TheBloke commited on
Commit
047d001
1 Parent(s): 7a16d98

Upload README.md

Browse files
Files changed (1) hide show
  1. README.md +468 -0
README.md ADDED
@@ -0,0 +1,468 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model: TigerResearch/tigerbot-70b-chat-v2
3
+ inference: false
4
+ language:
5
+ - zh
6
+ - en
7
+ license: apache-2.0
8
+ model_creator: Tiger Research
9
+ model_name: Tigerbot 70B Chat V2
10
+ model_type: llama
11
+ prompt_template: 'Below is an instruction that describes a task. Write a response
12
+ that appropriately completes the request.
13
+
14
+
15
+ ### Instruction:
16
+
17
+ {prompt}
18
+
19
+
20
+ ### Response:
21
+
22
+ '
23
+ quantized_by: TheBloke
24
+ ---
25
+ <!-- markdownlint-disable MD041 -->
26
+
27
+ <!-- header start -->
28
+ <!-- 200823 -->
29
+ <div style="width: auto; margin-left: auto; margin-right: auto">
30
+ <img src="https://i.imgur.com/EBdldam.jpg" alt="TheBlokeAI" style="width: 100%; min-width: 400px; display: block; margin: auto;">
31
+ </div>
32
+ <div style="display: flex; justify-content: space-between; width: 100%;">
33
+ <div style="display: flex; flex-direction: column; align-items: flex-start;">
34
+ <p style="margin-top: 0.5em; margin-bottom: 0em;"><a href="https://discord.gg/theblokeai">Chat & support: TheBloke's Discord server</a></p>
35
+ </div>
36
+ <div style="display: flex; flex-direction: column; align-items: flex-end;">
37
+ <p style="margin-top: 0.5em; margin-bottom: 0em;"><a href="https://www.patreon.com/TheBlokeAI">Want to contribute? TheBloke's Patreon page</a></p>
38
+ </div>
39
+ </div>
40
+ <div style="text-align:center; margin-top: 0em; margin-bottom: 0em"><p style="margin-top: 0.25em; margin-bottom: 0em;">TheBloke's LLM work is generously supported by a grant from <a href="https://a16z.com">andreessen horowitz (a16z)</a></p></div>
41
+ <hr style="margin-top: 1.0em; margin-bottom: 1.0em;">
42
+ <!-- header end -->
43
+
44
+ # Tigerbot 70B Chat V2 - GGUF
45
+ - Model creator: [Tiger Research](https://huggingface.co/TigerResearch)
46
+ - Original model: [Tigerbot 70B Chat V2](https://huggingface.co/TigerResearch/tigerbot-70b-chat-v2)
47
+
48
+ <!-- description start -->
49
+ ## Description
50
+
51
+ This repo contains GGUF format model files for [Tiger Research's Tigerbot 70B Chat V2](https://huggingface.co/TigerResearch/tigerbot-70b-chat-v2).
52
+
53
+ These files were quantised using hardware kindly provided by [Massed Compute](https://massedcompute.com/).
54
+
55
+ <!-- description end -->
56
+ <!-- README_GGUF.md-about-gguf start -->
57
+ ### About GGUF
58
+
59
+ 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.
60
+
61
+ Here is an incomplete list of clients and libraries that are known to support GGUF:
62
+
63
+ * [llama.cpp](https://github.com/ggerganov/llama.cpp). The source project for GGUF. Offers a CLI and a server option.
64
+ * [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.
65
+ * [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.
66
+ * [GPT4All](https://gpt4all.io/index.html), a free and open source local running GUI, supporting Windows, Linux and macOS with full GPU accel.
67
+ * [LM Studio](https://lmstudio.ai/), an easy-to-use and powerful local GUI for Windows and macOS (Silicon), with GPU acceleration. Linux available, in beta as of 27/11/2023.
68
+ * [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.
69
+ * [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.
70
+ * [llama-cpp-python](https://github.com/abetlen/llama-cpp-python), a Python library with GPU accel, LangChain support, and OpenAI-compatible API server.
71
+ * [candle](https://github.com/huggingface/candle), a Rust ML framework with a focus on performance, including GPU support, and ease of use.
