Automatic Speech Recognition
speechbrain
PyTorch
English
wav2vec2
CTC
Transformer
Eval Results (legacy)
Instructions to use speechbrain/asr-wav2vec2-commonvoice-14-en with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- speechbrain
How to use speechbrain/asr-wav2vec2-commonvoice-14-en with speechbrain:
from speechbrain.pretrained import EncoderASR model = EncoderASR.from_hparams( "speechbrain/asr-wav2vec2-commonvoice-14-en" ) model.transcribe_file("file.wav") - Notebooks
- Google Colab
- Kaggle
Update hyperparams.yaml
Browse files- hyperparams.yaml +1 -1
hyperparams.yaml
CHANGED
|
@@ -92,7 +92,7 @@ test_beam_search:
|
|
| 92 |
topk: 1
|
| 93 |
alpha: 1.0
|
| 94 |
beta: 0.5
|
| 95 |
-
|
| 96 |
|
| 97 |
pretrainer: !new:speechbrain.utils.parameter_transfer.Pretrainer
|
| 98 |
loadables:
|
|
|
|
| 92 |
topk: 1
|
| 93 |
alpha: 1.0
|
| 94 |
beta: 0.5
|
| 95 |
+
kenlm_model_path: speechbrain/asr-wav2vec2-commonvoice-14-en/en_5gram.arpa # uncomment this line if you do want to use ngram rescoring
|
| 96 |
|
| 97 |
pretrainer: !new:speechbrain.utils.parameter_transfer.Pretrainer
|
| 98 |
loadables:
|