Merge pull request #88 from hysunflower/update_scripts

Update scripts

python/paddle_fl/paddle_fl/examples/ctr_demo/fl_trainer.py

@@ -50,7 +50,7 @@ epoch_id = 0
 while not trainer.stop():
     if epoch_id > 15:
         break
-    print("{} epoch {} start train".format(time.strftime('%Y-%m-%d %H:%M:%S',time.localtime(time.time())), epoch_id))
+    print("{} Epoch {} start train".format(time.strftime('%Y-%m-%d %H:%M:%S',time.localtime(time.time())), epoch_id))
     train_step = 0
     for data in reader():
         trainer.run(feed=data, fetch=[])

python/paddle_fl/paddle_fl/examples/ctr_demo/run.sh

+#!/bin/bash
 unset http_proxy
 unset https_proxy
 ps -ef | grep -E fl_ | grep -v grep | awk '{print $2}' | xargs kill -9
 
-log_dir=${1:-$(pwd)}
+log_dir=${1:-"logs"}
 mkdir -p ${log_dir}
 
 python fl_master.py > ${log_dir}/master.log 2>&1 &

python/paddle_fl/paddle_fl/examples/dpsgd_demo/fl_trainer.py

@@ -20,6 +20,7 @@ import paddle
 import paddle.fluid as fluid
 import logging
 import math
+import time
 
 logging.basicConfig(
     filename="test.log",
@@ -72,9 +73,9 @@ epoch_id = 0
 step = 0
 while not trainer.stop():
     epoch_id += 1
-    if epoch_id > 40:
+    if epoch_id > 10:
         break
-    print("epoch %d start train" % (epoch_id))
+    print("{} Epoch {} start train".format(time.strftime('%Y-%m-%d %H:%M:%S',time.localtime(time.time())), epoch_id))
     for step_id, data in enumerate(train_reader()):
         acc = trainer.run(feeder.feed(data), fetch=["accuracy_0.tmp_0"])
         step += 1

python/paddle_fl/paddle_fl/examples/dpsgd_demo/run.sh

+#!/bin/bash
 unset http_proxy
 unset https_proxy
-python fl_master.py
+ps -ef | grep -E fl_ | grep -v grep | awk '{print $2}' | xargs kill -9
+
+log_dir=${1:-"logs"}
+mkdir -p ${log_dir}
+
+python fl_master.py > ${log_dir}/master.log 2>&1 &
 sleep 2
-python -u fl_scheduler.py >scheduler.log &
+python -u fl_scheduler.py > ${log_dir}/scheduler.log 2>&1 &
+sleep 5
+python -u fl_server.py > ${log_dir}/server0.log 2>&1 &
 sleep 2
-python -u fl_server.py >server0.log &
-sleep 2
-python -u fl_trainer.py 0 >trainer0.log &
-sleep 2
-python -u fl_trainer.py 1 >trainer1.log &
-sleep 2
-python -u fl_trainer.py 2 >trainer2.log &
-sleep 2
-python -u fl_trainer.py 3 >trainer3.log &
+for ((i=0;i<4;i++))
+do
+    python -u fl_trainer.py $i > ${log_dir}/trainer$i.log 2>&1 &
+    sleep 2
+done

python/paddle_fl/paddle_fl/examples/femnist_demo/fl_trainer.py

@@ -21,6 +21,7 @@ import paddle
 import paddle.fluid as fluid
 import logging
 import math
+import time
 
 logging.basicConfig(
     filename="test.log",
@@ -60,7 +61,7 @@ def train_test(train_test_program, train_test_feed, train_test_reader):
 
 epoch_id = 0
 step = 0
-epoch = 3000
+epoch = 10
 count_by_step = False
 if count_by_step:
     output_folder = "model_node%d" % trainer_id
@@ -72,7 +73,7 @@ while not trainer.stop():
     epoch_id += 1
     if epoch_id > epoch:
         break
-    print("epoch %d start train" % (epoch_id))
+    print("{} Epoch {} start train".format(time.strftime('%Y-%m-%d %H:%M:%S',time.localtime(time.time())), epoch_id))
     #train_data,test_data= data_generater(trainer_id,inner_step=trainer._step,batch_size=64,count_by_step=count_by_step)
     train_reader = paddle.batch(
         paddle.reader.shuffle(
@@ -97,7 +98,6 @@ while not trainer.stop():
             acc = trainer.run(feeder.feed(data), fetch=["accuracy_0.tmp_0"])
             step += 1
             count += 1
-            print(count)
             if count % trainer._step == 0:
                 break
     # print("acc:%.3f" % (acc[0]))

