Update README.md
Browse files
README.md
CHANGED
@@ -6,6 +6,8 @@ datasets:
|
|
6 |
- Ba2han/databricks-dolly_rated
|
7 |
- Open-Orca/OpenOrca
|
8 |
---
|
|
|
|
|
9 |
The training dataset consists of 2k (longest) examples from no_robots, reddit_instruct, dolly, OpenOrca plus two other personal datasets.
|
10 |
|
11 |
Please use with ChatML and the default system message or enter your own. It was trained with various system messages, the one in the config being the default one.
|
@@ -19,4 +21,8 @@ The model is:
|
|
19 |
|
20 |
- Not great with short text both in input and generation.
|
21 |
|
22 |
-
The aim is to see how the **"Long is More for Alignment"** paper holds. This is basically a combination of LIMA + LMA. There should be no benchmark contamination as far as I am aware of. Around 70% of the data is from the mentioned datasets. I am happy with how it turned out.
|
|
|
|
|
|
|
|
|
|
6 |
- Ba2han/databricks-dolly_rated
|
7 |
- Open-Orca/OpenOrca
|
8 |
---
|
9 |
+
|
10 |
+

|
11 |
The training dataset consists of 2k (longest) examples from no_robots, reddit_instruct, dolly, OpenOrca plus two other personal datasets.
|
12 |
|
13 |
Please use with ChatML and the default system message or enter your own. It was trained with various system messages, the one in the config being the default one.
|
|
|
21 |
|
22 |
- Not great with short text both in input and generation.
|
23 |
|
24 |
+
The aim is to see how the **"Long is More for Alignment"** paper holds. This is basically a combination of LIMA + LMA. There should be no benchmark contamination as far as I am aware of. Around 70% of the data is from the mentioned datasets. I am happy with how it turned out.
|
25 |
+
|
26 |
+
|
27 |
+

|
28 |
+
|