hezar package

Subpackages

• hezar.data package
  • Subpackages
    • hezar.data.datasets package
      • Submodules
      • Module contents
  • Submodules
    • hezar.data.data_collators module
      • CharLevelOCRDataCollator
      • ImageCaptioningDataCollator
      • SequenceLabelingDataCollator
      • SpeechRecognitionDataCollator
      • TextGenerationDataCollator
      • TextPaddingDataCollator
    • hezar.data.data_samplers module
      • RangedSampler
    • hezar.data.dataset_processors module
      • DatasetProcessor
      • ImageCaptioningDatasetProcessor
      • OCRDatasetProcessor
      • SequenceLabelingDatasetProcessor
      • SpeechRecognitionDatasetProcessor
      • TextClassificationDatasetProcessor
      • TextSummarizationDatasetProcessor
  • Module contents
• hezar.embeddings package
  • Submodules
    • hezar.embeddings.embedding module
      • Embedding
    • hezar.embeddings.fasttext module
      • FastText
      • FastTextConfig
    • hezar.embeddings.word2vec module
      • Word2Vec
      • Word2VecConfig
  • Module contents
• hezar.metrics package
  • Submodules
    • hezar.metrics.accuracy module
      • Accuracy
      • AccuracyConfig
    • hezar.metrics.bleu module
      • BLEU
      • BLEUConfig
    • hezar.metrics.cer module
      • CER
      • CERConfig
    • hezar.metrics.f1 module
      • F1
      • F1Config
    • hezar.metrics.metric module
      • Metric
    • hezar.metrics.precision module
      • Precision
      • PrecisionConfig
    • hezar.metrics.recall module
      • Recall
      • RecallConfig
    • hezar.metrics.rouge module
      • ROUGE
      • ROUGEConfig
    • hezar.metrics.seqeval module
      • Seqeval
      • SeqevalConfig
    • hezar.metrics.wer module
      • WER
      • WERConfig
  • Module contents
• hezar.models package
  • Subpackages
    • hezar.models.backbone package
      • Subpackages
      • Module contents
    • hezar.models.image2text package
      • Subpackages
      • Module contents
    • hezar.models.mask_filling package
      • Subpackages
      • Module contents
    • hezar.models.sequence_labeling package
      • Subpackages
      • Module contents
    • hezar.models.speech_recognition package
      • Subpackages
      • Module contents
    • hezar.models.text_classification package
      • Subpackages
      • Module contents
    • hezar.models.text_detection package
      • Subpackages
      • Module contents
    • hezar.models.text_generation package
      • Subpackages
      • Module contents
  • Submodules
    • hezar.models.model module
      • Model
    • hezar.models.model_outputs module
      • Image2TextOutput
      • MaskFillingOutput
      • ModelOutput
      • SequenceLabelingOutput
      • SpeechRecognitionOutput
      • TextClassificationOutput
      • TextDetectionOutput
      • TextGenerationOutput
  • Module contents
• hezar.preprocessors package
  • Subpackages
    • hezar.preprocessors.tokenizers package
      • Submodules
      • Module contents
  • Submodules
    • hezar.preprocessors.audio_feature_extractor module
      • AudioFeatureExtractor
      • AudioFeatureExtractorConfig
    • hezar.preprocessors.image_processor module
      • ImageProcessor
      • ImageProcessorConfig
    • hezar.preprocessors.preprocessor module
      • Preprocessor
      • PreprocessorsContainer
    • hezar.preprocessors.text_normalizer module
      • TextNormalizer
      • TextNormalizerConfig
  • Module contents
• hezar.trainer package
  • Submodules
    • hezar.trainer.metrics_handlers module
      • Image2TextMetricHandler
      • MetricsHandler
      • SequenceLabelingMetricsHandler
      • SpeechRecognitionMetricsHandler
      • TextClassificationMetricsHandler
      • TextGenerationMetricsHandler
    • hezar.trainer.trainer module
      • Trainer
    • hezar.trainer.trainer_utils module
      • AverageMeter
      • CSVLogger
      • MetricsTracker
      • TrainerState
      • get_distributed_logger()
      • get_lr_scheduler_type()
      • resolve_logdir()
      • write_to_tensorboard()
  • Module contents
• hezar.utils package
  • Submodules
    • hezar.utils.audio_utils module
      • amplitude_to_db()
      • hertz_to_mel()
      • load_audio_files()
      • mel_filter_bank()
      • mel_to_hertz()
      • power_to_db()
      • spectrogram()
      • window_function()
    • hezar.utils.common_utils module
      • colorize_text()
      • exec_timer
      • get_parents()
      • is_text_valid()
      • is_url()
      • permute_dict_list()
      • reverse_string_digits()
      • sanitize_function_parameters()
      • snake_case()
    • hezar.utils.data_utils module
      • convert_batch_dict_dtype()
      • dataloader_worker_init_fn()
      • flatten_dict()
      • get_non_numeric_keys()
      • pad_batch_items()
      • resolve_inputs_length_for_padding()
      • set_seed()
      • shift_tokens_right()
      • torch2numpy()
    • hezar.utils.hub_utils module
      • clean_cache()
      • clone_repo()
      • exists_in_cache()
      • exists_on_hub()
      • get_local_cache_path()
      • get_state_dict_from_hub()
      • list_repo_files()
    • hezar.utils.image_utils module
      • convert_image_type()
      • crop_boxes()
      • draw_boxes()
      • find_channels_axis_side()
      • gray_scale_image()
      • load_image()
      • mirror_image()
      • normalize_image()
      • pad_boxes()
      • rescale_image()
      • resize_image()
      • save_image()
      • show_image()
      • transpose_channels_axis_side()
    • hezar.utils.integration_utils module
      • check_dependencies()
      • get_lib_version()
      • is_backend_available()
      • verify_dependencies()
    • hezar.utils.logging module
      • Logger
    • hezar.utils.registry_utils module
      • get_module_class()
      • get_module_config_class()
      • get_registry_key_by_module_class()
      • get_registry_point()
      • list_available_datasets()
      • list_available_embeddings()
      • list_available_metrics()
      • list_available_models()
      • list_available_preprocessors()
  • Module contents