python/paddle_fl/paddle_fl/examples/femnist_demo/run.sh

+#!/bin/bash
 unset http_proxy
 unset https_proxy
-#killall python
-python fl_master.py
+ps -ef | grep -E fl_ | grep -v grep | awk '{print $2}' | xargs kill -9
+
+log_dir=${1:-"logs"}
+mkdir -p ${log_dir}
+
+python fl_master.py > ${log_dir}/master.log 2>&1 &
 sleep 2
-python -u fl_scheduler.py >scheduler.log &
-sleep 2
-python -u fl_server.py >server0.log &
+python -u fl_scheduler.py > ${log_dir}/scheduler.log 2>&1 &
+sleep 5
+python -u fl_server.py > ${log_dir}/server0.log 2>&1 &
 sleep 2
 for ((i=0;i<4;i++))
 do
-    python -u fl_trainer.py $i >trainer$i.log &
+    python -u fl_trainer.py $i > ${log_dir}/trainer$i.log 2>&1 &
     sleep 2
 done

python/paddle_fl/paddle_fl/examples/generate_job_from_program/README.md

@@ -17,6 +17,10 @@ pip install paddle_fl
 
 #### How to save a program
 
+```sh
+python program_saver.py
+```
+
 In program_saver.py, you can defind a model. And save the program in to 'load_file'
 
 ```python

python/paddle_fl/paddle_fl/examples/generate_job_from_program/fl_trainer.py

@@ -20,6 +20,7 @@ import paddle
 import paddle.fluid as fluid
 import logging
 import math
+import time
 
 logging.basicConfig(
     filename="test.log",
@@ -67,9 +68,9 @@ epoch_id = 0
 step = 0
 while not trainer.stop():
     epoch_id += 1
-    if epoch_id > 40:
+    if epoch_id > 10:
         break
-    print("epoch %d start train" % (epoch_id))
+    print("{} Epoch {} start train".format(time.strftime('%Y-%m-%d %H:%M:%S',time.localtime(time.time())), epoch_id))
     for step_id, data in enumerate(train_reader()):
         acc = trainer.run(feeder.feed(data), fetch=["accuracy_0.tmp_0"])
         step += 1

python/paddle_fl/paddle_fl/examples/generate_job_from_program/run.sh

+#!/bin/bash
 unset http_proxy
 unset https_proxy
-python program_saver.py
+ps -ef | grep -E fl_ | grep -v grep | awk '{print $2}' | xargs kill -9
+if [ ! -d load_file ]; then
+    python program_saver.py
+fi
 
-python fl_master.py
-sleep 2
-python -u fl_scheduler.py >scheduler.log &
-sleep 2
-python -u fl_server.py >server0.log &
-sleep 2
-python -u fl_trainer.py 0 >trainer0.log &
+log_dir=${1:-"logs"}
+mkdir -p ${log_dir}
+
+python fl_master.py > ${log_dir}/master.log 2>&1 &
 sleep 2
-python -u fl_trainer.py 1 > trainer1.log &
+python -u fl_scheduler.py > ${log_dir}/scheduler.log 2>&1 &
+sleep 5
+python -u fl_server.py > ${log_dir}/server0.log 2>&1 &
 sleep 2
+for ((i=0;i<2;i++))
+do
+    python -u fl_trainer.py $i > ${log_dir}/trainer$i.log 2>&1 &
+    sleep 2
+done

python/paddle_fl/paddle_fl/examples/gru4rec_demo/fl_trainer.py

@@ -20,6 +20,8 @@ import numpy as np
 import sys
 import os
 import logging
+import time
+
 logging.basicConfig(
     filename="test.log",
     filemode="w",
@@ -43,10 +45,9 @@ r = Gru4rec_Reader()
 train_reader = r.reader(train_file_dir, place, batch_size=125)
 
 output_folder = "model_node4"
-step_i = 0
+epoch_i = 0
 while not trainer.stop():
-    step_i += 1
-    print("batch %d start train" % (step_i))
+    epoch_i += 1
     train_step = 0
     for data in train_reader():
         #print(np.array(data['src_wordseq']))
@@ -56,10 +57,10 @@ while not trainer.stop():
             break
         avg_ppl = np.exp(ret_avg_cost[0])
         newest_ppl = np.mean(avg_ppl)
-        print("ppl:%.3f" % (newest_ppl))
