Module elpis.models.job

from __future__ import annotations

import json
from copy import copy
from dataclasses import dataclass, field
from pathlib import Path
from typing import Any, Dict, List, Optional

from transformers import HfArgumentParser, TrainingArguments


def list_field(default=None, metadata=None):
    return field(default_factory=lambda: default, metadata=metadata)


@dataclass
class ModelArguments:
    """
    Arguments pertaining to which model/config/tokenizer we are going to fine-tune from.
    """

    model_name_or_path: str = field(
        metadata={
            "help": "Path to pretrained model or model identifier from huggingface.co/models"
        }
    )
    tokenizer_name_or_path: Optional[str] = field(
        default=None,
        metadata={
            "help": "Path to pretrained tokenizer or tokenizer identifier from huggingface.co/models"
        },
    )
    cache_dir: Optional[str] = field(
        default=None,
        metadata={
            "help": "Where do you want to store the pretrained models downloaded from huggingface.co"
        },
    )
    freeze_feature_encoder: bool = field(
        default=True,
        metadata={"help": "Whether to freeze the feature encoder layers of the model."},
    )
    attention_dropout: float = field(
        default=0.0,
        metadata={"help": "The dropout ratio for the attention probabilities."},
    )
    activation_dropout: float = field(
        default=0.0,
        metadata={
            "help": "The dropout ratio for activations inside the fully connected layer."
        },
    )
    feat_proj_dropout: float = field(
        default=0.0, metadata={"help": "The dropout ratio for the projected features."}
    )
    hidden_dropout: float = field(
        default=0.0,
        metadata={
            "help": "The dropout probability for all fully connected layers in the embeddings, encoder, and pooler."
        },
    )
    final_dropout: float = field(
        default=0.0,
        metadata={"help": "The dropout probability for the final projection layer."},
    )
    mask_time_prob: float = field(
        default=0.05,
        metadata={
            "help": (
                "Probability of each feature vector along the time axis to be chosen as the start of the vector"
                "span to be masked. Approximately ``mask_time_prob * sequence_length // mask_time_length`` feature"
                "vectors will be masked along the time axis."
            )
        },
    )
    mask_time_length: int = field(
        default=10,
        metadata={"help": "Length of vector span to mask along the time axis."},
    )
    mask_feature_prob: float = field(
        default=0.0,
        metadata={
            "help": (
                "Probability of each feature vector along the feature axis to be chosen as the start of the vectorspan"
                " to be masked. Approximately ``mask_feature_prob * sequence_length // mask_feature_length`` feature"
                " bins will be masked along the time axis."
            )
        },
    )
    mask_feature_length: int = field(
        default=10,
        metadata={"help": "Length of vector span to mask along the feature axis."},
    )
    layerdrop: float = field(
        default=0.0, metadata={"help": "The LayerDrop probability."}
    )
    ctc_loss_reduction: Optional[str] = field(
        default="mean",
        metadata={
            "help": "The way the ctc loss should be reduced. Should be one of 'mean' or 'sum'."
        },
    )
    ctc_zero_infinity: bool = field(
        default=False,
        metadata={
            "help": "Whether to zero infinite losses and the associated gradients of `torch.nn.CTCLoss`. "
            "Infinite losses mainly occur when the inputs are too short to be aligned to the targets. "
            "Only relevant when training an instance of Wav2Vec2ForCTC."
        },
    )

    def to_dict(self) -> Dict[str, Any]:
        result = dict(self.__dict__)
        return result


@dataclass
class DataArguments:
    """
    Arguments pertaining to what data we are going to input our model for training and eval.

    Using `HfArgumentParser` we can turn this class
    into argparse arguments to be able to specify them on
    the command line.
    """