72
+ * [ctransformers](https://github.com/marella/ctransformers), a Python library with GPU accel, LangChain support, and OpenAI-compatible AI server. Note, as of time of writing (November 27th 2023), ctransformers has not been updated in a long time and does not support many recent models.
73
+
74
+ <!-- README_GGUF.md-about-gguf end -->
75
+ <!-- repositories-available start -->
76
+ ## Repositories available
77
+
78
+ * [AWQ model(s) for GPU inference.](https://huggingface.co/TheBloke/tigerbot-70B-chat-v2-AWQ)
79
+ * [GPTQ models for GPU inference, with multiple quantisation parameter options.](https://huggingface.co/TheBloke/tigerbot-70B-chat-v2-GPTQ)
80
+ * [2, 3, 4, 5, 6 and 8-bit GGUF models for CPU+GPU inference](https://huggingface.co/TheBloke/tigerbot-70B-chat-v2-GGUF)
81
+ * [Tiger Research's original unquantised fp16 model in pytorch format, for GPU inference and for further conversions](https://huggingface.co/TigerResearch/tigerbot-70b-chat-v2)
82
+ <!-- repositories-available end -->
83
+
84
+ <!-- prompt-template start -->
85
+ ## Prompt template: Alpaca
86
+
87
+ ```
88
+ Below is an instruction that describes a task. Write a response that appropriately completes the request.
89
+
90
+ ### Instruction:
91
+ {prompt}
92
+
93
+ ### Response:
94
+
95
+ ```
96
+
97
+ <!-- prompt-template end -->
98
+ <!-- licensing start -->
99
+ ## Licensing
100
+
101
+ The creator of the source model has listed its license as `apache-2.0`, and this quantization has therefore used that same license.
102
+
103
+ As this model is based on Llama 2, it is also subject to the Meta Llama 2 license terms, and the license files for that are additionally included. It should therefore be considered as being claimed to be licensed under both licenses. I contacted Hugging Face for clarification on dual licensing but they do not yet have an official position. Should this change, or should Meta provide any feedback on this situation, I will update this section accordingly.
104
+
105
+ In the meantime, any questions regarding licensing, and in particular how these two licenses might interact, should be directed to the original model repository: [Tiger Research's Tigerbot 70B Chat V2](https://huggingface.co/TigerResearch/tigerbot-70b-chat-v2).
106
+ <!-- licensing end -->
107
+ <!-- compatibility_gguf start -->
108
+ ## Compatibility
109
+
110
+ These quantised GGUFv2 files are compatible with llama.cpp from August 27th onwards, as of commit [d0cee0d](https://github.com/ggerganov/llama.cpp/commit/d0cee0d36d5be95a0d9088b674dbb27354107221)
111
+
112
+ They are also compatible with many third party UIs and libraries - please see the list at the top of this README.
113
+
114
+ ## Explanation of quantisation methods
115
+
116
+ <details>
117
+ <summary>Click to see details</summary>
118
+
119
+ The new methods available are:
120
+
121
+ * GGML_TYPE_Q2_K - "type-1" 2-bit quantization in super-blocks containing 16 blocks, each block having 16 weight. Block scales and mins are quantized with 4 bits. This ends up effectively using 2.5625 bits per weight (bpw)
122
+ * GGML_TYPE_Q3_K - "type-0" 3-bit quantization in super-blocks containing 16 blocks, each block having 16 weights. Scales are quantized with 6 bits. This end up using 3.4375 bpw.
123
+ * GGML_TYPE_Q4_K - "type-1" 4-bit quantization in super-blocks containing 8 blocks, each block having 32 weights. Scales and mins are quantized with 6 bits. This ends up using 4.5 bpw.
124
+ * GGML_TYPE_Q5_K - "type-1" 5-bit quantization. Same super-block structure as GGML_TYPE_Q4_K resulting in 5.5 bpw
125
+ * GGML_TYPE_Q6_K - "type-0" 6-bit quantization. Super-blocks with 16 blocks, each block having 16 weights. Scales are quantized with 8 bits. This ends up using 6.5625 bpw
126
+
127
+ Refer to the Provided Files table below to see what files use which methods, and how.
