image-match/python/py/Lib/site-packages/transformers/models/lasr/configuration_lasr.py

#                🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨
#           This file was automatically generated from src/transformers/models/lasr/modular_lasr.py.
#               Do NOT edit this file manually as any edits will be overwritten by the generation of
#             the file from the modular. If any change should be done, please apply the change to the
#                          modular_lasr.py file directly. One of our CI enforces this.
#                🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨
# Copyright 2025 The HuggingFace Inc. team and Google LLC. All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

from huggingface_hub.dataclasses import strict

from ...configuration_utils import PreTrainedConfig
from ...utils import auto_docstring


@auto_docstring(checkpoint="google/medasr")
@strict
class LasrEncoderConfig(PreTrainedConfig):
    r"""
    convolution_bias (`bool`, *optional*, defaults to `False`):
        Whether to use bias in convolutions of the conformer's convolution module.
    conv_kernel_size (`int`, *optional*, defaults to 32):
        The kernel size of the convolution layers in the Conformer block.
    subsampling_conv_channels (`int`, *optional*, defaults to 256):
        The number of channels in the subsampling convolution layers.
    subsampling_conv_kernel_size (`int`, *optional*, defaults to 5):
        The kernel size of the subsampling convolution layers.
    subsampling_conv_stride (`int`, *optional*, defaults to 2):
        The stride of the subsampling convolution layers.
    dropout_positions (`float`, *optional*, defaults to 0.0):
        The dropout ratio for the positions in the input sequence.
    feed_forward_residual_weights (`tuple[float, float]`, *optional*, defaults to `[1.5, 0.5]`):
        The residual weights for the feed forward layers.
    conv_residual_weights (`tuple[float, float]`, *optional*, defaults to `[2.0, 1.0]`):
        The residual weights for the convolution layers.
    batch_norm_momentum (`float`, *optional*, defaults to 0.01):
        The momentum for the batch normalization layers

    Example:
        ```python
        >>> from transformers import LasrEncoderModel, LasrEncoderConfig

        >>> # Initializing a `LasrEncoder` configuration
        >>> configuration = LasrEncoderConfig()

        >>> # Initializing a model from the configuration
        >>> model = LasrEncoderModel(configuration)

        >>> # Accessing the model configuration
        >>> configuration = model.config
        ```

    This configuration class is based on the LasrEncoder architecture from Google Health AI. You can find more details
    and pre-trained models at [TODO/TODO](https://huggingface.co/TODO/TODO).
    """

    model_type = "lasr_encoder"
    keys_to_ignore_at_inference = ["past_key_values"]

    hidden_size: int = 512
    num_hidden_layers: int = 17
    num_attention_heads: int = 8
    intermediate_size: int = 2048
    hidden_act: str = "silu"
    attention_bias: bool = False
    convolution_bias: bool = False
    conv_kernel_size: int = 32
    subsampling_conv_channels: int = 256
    num_mel_bins: int = 128
    subsampling_conv_kernel_size: int = 5
    subsampling_conv_stride: int = 2
    dropout: float | int = 0.1
    dropout_positions: float | int = 0.0
    layerdrop: float | int = 0.1
    activation_dropout: float | int = 0.1
    attention_dropout: float | int = 0.1
    max_position_embeddings: int = 10000
    initializer_range: float = 0.02
    layer_norm_eps: float = 1e-6
    feed_forward_residual_weights: list[float] | tuple[float, ...] = (1.5, 0.5)
    conv_residual_weights: list[float] | tuple[float, ...] = (2.0, 1.0)
    batch_norm_momentum: float = 0.01
    rope_parameters: dict | None = None

    def __post_init__(self, **kwargs):
        self.num_key_value_heads = self.num_attention_heads
        super().__post_init__(**kwargs)


@auto_docstring(checkpoint="google/medasr")
@strict
class LasrCTCConfig(PreTrainedConfig):
    r"""
    ctc_loss_reduction (`str`, *optional*, defaults to `"mean"`):
        Specifies the reduction to apply to the output of `torch.nn.CTCLoss`. Only relevant when training an
        instance of [`LasrForCTC`].
    ctc_zero_infinity (`bool`, *optional*, defaults to `True`):
        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 [`LasrForCTC`].

    Example:
        ```python
        >>> from transformers import LasrForCTC, LasrCTCConfig
        >>> # Initializing a Lasr configuration
        >>> configuration = LasrCTCConfig()
        >>> # Initializing a model from the configuration
        >>> model = LasrForCTC(configuration)
        >>> # Accessing the model configuration
        >>> configuration = model.config
        ```
    This configuration class is based on the Lasr CTC architecture from Google Health AI. You can find more details
    and pre-trained models at [TODO/TODO](https://huggingface.co/TODO/TODO).
    """

    model_type = "lasr_ctc"
    sub_configs = {"encoder_config": LasrEncoderConfig}

    vocab_size: int = 512
    ctc_loss_reduction: str = "mean"
    ctc_zero_infinity: bool = True
    encoder_config: dict | PreTrainedConfig | None = None
    pad_token_id: int = 0

    def __post_init__(self, **kwargs):
        if isinstance(self.encoder_config, dict):
            self.encoder_config = LasrEncoderConfig(**self.encoder_config)
        elif self.encoder_config is None:
            self.encoder_config = LasrEncoderConfig()
        self.initializer_range = self.encoder_config.initializer_range
        super().__post_init__(**kwargs)

    @property
    def inputs_to_logits_ratio(self):
        return self.encoder_config.subsampling_conv_stride**2


__all__ = ["LasrEncoderConfig", "LasrCTCConfig"]