[voice agent] Add support for nemotron-nano-v3, magpie-tts and misc improvements by stevehuang52 · Pull Request #15318 · NVIDIA-NeMo/NeMo

stevehuang52 · 2026-01-26T22:53:29Z

Important

The Update branch button must only be pressed in very rare occassions.
An outdated branch is never blocking the merge of a PR.
Please reach out to the automation team before pressing that button.

What does this PR do ?

Add support for nemotron-nano-v3, magpie-tts and misc improvements.

Changelog

Add nemotron-nano-v3 and magpie-tts.
Upgrade to pipecat==0.0.98
Refactor TTS code for cleaner style
Merge server.py with bot_websocket_server.py
Update README

Usage

PR Type:

New Feature
Bugfix
Documentation

Signed-off-by: stevehuang52 <heh@nvidia.com>

…lls overridden method Co-authored-by: Copilot Autofix powered by AI <62310815+github-advanced-security[bot]@users.noreply.github.com> Signed-off-by: He Huang (Steve) <105218074+stevehuang52@users.noreply.github.com>

Signed-off-by: He Huang (Steve) <105218074+stevehuang52@users.noreply.github.com>

Signed-off-by: stevehuang52 <heh@nvidia.com>

…a_nano3 Signed-off-by: stevehuang52 <heh@nvidia.com>

Signed-off-by: stevehuang52 <heh@nvidia.com>

tango4j

There are some minor but obvious mistakes.

Signed-off-by: He Huang (Steve) <105218074+stevehuang52@users.noreply.github.com>

Signed-off-by: He Huang <heh@nvidia.com>

Signed-off-by: stevehuang52 <stevehuang52@users.noreply.github.com>

github-actions · 2026-02-13T23:44:24Z

[🤖]: Hi @stevehuang52 👋,

We wanted to let you know that a CICD pipeline for this PR just finished successfully.

So it might be time to merge this PR or get some approvals.

//cc @chtruong814 @ko3n1g @pablo-garay @thomasdhc

tango4j

Changes confirmed.
Approving.

…mprovements (NVIDIA-NeMo#15318) * WIP namo-v3 Signed-off-by: stevehuang52 <heh@nvidia.com> * update tts tool prompt Signed-off-by: stevehuang52 <heh@nvidia.com> * add nano-v3 support Signed-off-by: stevehuang52 <heh@nvidia.com> * update for vllm Signed-off-by: stevehuang52 <heh@nvidia.com> * refactor and improve tts Signed-off-by: stevehuang52 <heh@nvidia.com> * update title Signed-off-by: stevehuang52 <heh@nvidia.com> * update readme Signed-off-by: stevehuang52 <heh@nvidia.com> * fix tts reset Signed-off-by: stevehuang52 <heh@nvidia.com> * add tts pre-fetch all Signed-off-by: stevehuang52 <heh@nvidia.com> * update tts and prompt Signed-off-by: stevehuang52 <heh@nvidia.com> * fix tts Signed-off-by: stevehuang52 <heh@nvidia.com> * upgrade to pipecat==0.0.98 Signed-off-by: stevehuang52 <heh@nvidia.com> * update for magpie-tts Signed-off-by: stevehuang52 <heh@nvidia.com> * update readme Signed-off-by: stevehuang52 <heh@nvidia.com> * update readme Signed-off-by: stevehuang52 <heh@nvidia.com> * drop redundant code and make websocket port configurable Signed-off-by: stevehuang52 <heh@nvidia.com> * clean up Signed-off-by: stevehuang52 <heh@nvidia.com> * fix linting Signed-off-by: stevehuang52 <heh@nvidia.com> * move tool calling folder Signed-off-by: stevehuang52 <heh@nvidia.com> * fix tts server config Signed-off-by: stevehuang52 <heh@nvidia.com> * Potential fix for code scanning alert no. 16871: `__init__` method calls overridden method Co-authored-by: Copilot Autofix powered by AI <62310815+github-advanced-security[bot]@users.noreply.github.com> Signed-off-by: He Huang (Steve) <105218074+stevehuang52@users.noreply.github.com> * Rename environment from 'nemo-voice-dev' to 'nemo-voice' Signed-off-by: He Huang (Steve) <105218074+stevehuang52@users.noreply.github.com> * fix dependency for nemo26.02 Signed-off-by: stevehuang52 <heh@nvidia.com> * clean up Signed-off-by: stevehuang52 <heh@nvidia.com> * upgrade min lhotse version Signed-off-by: stevehuang52 <heh@nvidia.com> * update Signed-off-by: stevehuang52 <heh@nvidia.com> * fix env yaml Signed-off-by: stevehuang52 <heh@nvidia.com> * replace with pinned version Signed-off-by: stevehuang52 <heh@nvidia.com> * fix en vname Signed-off-by: stevehuang52 <heh@nvidia.com> * update weather function Signed-off-by: stevehuang52 <heh@nvidia.com> * fix nano-v2 vllm config Signed-off-by: stevehuang52 <heh@nvidia.com> * fix env yaml Signed-off-by: stevehuang52 <heh@nvidia.com> * address reviews Signed-off-by: He Huang <heh@nvidia.com> * Apply isort and black reformatting Signed-off-by: stevehuang52 <stevehuang52@users.noreply.github.com> --------- Signed-off-by: stevehuang52 <heh@nvidia.com> Signed-off-by: He Huang (Steve) <105218074+stevehuang52@users.noreply.github.com> Signed-off-by: He Huang <heh@nvidia.com> Signed-off-by: stevehuang52 <stevehuang52@users.noreply.github.com> Co-authored-by: Copilot Autofix powered by AI <62310815+github-advanced-security[bot]@users.noreply.github.com> Co-authored-by: stevehuang52 <stevehuang52@users.noreply.github.com>

