Commit
·
7297758
1
Parent(s):
7a246af
Update README.md
Browse files
README.md
CHANGED
|
@@ -4,5 +4,7 @@ license: other
|
|
| 4 |
license_name: microsoft-research-license
|
| 5 |
---
|
| 6 |
This model is a fine-tuned version of microsoft/Orca-2-13b on a subset of the Vezora/Mini_Orca_Uncencored_Alpaca dataset, with some particularly spicy prompts added to reduce the risk of rejections.
|
| 7 |
-
|
|
|
|
|
|
|
| 8 |
I'll test it tomorrow when my GGUF quants are done.
|
|
|
|
| 4 |
license_name: microsoft-research-license
|
| 5 |
---
|
| 6 |
This model is a fine-tuned version of microsoft/Orca-2-13b on a subset of the Vezora/Mini_Orca_Uncencored_Alpaca dataset, with some particularly spicy prompts added to reduce the risk of rejections.
|
| 7 |
+
|
| 8 |
+
Only the q_proj and k_proj modules were targeted and a low rank (8) was used, in hopes of containing the adjustments to the prompt format and alignment. This is promising on paper, with the training's per-step loss averaging <0.9 for the last third of the run.
|
| 9 |
+
|
| 10 |
I'll test it tomorrow when my GGUF quants are done.
|