Generation

utils/generation

Classes, functions, and utilities for generation.

Todo

utils/generation.LogitsProcessorList ⇐ <code> Callable </code>

A class representing a list of logits processors. A logits processor is a function that modifies the logits output of a language model. This class provides methods for adding new processors and applying all processors to a batch of logits.

Kind: static class of utils/generation Extends: Callable

new LogitsProcessorList()

Constructs a new instance of LogitsProcessorList.

logitsProcessorList.push(item)

Adds a new logits processor to the list.

Kind: instance method of LogitsProcessorList

Param
Type
Description

item

LogitsProcessor

The logits processor function to add.

logitsProcessorList.extend(items)

Adds multiple logits processors to the list.

Kind: instance method of LogitsProcessorList

Param
Type
Description

items

Array.<LogitsProcessor>

The logits processor functions to add.

logitsProcessorList._call(input_ids, batchedLogits)

Applies all logits processors in the list to a batch of logits, modifying them in-place.

Kind: instance method of LogitsProcessorList

Param
Type
Description

input_ids

Array.<number>

The input IDs for the language model.

batchedLogits

Array.<Array<number>>

A 2D array of logits, where each row corresponds to a single input sequence in the batch.

utils/generation.LogitsProcessor ⇐ <code> Callable </code>

Base class for processing logits.

Kind: static class of utils/generation Extends: Callable

logitsProcessor._call(input_ids, logits)

Apply the processor to the input logits.

Kind: instance abstract method of LogitsProcessor Throws:

  • Error Throws an error if `_call` is not implemented in the subclass.

Param
Type
Description

input_ids

Array

The input ids.

logits

Tensor

The logits to process.

utils/generation.ForceTokensLogitsProcessor ⇐ <code> LogitsProcessor </code>

A logits processor that forces a specific token to be generated by the decoder.

Kind: static class of utils/generation Extends: LogitsProcessor

new ForceTokensLogitsProcessor(forced_decoder_ids)

Constructs a new instance of ForceTokensLogitsProcessor.

Param
Type
Description

forced_decoder_ids

Array

The ids of tokens that should be forced.

forceTokensLogitsProcessor._call(input_ids, logits) ⇒ <code> Tensor </code>

Apply the processor to the input logits.

Kind: instance method of ForceTokensLogitsProcessor Returns: Tensor - The processed logits.

Param
Type
Description

input_ids

Array

The input ids.

logits

Tensor

The logits to process.

utils/generation.ForcedBOSTokenLogitsProcessor ⇐ <code> LogitsProcessor </code>

A LogitsProcessor that forces a BOS token at the beginning of the generated sequence.

Kind: static class of utils/generation Extends: LogitsProcessor

new ForcedBOSTokenLogitsProcessor(bos_token_id)

Create a ForcedBOSTokenLogitsProcessor.

Param
Type
Description

bos_token_id

number

The ID of the beginning-of-sequence token to be forced.

forcedBOSTokenLogitsProcessor._call(input_ids, logits) ⇒ <code> Object </code>

Apply the BOS token forcing to the logits.

Kind: instance method of ForcedBOSTokenLogitsProcessor Returns: Object - The logits with BOS token forcing.

Param
Type
Description

input_ids

Array

The input IDs.

logits

Object

The logits.

utils/generation.ForcedEOSTokenLogitsProcessor ⇐ <code> LogitsProcessor </code>

A logits processor that forces end-of-sequence token probability to 1.

Kind: static class of utils/generation Extends: LogitsProcessor

new ForcedEOSTokenLogitsProcessor(max_length, forced_eos_token_id)

Create a ForcedEOSTokenLogitsProcessor.

Param
Type
Description

max_length

number

Max length of the sequence.

forced_eos_token_id

number | Array<number>

The ID of the end-of-sequence token to be forced.

forcedEOSTokenLogitsProcessor._call(input_ids, logits)

Apply the processor to input_ids and logits.

Kind: instance method of ForcedEOSTokenLogitsProcessor

Param
Type
Description

input_ids

Array.<number>

The input ids.

logits

Tensor

The logits tensor.

utils/generation.SuppressTokensAtBeginLogitsProcessor ⇐ <code> LogitsProcessor </code>

A LogitsProcessor that suppresses a list of tokens as soon as the generate function starts generating using begin_index tokens. This should ensure that the tokens defined by begin_suppress_tokens at not sampled at the begining of the generation.

