Issues
- 2
French Model
#169 opened by fatima-falconit - 10
- 3
Base on upstream docker
#131 opened by lissyx - 0
Fix `best_dev_checkpoint` path
#167 opened by wasertech - 0
Improve `ENGLISH_COMPATIBLE`
#165 opened by wasertech - 5
Use MLS dataset
#150 opened by lissyx - 3
OpenSLR 88
#135 opened by lissyx - 1
Expérimenter avec les augmentations (pitch, noise)
#92 opened by lissyx - 5
Numbers translated into words and other irreversible pre processing steps
#160 opened by funboarder13920 - 3
Can't build Dockerfile.train
#155 opened by wasertech - 2
- 6
Precise licence
#149 opened by metal3d - 3
Input data format
#146 opened by gandroz - 0
Upgrade to TensorFlow 1.15.4
#144 opened by lissyx - 0
invalid default lm alpha/beta
#139 opened by lissyx - 0
invalid scorer packaged
#138 opened by lissyx - 2
Évaluation des poids du LM
#127 opened by lissyx - 1
Évaluation des nouveaux hyper-paramètres
#126 opened by lissyx - 3
Mise à jour Lingua Libre
#129 opened by lissyx - 3
Mise à jour Common Voice FR v4
#128 opened by lissyx - 0
Update training code for 0.8
#132 opened by lissyx - 15
WER Modèle Français 0.4
#130 opened by zankerx - 1
Make scripts more generalized
#109 opened by lissyx - 0
Utiliser l'optimiseur de LM
#105 opened by lissyx - 0
Fill in model metadata
#115 opened by lissyx - 0
Set `uid/gid` for inside Docker
#110 opened by lissyx - 0
Produce LM as a packaged scorer
#114 opened by lissyx - 0
- 0
- 0
Install deepspech as a package like it is now required
#112 opened by lissyx - 0
Update KenLM (+Eigen) version
#113 opened by lissyx - 0
Update DeepSpeech to v0.7.4
#122 opened by lissyx - 0
More genericity
#124 opened by lissyx - 2
Integrate Corpora-Creator for easier experimenting
#116 opened by lissyx - 0
Split training / evaluation
#119 opened by lissyx - 0
- 5
Modèle v0.4
#88 opened by lissyx - 12
Comparaison Common Voice v3 FR -- Duplicatas
#102 opened by lissyx - 0
Utiliser et packager le checkpoint `best_dev`
#94 opened by lissyx - 1
Exporter les modèles depuis le `best_dev`
#95 opened by lissyx - 1
Améliorer le filtrage de Common Voice v3
#99 opened by lissyx - 2
- 0
Écrasement des TSVs Common Voice si fournis
#98 opened by lissyx - 2
- 0
Limitation de la taille du LM
#96 opened by lissyx - 17
Node.js DeepSpeech & modèle francophone
#93 opened by simon-tannai - 17
- 3
Fail to run import_trainingspeech.sh
#86 opened by GourmandGuedj - 1
Générer le modèle de langage avec `generate_lm.py`
#83 opened by lissyx - 10
Fail to run ./bin/run-tc-ldc93s1_new.sh
#84 opened by GourmandGuedj