TheBloke commited on
Commit
e2cbcef
1 Parent(s): 245bc51

Upload README.md

Browse files
Files changed (1) hide show
  1. README.md +103 -26
README.md CHANGED
@@ -1,13 +1,27 @@
1
  ---
 
2
  inference: false
3
  language:
4
  - en
5
- license: other
6
  model_creator: Meta Llama 2
7
- model_link: https://huggingface.co/meta-llama/Llama-2-13b-chat-hf
8
  model_name: Llama 2 13B Chat
9
  model_type: llama
10
  pipeline_tag: text-generation
 
 
 
 
 
 
 
 
 
 
 
 
 
 
11
  quantized_by: TheBloke
12
  tags:
13
  - facebook
@@ -38,34 +52,36 @@ tags:
38
  - Model creator: [Meta Llama 2](https://huggingface.co/meta-llama)
39
  - Original model: [Llama 2 13B Chat](https://huggingface.co/meta-llama/Llama-2-13b-chat-hf)
40
 
 
41
  ## Description
42
 
43
  This repo contains GGUF format model files for [Meta's Llama 2 13B-chat](https://huggingface.co/meta-llama/Llama-2-13b-chat-hf).
44
 
 
45
  <!-- README_GGUF.md-about-gguf start -->
46
  ### About GGUF
47
 
48
- GGUF is a new format introduced by the llama.cpp team on August 21st 2023. It is a replacement for GGML, which is no longer supported by llama.cpp.
49
 
50
- The key benefit of GGUF is that it is a extensible, future-proof format which stores more information about the model as metadata. It also includes significantly improved tokenization code, including for the first time full support for special tokens. This should improve performance, especially with models that use new special tokens and implement custom prompt templates.
51
 
52
- Here are a list of clients and libraries that are known to support GGUF:
53
- * [llama.cpp](https://github.com/ggerganov/llama.cpp).
54
- * [text-generation-webui](https://github.com/oobabooga/text-generation-webui), the most widely used web UI. Supports GGUF with GPU acceleration via the ctransformers backend - llama-cpp-python backend should work soon too.
55
- * [KoboldCpp](https://github.com/LostRuins/koboldcpp), now supports GGUF as of release 1.41! A powerful GGML web UI, with full GPU accel. Especially good for story telling.
56
- * [LM Studio](https://lmstudio.ai/), version 0.2.2 and later support GGUF. A fully featured local GUI with GPU acceleration on both Windows (NVidia and AMD), and macOS.
57
- * [LoLLMS Web UI](https://github.com/ParisNeo/lollms-webui), should now work, choose the `c_transformers` backend. A great web UI with many interesting features. Supports CUDA GPU acceleration.
58
- * [ctransformers](https://github.com/marella/ctransformers), now supports GGUF as of version 0.2.24! A Python library with GPU accel, LangChain support, and OpenAI-compatible AI server.
59
- * [llama-cpp-python](https://github.com/abetlen/llama-cpp-python), supports GGUF as of version 0.1.79. A Python library with GPU accel, LangChain support, and OpenAI-compatible API server.
60
- * [candle](https://github.com/huggingface/candle), added GGUF support on August 22nd. Candle is a Rust ML framework with a focus on performance, including GPU support, and ease of use.
61
 
62
  <!-- README_GGUF.md-about-gguf end -->
63
  <!-- repositories-available start -->
64
  ## Repositories available
65
 
 
66
  * [GPTQ models for GPU inference, with multiple quantisation parameter options.](https://huggingface.co/TheBloke/Llama-2-13B-chat-GPTQ)
67
  * [2, 3, 4, 5, 6 and 8-bit GGUF models for CPU+GPU inference](https://huggingface.co/TheBloke/Llama-2-13B-chat-GGUF)
68
- * [2, 3, 4, 5, 6 and 8-bit GGML models for CPU+GPU inference (deprecated)](https://huggingface.co/TheBloke/Llama-2-13B-chat-GGML)
69
  * [Meta Llama 2's original unquantised fp16 model in pytorch format, for GPU inference and for further conversions](https://huggingface.co/meta-llama/Llama-2-13B-chat-hf)
70
  <!-- repositories-available end -->
71
 
