TheBloke commited on
Commit
914ed5e
1 Parent(s): f2492e3

Upload README.md

Browse files
Files changed (1) hide show
  1. README.md +57 -42
README.md CHANGED
@@ -39,18 +39,24 @@ tags:
39
  - Model creator: [Voicelab](https://huggingface.co/Voicelab)
40
  - Original model: [Trurl 2 13B](https://huggingface.co/Voicelab/trurl-2-13b)
41
 
 
42
  ## Description
43
 
44
  This repo contains GPTQ model files for [Voicelab's Trurl 2 13B](https://huggingface.co/Voicelab/trurl-2-13b).
45
 
46
  Multiple GPTQ parameter permutations are provided; see Provided Files below for details of the options provided, their parameters, and the software used to create them.
47
 
 
 
48
  ## Repositories available
49
 
50
  * [GPTQ models for GPU inference, with multiple quantisation parameter options.](https://huggingface.co/TheBloke/Trurl-2-13B-GPTQ)
51
- * [2, 3, 4, 5, 6 and 8-bit GGML models for CPU+GPU inference](https://huggingface.co/TheBloke/Trurl-2-13B-GGML)
 
52
  * [Voicelab's original unquantised fp16 model in pytorch format, for GPU inference and for further conversions](https://huggingface.co/Voicelab/trurl-2-13b)
 
53
 
 
54
  ## Prompt template: Llama-2-Chat
55
 
56
  ```
@@ -58,22 +64,26 @@ Multiple GPTQ parameter permutations are provided; see Provided Files below for
58
  You 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.
59
  <</SYS>>
60
  {prompt}[/INST]
 
61
  ```
62
 
 
 
 
63
  ## Provided files and GPTQ parameters
64
 
65
  Multiple quantisation parameters are provided, to allow you to choose the best one for your hardware and requirements.
66
 
67
  Each separate quant is in a different branch. See below for instructions on fetching from different branches.
68
 
69
- All GPTQ files are made with AutoGPTQ.
70
 
71
  <details>
72
  <summary>Explanation of GPTQ parameters</summary>
73
 
74
  - Bits: The bit size of the quantised model.
75
  - GS: GPTQ group size. Higher numbers use less VRAM, but have lower quantisation accuracy. "None" is the lowest possible value.
76
- - Act Order: True or False. Also known as `desc_act`. True results in better quantisation accuracy. Some GPTQ clients have issues with models that use Act Order plus Group Size.
77
  - Damp %: A GPTQ parameter that affects how samples are processed for quantisation. 0.01 is default, but 0.1 results in slightly better accuracy.
78
  - GPTQ dataset: The dataset used for quantisation. Using a dataset more appropriate to the model's training can improve quantisation accuracy. Note that the GPTQ dataset is not the same as the dataset used to train the model - please refer to the original model repo for details of the training dataset(s).
79
  - Sequence Length: The length of the dataset sequences used for quantisation. Ideally this is the same as the model sequence length. For some very long sequence models (16+K), a lower sequence length may have to be used. Note that a lower sequence length does not limit the sequence length of the quantised model. It only impacts the quantisation accuracy on longer inference sequences.
@@ -90,6 +100,9 @@ All GPTQ files are made with AutoGPTQ.
90
  | [gptq-8bit--1g-actorder_True](https://huggingface.co/TheBloke/Trurl-2-13B-GPTQ/tree/gptq-8bit--1g-actorder_True) | 8 | None | Yes | 0.1 | [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. |
91
  | [gptq-8bit-128g-actorder_True](https://huggingface.co/TheBloke/Trurl-2-13B-GPTQ/tree/gptq-8bit-128g-actorder_True) | 8 | 128 | Yes | 0.1 | [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. |
92
 
 
 
