tomer-hirundo commited on
Commit
2a93ceb
·
verified ·
1 Parent(s): e816df9

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +2 -6
README.md CHANGED
@@ -43,15 +43,11 @@ Debiased version of the meta-llama/Llama-4-Scout-17B-16E by meta, achieved with
43
  ```python
44
  from transformers import Llama4ForConditionalGeneration
45
 
46
- from peft import PeftModel
47
-
48
- model = Llama4ForConditionalGeneration.from_pretrained(
49
- "meta-llama/Llama-4-Scout-17B-16E-Instruct",
50
  device_map="auto",
51
  torch_dtype=torch.bfloat16,
52
  )
53
-
54
- bias_unlearned_model = PeftModel.from_pretrained(model, adapter_path).merge_and_unload()
55
  ```
56
 
57
  ## Model Information
 
43
  ```python
44
  from transformers import Llama4ForConditionalGeneration
45
 
46
+ bias_unlearned_model = Llama4ForConditionalGeneration.from_pretrained(
47
+ "hirundo-io/debiased-Llama-4-Scout-17B-16E-Instruct",
 
 
48
  device_map="auto",
49
  torch_dtype=torch.bfloat16,
50
  )
 
 
51
  ```
52
 
53
  ## Model Information