QuartzNet15x5 checkpoints available here are trained using NeMo on the Ai-shell 2 Mandarin Chinese dataset.
It utilizes a character encoding scheme, and transcribes text in the standard character set that is provided in the Aishell-2 Mandard Corpus.
The Quartznet model is composed of multiple blocks with residual connections between them, trained with CTC loss. Each block consists of one or more modules with 1D time-channel separable convolutional layers, batch normalization, and ReLU layers.
The Quartznet 15x5 model consists of 79 layers and has a total of 18.9 million parameters, with five blocks that repeat fifteen times plus four additional convolutional layers .
This model was trained on the open source Aishell-2  corpus consisting of about 1000 hours transcribed Mandarin speech. The NeMo toolkit  was used for training this model over several hundred epochs on multiple GPUs.
The model has been fine-tuned with Room Impulse Response (RIR) and noise augmentation to make it more robust to noise and is also augmented with narrowband speech codecs to be more accurate with 8 kHz telephony data.
While training this model, we used the following datasets:
The performance of Automatic Speech Recognition models is measuring using Character Error Rate.
The model obtains the following scores on the following evaluation datasets -
Note that these scores on Aishell-2 are not particularly indicative of the quality of transcriptions that models trained on ASR Set will achieve, but they are a useful proxy.
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.
import nemo.collections.asr as nemo_asr asr_model = nemo_asr.models.EncDecCTCModel.from_pretrained(model_name="stt_zh_quartznet15x5")
python [NEMO_GIT_FOLDER]/examples/asr/transcribe_speech.py \ pretrained_name="stt_zh_quartznet15x5" \ audio_dir=""
This model accepts 16000 KHz Mono-channel Audio (wav files) as input.
This model provides transcribed speech as a string for a given audio sample.
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.