 
93
  ## How to download from branches
94
 
95
  - In text-generation-webui, you can add `:branch` to the end of the download name, eg `TheBloke/Trurl-2-13B-GPTQ:gptq-4bit-32g-actorder_True`
@@ -98,78 +111,78 @@ All GPTQ files are made with AutoGPTQ.
98
  git clone --single-branch --branch gptq-4bit-32g-actorder_True https://huggingface.co/TheBloke/Trurl-2-13B-GPTQ
99
  ```
100
  - In Python Transformers code, the branch is the `revision` parameter; see below.
101
-
 
102
  ## How to easily download and use this model in [text-generation-webui](https://github.com/oobabooga/text-generation-webui).
103
 
104
  Please make sure you're using the latest version of [text-generation-webui](https://github.com/oobabooga/text-generation-webui).
105
 
106
- It is strongly recommended to use the text-generation-webui one-click-installers unless you know how to make a manual install.
107
 
108
  1. Click the **Model tab**.
109
  2. Under **Download custom model or LoRA**, enter `TheBloke/Trurl-2-13B-GPTQ`.
110
  - To download from a specific branch, enter for example `TheBloke/Trurl-2-13B-GPTQ:gptq-4bit-32g-actorder_True`
111
  - see Provided Files above for the list of branches for each option.
112
  3. Click **Download**.
113
- 4. The model will start downloading. Once it's finished it will say "Done"
114
  5. In the top left, click the refresh icon next to **Model**.
115
  6. In the **Model** dropdown, choose the model you just downloaded: `Trurl-2-13B-GPTQ`
116
  7. The model will automatically load, and is now ready for use!
117
  8. If you want any custom settings, set them and then click **Save settings for this model** followed by **Reload the Model** in the top right.
118
- * Note that you do not need to set GPTQ parameters any more. These are set automatically from the file `quantize_config.json`.
119
  9. Once you're ready, click the **Text Generation tab** and enter a prompt to get started!
 
120
 
 
121
  ## How to use this GPTQ model from Python code
122
 
123
- First make sure you have [AutoGPTQ](https://github.com/PanQiWei/AutoGPTQ) 0.3.1 or later installed:
124
 
125
- ```
126
- pip3 install auto-gptq
127
- ```
128
 
129
- If you have problems installing AutoGPTQ, please build from source instead:
 
 
130
  ```
 
 
 
 
131
  pip3 uninstall -y auto-gptq
132
  git clone https://github.com/PanQiWei/AutoGPTQ
133
  cd AutoGPTQ
134
  pip3 install .
135
  ```
136
 
137
- Then try the following example code:
 
 
 
 
 
 
 
 
138
 
139
  ```python
140
- from transformers import AutoTokenizer, pipeline, logging
141
- from auto_gptq import AutoGPTQForCausalLM, BaseQuantizeConfig
142
 
143
  model_name_or_path = "TheBloke/Trurl-2-13B-GPTQ"
144
-
145
- use_triton = False
 
 
 
 
146
 
147
  tokenizer = AutoTokenizer.from_pretrained(model_name_or_path, use_fast=True)
148
 
149
- model = AutoGPTQForCausalLM.from_quantized(model_name_or_path,
150
- use_safetensors=True,
151
- trust_remote_code=False,
152
- device="cuda:0",
153
- use_triton=use_triton,
154
- quantize_config=None)
155
-
156
- """
157
- # To download from a specific branch, use the revision parameter, as in this example:
158
- # Note that `revision` requires AutoGPTQ 0.3.1 or later!
159
-
160
- model = AutoGPTQForCausalLM.from_quantized(model_name_or_path,
161
- revision="gptq-4bit-32g-actorder_True",
162
- use_safetensors=True,
163
- trust_remote_code=False,
164
- device="cuda:0",
165
- quantize_config=None)
166
- """
167
-
168
  prompt = "Tell me about AI"
169
  prompt_template=f'''[INST] <<SYS>>
170
  You 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.
171
  <</SYS>>
172
  {prompt}[/INST]
 
173
  '''
174
 
175
  print("\n\n*** Generate:")
@@ -180,9 +193,6 @@ print(tokenizer.decode(output[0]))
180
 
181
  # Inference can also be done using transformers' pipeline
182
 
183
- # Prevent printing spurious transformers error when using pipeline with AutoGPTQ
184
- logging.set_verbosity(logging.CRITICAL)
185
-
186
  print("*** Pipeline:")
187
  pipe = pipeline(
188
  "text-generation",
@@ -196,12 +206,17 @@ pipe = pipeline(
196
 
