Update README.md
Browse files
README.md
CHANGED
|
@@ -64,6 +64,23 @@ result = pipe("audio.wav",
|
|
| 64 |
print(result["text"])
|
| 65 |
```
|
| 66 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 67 |
## Training and evaluation data
|
| 68 |
|
| 69 |
- `training` = google/fleurs['train+validation'] + openslr['train']
|
|
|
|
| 64 |
print(result["text"])
|
| 65 |
```
|
| 66 |
|
| 67 |
+
|
| 68 |
+
## whisper.cpp
|
| 69 |
+
|
| 70 |
+
|
| 71 |
+
### 1. Transcode the input audio to 16kHz PCM
|
| 72 |
+
|
| 73 |
+
```shell
|
| 74 |
+
ffmpeg -i audio.ogg -ar 16000 -ac 1 -c:a pcm_s16le output.wav
|
| 75 |
+
```
|
| 76 |
+
|
| 77 |
+
### 2. Transcribe with whisper.cpp
|
| 78 |
+
|
| 79 |
+
```shell
|
| 80 |
+
./main -m ggml-model.bin -f output.wav --print-colors --language km
|
| 81 |
+
```
|
| 82 |
+
|
| 83 |
+
|
| 84 |
## Training and evaluation data
|
| 85 |
|
| 86 |
- `training` = google/fleurs['train+validation'] + openslr['train']
|