TheBloke commited on
Commit
0e624ca
1 Parent(s): 49f6f5d

Upload README.md

Browse files
Files changed (1) hide show
  1. README.md +97 -24
README.md CHANGED
@@ -1,4 +1,5 @@
1
  ---
 
2
  inference: false
3
  library_name: transformers
4
  license: llama2
@@ -18,9 +19,20 @@ model-index:
18
  task:
19
  type: text-generation
20
  model_creator: WizardLM
21
- model_link: https://huggingface.co/WizardLM/WizardCoder-Python-13B-V1.0
22
  model_name: WizardCoder Python 13B V1.0
23
  model_type: llama
 
 
 
 
 
 
 
 
 
 
 
 
24
  quantized_by: TheBloke
25
  tags:
26
  - code
@@ -47,23 +59,25 @@ tags:
47
  - Model creator: [WizardLM](https://huggingface.co/WizardLM)
48
  - Original model: [WizardCoder Python 13B V1.0](https://huggingface.co/WizardLM/WizardCoder-Python-13B-V1.0)
49
 
 
50
  ## Description
51
 
52
  This repo contains GGUF format model files for [WizardLM's WizardCoder Python 13B V1.0](https://huggingface.co/WizardLM/WizardCoder-Python-13B-V1.0).
53
 
 
54
  <!-- README_GGUF.md-about-gguf start -->
55
  ### About GGUF
56
 
57
- 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.
58
 
59
- 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.
60
 
61
- Here are a list of clients and libraries that are known to support GGUF:
62
- * [llama.cpp](https://github.com/ggerganov/llama.cpp).
63
- * [text-generation-webui](https://github.com/oobabooga/text-generation-webui), the most widely used web UI, with many features and powerful extensions.
64
- * [KoboldCpp](https://github.com/LostRuins/koboldcpp), a fully featured web UI, with full GPU accel across multiple platforms and GPU architectures. Especially good for story telling.
65
- * [LM Studio](https://lmstudio.ai/), an easy-to-use and powerful local GUI with GPU acceleration on both Windows (NVidia and AMD), and macOS.
66
  * [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.
 
67
  * [ctransformers](https://github.com/marella/ctransformers), a Python library with GPU accel, LangChain support, and OpenAI-compatible AI server.
68
  * [llama-cpp-python](https://github.com/abetlen/llama-cpp-python), a Python library with GPU accel, LangChain support, and OpenAI-compatible API server.
69
  * [candle](https://github.com/huggingface/candle), a Rust ML framework with a focus on performance, including GPU support, and ease of use.
@@ -72,9 +86,9 @@ Here are a list of clients and libraries that are known to support GGUF:
72
  <!-- repositories-available start -->
73
  ## Repositories available
74
 
 
75
  * [GPTQ models for GPU inference, with multiple quantisation parameter options.](https://huggingface.co/TheBloke/WizardCoder-Python-13B-V1.0-GPTQ)
76
  * [2, 3, 4, 5, 6 and 8-bit GGUF models for CPU+GPU inference](https://huggingface.co/TheBloke/WizardCoder-Python-13B-V1.0-GGUF)
77
- * [2, 3, 4, 5, 6 and 8-bit GGML models for CPU+GPU inference (deprecated)](https://huggingface.co/TheBloke/WizardCoder-Python-13B-V1.0-GGML)
78
  * [WizardLM's original unquantised fp16 model in pytorch format, for GPU inference and for further conversions](https://huggingface.co/WizardLM/WizardCoder-Python-13B-V1.0)
79
  <!-- repositories-available end -->
80
 
