File size: 5,325 Bytes
eb90a8a | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 | """
Experimental program for complete evaluation of similarity analyzer
using sequence or Tokens technique by 2-way siamese type DNN
Program analyzes all pairs of source code samples
Program operated in in multi-GPU mode
Uses:
- experimental Siameze CNN
- custom metric
"""
import sys
import os
import argparse
import pickle
import tensorflow as tf
main_dir = os.path.dirname(
os.path.dirname(os.path.realpath(__file__)))
sys.path.extend([f"{main_dir}/Dataset",
f"{main_dir}/CommonFunctions",
f"{main_dir}/PostProcess",
f"{main_dir}/ModelMaker"])
from SeqTokSim2WayComplDS import SeqTokSim2WayComplDS
from ProgramArguments import parseArguments
from Utilities import *
from SimilConfusion import SimilConfusAnalysis
from ExpSiamModel import (getLossFunction,
relaxedCrossEntropy, sinCrossEntropy,
cappedCrossEntropy)
def main(args):
"""
Main function of program for predicting similarity of
two source code samples
Parameters:
- args -- Parsed command line arguments
as object returned by ArgumentParser
"""
resetSeeds()
if not os.path.exists(args.out_dir):
os.makedirs(args.out_dir)
latest_checkpoint = getCheckpoint(args.ckpt_dir, args.ckpt)
_ds = SeqTokSim2WayComplDS(args.dataset,
short_code_th = 1,
max_seq_length = args.seq_len,
seq_len_to_pad = args.pad_len,
labels01 = not args.symmetric_labels)
#Indices of problems corresponding to subsequences of samples
_sample_probl_indices = _ds.sample_probl_indices
with open(f"{args.out_dir}/problem_indices.pcl", "wb") as _f:
pickle.dump(_sample_probl_indices, _f)
#Create parallelization strategy for multi GPU mode
#It also can be either MirroredStrategy or MultiWorkerMirroredStrategy
#But MultiWorkerMirroredStrategy works better
strategy = tf.distribute.MultiWorkerMirroredStrategy()
#distribute.MirroredStrategy()
print("Number of devices: {}".format(strategy.num_replicas_in_sync))
#Construct DNN in the scope of parallelization strategy.
with strategy.scope():
# Everything that creates variables should be under the strategy scope.
# In general this is only model construction & `compile()`.
print("Restoring from", latest_checkpoint)
if args.loss:
_loss = getLossFunction(args.loss)
with tf.keras.utils.custom_object_scope({'loss_function': _loss}):
_dnn = tf.keras.models.load_model(latest_checkpoint)
else:
_dnn = tf.keras.models.load_model(latest_checkpoint)
_test_ds = _ds.testDataset(batch = args.batch)
if args.evaluate:
_eval_loss, _eval_acc = _dnn.evaluate(_test_ds, verbose = args.progress)
print("\n")
print("Evaluation accuracy is {:5.2f}%".format(_eval_acc * 100))
print("Evaluation loss is {:5.2f}".format(_eval_loss))
_prob = _dnn.predict(_test_ds, verbose = args.progress)
_prob = _prob[:,0]
with open(f"{args.out_dir}/similarity_probabilities.pcl", "wb") as _f:
pickle.dump(_prob, _f)
"""
_confusion = SimilConfusAnalysis(_prob, _labels,
_ds.solution_names,
_ds.problems, _annotations)
"""
#_confusion.writeReport()
##############################################################################
# Args are described below
##############################################################################
if __name__ == '__main__':
print("\nEVALUATION OF SOURCE CODE SIMILARITY ANALYZER")
#Handle command-line arguments
parser = argparse.ArgumentParser(description =
"Evaluation of source code similarity analyzer")
parser.add_argument("dataset",
type=str, help="dataset directory")
parser.add_argument("ckpt_dir",
type=str, help="checkpoint directory")
parser.add_argument("out_dir",
type=str, help="directory for results")
parser.add_argument("--ckpt", default = None,
type=str, help="checkpoint file")
parser.add_argument("--batch", default=32, type=int,
help="batch size")
parser.add_argument('--seq_len', default=None, type=int,
help='maximum lengths of token sequence')
parser.add_argument('--pad_len', default=None, type=int,
help='sequence lengths to pad to')
parser.add_argument('--symmetric_labels', action='store_true',
default=False,
help="use symmetric labels: -1 and +1")
parser.add_argument("--loss", type=str, default=None,
help="loss function to optimize")
parser.add_argument('--evaluate', action='store_true',
default=False,
help="run evaluation")
parser.add_argument('--progress', default=2, type=int,
choices=[0, 1, 2],
help="mode of keras training progress bar")
args = parseArguments(parser)
main(args)
|