Hugging Face
Models
Datasets
Spaces
Posts
Docs
Solutions
Pricing
Log In
Sign Up
espnet
/
WavLabLM-MK-40k
like
0
ESPnet
fleurs
babel
voxpopuli
commonvoice
102 languages
audio
self-supervised-learning
speech-recognition
arxiv:
2309.15317
arxiv:
1804.00015
License:
cc-by-4.0
Model card
Files
Files and versions
Community
Use this model
main
WavLabLM-MK-40k
/
exp_li
/
hubert_iter2_train_ssl_torchaudiohubert_large_960h_pretrain_it2_wavlm_raw_layer_9
/
images
2 contributors
History:
1 commit
William Chen
init
921b02e
9 months ago
acc_m.png
33.1 kB
init
9 months ago
acc_u.png
31.1 kB
init
9 months ago
backward_time.png
38.9 kB
init
9 months ago
correct_m.png
32.7 kB
init
9 months ago
correct_u.png
31.5 kB
init
9 months ago
count_m.png
32.2 kB
init
9 months ago
count_u.png
30.5 kB
init
9 months ago
forward_time.png
42.9 kB
init
9 months ago
gpu_max_cached_mem_GB.png
38.6 kB
init
9 months ago
iter_time.png
30.3 kB
init
9 months ago
loss.png
37.7 kB
init
9 months ago
optim0_lr0.png
30.2 kB
init
9 months ago
optim_step_time.png
29.4 kB
init
9 months ago
train_time.png
32.4 kB
init
9 months ago