128
+ </details>
129
+ <!-- compatibility_gguf end -->
130
+
131
+ <!-- README_GGUF.md-provided-files start -->
132
+ ## Provided files
133
+
134
+ | Name | Quant method | Bits | Size | Max RAM required | Use case |
135
+ | ---- | ---- | ---- | ---- | ---- | ----- |
136
+ | [tigerbot-70b-chat-v2.Q2_K.gguf](https://huggingface.co/TheBloke/tigerbot-70B-chat-v2-GGUF/blob/main/tigerbot-70b-chat-v2.Q2_K.gguf) | Q2_K | 2 | 29.55 GB| 32.05 GB | smallest, significant quality loss - not recommended for most purposes |
137
+ | [tigerbot-70b-chat-v2.Q3_K_S.gguf](https://huggingface.co/TheBloke/tigerbot-70B-chat-v2-GGUF/blob/main/tigerbot-70b-chat-v2.Q3_K_S.gguf) | Q3_K_S | 3 | 30.21 GB| 32.71 GB | very small, high quality loss |
138
+ | [tigerbot-70b-chat-v2.Q3_K_M.gguf](https://huggingface.co/TheBloke/tigerbot-70B-chat-v2-GGUF/blob/main/tigerbot-70b-chat-v2.Q3_K_M.gguf) | Q3_K_M | 3 | 33.48 GB| 35.98 GB | very small, high quality loss |
139
+ | [tigerbot-70b-chat-v2.Q3_K_L.gguf](https://huggingface.co/TheBloke/tigerbot-70B-chat-v2-GGUF/blob/main/tigerbot-70b-chat-v2.Q3_K_L.gguf) | Q3_K_L | 3 | 36.44 GB| 38.94 GB | small, substantial quality loss |
140
+ | [tigerbot-70b-chat-v2.Q4_0.gguf](https://huggingface.co/TheBloke/tigerbot-70B-chat-v2-GGUF/blob/main/tigerbot-70b-chat-v2.Q4_0.gguf) | Q4_0 | 4 | 39.20 GB| 41.70 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
141
+ | [tigerbot-70b-chat-v2.Q4_K_S.gguf](https://huggingface.co/TheBloke/tigerbot-70B-chat-v2-GGUF/blob/main/tigerbot-70b-chat-v2.Q4_K_S.gguf) | Q4_K_S | 4 | 39.40 GB| 41.90 GB | small, greater quality loss |
142
+ | [tigerbot-70b-chat-v2.Q4_K_M.gguf](https://huggingface.co/TheBloke/tigerbot-70B-chat-v2-GGUF/blob/main/tigerbot-70b-chat-v2.Q4_K_M.gguf) | Q4_K_M | 4 | 41.75 GB| 44.25 GB | medium, balanced quality - recommended |
143
+ | [tigerbot-70b-chat-v2.Q5_0.gguf](https://huggingface.co/TheBloke/tigerbot-70B-chat-v2-GGUF/blob/main/tigerbot-70b-chat-v2.Q5_0.gguf) | Q5_0 | 5 | 47.81 GB| 50.31 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
144
+ | [tigerbot-70b-chat-v2.Q5_K_S.gguf](https://huggingface.co/TheBloke/tigerbot-70B-chat-v2-GGUF/blob/main/tigerbot-70b-chat-v2.Q5_K_S.gguf) | Q5_K_S | 5 | 47.81 GB| 50.31 GB | large, low quality loss - recommended |
145
+ | [tigerbot-70b-chat-v2.Q5_K_M.gguf](https://huggingface.co/TheBloke/tigerbot-70B-chat-v2-GGUF/blob/main/tigerbot-70b-chat-v2.Q5_K_M.gguf) | Q5_K_M | 5 | 49.11 GB| 51.61 GB | large, very low quality loss - recommended |
146
+ | tigerbot-70b-chat-v2.Q6_K.gguf | Q6_K | 6 | 56.97 GB| 59.47 GB | very large, extremely low quality loss |
147
+ | tigerbot-70b-chat-v2.Q8_0.gguf | Q8_0 | 8 | 73.79 GB| 76.29 GB | very large, extremely low quality loss - not recommended |
148
+
149
+ **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.
150
+
151
+ ### Q6_K and Q8_0 files are split and require joining
152
+
153
+ **Note:** HF does not support uploading files larger than 50GB. Therefore I have uploaded the Q6_K and Q8_0 files as split files.
