Model Overview

This collection contains the English FastConformer Hybrid (Transducer and CTC) Large model (around 114M parameters) with Punctuation and Capitalization on NeMo ASRSet En PC with around 8500 hours of English speech (SPGI 1k, VoxPopuli, MCV11, Europarl-ASR, Fisher, LibriSpeech, NSC1, MLS).

It utilizes a Google SentencePiece [1] tokenizer with a vocabulary size of 1024. It transcribes text in upper and lower case English alphabet along with spaces, periods, commas, question marks, and a few other characters.

Model Architecture

FastConformer is an optimized version of the Conformer model [2] with 8x depthwise-separable convolutional downsampling. The model is trained in a multitask setup with joint Transducer and CTC decoder loss. You may find more information on the details of FastConformer here: Fast-Conformer Model and about Hybrid Transducer-CTC training here: Hybrid Transducer-CTC.

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 PC) comprising several thousand hours of English speech:

• LibriSpeech (874 hrs)
• Fisher (998 hrs)
• MCV11 (1474 hrs)
• NSC1 (1381 hours)
• VCTK (82 hours)
• VoxPopuli (353 hours)
• Europarl-ASR (763 hours)
• MLS (1860 hours)
• SPGI (795 hours)

Tokenizer Construction

The tokenizer for this model was built using text corpus provided with the train dataset.

We build a Google Sentencepiece Tokenizer [1] with the following script :

python [NEMO_GIT_FOLDER]/scripts/tokenizers/process_asr_text_tokenizer.py \
  --manifest="train_manifest.json" \
  --data_root="<OUTPUT DIRECTORY FOR TOKENIZER>" \
  --vocab_size=1024 \
  --tokenizer="spe" \
  --spe_type="unigram" \
  --spe_character_coverage=1.0 \
  --no_lower_case \
  --log

Performance

The performance of Automatic Speech Recognition models is measured 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.

The following tables summarize 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.

a) On data without Punctuation and Capitalization with Transducer decoder

b) On data with Punctuation and Capitalization with Transducer decoder|

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 load the model from NGC

import nemo.collections.asr as nemo_asr
asr_model = nemo_asr.models.EncDecHybridRNNTCTCBPEModel.from_pretrained(model_name="stt_en_fastconformer_hybrid_large_pc")

Transcribing text with this model

Using Transducer mode inference:

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

Using CTC mode inference:

python [NEMO_GIT_FOLDER]/examples/asr/transcribe_speech.py \
  pretrained_name="stt_en_fastconformer_hybrid_large_pc" \
  audio_dir="<DIRECTORY CONTAINING AUDIO FILES>" \
  decoder_type="ctc"

Input

This model accepts 16 kHz Mono-channel Audio (wav files) as input.

Output

This model provides transcribed speech as a string for a given audio sample.

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. The model only outputs the punctuations: '.', ',', '?' and hence might not do well in scenarios where other punctuations are also expected.

References

[1] Google Sentencepiece Tokenizer

[2] Conformer: Convolution-augmented Transformer for Speech Recognition

[3] NVIDIA NeMo Toolkit

Licence

License to use this model is covered by the NGC TERMS OF USE unless another License/Terms Of Use/EULA is clearly specified. By downloading the public and release version of the model, you accept the terms and conditions of the NGC TERMS OF USE.