TheBloke commited on
Commit
37a0e16
1 Parent(s): 8df6fa9

Upload README.md

Browse files
Files changed (1) hide show
  1. README.md +33 -17
README.md CHANGED
@@ -1,12 +1,23 @@
1
  ---
 
2
  datasets:
3
  - rombodawg/LosslessMegaCodeTrainingV2_1m_Evol_Uncensored
4
  inference: false
5
  license: llama2
6
  model_creator: Rombo Dawg
7
- model_link: https://huggingface.co/rombodawg/LosslessMegaCoder-llama2-13b-mini
8
  model_name: LosslessMegaCoder Llama2 13B Mini
9
  model_type: llama
 
 
 
 
 
 
 
 
 
 
 
10
  quantized_by: TheBloke
11
  ---
12
 
@@ -42,9 +53,9 @@ Multiple GPTQ parameter permutations are provided; see Provided Files below for
42
  <!-- repositories-available start -->
43
  ## Repositories available
44
 
 
45
  * [GPTQ models for GPU inference, with multiple quantisation parameter options.](https://huggingface.co/TheBloke/LosslessMegaCoder-Llama2-13B-Mini-GPTQ)
46
  * [2, 3, 4, 5, 6 and 8-bit GGUF models for CPU+GPU inference](https://huggingface.co/TheBloke/LosslessMegaCoder-Llama2-13B-Mini-GGUF)
47
- * [2, 3, 4, 5, 6 and 8-bit GGML models for CPU+GPU inference (deprecated)](https://huggingface.co/TheBloke/LosslessMegaCoder-Llama2-13B-Mini-GGML)
48
  * [Rombo Dawg's original unquantised fp16 model in pytorch format, for GPU inference and for further conversions](https://huggingface.co/rombodawg/LosslessMegaCoder-llama2-13b-mini)
49
  <!-- repositories-available end -->
50
 
@@ -62,6 +73,7 @@ Multiple GPTQ parameter permutations are provided; see Provided Files below for
62
 
63
  <!-- prompt-template end -->
64
 
 
65
  <!-- README_GPTQ.md-provided-files start -->
66
  ## Provided files and GPTQ parameters
67
 
@@ -86,22 +98,22 @@ All recent GPTQ files are made with AutoGPTQ, and all files in non-main branches
86
 
87
  | Branch | Bits | GS | Act Order | Damp % | GPTQ Dataset | Seq Len | Size | ExLlama | Desc |
88
  | ------ | ---- | -- | --------- | ------ | ------------ | ------- | ---- | ------- | ---- |
89
- | [main](https://huggingface.co/TheBloke/LosslessMegaCoder-Llama2-13B-Mini-GPTQ/tree/main) | 4 | 128 | No | 0.1 | [Evol Instruct Code](https://huggingface.co/datasets/nickrosh/Evol-Instruct-Code-80k-v1) | 4096 | 7.26 GB | Yes | Most compatible option. Good inference speed in AutoGPTQ and GPTQ-for-LLaMa. Lower inference quality than other options. |
90
- | [gptq-4bit-32g-actorder_True](https://huggingface.co/TheBloke/LosslessMegaCoder-Llama2-13B-Mini-GPTQ/tree/gptq-4bit-32g-actorder_True) | 4 | 32 | Yes | 0.1 | [Evol Instruct Code](https://huggingface.co/datasets/nickrosh/Evol-Instruct-Code-80k-v1) | 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. |
91
- | [gptq-4bit-64g-actorder_True](https://huggingface.co/TheBloke/LosslessMegaCoder-Llama2-13B-Mini-GPTQ/tree/gptq-4bit-64g-actorder_True) | 4 | 64 | Yes | 0.1 | [Evol Instruct Code](https://huggingface.co/datasets/nickrosh/Evol-Instruct-Code-80k-v1) | 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. |
92
- | [gptq-4bit-128g-actorder_True](https://huggingface.co/TheBloke/LosslessMegaCoder-Llama2-13B-Mini-GPTQ/tree/gptq-4bit-128g-actorder_True) | 4 | 128 | Yes | 0.1 | [Evol Instruct Code](https://huggingface.co/datasets/nickrosh/Evol-Instruct-Code-80k-v1) | 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. |
93
- | [gptq-8bit--1g-actorder_True](https://huggingface.co/TheBloke/LosslessMegaCoder-Llama2-13B-Mini-GPTQ/tree/gptq-8bit--1g-actorder_True) | 8 | None | Yes | 0.1 | [Evol Instruct Code](https://huggingface.co/datasets/nickrosh/Evol-Instruct-Code-80k-v1) | 4096 | 13.36 GB | No | 8-bit, with Act Order. No group size, to lower VRAM requirements and to improve AutoGPTQ speed. |
94
- | [gptq-8bit-128g-actorder_True](https://huggingface.co/TheBloke/LosslessMegaCoder-Llama2-13B-Mini-GPTQ/tree/gptq-8bit-128g-actorder_True) | 8 | 128 | Yes | 0.1 | [Evol Instruct Code](https://huggingface.co/datasets/nickrosh/Evol-Instruct-Code-80k-v1) | 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. |
95
 
