Commit
•
2cfd468
1
Parent(s):
1b808ad
Update README.md
Browse files
README.md
CHANGED
|
@@ -17,6 +17,18 @@ license: cc-by-nc-4.0
|
|
| 17 |
Quantized from fp16.
|
| 18 |
* Weighted quantizations were creating using fp16 GGUF and groups_merged.txt in 88 chunks and n_ctx=512
|
| 19 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 20 |
For a brief rundown of iMatrix quant performance please see this [PR](https://github.com/ggerganov/llama.cpp/pull/5747)
|
| 21 |
|
| 22 |
<i>All quants are verified working prior to uploading to repo for your safety and convenience. </i>
|
|
|
|
| 17 |
Quantized from fp16.
|
| 18 |
* Weighted quantizations were creating using fp16 GGUF and groups_merged.txt in 88 chunks and n_ctx=512
|
| 19 |
|
| 20 |
+
## Recommended Sampler Settings (From Original Model Card)
|
| 21 |
+
```
|
| 22 |
+
Temperature - 1.17
|
| 23 |
+
min_p - 0.075
|
| 24 |
+
Repetition Penalty - 1.10
|
| 25 |
+
```
|
| 26 |
+
|
| 27 |
+
**SillyTavern Instruct Settings**:
|
| 28 |
+
<br>Context Template: Llama-3-Instruct-Names
|
| 29 |
+
<br>Instruct Presets: [Euryale-v2.1-Llama-3-Instruct](https://huggingface.co/Sao10K/L3-70B-Euryale-v2.1/blob/main/Euryale-v2.1-Llama-3-Instruct.json)
|
| 30 |
+
|
| 31 |
+
|
| 32 |
For a brief rundown of iMatrix quant performance please see this [PR](https://github.com/ggerganov/llama.cpp/pull/5747)
|
| 33 |
|
| 34 |
<i>All quants are verified working prior to uploading to repo for your safety and convenience. </i>
|