TheBloke commited on
Commit
98e6304
1 Parent(s): d755023

Upload README.md

Browse files
Files changed (1) hide show
  1. README.md +25 -17
README.md CHANGED
@@ -1,13 +1,16 @@
1
  ---
 
2
  inference: false
3
  language:
4
  - en
5
  license: llama2
6
  model_creator: Meta
7
- model_link: https://huggingface.co/meta-llama/Llama-2-13b-hf
8
  model_name: Llama 2 13B
9
  model_type: llama
10
  pipeline_tag: text-generation
 
 
 
11
  quantized_by: TheBloke
12
  tags:
13
  - facebook
@@ -49,9 +52,9 @@ Multiple GPTQ parameter permutations are provided; see Provided Files below for
49
  <!-- repositories-available start -->
50
  ## Repositories available
51
 
 
52
  * [GPTQ models for GPU inference, with multiple quantisation parameter options.](https://huggingface.co/TheBloke/Llama-2-13B-GPTQ)
53
  * [2, 3, 4, 5, 6 and 8-bit GGUF models for CPU+GPU inference](https://huggingface.co/TheBloke/Llama-2-13B-GGUF)
54
- * [2, 3, 4, 5, 6 and 8-bit GGML models for CPU+GPU inference (deprecated)](https://huggingface.co/TheBloke/Llama-2-13B-GGML)
55
  * [Meta's original unquantised fp16 model in pytorch format, for GPU inference and for further conversions](https://huggingface.co/meta-llama/Llama-2-13b-hf)
56
  <!-- repositories-available end -->
57
 
@@ -65,6 +68,7 @@ Multiple GPTQ parameter permutations are provided; see Provided Files below for
65
 
66
  <!-- prompt-template end -->
67
 
 
68
  <!-- README_GPTQ.md-provided-files start -->
69
  ## Provided files and GPTQ parameters
70
 
@@ -89,24 +93,24 @@ All recent GPTQ files are made with AutoGPTQ, and all files in non-main branches
89
 
90
  | Branch | Bits | GS | Act Order | Damp % | GPTQ Dataset | Seq Len | Size | ExLlama | Desc |
91
  | ------ | ---- | -- | --------- | ------ | ------------ | ------- | ---- | ------- | ---- |
92
- | [main](https://huggingface.co/TheBloke/Llama-2-13B-GPTQ/tree/main) | 4 | 128 | No | 0.01 | [wikitext](https://huggingface.co/datasets/wikitext/viewer/wikitext-2-v1/test) | 4096 | 7.26 GB | Yes | Most compatible option. Good inference speed in AutoGPTQ and GPTQ-for-LLaMa. Lower inference quality than other options. |
93
- | [gptq-4bit-32g-actorder_True](https://huggingface.co/TheBloke/Llama-2-13B-GPTQ/tree/gptq-4bit-32g-actorder_True) | 4 | 32 | Yes | 0.01 | [wikitext](https://huggingface.co/datasets/wikitext/viewer/wikitext-2-v1/test) | 4096 | 8.00 GB | Yes | 4-bit, with Act Order and group size 32g. Gives highest possible inference quality, with maximum VRAM usage. Poor AutoGPTQ CUDA speed. |
94
- | [gptq-4bit-64g-actorder_True](https://huggingface.co/TheBloke/Llama-2-13B-GPTQ/tree/gptq-4bit-64g-actorder_True) | 4 | 64 | Yes | 0.01 | [wikitext](https://huggingface.co/datasets/wikitext/viewer/wikitext-2-v1/test) | 4096 | 7.51 GB | Yes | 4-bit, with Act Order and group size 64g. Uses less VRAM than 32g, but with slightly lower accuracy. Poor AutoGPTQ CUDA speed. |
95
- | [gptq-4bit-128g-actorder_True](https://huggingface.co/TheBloke/Llama-2-13B-GPTQ/tree/gptq-4bit-128g-actorder_True) | 4 | 128 | Yes | 0.01 | [wikitext](https://huggingface.co/datasets/wikitext/viewer/wikitext-2-v1/test) | 4096 | 7.26 GB | Yes | 4-bit, with Act Order and group size 128g. Uses even less VRAM than 64g, but with slightly lower accuracy. Poor AutoGPTQ CUDA speed. |
96
- | [gptq-8bit-128g-actorder_True](https://huggingface.co/TheBloke/Llama-2-13B-GPTQ/tree/gptq-8bit-128g-actorder_True) | 8 | 128 | Yes | 0.01 | [wikitext](https://huggingface.co/datasets/wikitext/viewer/wikitext-2-v1/test) | 4096 | 13.65 GB | No | 8-bit, with group size 128g for higher inference quality and with Act Order for even higher accuracy. Poor AutoGPTQ CUDA speed. |
97
  | [gptq-8bit-64g-actorder_True](https://huggingface.co/TheBloke/Llama-2-13B-GPTQ/tree/gptq-8bit-64g-actorder_True) | 8 | 64 | Yes | 0.01 | [wikitext](https://huggingface.co/datasets/wikitext/viewer/wikitext-2-v1/test) | 4096 | 13.95 GB | No | 8-bit, with group size 64g and Act Order for even higher inference quality. Poor AutoGPTQ CUDA speed. |
98
  | [gptq-8bit-128g-actorder_False](https://huggingface.co/TheBloke/Llama-2-13B-GPTQ/tree/gptq-8bit-128g-actorder_False) | 8 | 128 | No | 0.01 | [wikitext](https://huggingface.co/datasets/wikitext/viewer/wikitext-2-v1/test) | 4096 | 13.65 GB | No | 8-bit, with group size 128g for higher inference quality and without Act Order to improve AutoGPTQ speed. |
99
- | [gptq-8bit--1g-actorder_True](https://huggingface.co/TheBloke/Llama-2-13B-GPTQ/tree/gptq-8bit--1g-actorder_True) | 8 | None | Yes | 0.01 | [wikitext](https://huggingface.co/datasets/wikitext/viewer/wikitext-2-v1/test) | 4096 | 13.36 GB | No | 8-bit, with Act Order. No group size, to lower VRAM requirements and to improve AutoGPTQ speed. |
100
 