96
  <!-- README_GPTQ.md-provided-files end -->
97
 
98
  <!-- README_GPTQ.md-download-from-branches start -->
99
  ## How to download from branches
100
 
101
- - In text-generation-webui, you can add `:branch` to the end of the download name, eg `TheBloke/LosslessMegaCoder-Llama2-13B-Mini-GPTQ:gptq-4bit-32g-actorder_True`
102
  - With Git, you can clone a branch with:
103
  ```
104
- git clone --single-branch --branch gptq-4bit-32g-actorder_True https://huggingface.co/TheBloke/LosslessMegaCoder-Llama2-13B-Mini-GPTQ
105
  ```
106
  - In Python Transformers code, the branch is the `revision` parameter; see below.
107
  <!-- README_GPTQ.md-download-from-branches end -->
@@ -114,7 +126,7 @@ It is strongly recommended to use the text-generation-webui one-click-installers
114
 
115
  1. Click the **Model tab**.
116
  2. Under **Download custom model or LoRA**, enter `TheBloke/LosslessMegaCoder-Llama2-13B-Mini-GPTQ`.
117
- - To download from a specific branch, enter for example `TheBloke/LosslessMegaCoder-Llama2-13B-Mini-GPTQ:gptq-4bit-32g-actorder_True`
118
  - see Provided Files above for the list of branches for each option.
119
  3. Click **Download**.
120
  4. The model will start downloading. Once it's finished it will say "Done".
@@ -162,10 +174,10 @@ from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline
162
 
163
  model_name_or_path = "TheBloke/LosslessMegaCoder-Llama2-13B-Mini-GPTQ"
164
  # To use a different branch, change revision
165
- # For example: revision="gptq-4bit-32g-actorder_True"
166
  model = AutoModelForCausalLM.from_pretrained(model_name_or_path,
167
- torch_dtype=torch.float16,
168
  device_map="auto",
 
169
  revision="main")
170
 
171
  tokenizer = AutoTokenizer.from_pretrained(model_name_or_path, use_fast=True)
@@ -182,7 +194,7 @@ prompt_template=f'''<|im_start|>system
182
  print("\n\n*** Generate:")
183
 
184
  input_ids = tokenizer(prompt_template, return_tensors='pt').input_ids.cuda()
185
- output = model.generate(inputs=input_ids, temperature=0.7, max_new_tokens=512)
186
  print(tokenizer.decode(output[0]))
187
 
188
  # Inference can also be done using transformers' pipeline
@@ -193,9 +205,11 @@ pipe = pipeline(
193
  model=model,
194
  tokenizer=tokenizer,
195
  max_new_tokens=512,
 
196
  temperature=0.7,
197
  top_p=0.95,
198
- repetition_penalty=1.15
 
199
  )
200
 
201
  print(pipe(prompt_template)[0]['generated_text'])
@@ -220,10 +234,12 @@ For further support, and discussions on these models and AI in general, join us
220
 
221
  [TheBloke AI's Discord server](https://discord.gg/theblokeai)
222
 
223
- ## Thanks, and how to contribute.
224
 
225
  Thanks to the [chirper.ai](https://chirper.ai) team!
226
 
 
 
