Update README.md
Browse files
README.md
CHANGED
|
@@ -52,5 +52,13 @@ We would like to take this opportunity to thank
|
|
| 52 |
- BAD: γγͺγγ―ββγγ§γγΎγ
|
| 53 |
- GOOD: γγͺγγ―ββγγγΎγ
|
| 54 |
|
| 55 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 56 |
If possible, we recommend inferring with llamacpp rather than Transformers.
|
|
|
|
| 52 |
- BAD: γγͺγγ―ββγγ§γγΎγ
|
| 53 |
- GOOD: γγͺγγ―ββγγγΎγ
|
| 54 |
|
| 55 |
+
## Merge recipe
|
| 56 |
+
|
| 57 |
+
|
| 58 |
+
- VT0.1 = Ninjav1 + Original Lora
|
| 59 |
+
- VT0.2 = Ninjav1 128k + Original Lora
|
| 60 |
+
- VT0.2on0.1 = VT0.1 + VT0.2
|
| 61 |
+
|
| 62 |
+
- VT1 = all VT Series + Lora + Ninja 128k and Normal
|
| 63 |
+
## Other points to keep in mind
|
| 64 |
If possible, we recommend inferring with llamacpp rather than Transformers.
|