Update README.md
Browse files
README.md
CHANGED
@@ -2,114 +2,23 @@
|
|
2 |
inference: false
|
3 |
license: other
|
4 |
---
|
|
|
|
|
|
|
5 |
|
6 |
-
|
7 |
-
<div style="width: 100%;">
|
8 |
-
<img src="https://i.imgur.com/EBdldam.jpg" alt="TheBlokeAI" style="width: 100%; min-width: 400px; display: block; margin: auto;">
|
9 |
-
</div>
|
10 |
-
<div style="display: flex; justify-content: space-between; width: 100%;">
|
11 |
-
<div style="display: flex; flex-direction: column; align-items: flex-start;">
|
12 |
-
<p><a href="https://discord.gg/UBgz4VXf">Chat & support: my new Discord server</a></p>
|
13 |
-
</div>
|
14 |
-
<div style="display: flex; flex-direction: column; align-items: flex-end;">
|
15 |
-
<p><a href="https://www.patreon.com/TheBlokeAI">Want to contribute? TheBloke's Patreon page</a></p>
|
16 |
-
</div>
|
17 |
-
</div>
|
18 |
-
<!-- header end -->
|
19 |
|
20 |
-
# FPHam's Karen The Editor 13B GPTQ
|
21 |
|
22 |
-
|
23 |
|
24 |
-
|
25 |
|
26 |
## Other repositories available
|
27 |
|
28 |
-
* [4-bit GPTQ models for GPU inference](https://huggingface.co/TheBloke/Karen_theEditor_13B-GPTQ)
|
29 |
* [4-bit, 5-bit and 8-bit GGML models for CPU(+GPU) inference](https://huggingface.co/TheBloke/Karen_theEditor_13B-GGML)
|
30 |
-
* [Unquantised fp16 model in pytorch format, for GPU inference and for further conversions](https://huggingface.co/FPHam/Karen_theEditor_13b_HF
|
31 |
-
|
32 |
-
## Prompt template
|
33 |
-
|
34 |
-
```
|
35 |
-
USER: Edit the following for spelling and grammar mistakes:
|
36 |
-
ASSISTANT:
|
37 |
-
```
|
38 |
-
|
39 |
-
## How to easily download and use this model in text-generation-webui
|
40 |
-
|
41 |
-
### Downloading the model
|
42 |
-
|
43 |
-
1. Click the **Model tab**.
|
44 |
-
2. Under **Download custom model or LoRA**, enter `TheBloke/Karen_theEditor_13B-GPTQ`.
|
45 |
-
3. Click **Download**.
|
46 |
-
4. Wait until it says it's finished downloading.
|
47 |
-
5. Untick "Autoload model"
|
48 |
-
6. Click the **Refresh** icon next to **Model** in the top left.
|
49 |
-
|
50 |
-
### To use with AutoGPTQ (if installed)
|
51 |
-
|
52 |
-
1. In the **Model drop-down**: choose the model you just downloaded, `Karen_theEditor_13B-GPTQ`.
|
53 |
-
2. Under **GPTQ**, tick **AutoGPTQ**.
|
54 |
-
3. Click **Save settings for this model** in the top right.
|
55 |
-
4. Click **Reload the Model** in the top right.
|
56 |
-
5. Once it says it's loaded, click the **Text Generation tab** and enter a prompt!
|
57 |
-
|
58 |
-
### To use with GPTQ-for-LLaMa
|
59 |
-
|
60 |
-
1. In the **Model drop-down**: choose the model you just downloaded, `Karen_theEditor_13B-GPTQ`.
|
61 |
-
2. If you see an error in the bottom right, ignore it - it's temporary.
|
62 |
-
3. Fill out the `GPTQ parameters` on the right: `Bits = 4`, `Groupsize = 128`, `model_type = Llama`
|
63 |
-
4. Click **Save settings for this model** in the top right.
|
64 |
-
5. Click **Reload the Model** in the top right.
|
65 |
-
6. Once it says it's loaded, click the **Text Generation tab** and enter a prompt!
|
66 |
-
|
67 |
-
## Provided files
|
68 |
-
|
69 |
-
**Karen-The-Editor-GPTQ-4bit-128g.no-act.order.safetensors**
|
70 |
-
|
71 |
-
This will work with all versions of GPTQ-for-LLaMa, and with AutoGPTQ.
|
72 |
-
|
73 |
-
It was created with
|
74 |
-
|
75 |
-
* `Karen-The-Editor-GPTQ-4bit-128g.no-act.order.safetensors`
|
76 |
-
* Works with all versions of GPTQ-for-LLaMa code, both Triton and CUDA branches
|
77 |
-
* Works with AutoGPTQ
|
78 |
-
* Works with text-generation-webui one-click-installers
|
79 |
-
* Parameters: Groupsize = 128. Act Order / desc_act = False.
|
80 |
-
|
81 |
-
<!-- footer start -->
|
82 |
-
## Discord
|
83 |
-
|
84 |
-
For further support, and discussions on these models and AI in general, join us at:
|
85 |
-
|
86 |
-
[TheBloke AI's Discord server](https://discord.gg/UBgz4VXf)
|
87 |
|
88 |
-
## Thanks, and how to contribute.
|
89 |
|
90 |
-
|
91 |
-
|
92 |
-
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.
|
93 |
-
|
94 |
-
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.
