Distilbert-punctuator is a python package provides a bert-based punctuator (fine-tuned model of pretrained huggingface DistilBertForTokenClassification) with following three components:
-
data process: funcs for processing user's data to prepare for training. If user perfer to fine-tune the model with his/her own data.
-
training: training pipeline and evaluation. User can fine-tune his/her own punctuator with the pipeline
-
inference: easy-to-use interface for user to use trained punctuator.
-
If user doesn't want to train a punctuator himself/herself, two pre-fined-tuned model from huggingface model hub
Qishuai/distilbert_punctuator_en๐ Model detailsQishuai/distilbert_punctuator_zh๐ Model details
-
model examples in huggingface web page.
- English model
- Simplified Chinese model
- Installing the package from pypi:
pip install distilbert-punctuatorfor directly usage of punctuator. - Installing the package with option to do data processing
pip install distilbert-punctuator[data_process]. - Installing the package with option to train and validate your own model
pip install distilbert-punctuator[training] - For development and contribution
- clone the repo
make install
Component for pre-processing the training data. To use this component, please install as pip install distilbert-punctuator[data_process]
The package is providing a simple pipeline for you to generate NER format training data.
examples/data_sample.py
Component for providing a training pipeline for fine-tuning a pretrained DistilBertForTokenClassification model from huggingface.
The latest version has the implementation of R-Drop enhanced training.
R-Drop github repo
Paper of R-Drop
examples/english_train_sample.py
Arguments required for the training pipeline.
-
training_corpus(List[List[str]]): list of sequences for training, longest sequence should be no longer than pretrained LM # noqa: E501validation_corpus(List[List[str]]): list of sequences for validation, longest sequence should be no longer than pretrained LM # noqa: E501training_tags(List[List[int]]): tags(int) for trainingvalidation_tags(List[List[int]]): tags(int) for validationmodel_name_or_path(str): name or path of pre-trained modeltokenizer_name(str): name of pretrained tokenizer
-
epoch(int): number of epochbatch_size(int): batch sizemodel_storage_dir(str): fine-tuned model storage pathlabel2id(Dict): the tags label and id mappingearly_stop_count(int): after how many epochs to early stop training if valid loss not become smaller. default 3 # noqa: E501gpu_device(int): specific gpu card index, default is the CUDA_VISIBLE_DEVICES from environwarm_up_steps(int): warm up steps.r_drop(bool): whether to train with r-dropr_alpha(int): alpha value for kl divengence in the loss, default is 0plot_steps(int): record training status to tensorboard among how many stepstensorboard_log_dir(Optional[str]): the tensorboard logs output directory, default is "runs"
-
addtional_model_config(Optional[Dict]): additional configuration for model
You can also train your own NER models with the trainer provided in this repo.
The example can be found in notebooks/R-drop NER.ipynb
Validation of fine-tuned model
examples/train_sample.py
evaluation_corpus(List[List[str]]): list of sequences for evaluation, longest sequence should be no longer than pretrained LM's max_position_embedding(512)evaluation_tags(List[List[int]]): tags(int) for evaluation (the GT)model_name_or_path(str): name or path of fine-tuned modeltokenizer_name(str): name of tokenizerbatch_size(int): batch sizelabel2id(Optional[Dict]): label2id. Default one is from model config. Pass in this argument if your model doesn't have a label2id inside configgpu_device(int): specific gpu card index, default is the CUDA_VISIBLE_DEVICES from environ
Component for providing an inference interface for user to use punctuator.
+----------------------+ (child process)
| user application | +-------------------+
+ + <---------->| punctuator server |
| +inference object | +-------------------+
+----------------------+
The punctuator will be deployed in a child process which communicates with main process through pipe connection.
Therefore user can initialize an inference object and call its punctuation function when needed. The punctuator will never block the main process unless doing punctuation.
There is a graceful shutdown methodology for the punctuator, hence user dosen't need to worry about the shutting-down.
examples/inference_sample.py
Arguments required for the inference pipeline.
model_name_or_path(str): name or path of pre-trained modeltokenizer_name(str): name of pretrained tokenizertag2punctuator(Dict[str, tuple]): tag to punctuator mapping. dbpunctuator.utils provides two default mappings for English and Chinesefor own fine-tuned model with different tags, pass in your own mappingNORMAL_TOKEN_TAG = "O" DEFAULT_ENGLISH_TAG_PUNCTUATOR_MAP = { NORMAL_TOKEN_TAG: ("", False), "COMMA": (",", False), "PERIOD": (".", True), "QUESTIONMARK": ("?", True), "EXLAMATIONMARK": ("!", True), } DEFAULT_CHINESE_TAG_PUNCTUATOR_MAP = { NORMAL_TOKEN_TAG: ("", False), "C_COMMA": ("๏ผ", False), "C_PERIOD": ("ใ", True), "C_QUESTIONMARK": ("? ", True), "C_EXLAMATIONMARK": ("! ", True), "C_DUNHAO": ("ใ", False), }
tag2id_storage_path(Optional[str]): tag2id storage path. Default one is from model config. Pass in this argument if your model doesn't have a tag2id inside config
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent