Commit Graph
64 Commits
Author SHA1 Message Date
Timothy Jaeryang Baek 5d5b42d3f5 refac/enh: openai tts additional params support 2025-10-02 03:52:29 -05:00
Timothy Jaeryang Baek 2e75c6dbdf refac/fix: azure audio escape 2025-10-01 23:01:08 -05:00
Timothy Jaeryang Baek c96252f7fe refac/fix: WHISPER_LANGUAGE 2025-09-17 10:46:11 -05:00
Timothy Jaeryang Baek 8339f59cdf fix: openai audio revert 2025-09-10 14:02:19 +04:00
Timothy Jaeryang Baek 485392fe63 chore: format 2025-09-09 18:19:31 +04:00
Adam 39ae9167ee removed test code. 2025-08-30 20:31:13 -04:00
Adam 20b6902b9f whitespace 2025-08-30 20:30:34 -04:00
Adam 08b958cfc9 re-add used var 2025-08-30 20:29:16 -04:00
Adam c62f30e22c remove whitespace 2025-08-30 20:12:46 -04:00
Adam 562710fe33 join the url instead of concatenating a string in case the user adds a slash to the end of their configured url. 2025-08-30 20:05:53 -04:00
Timothy Jaeryang Baek f23eb2a31c refac: audio lang fallback logic 2025-08-17 04:33:42 +04:00
_00_ ab8e54dd29 Update audio.py Fix Format error
Fix Format error
2025-07-23 08:49:00 +02:00
_00_ 51758c429b Update audio.py - fix FORMAT error
Fix  FORMAT error
2025-07-23 08:45:10 +02:00
_00_ 0613563644 FIX: STT default whisper trascription language
FIX: STT default whisper trascription language

Fix the transcripcion language used by default whisper, setting as WHISPER_LANGUAGE if it is setted in env var, even if a language is detected in the file's metadata.
It is understood that if a language is set as an environment variable for transcriptions, this should be the preferred one and the one that should be used for that purpose.

It would be advisable to add this variable as configurable in UI
2025-07-22 16:47:06 +02:00
Timothy Jaeryang Baek 19f1286cc7 fix: emoji call 2025-07-16 15:38:48 +04:00
Athanasios Oikonomou 96758176cc fix: don't over quote forwarded headers
Fix introduced on #15035 is over quoting headers.

Eg mails instead of user@example.com shown as user%40example.com
Eg names instead of First Last shown as First%20Last

Also we are spending some time quoting ids and roles without required.

Keep quote only on user name, initially had problem based on the discussion
https://github.com/open-webui/open-webui/discussions/14391

Also add space in safe characters, in order remove %20 from names.
2025-07-10 22:08:28 +03:00
Timothy Jaeryang Baek 0a1f9966ef refac: audio error handling 2025-07-06 14:20:38 +04:00
Timothy Jaeryang Baek 6186bbf337 refac/fix: stt supported type 2025-06-18 14:01:14 +04:00
priten f7920df870 Fix non-ascii error issue on ENABLE_FORWARD_USER_INFO_HEADERS 2025-06-16 12:33:11 -05:00
Timothy Jaeryang Baek 72df23ed79 refac 2025-06-16 17:24:55 +04:00
Timothy Jaeryang BaekandBryan Berns 7a1afa9c66 feat: custom stt content type
Co-Authored-By: Bryan Berns <berns@uwalumni.com>
2025-06-16 16:13:40 +04:00
Timothy Jaeryang Baek 8258dfb5af enh: enable deepgram smart_format 2025-06-16 12:34:01 +04:00
Timothy Jaeryang Baek 036ce12dd9 doc: changelog 2025-05-30 01:14:38 +04:00
Romain Dauby b12a493fe5 fix: only trust codec_name for audio conversion
Some files have .wav extension with incompatible OpenAI codec
2025-05-29 16:57:23 -04:00
Timothy Jaeryang Baek baaa285534 feat: user stt language 2025-05-24 00:36:30 +04:00
Timothy Jaeryang Baek 73e64fe7fb refac: audio upload handling 2025-05-19 02:52:48 +04:00
Timothy Jaeryang Baek b280f828b0 enh: very long audio transcription 2025-05-17 02:51:28 +04:00
Timothy Jaeryang Baek b143c71da2 refac: AIOHTTP_CLIENT_SESSION_SSL 2025-05-14 23:33:52 +04:00
Timothy Jaeryang Baek 549989e9ec refac 2025-05-10 19:04:40 +04:00
Timothy Jaeryang Baek 827326e1a2 refac: audio transcription issue 2025-05-08 22:57:48 +04:00
Timothy Jaeryang Baek bfa5550cc3 refac: openai already supports webm audio 2025-05-08 22:44:32 +04:00
Tim Jaeryang Baek 2a4dfc02a2 Merge pull request #13540 from NoMoreFood/dev
feat: Azure TTS Allow Base URL
2025-05-07 00:49:57 +04:00
Bryan Berns 5aabe21cbe Add Custom Azure TTS URL 2025-05-05 22:08:48 -04:00
Timothy Jaeryang Baek 7b36466c1c refac: audio transcribe supported filetype 2025-05-05 23:42:56 +04:00
Timothy Jaeryang Baek 4cfb99248d chore: format 2025-05-03 23:48:24 +04:00
Tim Jaeryang Baek 7b014e44ee Merge pull request #13376 from Thaniel94/add-whisper-language-constraint
feat: Added WHISPER_LANGUAGE env variable
2025-05-02 03:08:00 -07:00
nathaniel ef7acfbf3d WHISPER_LANGUAGE no longer a "PersistentConfig" variable (Was not appropriate with how WHISPER_LANGUAGE is currently configured). 2025-05-01 21:33:57 +01:00
Bryan Berns 6c8a9d000e Azure STT Allow Base URL & Max Speaker Setting 2025-04-30 08:51:01 -04:00
nathaniel 1efa708f83 Added WHISPER_LANGUAGE env variable. If set to a country's ISO2, constrains Whisper's stt to that language. Detects language as normal if unset 2025-04-27 05:58:06 +01:00
Timothy Jaeryang Baek e7332fd6fe refac 2025-04-13 23:39:38 -07:00
Tom 24367d459b Enable vad_filter to improve quality of transcription in faster-whisper model. 2025-04-13 13:03:57 +01:00
Timothy Jaeryang Baek bde89fd29e refac: audio 2025-04-12 18:40:09 -07:00
Timothy Jaeryang Baek 91a455a284 chore: format 2025-04-12 16:35:11 -07:00
Tim Jaeryang Baek 36ac81b229 Merge pull request #12727 from decent-engineer-decent-datascientist/main
feat: add Azure AI Speech STT provider
2025-04-10 16:50:40 -07:00
priten 9a55257c5b feat: add Azure AI Speech STT provider
- Add Azure STT configuration variables for API key, region and locales
- Implement Azure STT transcription endpoint with 200MB file size limit
- Update Audio settings UI to include Azure STT configuration fields
- Handle Azure API responses and error cases consistently
2025-04-10 15:38:59 -05:00
Timothy Jaeryang Baek 05aa9c6d9c refac 2025-04-10 12:27:11 -07:00
Thomas Rehn 4731e0d0e3 fix: convert webm to wav for OpenAI transcription endpoint 2025-04-10 09:00:51 +02:00
Thomas Rehn d99a883867 fix: convert ogg to wav for OpenAI transcription endpoint 2025-04-08 15:04:04 +02:00
Hermógenes Oliveira e936d7b53d fix: audio api endpoint filetype check
RFC2046 allows the Content-Type field to have additional parameters
after the main type/subtype information (Section 1).

Following RFC4281, many applications put codec information inside
parameters in the Content-Type. This is especially common for formats
that support many codecs, such as Ogg (RFC5334, Section 4).

The `/api/audio/transcriptions` endpoint is currently rejecting files
that contain parameters in the Content-Type field with a bad request
error.

This commit changes the current check in order to accept any
Content-Type field that begins with a supported type/subtype as listed
in the `supported_filetypes` tuple.

Since Content-Type here is provided by the user, I believe this check
is meant to prevent honest mistakes, like posting a PDF to an audio
processing endpoint, not as a security measure against possibly
malicious use. Therefore, I think it's OK not to validate the rest of
the field.
2025-03-08 18:03:30 -03:00
tidely b15814c42f chore: remove unnecessary Path conversions
Remove unnecessary `pathlib.Path` conversions. (CACHE_DIR and DATA_DIR)

Use `/` Path joining shorthand to ensure using platform specific Path separators (Windows: \\, Unix: /)
2025-03-04 19:53:52 +02:00