IMatrix | Is Split |
188
+ | -------- | ---------- | --------- | ------ | ------------ | -------- |
189
+ | llm-compiler-7b-ftd.BF16 | BF16 | - | ⏳ Processing | ⚪ Static | -
190
+ | llm-compiler-7b-ftd.FP16 | F16 | - | ⏳ Processing | ⚪ Static | -
191
+ | llm-compiler-7b-ftd.Q8_0 | Q8_0 | - | ⏳ Processing | ⚪ Static | -
192
+ | llm-compiler-7b-ftd.Q6_K | Q6_K | - | ⏳ Processing | ⚪ Static | -
193
+ | llm-compiler-7b-ftd.Q5_K | Q5_K | - | ⏳ Processing | ⚪ Static | -
194
+ | llm-compiler-7b-ftd.Q5_K_S | Q5_K_S | - | ⏳ Processing | ⚪ Static | -
195
+ | llm-compiler-7b-ftd.Q4_K | Q4_K | - | ⏳ Processing | 🟢 IMatrix | -
196
+ | llm-compiler-7b-ftd.Q4_K_S | Q4_K_S | - | ⏳ Processing | 🟢 IMatrix | -
197
+ | llm-compiler-7b-ftd.IQ4_NL | IQ4_NL | - | ⏳ Processing | 🟢 IMatrix | -
198
+ | llm-compiler-7b-ftd.IQ4_XS | IQ4_XS | - | ⏳ Processing | 🟢 IMatrix | -
199
+ | llm-compiler-7b-ftd.Q3_K | Q3_K | - | ⏳ Processing | 🟢 IMatrix | -
200
+ | llm-compiler-7b-ftd.Q3_K_L | Q3_K_L | - | ⏳ Processing | 🟢 IMatrix | -
201
+ | llm-compiler-7b-ftd.Q3_K_S | Q3_K_S | - | ⏳ Processing | 🟢 IMatrix | -
202
+ | llm-compiler-7b-ftd.IQ3_M | IQ3_M | - | ⏳ Processing | 🟢 IMatrix | -
203
+ | llm-compiler-7b-ftd.IQ3_S | IQ3_S | - | ⏳ Processing | 🟢 IMatrix | -
204
+ | llm-compiler-7b-ftd.IQ3_XS | IQ3_XS | - | ⏳ Processing | 🟢 IMatrix | -
205
+ | llm-compiler-7b-ftd.IQ3_XXS | IQ3_XXS | - | ⏳ Processing | 🟢 IMatrix | -
206
+ | llm-compiler-7b-ftd.Q2_K | Q2_K | - | ⏳ Processing | 🟢 IMatrix | -
207
+ | llm-compiler-7b-ftd.Q2_K_S | Q2_K_S | - | ⏳ Processing | 🟢 IMatrix | -
208
+ | llm-compiler-7b-ftd.IQ2_M | IQ2_M | - | ⏳ Processing | 🟢 IMatrix | -
209
+ | llm-compiler-7b-ftd.IQ2_S | IQ2_S | - | ⏳ Processing | 🟢 IMatrix | -
210
+ | llm-compiler-7b-ftd.IQ2_XS | IQ2_XS | - | ⏳ Processing | 🟢 IMatrix | -
211
+ | llm-compiler-7b-ftd.IQ2_XXS | IQ2_XXS | - | ⏳ Processing | 🟢 IMatrix | -
212
+ | llm-compiler-7b-ftd.IQ1_M | IQ1_M | - | ⏳ Processing | 🟢 IMatrix | -
213
+ | llm-compiler-7b-ftd.IQ1_S | IQ1_S | - | ⏳ Processing | 🟢 IMatrix | -
214
+
215
+
216
+ ## Downloading using huggingface-cli
217
+ If you do not have hugginface-cli installed:
218
+ ```
219
+ pip install -U "huggingface_hub[cli]"
220
+ ```
221
+ Download the specific file you want:
222
+ ```
223
+ huggingface-cli download legraphista/llm-compiler-7b-ftd-IMat-GGUF --include "llm-compiler-7b-ftd.Q8_0.gguf" --local-dir ./
224
+ ```
225
+ If the model file is big, it has been split into multiple files. In order to download them all to a local folder, run:
226
+ ```
227
+ huggingface-cli download legraphista/llm-compiler-7b-ftd-IMat-GGUF --include "llm-compiler-7b-ftd.Q8_0/*" --local-dir ./
228
+ # see FAQ for merging GGUF's
229
+ ```
230
+
231
+ ---
232
+
233
+ ## Inference
234
+
235
+ ### Llama.cpp
236
+ ```
237
+ llama.cpp/main -m llm-compiler-7b-ftd.Q8_0.gguf --color -i -p "prompt here"
238
+ ```
239
+
240
+ ---
241
+
242
+ ## FAQ
243
+
244
+ ### Why is the IMatrix not applied everywhere?
245
+ 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).
246
+
247
+ ### How do I merge a split GGUF?
248
+ 1. Make sure you have `gguf-split` available
249
+ - To get hold of `gguf-split`, navigate to https://github.com/ggerganov/llama.cpp/releases
250
+ - Download the appropriate zip for your system from the latest release
251
+ - Unzip the archive and you should be able to find `gguf-split`
252
+ 2. Locate your GGUF chunks folder (ex: `llm-compiler-7b-ftd.Q8_0`)
253
+ 3. Run `gguf-split --merge llm-compiler-7b-ftd.Q8_0/llm-compiler-7b-ftd.Q8_0-00001-of-XXXXX.gguf llm-compiler-7b-ftd.Q8_0.gguf`
254
+ - Make sure to point `gguf-split` to the first chunk of the split.
255
+
256
+ ---
257
+
258
+ Got a suggestion? Ping me [@legraphista](https://x.com/legraphista)!