Instructions to use GetmanY1/wav2vec2-base-fi-150k with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use GetmanY1/wav2vec2-base-fi-150k with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("automatic-speech-recognition", model="GetmanY1/wav2vec2-base-fi-150k")# Load model directly from transformers import AutoProcessor, AutoModelForPreTraining processor = AutoProcessor.from_pretrained("GetmanY1/wav2vec2-base-fi-150k") model = AutoModelForPreTraining.from_pretrained("GetmanY1/wav2vec2-base-fi-150k", device_map="auto") - Notebooks
- Google Colab
- Kaggle
Update README.md
Browse files
README.md
CHANGED
|
@@ -43,10 +43,10 @@ This model was pre-trained with 158k hours of Finnish speech data from the follo
|
|
| 43 |
| [Lahjoita puhetta](https://www.kielipankki.fi/donate-speech/) | 2740 h | 1.74 % |
|
| 44 |
| [Finnish parliament](http://urn.fi/urn:nbn:fi:lb-2016042502) | 2692 h | 1.71 % |
|
| 45 |
| [VoxPopuli Finnish](https://github.com/facebookresearch/voxpopuli) | 14264 h | 9.04 % |
|
| 46 |
-
| [YlePuhe](
|
| 47 |
-
| [MTV3](
|
| 48 |
-
| [YleTV1](
|
| 49 |
-
| [AlfaTV](
|
| 50 |
|
| 51 |
Datasets were filtered to include a maximum length of 60 seconds long audio samples.
|
| 52 |
|
|
|
|
| 43 |
| [Lahjoita puhetta](https://www.kielipankki.fi/donate-speech/) | 2740 h | 1.74 % |
|
| 44 |
| [Finnish parliament](http://urn.fi/urn:nbn:fi:lb-2016042502) | 2692 h | 1.71 % |
|
| 45 |
| [VoxPopuli Finnish](https://github.com/facebookresearch/voxpopuli) | 14264 h | 9.04 % |
|
| 46 |
+
| [YlePuhe](https://kavi.fi/en/radio-ja-televisioarkistointia-vuodesta-2008/) | 95478 h | 60.52 % |
|
| 47 |
+
| [MTV3](https://kavi.fi/en/radio-ja-televisioarkistointia-vuodesta-2008/) | 16723 h | 10.60 % |
|
| 48 |
+
| [YleTV1](https://kavi.fi/en/radio-ja-televisioarkistointia-vuodesta-2008/) | 13947 h | 8.84 % |
|
| 49 |
+
| [AlfaTV](https://kavi.fi/en/radio-ja-televisioarkistointia-vuodesta-2008/) | 11933 h | 7.56 % |
|
| 50 |
|
| 51 |
Datasets were filtered to include a maximum length of 60 seconds long audio samples.
|
| 52 |
|