197
  print(pipe(prompt_template)[0]['generated_text'])
198
  ```
 
199
 
 
200
  ## Compatibility
201
 
202
- The files provided will work with AutoGPTQ (CUDA and Triton modes), GPTQ-for-LLaMa (only CUDA has been tested), and Occ4m's GPTQ-for-LLaMa fork.
 
 
203
 
204
- ExLlama works with Llama models in 4-bit. Please see the Provided Files table above for per-file compatibility.
 
205
 
206
  <!-- footer start -->
207
  <!-- 200823 -->
@@ -226,7 +241,7 @@ Donaters will get priority support on any and all AI/LLM/model questions and req
226
 
227
  **Special thanks to**: Aemon Algiz.
228
 
229
- **Patreon special mentions**: Sam, theTransient, Jonathan Leane, Steven Wood, webtim, Johann-Peter Hartmann, Geoffrey Montalvo, Gabriel Tamborski, Willem Michiel, John Villwock, Derek Yates, Mesiah Bishop, Eugene Pentland, Pieter, Chadd, Stephen Murray, Daniel P. Andersen, terasurfer, Brandon Frisco, Thomas Belote, Sid, Nathan LeClaire, Magnesian, Alps Aficionado, Stanislav Ovsiannikov, Alex, Joseph William Delisle, Nikolai Manek, Michael Davis, Junyu Yang, K, J, Spencer Kim, Stefan Sabev, Olusegun Samson, transmissions 11, Michael Levine, Cory Kujawski, Rainer Wilmers, zynix, Kalila, Luke @flexchar, Ajan Kanaga, Mandus, vamX, Ai Maven, Mano Prime, Matthew Berman, subjectnull, Vitor Caleffi, Clay Pascal, biorpg, alfie_i, 阿明, Jeffrey Morgan, ya boyyy, Raymond Fosdick, knownsqashed, Olakabola, Leonard Tan, ReadyPlayerEmma, Enrico Ros, Dave, Talal Aujan, Illia Dulskyi, Sean Connelly, senxiiz, Artur Olbinski, Elle, Raven Klaugh, Fen Risland, Deep Realms, Imad Khwaja, Fred von Graf, Will Dee, usrbinkat, SuperWojo, Alexandros Triantafyllidis, Swaroop Kallakuri, Dan Guido, John Detwiler, Pedro Madruga, Iucharbius, Viktor Bowallius, Asp the Wyvern, Edmond Seymore, Trenton Dambrowitz, Space Cruiser, Spiking Neurons AB, Pyrater, LangChain4j, Tony Hughes, Kacper Wikieł, Rishabh Srivastava, David Ziegler, Luke Pendergrass, Andrey, Gabriel Puliatti, Lone Striker, Sebastain Graf, Pierre Kircher, Randy H, NimbleBox.ai, Vadim, danny, Deo Leter
230
 
231
 
232
  Thank you to all my generous patrons and donaters!
 
39
  - Model creator: [Voicelab](https://huggingface.co/Voicelab)
40
  - Original model: [Trurl 2 13B](https://huggingface.co/Voicelab/trurl-2-13b)
41
 
42
+ <!-- description start -->
43
  ## Description
44
 
45
  This repo contains GPTQ model files for [Voicelab's Trurl 2 13B](https://huggingface.co/Voicelab/trurl-2-13b).
46
 
47
  Multiple GPTQ parameter permutations are provided; see Provided Files below for details of the options provided, their parameters, and the software used to create them.
48
 
49
+ <!-- description end -->
50
+ <!-- repositories-available start -->
51
  ## Repositories available
52
 
53
  * [GPTQ models for GPU inference, with multiple quantisation parameter options.](https://huggingface.co/TheBloke/Trurl-2-13B-GPTQ)
54
+ * [2, 3, 4, 5, 6 and 8-bit GGUF models for CPU+GPU inference](https://huggingface.co/TheBloke/Trurl-2-13B-GGUF)
55
+ * [2, 3, 4, 5, 6 and 8-bit GGML models for CPU+GPU inference (deprecated)](https://huggingface.co/TheBloke/Trurl-2-13B-GGML)
56
  * [Voicelab's original unquantised fp16 model in pytorch format, for GPU inference and for further conversions](https://huggingface.co/Voicelab/trurl-2-13b)
57
+ <!-- repositories-available end -->
58
 
59
+ <!-- prompt-template start -->
60
  ## Prompt template: Llama-2-Chat
61
 
62
  ```
 
