Update README.md
Browse files
README.md
CHANGED
|
@@ -37,10 +37,15 @@ Violet_Twilight's models were trained with the Mistral Instruct template, theref
|
|
| 37 |
*** NOTE *** <br/>
|
| 38 |
There have been reports of the quantized model misbehaving with the mistral prompt, if you are seeing issues it may be worth trying ChatML Context and Instruct templates.
|
| 39 |
If you are using GGUF I strongly advise using ChatML, for some reason that quantization performs better using ChatML.
|
|
|
|
| 40 |
### Current Top Sampler Settings
|
| 41 |
-
[
|
|
|
|
| 42 |
[Crimson_Dawn-Magnum-Style](https://files.catbox.moe/lc59dn.json)
|
| 43 |
|
|
|
|
|
|
|
|
|
|
| 44 |
## Merging
|
| 45 |
The following config was used to merge Azure Dusk and Crimson Dawn
|
| 46 |
```yaml
|
|
|
|
| 37 |
*** NOTE *** <br/>
|
| 38 |
There have been reports of the quantized model misbehaving with the mistral prompt, if you are seeing issues it may be worth trying ChatML Context and Instruct templates.
|
| 39 |
If you are using GGUF I strongly advise using ChatML, for some reason that quantization performs better using ChatML.
|
| 40 |
+
|
| 41 |
### Current Top Sampler Settings
|
| 42 |
+
[Violet_Twilight-Nitral-Special](https://files.catbox.moe/ot54u3.json)- Considered the best settings! <br/>
|
| 43 |
+
[Crimson_Dawn-Nitral-Special](https://files.catbox.moe/8xjxht.json) <br/>
|
| 44 |
[Crimson_Dawn-Magnum-Style](https://files.catbox.moe/lc59dn.json)
|
| 45 |
|
| 46 |
+
### Tokenizer
|
| 47 |
+
If you are using SillyTavern, please set the tokenizer to API (WebUI/ koboldcpp)
|
| 48 |
+
|
| 49 |
## Merging
|
| 50 |
The following config was used to merge Azure Dusk and Crimson Dawn
|
| 51 |
```yaml
|