Update README.md
Browse files
README.md
CHANGED
|
@@ -21,6 +21,9 @@ Qwen 2.5 presumably uses enough synthetic data that she's not as inclined to sha
|
|
| 21 |
(Additionally I did have a higher proportion of nontoxic data in the DPO, and trained based on a non-Unleashed model.
|
| 22 |
Then undid a lot of my work with the SLERP, technically. ;) )
|
| 23 |
|
|
|
|
|
|
|
|
|
|
| 24 |
But that doesn't matter because she got 78.0391 in EQ-Bench and 100% parseable. :D Clearly worthy in her own right.
|
| 25 |
|
| 26 |
In addition to basing the model on a variation of Lumen that I attempted to rebase and heal from the damage of original instruct,
|
|
|
|
| 21 |
(Additionally I did have a higher proportion of nontoxic data in the DPO, and trained based on a non-Unleashed model.
|
| 22 |
Then undid a lot of my work with the SLERP, technically. ;) )
|
| 23 |
|
| 24 |
+
Edit: Tagging with not for all audiences because of the dataset, I guess, but she's actually still aligned?
|
| 25 |
+
Not sure what niche she falls into exactly.
|
| 26 |
+
|
| 27 |
But that doesn't matter because she got 78.0391 in EQ-Bench and 100% parseable. :D Clearly worthy in her own right.
|
| 28 |
|
| 29 |
In addition to basing the model on a variation of Lumen that I attempted to rebase and heal from the damage of original instruct,
|