227
  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.
228
 
229
  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.
@@ -235,7 +251,7 @@ Donaters will get priority support on any and all AI/LLM/model questions and req
235
 
236
  **Special thanks to**: Aemon Algiz.
237
 
238
- **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
239
 
240
 
241
  Thank you to all my generous patrons and donaters!
 
1
  ---
2
+ base_model: https://huggingface.co/rombodawg/LosslessMegaCoder-llama2-13b-mini
3
  datasets:
4
  - rombodawg/LosslessMegaCodeTrainingV2_1m_Evol_Uncensored
5
  inference: false
6
  license: llama2
7
  model_creator: Rombo Dawg
 
8
  model_name: LosslessMegaCoder Llama2 13B Mini
9
  model_type: llama
10
+ prompt_template: '<|im_start|>system
11
+
12
+ {system_message}<|im_end|>
13
+
14
+ <|im_start|>user
15
+
16
+ {prompt}<|im_end|>
17
+
18
+ <|im_start|>assistant
19
+
20
+ '
21
  quantized_by: TheBloke
22
  ---
23
 
 
53
  <!-- repositories-available start -->
54
  ## Repositories available
55
 
56
+ * [AWQ model(s) for GPU inference.](https://huggingface.co/TheBloke/LosslessMegaCoder-Llama2-13B-Mini-AWQ)
57
  * [GPTQ models for GPU inference, with multiple quantisation parameter options.](https://huggingface.co/TheBloke/LosslessMegaCoder-Llama2-13B-Mini-GPTQ)
58
  * [2, 3, 4, 5, 6 and 8-bit GGUF models for CPU+GPU inference](https://huggingface.co/TheBloke/LosslessMegaCoder-Llama2-13B-Mini-GGUF)
 
59
  * [Rombo Dawg's original unquantised fp16 model in pytorch format, for GPU inference and for further conversions](https://huggingface.co/rombodawg/LosslessMegaCoder-llama2-13b-mini)
60
  <!-- repositories-available end -->
61
 
 
73
 
74
  <!-- prompt-template end -->
75
 
76
+
77
  <!-- README_GPTQ.md-provided-files start -->
78
  ## Provided files and GPTQ parameters
79
 
 
98
 
99
  | Branch | Bits | GS | Act Order | Damp % | GPTQ Dataset | Seq Len | Size | ExLlama | Desc |
100
  | ------ | ---- | -- | --------- | ------ | ------------ | ------- | ---- | ------- | ---- |
101
+ | [main](https://huggingface.co/TheBloke/LosslessMegaCoder-Llama2-13B-Mini-GPTQ/tree/main) | 4 | 128 | No | 0.1 | [Evol Instruct Code](https://huggingface.co/datasets/nickrosh/Evol-Instruct-Code-80k-v1) | 4096 | 7.26 GB | Yes | 4-bit, without Act Order and group size 128g. |
102
+ | [gptq-4bit-32g-actorder_True](https://huggingface.co/TheBloke/LosslessMegaCoder-Llama2-13B-Mini-GPTQ/tree/gptq-4bit-32g-actorder_True) | 4 | 32 | Yes | 0.1 | [Evol Instruct Code](https://huggingface.co/datasets/nickrosh/Evol-Instruct-Code-80k-v1) | 4096 | 8.00 GB | Yes | 4-bit, with Act Order and group size 32g. Gives highest possible inference quality, with maximum VRAM usage. |
103
+ | [gptq-4bit-64g-actorder_True](https://huggingface.co/TheBloke/LosslessMegaCoder-Llama2-13B-Mini-GPTQ/tree/gptq-4bit-64g-actorder_True) | 4 | 64 | Yes | 0.1 | [Evol Instruct Code](https://huggingface.co/datasets/nickrosh/Evol-Instruct-Code-80k-v1) | 4096 | 7.51 GB | Yes | 4-bit, with Act Order and group size 64g. Uses less VRAM than 32g, but with slightly lower accuracy. |
104
+ | [gptq-4bit-128g-actorder_True](https://huggingface.co/TheBloke/LosslessMegaCoder-Llama2-13B-Mini-GPTQ/tree/gptq-4bit-128g-actorder_True) | 4 | 128 | Yes | 0.1 | [Evol Instruct Code](https://huggingface.co/datasets/nickrosh/Evol-Instruct-Code-80k-v1) | 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. |
105
+ | [gptq-8bit--1g-actorder_True](https://huggingface.co/TheBloke/LosslessMegaCoder-Llama2-13B-Mini-GPTQ/tree/gptq-8bit--1g-actorder_True) | 8 | None | Yes | 0.1 | [Evol Instruct Code](https://huggingface.co/datasets/nickrosh/Evol-Instruct-Code-80k-v1) | 4096 | 13.36 GB | No | 8-bit, with Act Order. No group size, to lower VRAM requirements. |
106
+ | [gptq-8bit-128g-actorder_True](https://huggingface.co/TheBloke/LosslessMegaCoder-Llama2-13B-Mini-GPTQ/tree/gptq-8bit-128g-actorder_True) | 8 | 128 | Yes | 0.1 | [Evol Instruct Code](https://huggingface.co/datasets/nickrosh/Evol-Instruct-Code-80k-v1) | 4096 | 13.65 GB | No | 8-bit, with group size 128g for higher inference quality and with Act Order for even higher accuracy. |
107
 
