issues
search
JamesOwers
/
midi_degradation_toolkit
A toolkit for generating datasets of midi files which have been degraded to be 'un-musical'.
MIT License
38
stars
5
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
Update from master
#68
apmcleod
closed
5 years ago
0
Gpu fixes
#67
JamesOwers
closed
5 years ago
1
If command/pianoroll datasets only partially made, they are not re-made in full
#66
JamesOwers
closed
5 years ago
1
Fix the pip install
#65
JamesOwers
closed
5 years ago
7
Speed up overlaps check
#64
JamesOwers
closed
4 years ago
1
Implement Evaluation
#63
apmcleod
closed
5 years ago
0
Baseline
#62
apmcleod
closed
5 years ago
9
Ensure padding and masking is done correctly
#61
apmcleod
closed
5 years ago
1
Overlaps
#60
apmcleod
closed
5 years ago
5
Degraded excerpts may not start from time 0.
#59
apmcleod
closed
5 years ago
0
Models
#58
JamesOwers
closed
5 years ago
1
Fix overlaps on input
#57
apmcleod
closed
5 years ago
4
Before release: DQA of the released data!
#56
JamesOwers
opened
5 years ago
4
Adapt `make_dataset.py` such that you can supply a flag to make it do just single parts
#55
JamesOwers
closed
5 years ago
2
Have make_dataset not rely on torch
#54
apmcleod
closed
5 years ago
2
Datasets and loaders
#53
JamesOwers
closed
5 years ago
3
Fix the absolute diatribe of warnings and logs spewed out for make_dataset.py
#52
JamesOwers
closed
5 years ago
0
Enforce common tempo and pitch basis in downloads.
#51
apmcleod
closed
5 years ago
4
Update bugfix from master
#50
apmcleod
closed
5 years ago
0
Degradation error - just noting in case you know...
#49
JamesOwers
closed
5 years ago
2
Acme
#48
apmcleod
closed
5 years ago
8
Degradations
#47
apmcleod
closed
5 years ago
0
Decide how to ignore tracks in degradations
#46
apmcleod
closed
5 years ago
1
Update from master
#45
apmcleod
closed
5 years ago
0
Add flag to force time-based degradations to choose from existing durations (and onset times?)
#44
apmcleod
closed
5 years ago
4
Make dataset script
#43
JamesOwers
closed
5 years ago
2
Decide what datasets to include in ACME v1.0, and what parameters to use for the degs.
#42
apmcleod
closed
5 years ago
1
Piano midi
#41
apmcleod
closed
5 years ago
2
downloads.extract_zip contains assumptions specific to PPD data
#40
apmcleod
opened
5 years ago
3
Enable join_notes to join more than 2 notes together, if possible.
#39
apmcleod
closed
5 years ago
1
Update from master
#38
apmcleod
closed
5 years ago
0
Enforce DataFrame formatting
#37
apmcleod
closed
4 years ago
8
Degradation optimizations
#36
apmcleod
closed
5 years ago
0
Investigate use of `inplace` operations for efficiency in degradations
#35
JamesOwers
closed
5 years ago
1
Iterate by index over pandas dataframes is not efficient - edit if possible
#34
JamesOwers
closed
5 years ago
0
df_to_csv and csv_to_df should probably be in the same place
#33
apmcleod
closed
4 years ago
2
Support excerpts with multiple degradations
#32
apmcleod
opened
5 years ago
4
Allow the user to specify a desired window size for excerpts
#31
apmcleod
closed
5 years ago
1
Code to measure types of errors present in a transcription
#30
apmcleod
closed
4 years ago
1
New degradation parameters
#29
apmcleod
opened
5 years ago
0
Degradations
#28
apmcleod
closed
5 years ago
6
Degradation input validation is not explicitly performed.
#27
apmcleod
opened
5 years ago
3
Convert degradations to integers
#26
apmcleod
closed
5 years ago
4
Degradation input format
#25
apmcleod
closed
5 years ago
2
The overlap check in read_note_csv is dog slow...
#24
JamesOwers
closed
5 years ago
1
Update from master
#23
apmcleod
closed
5 years ago
0
Ensure Randomness and reproducibility
#22
apmcleod
opened
5 years ago
3
Set rounding level for onset and durations, and set warning strategy for df checking.
#21
apmcleod
closed
5 years ago
0
Decide what to do with "invalid" MIDI
#20
apmcleod
closed
5 years ago
4
Use data structures
#19
JamesOwers
closed
5 years ago
0
Previous
Next