Commits

Branch selector

 main

User selector

All users

Datepicker

All time

Commit History

Commits on Feb 1, 2025

• List all export library symbols at the end of the build.

  Show description for f6662ad
  petterreinholdtsen

  committed
  f6662ad

  Copy full SHA for f6662ad

• Fix -Wstrict-prototypes

  Show description for c93d8ae
  

  thesamesam

  authored and

  petterreinholdtsen

  committed
  c93d8ae

  Copy full SHA for c93d8ae

• m4: fix -Wstrict-protoypes

  Show description for dee8a49
  

  thesamesam

  authored and

  petterreinholdtsen

  committed
  dee8a49

  Copy full SHA for dee8a49

• Added gitlab CI script building with GCC and clang with a few configure options.

  Show description for bcb70b0
  petterreinholdtsen

  committed
  bcb70b0

  Copy full SHA for bcb70b0

Commits on Jan 30, 2025

• Augmentation/loss tuning, adding more noise data

  jmvalin

  committed
  130913d

  Copy full SHA for 130913d

• Fix VAD fading bug

  Show description for 1685549
  jmvalin

  committed
  1685549

  Copy full SHA for 1685549

Commits on Jan 1, 2025

• Fix typo "mix us data" to "mix of data"

  Show description for 751dc33
  

  codesoap

  authored and

  rillian

  committed
  751dc33

  Copy full SHA for 751dc33

Commits on Aug 24, 2024

• Update training instructions

  Show description for d983458
  jmvalin

  committed
  d983458

  Copy full SHA for d983458

• Add skip connections

  jmvalin

  committed
  24736fb

  Copy full SHA for 24736fb

• Add sparse option, loss tuning

  Show description for 9969fa7
  jmvalin

  committed
  9969fa7

  Copy full SHA for 9969fa7

Commits on Aug 18, 2024

• Minor tweaks to augmentation

  jmvalin

  committed
  2ea0e31

  Copy full SHA for 2ea0e31

Commits on Aug 17, 2024

• Better signal/noise normalization

  jmvalin

  committed
  eca43e8

  Copy full SHA for eca43e8

Commits on Aug 12, 2024

• New models with updated augmentation

  Show description for 57f75a7
  jmvalin

  committed
  57f75a7

  Copy full SHA for 57f75a7

Commits on Aug 11, 2024

• Consider energy changes in gain decay cap

  Show description for bb18d2f
  jmvalin

  committed
  bb18d2f

  Copy full SHA for bb18d2f

• Augment with saturation and quantization

  jmvalin

  committed
  477197c

  Copy full SHA for 477197c

Commits on Aug 10, 2024

• Add Viterbi-based VAD for training

  Show description for eea0493
  jmvalin

  committed
  eea0493

  Copy full SHA for eea0493

• Tuning noise augmentation

  jmvalin

  committed
  d951101

  Copy full SHA for d951101

• Randomize filter based on pole/zero location

  jmvalin

  committed
  1b982a6

  Copy full SHA for 1b982a6

• Augmentation using foreground noise

  jmvalin

  committed
  ddc4686

  Copy full SHA for ddc4686

Commits on May 20, 2024

• Added sha256 check for downloaded data

  Show description for 2e3c812
  jmvalin

  committed
  2e3c812

  Copy full SHA for 2e3c812

Commits on Apr 16, 2024

• Fix compilation errors.

  Show description for 372f7b4
  

  Timothy B. Terriberry

  committed
  372f7b4

  Copy full SHA for 372f7b4

Commits on Apr 15, 2024

• Add "little" model

  jmvalin

  committed
  904a876

  Copy full SHA for 904a876

• Updating copyright notices

  jmvalin

  committed
  0b50c45

  Copy full SHA for 0b50c45

• Adding more ways to load model

  Show description for 4ad7669
  jmvalin

  committed
  4ad7669

  Copy full SHA for 4ad7669

Commits on Apr 14, 2024

• Add missing file

  jmvalin

  committed
  ec979ef

  Copy full SHA for ec979ef

• Script for shrinking C models

  Show description for 12639df
  jmvalin

  committed
  12639df

  Copy full SHA for 12639df

• Updated README

  jmvalin

  committed
  31a4a23

  Copy full SHA for 31a4a23

• Reviving model loading based on Opus code

  jmvalin

  committed
  c7845a3

  Copy full SHA for c7845a3

• Initial sparse matrix support copied from Opus

  Show description for f4700a6
  jmvalin

  committed
  f4700a6

  Copy full SHA for f4700a6

• Adding one extra frame of delay

  jmvalin

  committed
  84fe83b

  Copy full SHA for 84fe83b

• Noise data

  jmvalin

  committed
  37cf35f

  Copy full SHA for 37cf35f

• Forgot weight dumping script

  jmvalin

  committed
  0ef02af

  Copy full SHA for 0ef02af

Commits on Apr 11, 2024

• Model update

  Show description for 5917870
  jmvalin

  committed
  5917870

  Copy full SHA for 5917870

• More details for TRAINING

  jmvalin

  committed
  f56003f

  Copy full SHA for f56003f

Commits on Apr 10, 2024

• Default to no float debug

  jmvalin

  committed
  5944647

  Copy full SHA for 5944647

Pagination

Previous

Next