mradermacher commited on
Commit
491430e
1 Parent(s): 079f28d

auto-patch README.md

Browse files
Files changed (1) hide show
  1. README.md +222 -0
README.md CHANGED
@@ -1,6 +1,228 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  <!-- ### quantize_version: 2 -->
2
  <!-- ### output_tensor_quantised: 1 -->
3
  <!-- ### convert_type: hf -->
4
  <!-- ### vocab_type: -->
5
  <!-- ### tags: nicoboss -->
6
  static quants of https://huggingface.co/ExAi/Meta-Llama-3-MoE-4x8B-Instruct
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model: ExAi/Meta-Llama-3-MoE-4x8B-Instruct
3
+ extra_gated_button_content: Submit
4
+ extra_gated_fields:
5
+ Affiliation: text
6
+ ? By clicking Submit below I accept the terms of the license and acknowledge that
7
+ the information I provide will be collected stored processed and shared in accordance
8
+ with the Meta Privacy Policy
9
+ : checkbox
10
+ Country: country
11
+ Date of birth: date_picker
12
+ First Name: text
13
+ Last Name: text
14
+ geo: ip_location
15
+ extra_gated_prompt: "### META LLAMA 3 COMMUNITY LICENSE AGREEMENT\nMeta Llama 3 Version
16
+ Release Date: April 18, 2024\n\"Agreement\" means the terms and conditions for use,
17
+ reproduction, distribution and modification of the Llama Materials set forth herein.\n\"Documentation\"
18
+ means the specifications, manuals and documentation accompanying Meta Llama 3 distributed
19
+ by Meta at https://llama.meta.com/get-started/.\n\"Licensee\" or \"you\" means you,
20
+ or your employer or any other person or entity (if you are entering into this Agreement
21
+ on such person or entity’s behalf), of the age required under applicable laws, rules
22
+ or regulations to provide legal consent and that has legal authority to bind your
23
+ employer or such other person or entity if you are entering in this Agreement on
24
+ their behalf.\n\"Meta Llama 3\" means the foundational large language models and
25
+ software and algorithms, including machine-learning model code, trained model weights,
26
+ inference-enabling code, training-enabling code, fine-tuning enabling code and other
27
+ elements of the foregoing distributed by Meta at https://llama.meta.com/llama-downloads.\n\"Llama
28
+ Materials\" means, collectively, Meta’s proprietary Meta Llama 3 and Documentation
29
+ (and any portion thereof) made available under this Agreement.\n\"Meta\" or \"we\"
30
+ means Meta Platforms Ireland Limited (if you are located in or, if you are an entity,
31
+ your principal place of business is in the EEA or Switzerland) and Meta Platforms,
32
+ Inc. (if you are located outside of the EEA or Switzerland).\n \n1. License Rights
33
+ and Redistribution.\na. Grant of Rights. You are granted a non-exclusive, worldwide,
34
+ non-transferable and royalty-free limited license under Meta’s intellectual property
35
+ or other rights owned by Meta embodied in the Llama Materials to use, reproduce,
36
+ distribute, copy, create derivative works of, and make modifications to the Llama
37
+ Materials.\nb. Redistribution and Use.\ni. If you distribute or make available the
38
+ Llama Materials (or any derivative works thereof), or a product or service that
39
+ uses any of them, including another AI model, you shall (A) provide a copy of this
40
+ Agreement with any such Llama Materials; and (B) prominently display “Built with
41
+ Meta Llama 3” on a related website, user interface, blogpost, about page, or product
42
+ documentation. If you use the Llama Materials to create, train, fine tune, or otherwise
43
+ improve an AI model, which is distributed or made available, you shall also include
44
+ “Llama 3” at the beginning of any such AI model name.\nii. If you receive Llama
45
+ Materials, or any derivative works thereof, from a Licensee as part of an integrated
46
+ end user product, then Section 2 of this Agreement will not apply to you.\niii.
