TarteelAI/tarteel-ml

Issues

• pip or conda ?

  #55 opened 4 years ago by Karim-53
  5

• Update tutorials

  #57 opened 4 years ago by Karim-53
  0

• Create a Dockerfile to support repo on multiple OS like Windows

  #53 opened 4 years ago by naushervan
  2

• Same number of audio files with or without surah [-s] argument

  #46 opened 4 years ago by omerasif-itu
  2

• Logging exception due to invalid arguments

  #49 opened 4 years ago by murtraja
  1

• Missing audio_processing directory

  #54 opened 4 years ago by shehrbano1248
  0

• Write a script to download, convert to features, and delete audio by ayah.

  #12 opened 6 years ago by hmzh-khn
  2

• How to pass a live stream of audio to the ML

  #48 opened 4 years ago by Nouraaaawwad
  2

• How to Add/Edit Wiki pages?

  #51 opened 4 years ago by Ahmed-H-N
  0

• `dataset_csv_url` : A ghost argument?

  #45 opened 5 years ago by omerasif-itu
  1

• Error when creating conda env, sox=14.4.2 not found.

  #44 opened 5 years ago by sdqmd
  1

• Update training utils and models with OpenNMT based methods.

  #27 opened 5 years ago by hmzh-khn
  0

• the files for the fattiha overfitting experiment

  #40 opened 5 years ago by Nouraaaawwad
  4

• Error when trying to do "Fatihah overfitting experiment" but with surat Al-Ekhlass

  #39 opened 5 years ago by Nouraaaawwad
  1

• how to generate language model?

  #38 opened 5 years ago by lerndeep
  1

• Move feature generation files to an archived code location.

  #28 opened 5 years ago by hmzh-khn
  0

• Update the documentation for how to download the dataset and its audio.

  #30 opened 5 years ago by hmzh-khn
  0

• Refactor utilities related to the Quranic dataset.

  #29 opened 5 years ago by hmzh-khn
  1

• Train a model for gender prediction

  #8 opened 6 years ago by abidlabs
  1

• Some demographic metadata missing from Tarteel Dataset CSV

  #19 opened 5 years ago by abidlabs
  2

• get error IndexError: too many indices for array

  #22 opened 5 years ago by Mohamedmtc
  1

• Train an initial seq2seq model for recitation2text using the v1 dataset.

  #15 opened 5 years ago by hmzh-khn
  3

• Create an overfitted model using ~100 pieces of MFCC data.

  #13 opened 5 years ago by hmzh-khn
  1

• Convert all v1 recordings to filter bank representations.

  #14 opened 5 years ago by hmzh-khn
  1

• get Invalid wave header found

  #23 opened 5 years ago by MahmoudMabrok
  2

• Make improvements to the args passed to `download.py` for greater specification of what to download.

  #6 opened 6 years ago by hmzh-khn
  2

• Validate that mfcc feature generation script works by reconstructing the signal and listening to it.

  #9 opened 5 years ago by hmzh-khn
  1

• Migrate `has_speech` function to use evaluation data once it is ready.

  #7 opened 5 years ago by hmzh-khn
  1

• Downloading Surah Al-Fatihah only took long time

  #25 opened 5 years ago by SamahZaro
  7

• Conda requirements are OSX specific

  #16 opened 6 years ago by AymenQ
  2

• Write helper functions to read the .npy files that hold the features.

  #10 opened 6 years ago by hmzh-khn
  1

• Write helper functions for working with the one-hot encoding.

  #11 opened 6 years ago by hmzh-khn
  1

• Convert all v1 recordings to MFCC representations.

  #3 opened 6 years ago by hmzh-khn
  4

• Create a train-test-validation split for the recordings by verse.

  #5 opened 6 years ago by hmzh-khn
  1

• Complete MFCC and Filter Bank script.

  #2 opened 6 years ago by hmzh-khn
  2

• Create a one-hot output encoding for the Arabic alphabet, including harakat and madd.

  #4 opened 6 years ago by hmzh-khn
  1

• Provide instructions for downloading labelled data

  #1 opened 6 years ago by kareemn
  1