Pretergeek commited on
Commit
1bc1594
1 Parent(s): 51d4169

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +6 -2
README.md CHANGED
@@ -14,12 +14,16 @@ license: apache-2.0
14
 
15
  # OpenChat-3.5-0106_10.7B_48Layers-Interleaved
16
 
17
- This is a merge of pre-trained language models created using [mergekit](https://github.com/cg123/mergekit).
18
 
19
  ## Merge Details
20
  ### Merge Method
21
 
22
- This model was merged using the passthrough merge method.
 
 
 
 
23
 
24
  ### Models Merged
25
 
 
14
 
15
  # OpenChat-3.5-0106_10.7B_48Layers-Interleaved
16
 
17
+ This is NOT your usual frankenmerge created using [mergekit](https://github.com/cg123/mergekit).
18
 
19
  ## Merge Details
20
  ### Merge Method
21
 
22
+ This model was merged using the passthrough merge method, but employing the Block Expansion method described in the paper [LLaMA Pro: Progressive LLaMA with Block Expansion](https://arxiv.org/abs/2401.02415).
23
+
24
+ The authors of the paper added new layers interleaved in between the original layers of the model, setting the parameters of the o_proj and down_proj layers to zero. This effectively adds layers that will just output their input (as if they were "transparent") allowing the model to remain functional even without further training. These new layers can then be targeted during training or fine-tuning without risking catastrophic forgetting, if you follow the author's training method to freeze the original layers and only train the new layers.
25
+
26
+ This model has not yet received additional training, so it should perform close to the original model.
27
 
28
  ### Models Merged
29