Create README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,40 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
language:
|
| 4 |
+
- cy
|
| 5 |
+
- en
|
| 6 |
+
base_model:
|
| 7 |
+
- techiaith/whisper-large-v3-ft-verbatim-cy-en
|
| 8 |
+
pipeline_tag: automatic-speech-recognition
|
| 9 |
+
tags:
|
| 10 |
+
- faster-whisper
|
| 11 |
+
---
|
| 12 |
+
|
| 13 |
+
**Model Name:** ExampleTransformer-CT2
|
| 14 |
+
|
| 15 |
+
**Model Description:**
|
| 16 |
+
|
| 17 |
+
This model card describes `whisper-large-v3-ft-verbatim-cy-en-ct2`, a conversion of the `techiaith/whisper-large-v3-ft-verbatim-cy-en`
|
| 18 |
+
fine-tuned OpenAI whisper model to the CTranslate2 format. This conversion allows for significantly faster and more efficient
|
| 19 |
+
inference, particularly on CPU and with batching.
|
| 20 |
+
|
| 21 |
+
|
| 22 |
+
**How to Use:**
|
| 23 |
+
|
| 24 |
+
```python
|
| 25 |
+
from faster_whisper import WhisperModel
|
| 26 |
+
|
| 27 |
+
audio_file_path=<path to your audio file>
|
| 28 |
+
|
| 29 |
+
model = WhisperModel("techiaith/whisper-large-v3-ft-verbatim-cy-en-ct2")
|
| 30 |
+
segments, info = model.transcribe(audio_file_path, beam_size=5)
|
| 31 |
+
|
| 32 |
+
print("Detected language '%s' with probability %f" % (info.language, info.language_probability))
|
| 33 |
+
for segment in segments:
|
| 34 |
+
print("[%.2fs -> %.2fs] %s" % (segment.start, segment.end, segment.text))
|
| 35 |
+
```
|
| 36 |
+
|
| 37 |
+
```
|
| 38 |
+
Detected language 'cy' with probability 0.999987
|
| 39 |
+
[0.00s -> 4.24s] Dwi teimlo weithie unwaith ti'n cyfadda bo' na rwbath yn bod ma'n wir wedyn dydi?
|
| 40 |
+
```
|