@@ -81,12 +97,14 @@ You are a helpful, respectful and honest assistant. Always answer as helpfully a
81
  ```
82
 
83
  <!-- prompt-template end -->
 
 
84
  <!-- compatibility_gguf start -->
85
  ## Compatibility
86
 
87
- These quantised GGUF files are compatible with llama.cpp from August 21st 2023 onwards, as of commit [6381d4e110bd0ec02843a60bbeb8b6fc37a9ace9](https://github.com/ggerganov/llama.cpp/commit/6381d4e110bd0ec02843a60bbeb8b6fc37a9ace9)
88
 
89
- They are now also compatible with many third party UIs and libraries - please see the list at the top of the README.
90
 
91
  ## Explanation of quantisation methods
92
  <details>
@@ -122,23 +140,80 @@ Refer to the Provided Files table below to see what files use which methods, and
122
  | [llama-2-13b-chat.Q8_0.gguf](https://huggingface.co/TheBloke/Llama-2-13B-chat-GGUF/blob/main/llama-2-13b-chat.Q8_0.gguf) | Q8_0 | 8 | 13.83 GB| 16.33 GB | very large, extremely low quality loss - not recommended |
123
 
124
  **Note**: the above RAM figures assume no GPU offloading. If layers are offloaded to the GPU, this will reduce RAM usage and use VRAM instead.
 
 
 
125
  <!-- README_GGUF.md-provided-files end -->
126
 
127
- <!-- README_GGUF.md-how-to-run start -->
128
- ## Example `llama.cpp` command
 
 
 
 
 
 
 
 
 
 
 
129
 
130
- Make sure you are using `llama.cpp` from commit [6381d4e110bd0ec02843a60bbeb8b6fc37a9ace9](https://github.com/ggerganov/llama.cpp/commit/6381d4e110bd0ec02843a60bbeb8b6fc37a9ace9) or later.
131
 
132
- For compatibility with older versions of llama.cpp, or for any third-party libraries or clients that haven't yet updated for GGUF, please use GGML files instead.
133
 
 
 
 
 
134
  ```
135
- ./main -t 10 -ngl 32 -m llama-2-13b-chat.q4_K_M.gguf --color -c 4096 --temp 0.7 --repeat_penalty 1.1 -n -1 -p "[INST] <<SYS>>\nYou are a helpful, respectful and honest assistant. Always answer as helpfully as possible, while being safe. Your answers should not include any harmful, unethical, racist, sexist, toxic, dangerous, or illegal content. Please ensure that your responses are socially unbiased and positive in nature. If a question does not make any sense, or is not factually coherent, explain why instead of answering something not correct. If you don't know the answer to a question, please don't share false information.\n<</SYS>>\nWrite a story about llamas[/INST]"
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
136
  ```
137
- Change `-t 10` to the number of physical CPU cores you have. For example if your system has 8 cores/16 threads, use `-t 8`. If offloading all layers to GPU, set `-t 1`.
138
 
139
  Change `-ngl 32` to the number of layers to offload to GPU. Remove it if you don't have GPU acceleration.
140
 
141
- Change `-c 4096` to the desired sequence length for this model. For extended sequence models - eg 8K, 16K, 32K - the necessary RoPE scaling parameters are read from the GGUF file and set by llama.cpp automatically.
142
 
143
  If you want to have a chat-style conversation, replace the `-p <PROMPT>` argument with `-i -ins`
144
 
@@ -173,7 +248,7 @@ CT_METAL=1 pip install ctransformers>=0.2.24 --no-binary ctransformers
173
  from ctransformers import AutoModelForCausalLM
174
 
175
  # Set gpu_layers to the number of layers to offload to GPU. Set to 0 if no GPU acceleration is available on your system.
176
- llm = AutoModelForCausalLM.from_pretrained("TheBloke/Llama-2-13B-chat-GGML", model_file="llama-2-13b-chat.q4_K_M.gguf", model_type="llama", gpu_layers=50)
177
 
178
  print(llm("AI is going to"))
179
  ```
@@ -195,10 +270,12 @@ For further support, and discussions on these models and AI in general, join us
195
 