154
+
155
+ <details>
156
+ <summary>Click for instructions regarding Q6_K and Q8_0 files</summary>
157
+
158
+ ### q6_K
159
+ Please download:
160
+ * `tigerbot-70b-chat-v2.Q6_K.gguf-split-a`
161
+ * `tigerbot-70b-chat-v2.Q6_K.gguf-split-b`
162
+
163
+ ### q8_0
164
+ Please download:
165
+ * `tigerbot-70b-chat-v2.Q8_0.gguf-split-a`
166
+ * `tigerbot-70b-chat-v2.Q8_0.gguf-split-b`
167
+
168
+ To join the files, do the following:
169
+
170
+ Linux and macOS:
171
+ ```
172
+ cat tigerbot-70b-chat-v2.Q6_K.gguf-split-* > tigerbot-70b-chat-v2.Q6_K.gguf && rm tigerbot-70b-chat-v2.Q6_K.gguf-split-*
173
+ cat tigerbot-70b-chat-v2.Q8_0.gguf-split-* > tigerbot-70b-chat-v2.Q8_0.gguf && rm tigerbot-70b-chat-v2.Q8_0.gguf-split-*
174
+ ```
175
+ Windows command line:
176
+ ```
177
+ COPY /B tigerbot-70b-chat-v2.Q6_K.gguf-split-a + tigerbot-70b-chat-v2.Q6_K.gguf-split-b tigerbot-70b-chat-v2.Q6_K.gguf
178
+ del tigerbot-70b-chat-v2.Q6_K.gguf-split-a tigerbot-70b-chat-v2.Q6_K.gguf-split-b
179
+
180
+ COPY /B tigerbot-70b-chat-v2.Q8_0.gguf-split-a + tigerbot-70b-chat-v2.Q8_0.gguf-split-b tigerbot-70b-chat-v2.Q8_0.gguf
181
+ del tigerbot-70b-chat-v2.Q8_0.gguf-split-a tigerbot-70b-chat-v2.Q8_0.gguf-split-b
182
+ ```
183
+
184
+ </details>
185
+ <!-- README_GGUF.md-provided-files end -->
186
+
187
+ <!-- README_GGUF.md-how-to-download start -->
188
+ ## How to download GGUF files
189
+
190
+ **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.
191
+
192
+ The following clients/libraries will automatically download models for you, providing a list of available models to choose from:
193
+
194
+ * LM Studio
195
+ * LoLLMS Web UI
196
+ * Faraday.dev
197
+
198
+ ### In `text-generation-webui`
199
+
200
+ Under Download Model, you can enter the model repo: TheBloke/tigerbot-70B-chat-v2-GGUF and below it, a specific filename to download, such as: tigerbot-70b-chat-v2.Q4_K_M.gguf.
201
+
202
+ Then click Download.
203
+
204
+ ### On the command line, including multiple files at once
205
+
206
+ I recommend using the `huggingface-hub` Python library:
207
+
208
+ ```shell
209
+ pip3 install huggingface-hub
210
+ ```
211
+
212
+ Then you can download any individual model file to the current directory, at high speed, with a command like this:
213
+
214
+ ```shell
215
+ huggingface-cli download TheBloke/tigerbot-70B-chat-v2-GGUF tigerbot-70b-chat-v2.Q4_K_M.gguf --local-dir . --local-dir-use-symlinks False
216
+ ```
217
+
218
+ <details>
219
+ <summary>More advanced huggingface-cli download usage (click to read)</summary>
220
+
221
+ You can also download multiple files at once with a pattern:
222
+
223
+ ```shell
224
+ huggingface-cli download TheBloke/tigerbot-70B-chat-v2-GGUF --local-dir . --local-dir-use-symlinks False --include='*Q4_K*gguf'
225
+ ```
226
+
227
+ 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).
228
+
229
+ To accelerate downloads on fast connections (1Gbit/s or higher), install `hf_transfer`:
230
+
231
+ ```shell
232
+ pip3 install hf_transfer
233
+ ```
234
+
235
+ And set environment variable `HF_HUB_ENABLE_HF_TRANSFER` to `1`:
236
+
237
+ ```shell
238
+ HF_HUB_ENABLE_HF_TRANSFER=1 huggingface-cli download TheBloke/tigerbot-70B-chat-v2-GGUF tigerbot-70b-chat-v2.Q4_K_M.gguf --local-dir . --local-dir-use-symlinks False
239
+ ```
240
+
241
+ Windows Command Line users: You can set the environment variable by running `set HF_HUB_ENABLE_HF_TRANSFER=1` before the download command.