-    save_dir = (output_folder + "/epoch_%d") % step_i
+        print("{} Epoch {} start train, train_step {}, ppl {}".format (time.strftime('%Y-%m-%d %H:%M:%S',time.localtime(time.time())), epoch_i, train_step, newest_ppl))
+    save_dir = (output_folder + "/epoch_%d") % epoch_i
     if trainer_id == 0:
         print("start save")
         trainer.save_inference_program(save_dir)
-    if step_i >= 40:
+    if epoch_i >= 5:
         break

python/paddle_fl/paddle_fl/examples/gru4rec_demo/run.sh

+#!/bin/bash
 unset http_proxy
 unset https_proxy
-python fl_master.py
+ps -ef | grep -E fl_ | grep -v grep | awk '{print $2}' | xargs kill -9
+
+if [ ! -d mid_data ];then
+    sh download.sh
+fi
+
+log_dir=${1:-"logs"}
+mkdir -p ${log_dir}
+
+python fl_master.py > ${log_dir}/master.log 2>&1 &
 sleep 2
-python -u fl_scheduler.py >scheduler.log &
-python -u fl_server.py >server0.log &
+python -u fl_scheduler.py > ${log_dir}/scheduler.log 2>&1 &
+sleep 5
+python -u fl_server.py > ${log_dir}/server0.log 2>&1 &
 sleep 2
-python -u fl_trainer.py 0 >trainer0.log &
-sleep 2
-python -u fl_trainer.py 1 >trainer1.log &
-sleep 2
-python -u fl_trainer.py 2 >trainer2.log &
-sleep 2
-python -u fl_trainer.py 3 >trainer3.log &
+for ((i=0;i<4;i++))
+do
+    python -u fl_trainer.py $i > ${log_dir}/trainer$i.log 2>&1 &
+    sleep 2
+done

python/paddle_fl/paddle_fl/examples/secagg_demo/fl_trainer.py

@@ -84,21 +84,16 @@ def train_test(train_test_program, train_test_feed, train_test_reader):
 
 
 # for test
-
 while not trainer.stop():
     epoch_id += 1
-    print("epoch %d start train" % (epoch_id))
-
     for data in train_reader():
         step_i += 1
         trainer.step_id = step_i
         accuracy, = trainer.run(feed=feeder.feed(data),
                                 fetch=["accuracy_0.tmp_0"])
         if step_i % 100 == 0:
-            print("Epoch: {0}, step: {1}, accuracy: {2}".format(
+            print("{} Epoch {} start train, step: {}, accuracy: {}".format(time.strftime('%Y-%m-%d %H:%M:%S',time.localtime(time.time())), 
                 epoch_id, step_i, accuracy[0]))
-
-    print(step_i)
     avg_loss_val, acc_val = train_test(
         train_test_program=test_program,
         train_test_reader=test_reader,
@@ -106,7 +101,7 @@ while not trainer.stop():
     print("Test with Epoch %d, avg_cost: %s, acc: %s" %
           (epoch_id, avg_loss_val, acc_val))
 
-    if epoch_id > 40:
+    if epoch_id > 5:
         break
     if epoch_id % 5 == 0:
         trainer.save_inference_program(output_folder)

python/paddle_fl/paddle_fl/examples/secagg_demo/run.sh

+#!/bin/bash
 unset http_proxy
 unset https_proxy
+ps -ef | grep -E fl_ | grep -v grep | awk '{print $2}' | xargs kill -9
 
-if [ ! -d log ];then
-    mkdir log
-fi
+log_dir=${1:-"logs"}
+mkdir -p ${log_dir}
 
-python fl_master.py
+python fl_master.py > ${log_dir}/master.log 2>&1 &
 sleep 2
-python -u fl_server.py >log/server0.log &
+python -u fl_scheduler.py > ${log_dir}/scheduler.log 2>&1 &
+sleep 5
+python -u fl_server.py > ${log_dir}/server0.log 2>&1 &
 sleep 2
-python -u fl_scheduler.py > log/scheduler.log &
-sleep 2
-python -u fl_trainer.py 0 >log/trainer0.log &
-sleep 2
-python -u fl_trainer.py 1 >log/trainer1.log &
+for ((i=0;i<2;i++))
+do
+    python -u fl_trainer.py $i > ${log_dir}/trainer$i.log 2>&1 &
+    sleep 2
+done