diff --git a/bin/process_CDR/process_CDR.sh b/bin/process_CDR/process_CDR.sh
index 6378626..d4186c5 100755
--- a/bin/process_CDR/process_CDR.sh
+++ b/bin/process_CDR/process_CDR.sh
@@ -1,6 +1,6 @@
 #!/usr/bin/env bash
 
-word_piece_vocab=data/cdr/word_piece_vocabs/just_train_2500/word_pieces.txt
+word_piece_vocab=${CDR_IE_ROOT}/data/cdr/word_piece_vocabs/just_train_2500/word_pieces.txt
 input_dir=${CDR_IE_ROOT}/data/cdr
 processed_dir=${input_dir}/processed/just_train_2500
 proto_dir=${processed_dir}/protos
@@ -21,7 +21,7 @@ echo "Processing Test data"
 python ${CDR_IE_ROOT}/src/processing/utils/process_CDR_data.py --input_file ${input_dir}/CDR_TestSet.PubTator.txt.gz --output_dir ${processed_dir} --output_file_suffix CDR_test.txt --max_seq ${max_len} --full_abstract True --word_piece_codes ${word_piece_vocab}
 
 for f in CDR_dev CDR_train CDR_test; do
-python ${CDR_IE_ROOT}/src/processing/utils/filter_hypernyms.py -p ${processed_dir}/positive_0_${f}.txt -n ${processed_dir}/negative_0_${f}.txt -m data/2017MeshTree.txt -o ${processed_dir}/negative_0_${f}_filtered.txt
+python ${CDR_IE_ROOT}/src/processing/utils/filter_hypernyms.py -p ${processed_dir}/positive_0_${f}.txt -n ${processed_dir}/negative_0_${f}.txt -m ${CDR_IE_ROOT}/data/2017MeshTree.txt -o ${processed_dir}/negative_0_${f}_filtered.txt
 done
 
 
@@ -33,4 +33,4 @@ mv ${proto_dir}/negative_0_CDR_dev_filtered.txt.proto ${proto_dir}/negative_0_CD
 mv ${proto_dir}/positive_0_CDR_dev.txt.proto ${proto_dir}/positive_0_CDR_train_dev.txt.proto
 
 # convert ner data to tf protos
-python ${CDR_IE_ROOT}/src/processing/ner_to_tfrecords.py --in_files ${processed_dir}/ner_\* --out_dir ${proto_dir} --load_vocab ${proto_dir} --num_threads 5
\ No newline at end of file
+python ${CDR_IE_ROOT}/src/processing/ner_to_tfrecords.py --in_files ${processed_dir}/ner_\* --out_dir ${proto_dir} --load_vocab ${proto_dir} --num_threads 5