Update README.md

README.md

@@ -52,26 +52,27 @@ output # 'CCN(CC)CCN=C=S.Cc1cnc2c(c1)CCCC2N'
 ### Training Procedure 
 
 <!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
-We used the Open Reaction Database (ORD) dataset for model training.
+We used the [USPTO_50k dataset](https://yzhang.hpc.nyu.edu/T5Chem/index.html) for model finetuning.
 The command used for training is the following. For more information, please refer to the paper and GitHub repository.
 
 ```python
-python train_without_duplicates.py \
-    --model='t5' \
-    --epochs=80 \
-    --lr=2e-4 \
+cd task_retrosynthesis
+python finetune.py \
+    --output_dir='t5' \
+    --epochs=20 \
+    --lr=2e-5 \
     --batch_size=32 \
-    --input_max_len=100 \
+    --input_max_len=150 \
     --target_max_len=150 \
     --weight_decay=0.01 \
     --evaluation_strategy='epoch' \
     --save_strategy='epoch' \
     --logging_strategy='epoch' \
-    --train_data_path='/home/acf15718oa/ReactionT5_neword/data/all_ord_reaction_uniq_with_attr20240506_v3_train.csv' \
-    --valid_data_path='/home/acf15718oa/ReactionT5_neword/data/all_ord_reaction_uniq_with_attr20240506_v3_valid.csv' \
-    --test_data_path='/home/acf15718oa/ReactionT5_neword/data/all_ord_reaction_uniq_with_attr20240506_v3_test.csv' \
-    --USPTO_test_data_path='/home/acf15718oa/ReactionT5_neword/data/USPTO_50k/test.csv' \
-    --pretrained_model_name_or_path='sagawa/CompoundT5'
+    --save_total_limit=10 \
+    --train_data_path='../data/USPTO_50k/train.csv' \
+    --valid_data_path='../data/USPTO_50k/val.csv' \
+    --disable_tqdm \
+    --model_name_or_path='sagawa/ReactionT5v2-retrosynthesis'
 ```
 
 ### Results