@@ -92,12 +106,14 @@ Below is an instruction that describes a task. Write a response that appropriate
92
  ```
93
 
94
  <!-- prompt-template end -->
 
 
95
  <!-- compatibility_gguf start -->
96
  ## Compatibility
97
 
98
- 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)
99
 
100
- They are now also compatible with many third party UIs and libraries - please see the list at the top of the README.
101
 
102
  ## Explanation of quantisation methods
103
  <details>
@@ -138,21 +154,75 @@ Refer to the Provided Files table below to see what files use which methods, and
138
 
139
  <!-- README_GGUF.md-provided-files end -->
140
 
141
- <!-- README_GGUF.md-how-to-run start -->
142
- ## Example `llama.cpp` command
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
143
 
144
- Make sure you are using `llama.cpp` from commit [6381d4e110bd0ec02843a60bbeb8b6fc37a9ace9](https://github.com/ggerganov/llama.cpp/commit/6381d4e110bd0ec02843a60bbeb8b6fc37a9ace9) or later.
145
 
146
- 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.
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
147
 
 
 
148
  ```
149
- ./main -t 10 -ngl 32 -m wizardcoder-python-13b-v1.0.q4_K_M.gguf --color -c 4096 --temp 0.7 --repeat_penalty 1.1 -n -1 -p "Below is an instruction that describes a task. Write a response that appropriately completes the request.\n\n### Instruction:\n{prompt}\n\n### Response:"
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
150
  ```
151
- 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`.
152
 
153
  Change `-ngl 32` to the number of layers to offload to GPU. Remove it if you don't have GPU acceleration.
154
 
155
- 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.
156
 
157
  If you want to have a chat-style conversation, replace the `-p <PROMPT>` argument with `-i -ins`
158
 
@@ -209,10 +279,12 @@ For further support, and discussions on these models and AI in general, join us
209
 
210
  [TheBloke AI's Discord server](https://discord.gg/theblokeai)
211
 
212
- ## Thanks, and how to contribute.
213
 
214
  Thanks to the [chirper.ai](https://chirper.ai) team!
215
 
 
 
216
  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.
217
 
218
  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.
@@ -224,7 +296,7 @@ Donaters will get priority support on any and all AI/LLM/model questions and req
224
 
225
  **Special thanks to**: Aemon Algiz.
226
 
227
- **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
228
 
229
 
230
  Thank you to all my generous patrons and donaters!
@@ -314,10 +386,11 @@ Note: This script supports `WizardLM/WizardCoder-Python-34B/13B/7B-V1.0`. If you
314
  Please cite the repo if you use the data, method or code in this repo.
315
 
316
  ```
317
- @misc{luo2023wizardcoder,
318
- title={WizardCoder: Empowering Code Large Language Models with Evol-Instruct},
319
- author={Ziyang Luo and Can Xu and Pu Zhao and Qingfeng Sun and Xiubo Geng and Wenxiang Hu and Chongyang Tao and Jing Ma and Qingwei Lin and Daxin Jiang},
320
- year={2023},
 
321
  }
322
  ```
323
 
 
1
  ---
2
+ base_model: https://huggingface.co/WizardLM/WizardCoder-Python-13B-V1.0
3
  inference: false
4
  library_name: transformers
5
  license: llama2
 
19
  task:
20
  type: text-generation
21
  model_creator: WizardLM
 
22
  model_name: WizardCoder Python 13B V1.0
23
  model_type: llama
24
+ prompt_template: 'Below is an instruction that describes a task. Write a response
25
+ that appropriately completes the request.
26
+
27
+
28
+ ### Instruction:
29
+
30
+ {prompt}
31
+
32
+
33
+ ### Response:
34
+
35
+ '
36
  quantized_by: TheBloke
37
  tags:
38
  - code
 
59
  - Model creator: [WizardLM](https://huggingface.co/WizardLM)
60
  - Original model: [WizardCoder Python 13B V1.0](https://huggingface.co/WizardLM/WizardCoder-Python-13B-V1.0)
61
 
62
+ <!-- description start -->
63
  ## Description
64
 
65
  This repo contains GGUF format model files for [WizardLM's WizardCoder Python 13B V1.0](https://huggingface.co/WizardLM/WizardCoder-Python-13B-V1.0).
66
 
67
+ <!-- description end -->
68
  <!-- README_GGUF.md-about-gguf start -->
69
  ### About GGUF
70
 
71
+ 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.
72
 
73
+ Here is an incomplate list of clients and libraries that are known to support GGUF:
74
 
75
+ * [llama.cpp](https://github.com/ggerganov/llama.cpp). The source project for GGUF. Offers a CLI and a server option.
76
+ * [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.
77
+ * [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.
78
+ * [LM Studio](https://lmstudio.ai/), an easy-to-use and powerful local GUI for Windows and macOS (Silicon), with GPU acceleration.
 
79
  * [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.
80
+ * [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.
81
  * [ctransformers](https://github.com/marella/ctransformers), a Python library with GPU accel, LangChain support, and OpenAI-compatible AI server.
82
  * [llama-cpp-python](https://github.com/abetlen/llama-cpp-python), a Python library with GPU accel, LangChain support, and OpenAI-compatible API server.
83
  * [candle](https://github.com/huggingface/candle), a Rust ML framework with a focus on performance, including GPU support, and ease of use.
 
86
  <!-- repositories-available start -->
87
  ## Repositories available
88
 
89
+ * [AWQ model(s) for GPU inference.](https://huggingface.co/TheBloke/WizardCoder-Python-13B-V1.0-AWQ)
90
  * [GPTQ models for GPU inference, with multiple quantisation parameter options.](https://huggingface.co/TheBloke/WizardCoder-Python-13B-V1.0-GPTQ)
91
  * [2, 3, 4, 5, 6 and 8-bit GGUF models for CPU+GPU inference](https://huggingface.co/TheBloke/WizardCoder-Python-13B-V1.0-GGUF)
 
92
  * [WizardLM's original unquantised fp16 model in pytorch format, for GPU inference and for further conversions](https://huggingface.co/WizardLM/WizardCoder-Python-13B-V1.0)
93
  <!-- repositories-available end -->
94
 
 
106
  ```
107
 
108
  <!-- prompt-template end -->
109
+
110
+
111
  <!-- compatibility_gguf start -->
112
  ## Compatibility
113
 
114
+ 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)
115
 
116
+ They are also compatible with many third party UIs and libraries - please see the list at the top of this README.
117
 
