def main(_):
params=cf.get_params()
params['mfile']="-11"
model_path=params["wd"]+"/cp/"+params['mfile']
data=du.load_pose(params)
data_train=(data[0],data[1])
data_test=(data[2],data[3])
with tf.Graph().as_default(), tf.Session() as session:
initializer = tf.random_uniform_initializer(-params["init_scale"],params["init_scale"])
# m = mp.get_model(is_training=True,params=params)
mtest = mp.get_model(is_training=False,params=params)
saver = tf.train.Saver()
# tf.initialize_all_variables().run()
saver.restore(sess=session,save_path=model_path)
test_err = run_epoch(session, mtest,tf.no_op(),params, data_test,is_training=False)
print("Test Err: %.5f" % test_err)
loss3d =u.get_loss(params,y,pred)
batch_loss3d.append(loss3d)
x=[]
y=[]
(sid,H,C,x,y) = async_b.get() # get the return value from your function.
if(minibatch_index==n_train_batches-1):
pred,H,C= model.predictions(x,is_train,H,C)
loss3d =u.get_loss(params,y,pred)
batch_loss3d.append(loss3d)
batch_loss3d=np.nanmean(batch_loss3d)
if(batch_loss3d<best_loss):
best_loss=batch_loss3d
ext=str(epoch_counter)+"_"+str(batch_loss3d)+"_best.p"
u.write_params(model.params,params,ext)
else:
ext=str(val_counter%2)+".p"
u.write_params(model.params,params,ext)
val_counter+=1#0.08
s ='VAL--> epoch %i | error %f, %f'%(val_counter,batch_loss3d,n_test_batches)
u.log_write(s,params)
params= config.get_params()
parser = argparse.ArgumentParser(description='Training the module')
parser.add_argument('-m','--model',help='Model: lstm, lstm2, erd current('+params["model"]+')',default=params["model"])
args = vars(parser.parse_args())
params["model"]=args["model"]
params=config.update_params(params)
train_rnn(params)
import math
import numpy as np
import tensorflow as tf
from nets import inception
import urllib2
from datetime import datetime
from helper import config
from helper import utils as ut
from helper import dt_utils as dut
from helper.preprocessing import human36m_preprocessing
from PIL import Image
import numpy as np
params = config.get_params()
slim = tf.contrib.slim
num_examples = 100
subset = 'validation'
is_training = False
def eval(params):
# batch_size = params['batch_size']
# num_examples = len(params['test_files'][0])
with tf.Graph().as_default() as g:
url = '/home/coskun/PycharmProjects/data/pose/mv_val/img/S9/Discussion 1.54138969/frame_00010.png'
filename_queue = tf.train.string_input_producer(
[url]) # list of files to read
reader = tf.WholeFileReader()
save_path = base_cp_path + model_name
saved_path = saver.save(sess, save_path)
else:
if e % 3.0 == 0:
saved_path = saver.save(sess, save_path)
if saved_path != "":
s = 'MODEL_Saved --> epoch %i | error %f path %s' % (e, total_loss, saved_path)
ut.log_write(s, params)
rnn_keep_prob_lst=[0.8]
rnn_input_prob_lst=[1.0]
seq_lst=[50]
reset_state=[5,100,20]
normalise_data_lst=[3]
# To get current status of params#
params = config.get_params() # To get current status of params#
###############################
params["mfile"]='/mnt/Data1/hc/tt/cp/lstm_nostate1/cp/' # adding more values to params#
rnn_keep_prob=0.8
input_keep_prob=1.0
params['rnn_keep_prob']=rnn_keep_prob
params['input_keep_prob']=input_keep_prob
seq=50 #what does this value signify?
res=5 #what does this value signify?
with tf.Graph().as_default():
print "seq: ============== %s ============" % seq
print "reset_state: ============== %s ============" % res
print "rnn_keep_prob: ============== %s ============" % rnn_keep_prob
params['normalise_data'] = 4 # adding more values to params, what does this value signify? #
params['reset_state']=res # adding more values to params, what does this value signify? #
params['seq_length']=seq # adding more values to params, what does this value signify? #
