diff --git a/fluid/machine_reading_comprehension/.run_ce.sh b/fluid/machine_reading_comprehension/.run_ce.sh
deleted file mode 100644
index f4134481591732287d315f5e1f6288a48a75668d..0000000000000000000000000000000000000000
--- a/fluid/machine_reading_comprehension/.run_ce.sh
+++ /dev/null
@@ -1,40 +0,0 @@
-#!/bin/bash
-
-DATA_PATH=./data
-if [ ! -e $DATA_PATH/demo ] ; then
-    mkdir -p $DATA_PATH/demo
-    tar -zxf $DATA_PATH/demo.tgz -C $DATA_PATH/demo
-fi
-
-train(){
-python  -u run.py   \
-	--trainset 'data/demo/search.train.json' \
-	--devset 'data/demo/search.dev.json' \
-	--testset 'data/demo/search.test.json' \
-	--vocab_dir 'data/demo/' \
-	--use_gpu true \
-	--save_dir ./models \
-	--pass_num 1 \
-	--learning_rate 0.001 \
-	--batch_size 32 \
-	--embed_size 300 \
-	--hidden_size 150 \
-	--max_p_num 5 \
-	--max_p_len 500 \
-	--max_q_len 60 \
-	--max_a_len 200 \
-	--drop_rate 0.2 \
-	--log_interval 1 \
-        --enable_ce \
-        --train 
-}
-
-cudaid=${transformer:=2} # use 0-th card as default
-export CUDA_VISIBLE_DEVICES=$cudaid
-
-train | python _ce.py
-
-cudaid=${transformer_m:=2,4} # use 0,1,2,3 card as default
-export CUDA_VISIBLE_DEVICES=$cudaid
-
-train | python _ce.py
diff --git a/fluid/machine_reading_comprehension/_ce.py b/fluid/machine_reading_comprehension/_ce.py
deleted file mode 100644
index 24243846ee1f321cb449573f1a3729346edd1035..0000000000000000000000000000000000000000
--- a/fluid/machine_reading_comprehension/_ce.py
+++ /dev/null
@@ -1,68 +0,0 @@
-####this file is only used for continuous evaluation test!
-
-import os
-import sys
-#sys.path.insert(0, os.environ['ceroot'])
-from kpi import CostKpi, DurationKpi, AccKpi
-
-#### NOTE kpi.py should shared in models in some way!!!!
-
-train_cost_card1_kpi = CostKpi('train_cost_card1', 0.02, 0, actived=True)
-test_cost_card1_kpi = CostKpi('test_cost_card1', 0.005, 0, actived=True)
-train_duration_card1_kpi = DurationKpi(
-    'train_duration_card1', 0.06, 0, actived=True)
-train_cost_card4_kpi = CostKpi('train_cost_card2', 0.01, 0, actived=True)
-test_cost_card4_kpi = CostKpi('test_cost_card2', 0.005, 0, actived=True)
-train_duration_card4_kpi = DurationKpi(
-    'train_duration_card2', 0.06, 0, actived=True)
-
-tracking_kpis = [
-    train_cost_card1_kpi,
-    test_cost_card1_kpi,
-    train_duration_card1_kpi,
-    train_cost_card4_kpi,
-    test_cost_card4_kpi,
-    train_duration_card4_kpi,
-]
-
-
-def parse_log(log):
-    '''
-    This method should be implemented by model developers.
-    The suggestion:
-    each line in the log should be key, value, for example:
-    "
-    train_cost\t1.0
-    test_cost\t1.0
-    train_cost\t1.0
-    train_cost\t1.0
-    train_acc\t1.2
-    "
-    '''
-    for line in log.split('\n'):
-        fs = line.strip().split('\t')
-        print(fs)
-        if len(fs) == 3 and fs[0] == 'kpis':
-            print("-----%s" % fs)
-            kpi_name = fs[1]
-            kpi_value = float(fs[2])
-            yield kpi_name, kpi_value
-
-
-def log_to_ce(log):
-    kpi_tracker = {}
-    for kpi in tracking_kpis:
-        kpi_tracker[kpi.name] = kpi
-
-    for (kpi_name, kpi_value) in parse_log(log):
-        print(kpi_name, kpi_value)
-        kpi_tracker[kpi_name].add_record(kpi_value)
-        kpi_tracker[kpi_name].persist()
-
-
-if __name__ == '__main__':
-    log = sys.stdin.read()
-    print("*****")
-    print(log)
-    print("****")
-    log_to_ce(log)