Quentin Fuxa
commited on
Commit
·
519ccaa
1
Parent(s):
38f752b
huggingface steps added. Solve https://github.com/QuentinFuxa/whisper_streaming_web/issues/59 & https://github.com/QuentinFuxa/whisper_streaming_web/issues/60
Browse files
README.md
CHANGED
|
@@ -25,7 +25,7 @@ This project is based on [Whisper Streaming](https://github.com/ufal/whisper_str
|
|
| 25 |
|
| 26 |
#### 🚀 **Coming Soon**
|
| 27 |
|
| 28 |
-
- **Faster Word Validation** – Accelerate real-time transcription by validating high-confidence words immediately upon first appearance for whisper backends that
|
| 29 |
- **Enhanced Diarization Performance** – Optimize speaker identification by implementing longer steps for Diart processing and leveraging language-specific segmentation patterns to improve speaker boundary detection
|
| 30 |
|
| 31 |
|
|
@@ -78,6 +78,8 @@ This project is based on [Whisper Streaming](https://github.com/ufal/whisper_str
|
|
| 78 |
diart
|
| 79 |
```
|
| 80 |
|
|
|
|
|
|
|
| 81 |
|
| 82 |
3. **Run the FastAPI Server**:
|
| 83 |
|
|
|
|
| 25 |
|
| 26 |
#### 🚀 **Coming Soon**
|
| 27 |
|
| 28 |
+
- **Faster Word Validation** – Accelerate real-time transcription by validating high-confidence words immediately upon first appearance for whisper backends that return word & segment probabilities
|
| 29 |
- **Enhanced Diarization Performance** – Optimize speaker identification by implementing longer steps for Diart processing and leveraging language-specific segmentation patterns to improve speaker boundary detection
|
| 30 |
|
| 31 |
|
|
|
|
| 78 |
diart
|
| 79 |
```
|
| 80 |
|
| 81 |
+
Diart uses by default [pyannote.audio](https://github.com/pyannote/pyannote-audio) models from the _huggingface hub_. To use them, please follow the steps described [here](https://github.com/juanmc2005/diart?tab=readme-ov-file#get-access-to--pyannote-models).
|
| 82 |
+
|
| 83 |
|
| 84 |
3. **Run the FastAPI Server**:
|
| 85 |
|