Kind: static class of utils/generation Extends: LogitsProcessor

new SuppressTokensAtBeginLogitsProcessor(begin_suppress_tokens, begin_index)

Create a SuppressTokensAtBeginLogitsProcessor.

Param
Type
Description

begin_suppress_tokens

Array.<number>

The IDs of the tokens to suppress.

begin_index

number

The number of tokens to generate before suppressing tokens.

suppressTokensAtBeginLogitsProcessor._call(input_ids, logits) ⇒ <code> Object </code>

Apply the BOS token forcing to the logits.

Kind: instance method of SuppressTokensAtBeginLogitsProcessor Returns: Object - The logits with BOS token forcing.

Param
Type
Description

input_ids

Array

The input IDs.

logits

Object

The logits.

utils/generation.WhisperTimeStampLogitsProcessor ⇐ <code> LogitsProcessor </code>

A LogitsProcessor that handles adding timestamps to generated text.

Kind: static class of utils/generation Extends: LogitsProcessor

new WhisperTimeStampLogitsProcessor(generate_config)

Constructs a new WhisperTimeStampLogitsProcessor.

Param
Type
Description

generate_config

Object

The config object passed to the generate() method of a transformer model.

generate_config.eos_token_id

number

The ID of the end-of-sequence token.

generate_config.no_timestamps_token_id

number

The ID of the token used to indicate that a token should not have a timestamp.

[generate_config.forced_decoder_ids]

Array.<Array<number>>

An array of two-element arrays representing decoder IDs that are forced to appear in the output. The second element of each array indicates whether the token is a timestamp.

[generate_config.max_initial_timestamp_index]

number

The maximum index at which an initial timestamp can appear.

whisperTimeStampLogitsProcessor._call(input_ids, logits) ⇒ <code> Tensor </code>

Modify the logits to handle timestamp tokens.

Kind: instance method of WhisperTimeStampLogitsProcessor Returns: Tensor - The modified logits.

Param
Type
Description

input_ids

Array

The input sequence of tokens.

logits

Tensor

The logits output by the model.

utils/generation.NoRepeatNGramLogitsProcessor ⇐ <code> LogitsProcessor </code>

A logits processor that disallows ngrams of a certain size to be repeated.

Kind: static class of utils/generation Extends: LogitsProcessor

new NoRepeatNGramLogitsProcessor(no_repeat_ngram_size)

Create a NoRepeatNGramLogitsProcessor.

Param
Type
Description

no_repeat_ngram_size

number

The no-repeat-ngram size. All ngrams of this size can only occur once.

noRepeatNGramLogitsProcessor.getNgrams(prevInputIds) ⇒ <code> Map. < string, Array < number > > </code>

Generate n-grams from a sequence of token ids.

Kind: instance method of NoRepeatNGramLogitsProcessor Returns: Map.<string, Array<number>> - Map of generated n-grams

Param
Type
Description

prevInputIds

Array.<number>

List of previous input ids

noRepeatNGramLogitsProcessor.getGeneratedNgrams(bannedNgrams, prevInputIds) ⇒ <code> Array. < number > </code>

Generate n-grams from a sequence of token ids.

Kind: instance method of NoRepeatNGramLogitsProcessor Returns: Array.<number> - Map of generated n-grams

Param
Type
Description

bannedNgrams

Map.<string, Array<number>>

Map of banned n-grams

prevInputIds

Array.<number>

List of previous input ids

noRepeatNGramLogitsProcessor.calcBannedNgramTokens(prevInputIds) ⇒ <code> Array. < number > </code>

Calculate banned n-gram tokens

Kind: instance method of NoRepeatNGramLogitsProcessor Returns: Array.<number> - Map of generated n-grams

Param
Type
Description

prevInputIds

Array.<number>

List of previous input ids

noRepeatNGramLogitsProcessor._call(input_ids, logits) ⇒ <code> Object </code>

Apply the no-repeat-ngram processor to the logits.

Kind: instance method of NoRepeatNGramLogitsProcessor Returns: Object - The logits with no-repeat-ngram processing.

Param
Type
Description

input_ids

Array

The input IDs.

logits

Object

The logits.

utils/generation.RepetitionPenaltyLogitsProcessor ⇐ <code> LogitsProcessor </code>

A logits processor that penalises repeated output tokens.

Kind: static class of utils/generation Extends: LogitsProcessor

new RepetitionPenaltyLogitsProcessor(penalty)

Create a RepetitionPenaltyLogitsProcessor.