196
  [TheBloke AI's Discord server](https://discord.gg/theblokeai)
197
 
198
- ## Thanks, and how to contribute.
199
 
200
  Thanks to the [chirper.ai](https://chirper.ai) team!
201
 
 
 
202
  I've had a lot of people ask if they can contribute. I enjoy providing models and helping people, and would love to be able to spend even more time doing it, as well as expanding into new projects like fine tuning/training.
203
 
204
  If you're able and willing to contribute it will be most gratefully received and will help me to keep providing more models, and to start work on new AI projects.
@@ -210,7 +287,7 @@ Donaters will get priority support on any and all AI/LLM/model questions and req
210
 
211
  **Special thanks to**: Aemon Algiz.
212
 
213
- **Patreon special mentions**: Russ Johnson, J, alfie_i, Alex, NimbleBox.ai, Chadd, Mandus, Nikolai Manek, Ken Nordquist, ya boyyy, Illia Dulskyi, Viktor Bowallius, vamX, Iucharbius, zynix, Magnesian, Clay Pascal, Pierre Kircher, Enrico Ros, Tony Hughes, Elle, Andrey, knownsqashed, Deep Realms, Jerry Meng, Lone Striker, Derek Yates, Pyrater, Mesiah Bishop, James Bentley, Femi Adebogun, Brandon Frisco, SuperWojo, Alps Aficionado, Michael Dempsey, Vitor Caleffi, Will Dee, Edmond Seymore, usrbinkat, LangChain4j, Kacper Wikieł, Luke Pendergrass, John Detwiler, theTransient, Nathan LeClaire, Tiffany J. Kim, biorpg, Eugene Pentland, Stanislav Ovsiannikov, Fred von Graf, terasurfer, Kalila, Dan Guido, Nitin Borwankar, 阿明, Ai Maven, John Villwock, Gabriel Puliatti, Stephen Murray, Asp the Wyvern, danny, Chris Smitley, ReadyPlayerEmma, S_X, Daniel P. Andersen, Olakabola, Jeffrey Morgan, Imad Khwaja, Caitlyn Gatomon, webtim, Alicia Loh, Trenton Dambrowitz, Swaroop Kallakuri, Erik Bjäreholt, Leonard Tan, Spiking Neurons AB, Luke @flexchar, Ajan Kanaga, Thomas Belote, Deo Leter, RoA, Willem Michiel, transmissions 11, subjectnull, Matthew Berman, Joseph William Delisle, David Ziegler, Michael Davis, Johann-Peter Hartmann, Talal Aujan, senxiiz, Artur Olbinski, Rainer Wilmers, Spencer Kim, Fen Risland, Cap'n Zoog, Rishabh Srivastava, Michael Levine, Geoffrey Montalvo, Sean Connelly, Alexandros Triantafyllidis, Pieter, Gabriel Tamborski, Sam, Subspace Studios, Junyu Yang, Pedro Madruga, Vadim, Cory Kujawski, K, Raven Klaugh, Randy H, Mano Prime, Sebastain Graf, Space Cruiser
214
 
215
 
216
  Thank you to all my generous patrons and donaters!
 
1
  ---
2
+ base_model: https://huggingface.co/meta-llama/Llama-2-13b-chat-hf
3
  inference: false
4
  language:
5
  - en
6
+ license: llama2
7
  model_creator: Meta Llama 2
 
8
  model_name: Llama 2 13B Chat
9
  model_type: llama
10
  pipeline_tag: text-generation
11
+ prompt_template: '[INST] <<SYS>>
12
+
13
+ You are a helpful, respectful and honest assistant. Always answer as helpfully as
14
+ possible, while being safe. Your answers should not include any harmful, unethical,
15
+ racist, sexist, toxic, dangerous, or illegal content. Please ensure that your responses
16
+ are socially unbiased and positive in nature. If a question does not make any sense,
17
+ or is not factually coherent, explain why instead of answering something not correct.
18
+ If you don''t know the answer to a question, please don''t share false information.
19
+
20
+ <</SYS>>
21
+
22
+ {prompt}[/INST]
23
+
24
+ '
25
  quantized_by: TheBloke
26
  tags:
27
  - facebook
 
