bzamecnik ml issues - Githubissues

bzamecnik / ml

Machine learning projects, often on audio datasets

MIT License

92 stars 39 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

Why are you using binary cross entropy

#46 aribornstein closed 6 years ago
1
extract_features.py gives very weird error.

#45 JBloodless closed 6 years ago
5
predict.py dropping different errors

#44 JBloodless closed 6 years ago
2
[chord-recognition]: Try CTC loss for the sequence label prediction task.

#43 bzamecnik opened 6 years ago
0
There's missing .py tools

#42 denisb411 opened 6 years ago
2
TypeError: unsupported operand type(s) for +: 'NoneType' and 'str'

#41 sd12832 opened 6 years ago
7
Update the chord classifier to current versions of Keras/TensorFlow

#40 bzamecnik opened 7 years ago
0
Use Seaborn heatmap() instead of corrplot()

#39 bzamecnik opened 7 years ago
0
The corrplot function has been deprecated in v0.6

#38 mwprog closed 7 years ago
1
Update the code to use tfr-0.2.1 API.

#37 bzamecnik closed 8 years ago
0
use tfr instead of music-processing-experiments

#36 bzamecnik closed 8 years ago
0
deploy in pure client-side via keras-js!

#35 bzamecnik opened 8 years ago
1
avoid duplication between the ml and deep-instrument-heroku repos

#34 bzamecnik opened 8 years ago
0
write blog post describing the instrument classification task

#33 bzamecnik opened 8 years ago
0
package the chromagram library

#32 bzamecnik closed 7 years ago
1
start the pitch classification task

#31 bzamecnik opened 8 years ago
0
generate audio data with chords

#30 bzamecnik opened 8 years ago
0
generate audio data with sequence of tones

#29 bzamecnik opened 8 years ago
0
generalize to variable length audio samples

#28 bzamecnik opened 8 years ago
1
visualize the chromagrams

#27 bzamecnik opened 8 years ago
1
rewrite the model to the Keras functional API

#26 bzamecnik opened 8 years ago
0
compute ranking of models - which is the best so far?

#25 bzamecnik opened 8 years ago
0
separate input data directory and model directory

#24 bzamecnik opened 8 years ago
0
put the models into unique directories

#23 bzamecnik opened 8 years ago
0
Analyze error with respect to pitch.

#22 bzamecnik opened 8 years ago
0
Deploy the instrument classsification model webapp publicly.

#21 bzamecnik closed 8 years ago
1
Compute the chromagram features as a sklearn transformer

#20 bzamecnik opened 8 years ago
0
Prepare a library for loading and saving data

#19 bzamecnik opened 8 years ago
0
Wrap the model into a web application.

#18 bzamecnik opened 8 years ago
1
Analyze training via TensorBoard.

#17 bzamecnik opened 8 years ago
0
Produce a model evaluation report as a generated IPython notebook.

#16 bzamecnik opened 8 years ago
0
Produce a model evaluation report.

#15 bzamecnik opened 8 years ago
0
Analyze and describe the dataset.

#14 bzamecnik opened 8 years ago
0
Convert the model notebook into a script

#13 bzamecnik opened 8 years ago
1
Separate scripts for data preparation, training and evaluation.

#12 bzamecnik opened 8 years ago
1
Wrap the model into an end-to-end tool.

#11 bzamecnik opened 8 years ago
0
Git submodule init doesn't work

#10 bzamecnik closed 8 years ago
1
Compare the model performance with trivial models

#9 bzamecnik opened 8 years ago
0
Publish the dataset of computed chromagram features

#8 bzamecnik opened 8 years ago
1
Describe how to prepare the features

#7 bzamecnik opened 8 years ago
0
Evaluate the errors in pitch class vector prediction

#6 bzamecnik opened 8 years ago
0
Make it easy to use the predicted pitch classes eg. in Sonic Visualizer

#5 bzamecnik opened 8 years ago
0
Add music-processing-experiments as git submodule

#4 bzamecnik closed 8 years ago
0
Write a proper README

#3 bzamecnik opened 8 years ago
0
Compare a few approaches to binary pitch class vector classification

#2 bzamecnik opened 8 years ago
0
beatles time map - inverse polar mapping

#1 bzamecnik opened 9 years ago
0