issues
search
castorini
/
howl
Wake word detection modeling toolkit for Firefox Voice, supporting open datasets like Speech Commands and Common Voice.
Mozilla Public License 2.0
194
stars
28
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
doxygen setup
#78
ljj7975
closed
3 years ago
0
Update demo to use the HowlClient and read settings from workspace
#77
edwinzhng
closed
3 years ago
0
fixing phone class
#76
ljj7975
closed
3 years ago
0
refactor inference (taking in context)
#75
ljj7975
closed
3 years ago
1
using the right inference engine
#74
ljj7975
closed
3 years ago
1
Inference weight now requires extra weight for OOV
#73
ljj7975
closed
3 years ago
0
Filter out bad stitched samples by checking the speech to text transcription
#72
ljj7975
closed
3 years ago
0
Document codebase and generate web docs with Sphinx
#71
daemon
opened
3 years ago
0
Setting up Ci with pre-commit
#70
ljj7975
closed
3 years ago
0
Implementing auto stopping
#69
ljj7975
opened
3 years ago
0
Integrating hyper parameter tuning framework
#68
ljj7975
opened
3 years ago
0
French dataset
#67
taminhtoan2601
opened
3 years ago
2
stitching vocab to generate wakeword samples
#66
ljj7975
closed
3 years ago
0
restructuring/refactoring files for howl/training
#65
ljj7975
opened
3 years ago
1
missing changes for dataset stat change
#64
ljj7975
closed
3 years ago
0
Howl on CPU
#63
ljj7975
closed
3 years ago
1
including vocab sample in different form (ex. singular,plural)
#62
ljj7975
opened
3 years ago
0
Print datasets stat
#61
ljj7975
closed
3 years ago
3
Setting up CI and implement testing
#60
ljj7975
closed
3 years ago
0
stitching audio samples to generate diverse positive dataset
#59
ljj7975
opened
3 years ago
10
improving vocab searching logic
#58
ljj7975
closed
3 years ago
1
printing out how many instances of vocab is captured from dataset generation
#57
ljj7975
closed
3 years ago
4
speeding up create_raw_dataset.py
#56
ljj7975
opened
3 years ago
4
scripts for generating new dataset with custom command
#55
ljj7975
closed
3 years ago
0
minor bug fix
#54
ljj7975
closed
3 years ago
0
Refactoring InferenceContext and fixing incorrect ctc blank index
#53
ljj7975
closed
3 years ago
2
Confirming the functionality of token type == 'phone'
#52
ljj7975
opened
3 years ago
0
renaming howl_training folder to training
#51
ljj7975
closed
3 years ago
0
fixing issues with demo.py for the model trained with ctc loss
#50
ljj7975
closed
3 years ago
0
Update README and requirements for PyPI release
#49
edwinzhng
closed
3 years ago
0
Fix tensor division error
#48
edwinzhng
closed
3 years ago
0
Fix training with ctc loss
#47
ljj7975
closed
3 years ago
2
pip install howl does not work
#46
ayeganov
closed
3 years ago
8
Add flag to prevent duplicate inference and refactor PyTorch hub config
#45
edwinzhng
closed
3 years ago
1
demo: allow to run on CPU
#44
Mic92
closed
3 years ago
2
fixing the link to the datasets
#43
ljj7975
closed
3 years ago
0
Example of the CTC loss usage
#42
joemayer1
closed
3 years ago
3
Create PyPI Package and Separate Training/Client Code
#41
edwinzhng
closed
3 years ago
0
Scripts for reproducing the paper results
#40
ljj7975
closed
3 years ago
0
including pretrained model and instruction on how to run demo with it
#39
ljj7975
closed
3 years ago
2
missing prerequisites
#38
idog-ceva
closed
3 years ago
3
Model trained sucessfully but not detection result
#37
Alex-Beh
opened
3 years ago
12
UnicodeDecodeError: 'utf-8' codec can't decode byte 0xc9 in position 7089: invalid continuation byte
#36
Alex-Beh
closed
3 years ago
0
Scripts for generating dataset for precise from existing dataset
#35
ljj7975
closed
3 years ago
0
Requirements conflict: torchvision requires torch==1.5.1
#34
krisgesling
closed
3 years ago
1
size mismatch RuntimeError running CLI demo
#33
krisgesling
closed
3 years ago
3
Fixing the path argument issue
#32
ljj7975
closed
4 years ago
0
Hey snips dataset support
#31
ljj7975
closed
3 years ago
0
Errors when reproducing Hey Firefox results
#30
edwinzhng
closed
4 years ago
3
Training without GPU
#29
codeghees
closed
4 years ago
1
Previous
Next