118
  ## Explanation of quantisation methods
119
  <details>
 
154
 
155
  <!-- README_GGUF.md-provided-files end -->
156
 
157
+ <!-- README_GGUF.md-how-to-download start -->
158
+ ## How to download GGUF files
159
+
160
+ **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.
161
+
162
+ The following clients/libraries will automatically download models for you, providing a list of available models to choose from:
163
+ - LM Studio
164
+ - LoLLMS Web UI
165
+ - Faraday.dev
166
+
167
+ ### In `text-generation-webui`
168
+
169
+ Under Download Model, you can enter the model repo: TheBloke/WizardCoder-Python-13B-V1.0-GGUF and below it, a specific filename to download, such as: wizardcoder-python-13b-v1.0.q4_K_M.gguf.
170
+
171
+ Then click Download.
172
+
173
+ ### On the command line, including multiple files at once
174
+
175
+ I recommend using the `huggingface-hub` Python library:
176
+
177
+ ```shell
178
+ pip3 install huggingface-hub>=0.17.1
179
+ ```
180
 
181
+ Then you can download any individual model file to the current directory, at high speed, with a command like this:
182
 
183
+ ```shell
184
+ huggingface-cli download TheBloke/WizardCoder-Python-13B-V1.0-GGUF wizardcoder-python-13b-v1.0.q4_K_M.gguf --local-dir . --local-dir-use-symlinks False
185
+ ```
186
+
187
+ <details>
188
+ <summary>More advanced huggingface-cli download usage</summary>
189
+
190
+ You can also download multiple files at once with a pattern:
191
+
192
+ ```shell
193
+ huggingface-cli download TheBloke/WizardCoder-Python-13B-V1.0-GGUF --local-dir . --local-dir-use-symlinks False --include='*Q4_K*gguf'
194
+ ```
195
+
196
+ 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).
197
+
198
+ To accelerate downloads on fast connections (1Gbit/s or higher), install `hf_transfer`:
199
 
200
+ ```shell
201
+ pip3 install hf_transfer
202
  ```
203
+
204
+ And set environment variable `HF_HUB_ENABLE_HF_TRANSFER` to `1`:
205
+
206
+ ```shell
207
+ HUGGINGFACE_HUB_ENABLE_HF_TRANSFER=1 huggingface-cli download TheBloke/WizardCoder-Python-13B-V1.0-GGUF wizardcoder-python-13b-v1.0.q4_K_M.gguf --local-dir . --local-dir-use-symlinks False
208
+ ```
209
+
210
+ Windows CLI users: Use `set HUGGINGFACE_HUB_ENABLE_HF_TRANSFER=1` before running the download command.
211
+ </details>
212
+ <!-- README_GGUF.md-how-to-download end -->
213
+
214
+ <!-- README_GGUF.md-how-to-run start -->
215
+ ## Example `llama.cpp` command
216
+
217
+ Make sure you are using `llama.cpp` from commit [d0cee0d36d5be95a0d9088b674dbb27354107221](https://github.com/ggerganov/llama.cpp/commit/d0cee0d36d5be95a0d9088b674dbb27354107221) or later.
218
+
219
+ ```shell
220
+ ./main -ngl 32 -m wizardcoder-python-13b-v1.0.q4_K_M.gguf --color -c 4096 --temp 0.7 --repeat_penalty 1.1 -n -1 -p "Below is an instruction that describes a task. Write a response that appropriately completes the request.\n\n### Instruction:\n{prompt}\n\n### Response:"
221
  ```
 
222
 
223
  Change `-ngl 32` to the number of layers to offload to GPU. Remove it if you don't have GPU acceleration.
224
 
225
+ 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.
226
 
227
  If you want to have a chat-style conversation, replace the `-p <PROMPT>` argument with `-i -ins`
228
 
 
279
 
280
  [TheBloke AI's Discord server](https://discord.gg/theblokeai)
281
 
282
+ ## Thanks, and how to contribute
283
 
284
  Thanks to the [chirper.ai](https://chirper.ai) team!
285
 
286
+ Thanks to Clay from [gpus.llm-utils.org](llm-utils)!
287
+
288
  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.
289
 
290
  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.
 
296
 
297
  **Special thanks to**: Aemon Algiz.
298
 
299
+ **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
300
 
301
 
302
  Thank you to all my generous patrons and donaters!
 
386
  Please cite the repo if you use the data, method or code in this repo.
387
 
388
  ```
389
+ @article{luo2023wizardcoder,
390
+ title={WizardCoder: Empowering Code Large Language Models with Evol-Instruct},
391
+ author={Luo, Ziyang and Xu, Can and Zhao, Pu and Sun, Qingfeng and Geng, Xiubo and Hu, Wenxiang and Tao, Chongyang and Ma, Jing and Lin, Qingwei and Jiang, Daxin},
392
+ journal={arXiv preprint arXiv:2306.08568},
393
+ year={2023}
394
  }
395
  ```
396