52
  - Model creator: [Meta Llama 2](https://huggingface.co/meta-llama)
53
  - Original model: [Llama 2 13B Chat](https://huggingface.co/meta-llama/Llama-2-13b-chat-hf)
54
 
55
+ <!-- description start -->
56
  ## Description
57
 
58
  This repo contains GGUF format model files for [Meta's Llama 2 13B-chat](https://huggingface.co/meta-llama/Llama-2-13b-chat-hf).
59
 
60
+ <!-- description end -->
61
  <!-- README_GGUF.md-about-gguf start -->
62
  ### About GGUF
63
 
64
+ GGUF is a new format introduced by the llama.cpp team on August 21st 2023. It is a replacement for GGML, which is no longer supported by llama.cpp. GGUF offers numerous advantages over GGML, such as better tokenisation, and support for special tokens. It is also supports metadata, and is designed to be extensible.
65
 
66
+ Here is an incomplate list of clients and libraries that are known to support GGUF:
67
 
68
+ * [llama.cpp](https://github.com/ggerganov/llama.cpp). The source project for GGUF. Offers a CLI and a server option.
69
+ * [text-generation-webui](https://github.com/oobabooga/text-generation-webui), the most widely used web UI, with many features and powerful extensions. Supports GPU acceleration.
70
+ * [KoboldCpp](https://github.com/LostRuins/koboldcpp), a fully featured web UI, with GPU accel across all platforms and GPU architectures. Especially good for story telling.
71
+ * [LM Studio](https://lmstudio.ai/), an easy-to-use and powerful local GUI for Windows and macOS (Silicon), with GPU acceleration.
72
+ * [LoLLMS Web UI](https://github.com/ParisNeo/lollms-webui), a great web UI with many interesting and unique features, including a full model library for easy model selection.
73
+ * [Faraday.dev](https://faraday.dev/), an attractive and easy to use character-based chat GUI for Windows and macOS (both Silicon and Intel), with GPU acceleration.
74
+ * [ctransformers](https://github.com/marella/ctransformers), a Python library with GPU accel, LangChain support, and OpenAI-compatible AI server.
75
+ * [llama-cpp-python](https://github.com/abetlen/llama-cpp-python), a Python library with GPU accel, LangChain support, and OpenAI-compatible API server.
76
+ * [candle](https://github.com/huggingface/candle), a Rust ML framework with a focus on performance, including GPU support, and ease of use.
77
 
78
  <!-- README_GGUF.md-about-gguf end -->
79
  <!-- repositories-available start -->
80
  ## Repositories available
81
 
82
+ * [AWQ model(s) for GPU inference.](https://huggingface.co/TheBloke/Llama-2-13B-chat-AWQ)
83
  * [GPTQ models for GPU inference, with multiple quantisation parameter options.](https://huggingface.co/TheBloke/Llama-2-13B-chat-GPTQ)
84
  * [2, 3, 4, 5, 6 and 8-bit GGUF models for CPU+GPU inference](https://huggingface.co/TheBloke/Llama-2-13B-chat-GGUF)
 
85
  * [Meta Llama 2's original unquantised fp16 model in pytorch format, for GPU inference and for further conversions](https://huggingface.co/meta-llama/Llama-2-13B-chat-hf)
86
  <!-- repositories-available end -->
87
 