101
  <!-- README_GPTQ.md-provided-files end -->
102
 
103
  <!-- README_GPTQ.md-download-from-branches start -->
104
  ## How to download from branches
105
 
106
- - In text-generation-webui, you can add `:branch` to the end of the download name, eg `TheBloke/Llama-2-13B-GPTQ:gptq-4bit-32g-actorder_True`
107
  - With Git, you can clone a branch with:
108
  ```
109
- git clone --single-branch --branch gptq-4bit-32g-actorder_True https://huggingface.co/TheBloke/Llama-2-13B-GPTQ
110
  ```
111
  - In Python Transformers code, the branch is the `revision` parameter; see below.
112
  <!-- README_GPTQ.md-download-from-branches end -->
@@ -119,7 +123,7 @@ It is strongly recommended to use the text-generation-webui one-click-installers
119
 
120
  1. Click the **Model tab**.
121
  2. Under **Download custom model or LoRA**, enter `TheBloke/Llama-2-13B-GPTQ`.
122
- - To download from a specific branch, enter for example `TheBloke/Llama-2-13B-GPTQ:gptq-4bit-32g-actorder_True`
123
  - see Provided Files above for the list of branches for each option.
124
  3. Click **Download**.
125
  4. The model will start downloading. Once it's finished it will say "Done".
@@ -167,10 +171,10 @@ from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline
167
 
168
  model_name_or_path = "TheBloke/Llama-2-13B-GPTQ"
169
  # To use a different branch, change revision
170
- # For example: revision="gptq-4bit-32g-actorder_True"
171
  model = AutoModelForCausalLM.from_pretrained(model_name_or_path,
172
- torch_dtype=torch.float16,
173
  device_map="auto",
 
174
  revision="main")
175
 
176
  tokenizer = AutoTokenizer.from_pretrained(model_name_or_path, use_fast=True)
@@ -183,7 +187,7 @@ prompt_template=f'''{prompt}
183
  print("\n\n*** Generate:")
184
 
185
  input_ids = tokenizer(prompt_template, return_tensors='pt').input_ids.cuda()
186
- output = model.generate(inputs=input_ids, temperature=0.7, max_new_tokens=512)
187
  print(tokenizer.decode(output[0]))
188
 
189
  # Inference can also be done using transformers' pipeline
@@ -194,9 +198,11 @@ pipe = pipeline(
194
  model=model,
195
  tokenizer=tokenizer,
196
  max_new_tokens=512,
 
197
  temperature=0.7,
198
  top_p=0.95,
199
- repetition_penalty=1.15
 
200
  )
201
 
202
  print(pipe(prompt_template)[0]['generated_text'])
@@ -221,10 +227,12 @@ For further support, and discussions on these models and AI in general, join us
221
 