Param
Type
Description

penalty

number

The penalty to apply for repeated tokens.

repetitionPenaltyLogitsProcessor._call(input_ids, logits) ⇒ <code> Object </code>

Apply the repetition penalty to the logits.

Kind: instance method of RepetitionPenaltyLogitsProcessor Returns: Object - The logits with repetition penalty processing.

Param
Type
Description

input_ids

Array

The input IDs.

logits

Object

The logits.

utils/generation.MinLengthLogitsProcessor ⇐ <code> LogitsProcessor </code>

A logits processor that enforces a minimum number of tokens.

Kind: static class of utils/generation Extends: LogitsProcessor

new MinLengthLogitsProcessor(min_length, eos_token_id)

Create a MinLengthLogitsProcessor.

Param
Type
Description

min_length

number

The minimum length below which the score of eos_token_id is set to negative infinity.

eos_token_id

number | Array<number>

The ID/IDs of the end-of-sequence token.

minLengthLogitsProcessor._call(input_ids, logits) ⇒ <code> Object </code>

Apply logit processor.

Kind: instance method of MinLengthLogitsProcessor Returns: Object - The processed logits.

Param
Type
Description

input_ids

Array

The input IDs.

logits

Object

The logits.

utils/generation.MinNewTokensLengthLogitsProcessor ⇐ <code> LogitsProcessor </code>

A logits processor that enforces a minimum number of new tokens.

Kind: static class of utils/generation Extends: LogitsProcessor

new MinNewTokensLengthLogitsProcessor(prompt_length_to_skip, min_new_tokens, eos_token_id)

Create a MinNewTokensLengthLogitsProcessor.

Param
Type
Description

prompt_length_to_skip

number

The input tokens length.

min_new_tokens

number

The minimum new tokens length below which the score of eos_token_id is set to negative infinity.

eos_token_id

number | Array<number>

The ID/IDs of the end-of-sequence token.

minNewTokensLengthLogitsProcessor._call(input_ids, logits) ⇒ <code> Object </code>

Apply logit processor.

Kind: instance method of MinNewTokensLengthLogitsProcessor Returns: Object - The processed logits.

Param
Type
Description

input_ids

Array

The input IDs.

logits

Object

The logits.

utils/generation.GenerationConfig

Class that holds a configuration for a generation task.

Kind: static class of utils/generation

new GenerationConfig([kwargs])

Create a GenerationConfig object

Param
Type
Default
Description

[kwargs]

Object

{}

The configuration parameters. If not set, the default values are used.

[kwargs.max_length]

number

20

The maximum length the generated tokens can have. Corresponds to the length of the input prompt + max_new_tokens. Its effect is overridden by max_new_tokens, if also set.

[kwargs.max_new_tokens]

number

The maximum numbers of tokens to generate, ignoring the number of tokens in the prompt.

[kwargs.min_length]

number

0

The minimum length of the sequence to be generated. Corresponds to the length of the input prompt + min_new_tokens. Its effect is overridden by min_new_tokens, if also set.

[kwargs.min_new_tokens]

number

The minimum numbers of tokens to generate, ignoring the number of tokens in the prompt.

[kwargs.early_stopping]

boolean | "never"

false

Controls the stopping condition for beam-based methods, like beam-search. It accepts the following values:

  • true, where the generation stops as soon as there are num_beams complete candidates;

  • false, where an heuristic is applied and the generation stops when is it very unlikely to find better candidates;

  • "never", where the beam search procedure only stops when there cannot be better candidates (canonical beam search algorithm).

[kwargs.max_time]

number

The maximum amount of time you allow the computation to run for in seconds. Generation will still finish the current pass after allocated time has been passed.

[kwargs.do_sample]

boolean

false

Whether or not to use sampling; use greedy decoding otherwise.

[kwargs.num_beams]

number

1

Number of beams for beam search. 1 means no beam search.

[kwargs.num_beam_groups]

number

1

Number of groups to divide num_beams into in order to ensure diversity among different groups of beams. See this paper for more details.

[kwargs.penalty_alpha]

number

The values balance the model confidence and the degeneration penalty in contrastive search decoding.

[kwargs.use_cache]

boolean

true

Whether or not the model should use the past last key/values attentions (if applicable to the model) to speed up decoding.

[kwargs.temperature]

number

1.0

The value used to modulate the next token probabilities.