stevehuang52 added 18 commits December 30, 2025 10:24

WIP namo-v3

026177f

Signed-off-by: stevehuang52 <heh@nvidia.com>

Merge remote-tracking branch 'origin/main' into heh/va_nano3

c871eb7

Signed-off-by: stevehuang52 <heh@nvidia.com>

Merge remote-tracking branch 'origin/main' into heh/va_nano3

1e79403

update tts tool prompt

1ab37a3

Signed-off-by: stevehuang52 <heh@nvidia.com>

add nano-v3 support

921b4c3

Signed-off-by: stevehuang52 <heh@nvidia.com>

update for vllm

49f6d2b

Signed-off-by: stevehuang52 <heh@nvidia.com>

refactor and improve tts

6d1d953

Signed-off-by: stevehuang52 <heh@nvidia.com>

update title

9b21333

Signed-off-by: stevehuang52 <heh@nvidia.com>

update readme

bb8799e

Signed-off-by: stevehuang52 <heh@nvidia.com>

fix tts reset

2f59910

Signed-off-by: stevehuang52 <heh@nvidia.com>

add tts pre-fetch all

63d3891

Signed-off-by: stevehuang52 <heh@nvidia.com>

update tts and prompt

8f01fee

Signed-off-by: stevehuang52 <heh@nvidia.com>

fix tts

a206083

Signed-off-by: stevehuang52 <heh@nvidia.com>

upgrade to pipecat==0.0.98

d1fc2a5

Signed-off-by: stevehuang52 <heh@nvidia.com>

update for magpie-tts

7e75d38

Signed-off-by: stevehuang52 <heh@nvidia.com>

update readme

26bf617

Signed-off-by: stevehuang52 <heh@nvidia.com>

update readme

8cfe37a

Signed-off-by: stevehuang52 <heh@nvidia.com>

drop redundant code and make websocket port configurable

b41cdf3

Signed-off-by: stevehuang52 <heh@nvidia.com>

stevehuang52 requested a review from tango4j January 26, 2026 22:53

github-advanced-security AI found potential problems Jan 26, 2026

View reviewed changes

Comment thread nemo/agents/voice_agent/pipecat/services/nemo/tts.py Fixed

stevehuang52 added 3 commits January 26, 2026 17:59

clean up

48f1b52

Signed-off-by: stevehuang52 <heh@nvidia.com>

fix linting

9be6f68

Signed-off-by: stevehuang52 <heh@nvidia.com>

move tool calling folder

59aaf3b

Signed-off-by: stevehuang52 <heh@nvidia.com>

github-advanced-security AI found potential problems Jan 27, 2026

View reviewed changes

Comment thread nemo/agents/voice_agent/pipecat/services/nemo/tts.py Fixed

stevehuang52 and others added 6 commits January 27, 2026 10:27

fix tts server config

ddb0619

Signed-off-by: stevehuang52 <heh@nvidia.com>

Rename environment from 'nemo-voice-dev' to 'nemo-voice'

8c9a189

Signed-off-by: He Huang (Steve) <105218074+stevehuang52@users.noreply.github.com>

fix dependency for nemo26.02

03d4ad2

Signed-off-by: stevehuang52 <heh@nvidia.com>

clean up

480d576

Signed-off-by: stevehuang52 <heh@nvidia.com>

upgrade min lhotse version

