add ce for neural_machine_translation/transformer (#2538)

* add ce for neural_machine_translation/transformer

* add ce for neural_machine_translation/transformer

* add ce for neural_machine_translation/transformer

PaddleNLP/neural_machine_translation/transformer/.run_ce.sh

+#!/bin/bash
+
+sed -i '$a\dropout_seed = 1000' ../../models/neural_machine_translation/transformer/desc.py
+
+DATA_PATH=./dataset/wmt16
+
+train(){
+    python -u train.py \
+        --src_vocab_fpath $DATA_PATH/en_10000.dict \
+        --trg_vocab_fpath $DATA_PATH/de_10000.dict \
+        --special_token '<s>' '<e>' '<unk>' \
+        --train_file_pattern $DATA_PATH/wmt16/train \
+        --val_file_pattern $DATA_PATH/wmt16/val \
+        --use_token_batch True \
+        --batch_size 2048 \
+        --sort_type pool \
+        --pool_size 10000 \
+        --enable_ce True \
+        --fetch_steps 1 \
+        weight_sharing False \
+        pass_num 20
+}
+
+cudaid=${transformer:=0} # use 0-th card as default
+export CUDA_VISIBLE_DEVICES=$cudaid
+
+train | python _ce.py
+
+cudaid=${transformer_m:=0,1,2,3} # use 0,1,2,3 card as default
+export CUDA_VISIBLE_DEVICES=$cudaid
+
+train | python _ce.py

PaddleNLP/neural_machine_translation/transformer/_ce.py

+####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', 0.002, 0, actived=True)
+test_cost_card1_kpi = CostKpi('test_cost_card1', 0.008, 0, actived=True)
+train_duration_card1_kpi = DurationKpi(
+    'train_duration_card1', 0.006, 0, actived=True)
+train_cost_card4_kpi = CostKpi('train_cost_card4', 0.001, 0, actived=True)
+test_cost_card4_kpi = CostKpi('test_cost_card4', 0.001, 0, actived=True)
+train_duration_card4_kpi = DurationKpi(
+    'train_duration_card4', 0.02, 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)