 
97
  ```
98
 
99
  <!-- prompt-template end -->
100
+
101
+
102
  <!-- compatibility_gguf start -->
103
  ## Compatibility
104
 
105
+ These quantised GGUFv2 files are compatible with llama.cpp from August 27th onwards, as of commit [d0cee0d36d5be95a0d9088b674dbb27354107221](https://github.com/ggerganov/llama.cpp/commit/d0cee0d36d5be95a0d9088b674dbb27354107221)
106
 
107
+ They are also compatible with many third party UIs and libraries - please see the list at the top of this README.
108
 
109
  ## Explanation of quantisation methods
110
  <details>
 
140
  | [llama-2-13b-chat.Q8_0.gguf](https://huggingface.co/TheBloke/Llama-2-13B-chat-GGUF/blob/main/llama-2-13b-chat.Q8_0.gguf) | Q8_0 | 8 | 13.83 GB| 16.33 GB | very large, extremely low quality loss - not recommended |
141
 
142
  **Note**: the above RAM figures assume no GPU offloading. If layers are offloaded to the GPU, this will reduce RAM usage and use VRAM instead.
143
+
144
+
145
+
146
  <!-- README_GGUF.md-provided-files end -->
147
 
148
+ <!-- README_GGUF.md-how-to-download start -->
149
+ ## How to download GGUF files
150
+
151
+ **Note for manual downloaders:** You almost never want to clone the entire repo! Multiple different quantisation formats are provided, and most users only want to pick and download a single file.
152
+
153
+ The following clients/libraries will automatically download models for you, providing a list of available models to choose from:
154
+ - LM Studio
155
+ - LoLLMS Web UI
156
+ - Faraday.dev
157
+
158
+ ### In `text-generation-webui`
159
+
160
+ Under Download Model, you can enter the model repo: TheBloke/Llama-2-13B-chat-GGUF and below it, a specific filename to download, such as: llama-2-13b-chat.q4_K_M.gguf.
161
 
162
+ Then click Download.
163
 
164
+ ### On the command line, including multiple files at once
165
 
166
+ I recommend using the `huggingface-hub` Python library:
167
+
168
+ ```shell
169
+ pip3 install huggingface-hub>=0.17.1
170
  ```
171
+
172
+ Then you can download any individual model file to the current directory, at high speed, with a command like this:
173
+
174
+ ```shell
175
+ huggingface-cli download TheBloke/Llama-2-13B-chat-GGUF llama-2-13b-chat.q4_K_M.gguf --local-dir . --local-dir-use-symlinks False
176
+ ```
177
+
178
+ <details>
179
+ <summary>More advanced huggingface-cli download usage</summary>
180
+
181
+ You can also download multiple files at once with a pattern:
182
+
183
+ ```shell
184
+ huggingface-cli download TheBloke/Llama-2-13B-chat-GGUF --local-dir . --local-dir-use-symlinks False --include='*Q4_K*gguf'
185
+ ```
186
+
187
+ For more documentation on downloading with `huggingface-cli`, please see: [HF -> Hub Python Library -> Download files -> Download from the CLI](https://huggingface.co/docs/huggingface_hub/guides/download#download-from-the-cli).
188
+
189
+ To accelerate downloads on fast connections (1Gbit/s or higher), install `hf_transfer`:
190
+
191
+ ```shell
192
+ pip3 install hf_transfer
193
+ ```
194
+
195
+ And set environment variable `HF_HUB_ENABLE_HF_TRANSFER` to `1`:
196
+
197
+ ```shell
198
+ HUGGINGFACE_HUB_ENABLE_HF_TRANSFER=1 huggingface-cli download TheBloke/Llama-2-13B-chat-GGUF llama-2-13b-chat.q4_K_M.gguf --local-dir . --local-dir-use-symlinks False
199
+ ```
200
+
201
+ Windows CLI users: Use `set HUGGINGFACE_HUB_ENABLE_HF_TRANSFER=1` before running the download command.
202
+ </details>
203
+ <!-- README_GGUF.md-how-to-download end -->
204
+
205
+ <!-- README_GGUF.md-how-to-run start -->
206
+ ## Example `llama.cpp` command
207
+
208
+ Make sure you are using `llama.cpp` from commit [d0cee0d36d5be95a0d9088b674dbb27354107221](https://github.com/ggerganov/llama.cpp/commit/d0cee0d36d5be95a0d9088b674dbb27354107221) or later.
209
+
210
+ ```shell
211
+ ./main -ngl 32 -m llama-2-13b-chat.q4_K_M.gguf --color -c 4096 --temp 0.7 --repeat_penalty 1.1 -n -1 -p "[INST] <<SYS>>\nYou are a helpful, respectful and honest assistant. Always answer as helpfully as possible, while being safe. Your answers should not include any harmful, unethical, racist, sexist, toxic, dangerous, or illegal content. Please ensure that your responses are socially unbiased and positive in nature. If a question does not make any sense, or is not factually coherent, explain why instead of answering something not correct. If you don't know the answer to a question, please don't share false information.\n<</SYS>>\n{prompt}[/INST]"
212
  ```
 
213
 
214
  Change `-ngl 32` to the number of layers to offload to GPU. Remove it if you don't have GPU acceleration.
215
 
216
+ Change `-c 4096` to the desired sequence length. For extended sequence models - eg 8K, 16K, 32K - the necessary RoPE scaling parameters are read from the GGUF file and set by llama.cpp automatically.
217
 
218
  If you want to have a chat-style conversation, replace the `-p <PROMPT>` argument with `-i -ins`
219
 
 
248
  from ctransformers import AutoModelForCausalLM
249
 
250
  # Set gpu_layers to the number of layers to offload to GPU. Set to 0 if no GPU acceleration is available on your system.
251
+ llm = AutoModelForCausalLM.from_pretrained("TheBloke/Llama-2-13B-chat-GGUF", model_file="llama-2-13b-chat.q4_K_M.gguf", model_type="llama", gpu_layers=50)
252
 
253
  print(llm("AI is going to"))
254
  ```
 