47
+ You must retain in all copies of the Llama Materials that you distribute the following
48
+ attribution notice within a “Notice” text file distributed as a part of such copies:
49
+ “Meta Llama 3 is licensed under the Meta Llama 3 Community License, Copyright ©
50
+ Meta Platforms, Inc. All Rights Reserved.”\niv. Your use of the Llama Materials
51
+ must comply with applicable laws and regulations (including trade compliance laws
52
+ and regulations) and adhere to the Acceptable Use Policy for the Llama Materials
53
+ (available at https://llama.meta.com/llama3/use-policy), which is hereby incorporated
54
+ by reference into this Agreement.\nv. You will not use the Llama Materials or any
55
+ output or results of the Llama Materials to improve any other large language model
56
+ (excluding Meta Llama 3 or derivative works thereof).\n2. Additional Commercial
57
+ Terms. If, on the Meta Llama 3 version release date, the monthly active users of
58
+ the products or services made available by or for Licensee, or Licensee’s affiliates,
59
+ is greater than 700 million monthly active users in the preceding calendar month,
60
+ you must request a license from Meta, which Meta may grant to you in its sole discretion,
61
+ and you are not authorized to exercise any of the rights under this Agreement unless
62
+ or until Meta otherwise expressly grants you such rights.\n3. Disclaimer of Warranty.
63
+ UNLESS REQUIRED BY APPLICABLE LAW, THE LLAMA MATERIALS AND ANY OUTPUT AND RESULTS
64
+ THEREFROM ARE PROVIDED ON AN “AS IS” BASIS, WITHOUT WARRANTIES OF ANY KIND, AND
65
+ META DISCLAIMS ALL WARRANTIES OF ANY KIND, BOTH EXPRESS AND IMPLIED, INCLUDING,
66
+ WITHOUT LIMITATION, ANY WARRANTIES OF TITLE, NON-INFRINGEMENT, MERCHANTABILITY,
67
+ OR FITNESS FOR A PARTICULAR PURPOSE. YOU ARE SOLELY RESPONSIBLE FOR DETERMINING
68
+ THE APPROPRIATENESS OF USING OR REDISTRIBUTING THE LLAMA MATERIALS AND ASSUME ANY
69
+ RISKS ASSOCIATED WITH YOUR USE OF THE LLAMA MATERIALS AND ANY OUTPUT AND RESULTS.\n4.
70
+ Limitation of Liability. IN NO EVENT WILL META OR ITS AFFILIATES BE LIABLE UNDER
71
+ ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, TORT, NEGLIGENCE, PRODUCTS LIABILITY,
72
+ OR OTHERWISE, ARISING OUT OF THIS AGREEMENT, FOR ANY LOST PROFITS OR ANY INDIRECT,
73
+ SPECIAL, CONSEQUENTIAL, INCIDENTAL, EXEMPLARY OR PUNITIVE DAMAGES, EVEN IF META
74
+ OR ITS AFFILIATES HAVE BEEN ADVISED OF THE POSSIBILITY OF ANY OF THE FOREGOING.\n5.
75
+ Intellectual Property.\na. No trademark licenses are granted under this Agreement,
76
+ and in connection with the Llama Materials, neither Meta nor Licensee may use any
77
+ name or mark owned by or associated with the other or any of its affiliates, except
78
+ as required for reasonable and customary use in describing and redistributing the
79
+ Llama Materials or as set forth in this Section 5(a). Meta hereby grants you a license
80
+ to use “Llama 3” (the “Mark”) solely as required to comply with the last sentence
81
+ of Section 1.b.i. You will comply with Meta’s brand guidelines (currently accessible
82
+ at https://about.meta.com/brand/resources/meta/company-brand/ ). All goodwill arising
83
+ out of your use of the Mark will inure to the benefit of Meta.\nb. Subject to Meta’s
84
+ ownership of Llama Materials and derivatives made by or for Meta, with respect to
85
+ any derivative works and modifications of the Llama Materials that are made by you,
86
+ as between you and Meta, you are and will be the owner of such derivative works
87
+ and modifications.\nc. If you institute litigation or other proceedings against
88
+ Meta or any entity (including a cross-claim or counterclaim in a lawsuit) alleging
89
+ that the Llama Materials or Meta Llama 3 outputs or results, or any portion of any
90
+ of the foregoing, constitutes infringement of intellectual property or other rights
91
+ owned or licensable by you, then any licenses granted to you under this Agreement
92
+ shall terminate as of the date such litigation or claim is filed or instituted.