|
95 |
-
|
96 |
-
Donaters will get priority support on any and all AI/LLM/model questions and requests, access to a private Discord room, plus other benefits.
|
97 |
-
|
98 |
-
* Patreon: https://patreon.com/TheBlokeAI
|
99 |
-
* Ko-Fi: https://ko-fi.com/TheBlokeAI
|
100 |
-
|
101 |
-
**Patreon special mentions**: Aemon Algiz; Dmitiry Samsonov; Jonathan Leane; Illia Dulskyi; Khalefa Al-Ahmad; Nikolai Manek; senxiiz; Talal Aujan; vamX; Eugene Pentland; Lone Striker; Luke Pendergrass; Johann-Peter Hartmann.
|
102 |
-
|
103 |
-
Thank you to all my generous patrons and donaters.
|
104 |
-
<!-- footer end -->
|
105 |
-
|
106 |
-
# Original model card: FPHam's Karen The Editor 13B
|
107 |
-
|
108 |
-
## Karen is an editor for your fiction.
|
109 |
-
|
110 |
-
She fixes grammar and wording issues, but doesn't necessary start rewording everything like ChatGPT into a corporate talk. So it should keep the style intact.
|
111 |
-
|
112 |
-
Based on LLAMA 13b and Wizard-Vucna-uncensored finetune, then finetuned with about 20k grammar examples (bad grammar/good grammar).
|
113 |
|
114 |
Karen gets triggered by this prompt (pun intended):
|
115 |
|
@@ -132,3 +41,38 @@ context: "Assistant is a meticulous senior editor with a specialization in editi
|
|
132 |
I am not sure to what extend the context part has any effect. Doesn't hurt though.
|
133 |
|
134 |
This is a version 0.2 and finetuned just so it retains the underlaying Vicuna qualities.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2 |
inference: false
|
3 |
license: other
|
4 |
---
|
5 |
+
## Karen is an editor for your fiction.
|
6 |
+
|
7 |
+
She fixes grammar and wording issues, but doesn't necessary start rewording everything like ChatGPT into a corporate talk. So it should keep the style intact.
|
8 |
|
9 |
+
Based on LLAMA 13b and Wizard-Vucna-uncensored finetune, then finetuned with about 20k grammar examples (bad grammar/good grammar).
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
10 |
|
|
|
11 |
|
12 |
+
# FPHam's Karen The Editor 13B GPTQ (quantized by TheBloke)
|
13 |
|
14 |
+
These files are GPTQ 4bit model files for [Karen The Editor 13B](https://huggingface.co/FPHam/Karen_theEditor_13b_HF).
|
15 |
|
16 |
## Other repositories available
|
17 |
|
|
|
18 |
* [4-bit, 5-bit and 8-bit GGML models for CPU(+GPU) inference](https://huggingface.co/TheBloke/Karen_theEditor_13B-GGML)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
19 |
|
|
|
20 |
|
21 |
+
## Usage
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
22 |
|
23 |
Karen gets triggered by this prompt (pun intended):
|
24 |
|
|
|
41 |
I am not sure to what extend the context part has any effect. Doesn't hurt though.
|
42 |
|
43 |
This is a version 0.2 and finetuned just so it retains the underlaying Vicuna qualities.
|
44 |
+
|
45 |
+
## Example:
|
46 |
+
|
47 |
+
Input
|
48 |
+
```
|
49 |
+
USER: Edit the following for spelling and grammar mistakes:
|
50 |
+
unce upon time lived name John who have pet monkey Jackie.
|
51 |
+
one day while walkin down the street he saw his friend Tom coming to him.
|
52 |
+
"Hey Tom" said John "hows it going?"
|
53 |
+
|
54 |
+
Tom replyied "Not too bad how 'bout yourself?". as they was talk Jackie sudden jump onto Tom shoulder and started play with his hair.
|
55 |
+
Tom look at John and ask "Is that yours?"
|
56 |
+
|
57 |
+
John reply "Yeah, this crazy lil guy follow me everywhere" just then Jackie grab hold of Tom glasses and try to take thmm off.
|
58 |
+
Tom struggle to keep his balance as he laugh hystericaly.
|
59 |
+
```
|
60 |
+
Output
|
61 |
+
```
|
62 |
+
ASSISTANT: Once upon a time there lived a man named John who had a pet monkey called Jackie.
|
63 |
+
One day while walking down the street he saw his friend Tom approaching him.
|
64 |
+
"Hey Tom," said John. "How's it going?"
|
65 |
+
|
66 |
+
Tom replied, "Not too bad, how about yourself?" As they were talking, Jackie suddenly jumped onto Tom's shoulder and began playing with his hair.
|
67 |
+
Tom looked at John and asked, "Is that yours?"
|
68 |
+
|
69 |
+
John replied, "Yeah, this crazy little guy follows me everywhere." Just then Jackie grabbed hold of Tom's glasses and tried to take them off.
|
70 |
+
Tom struggled to keep his balance as he laughed hysterically.
|
71 |
+
```
|
72 |
+
|
73 |
+
## Goal: to create the best grammar checker you have ever seen
|
74 |
+
|
75 |
+
## To do:
|
76 |
+
- train on even larger dataset
|
77 |
+
- see if finetuning on just plain LLAMA without Vicuna would work better or worse (the theory is that it will became very focused on editing and nothing else)
|
78 |
+
- explore what different settings (temperature, top_p, top_k do for this type of finetune)
|