Commit
•
329077d
1
Parent(s):
c294b06
Update README.md
Browse files
README.md
CHANGED
|
@@ -15,7 +15,9 @@ Experimental model, using a limarp qlora trained at 10k ctx length (greater than
|
|
| 15 |
|
| 16 |
Note that all modules were trained, including 'gate'. There are some reports that perhaps training the 'gate' module isn't fully functional at the moment. In cursory testing this appears to obey the limarp alpaca prompt format correctly.
|
| 17 |
|
| 18 |
-
Not extensively tested for quality, YMMV.
|
|
|
|
|
|
|
| 19 |
|
| 20 |
## Usage:
|
| 21 |
The intended prompt format is the Alpaca instruction format of LimaRP v3:
|
|
|
|
| 15 |
|
| 16 |
Note that all modules were trained, including 'gate'. There are some reports that perhaps training the 'gate' module isn't fully functional at the moment. In cursory testing this appears to obey the limarp alpaca prompt format correctly.
|
| 17 |
|
| 18 |
+
Not extensively tested for quality, YMMV. Would try with temp ~1.5-2 and min-p of ~0.03-0.05 since mixtral does appear to be highly confident on its responses.
|
| 19 |
+
|
| 20 |
+
[EXL2 Quants](https://huggingface.co/Doctor-Shotgun/Mixtral-8x7B-Instruct-limarp-v0.1-exl2)
|
| 21 |
|
| 22 |
## Usage:
|
| 23 |
The intended prompt format is the Alpaca instruction format of LimaRP v3:
|