222
  [TheBloke AI's Discord server](https://discord.gg/theblokeai)
223
 
224
- ## Thanks, and how to contribute.
225
 
226
  Thanks to the [chirper.ai](https://chirper.ai) team!
227
 
 
 
228
  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.
229
 
230
  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.
@@ -236,7 +244,7 @@ Donaters will get priority support on any and all AI/LLM/model questions and req
236
 
237
  **Special thanks to**: Aemon Algiz.
238
 
239
- **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
240
 
241
 
242
  Thank you to all my generous patrons and donaters!
 
1
  ---
2
+ base_model: https://huggingface.co/meta-llama/Llama-2-13b-hf
3
  inference: false
4
  language:
5
  - en
6
  license: llama2
7
  model_creator: Meta
 
8
  model_name: Llama 2 13B
9
  model_type: llama
10
  pipeline_tag: text-generation
11
+ prompt_template: '{prompt}
12
+
13
+ '
14
  quantized_by: TheBloke
15
  tags:
16
  - facebook
 
52
  <!-- repositories-available start -->
53
  ## Repositories available
54
 
55
+ * [AWQ model(s) for GPU inference.](https://huggingface.co/TheBloke/Llama-2-13B-AWQ)
56
  * [GPTQ models for GPU inference, with multiple quantisation parameter options.](https://huggingface.co/TheBloke/Llama-2-13B-GPTQ)
57
  * [2, 3, 4, 5, 6 and 8-bit GGUF models for CPU+GPU inference](https://huggingface.co/TheBloke/Llama-2-13B-GGUF)
 
58
  * [Meta's original unquantised fp16 model in pytorch format, for GPU inference and for further conversions](https://huggingface.co/meta-llama/Llama-2-13b-hf)
59
  <!-- repositories-available end -->
60
 
 
68
 
69
  <!-- prompt-template end -->
70
 
71
+
72
  <!-- README_GPTQ.md-provided-files start -->
73
  ## Provided files and GPTQ parameters
74
 
 
93
 
94
  | Branch | Bits | GS | Act Order | Damp % | GPTQ Dataset | Seq Len | Size | ExLlama | Desc |
95
  | ------ | ---- | -- | --------- | ------ | ------------ | ------- | ---- | ------- | ---- |
96
+ | [main](https://huggingface.co/TheBloke/Llama-2-13B-GPTQ/tree/main) | 4 | 128 | No | 0.01 | [wikitext](https://huggingface.co/datasets/wikitext/viewer/wikitext-2-v1/test) | 4096 | 7.26 GB | Yes | 4-bit, without Act Order and group size 128g. |
97
+ | [gptq-4bit-32g-actorder_True](https://huggingface.co/TheBloke/Llama-2-13B-GPTQ/tree/gptq-4bit-32g-actorder_True) | 4 | 32 | Yes | 0.01 | [wikitext](https://huggingface.co/datasets/wikitext/viewer/wikitext-2-v1/test) | 4096 | 8.00 GB | Yes | 4-bit, with Act Order and group size 32g. Gives highest possible inference quality, with maximum VRAM usage. |
98
+ | [gptq-4bit-64g-actorder_True](https://huggingface.co/TheBloke/Llama-2-13B-GPTQ/tree/gptq-4bit-64g-actorder_True) | 4 | 64 | Yes | 0.01 | [wikitext](https://huggingface.co/datasets/wikitext/viewer/wikitext-2-v1/test) | 4096 | 7.51 GB | Yes | 4-bit, with Act Order and group size 64g. Uses less VRAM than 32g, but with slightly lower accuracy. |
99
+ | [gptq-4bit-128g-actorder_True](https://huggingface.co/TheBloke/Llama-2-13B-GPTQ/tree/gptq-4bit-128g-actorder_True) | 4 | 128 | Yes | 0.01 | [wikitext](https://huggingface.co/datasets/wikitext/viewer/wikitext-2-v1/test) | 4096 | 7.26 GB | Yes | 4-bit, with Act Order and group size 128g. Uses even less VRAM than 64g, but with slightly lower accuracy. |
100
+ | [gptq-8bit-128g-actorder_True](https://huggingface.co/TheBloke/Llama-2-13B-GPTQ/tree/gptq-8bit-128g-actorder_True) | 8 | 128 | Yes | 0.01 | [wikitext](https://huggingface.co/datasets/wikitext/viewer/wikitext-2-v1/test) | 4096 | 13.65 GB | No | 8-bit, with group size 128g for higher inference quality and with Act Order for even higher accuracy. |
101
  | [gptq-8bit-64g-actorder_True](https://huggingface.co/TheBloke/Llama-2-13B-GPTQ/tree/gptq-8bit-64g-actorder_True) | 8 | 64 | Yes | 0.01 | [wikitext](https://huggingface.co/datasets/wikitext/viewer/wikitext-2-v1/test) | 4096 | 13.95 GB | No | 8-bit, with group size 64g and Act Order for even higher inference quality. Poor AutoGPTQ CUDA speed. |
102
  | [gptq-8bit-128g-actorder_False](https://huggingface.co/TheBloke/Llama-2-13B-GPTQ/tree/gptq-8bit-128g-actorder_False) | 8 | 128 | No | 0.01 | [wikitext](https://huggingface.co/datasets/wikitext/viewer/wikitext-2-v1/test) | 4096 | 13.65 GB | No | 8-bit, with group size 128g for higher inference quality and without Act Order to improve AutoGPTQ speed. |
103
+ | [gptq-8bit--1g-actorder_True](https://huggingface.co/TheBloke/Llama-2-13B-GPTQ/tree/gptq-8bit--1g-actorder_True) | 8 | None | Yes | 0.01 | [wikitext](https://huggingface.co/datasets/wikitext/viewer/wikitext-2-v1/test) | 4096 | 13.36 GB | No | 8-bit, with Act Order. No group size, to lower VRAM requirements. |
104
 
