This class can be used to create a list of [`LogitsProcessor`] or [`LogitsWarper`] to subsequently process a `scores` input tensor. This clas

This class can be used to create a list of [`LogitsProcessor`] or [`LogitsWarper`] to subsequently process a `scores` input tensor. This clas

This class can be used to create a list of [`TFLogitsProcessor`] to subsequently process a `scores` input tensor. This class inherits from li

This class can be used to create a list of [`TFLogitsProcessor`] to subsequently process a `scores` input tensor. This class inherits from li

r""" [`BeamScorer`] implementing standard beam search decoding. Adapted in part from [Facebook's XLM beam search code](https://github.com

Transformer decoder consisting of *config.num_hidden_layers* layers. Each layer is a [`LlamaDecoderLayer`] Args: config: LlamaConfig

r""" [`LogitsWarper`] and [`LogitsProcessor`] for normalizing the scores using log-softmax. It's important to normalize the scores during beam

r""" [`LogitsWarper`] that performs top-k, i.e. restricting to the k highest probability elements. Often used together with [`TemperatureLogit

[`LogitsWarper`] that performs top-p, i.e. restricting to top tokens summing to prob_cut_off <= prob_cut_off. Often used together with [`Temp

Base class for outputs of decoder-only generation models using beam search. Args: sequences (`torch.LongTensor` of shape `(batch_siz

Base class for outputs of decoder-only generation models using beam search. Args: sequences (`torch.LongTensor` of shape `(batch_siz

Base class for outputs of decoder-only generation models using beam search. Args: sequences (`torch.LongTensor` of shape `(batch_siz

Base class for outputs of encoder-decoder generation models using beam search. Hidden states and attention weights of the decoder (respective

Base class for outputs of encoder-decoder generation models using beam search. Hidden states and attention weights of the decoder (respective

Base class for outputs of encoder-decoder generation models using beam search. Hidden states and attention weights of the decoder (respective

r""" [`BeamScorer`] implementing standard beam search decoding. Adapted in part from [Facebook's XLM beam search code](https://github.com

r""" A special [`Constraint`] that is fulfilled by fulfilling just one of several constraints. Args: nested_token_ids (`List[List[int

This class can be used to create a list of [`FlaxLogitsProcessor`] or [`FlaxLogitsWarper`] to subsequently process a `scores` input tensor. T

This class can be used to create a list of [`FlaxLogitsProcessor`] or [`FlaxLogitsWarper`] to subsequently process a `scores` input tensor. T

Transformer decoder consisting of *config.num_hidden_layers* layers. Each layer is a [`Gemma2DecoderLayer`] Args: config: Gemma2Conf

Transformer decoder consisting of *config.num_hidden_layers* layers. Each layer is a [`Gemma2DecoderLayer`] Args: config: Gemma2Conf

r""" Class that holds a configuration for a generation task. A `generate` call supports the following generation methods for text-decoder, tex

Transformer decoder consisting of *config.num_hidden_layers* layers. Each layer is a [`LlamaDecoderLayer`] Args: config: LlamaConfig

Transformer decoder consisting of *config.num_hidden_layers* layers. Each layer is a [`LlamaDecoderLayer`] Args: config: LlamaConfig

This class can be used to stop generation whenever the full generated number of tokens exceeds `max_length`. Keep in mind for decoder-only ty

Transformer decoder consisting of *config.num_hidden_layers* layers. Each layer is a [`MistralDecoderLayer`] Args: config: MistralCo

Transformer decoder consisting of *config.num_hidden_layers* layers. Each layer is a [`MistralDecoderLayer`] Args: config: MistralCo

r""" [`Constraint`] enforcing that an ordered sequence of tokens is included in the output. Args: token_ids (`List[int]`):

Transformer decoder consisting of *config.num_hidden_layers* layers. Each layer is a [`MistralDecoderLayer`] Args: config: MistralCo

Transformer decoder consisting of *config.num_hidden_layers* layers. Each layer is a [`MistralDecoderLayer`] Args: config: MistralCo

r""" [`BeamScorer`] implementing constrained beam search decoding. Args: batch_size (`int`): Batch Size of `input_ids` f

r""" A class for beam scorers to track its progress through a list of constraints. Args: constraints (`List[Constraint]`):

r""" A class for beam scorers to track its progress through a list of constraints. Args: constraints (`List[Constraint]`):

r""" A special [`Constraint`] that is fulfilled by fulfilling just one of several constraints. Args: nested_token_ids (`List[List[int

r""" [`LogitsProcessor`] that works similarly to [`NoRepeatNGramLogitsProcessor`], but applied exclusively to prevent the repetition of n-gram

