Skip to content

wip: advanced preprocessing [english] #3927

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Draft
wants to merge 5 commits into
base: main
Choose a base branch
from
+1,081 −7
Draft

wip: advanced preprocessing [english] #3927

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
169c958
WIP
Hormold Nov 14, 2025
c817739
remove unnecessary filter
Hormold Nov 14, 2025
6278167
update filters and tests
Hormold Nov 14, 2025
2b1a7ff
fix linter issue
Hormold Nov 14, 2025
04d0154
fix
Hormold Nov 14, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
26 changes: 23 additions & 3 deletions livekit-agents/livekit/agents/voice/agent_session.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -144,7 +144,22 @@ def __call__(self, frame: rtc.VideoFrame, session: AgentSession) -> bool:
return False


DEFAULT_TTS_TEXT_TRANSFORMS: list[TextTransforms] = ["filter_markdown", "filter_emoji"]
DEFAULT_TTS_TEXT_TRANSFORMS: list[TextTransforms] = [
"filter_markdown",
"filter_emoji",
"remove_angle_bracket_content",
"replace_newlines_with_periods",
"format_dates",
"format_times",
"format_emails",
"format_phone_numbers",
"format_acronyms",
"format_dollar_amounts",
"format_distances",
"format_units",
"format_percentages",
"format_numbers", # Must be LAST to not interfere with other formatters
]


class AgentSession(rtc.EventEmitter[EventTypes], Generic[Userdata_T]):
Expand Down Expand Up @@ -250,8 +265,13 @@ def __init__(
if ``TTS.capabilities.aligned_transcript`` is ``True`` or ``streaming``
is ``False``. When NOT_GIVEN, it's disabled.
tts_text_transforms (Sequence[TextTransforms], optional): The transforms to apply
to the tts input text, available built-in transforms: ``"filter_markdown"``, ``"filter_emoji"``.
Set to ``None`` to disable. When NOT_GIVEN, all filters will be applied.
to the tts input text. Available built-in transforms include: ``"filter_markdown"``,
``"filter_emoji"``, ``"format_numbers"``, ``"format_dates"``, ``"format_times"``,
``"format_emails"``, ``"format_phone_numbers"``, ``"format_acronyms"``,
``"format_dollar_amounts"``, ``"format_distances"``, ``"format_units"``,
``"format_percentages"``, ``"remove_angle_bracket_content"``,
``"replace_newlines_with_periods"``.
Set to ``None`` to disable. When NOT_GIVEN, a default set of filters will be applied.
preemptive_generation (bool):
Whether to speculatively begin LLM and TTS requests before an end-of-turn is
detected. When True, the agent sends inference calls as soon as a user
Expand Down
Loading