64
  You 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.
65
  <</SYS>>
66
  {prompt}[/INST]
67
+
68
  ```
69
 
70
+ <!-- prompt-template end -->
71
+
72
+ <!-- README_GPTQ.md-provided-files start -->
73
  ## Provided files and GPTQ parameters
74
 
75
  Multiple quantisation parameters are provided, to allow you to choose the best one for your hardware and requirements.
76
 
77
  Each separate quant is in a different branch. See below for instructions on fetching from different branches.
78
 
79
+ All recent GPTQ files are made with AutoGPTQ, and all files in non-main branches are made with AutoGPTQ. Files in the `main` branch which were uploaded before August 2023 were made with GPTQ-for-LLaMa.
80
 
81
  <details>
82
  <summary>Explanation of GPTQ parameters</summary>
83
 
84
  - Bits: The bit size of the quantised model.
85
  - GS: GPTQ group size. Higher numbers use less VRAM, but have lower quantisation accuracy. "None" is the lowest possible value.
86
+ - Act Order: True or False. Also known as `desc_act`. True results in better quantisation accuracy. Some GPTQ clients have had issues with models that use Act Order plus Group Size, but this is generally resolved now.
87
  - Damp %: A GPTQ parameter that affects how samples are processed for quantisation. 0.01 is default, but 0.1 results in slightly better accuracy.
88
  - GPTQ dataset: The dataset used for quantisation. Using a dataset more appropriate to the model's training can improve quantisation accuracy. Note that the GPTQ dataset is not the same as the dataset used to train the model - please refer to the original model repo for details of the training dataset(s).
89
  - Sequence Length: The length of the dataset sequences used for quantisation. Ideally this is the same as the model sequence length. For some very long sequence models (16+K), a lower sequence length may have to be used. Note that a lower sequence length does not limit the sequence length of the quantised model. It only impacts the quantisation accuracy on longer inference sequences.
 
100
  | [gptq-8bit--1g-actorder_True](https://huggingface.co/TheBloke/Trurl-2-13B-GPTQ/tree/gptq-8bit--1g-actorder_True) | 8 | None | Yes | 0.1 | [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. |
101
  | [gptq-8bit-128g-actorder_True](https://huggingface.co/TheBloke/Trurl-2-13B-GPTQ/tree/gptq-8bit-128g-actorder_True) | 8 | 128 | Yes | 0.1 | [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. |
102
 
103
+ <!-- README_GPTQ.md-provided-files end -->
104
+
105
+ <!-- README_GPTQ.md-download-from-branches start -->
106
  ## How to download from branches
107
 
108
  - In text-generation-webui, you can add `:branch` to the end of the download name, eg `TheBloke/Trurl-2-13B-GPTQ:gptq-4bit-32g-actorder_True`
 
111
  git clone --single-branch --branch gptq-4bit-32g-actorder_True https://huggingface.co/TheBloke/Trurl-2-13B-GPTQ
112
  ```
113
  - In Python Transformers code, the branch is the `revision` parameter; see below.
114
+ <!-- README_GPTQ.md-download-from-branches end -->
115
+ <!-- README_GPTQ.md-text-generation-webui start -->
116
  ## How to easily download and use this model in [text-generation-webui](https://github.com/oobabooga/text-generation-webui).
117
 
118
  Please make sure you're using the latest version of [text-generation-webui](https://github.com/oobabooga/text-generation-webui).
119
 
120
+ It is strongly recommended to use the text-generation-webui one-click-installers unless you're sure you know how to make a manual install.
121
 