692b188

Signed-off-by: stevehuang52 <heh@nvidia.com>

stevehuang52 added 8 commits February 10, 2026 11:26

update

c4c93c5

Signed-off-by: stevehuang52 <heh@nvidia.com>

Merge remote-tracking branch 'origin/heh/va_nemo26.02_fix' into heh/v…

d99da54

…a_nano3 Signed-off-by: stevehuang52 <heh@nvidia.com>

fix env yaml

27d0b5e

Signed-off-by: stevehuang52 <heh@nvidia.com>

replace with pinned version

0fa8a5e

Signed-off-by: stevehuang52 <heh@nvidia.com>

fix en vname

ffb55ed

Signed-off-by: stevehuang52 <heh@nvidia.com>

update weather function

b6aff89

Signed-off-by: stevehuang52 <heh@nvidia.com>

fix nano-v2 vllm config

97fe6a7

Signed-off-by: stevehuang52 <heh@nvidia.com>

fix env yaml

3525f32

Signed-off-by: stevehuang52 <heh@nvidia.com>

tango4j requested changes Feb 11, 2026

View reviewed changes

stevehuang52 and others added 4 commits February 13, 2026 17:44

Merge branch 'main' into heh/va_nano3

e07cc9b

Signed-off-by: He Huang (Steve) <105218074+stevehuang52@users.noreply.github.com>

address reviews

89638fb

Signed-off-by: He Huang <heh@nvidia.com>

Merge remote-tracking branch 'origin/heh/va_nano3' into heh/va_nano3

acf0bcf

Apply isort and black reformatting

1f64ccb

Signed-off-by: stevehuang52 <stevehuang52@users.noreply.github.com>

stevehuang52 force-pushed the heh/va_nano3 branch from c950248 to 1f64ccb Compare February 13, 2026 23:17

tango4j added Run CICD and removed Run CICD labels Feb 13, 2026

github-actions Bot removed the Run CICD label Feb 13, 2026

tango4j approved these changes Feb 13, 2026

View reviewed changes

stevehuang52 merged commit 0ae9f88 into main Feb 14, 2026
68 of 74 checks passed

stevehuang52 deleted the heh/va_nano3 branch February 14, 2026 01:56

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

[voice agent] Add support for nemotron-nano-v3, magpie-tts and misc improvements#15318

[voice agent] Add support for nemotron-nano-v3, magpie-tts and misc improvements#15318
stevehuang52 merged 39 commits intomainfrom
heh/va_nano3

stevehuang52 commented Jan 26, 2026

Uh oh!

Uh oh!

Uh oh!

tango4j left a comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

github-actions Bot commented Feb 13, 2026

Uh oh!

tango4j left a comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Conversation

stevehuang52 commented Jan 26, 2026

What does this PR do ?

Changelog

Usage

Uh oh!

Uh oh!

Uh oh!

tango4j left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

github-actions Bot commented Feb 13, 2026

Uh oh!

tango4j left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants