ssmits commited on
Commit
884eade
1 Parent(s): b4963d1

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +2 -2
README.md CHANGED
@@ -13,9 +13,9 @@ language:
13
  ---
14
  ## Why prune?
15
 
16
- Falcon-11B is still undertrained, as can be seen by this graph:
17
  ![image/png](https://cdn-uploads.huggingface.co/production/uploads/660c0a02cf274b3ab77dd6b7/QeaL9bOrPskustzFpjMUP.png)
18
- This is why the choice is made by prune 50% of the layers.
19
  Note that \~1B of continued pre-training (\~1M rows of 1k tokens) is still required to restore the perplexity of this model in the desired language.
20
  I'm planning on doing that for certain languages, depending on how much compute will be available.
21
 
 
13
  ---
14
  ## Why prune?
15
 
16
+ Even though [Falcon-11B](https://huggingface.co/tiiuae/falcon-11B) is trained on 5T tokens, it is still undertrained, as can be seen by this graph:
17
  ![image/png](https://cdn-uploads.huggingface.co/production/uploads/660c0a02cf274b3ab77dd6b7/QeaL9bOrPskustzFpjMUP.png)
18
+ This is why the choice is made to prune 50% of the layers.
19
  Note that \~1B of continued pre-training (\~1M rows of 1k tokens) is still required to restore the perplexity of this model in the desired language.
20
  I'm planning on doing that for certain languages, depending on how much compute will be available.
21