Update README.md

Author: shivank21

README.md

@@ -23,33 +23,33 @@ We have gone beyond traditional fine-tuning methods by exploring the impact of s
 One of the following commands may be run to train the chosen pre-trained model followed by pruning.
 
 **OMP**
-
+```
 python -u main_imp.py --data ./data --dataset ${data} --arch ${arch} --prune_type rewind_lt --rewind_epoch 8 --save_dir ${save_dir} --rate ${rate} --pruning_times 2 --num_workers 8
-
+```
 **IMP**
-
+```
 python -u main_imp.py --data ./data --dataset ${data} --arch ${arch} --prune_type rewind_lt --rewind_epoch 8 --save_dir ${save_dir} --rate 0.2 --pruning_times ${pruning_times} --num_workers 8
-
+```
 ## 4. Unlearning
 
 Following pruning one of these may be run to carry out unlearning.
 
 **FT**
-
+```
 python -u main_forget.py --save_dir ${save_dir} --mask ${mask_path} --unlearn FT --num_indexes_to_replace 4500 --unlearn_lr 0.01 --unlearn_epochs 10
-
+```
 **GA**
-
+```
 python -u main_forget.py --save_dir ${save_dir} --mask ${mask_path} --unlearn GA --num_indexes_to_replace 4500 --unlearn_lr 0.0001 --unlearn_epochs 5
-
+```
 **FF**
-
+```
 python -u main_forget.py --save_dir ${save_dir} --mask ${mask_path} --unlearn fisher_new --num_indexes_to_replace 4500 --alpha ${alpha}
-
+```
 **IU**
-
+```
 python -u main_forget.py --save_dir ${save_dir} --mask ${mask_path} --unlearn wfisher --num_indexes_to_replace 4500 --alpha ${alpha}
-
+```
 ## 5. Conclusion
 
  We discovered that the introduction of sparsity can effectively enhance the accuracy of the unlearning algorithm. Moreover, the application of fine-tuning, combined with LoRA, not only substantiates this concept but also expedites the entire process.