r""" [`LogitsProcessor`] that works similarly to [`RepetitionPenaltyLogitsProcessor`], but with an *inverse* penalty that is applied to the to

r""" [`LogitsWarper`] that performs epsilon-sampling, i.e. restricting to tokens with `prob >= epsilon`. Takes the largest min_tokens_to_keep

r""" [`LogitsWarper`] that performs eta-sampling, a technique to filter out tokens with probabilities below a dynamic cutoff value, `eta`, whi

r""" [`LogitsProcessor`] that exponentially increases the score of the `eos_token_id` after `start_index` has been reached. This allows genera

r""" This processor takes a list of pairs of integers which indicates a mapping from generation indices to token indices that will be forced b

r""" [`LogitsProcessor`] that enforces the specified token as the first generated token. Used with encoder-decoder models. Args:

r""" [`LogitsProcessor`] that enforces the specified token as the last generated token when `max_length` is reached. Args: max_length

r""" Class that holds a configuration for a generation task. A `generate` call supports the following generation methods for text-decoder, tex

Base class for outputs of decoder-only generation models using greedy search. Args: sequences (`torch.LongTensor` of shape `(batch_

Base class for outputs of encoder-decoder generation models using greedy search. Hidden states and attention weights of the decoder (respecti

r""" [`LogitsProcessor`] that enforces diverse beam search. Note that this logits processor is only effective for [`PreTrainedModel.group_bea

r""" [`LogitsProcessor`] that removes all `nan` and `inf` values to avoid the generation method to fail. Note that using the logits processor

r""" [`LogitsWarper`] and [`LogitsProcessor`] for normalizing the scores using log-softmax. It's important to normalize the scores during beam

This class can be used to stop generation whenever the full generated number of tokens exceeds `max_length`. Keep in mind for decoder-only ty

This class can be used to stop generation whenever the full generation exceeds some amount of time. By default, the time will start being cou

r""" [`LogitsProcessor`] enforcing a min-length by setting EOS probability to 0. Note that, for decoder-only models like most LLMs, the length

r""" [`LogitsProcessor`] enforcing a min-length of new tokens by setting EOS (End-Of-Sequence) token probability to 0. Contrarily to [`MinLeng

[`LogitsProcessor`] that enforces that specified sequences will never be selected. <Tip> In order to get the token ids of the words tha

r""" N-grams are groups of "n" consecutive words, characters, or tokens taken from a sequence of text. Given the sentence: "She runs fast", th

r""" [`Constraint`] enforcing that an ordered sequence of tokens is included in the output. Args: token_ids (`List[int]`):

r""" [`LogitsProcessor`] that enforces constrained generation and is useful for prefix-conditioned constrained generation. See [Autoregressive

Transformer decoder consisting of *config.num_hidden_layers* layers. Each layer is a [`Qwen2DecoderLayer`] Args: config: Qwen2Config

r""" [`LogitsProcessor`] that prevents the repetition of previous tokens through a penalty. This penalty is applied at most once per token. No

r""" [`SuppressTokensAtBeginLogitsProcessor`] supresses a list of tokens as soon as the `generate` function starts generating using `begin_ind

r""" This processor can be used to suppress a list of tokens. The processor will set their log probs to `-inf` so that they are not generated.

r""" [`LogitsWarper`] for temperature (exponential scaling output probability distribution), which effectively means that it can control the r

r""" [`LogitsWarper`] that performs top-k, i.e. restricting to the k highest probability elements. Often used together with [`TemperatureLogit

[`LogitsWarper`] that performs top-p, i.e. restricting to top tokens summing to prob_cut_off <= prob_cut_off. Often used together with [`Temp

r""" [`LogitsWarper`] that performs typical decoding. Inspired on how humans use language, it prioritizes tokens whose log probability is clos

Base class for outputs of decoder-only generation models using beam sample. Args: sequences (`torch.LongTensor` of shape `(batch_siz

Base class for outputs of decoder-only generation models using beam sample. Args: sequences (`torch.LongTensor` of shape `(batch_siz

Base class for outputs of decoder-only generation models using beam sample. Args: sequences (`torch.LongTensor` of shape `(batch_siz

Base class for outputs of encoder-decoder generation models using beam sampling. Hidden states and attention weights of the decoder (respecti

Base class for outputs of encoder-decoder generation models using beam sampling. Hidden states and attention weights of the decoder (respecti

Base class for outputs of encoder-decoder generation models using beam sampling. Hidden states and attention weights of the decoder (respecti