This model was published in HF papers on 2025-09-17 and contributed to Hugging Face Transformers on 2026-08-31.

Canary

Overview

Canary-1B-v2 was proposed in Canary-1B-v2 & Parakeet-TDT-0.6B-v3: Efficient and High-Performance Models for Multilingual ASR and AST by Monica Sekoyan, Nithin Rao Koluguri, Nune Tadevosyan, Piotr Zelasko, Travis Bartley, Nikolay Karpov, Jagadeesh Balam, and Boris Ginsburg.

The abstract from the paper is the following:

This report introduces Canary-1B-v2, a fast, robust multilingual model for Automatic Speech Recognition (ASR) and Speech-to-Text Translation (AST). Built with a FastConformer encoder and Transformer decoder, it supports 25 European languages. The model was trained on 1.7M hours of total data samples, including Granary and NeMo ASR Set 3.0, with non-speech audio added to reduce hallucinations for ASR and AST. We describe its two-stage pre-training and fine-tuning process with dynamic data balancing, as well as experiments with an nGPT encoder. Results show nGPT scales well with massive data, while FastConformer excels after fine-tuning. For timestamps, Canary-1B-v2 uses the NeMo Forced Aligner (NFA) with an auxiliary CTC model, providing reliable segment-level timestamps for ASR and AST. Evaluations show Canary-1B-v2 outperforms Whisper-large-v3 on English ASR while being 10× faster, and delivers competitive multilingual ASR and AST performance against larger models like Seamless-M4T-v2-large and LLM-based systems. We also release Parakeet-TDT-0.6B-v3, a successor to v2, offering multilingual ASR across the same 25 languages with just 600M parameters.

Canary reuses the Fast Conformer encoder from Parakeet (loaded through ParakeetEncoder / ParakeetEncoderConfig) and pairs it with a Transformer decoder that uses fixed sinusoidal positional embeddings, cross-attention to the encoder outputs and tied input/output embeddings. The task is selected through a decoder prompt prefix built by CanaryProcessor of the form <|startofcontext|> <|startoftranscript|> <|emo:undefined|> <source_lang> <target_lang> <pnc|nopnc> <|noitn|> <|notimestamp|> <|nodiarize|>, where source_lang == target_lang selects transcription and otherwise selects translation.

The original implementation can be found in NVIDIA NeMo. A model checkpoint is available at nvidia/canary-1b-v2.

This model was contributed by Harshal Janjani.

Segment-level timestamps for Canary-1B-v2 are produced by the external NeMo Forced Aligner (NFA) with an auxiliary CTC model, not by the decoder, so they are not part of the generate output.

Usage

Transcription

The simplest way to transcribe audio is with apply_transcription_request, which builds the multitask decoder prompt for you (it is a convenience wrapper for apply_chat_template).

from datasets import load_dataset, Audio
from transformers import AutoProcessor, AutoModelForSpeechSeq2Seq

processor = AutoProcessor.from_pretrained("nvidia/canary-1b-v2")
model = AutoModelForSpeechSeq2Seq.from_pretrained("nvidia/canary-1b-v2", device_map="auto")

ds = load_dataset("hf-internal-testing/librispeech_asr_dummy", "clean", split="validation")
ds = ds.cast_column("audio", Audio(sampling_rate=processor.feature_extractor.sampling_rate))

inputs = processor.apply_transcription_request(audio=ds[0]["audio"]["array"], source_language="en").to(model.device)
generated_ids = model.generate(**inputs, max_new_tokens=128)
print(processor.decode(generated_ids, skip_special_tokens=True)[0])

Translation

Set target_language to a different language than source_language for speech-to-text translation.

...

inputs = processor.apply_transcription_request(
    audio=ds[0]["audio"]["array"], source_language="en", target_language="de"
).to(model.device)
generated_ids = model.generate(**inputs, max_new_tokens=128)
print(processor.decode(generated_ids, skip_special_tokens=True)[0])

Batch inference

Pass a list of audios and, optionally, a list of source_language / target_language.

...

audios = [ds[0]["audio"]["array"], ds[1]["audio"]["array"]]
# single entries get broadcasted to list
inputs = processor.apply_transcription_request(
    audio=audios, source_language="en", target_language=["en", "de"]
).to(model.device)
generated_ids = model.generate(**inputs, max_new_tokens=128)
for text in processor.decode(generated_ids, skip_special_tokens=True):
    print(text)

Torch compile

For autoregressive transcription, torch.compile accelerates the per-token forward passes inside generate by providing a CompileConfig object.

...

from transformers import CompileConfig

inputs = processor.apply_transcription_request(audio=ds[0]["audio"]["array"], source_language="en").to(model.device)
compile_config = CompileConfig()

# Warmup
for _ in range(3):
    _ = model.generate(**inputs, max_new_tokens=128, cache_implementation="static", compile_config=compile_config)

# Apply model
generated_ids = model.generate(**inputs, max_new_tokens=128, cache_implementation="static", compile_config=compile_config)
print(processor.decode(generated_ids, skip_special_tokens=True)[0])

Training

Canary can be trained with the loss outputted by the model. Put the target transcript in the assistant turn and pass output_labels=True. Padding positions are masked automatically.

