enne2/stt-server
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
4 Commits 1 Branch 0 Tags
master
T
Clone
Open with VS Code Open with VSCodium Open with Intellij IDEA
Download ZIP Download TAR.GZ Download BUNDLE
Matteo Benedetto 536380e793 Trim silenzio iniziale/finale (VAD energia RMS, numpy+soundfile): meno audio da processare, meno allucinazioni; timestamp riallineati con offset; TRIM_SILENCE=0 per disattivare, TRIM_THRESHOLD_DB/TRIM_PADDING_MS configurabili; flag --no-trim
2026-08-21 15:12:27 +02:00
.gitignore
stt-server: FastAPI whisper.cpp Vulkan ASR + pyannote diarization, containerizzato
2026-08-13 22:29:30 +02:00
docker-compose.yml
initial_prompt personale + glossario: POST /transcribe accetta prompt e glossary (normalizzazione alias->canonical post-ASR, raw_text/segments[].raw), GET/PUT /glossary con GLOSSARY_FILE, glossary.example.json, --prompt/--carry-initial-prompt in transcribe.py
2026-08-21 12:55:53 +02:00
Dockerfile
initial_prompt personale + glossario: POST /transcribe accetta prompt e glossary (normalizzazione alias->canonical post-ASR, raw_text/segments[].raw), GET/PUT /glossary con GLOSSARY_FILE, glossary.example.json, --prompt/--carry-initial-prompt in transcribe.py
2026-08-21 12:55:53 +02:00
entrypoint.sh
stt-server: FastAPI whisper.cpp Vulkan ASR + pyannote diarization, containerizzato
2026-08-13 22:29:30 +02:00
glossary.example.json
initial_prompt personale + glossario: POST /transcribe accetta prompt e glossary (normalizzazione alias->canonical post-ASR, raw_text/segments[].raw), GET/PUT /glossary con GLOSSARY_FILE, glossary.example.json, --prompt/--carry-initial-prompt in transcribe.py
2026-08-21 12:55:53 +02:00
requirements.txt
stt-server: FastAPI whisper.cpp Vulkan ASR + pyannote diarization, containerizzato
2026-08-13 22:29:30 +02:00
server.py
initial_prompt personale + glossario: POST /transcribe accetta prompt e glossary (normalizzazione alias->canonical post-ASR, raw_text/segments[].raw), GET/PUT /glossary con GLOSSARY_FILE, glossary.example.json, --prompt/--carry-initial-prompt in transcribe.py
2026-08-21 12:55:53 +02:00
transcribe.py
Trim silenzio iniziale/finale (VAD energia RMS, numpy+soundfile): meno audio da processare, meno allucinazioni; timestamp riallineati con offset; TRIM_SILENCE=0 per disattivare, TRIM_THRESHOLD_DB/TRIM_PADDING_MS configurabili; flag --no-trim
2026-08-21 15:12:27 +02:00
S
Description
STT server: whisper.cpp Vulkan ASR + pyannote Community-1 diarization (FastAPI)
48 KiB
Languages
Python 83.7%
Dockerfile 15%
Shell 1.3%