270
 
271
  [TheBloke AI's Discord server](https://discord.gg/theblokeai)
272
 
273
+ ## Thanks, and how to contribute
274
 
275
  Thanks to the [chirper.ai](https://chirper.ai) team!
276
 
277
+ Thanks to Clay from [gpus.llm-utils.org](llm-utils)!
278
+
279
  I've had a lot of people ask if they can contribute. I enjoy providing models and helping people, and would love to be able to spend even more time doing it, as well as expanding into new projects like fine tuning/training.
280
 
281
  If you're able and willing to contribute it will be most gratefully received and will help me to keep providing more models, and to start work on new AI projects.
 
287
 
288
  **Special thanks to**: Aemon Algiz.
289
 
290
+ **Patreon special mentions**: Alicia Loh, Stephen Murray, K, Ajan Kanaga, RoA, Magnesian, Deo Leter, Olakabola, Eugene Pentland, zynix, Deep Realms, Raymond Fosdick, Elijah Stavena, Iucharbius, Erik Bjäreholt, Luis Javier Navarrete Lozano, Nicholas, theTransient, John Detwiler, alfie_i, knownsqashed, Mano Prime, Willem Michiel, Enrico Ros, LangChain4j, OG, Michael Dempsey, Pierre Kircher, Pedro Madruga, James Bentley, Thomas Belote, Luke @flexchar, Leonard Tan, Johann-Peter Hartmann, Illia Dulskyi, Fen Risland, Chadd, S_X, Jeff Scroggin, Ken Nordquist, Sean Connelly, Artur Olbinski, Swaroop Kallakuri, Jack West, Ai Maven, David Ziegler, Russ Johnson, transmissions 11, John Villwock, Alps Aficionado, Clay Pascal, Viktor Bowallius, Subspace Studios, Rainer Wilmers, Trenton Dambrowitz, vamX, Michael Levine, 준교 김, Brandon Frisco, Kalila, Trailburnt, Randy H, Talal Aujan, Nathan Dryer, Vadim, 阿明, ReadyPlayerEmma, Tiffany J. Kim, George Stoitzev, Spencer Kim, Jerry Meng, Gabriel Tamborski, Cory Kujawski, Jeffrey Morgan, Spiking Neurons AB, Edmond Seymore, Alexandros Triantafyllidis, Lone Striker, Cap'n Zoog, Nikolai Manek, danny, ya boyyy, Derek Yates, usrbinkat, Mandus, TL, Nathan LeClaire, subjectnull, Imad Khwaja, webtim, Raven Klaugh, Asp the Wyvern, Gabriel Puliatti, Caitlyn Gatomon, Joseph William Delisle, Jonathan Leane, Luke Pendergrass, SuperWojo, Sebastain Graf, Will Dee, Fred von Graf, Andrey, Dan Guido, Daniel P. Andersen, Nitin Borwankar, Elle, Vitor Caleffi, biorpg, jjj, NimbleBox.ai, Pieter, Matthew Berman, terasurfer, Michael Davis, Alex, Stanislav Ovsiannikov
291
 
292
 
293
  Thank you to all my generous patrons and donaters!