Submodules

• hezar.builders module
  • build_dataset()
  • build_embedding()
  • build_metric()
  • build_model()
  • build_preprocessor()
• hezar.configs module
  • Config
    • Config.config_type
    • Config.dict()
    • Config.fields()
    • Config.from_dict()
    • Config.get()
    • Config.keys()
    • Config.load()
    • Config.name
    • Config.push_to_hub()
    • Config.save()
    • Config.update()
  • DatasetConfig
    • DatasetConfig.config_type
    • DatasetConfig.hf_load_kwargs
    • DatasetConfig.max_size
    • DatasetConfig.name
    • DatasetConfig.path
    • DatasetConfig.task
  • EmbeddingConfig
    • EmbeddingConfig.bypass_version_check
    • EmbeddingConfig.config_type
    • EmbeddingConfig.name
  • MetricConfig
    • MetricConfig.config_type
    • MetricConfig.n_decimals
    • MetricConfig.name
    • MetricConfig.objective
    • MetricConfig.output_keys
  • ModelConfig
    • ModelConfig.config_type
    • ModelConfig.name
  • PreprocessorConfig
    • PreprocessorConfig.config_type
    • PreprocessorConfig.name
  • TrainerConfig
    • TrainerConfig.batch_size
    • TrainerConfig.checkpoints_dir
    • TrainerConfig.config_type
    • TrainerConfig.dataloader_shuffle
    • TrainerConfig.device
    • TrainerConfig.distributed
    • TrainerConfig.do_evaluate
    • TrainerConfig.eval_batch_size
    • TrainerConfig.evaluate_with_generate
    • TrainerConfig.gradient_accumulation_steps
    • TrainerConfig.init_weights_from
    • TrainerConfig.learning_rate
    • TrainerConfig.log_steps
    • TrainerConfig.logs_dir
    • TrainerConfig.lr_scheduler
    • TrainerConfig.lr_scheduler_kwargs
    • TrainerConfig.lr_scheduling_steps
    • TrainerConfig.max_steps
    • TrainerConfig.metric_for_best_model
    • TrainerConfig.metrics
    • TrainerConfig.mixed_precision
    • TrainerConfig.name
    • TrainerConfig.num_dataloader_workers
    • TrainerConfig.num_epochs
    • TrainerConfig.optimizer
    • TrainerConfig.output_dir
    • TrainerConfig.resume_from_checkpoint
    • TrainerConfig.save_enabled
    • TrainerConfig.save_freq
    • TrainerConfig.save_steps
    • TrainerConfig.seed
    • TrainerConfig.task
    • TrainerConfig.use_cpu
    • TrainerConfig.weight_decay
• hezar.constants module
  • Backends
    • Backends.ACCELERATE
    • Backends.DATASETS
    • Backends.GENSIM
    • Backends.JIWER
    • Backends.LIBROSA
    • Backends.NLTK
    • Backends.OPENCV
    • Backends.PILLOW
    • Backends.PYTORCH
    • Backends.ROUGE
    • Backends.SCIKIT
    • Backends.SEQEVAL
    • Backends.SOUNDFILE
    • Backends.TENSORBOARD
    • Backends.TOKENIZERS
    • Backends.TORCHVISION
    • Backends.TRANSFORMERS
    • Backends.WANDB
  • ChannelsAxisSide
    • ChannelsAxisSide.FIRST
    • ChannelsAxisSide.LAST
  • Color
    • Color.BLUE