[kwargs.top_k]

number

50

The number of highest probability vocabulary tokens to keep for top-k-filtering.

[kwargs.top_p]

number

1.0

If set to float < 1, only the smallest set of most probable tokens with probabilities that add up to top_p or higher are kept for generation.

[kwargs.typical_p]

number

1.0

Local typicality measures how similar the conditional probability of predicting a target token next is to the expected conditional probability of predicting a random token next, given the partial text already generated. If set to float < 1, the smallest set of the most locally typical tokens with probabilities that add up to typical_p or higher are kept for generation. See this paper for more details.

[kwargs.epsilon_cutoff]

number

0.0

If set to float strictly between 0 and 1, only tokens with a conditional probability greater than epsilon_cutoff will be sampled. In the paper, suggested values range from 3e-4 to 9e-4, depending on the size of the model. See Truncation Sampling as Language Model Desmoothing for more details.

[kwargs.eta_cutoff]

number

0.0

Eta sampling is a hybrid of locally typical sampling and epsilon sampling. If set to float strictly between 0 and 1, a token is only considered if it is greater than either eta_cutoff or sqrt(eta_cutoff) * exp(-entropy(softmax(next_token_logits))). The latter term is intuitively the expected next token probability, scaled by sqrt(eta_cutoff). In the paper, suggested values range from 3e-4 to 2e-3, depending on the size of the model. See Truncation Sampling as Language Model Desmoothing for more details.

[kwargs.diversity_penalty]

number

0.0

This value is subtracted from a beam's score if it generates a token same as any beam from other group at a particular time. Note that diversity_penalty is only effective if group beam search is enabled.

[kwargs.repetition_penalty]

number

1.0

The parameter for repetition penalty. 1.0 means no penalty. See this paper for more details.

[kwargs.encoder_repetition_penalty]

number

1.0

The paramater for encoder_repetition_penalty. An exponential penalty on sequences that are not in the original input. 1.0 means no penalty.

[kwargs.length_penalty]

number

1.0

Exponential penalty to the length that is used with beam-based generation. It is applied as an exponent to the sequence length, which in turn is used to divide the score of the sequence. Since the score is the log likelihood of the sequence (i.e. negative), length_penalty > 0.0 promotes longer sequences, while length_penalty < 0.0 encourages shorter sequences.

[kwargs.no_repeat_ngram_size]

number

0

If set to int > 0, all ngrams of that size can only occur once.

[kwargs.bad_words_ids]

Array.<Array<number>>

List of token ids that are not allowed to be generated. In order to get the token ids of the words that should not appear in the generated text, use (await tokenizer(bad_words, {add_prefix_space: true, add_special_tokens: false})).input_ids.

[kwargs.force_words_ids]

Array<Array<number>> | Array<Array<Array<number>>>

List of token ids that must be generated. If given a number[][], this is treated as a simple list of words that must be included, the opposite to bad_words_ids. If given number[][][], this triggers a disjunctive constraint, where one can allow different forms of each word.

[kwargs.renormalize_logits]

boolean

false

Whether to renormalize the logits after applying all the logits processors or warpers (including the custom ones). It's highly recommended to set this flag to true as the search algorithms suppose the score logits are normalized but some logit processors or warpers break the normalization.

[kwargs.constraints]

Array.<Object>

Custom constraints that can be added to the generation to ensure that the output will contain the use of certain tokens as defined by Constraint objects, in the most sensible way possible.

[kwargs.forced_bos_token_id]

number

The id of the token to force as the first generated token after the decoder_start_token_id. Useful for multilingual models like mBART where the first generated token needs to be the target language token.

[kwargs.forced_eos_token_id]

number | Array<number>

The id of the token to force as the last generated token when max_length is reached. Optionally, use a list to set multiple end-of-sequence tokens.

[kwargs.remove_invalid_values]

boolean

false

Whether to remove possible nan and inf outputs of the model to prevent the generation method to crash. Note that using remove_invalid_values can slow down generation.

[kwargs.exponential_decay_length_penalty]

Array.<number>

This Tuple adds an exponentially increasing length penalty, after a certain amount of tokens have been generated. The tuple shall consist of: (start_index, decay_factor) where start_index indicates where penalty starts and decay_factor represents the factor of exponential decay.

[kwargs.suppress_tokens]

Array.<number>

A list of tokens that will be suppressed at generation. The SupressTokens logit processor will set their log probs to -inf so that they are not sampled.

