|
--- |
|
library_name: transformers |
|
license: apache-2.0 |
|
language: |
|
- en |
|
tags: |
|
- gguf |
|
- quantized |
|
- roleplay |
|
- imatrix |
|
- mistral |
|
- merge |
|
- nsfw |
|
inference: false |
|
base_model: |
|
- ResplendentAI/Datura_7B |
|
- ChaoticNeutrals/Eris_Floramix_DPO_7B |
|
--- |
|
|
|
> [!TIP] |
|
> **Support:** <br> |
|
> My upload speeds have been cooked and unstable lately. <br> |
|
> Realistically I'd need to move to get a better provider. <br> |
|
> If you **want** and you are able to... <br> |
|
> [**You can support my various endeavors here (Ko-fi).**](https://ko-fi.com/Lewdiculous) <br> |
|
> I apologize for disrupting your experience. |
|
|
|
|
|
This repository hosts GGUF-Imatrix quantizations for [Endevor/InfinityRP-v1-7B](https://huggingface.co/Endevor/InfinityRP-v1-7B). |
|
|
|
The supported --contextsize is 8192. |
|
|
|
**What does "Imatrix" mean?** |
|
|
|
It stands for **Importance Matrix**, a technique used to improve the quality of quantized models. |
|
The **Imatrix** is calculated based on calibration data, and it helps determine the importance of different model activations during the quantization process. |
|
The idea is to preserve the most important information during quantization, which can help reduce the loss of model performance, especially when the calibration data is diverse. |
|
[[1]](https://github.com/ggerganov/llama.cpp/discussions/5006) [[2]](https://github.com/ggerganov/llama.cpp/discussions/5263#discussioncomment-8395384) |
|
|
|
**Steps:** |
|
``` |
|
Base⇢ GGUF(F16)⇢ Imatrix-Data(F16)⇢ GGUF(Imatrix-Quants) |
|
``` |
|
**Quants:** |
|
```python |
|
quantization_options = [ |
|
"Q4_K_M", "IQ4_XS", "Q5_K_M", "Q5_K_S", "Q6_K", |
|
"Q8_0", "IQ3_M", "IQ3_S", "IQ3_XXS" |
|
] |
|
``` |
|
|
|
If you want anything that's not here or another model, feel free to request. |
|
|
|
**This is experimental.** |
|
|
|
For imatrix data generation, kalomaze's `groups_merged.txt` with added roleplay chats was used, you can find it [here](https://huggingface.co/Lewdiculous/Datura_7B-GGUF-Imatrix/blob/main/imatrix-with-rp-format-data.txt). |
|
|
|
**Original model information:** |
|
|
|
 |
|
|
|
This is an experimental model I currently use. It's far from great as I'm still working on it, but I leave it here for people to try if interested in this format. |
|
This model was basically made to stop some upsetting hallucinations, so {{char}} mostly and occasionally will wait {{user}} response instead of responding itself or deciding for {{user}}, also, my primary idea was to create a cozy model that thinks.* |
|
|
|
Inspired by [lemonilia/Limamono-Mistral-7B-v0.50](https://huggingface.co/lemonilia/Limamono-Mistral-7B-v0.50) |
|
### Style details: |
|
- Quotes are used for character dialogs. |
|
- `"Hey, Anon... What do you think about my style?"` |
|
- Asterisks can be used for narration, but it's optional, it's recommended to use default novel format. |
|
- `*Her cheeks blush slightly, she tries to hide.*` |
|
- Character thoughts are wrapped with ` marks. **This may often spontaneously occur.** |
|
- `My heart skips a beat hearing him call me pretty!` |
|
|
|
*If you want thoughts to appear more often, just add something like this to your system prompt: ```"{{char}} internal thoughts are wrapped with ` marks."```* |
|
|
|
- Accepted response lengths: ***tiny, short, medium, long, huge*** |
|
- |
|
For example: ### Response: (length = medium) |
|
|
|
Note: Apparently ***humongous***, ***extreme*** and ***unlimited*** may not work at moment. Not fully tested. |
|
|
|
### Prompt format: |
|
Extended Alpaca, as always. |
|
|
|
``"You are now in roleplay chat mode. Engage in an endless chat with {{user}}. Always wait {{user}} turn, next actions and responses."`` |
|
|
|
## Example: |
|
|
|
 |
|
|