242
+ </details>
243
+ <!-- README_GGUF.md-how-to-download end -->
244
+
245
+ <!-- README_GGUF.md-how-to-run start -->
246
+ ## Example `llama.cpp` command
247
+
248
+ Make sure you are using `llama.cpp` from commit [d0cee0d](https://github.com/ggerganov/llama.cpp/commit/d0cee0d36d5be95a0d9088b674dbb27354107221) or later.
249
+
250
+ ```shell
251
+ ./main -ngl 35 -m tigerbot-70b-chat-v2.Q4_K_M.gguf --color -c 2048 --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:"
252
+ ```
253
+
254
+ Change `-ngl 32` to the number of layers to offload to GPU. Remove it if you don't have GPU acceleration.
255
+
256
+ Change `-c 2048` 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. Note that longer sequence lengths require much more resources, so you may need to reduce this value.
257
+
258
+ If you want to have a chat-style conversation, replace the `-p <PROMPT>` argument with `-i -ins`
259
+
260
+ For other parameters and how to use them, please refer to [the llama.cpp documentation](https://github.com/ggerganov/llama.cpp/blob/master/examples/main/README.md)
261
+
262
+ ## How to run in `text-generation-webui`
263
+
264
+ Further instructions can be found in the text-generation-webui documentation, here: [text-generation-webui/docs/04 ‐ Model Tab.md](https://github.com/oobabooga/text-generation-webui/blob/main/docs/04%20%E2%80%90%20Model%20Tab.md#llamacpp).
265
+
266
+ ## How to run from Python code
267
+
268
+ You can use GGUF models from Python using the [llama-cpp-python](https://github.com/abetlen/llama-cpp-python) or [ctransformers](https://github.com/marella/ctransformers) libraries. Note that at the time of writing (Nov 27th 2023), ctransformers has not been updated for some time and is not compatible with some recent models. Therefore I recommend you use llama-cpp-python.
269
+
270
+ ### How to load this model in Python code, using llama-cpp-python
271
+
272
+ For full documentation, please see: [llama-cpp-python docs](https://abetlen.github.io/llama-cpp-python/).
273
+
274
+ #### First install the package
275
+
276
+ Run one of the following commands, according to your system:
277
+
278
+ ```shell
279
+ # Base ctransformers with no GPU acceleration
280
+ pip install llama-cpp-python
281
+ # With NVidia CUDA acceleration
282
+ CMAKE_ARGS="-DLLAMA_CUBLAS=on" pip install llama-cpp-python
283
+ # Or with OpenBLAS acceleration
284
+ CMAKE_ARGS="-DLLAMA_BLAS=ON -DLLAMA_BLAS_VENDOR=OpenBLAS" pip install llama-cpp-python
285
+ # Or with CLBLast acceleration
286
+ CMAKE_ARGS="-DLLAMA_CLBLAST=on" pip install llama-cpp-python
287
+ # Or with AMD ROCm GPU acceleration (Linux only)
288
+ CMAKE_ARGS="-DLLAMA_HIPBLAS=on" pip install llama-cpp-python
289
+ # Or with Metal GPU acceleration for macOS systems only
290
+ CMAKE_ARGS="-DLLAMA_METAL=on" pip install llama-cpp-python
291
+
292
+ # In windows, to set the variables CMAKE_ARGS in PowerShell, follow this format; eg for NVidia CUDA:
293
+ $env:CMAKE_ARGS = "-DLLAMA_OPENBLAS=on"
294
+ pip install llama-cpp-python
295
+ ```
296
+
297
+ #### Simple llama-cpp-python example code
298
+
299
+ ```python
300
+ from llama_cpp import Llama
301
+
302
+ # Set gpu_layers to the number of layers to offload to GPU. Set to 0 if no GPU acceleration is available on your system.