[kwargs.begin_suppress_tokens]

Array.<number>

A list of tokens that will be suppressed at the beginning of the generation. The SupressBeginTokens logit processor will set their log probs to -inf so that they are not sampled.

[kwargs.forced_decoder_ids]

Array.<Array<number>>

A list of pairs of integers which indicates a mapping from generation indices to token indices that will be forced before sampling. For example, [[1, 123]] means the second generated token will always be a token of index 123.

[kwargs.num_return_sequences]

number

1

The number of independently computed returned sequences for each element in the batch.

[kwargs.output_attentions]

boolean

false

Whether or not to return the attentions tensors of all attention layers. See attentions under returned tensors for more details.

[kwargs.output_hidden_states]

boolean

false

Whether or not to return the hidden states of all layers. See hidden_states under returned tensors for more details.

[kwargs.output_scores]

boolean

false

Whether or not to return the prediction scores. See scores under returned tensors for more details.

[kwargs.return_dict_in_generate]

boolean

false

Whether or not to return a ModelOutput instead of a plain tuple.

[kwargs.pad_token_id]

number

The id of the padding token.

[kwargs.bos_token_id]

number

The id of the beginning-of-sequence token.

[kwargs.eos_token_id]

number | Array<number>

The id of the end-of-sequence token. Optionally, use a list to set multiple end-of-sequence tokens.

[kwargs.encoder_no_repeat_ngram_size]

number

0

If set to int > 0, all ngrams of that size that occur in the encoder_input_ids cannot occur in the decoder_input_ids.

[kwargs.decoder_start_token_id]

number

If an encoder-decoder model starts decoding with a different token than bos, the id of that token.

[kwargs.generation_kwargs]

Object

{}

Additional generation kwargs will be forwarded to the generate function of the model. Kwargs that are not present in generate's signature will be used in the model forward pass.

utils/generation.Sampler

Sampler is a base class for all sampling methods used for text generation.

Kind: static class of utils/generation

new Sampler(generation_config)

Creates a new Sampler object with the specified generation config.

Param
Type
Description

generation_config

GenerationConfig

The generation config.

sampler._call(logits, index) ⇒ <code> void </code>

Executes the sampler, using the specified logits.

Kind: instance method of Sampler

Param
Type

logits

Tensor

index

number

sampler.sample(logits, index)

Abstract method for sampling the logits.

Kind: instance method of Sampler Throws:

  • Error

Param
Type

logits

Tensor

index

number

sampler.getLogits(logits, index) ⇒ <code> Array </code>

Returns the specified logits as an array, with temperature applied.

Kind: instance method of Sampler

Param
Type

logits

Tensor

index

number

sampler.randomSelect(probabilities) ⇒ <code> number </code>

Selects an item randomly based on the specified probabilities.

Kind: instance method of Sampler Returns: number - The index of the selected item.

Param
Type
Description

probabilities

Array

An array of probabilities to use for selection.

Sampler.getSampler(generation_config) ⇒ <code> Sampler </code>

Returns a Sampler object based on the specified options.

Kind: static method of Sampler Returns: Sampler - A Sampler object.

Param
Type
Description

generation_config

GenerationConfig

An object containing options for the sampler.

utils/generation~GreedySampler ⇐ <code> Sampler </code>

Class representing a Greedy Sampler.

Kind: inner class of utils/generation Extends: Sampler

greedySampler.sample(logits, [index]) ⇒ <code> Array </code>

Sample the maximum probability of a given logits tensor.

Kind: instance method of GreedySampler Returns: Array - An array with a single tuple, containing the index of the maximum value and a meaningless score (since this is a greedy search).

Param
Type
Default

logits

Tensor

[index]

number

-1

utils/generation~MultinomialSampler ⇐ <code> Sampler </code>

Class representing a MultinomialSampler.

Kind: inner class of utils/generation Extends: Sampler

multinomialSampler.sample(logits, index) ⇒ <code> Array </code>

Sample from the logits.

Kind: instance method of MultinomialSampler

Param
Type

logits

Tensor

index

number

utils/generation~BeamSearchSampler ⇐ <code> Sampler </code>

Class representing a BeamSearchSampler.

Kind: inner class of utils/generation Extends: Sampler

beamSearchSampler.sample(logits, index) ⇒ <code> Array </code>

Sample from the logits.

Kind: instance method of BeamSearchSampler

Param
Type

logits

Tensor

index

number

PreviousMathsNextData Structures

Last updated