๐ NVIDIA FastConformer-CTC Large (en)
This model is designed for automatic speech recognition, transcribing speech into lower - case English text. It's a large - scale FastConformer CTC model with around 115M parameters, offering high - performance speech recognition capabilities.
๐ Quick Start
Installation
To train, fine - tune or play with the model you will need to install NVIDIA NeMo. We recommend you install it after you've installed latest Pytorch version.
pip install nemo_toolkit['all']
Usage
The model is available for use in the NeMo toolkit and can be used as a pre - trained checkpoint for inference or for fine - tuning on another dataset.
Automatically instantiate the model
import nemo.collections.asr as nemo_asr
asr_model = nemo_asr.models.EncDecCTCModelBPE.from_pretrained(model_name="nvidia/stt_en_fastconformer_ctc_large")
Transcribing using Python
First, let's get a sample
wget https://dldata-public.s3.us-east-2.amazonaws.com/2086-149220-0033.wav
Then simply do:
output = asr_model.transcribe(['2086-149220-0033.wav'])
print(output[0].text)
Transcribing many audio files
python [NEMO_GIT_FOLDER]/examples/asr/transcribe_speech.py
pretrained_name="nvidia/stt_en_fastconformer_ctc_large"
audio_dir="<DIRECTORY CONTAINING AUDIO FILES>"
โจ Features
- High - Performance Architecture: FastConformer is an optimized version of the Conformer model with 8x depthwise - separable convolutional downsampling, trained in a multitask setup with a Transducer decoder loss.
- Multidomain Adaptability: Trained on a large composite dataset of English speech, it can perform well in various scenarios.
๐ป Usage Examples
Basic Usage
Automatically instantiate the model:
import nemo.collections.asr as nemo_asr
asr_model = nemo_asr.models.EncDecCTCModelBPE.from_pretrained(model_name="nvidia/stt_en_fastconformer_ctc_large")
Advanced Usage
Transcribing many audio files:
python [NEMO_GIT_FOLDER]/examples/asr/transcribe_speech.py
pretrained_name="nvidia/stt_en_fastconformer_ctc_large"
audio_dir="<DIRECTORY CONTAINING AUDIO FILES>"
๐ Documentation
Input
This model accepts 16000 Hz Mono - channel Audio (wav files) as input.
Output
This model provides transcribed speech as a string for a given audio sample.
Model Architecture
FastConformer [1] is an optimized version of the Conformer model with 8x depthwise - separable convolutional downsampling. The model is trained in a multitask setup with a Transducer decoder loss. You may find more information on the details of FastConformer here: [Fast - Conformer Model](https://docs.nvidia.com/deeplearning/nemo/user-guide/docs/en/main/asr/models.html#fast - conformer).
Training
The NeMo toolkit [3] was used for training the models for over several hundred epochs. These model are trained with this example script and this [base config](https://github.com/NVIDIA/NeMo/blob/main/examples/asr/conf/fastconformer/fast - conformer_ctc_bpe.yaml).
The tokenizers for these models were built using the text transcripts of the train set with this script.
Datasets
The model in this collection is trained on a composite dataset (NeMo ASRSet En) comprising several thousand hours of English speech:
- Librispeech 960 hours of English speech
- Fisher Corpus
- Switchboard - 1 Dataset
- WSJ - 0 and WSJ - 1
- National Speech Corpus (Part 1, Part 6)
- VCTK
- VoxPopuli (EN)
- Europarl - ASR (EN)
- Multilingual Librispeech (MLS EN) - 2,000 hrs subset
- Mozilla Common Voice (v7.0)
- People's Speech - 12,000 hrs subset
Performance
The performance of Automatic Speech Recognition models is measuring using Word Error Rate. Since this dataset is trained on multiple domains and a much larger corpus, it will generally perform better at transcribing audio in general.
The following tables summarizes the performance of the available models in this collection with the Transducer decoder. Performances of the ASR models are reported in terms of Word Error Rate (WER%) with greedy decoding.
Version |
Tokenizer |
Vocabulary Size |
LS test - other |
LS test - clean |
WSJ Eval92 |
WSJ Dev93 |
NSC Part 1 |
MLS Test |
MCV Test 7.0 |
Train Dataset |
1.18.0 |
SentencePiece Unigram |
1024 |
4.2 |
2.1 |
1.6 |
2.5 |
6.3 |
6.4 |
8.3 |
NeMo ASRSET 3.0 |
๐ง Technical Details
FastConformer is an optimized version of the Conformer model. It uses 8x depthwise - separable convolutional downsampling and is trained in a multitask setup with a Transducer decoder loss. For more detailed information, refer to [Fast - Conformer Model](https://docs.nvidia.com/deeplearning/nemo/user-guide/docs/en/main/asr/models.html#fast - conformer).
๐ License
License to use this model is covered by the CC - BY - 4.0. By downloading the public and release version of the model, you accept the terms and conditions of the CC - BY - 4.0 license.
Additional Information
Limitations
โ ๏ธ Important Note
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.
NVIDIA Riva: Deployment
NVIDIA Riva, is an accelerated speech AI SDK deployable on - prem, in all clouds, multi - cloud, hybrid, on edge, and embedded.
Additionally, Riva provides:
- World - class out - of - the - box accuracy for the most common languages with model checkpoints trained on proprietary data with hundreds of thousands of GPU - compute hours
- Best in class accuracy with run - time word boosting (e.g., brand and product names) and customization of acoustic model, language model, and inverse text normalization
- Streaming speech recognition, Kubernetes compatible scaling, and enterprise - grade support
Although this model isnโt supported yet by Riva, the list of supported models is here.
Check out Riva live demo.
References
[1] Fast Conformer with Linearly Scalable Attention for Efficient Speech Recognition
[2] Google Sentencepiece Tokenizer
[3] NVIDIA NeMo Toolkit