forked from magenta/symbolic-music-diffusion
-
Notifications
You must be signed in to change notification settings - Fork 0
/
calculate_metrics.py
57 lines (42 loc) · 1.4 KB
/
calculate_metrics.py
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
"""Calculates OA metrics for """
import glob
import os
import pickle
from absl import app, flags, logging
from IPython import embed
from utils import metrics
FLAGS = flags.FLAGS
flags.DEFINE_string(
"input", "dataset/subset/midi", "Location of MIDI output from sample_audio."
)
def _load_pkl(fname):
with open(fname, "rb") as f:
return pickle.load(f)
def main(argv):
# load real and generated note_seqs
id2note_seqs = {}
real = ["unedited"]
fake = ["generated", "edited"]
for id in real + fake:
path = os.path.join(FLAGS.input, id, "ns")
# load all the pickles
id2note_seqs[id] = [_load_pkl(fname) for fname in glob.glob(f"{path}/*.pkl")]
# calculate metrics for each output
for fake_id in fake:
logging.info(f"calculating metrics for {fake_id}")
(
pitch_consistency,
pitch_variance,
duration_consistency,
duration_variance,
) = metrics.framewise_self_sim(
id2note_seqs[real[0]][:50], id2note_seqs[fake_id]
)
results = f"{pitch_consistency = }, {pitch_variance = }, {duration_consistency =}, {duration_variance = }"
print(results)
# write to disk
output = os.path.join(FLAGS.input, f"{fake_id}_self_sim.txt")
with open(output, "w") as f:
f.write(f"{results}\n")
if __name__ == "__main__":
app.run(main)