NVIDIA FastConformer-Transducer Large (en)

| | |

This model transcribes speech in lower case English alphabet. It is a "large" version of FastConformer Transducer (around 114M parameters) model. See the model architecture section and NeMo documentation for complete architecture details.

NVIDIA NeMo: Training

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']

How to Use this Model

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.

Automatically instantiate the model

import nemo.collections.asr as nemo_asr
asr_model = nemo_asr.models.EncDecRNNTBPEModel.from_pretrained(model_name="nvidia/stt_en_fastconformer_transducer_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:

asr_model.transcribe(['2086-149220-0033.wav'])

Transcribing many audio files

python [NEMO_GIT_FOLDER]/examples/asr/transcribe_speech.py 
 pretrained_name="nvidia/stt_en_fastconformer_transducer_large" 
 audio_dir="<DIRECTORY CONTAINING AUDIO FILES>"

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 .

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 .

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	3.8	1.8	1.4	2.4	5.5	5.8	7.5	NeMo ASRSET 3.0

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.

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 isn't 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

Licence

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.

.hf-sanitized.hf-sanitized-liB6pS1nHo9xrMeAgwHYI img {display: inline;} 本模型支持在NeMo工具包 [3] 中使用，并可用作推断的预训练检查点，或用于对另一个数据集进行微调。有关FastConformer的详细信息，请单击此处 Fast-Conformer Model 。该模型使用NeMo工具包[3]进行了数百个时期的训练。这些模型使用此 example script 和此 base config 进行训练。这些模型的标记器是使用训练集的文本转录进行构建的，具体步骤请参阅此 script 。该模型在NeMo ASRSet En的复合数据集上进行了训练，包括数千小时的英语语音。这些数据集包括：Librispeech 960小时的英语语音、Fisher语料库、Switchboard-1数据集、WSJ-0和WSJ-1、National Speech Corpus (Part 1, Part 6)、VCTK、VoxPopuli (EN)、Europarl-ASR (EN)、Multilingual Librispeech (MLS EN) – 2,000小时子集、Mozilla Common Voice (v7.0)和People's Speech – 12,000小时子集。该模型在不同领域和更大语料库上进行训练，因此在一般情况下，其在音频转录方面的性能更好。根据贪婪解码，以下表格总结了该系列中可用模型的性能，性能以词错误率 (Word Error Rate, WER%) 表示。由于该模型是基于公共可用语音数据集进行训练，因此对包含技术术语或模型未经训练的方言的语音的性能可能会降低。该模型对带有口音的语音的性能可能会更差。另外，NVIDIA Riva是一个可在本地、所有云上、多云、混合、边缘和嵌入式部署的加速语音AI SDK。Riva提供以下功能：1.基于专有数据训练的模型检查点，提供最常见语言的开箱即用精度；2.具有运行时词权重提升（例如品牌和产品名称）以及自定义声学模型、语言模型和逆文本规范化的最佳精度；3.流式语音识别、与Kubernetes兼容的缩放和企业级支持。此模型目前还不受Riva支持，但请查看 Riva live demo 以获取更多信息。参考资料：[1] Fast Conformer with Linearly Scalable Attention for Efficient Speech Recognition ，[2] Google Sentencepiece Tokenizer ，[3] NVIDIA NeMo Toolkit 。该模型的使用许可受到 CC-BY-4.0 的约束。下载和使用该模型的公共和发布版本即表示您接受 CC-BY-4.0 许可的条款和条件。.hf-sanitized.hf-sanitized-liB6pS1nHo9xrMeAgwHYI img {display: inline;}