Lewdiculous
commited on
Commit
•
d460a1a
1
Parent(s):
ff0ad29
att --quantkv mention
Browse files
README.md
CHANGED
@@ -1,8 +1,11 @@
|
|
1 |
---
|
|
|
|
|
|
|
2 |
license: cc-by-nc-4.0
|
|
|
3 |
language:
|
4 |
- en
|
5 |
-
inference: false
|
6 |
tags:
|
7 |
- roleplay
|
8 |
- llama3
|
@@ -23,13 +26,16 @@ I recommend checking their page for feedback and support.
|
|
23 |
> If you noticed any issues let me know in the discussions.
|
24 |
|
25 |
> [!NOTE]
|
26 |
-
> **General usage:** <br>
|
27 |
-
> For **8GB VRAM** GPUs, I recommend the **Q4_K_M-imat** (4.89 BPW) quant for up to 12288 context sizes. <br>
|
28 |
-
>
|
29 |
> **Presets:** <br>
|
30 |
> Some compatible SillyTavern presets can be found [**here (Virt's Roleplay Presets - v1.9)**](https://huggingface.co/Virt-io/SillyTavern-Presets). <br>
|
|
|
31 |
> Lower temperatures are recommended by the authors, so make sure to experiment. <br>
|
32 |
-
>
|
|
|
|
|
|
|
|
|
|
|
33 |
|
34 |
<details>
|
35 |
<summary>⇲ Click here to expand/hide information – General chart with relative quant parformances.</summary>
|
|
|
1 |
---
|
2 |
+
base_model: NeverSleep/Lumimaid-v0.2-8B
|
3 |
+
quantized_by: Lewdiculous
|
4 |
+
library_name: transformers
|
5 |
license: cc-by-nc-4.0
|
6 |
+
inference: false
|
7 |
language:
|
8 |
- en
|
|
|
9 |
tags:
|
10 |
- roleplay
|
11 |
- llama3
|
|
|
26 |
> If you noticed any issues let me know in the discussions.
|
27 |
|
28 |
> [!NOTE]
|
|
|
|
|
|
|
29 |
> **Presets:** <br>
|
30 |
> Some compatible SillyTavern presets can be found [**here (Virt's Roleplay Presets - v1.9)**](https://huggingface.co/Virt-io/SillyTavern-Presets). <br>
|
31 |
+
> Check [**discussions such as this one**](https://huggingface.co/Virt-io/SillyTavern-Presets/discussions/5#664d6fb87c563d4d95151baa) and [**this one**](https://www.reddit.com/r/SillyTavernAI/comments/1dff2tl/my_personal_llama3_stheno_presets/) for other presets and samplers recommendations. <br>
|
32 |
> Lower temperatures are recommended by the authors, so make sure to experiment. <br>
|
33 |
+
>
|
34 |
+
> **General usage with KoboldCpp:** <br>
|
35 |
+
> For **8GB VRAM** GPUs, I recommend the **Q4_K_M-imat** (4.89 BPW) quant for up to 12288 context sizes without the use of `--quantkv`. <br>
|
36 |
+
> Using `--quantkv 1` (≈Q8) or even `--quantkv 2` (≈Q4) can get you to 32K context sizes with the caveat of not being compatible with Context Shifting, only relevant if you can manage to fill up that much context. <br>
|
37 |
+
> [**Read more about it in the release here**](https://github.com/LostRuins/koboldcpp/releases/tag/v1.67).
|
38 |
+
|
39 |
|
40 |
<details>
|
41 |
<summary>⇲ Click here to expand/hide information – General chart with relative quant parformances.</summary>
|