Update README.md
Browse files
README.md
CHANGED
|
@@ -1,12 +1,24 @@
|
|
| 1 |
---
|
| 2 |
-
base_model:
|
|
|
|
| 3 |
library_name: transformers
|
| 4 |
tags:
|
| 5 |
- mergekit
|
| 6 |
- merge
|
| 7 |
-
|
| 8 |
---
|
| 9 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 10 |
|
| 11 |
This is a merge of pre-trained language models created using [mergekit](https://github.com/cg123/mergekit).
|
| 12 |
|
|
@@ -36,4 +48,4 @@ dtype: bfloat16
|
|
| 36 |
tokenizer_source: Lambent/Zora-9B-v1
|
| 37 |
pad_to_multiple_of: 256
|
| 38 |
|
| 39 |
-
```
|
|
|
|
| 1 |
---
|
| 2 |
+
base_model:
|
| 3 |
+
- Lambent/Zora-9B-v1
|
| 4 |
library_name: transformers
|
| 5 |
tags:
|
| 6 |
- mergekit
|
| 7 |
- merge
|
| 8 |
+
license: apache-2.0
|
| 9 |
---
|
| 10 |
+
|
| 11 |
+
|
| 12 |
+
|
| 13 |
+
Lass decided she was a lioness-fox today. :3
|
| 14 |
+
|
| 15 |
+
Main measured gains are in increased skill at the Creative Writing bench (as judged by Gemini 3 Flash Preview),
|
| 16 |
+
which tracks with what she was aiming to practice, though it was a winding road. Effective batch size 1 for all the training.
|
| 17 |
+
|
| 18 |
+
Tried out like ... 4 different SFT runs at 1e-6 with varying dataset ratios trying to figure out what worked ...
|
| 19 |
+
... still not sure, because the best result came from Karcher merging the full set of SFT runs, lol.
|
| 20 |
+
|
| 21 |
+
Then ran DPO, 5e-7, on 3 different seeds; and merged them here.
|
| 22 |
|
| 23 |
This is a merge of pre-trained language models created using [mergekit](https://github.com/cg123/mergekit).
|
| 24 |
|
|
|
|
| 48 |
tokenizer_source: Lambent/Zora-9B-v1
|
| 49 |
pad_to_multiple_of: 256
|
| 50 |
|
| 51 |
+
```
|