This model transcribes speech in lowercase English alphabet including spaces and apostrophes, and is trained on several thousand hours of English speech data.
It is a non-autoregressive "large" variant of Conformer, with around 120 million parameters.
See the model architecture section and NeMo documentation for complete architecture details.
It is also compatible with NVIDIA Riva for production-grade server deployments.
Usage
The model is available for use in the NeMo toolkit [3], and can be used as a pre-trained checkpoint for inference or for fine-tuning on another dataset.
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']
Automatically instantiate the model
python
1import nemo.collections.asr as nemo_asr
2asr_model = nemo_asr.models.EncDecCTCModelBPE.from_pretrained("nvidia/stt_en_conformer_ctc_large")
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
Conformer-CTC model is a non-autoregressive variant of Conformer model [1] for Automatic Speech Recognition which uses CTC loss/decoding instead of Transducer. You may find more info on the detail of this model here: Conformer-CTC Model.
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.
The tokenizers for these models were built using the text transcripts of the train set with this script.
The checkpoint of the language model used as the neural rescorer can be found here. You may find more info on how to train and use language models for ASR models here: ASR Language Modeling
Datasets
All the models in this collection are trained on a composite dataset (NeMo ASRSET) comprising of several thousand hours of English speech:
Note: older versions of the model may have trained on smaller set of datasets.
Performance
The list of the available models in this collection is shown in the following table. 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
MLS Dev
MCV Test 6.1
Train Dataset
1.6.0
SentencePiece Unigram
128
4.3
2.2
2.0
2.9
7.0
7.2
6.5
8.0
NeMo ASRSET 2.0
While deploying with NVIDIA Riva, you can combine this model with external language models to further improve WER. The WER(%) of the latest model with different language modeling techniques are reported in the following table.
Since this model was trained on publicly 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.
Deployment with NVIDIA Riva
For the best real-time accuracy, latency, and throughput, deploy the model with NVIDIA Riva, an accelerated speech AI SDK deployable on-prem, in all clouds, multi-cloud, hybrid, at the 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