303
+ llm = Llama(
304
+ model_path="./tigerbot-70b-chat-v2.Q4_K_M.gguf", # Download the model file first
305
+ n_ctx=2048, # The max sequence length to use - note that longer sequence lengths require much more resources
306
+ n_threads=8, # The number of CPU threads to use, tailor to your system and the resulting performance
307
+ n_gpu_layers=35 # The number of layers to offload to GPU, if you have GPU acceleration available
308
+ )
309
+
310
+ # Simple inference example
311
+ output = llm(
312
+ "Below is an instruction that describes a task. Write a response that appropriately completes the request.\n\n### Instruction:\n{prompt}\n\n### Response:", # Prompt
313
+ max_tokens=512, # Generate up to 512 tokens
314
+ stop=["</s>"], # Example stop token - not necessarily correct for this specific model! Please check before using.
315
+ echo=True # Whether to echo the prompt
316
+ )
317
+
318
+ # Chat Completion API
319
+
320
+ llm = Llama(model_path="./tigerbot-70b-chat-v2.Q4_K_M.gguf", chat_format="llama-2") # Set chat_format according to the model you are using
321
+ llm.create_chat_completion(
322
+ messages = [
323
+ {"role": "system", "content": "You are a story writing assistant."},
324
+ {
325
+ "role": "user",
326
+ "content": "Write a story about llamas."
327
+ }
328
+ ]
329
+ )
330
+ ```
331
+
332
+ ## How to use with LangChain
333
+
334
+ Here are guides on using llama-cpp-python and ctransformers with LangChain:
335
+
336
+ * [LangChain + llama-cpp-python](https://python.langchain.com/docs/integrations/llms/llamacpp)
337
+ * [LangChain + ctransformers](https://python.langchain.com/docs/integrations/providers/ctransformers)
338
+
339
+ <!-- README_GGUF.md-how-to-run end -->
340
+
341
+ <!-- footer start -->
342
+ <!-- 200823 -->
343
+ ## Discord
344
+
345
+ For further support, and discussions on these models and AI in general, join us at:
346
+
347
+ [TheBloke AI's Discord server](https://discord.gg/theblokeai)
348
+
349
+ ## Thanks, and how to contribute
350
+
351
+ Thanks to the [chirper.ai](https://chirper.ai) team!
352
+
353
+ Thanks to Clay from [gpus.llm-utils.org](llm-utils)!
354
+
355
+ 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.
356
+
357
+ 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.
358
+
359
+ Donaters will get priority support on any and all AI/LLM/model questions and requests, access to a private Discord room, plus other benefits.
360
+
361
+ * Patreon: https://patreon.com/TheBlokeAI
362
+ * Ko-Fi: https://ko-fi.com/TheBlokeAI
363
+
364
+ **Special thanks to**: Aemon Algiz.
365
+
366
+ **Patreon special mentions**: Brandon Frisco, LangChain4j, Spiking Neurons AB, transmissions 11, Joseph William Delisle, Nitin Borwankar, Willem Michiel, Michael Dempsey, vamX, Jeffrey Morgan, zynix, jjj, Omer Bin Jawed, Sean Connelly, jinyuan sun, Jeromy Smith, Shadi, Pawan Osman, Chadd, Elijah Stavena, Illia Dulskyi, Sebastain Graf, Stephen Murray, terasurfer, Edmond Seymore, Celu Ramasamy, Mandus, Alex, biorpg, Ajan Kanaga, Clay Pascal, Raven Klaugh, 阿明, K, ya boyyy, usrbinkat, Alicia Loh, John Villwock, ReadyPlayerEmma, Chris Smitley, Cap'n Zoog, fincy, GodLy, S_X, sidney chen, Cory Kujawski, OG, Mano Prime, AzureBlack, Pieter, Kalila, Spencer Kim, Tom X Nguyen, Stanislav Ovsiannikov, Michael Levine, Andrey, Trailburnt, Vadim, Enrico Ros, Talal Aujan, Brandon Phillips, Jack West, Eugene Pentland, Michael Davis, Will Dee, webtim, Jonathan Leane, Alps Aficionado, Rooh Singh, Tiffany J. Kim, theTransient, Luke @flexchar, Elle, Caitlyn Gatomon, Ari Malik, subjectnull, Johann-Peter Hartmann, Trenton Dambrowitz, Imad Khwaja, Asp the Wyvern, Emad Mostaque, Rainer Wilmers, Alexandros Triantafyllidis, Nicholas, Pedro Madruga, SuperWojo, Harry Royden McLaughlin, James Bentley, Olakabola, David Ziegler, Ai Maven, Jeff Scroggin, Nikolai Manek, Deo Leter, Matthew Berman, Fen Risland, Ken Nordquist, Manuel Alberto Morcote, Luke Pendergrass, TL, Fred von Graf, Randy H, Dan Guido, NimbleBox.ai, Vitor Caleffi, Gabriel Tamborski, knownsqashed, Lone Striker, Erik Bjäreholt, John Detwiler, Leonard Tan, Iucharbius
367
+
368
+
369
+ Thank you to all my generous patrons and donaters!
