-
Notifications
You must be signed in to change notification settings - Fork 3.6k
Pull requests: NVIDIA-NeMo/Speech
Author
Label
Projects
Milestones
Reviews
Assignee
Sort
Pull requests list
Add DFlash speculative decoding support to NeMo SpeechLM vLLM plugin
#16239
opened Sep 10, 2026 by
Slyne
Contributor
•
6/6
Loading…
4 of 5 tasks
optional tail silence to flush the decoder
ASR
#16237
opened Sep 9, 2026 by
naymaraq
Collaborator
Loading…
8 tasks
Fix negative offsets in GPUBiasingMultiModel.remove_model
ASR
community-request
#16236
opened Sep 9, 2026 by
kzos
Loading…
[Fix] run_nvrtc corrupts interned 1-byte bytes singleton on empty NVRTC log (refresh main)
community-request
core
Changes to NeMo Core
#16234
opened Sep 9, 2026 by
yushuosun
Contributor
Loading…
[audio codec] Apply MMD loss to pre-dropout embeddings
TTS
#16233
opened Sep 9, 2026 by
rfejgin
Collaborator
Loading…
Speed up the cache-aware streaming RNNT pipeline
ASR
common
#16231
opened Sep 8, 2026 by
naymaraq
Collaborator
Loading…
3 of 8 tasks
Updates for Sortformer training
ASR
common
Speaker Tasks
#16227
opened Sep 7, 2026 by
ipmedenn
Collaborator
Loading…
4 of 8 tasks
fix(asr): SubsamplingReductionModule pooling computes lengths for a single MaxPool1d pass
ASR
community-request
waiting-on-maintainers
Waiting on maintainers to respond
#16226
opened Sep 7, 2026 by
ManoharPaturi
Loading…
fix(asr): repair ConvSubsampling forward paths missed by the MaskedConvSequential refactor
ASR
community-request
waiting-on-maintainers
Waiting on maintainers to respond
#16225
opened Sep 7, 2026 by
ManoharPaturi
Loading…
docs: exclude 16 recurring broken Lightning doc links from linkcheck
community-request
waiting-on-maintainers
Waiting on maintainers to respond
#16222
opened Sep 7, 2026 by
VaggelisGian
Contributor
Loading…
4 of 5 tasks
feat(tts): Optimize and harden EasyMagpie vLLM-Omni streaming at BS 32, 64, 128
community-request
TTS
waiting-on-customer
Waiting on the original author to respond
#16220
opened Sep 6, 2026 by
styagi130
Contributor
Loading…
fix(voice_agent): recognize a sentence ending in a decimal number as complete
community-request
waiting-on-maintainers
Waiting on maintainers to respond
#16219
opened Sep 5, 2026 by
udsy19
Contributor
Loading…
3 of 7 tasks
fix(tts): restore Lightning-2.x on_validation_epoch_end signature in SSL models
community-request
TTS
waiting-on-maintainers
Waiting on maintainers to respond
#16217
opened Sep 5, 2026 by
udsy19
Contributor
Loading…
3 of 5 tasks
docs: add adalat-ai/koyal-indic-600m-1.0 to featured community checkp…
ASR
community-request
waiting-on-maintainers
Waiting on maintainers to respond
#16215
opened Sep 5, 2026 by
janaab11
Loading…
3 of 7 tasks
Add SLURM-aware exp manager run timer
#16214
opened Sep 5, 2026 by
pzelasko
Collaborator
Loading…
6 of 7 tasks
Add per-stream speaker limits to Sortformer sessions
ASR
community-request
Speaker Tasks
waiting-on-maintainers
Waiting on maintainers to respond
#16210
opened Sep 2, 2026 by
tianyi-zhang-02
Loading…
6 of 7 tasks
Extract shared Waiting on maintainers to respond
segment_wav into tts_dataset_utils.py.
community-request
TTS
waiting-on-maintainers
#16209
opened Sep 2, 2026 by
StarsExpress
Loading…
4 of 6 tasks
feat(tts): vendor Perth watermarking in EasyMagpie codec stage
community-request
TTS
waiting-on-customer
Waiting on the original author to respond
#16208
opened Sep 2, 2026 by
yhayarannvidia
Loading…
fix(asr): stop evaluate_eou returning negative false_negatives and missing
ASR
community-request
waiting-on-maintainers
Waiting on maintainers to respond
#16201
opened Sep 2, 2026 by
Kayvan-Zahiri
Loading…
fix(asr): end the final VAD speech segment at the last frame's end
ASR
community-request
waiting-on-maintainers
Waiting on maintainers to respond
#16200
opened Sep 2, 2026 by
Kayvan-Zahiri
Loading…
3 of 7 tasks
[speechlm2] Add streaming inference pipeline for NemotronVoiceChat
#16198
opened Sep 2, 2026 by
erastorgueva-nv
Collaborator
Loading…
6 of 8 tasks
feat(speechlm2): add native MTP speculative decoding in vLLM
#16196
opened Sep 1, 2026 by
pzelasko
Collaborator
•
5/6
Loading…
feat(speechlm2): export portable checkpoints and serve them with vLLM
ASR
audio
#16195
opened Sep 1, 2026 by
pzelasko
Collaborator
•
4/6
Loading…
feat(speechlm2): add padding-free packed speech training
ASR
audio
#16194
opened Sep 1, 2026 by
pzelasko
Collaborator
•
3/6
Loading…
feat(data): add indexed resumable multimodal loading and exact token budgets
ASR
common
#16193
opened Sep 1, 2026 by
pzelasko
Collaborator
•
2/6
Loading…
Previous Next
ProTip!
Mix and match filters to narrow down what you’re looking for.