def test_read_tag_logits_with_pattern_v2(self):
# Create fake logs for the read_tag_logits() function to consume.
tempdir = self.get_temp_dir()
writer = tf2.summary.create_file_writer(tempdir, max_queue=0)
with writer.as_default():
# Events matching pattern v2.
tf2.summary.scalar('rltaglogits/op_indices_0/0', 1.0, step=42)
tf2.summary.scalar('rltaglogits/op_indices_0/1', 2.0, step=42)
tf2.summary.scalar('rltaglogits/op_indices_0/2', 3.0, step=42)
tf2.summary.scalar('rltaglogits/op_indices_1/0', 4.0, step=42)
tf2.summary.scalar('rltaglogits/op_indices_1/1', 5.0, step=42)
tf2.summary.scalar('rltaglogits/op_indices_1/2', 6.0, step=42)
tf2.summary.scalar('rltaglogits/filters_indices_0/0', 7.0, step=42)
# Events not matching any pattern.
tf2.summary.scalar('global_step/sec', 10.0, step=42)
self.evaluate(writer.init())
self.evaluate(tf.summary.all_v2_summary_ops())
self.evaluate(writer.flush())
# Try to read the events from file.
self.assertAllClose(
analyze_mobile_search_lib.read_tag_logits(tempdir),
{
42: {
'op_indices': [[1.0, 2.0, 3.0], [4.0, 5.0, 6.0]],
'filters_indices': [[7.0]],
}
})
def test_read_tag_logits_with_two_valid_steps(self):
# Create fake logs for the read_tag_logits() function to consume.
tempdir = self.get_temp_dir()
writer = tf2.summary.create_file_writer(tempdir, max_queue=0)
values = tf.placeholder(tf.float32, [8])
global_step = tf.placeholder(tf.int64, ())
with writer.as_default():
tf2.summary.scalar(
'rltaglogits/op_indices_0/0', values[0], step=global_step)
tf2.summary.scalar(
'rltaglogits/op_indices_0/1', values[1], step=global_step)
tf2.summary.scalar(
'rltaglogits/op_indices_0/2', values[2], step=global_step)
tf2.summary.scalar(
'rltaglogits/op_indices_1/0', values[3], step=global_step)
tf2.summary.scalar(
'rltaglogits/op_indices_1/1', values[4], step=global_step)
tf2.summary.scalar(
'rltaglogits/op_indices_1/2', values[5], step=global_step)
tf2.summary.scalar(
'rltaglogits/filters_indices_0/0', values[6], step=global_step)
tf2.summary.scalar(
'rltaglogits/filters_indices_0/1', values[7], step=global_step)
self.evaluate(writer.init())
summary_op = tf.summary.all_v2_summary_ops()
flush_op = writer.flush()
with self.cached_session() as sess:
sess.run(summary_op, {
values: [1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0],
global_step: 31,
})
sess.run(flush_op)
sess.run(summary_op, {
values: [9.0, 10.0, 11.0, 12.0, 13.0, 14.0, 15.0, 16.0],
global_step: 42,
})
sess.run(flush_op)
# Now check that read_tag_logits() processes the events correctly.
self.assertAllClose(
analyze_mobile_search_lib.read_tag_logits(tempdir),
{
31: {
'op_indices': [[1.0, 2.0, 3.0], [4.0, 5.0, 6.0]],
'filters_indices': [[7.0, 8.0]],
},
42: {
'op_indices': [[9.0, 10.0, 11.0], [12.0, 13.0, 14.0]],
'filters_indices': [[15.0, 16.0]],
},
})
def test_read_tag_logits_with_invalid_entry(self):
# Create fake logs for the read_tag_logits() function to consume.
tempdir = self.get_temp_dir()
writer = tf2.summary.create_file_writer(tempdir, max_queue=0)
with writer.as_default():
tf2.summary.scalar('rltaglogits/op_indices_0/0', 1.0, step=42)
tf2.summary.scalar('rltaglogits/op_indices_0/1', 2.0, step=42)
tf2.summary.scalar('rltaglogits/op_indices_0/2', 3.0, step=42)
tf2.summary.scalar('rltaglogits/op_indices_1/0', 4.0, step=42)
tf2.summary.scalar('rltaglogits/op_indices_1/1', 5.0, step=42)
tf2.summary.scalar('rltaglogits/op_indices_1/2', 6.0, step=42)
# 'rltaglogits/filters_indices_0/1' is missing from the logs.
tf2.summary.scalar('rltaglogits/filters_indices_0/1', 8.0, step=42)
self.evaluate(writer.init())
self.evaluate(tf.summary.all_v2_summary_ops())
self.evaluate(writer.flush())
# Try to read the events from file. The events from Step 42 should be
# skipped, since some of the data is incomplete.
self.assertEmpty(analyze_mobile_search_lib.read_tag_logits(tempdir))