370
+
371
+ And thank you again to a16z for their generous grant.
372
+
373
+ <!-- footer end -->
374
+
375
+ <!-- original-model-card start -->
376
+ # Original model card: Tiger Research's Tigerbot 70B Chat V2
377
+
378
+ <div style="width: 100%;">
379
+ <p align="center" width="20%">
380
+ <img src="http://x-pai.algolet.com/bot/img/logo_core.png" alt="TigerBot" width="20%", style="display: block; margin: auto;"></img>
381
+ </p>
382
+ </div>
383
+ <p align="center">
384
+ <font face="黑体" size=5"> A cutting-edge foundation for your very own LLM. </font>
385
+ </p>
386
+ <p align="center">
387
+ 💻<a href="https://github.com/TigerResearch/TigerBot" target="_blank">Github</a> • 🌐 <a href="https://tigerbot.com/" target="_blank">TigerBot</a> • 🤗 <a href="https://huggingface.co/TigerResearch" target="_blank">Hugging Face</a>
388
+ </p>
389
+
390
+ # 快速开始
391
+
392
+ - 方法1,通过transformers使用
393
+
394
+ - 下载 TigerBot Repo
395
+
396
+ ```shell
397
+ git clone https://github.com/TigerResearch/TigerBot.git
398
+ ```
399
+
400
+ - 启动infer代码
401
+
402
+ ```shell
403
+ python infer.py --model_path TigerResearch/tigerbot-70b-chat-v2
404
+ ```
405
+
406
+ - 方法2:
407
+
408
+ - 下载 TigerBot Repo
409
+
410
+ ```shell
411
+ git clone https://github.com/TigerResearch/TigerBot.git
412
+ ```
413
+
414
+ - 安装git lfs: `git lfs install`
415
+
416
+ - 通过huggingface或modelscope平台下载权重
417
+ ```shell
418
+ git clone https://huggingface.co/TigerResearch/tigerbot-70b-chat-v2
419
+ git clone https://www.modelscope.cn/TigerResearch/tigerbot-70b-chat-v2.git
420
+ ```
421
+
422
+ - 启动infer代码
423
+
424
+ ```shell
425
+ python infer.py --model_path tigerbot-70b-chat-v2
426
+ ```
427
+
428
+ ------
429
+
430
+ # Quick Start
431
+
432
+ - Method 1, use through transformers
433
+
434
+ - Clone TigerBot Repo
435
+
436
+ ```shell
437
+ git clone https://github.com/TigerResearch/TigerBot.git
438
+ ```
439
+
440
+ - Run infer script
441
+
442
+ ```shell
443
+ python infer.py --model_path TigerResearch/tigerbot-70b-chat-v2
444
+ ```
445
+
446
+ - Method 2:
447
+
448
+ - Clone TigerBot Repo
449
+
450
+ ```shell
451
+ git clone https://github.com/TigerResearch/TigerBot.git
452
+ ```
453
+
454
+ - install git lfs: `git lfs install`
455
+
456
+ - Download weights from huggingface or modelscope
457
+ ```shell
458
+ git clone https://huggingface.co/TigerResearch/tigerbot-70b-chat-v2
459
+ git clone https://www.modelscope.cn/TigerResearch/tigerbot-70b-chat-v2.git
460
+ ```
461
+
462
+ - Run infer script
463
+
464
+ ```shell
465
+ python infer.py --model_path tigerbot-70b-chat-v2
466
+ ```
467
+
468
+ <!-- original-model-card end -->