File size: 3,135 Bytes
b87ee07 c587a36 c165b32 c587a36 bd9a43b c165b32 31b09b5 b87ee07 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 |
---
language:
- pt
license: cc-by-4.0
library_name: nemo
datasets:
- mozilla-foundation/common_voice_9_0
thumbnail: null
tags:
- automatic-speech-recognition
- speech
- audio
- CTC
- QuartzNet
- Transformer
- NeMo
- pytorch
model-index:
- name: stt_pt_quartznet15x5_ctc_small
results:
- task:
type: automatic-speech-recognition
dataset:
type: common_voice
name: Common Voice Portuguese
config: clean
split: test
args:
language: pt
metrics:
- type: wer
value: 49.17
name: Test WER
- type: cer
value: 18.59
name: Test CER
---
## Model Overview
This model transcribes speech in lower case Portuguese alphabet along with spaces. It is a "small" versions of QuartzNet-CTC model.
## NVIDIA NeMo: Training
To train, fine-tune or play with the model you will need to install [NVIDIA NeMo](https://github.com/NVIDIA/NeMo). We recommend you install it after you've installed latest Pytorch version.
```
pip install nemo_toolkit['all']
```
## How to Use this Model
The model is available for use in the NeMo toolkit [1], and can be used as a pre-trained checkpoint for inference or for fine-tuning on another dataset.
### Automatically instantiate the model
```python
import nemo.collections.asr as nemo_asr
asr_model = nemo_asr.models.ASRModel.from_pretrained("dominguesm/stt_pt_quartznet15x5_ctc_small")
```
### Transcribing using Python
First, let's get a sample
```
wget https://github.com/DominguesM/stt_pt_quartznet15x5_ctc_small/raw/main/audios/common_voice_pt_25555332.mp3
```
Then simply do:
```
asr_model.transcribe(['common_voice_pt_25555332.mp3'])
```
### Transcribing many audio files
```shell
python [NEMO_GIT_FOLDER]/examples/asr/transcribe_speech.py pretrained_name="dominguesm/stt_pt_quartznet15x5_ctc_small" audio_dir="<DIRECTORY CONTAINING AUDIO FILES>"
```
### Input
This model accepts 16000 KHz Mono-channel Audio (wav files) as input.
### Output
This model provides transcribed speech as a string for a given audio sample.
## Model Architecture
This model are based on the QuartzNet architecture, which is a variant of Jasper that uses 1D time-channel separable convolutional layers in its convolutional residual blocks and are therefore smaller than Jasper models.
QuartzNet models take in audio segments and transcribe them to letter, byte pair, or word piece sequences.
## Training
All training scripts will be available at: [DominguesM/stt_pt_quartznet15x5_ctc_small](https://github.com/DominguesM/stt_pt_quartznet15x5_ctc_small)
**Soon more information**
### Datasets
The model was trained with a part of the Common Voices 9.0 dataset in Portuguese, totaling 26 hours of audio.
* Mozilla Common Voice (v9.0)
## Performance
**Coming soon**
## Limitations
Since this model was trained on publically available speech datasets, the performance of this model might degrade for speech which includes technical terms, or vernacular that the model has not been trained on. The model might also perform worse for accented speech.
## References
[1] [NVIDIA NeMo Toolkit](https://github.com/NVIDIA/NeMo)
|