Commit Graph

30 Commits

Author SHA1 Message Date
Filipi Fuchter
1330ef3ad6 Enhanced error handling across the framework.
Co-authored-by: Mark Backman <m.backman@gmail.com>
2025-11-26 18:34:25 -03:00
Hwuiwon Kim
1cc69d475d feat: Add speaking rate control to Inworld TTS service & fix param cases 2025-11-19 22:57:53 -05:00
Paul Kompfner
f3b254e335 D'oh! My TTS "inter-frame-spaces" logic was *way* overcomplicated (and fundamentally mistaken, though it happened to work)
Now:
- For TTS word-by-word output and `TTSSpeakFrames`: `TTSTextFrame`s' have `includes_inter_frame_spaces=False`.
- For all other TTS output: `TTSTextFrame` pass through the received text frames' `includes_inter_frame_spaces` value. So far, this value has always been `True`: LLMs send text chunks already containing all necessary spaces.
- `LLMTextFrame`s set `includes_inter_frame_spaces=False` at init time, per the aforementioned assumption.
2025-11-17 10:14:28 -05:00
Angad Singh
d1116d149e feat: Add ErrorFrame emission to TTS/STT services for pipeline error detection (#2881)
* feat: Add ErrorFrame emission to TTS/STT services for pipeline error detection

- Add ErrorFrame emission to all major TTS/STT services during initialization and runtime failures
- Services updated: Cartesia, ElevenLabs, Deepgram, AssemblyAI, Rime, Azure
- ErrorFrame objects emitted with fatal=False for graceful degradation
- Enables on_pipeline_error event handler to detect service failures programmatically
- Add comprehensive pytest test suite to verify ErrorFrame emission
- Fixes issue where services failed gracefully but didn't emit ErrorFrame objects

This allows developers to implement real-time error monitoring and alerting
using the on_pipeline_error event handler introduced in v0.0.90.

* Update STT and TTS services to use consistent error handling pattern

- Improves error handling consistency across all services

* Add changelog entry for STT/TTS error handling improvements

* Linting issues Resolved

* Azure STT ErrorFrames added with consistent patterns

* Cartesia STT and Deepgram STT; additional fixes made

* Removed Fatal Flags across services, removed duplication

* Moving the changelog entry to the correct place.

* Refactoring some classes to use yield instead of push_error directly.

* Fixing ruff format.

---------

Co-authored-by: Filipi Fuchter <filipi87@gmail.com>
2025-11-14 15:03:05 -03:00
Paul Kompfner
5222ff99de Apply includes_inter_frame_spaces = True in all LLM and TTS services that need it.
Note that for `LLMTextFrame`s, the right behavior is pretty much always `includes_inter_frame_spaces = True`. I decided *not* to go ahead and make that the default for `LLMTextFrame`s, though, simply to not introduce a subtle behavior change for creative/unexpected use-cases that were relying on text in hand-crafted `LLMTextFrame`s being handled a certain way. Ditto for `TTSTextFrame`s.

Also, fix an issue in `NeuphonicTTSService` where it wasn't pushing `TTSTextFrame`s.

Also, fix the broken `SarvamHttpTTSService` example.

Also, add a couple of missing examples.
2025-11-12 15:10:11 -05:00
Filipi Fuchter
9e098b5f79 Changing InworldTTSService default temperature to 1.1 2025-09-19 17:03:53 -03:00
padillamt
e140bd6960 mtpadilla: moved model and voice id setting into the class constructor 2025-07-25 14:04:49 -07:00
padillamt
4a9bec5b35 mtpadilla: stop metrics at result chunk 2025-07-25 11:14:20 -07:00
padillamt
37361391d9 mtpadilla: removed ability to set base_url via constructor, set internally based on streaming variable 2025-07-25 09:16:56 -07:00
padillamt
acc5b9f210 inworld: change to function that stops all processing metrics
Co-authored-by: Mark Backman <m.backman@gmail.com>
2025-07-24 22:07:15 -07:00
padillamt
5fb1899aeb inworld: removal of unnecessary default assignment as already handled 2025-07-24 21:42:42 -07:00
padillamt
7483422bd9 inworld: change set_voice uto use self._settings
Co-authored-by: Mark Backman <m.backman@gmail.com>
2025-07-24 21:23:03 -07:00
padillamt
16c20f3a99 inworld: removal of unnecessary default assignment since already done
Co-authored-by: Mark Backman <m.backman@gmail.com>
2025-07-24 21:15:34 -07:00
padillamt
d248c102c8 inworld: removal of unnecessary default assignment since already done
Co-authored-by: Mark Backman <m.backman@gmail.com>
2025-07-24 21:15:20 -07:00
padillamt
662550cc5e mtpadilla: remove unused imports 2025-07-24 21:05:22 -07:00
padillamt
f6440ee6e1 mtpadilla: correct Examples header in comments 2025-07-24 13:36:40 -07:00
padillamt
b6367965cb mtpadilla: consolidate streaming and non-streaming options into a single class with common API, with boolean switch variable added (streaming) 2025-07-23 16:50:32 -07:00
padillamt
147bf9cfe8 mtpadilla: addition of non-streaming option with own dedicated class, and related additional non-streaming test option 2025-07-23 15:28:43 -07:00
padillamt
a5d353030e mtpadilla: small formatting fix to comments 2025-07-23 12:02:58 -07:00
padillamt
f29024bcc0 mtpadilla: update coments regarding temperature parameter 2025-07-23 11:47:26 -07:00
padillamt
1915407ff7 inworld: removed unreferenced is_first_chunk variable 2025-07-21 15:30:48 -07:00
padillamt
076a675a75 inworld: Fix...Set sample_rate=None in InworldHttpTTSService to match Cartesia pattern 2025-07-21 13:50:36 -07:00
padillamt
8eda2435a2 inworld: removed explicit references to language since our models currently infer that from the text. 2025-07-21 13:24:10 -07:00
padillamt
54ff946976 inworld: largely adjustments for docstring compatibility 2025-07-21 12:07:58 -07:00
padillamt
e3711f96a3 inworld: added detailed comments 2025-07-20 17:06:35 -07:00
padillamt
1bc442e329 inworld: docstring fix 2025-07-18 15:13:19 -07:00
padillamt
2b76823b01 inworld: added comments to track a few things to confirm 2025-07-15 18:17:30 -07:00
padillamt
913dba3b74 inworld: class name change 2025-07-15 17:15:57 -07:00
padillamt
384838147a inworld: removed unnecessary code from stop() and cancel() 2025-07-15 16:56:18 -07:00
padillamt
7861b911c0 inworld: first commit of __init__ and tts.py files 2025-07-15 16:50:50 -07:00