Update README.md
Browse files
README.md
CHANGED
@@ -9,10 +9,14 @@ tags:
|
|
9 |
---
|
10 |
# Llama-3-6B-Instruct-pruned
|
11 |
*Experimental*
|
|
|
12 |
Using [PruneMe](https://github.com/arcee-ai/PruneMe) to find minimal average distance. Thank you for awesome toolkit @arcee-ai !
|
13 |
<img src="./distance.png" alt="distance" width="390"/>
|
14 |
*It shows pruning the 22-30 layer is the best option, but I'm worried about drasitical change between 22 to 23.*
|
15 |
|
|
|
|
|
|
|
16 |
This is a merge of pre-trained language models created using [mergekit](https://github.com/cg123/mergekit).
|
17 |
|
18 |
## Merge Details
|
|
|
9 |
---
|
10 |
# Llama-3-6B-Instruct-pruned
|
11 |
*Experimental*
|
12 |
+
|
13 |
Using [PruneMe](https://github.com/arcee-ai/PruneMe) to find minimal average distance. Thank you for awesome toolkit @arcee-ai !
|
14 |
<img src="./distance.png" alt="distance" width="390"/>
|
15 |
*It shows pruning the 22-30 layer is the best option, but I'm worried about drasitical change between 22 to 23.*
|
16 |
|
17 |
+
### Disclaimer
|
18 |
+
I haven't done any post-training (called 'healing' process as the [paper](https://arxiv.org/abs/2403.17887) suggests), will do it later but no guarantee at all.
|
19 |
+
|
20 |
This is a merge of pre-trained language models created using [mergekit](https://github.com/cg123/mergekit).
|
21 |
|
22 |
## Merge Details
|