Commit graph

66 commits

Author SHA1 Message Date
Maicon Moreira
7c04ee85a7
feat: add “mute while recording” setting (#257)
* feat: mute while recording

* refactor: move MuteWhileRecording from AdvancedSettings to DebugSettings
2025-10-28 17:49:33 -07:00
CJ Pais
65c1e2dbcb
parakeet v2 support (#116)
* parakeet v2 support

* Update model.rs

* fix parakeet size

* new onboarding UI for parakeet v2 introduction

* tweak height
2025-10-28 16:32:26 -07:00
CJ Pais
a9a3e7bfd8
display download size + refactor (#221) 2025-10-17 16:57:51 -07:00
Jackson
93215dc097
A few optimizations on startup (#182)
* Use an optimal release profile... load managers & hotkeys first, add a
splash, simplify app handler

* removed splashscreen, hopefully app boots fast enough anyhow

* No need to load transcription model on start now that its loaded on
demand & cleanup

* Make an unloaded model status

* tiny modification to unloading

* update comment

---------

Co-authored-by: CJ Pais <cj@cjpais.com>
2025-10-17 16:05:43 -07:00
Jackson
83970b0f44
Load model in parallel (#181) 2025-10-08 16:44:55 -07:00
CJ Pais
18f32c7c8d simplify history slightly 2025-10-08 16:28:01 -07:00
becker
d0e6a17c17
Add history limit setting (#150)
* Add history limit setting

Dearest Reviewer,

I have added a history limit setting. The limit change is applied instantly.
If the limit is 0 then no recordings are saved.

Becker

* remove extra save

save happens on exit

* move to debug menu

* use the text color

---------

Co-authored-by: CJ Pais <cj@cjpais.com>
2025-10-08 13:01:09 -07:00
segudev
03289f84d7
fix: restore translate-to-english functionality (#173)
* fix: restore translate-to-english functionality

The translate option broke when migrating from whisper-rs to transcribe-rs.
This fix patches transcribe-rs to add translate parameter support and passes
the translate_to_english setting to the Whisper engine.

Additionally fixed AccessibilityPermissions API calls to match updated library.

Translation only works with Whisper multilingual models (not .en models).
Parakeet models don't support translation as they are transcription-only.

Fixes #152

* bump transcribe rs

---------

Co-authored-by: CJ Pais <cj@cjpais.com>
2025-10-08 11:57:47 -07:00
Han Pham
9185b3db08
feat: delete a single audio entry (#119)
* feat: delete a single audio entry

* fix(HistorySetting): remove unnecessary loading state
2025-09-19 12:30:50 -07:00
CJ Pais
338d15d2c4
tweak onboarding (#112) 2025-09-09 18:59:09 -07:00
CJ Pais
45a473dc9f
Parakeet V3 (#111)
* wip: model manager support for parakeet

* working parakeet??git status

* Update LanguageSelector.tsx

* Use the published version of the crate.

* move word correction into audio toolkit
2025-09-09 18:33:57 -07:00
CJ Pais
cdfe15c898
tweak sampling strat for whisper to hopefully decrease hallucination (#109) 2025-09-08 15:43:07 -07:00
CJ Pais
7638f6b827
basic working history (#104) 2025-09-05 16:25:38 -07:00
CJ Pais
ae5f640da5
Add VRAM Inactivity Timer (#101)
* init vram unload