    dataset_name_or_path: str = field(
        metadata={
            "help": "If a path, the path to a directory containing the dataset files. "
            "Otherwise- the name of the dataset to use (via the datasets library)."
        }
    )
    dataset_config_name: Optional[str] = field(
        default=None,
        metadata={
            "help": "The configuration name of the dataset to use (via the datasets library)."
        },
    )
    stream_dataset: bool = field(
        default=False,
        metadata={
            "help": "Whether to stream the dataset as opposed to downloading it all at once."
        },
    )
    train_split_name: str = field(
        default="train+validation",
        metadata={
            "help": (
                "The name of the training data set split to use (via the datasets library). Defaults to "
                "'train+validation'"
            )
        },
    )
    eval_split_name: str = field(
        default="test",
        metadata={
            "help": "The name of the evaluation data set split to use (via the datasets library). Defaults to 'test'"
        },
    )
    audio_column_name: str = field(
        default="audio",
        metadata={
            "help": "The name of the dataset column containing the audio data. Defaults to 'audio'"
        },
    )
    text_column_name: str = field(
        default="text",
        metadata={
            "help": "The name of the dataset column containing the text data. Defaults to 'text'"
        },
    )
    overwrite_cache: bool = field(
        default=False,
        metadata={"help": "Overwrite the cached preprocessed datasets or not."},
    )
    preprocessing_num_workers: Optional[int] = field(
        default=None,
        metadata={"help": "The number of processes to use for the preprocessing."},
    )
    max_train_samples: Optional[int] = field(
        default=None,
        metadata={
            "help": (
                "For debugging purposes or quicker training, truncate the number of training examples to this "
                "value if set."
            )
        },
    )
    max_eval_samples: Optional[int] = field(
        default=None,
        metadata={
            "help": (
                "For debugging purposes or quicker training, truncate the number of validation examples to this "
                "value if set."
            )
        },
    )
    do_clean: bool = field(
        default=True,
        metadata={"help": "True if the dataset should be cleaned before use."},
    )
    words_to_remove: Optional[List[str]] = list_field(
        default=[],
        metadata={
            "help": "A list of words to remove from the transcripts during dataset cleaning."
        },
    )
    chars_to_remove: Optional[List[str]] = list_field(
        default=[],
        metadata={
            "help": "A list of characters to remove from the transcripts during dataset cleaning."
        },
    )
    chars_to_explode: Optional[List[str]] = list_field(
        default=[],
        metadata={
            "help": "A list of characters to replace with spaces in the transcripts during dataset cleaning."
        },
    )
    do_lower_case: Optional[bool] = field(
        default=None,
        metadata={"help": "Whether the target text should be lower cased."},
    )
    eval_metrics: List[str] = list_field(  # type: ignore
        default=["wer", "cer"],
        metadata={
            "help": "A list of metrics the model should be evaluated on. E.g. `('wer', 'cer')`"
        },
    )
    max_duration_in_seconds: float = field(
        default=20.0,
        metadata={
            "help": (
                "Filter audio files that are longer than `max_duration_in_seconds` seconds to"
                " 'max_duration_in_seconds`"
            )
        },
    )
    min_duration_in_seconds: float = field(
        default=0.0,
        metadata={
            "help": "Filter audio files that are shorter than `min_duration_in_seconds` seconds"
        },
    )
    preprocessing_only: bool = field(
        default=False,
        metadata={
            "help": (
                "Whether to only do data preprocessing and skip training. This is especially useful when data"
                " preprocessing errors out in distributed training due to timeout. In this case, one should run the"
                " preprocessing in a non-distributed setup with `preprocessing_only=True` so that the cached datasets"
                " can consequently be loaded in distributed training"
            )
        },
    )
    token: Optional[str] = field(
        default=None,
        metadata={
            "help": (
                "The token to use as HTTP bearer authorization for remote files. If not specified, will use the token "
                "generated when running `huggingface-cli login` (stored in `~/.huggingface`)."
            )
        },
    )
    use_auth_token: Optional[bool] = field(
        default=None,
        metadata={
            "help": "The `use_auth_token` argument is deprecated and will be removed in v4.34. Please use `token`."
        },
    )
    trust_remote_code: bool = field(
        default=False,
        metadata={
            "help": (
                "Whether or not to allow for custom models defined on the Hub in their own modeling files. This option"
                "should only be set to `True` for repositories you trust and in which you have read the code, as it will"
                "execute code present on the Hub on your local machine."
            )
        },
    )
    unk_token: str = field(
        default="[UNK]",
        metadata={"help": "The unk token for the tokenizer"},
    )
    pad_token: str = field(
        default="[PAD]",
        metadata={"help": "The padding token for the tokenizer"},
    )
    word_delimiter_token: str = field(
        default="|",
        metadata={"help": "The word delimiter token for the tokenizer"},
    )
    phoneme_language: Optional[str] = field(
        default=None,
        metadata={
            "help": (
                "The target language that should be used be"
                " passed to the tokenizer for tokenization. Note that"
                " this is only relevant if the model classifies the"
                " input audio to a sequence of phoneme sequences."
            )
        },
    )

    def to_dict(self) -> Dict[str, Any]:
        result = dict(self.__dict__)
        return result


@dataclass
class Job:
    """Generic class which encapsulates elpis training functionality"""

    model_args: ModelArguments
    data_args: DataArguments
    training_args: TrainingArguments

    @staticmethod
    def parser():
        return HfArgumentParser((ModelArguments, DataArguments, TrainingArguments))  # type: ignore

    @classmethod
    def from_args(cls, args=None) -> Job:
        (
            model_args,
            data_args,
            training_args,
        ) = Job.parser().parse_args_into_dataclasses(args)
        return cls(
            model_args=model_args, data_args=data_args, training_args=training_args
        )

    @classmethod
    def from_json(cls, file: Path) -> Job:
        (
            model_args,
            data_args,
            training_args,
        ) = Job.parser().parse_json_file(str(file))
        return cls(
            model_args=model_args, data_args=data_args, training_args=training_args
        )

    def save(self, path: Path, overwrite=True) -> None:
        if not overwrite and path.is_file():
            return

        with open(path, "w") as out_file:
            json.dump(self.to_dict(), out_file)

    @classmethod
    def from_dict(cls, data: Dict[str, Any]) -> Job:
        (
            model_args,
            data_args,
            training_args,
        ) = Job.parser().parse_dict(data)
        return cls(
            model_args=model_args, data_args=data_args, training_args=training_args
        )

    def to_dict(self) -> Dict[str, Any]:
        return (
            self.training_args.to_dict()
            | self.data_args.to_dict()
            | self.model_args.to_dict()
        )

    def __eq__(self, __value: object) -> bool:
        if not isinstance(__value, Job):
            return False

        job = __value

        return (
            self.training_args.to_dict() == job.training_args.to_dict()
            and self.model_args == job.model_args
            and self.data_args == job.data_args
        )