Spaces:
Sleeping
Sleeping
Commit History
-
tests: add CURL test
baf9ef9 -
chore: update gitignore
c243eda -
fix: pass original voice name to generate_speech instead of display name
Update voice dropdown choices to use tuples of (display_name, original_name) so the function receives the correct voice identifier while showing user-friendly labels in the UI.b0728cf
-
feat: Add voice notes
08036d2 -
Change the word “demo” to public preview.
15874d6
-
feat: Add new Common Voice voices
501d7ce -
Remove vocies: Fabio, Ruben
- Fabio sounds half Brazilian. A lot of words are pronounced Brazilian Portuguese, not European. - Ruben: Has Brazilian Portuguese more than European.. wave one gets slower and slower as it plays. Wave 2 is leaning more towards Latin Spanish pronunciation too. There is also a robotic/metallic undertone to this voice.282a489 -
feat: add more voices
ed15dd4 -
UI: update explanatory sections
5042682
-
feat: add server readiness check with dynamic UI updates
- Add check_readiness function to poll /health endpoint every 1 second - Disable submit button by default until server and models are ready - Update button text dynamically based on readiness state5f4dfb8 -
UI: Update
8077b71
-
UI: update explanatory sections
removed the coming soon part from API Access and Fine-tuning sections Translated into pt-PT and added extra English version in smaller font1e3ef59 -
UI: update UI and add space info
e280b3f -
chore: update requirements
11abd56
-
fix: Keep only one voice, Andre
4064c42 -
fix: Keep only one voice, Andre
cf4b6ae -
fix: Remove not used fine_tuning_version parameter
57346d0
-
chore(.gitignore): update
e9f6fe9 -
fix: Update voice_id_2_name import to be from deployment_options
6fa9379 -
feat: Rename config file for API deployment options
994fd9a -
fix: Remove type="filepath" from Audio component - that's only for inputs
c1657c4 -
fix: save audio to regular file instead of tempfile for Gradio
- Write WAV output to regular file in current directory - Gradio security restricts tempfile access causing upload errors4f4b157 -
fix: save audio to temporary file for Gradio compatibility
- Write WAV bytes to temporary file instead of returning raw bytes - Resolves "Upload failed: unsupported file type" error in Gradio - Use tempfile.NamedTemporaryFile with .wav suffix for proper format detection9fe6994
-
fix: wait until the serer generates the speech
39a7318 -
fix: update voice_id_map initialization to populate from voice_id_2_name
a2c9da9 -
UI: remove the word 'Public' from 'Public Demo'
5cf98d0 -
feat: update voice_id_map
4fa020c
-
feat: connect Gradio UI to private TTS endpoint
- Add endpoint authentication using endpoint_READ token from env - Implement streaming POST request to /v1/tts/{voice_id} - Return raw WAV bytes directly to Gradio audio component - Map voice names to voice IDs (André -> id59) - Add proper error handling for missing token and request failuresd330639 -
fix: remove deprecated show_api kwarg from Blocks.launch()
Parameter was removed in Gradio 6.xda8df4a -
Add simple public Gradio demo calling private /test-numpy endpoint
3013bc4 -
initial commit
Verified signature032540c