* logging

* add immediate unloading

* thread fixes
2025-09-03 19:00:36 -07:00
CJ Pais
308e001661 downgrade whisper-rs to 0.13.2 to continue to support older versions of
macos
2025-08-19 16:12:13 -07:00
CJ Pais
8bbcfc65fe
Merge branch 'main' into upgrade-whisper 2025-08-18 15:58:06 -07:00
CJ Pais
934edaa9e1 Maybe slightly better, no speech detection. 2025-08-18 13:54:55 -07:00
CJ Pais
bf35d88c03 upgrade whisper-rs (broken on macos) 2025-08-15 19:53:07 -07:00
CJ Pais
04b006d5b8 new default correction threshold + debug menu 2025-08-15 11:55:15 -07:00
CJ Pais
c2b3815f56 'custom words' everywhere instead of 'correct words' 2025-08-11 15:19:48 -07:00
Vlad Gerasimov
6404ce7487 refactor(transcription): optimize word correction performance 2025-08-08 19:25:21 +04:00
Vlad Gerasimov
003bf23402 feat(transcription): add word correction with phonetic matching 2025-08-08 17:50:31 +04:00
CJ Pais
86a1205619 add partial/resumable download capability 2025-08-02 18:25:30 -07:00
CJ Pais
6ce848dc66
Merge pull request #59 from cjpais/recording-pill
Add Recording Overlay
2025-08-01 15:35:11 -07:00
CJ Pais
a7c5b199fb
Merge branch 'main' into cancel-actions-in-menu 2025-08-01 15:25:57 -07:00
CJ Pais
dd957a894d add setting to enable or disable showing the overlay 2025-08-01 14:29:02 -07:00
CJ Pais
ec2fefb9fe Audio visualizer generally working nicely. 2025-08-01 14:15:02 -07:00
CJ Pais
51105dde15 more wip but roughly is there 2025-08-01 14:15:02 -07:00
CJ Pais
d5f23cdfa5 Try not to focus it first. It doesn't quite work though. 2025-08-01 14:15:02 -07:00
CJ Pais
8731881837 wip 2025-08-01 14:15:02 -07:00
Václav Škorpil
b767eca6fb Add language select to settings 2025-07-28 19:52:37 +02:00
Vlad Gerasimov
6c26d6afb8 feat(tray): add cancel recording/transcribing actions 2025-07-27 22:46:59 +04:00
CJ Pais
21e96083df add ui toggle back 2025-07-11 18:19:26 -07:00
segudev
c3aa2d7c1b Add "Translate to English" setting: Introduced a new setting to enable automatic translation of speech to English. Updated relevant components, settings modules, and transcription logic to support this feature. Adjusted accessibility permission calls for consistency. 2025-07-11 18:19:15 -07:00
CJ Pais
dc90edadb1 Add hysteresis to VAD. 2025-07-11 15:40:47 -07:00
CJ Pais
c23befd98a big refactor of settings + switch audio devices 2025-07-10 21:10:46 -07:00
CJ Pais
59541b8ec4 move audio toolkit into src-tauri to fix permissions 2025-07-10 12:38:06 -07:00
CJ Pais
a60cf57874 use new backend 2025-07-10 11:59:05 -07:00
CJ Pais
78ec8e8918 use unquantized version of whisper turbo, and change whisper medium
quant to 4_1
2025-07-03 14:14:07 -07:00
CJ Pais
0805442bf2 fix not loading model on start if available 2025-07-01 14:57:02 -07:00
CJ Pais
a479ae3625 make vad less permissive 2025-06-30 15:58:13 -07:00
CJ Pais
6ad9534ab7 activate the microphone on demand 2025-06-30 15:18:59 -07:00
CJ Pais
2ab3f2ce98 tweak language 2025-06-30 14:43:34 -07:00
CJ Pais
9036c05088 initial working version of multiple models + cleanup and onboarding 2025-06-30 14:27:21 -07:00
CJ Pais
428e515129 Set language to auto. 2025-06-17 20:56:53 -07:00
CJ Pais
6e8128defd remove some debug logging, plus some general minor cleanup 2025-05-16 09:06:46 -07:00
CJ Pais
4bab3f11a9 fix short words being thrown out 2025-05-15 15:48:22 -07:00
CJ Pais
821e12bdf6 handle app close uniformly xplatform 2025-05-15 13:20:54 -07:00
CJ Pais
e307da8544 fix bugs on linux (mix two channels down to 1, build with vulkan) 2025-05-15 13:00:16 -07:00
CJ Pais
113edbab5c theoretically have different sample formats working as expected. 2025-05-13 13:31:51 -07:00