108
  <!-- README_GPTQ.md-provided-files end -->
109
 
110
  <!-- README_GPTQ.md-download-from-branches start -->
111
  ## How to download from branches
112
 
113
+ - In text-generation-webui, you can add `:branch` to the end of the download name, eg `TheBloke/LosslessMegaCoder-Llama2-13B-Mini-GPTQ:main`
114
  - With Git, you can clone a branch with:
115
  ```
116
+ git clone --single-branch --branch main https://huggingface.co/TheBloke/LosslessMegaCoder-Llama2-13B-Mini-GPTQ
117
  ```
118
  - In Python Transformers code, the branch is the `revision` parameter; see below.
119
  <!-- README_GPTQ.md-download-from-branches end -->
 
126
 
127
  1. Click the **Model tab**.
128
  2. Under **Download custom model or LoRA**, enter `TheBloke/LosslessMegaCoder-Llama2-13B-Mini-GPTQ`.
129
+ - To download from a specific branch, enter for example `TheBloke/LosslessMegaCoder-Llama2-13B-Mini-GPTQ:main`
130
  - see Provided Files above for the list of branches for each option.
131
  3. Click **Download**.
132
  4. The model will start downloading. Once it's finished it will say "Done".
 
174
 
175
  model_name_or_path = "TheBloke/LosslessMegaCoder-Llama2-13B-Mini-GPTQ"
176
  # To use a different branch, change revision
177
+ # For example: revision="main"
178
  model = AutoModelForCausalLM.from_pretrained(model_name_or_path,
 
179
  device_map="auto",
180
+ trust_remote_code=False,
181
  revision="main")
182
 
183
  tokenizer = AutoTokenizer.from_pretrained(model_name_or_path, use_fast=True)
 
194
  print("\n\n*** Generate:")
195
 
196
  input_ids = tokenizer(prompt_template, return_tensors='pt').input_ids.cuda()
197
+ output = model.generate(inputs=input_ids, temperature=0.7, do_sample=True, top_p=0.95, top_k=40, max_new_tokens=512)
198
  print(tokenizer.decode(output[0]))
199
 
200
  # Inference can also be done using transformers' pipeline
 
205
  model=model,
206
  tokenizer=tokenizer,
207
  max_new_tokens=512,
208
+ do_sample=True,
209
  temperature=0.7,
210
  top_p=0.95,
211
+ top_k=40,
212
+ repetition_penalty=1.1
213
  )
214
 
215
  print(pipe(prompt_template)[0]['generated_text'])
 
234
 
235
  [TheBloke AI's Discord server](https://discord.gg/theblokeai)
236
 
237
+ ## Thanks, and how to contribute
238
 
239
  Thanks to the [chirper.ai](https://chirper.ai) team!
240
 
241
+ Thanks to Clay from [gpus.llm-utils.org](llm-utils)!
242
+
243
  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.
244
 
245
  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.
 
251
 
252
  **Special thanks to**: Aemon Algiz.
253
 
254
+ **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
255
 
256
 
257
  Thank you to all my generous patrons and donaters!