athirdpath
commited on
Commit
•
5046913
1
Parent(s):
1b97a04
Update README.md
Browse files
README.md
CHANGED
@@ -6,6 +6,8 @@ Trained from athirdpath/Llama-3-15b-Instruct on 0.6 epochs of athirdpath/Merge_G
|
|
6 |
|
7 |
I cut it off pretty early because I didn't want the model taking on the tone/logic of the largely GPT 3.5/4 data, just to heal the grafts
|
8 |
|
|
|
|
|
9 |
![image/png](https://cdn-uploads.huggingface.co/production/uploads/633a809fa4a8f33508dce32c/5n_J2cQxxtJ-Kj5nlVYGz.png)
|
10 |
|
11 |
<p align="center"><font size="5"> <b>Date Writing Example @ q6_k</b> </font></p>
|
|
|
6 |
|
7 |
I cut it off pretty early because I didn't want the model taking on the tone/logic of the largely GPT 3.5/4 data, just to heal the grafts
|
8 |
|
9 |
+
It amazes me how, like the human brain, an LLM can recover from massive trauma by just being given an opportunity to reroute. Neuroplasticity is magic.
|
10 |
+
|
11 |
![image/png](https://cdn-uploads.huggingface.co/production/uploads/633a809fa4a8f33508dce32c/5n_J2cQxxtJ-Kj5nlVYGz.png)
|
12 |
|
13 |
<p align="center"><font size="5"> <b>Date Writing Example @ q6_k</b> </font></p>
|