History

Hui Zhang 69055698a2 transformer using batch data loader		3 years ago
..
conf	fix the bug of chooing dataloader, remove the log of downloads lm, change the epoch in tiny	3 years ago
local	transformer using batch data loader	3 years ago
.gitignore	E2E/Streaming Transformer/Conformer ASR (#578 )	4 years ago
README.md	fix textfrontend readme, fix imgs link	3 years ago
path.sh	merge deepspeech, parakeet and text_processing into paddlespeech	3 years ago
run.sh	fix recipe train and avg shell	3 years ago

Tiny Example

Steps

Prepare the data
```
bash local/data.sh
```
data.sh will download dataset, generate manifests, collect normalizer's statistics and build vocabulary. Once the data preparation is done, you will find the data (only part of LibriSpeech) downloaded in ${MAIN_ROOT}/dataset/librispeech and the corresponding manifest files generated in ${PWD}/data as well as a mean stddev file and a vocabulary file. It has to be run for the very first time you run this dataset and is reusable for all further experiments.
Train your own ASR model
```
bash local/train.sh
```
train.sh will start a training job, with training logs printed to stdout and model checkpoint of every pass/epoch saved to ${PWD}/checkpoints. These checkpoints could be used for training resuming, inference, evaluation and deployment.
Case inference with an existing model
Evaluate an existing model
```
bash local/test.sh
```
test.sh will evaluate the model with Word Error Rate (or Character Error Rate) measurement. Similarly, you can also download a well-trained model and test its performance:

Export jit model

bash local/export.sh ckpt_path saved_jit_model_path