93
+ You will indemnify and hold harmless Meta from and against any claim by any third
94
+ party arising out of or related to your use or distribution of the Llama Materials.\n6.
95
+ Term and Termination. The term of this Agreement will commence upon your acceptance
96
+ of this Agreement or access to the Llama Materials and will continue in full force
97
+ and effect until terminated in accordance with the terms and conditions herein.
98
+ Meta may terminate this Agreement if you are in breach of any term or condition
99
+ of this Agreement. Upon termination of this Agreement, you shall delete and cease
100
+ use of the Llama Materials. Sections 3, 4 and 7 shall survive the termination of
101
+ this Agreement.\n7. Governing Law and Jurisdiction. This Agreement will be governed
102
+ and construed under the laws of the State of California without regard to choice
103
+ of law principles, and the UN Convention on Contracts for the International Sale
104
+ of Goods does not apply to this Agreement. The courts of California shall have exclusive
105
+ jurisdiction of any dispute arising out of this Agreement.\n### Meta Llama 3 Acceptable
106
+ Use Policy\nMeta is committed to promoting safe and fair use of its tools and features,
107
+ including Meta Llama 3. If you access or use Meta Llama 3, you agree to this Acceptable
108
+ Use Policy (“Policy”). The most recent copy of this policy can be found at [https://llama.meta.com/llama3/use-policy](https://llama.meta.com/llama3/use-policy)\n####
109
+ Prohibited Uses\nWe want everyone to use Meta Llama 3 safely and responsibly. You
110
+ agree you will not use, or allow others to use, Meta Llama 3 to: 1. Violate the
111
+ law or others’ rights, including to:\n 1. Engage in, promote, generate, contribute
112
+ to, encourage, plan, incite, or further illegal or unlawful activity or content,
113
+ such as:\n 1. Violence or terrorism\n 2. Exploitation or harm to children,
114
+ including the solicitation, creation, acquisition, or dissemination of child exploitative
115
+ content or failure to report Child Sexual Abuse Material\n 3. Human trafficking,
116
+ exploitation, and sexual violence\n 4. The illegal distribution of information
117
+ or materials to minors, including obscene materials, or failure to employ legally
118
+ required age-gating in connection with such information or materials.\n 5.
119
+ Sexual solicitation\n 6. Any other criminal activity\n 2. Engage in, promote,
120
+ incite, or facilitate the harassment, abuse, threatening, or bullying of individuals
121
+ or groups of individuals\n 3. Engage in, promote, incite, or facilitate discrimination
122
+ or other unlawful or harmful conduct in the provision of employment, employment
123
+ benefits, credit, housing, other economic benefits, or other essential goods and
124
+ services\n 4. Engage in the unauthorized or unlicensed practice of any profession
125
+ including, but not limited to, financial, legal, medical/health, or related professional
126
+ practices\n 5. Collect, process, disclose, generate, or infer health, demographic,
127
+ or other sensitive personal or private information about individuals without rights
128
+ and consents required by applicable laws\n 6. Engage in or facilitate any action
129
+ or generate any content that infringes, misappropriates, or otherwise violates any
130
+ third-party rights, including the outputs or results of any products or services
131
+ using the Llama Materials\n 7. Create, generate, or facilitate the creation of
132
+ malicious code, malware, computer viruses or do anything else that could disable,
133
+ overburden, interfere with or impair the proper working, integrity, operation or
134
+ appearance of a website or computer system\n2. Engage in, promote, incite, facilitate,
135
+ or assist in the planning or development of activities that present a risk of death
136
+ or bodily harm to individuals, including use of Meta Llama 3 related to the following:\n
137
+ \ 1. Military, warfare, nuclear industries or applications, espionage, use for
138
+ materials or activities that are subject to the International Traffic Arms Regulations
139
+ (ITAR) maintained by the United States Department of State\n 2. Guns and illegal
140
+ weapons (including weapon development)\n 3. Illegal drugs and regulated/controlled
141
+ substances\n 4. Operation of critical infrastructure, transportation technologies,
142
+ or heavy machinery\n 5. Self-harm or harm to others, including suicide, cutting,
143
+ and eating disorders\n 6. Any content intended to incite or promote violence,
144
+ abuse, or any infliction of bodily harm to an individual\n3. Intentionally deceive
145
+ or mislead others, including use of Meta Llama 3 related to the following:\n 1.