105
  <!-- README_GPTQ.md-provided-files end -->
106
 
107
  <!-- README_GPTQ.md-download-from-branches start -->
108
  ## How to download from branches
109
 
110
+ - In text-generation-webui, you can add `:branch` to the end of the download name, eg `TheBloke/Llama-2-13B-GPTQ:main`
111
  - With Git, you can clone a branch with:
112
  ```
113
+ git clone --single-branch --branch main https://huggingface.co/TheBloke/Llama-2-13B-GPTQ
114
  ```
115
  - In Python Transformers code, the branch is the `revision` parameter; see below.
116
  <!-- README_GPTQ.md-download-from-branches end -->
 
123
 
124
  1. Click the **Model tab**.
125
  2. Under **Download custom model or LoRA**, enter `TheBloke/Llama-2-13B-GPTQ`.
126
+ - To download from a specific branch, enter for example `TheBloke/Llama-2-13B-GPTQ:main`
127
  - see Provided Files above for the list of branches for each option.
128
  3. Click **Download**.
129
  4. The model will start downloading. Once it's finished it will say "Done".
 
171
 
172
  model_name_or_path = "TheBloke/Llama-2-13B-GPTQ"
173
  # To use a different branch, change revision
174
+ # For example: revision="main"
175
  model = AutoModelForCausalLM.from_pretrained(model_name_or_path,
 
176
  device_map="auto",
177
+ trust_remote_code=False,
178
  revision="main")
179
 
180
  tokenizer = AutoTokenizer.from_pretrained(model_name_or_path, use_fast=True)
 
187
  print("\n\n*** Generate:")
188
 
189
  input_ids = tokenizer(prompt_template, return_tensors='pt').input_ids.cuda()
190
+ output = model.generate(inputs=input_ids, temperature=0.7, do_sample=True, top_p=0.95, top_k=40, max_new_tokens=512)
191
  print(tokenizer.decode(output[0]))
192
 
193
  # Inference can also be done using transformers' pipeline
 
198
  model=model,
199
  tokenizer=tokenizer,
200
  max_new_tokens=512,
201
+ do_sample=True,
202
  temperature=0.7,
203
  top_p=0.95,
204
+ top_k=40,
205
+ repetition_penalty=1.1
206
  )
207
 
208
  print(pipe(prompt_template)[0]['generated_text'])
 
227
 
228
  [TheBloke AI's Discord server](https://discord.gg/theblokeai)
229
 
230
+ ## Thanks, and how to contribute
231
 
232
  Thanks to the [chirper.ai](https://chirper.ai) team!
233
 
234
+ Thanks to Clay from [gpus.llm-utils.org](llm-utils)!
235
+
236
  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.
237
 
238
  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.
 
244
 
245
  **Special thanks to**: Aemon Algiz.
246
 
247
+ **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
248
 
249
 
250
  Thank you to all my generous patrons and donaters!