    • Color.BOLD
    • Color.CYAN
    • Color.GREEN
    • Color.GREY
    • Color.HEADER
    • Color.ITALIC
    • Color.NORMAL
    • Color.RED
    • Color.UNDERLINE
    • Color.YELLOW
  • ConfigType
    • ConfigType.BASE
    • ConfigType.CRITERION
    • ConfigType.DATASET
    • ConfigType.EMBEDDING
    • ConfigType.LR_SCHEDULER
    • ConfigType.METRIC
    • ConfigType.MODEL
    • ConfigType.OPTIMIZER
    • ConfigType.PREPROCESSOR
    • ConfigType.TRAINER
  • ExplicitEnum
    • ExplicitEnum.list()
  • ImageType
    • ImageType.NUMPY
    • ImageType.PILLOW
    • ImageType.TORCH
  • LRSchedulerType
    • LRSchedulerType.CONSTANT
    • LRSchedulerType.COSINE_ANEALING
    • LRSchedulerType.CYCLIC
    • LRSchedulerType.EXPONENTIAL
    • LRSchedulerType.LAMBDA
    • LRSchedulerType.LINEAR
    • LRSchedulerType.MULTI_STEP
    • LRSchedulerType.ONE_CYCLE
    • LRSchedulerType.POLYNOMIAL
    • LRSchedulerType.REDUCE_ON_PLATEAU
    • LRSchedulerType.SEQUENTIAL
    • LRSchedulerType.STEP
  • LossType
    • LossType.BCE
    • LossType.BCE_WITH_LOGITS
    • LossType.CROSS_ENTROPY
    • LossType.CTC
    • LossType.GAUSSIAN_NLL
    • LossType.L1
    • LossType.MSE
    • LossType.NLL
    • LossType.NLL_2D
    • LossType.POISSON_NLL
    • LossType.TRIPLE_MARGIN
  • MetricType
    • MetricType.ACCURACY
    • MetricType.BLEU
    • MetricType.CER
    • MetricType.F1
    • MetricType.PRECISION
    • MetricType.RECALL
    • MetricType.ROUGE
    • MetricType.SEQEVAL
    • MetricType.WER
  • OptimizerType
    • OptimizerType.ADAM
    • OptimizerType.ADAMW
    • OptimizerType.SDG
  • PaddingType
    • PaddingType.LONGEST
    • PaddingType.MAX_LENGTH
  • PrecisionType
    • PrecisionType.BF16
    • PrecisionType.FP16
    • PrecisionType.FP8
    • PrecisionType.NO
  • RegistryType
    • RegistryType.CRITERION
    • RegistryType.DATASET
    • RegistryType.EMBEDDING
    • RegistryType.LR_SCHEDULER
    • RegistryType.METRIC
    • RegistryType.MODEL
    • RegistryType.OPTIMIZER
    • RegistryType.PREPROCESSOR
    • RegistryType.TRAINER
  • RepoType
    • RepoType.DATASET
    • RepoType.MODEL
  • SplitType
    • SplitType.EVAL
    • SplitType.TEST
    • SplitType.TRAIN
    • SplitType.VALID
  • TaskType
    • TaskType.AUDIO_CLASSIFICATION
    • TaskType.BACKBONE
    • TaskType.IMAGE2TEXT
    • TaskType.LANGUAGE_MODELING
    • TaskType.MASK_FILLING
    • TaskType.SEQUENCE_LABELING
    • TaskType.SPEECH_RECOGNITION
    • TaskType.TEXT_CLASSIFICATION
    • TaskType.TEXT_DETECTION
    • TaskType.TEXT_GENERATION
• hezar.registry module
  • Registry
    • Registry.config_class
    • Registry.description
    • Registry.module_class
  • register_dataset()
  • register_embedding()
  • register_metric()
  • register_model()
  • register_preprocessor()

Module contents