146
+ Generating, promoting, or furthering fraud or the creation or promotion of disinformation\n
147
+ \ 2. Generating, promoting, or furthering defamatory content, including the creation
148
+ of defamatory statements, images, or other content\n 3. Generating, promoting,
149
+ or further distributing spam\n 4. Impersonating another individual without consent,
150
+ authorization, or legal right\n 5. Representing that the use of Meta Llama 3
151
+ or outputs are human-generated\n 6. Generating or facilitating false online engagement,
152
+ including fake reviews and other means of fake online engagement\n4. Fail to appropriately
153
+ disclose to end users any known dangers of your AI system\nPlease report any violation
154
+ of this Policy, software “bug,” or other problems that could lead to a violation
155
+ of this Policy through one of the following means:\n * Reporting issues with
156
+ the model: [https://github.com/meta-llama/llama3](https://github.com/meta-llama/llama3)\n
157
+ \ * Reporting risky content generated by the model:\n developers.facebook.com/llama_output_feedback\n
158
+ \ * Reporting bugs and security concerns: facebook.com/whitehat/info\n * Reporting
159
+ violations of the Acceptable Use Policy or unlicensed uses of Meta Llama 3: LlamaUseReport@meta.com"
160
+ language:
161
+ - en
162
+ library_name: transformers
163
+ license: other
164
+ license_link: LICENSE
165
+ license_name: llama3
166
+ quantized_by: mradermacher
167
+ tags:
168
+ - facebook
169
+ - meta
170
+ - pytorch
171
+ - llama
172
+ - llama-3
173
+ ---
174
+ ## About
175
+
176
  <!-- ### quantize_version: 2 -->
177
  <!-- ### output_tensor_quantised: 1 -->
178
  <!-- ### convert_type: hf -->
179
  <!-- ### vocab_type: -->
180
  <!-- ### tags: nicoboss -->
181
  static quants of https://huggingface.co/ExAi/Meta-Llama-3-MoE-4x8B-Instruct
182
+
183
+ <!-- provided-files -->
184
+ weighted/imatrix quants seem not to be available (by me) at this time. If they do not show up a week or so after the static ones, I have probably not planned for them. Feel free to request them by opening a Community Discussion.
185
+ ## Usage
186
+
187
+ If you are unsure how to use GGUF files, refer to one of [TheBloke's
188
+ READMEs](https://huggingface.co/TheBloke/KafkaLM-70B-German-V0.1-GGUF) for
189
+ more details, including on how to concatenate multi-part files.