122
  1. Click the **Model tab**.
123
  2. Under **Download custom model or LoRA**, enter `TheBloke/Trurl-2-13B-GPTQ`.
124
  - To download from a specific branch, enter for example `TheBloke/Trurl-2-13B-GPTQ:gptq-4bit-32g-actorder_True`
125
  - see Provided Files above for the list of branches for each option.
126
  3. Click **Download**.
127
+ 4. The model will start downloading. Once it's finished it will say "Done".
128
  5. In the top left, click the refresh icon next to **Model**.
129
  6. In the **Model** dropdown, choose the model you just downloaded: `Trurl-2-13B-GPTQ`
130
  7. The model will automatically load, and is now ready for use!
131
  8. If you want any custom settings, set them and then click **Save settings for this model** followed by **Reload the Model** in the top right.
132
+ * Note that you do not need to and should not set manual GPTQ parameters any more. These are set automatically from the file `quantize_config.json`.
133
  9. Once you're ready, click the **Text Generation tab** and enter a prompt to get started!
134
+ <!-- README_GPTQ.md-text-generation-webui end -->
135
 
136
+ <!-- README_GPTQ.md-use-from-python start -->
137
  ## How to use this GPTQ model from Python code
138
 
139
+ ### Install the necessary packages
140
 
141
+ Requires: Transformers 4.32.0 or later, Optimum 1.12.0 or later, and AutoGPTQ 0.4.2 or later.
 
 
142
 
143
+ ```shell
144
+ pip3 install transformers>=4.32.0 optimum>=1.12.0
145
+ pip3 install auto-gptq --extra-index-url https://huggingface.github.io/autogptq-index/whl/cu118/ # Use cu117 if on CUDA 11.7
146
  ```
147
+
148
+ If you have problems installing AutoGPTQ using the pre-built wheels, install it from source instead:
149
+
150
+ ```shell
151
  pip3 uninstall -y auto-gptq
152
  git clone https://github.com/PanQiWei/AutoGPTQ
153
  cd AutoGPTQ
154
  pip3 install .
155
  ```
156
 
157
+ ### For CodeLlama models only: you must use Transformers 4.33.0 or later.
158
+
159
+ If 4.33.0 is not yet released when you read this, you will need to install Transformers from source:
160
+ ```shell
161
+ pip3 uninstall -y transformers
162
+ pip3 install git+https://github.com/huggingface/transformers.git
163
+ ```
164
+
165
+ ### You can then use the following code
166
 
167
  ```python
168
+ from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline
 
169
 
170
  model_name_or_path = "TheBloke/Trurl-2-13B-GPTQ"
171
+ # To use a different branch, change revision
172
+ # For example: revision="gptq-4bit-32g-actorder_True"
173
+ model = AutoModelForCausalLM.from_pretrained(model_name_or_path,
174
+ torch_dtype=torch.float16,
175
+ device_map="auto",
176
+ revision="main")
177
 
178
  tokenizer = AutoTokenizer.from_pretrained(model_name_or_path, use_fast=True)
179
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
180
  prompt = "Tell me about AI"
181
  prompt_template=f'''[INST] <<SYS>>
182
  You 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.
183
  <</SYS>>
184
  {prompt}[/INST]
185
+
186
  '''
187
 
188
  print("\n\n*** Generate:")
 
193
 
194
  # Inference can also be done using transformers' pipeline
195
 
 
 
 
196
  print("*** Pipeline:")
197
  pipe = pipeline(
198
  "text-generation",
 
206
 
207
  print(pipe(prompt_template)[0]['generated_text'])
208
  ```
209
+ <!-- README_GPTQ.md-use-from-python end -->
210
 
211
+ <!-- README_GPTQ.md-compatibility start -->
212
  ## Compatibility
213
 
214
+ The files provided are tested to work with AutoGPTQ, both via Transformers and using AutoGPTQ directly. They should also work with [Occ4m's GPTQ-for-LLaMa fork](https://github.com/0cc4m/KoboldAI).
215
+
216
+ [ExLlama](https://github.com/turboderp/exllama) is compatible with Llama models in 4-bit. Please see the Provided Files table above for per-file compatibility.
217
 
218
+ [Huggingface Text Generation Inference (TGI)](https://github.com/huggingface/text-generation-inference) is compatible with all GPTQ models.
219
+ <!-- README_GPTQ.md-compatibility end -->
220
 
221
  <!-- footer start -->
222
  <!-- 200823 -->
 
241
 
242
  **Special thanks to**: Aemon Algiz.
243
 
244
+ **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
245
 
246
 
247
  Thank you to all my generous patrons and donaters!