nelmo-ux
/

mod-sencevoice

Automatic Speech Recognition

chunk-inference

Model card Files Files and versions

mod-sencevoice / eval

98.3 kB

Ctrl+K

Ctrl+K

1 contributor

History: 3 commits

nelmo-ux's picture

v3: 814h round-2 finetune (ep3) - chunk CER 0.1623, full CER 0.1518

1c7a359 verified 7 days ago

baseline.json

7.25 kB
Add Japanese chunk-streaming SenseVoiceSmall (ep4) with eval results 8 days ago
ep1.json

11.2 kB
Add Japanese chunk-streaming SenseVoiceSmall (ep4) with eval results 8 days ago
ep2.json

11.2 kB
Add Japanese chunk-streaming SenseVoiceSmall (ep4) with eval results 8 days ago
ep3.json

11.2 kB
Add Japanese chunk-streaming SenseVoiceSmall (ep4) with eval results 8 days ago
ep4.json

11.2 kB
Add Japanese chunk-streaming SenseVoiceSmall (ep4) with eval results 8 days ago
eval_full_lr2e4_ep4.json

9.25 kB
v2: 300h cluster finetune (H100, bf16) - chunk CER 0.1679, gap 0.0095 8 days ago
eval_full_lr6e4_ep4.json

9.28 kB
v2: 300h cluster finetune (H100, bf16) - chunk CER 0.1679, gap 0.0095 8 days ago
round2_eval_fullval_ep2.json

9.31 kB
v3: 814h round-2 finetune (ep3) - chunk CER 0.1623, full CER 0.1518 7 days ago
round2_eval_fullval_ep3.json

9.31 kB
v3: 814h round-2 finetune (ep3) - chunk CER 0.1623, full CER 0.1518 7 days ago
round2_eval_fullval_ep4.json

9.23 kB
v3: 814h round-2 finetune (ep3) - chunk CER 0.1623, full CER 0.1518 7 days ago