190
+
191
+ ## Provided Quants
192
+
193
+ (sorted by size, not necessarily quality. IQ-quants are often preferable over similar sized non-IQ quants)
194
+
195
+ | Link | Type | Size/GB | Notes |
196
+ |:-----|:-----|--------:|:------|
197
+ | [GGUF](https://huggingface.co/mradermacher/Meta-Llama-3-MoE-4x8B-Instruct-GGUF/resolve/main/Meta-Llama-3-MoE-4x8B-Instruct.Q2_K.gguf) | Q2_K | 9.4 | |
198
+ | [GGUF](https://huggingface.co/mradermacher/Meta-Llama-3-MoE-4x8B-Instruct-GGUF/resolve/main/Meta-Llama-3-MoE-4x8B-Instruct.Q3_K_S.gguf) | Q3_K_S | 11.0 | |
199
+ | [GGUF](https://huggingface.co/mradermacher/Meta-Llama-3-MoE-4x8B-Instruct-GGUF/resolve/main/Meta-Llama-3-MoE-4x8B-Instruct.Q3_K_M.gguf) | Q3_K_M | 12.2 | lower quality |
200
+ | [GGUF](https://huggingface.co/mradermacher/Meta-Llama-3-MoE-4x8B-Instruct-GGUF/resolve/main/Meta-Llama-3-MoE-4x8B-Instruct.Q3_K_L.gguf) | Q3_K_L | 13.1 | |
201
+ | [GGUF](https://huggingface.co/mradermacher/Meta-Llama-3-MoE-4x8B-Instruct-GGUF/resolve/main/Meta-Llama-3-MoE-4x8B-Instruct.IQ4_XS.gguf) | IQ4_XS | 13.7 | |
202
+ | [GGUF](https://huggingface.co/mradermacher/Meta-Llama-3-MoE-4x8B-Instruct-GGUF/resolve/main/Meta-Llama-3-MoE-4x8B-Instruct.Q4_K_S.gguf) | Q4_K_S | 14.4 | fast, recommended |
203
+ | [GGUF](https://huggingface.co/mradermacher/Meta-Llama-3-MoE-4x8B-Instruct-GGUF/resolve/main/Meta-Llama-3-MoE-4x8B-Instruct.Q4_K_M.gguf) | Q4_K_M | 15.3 | fast, recommended |
204
+ | [GGUF](https://huggingface.co/mradermacher/Meta-Llama-3-MoE-4x8B-Instruct-GGUF/resolve/main/Meta-Llama-3-MoE-4x8B-Instruct.Q5_K_S.gguf) | Q5_K_S | 17.3 | |
205
+ | [GGUF](https://huggingface.co/mradermacher/Meta-Llama-3-MoE-4x8B-Instruct-GGUF/resolve/main/Meta-Llama-3-MoE-4x8B-Instruct.Q5_K_M.gguf) | Q5_K_M | 17.8 | |
206
+ | [GGUF](https://huggingface.co/mradermacher/Meta-Llama-3-MoE-4x8B-Instruct-GGUF/resolve/main/Meta-Llama-3-MoE-4x8B-Instruct.Q6_K.gguf) | Q6_K | 20.6 | very good quality |
207
+ | [GGUF](https://huggingface.co/mradermacher/Meta-Llama-3-MoE-4x8B-Instruct-GGUF/resolve/main/Meta-Llama-3-MoE-4x8B-Instruct.Q8_0.gguf) | Q8_0 | 26.6 | fast, best quality |
208
+
209
+ Here is a handy graph by ikawrakow comparing some lower-quality quant
210
+ types (lower is better):
211
+
212
+ ![image.png](https://www.nethype.de/huggingface_embed/quantpplgraph.png)
213
+
214
+ And here are Artefact2's thoughts on the matter:
215
+ https://gist.github.com/Artefact2/b5f810600771265fc1e39442288e8ec9
216
+
217
+ ## FAQ / Model Request
218
+
219
+ See https://huggingface.co/mradermacher/model_requests for some answers to
220
+ questions you might have and/or if you want some other model quantized.
221
+
222
+ ## Thanks
223
+
224
+ I thank my company, [nethype GmbH](https://www.nethype.de/), for letting
225
+ me use its servers and providing upgrades to my workstation to enable
226
+ this work in my free time. Additional thanks to [@nicoboss](https://huggingface.co/nicoboss) for giving me access to his private supercomputer, enabling me to provide many more imatrix quants, at much higher quality, than I would otherwise be able to.
227
+
228
+ <!-- end -->