indicate package

Submodules

indicate.base module

indicate.decoder module

class indicate.decoder.Decoder(*args, **kwargs)[source]

Bases: Model

__init__(vocab_size: int, embedding_dim: int, dec_units: int, batch_sz: int, max_length_input: int, max_length_output: int, attention_type: str = 'luong') → None[source]

build_attention_mechanism(memory: Tensor | None) → Layer[source]

call(inputs: Tensor, initial_state: List[Tensor], encoder_outputs: Tensor) → Tuple[Tensor, Tuple[Tensor, Tensor], Tensor][source]

build_initial_state(batch_sz: int, encoder_state: List[Tensor]) → List[Tensor][source]

indicate.encoder module

class indicate.encoder.Encoder(*args, **kwargs)[source]

Bases: Model

__init__(vocab_size: int, embedding_dim: int, enc_units: int, batch_sz: int) → None[source]

call(x: Tensor, hidden: List[Tensor], training: bool = None) → Tuple[Tensor, Tensor, Tensor][source]

initialize_hidden_state() → List[Tensor][source]

indicate.hindi2english module

class indicate.hindi2english.HindiToEnglish[source]

Bases: object

MODELFN: str = 'data/model/hindi_to_english/saved_weights/'

INPUT_VOCAB: str = 'data/model/hindi_to_english/hindi_tokens.json'

TARGET_VOCAB: str = 'data/model/hindi_to_english/english_tokens.json'

embedding_dim: int = 256

units: int = 1024

BATCH_SIZE: int = 64

BUFFER_SIZE: int = 120000

max_length_input: int = 47

max_length_output: int = 173

START_TOKEN: str = '^'

END_TOKEN: str = '$'

input_lang_tokenizer: Any | None = None

target_lang_tokenizer: Any | None = None

encoder: Encoder | None = None

decoder: Decoder | None = None

classmethod get_model_path() → str[source]

classmethod get_input_vocab() → str[source]

classmethod get_target_vocab() → str[source]

classmethod transliterate(input: str) → str[source]

Transliterate from Hindi to English.

Parameters:

input (str) – Hindi text

Returns:

English text

Return type:

output (str)

Raises:

ValueError – If input is empty or not a string
RuntimeError – If model loading fails

indicate.logging module

indicate.logging.get_logger() → Logger[source]

indicate.transliterate module

indicate.transliterate.main(argv=['-M', 'html', 'source', 'build'])[source]

indicate.utils module

indicate.utils.sequence_to_chars(tokenizer: Any, sequence: Tensor) → str[source]: Convert a sequence of indices back to characters.

indicate.utils.evaluate_sentence(sentence: str, units: int, input_lang_tokenizer: Any, target_lang_tokenizer: Any, encoder: Any, decoder: Any, max_length_input: int) → Tensor[source]: Evaluate/translate a single sentence.

indicate.utils.translate(sentence: str, units: int, input_lang_tokenizer: Any, target_lang_tokenizer: Any, encoder: Any, decoder: Any, max_length_input: int) → str[source]: Translate a sentence from source to target language.

Module contents

indicate.hindi2english(input: str) → str