...

model.train()
transcription = "mister Quilter is the apostle of the middle classes, and we are glad to welcome his gospel."

conversation = [
    [
        {
            "role": "user",
            "content": [
                {"type": "audio", "audio": ds[0]["audio"]["array"]},
                {"type": "text", "source_language": "en", "target_language": "en", "punctuation": True},
            ],
        },
        {"role": "assistant", "content": transcription},
    ]
]

inputs = processor.apply_chat_template(
    conversation,
    tokenize=True,
    return_dict=True,
    processor_kwargs={"output_labels": True},
).to(model.device)

outputs = model(**inputs)
outputs.loss.backward()

CanaryConfig

class transformers.CanaryConfig

< source >

( transformers_version: str | None = Nonearchitectures: list[str] | None = Noneoutput_hidden_states: bool | None = Falsereturn_dict: bool | None = Truedtype: typing.Union[str, ForwardRef('torch.dtype'), NoneType] = Nonechunk_size_feed_forward: int = 0id2label: dict[int, str] | dict[str, str] | None = Nonelabel2id: dict[str, int] | dict[str, str] | None = Noneproblem_type: typing.Optional[typing.Literal['regression', 'single_label_classification', 'multi_label_classification']] = Noneis_encoder_decoder: bool = Trueencoder_config: dict | transformers.configuration_utils.PreTrainedConfig | None = Nonedecoder_config: dict | transformers.configuration_utils.PreTrainedConfig | None = Noneuse_cache: bool = Truetie_word_embeddings: bool = Truepad_token_id: int | None = 2bos_token_id: int | None = 4eos_token_id: int | None = 3decoder_start_token_id: int | None = 7initializer_range: float = 0.02vocab_size: int = 16384 )

Parameters

is_encoder_decoder (bool, optional, defaults to True) — Whether the model is used as an encoder/decoder or not.
encoder_config (Union[dict, ParakeetEncoderConfig], optional) — The config object or dictionary of the FastConformer encoder (ParakeetEncoderConfig).
decoder_config (Union[dict, CanaryDecoderConfig], optional) — The config object or dictionary of the Transformer decoder (CanaryDecoderConfig).
use_cache (bool, optional, defaults to True) — Whether or not the model should return the last key/values attentions (not used by all models). Only relevant if config.is_decoder=True or when the model is a decoder-only generative model.
tie_word_embeddings (bool, optional, defaults to True) — Whether to tie weight embeddings according to model’s tied_weights_keys mapping.
pad_token_id (int, optional, defaults to 2) — Token id used for padding in the vocabulary.
bos_token_id (int, optional, defaults to 4) — Token id used for beginning-of-stream in the vocabulary.
eos_token_id (int, optional, defaults to 3) — Token id used for end-of-stream in the vocabulary.
decoder_start_token_id (int, optional, defaults to 7) — The token id that starts decoding (<|startofcontext|>, the first token of the multitask prompt).
initializer_range (float, optional, defaults to 0.02) — The standard deviation of the truncated_normal_initializer for initializing all weight matrices.
vocab_size (int, optional, defaults to 16384) — Vocabulary size of the model. Defines the number of different tokens that can be represented by the input_ids.

This is the configuration class to store the configuration of a CanaryModel. It is used to instantiate a Canary model according to the specified arguments, defining the model architecture. Instantiating a configuration with the defaults will yield a similar configuration to that of the nvidia/canary-1b-v2

Configuration objects inherit from PreTrainedConfig and can be used to control the model outputs. Read the documentation from PreTrainedConfig for more information.

Example:

>>> from transformers import CanaryForConditionalGeneration, CanaryConfig

>>> # Initializing a Canary configuration
>>> configuration = CanaryConfig()

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

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

CanaryDecoderConfig

class transformers.CanaryDecoderConfig

< source >

( transformers_version: str | None = Nonearchitectures: list[str] | None = Noneoutput_hidden_states: bool | None = Falsereturn_dict: bool | None = Truedtype: typing.Union[str, ForwardRef('torch.dtype'), NoneType] = Nonechunk_size_feed_forward: int = 0id2label: dict[int, str] | dict[str, str] | None = Nonelabel2id: dict[str, int] | dict[str, str] | None = Noneproblem_type: typing.Optional[typing.Literal['regression', 'single_label_classification', 'multi_label_classification']] = Noneis_encoder_decoder: bool = Truevocab_size: int = 16384hidden_size: int = 1024intermediate_size: int = 4096num_hidden_layers: int = 8num_attention_heads: int = 8num_key_value_heads: int = 8hidden_act: str = 'relu'max_position_embeddings: int = 1024initializer_range: float = 0.02use_cache: bool = Truepad_token_id: int | None = 2bos_token_id: int | None = 4eos_token_id: int | None = 3attention_bias: bool = Trueattention_dropout: int | float | None = 0.0head_dim: int = 128 )

Parameters

is_encoder_decoder (bool, optional, defaults to True) — Whether the model is used as an encoder/decoder or not.
vocab_size (int, optional, defaults to 16384) — Vocabulary size of the model. Defines the number of different tokens that can be represented by the input_ids.
hidden_size (int, optional, defaults to 1024) — Dimension of the hidden representations.
intermediate_size (int, optional, defaults to 4096) — Dimension of the MLP representations.
num_hidden_layers (int, optional, defaults to 8) — Number of hidden layers in the Transformer decoder.
num_attention_heads (int, optional, defaults to 8) — Number of attention heads for each attention layer in the Transformer decoder.
num_key_value_heads (int, optional, defaults to 8) — This is the number of key_value heads that should be used to implement Grouped Query Attention. If num_key_value_heads=num_attention_heads, the model will use Multi Head Attention (MHA), if num_key_value_heads=1 the model will use Multi Query Attention (MQA) otherwise GQA is used. When converting a multi-head checkpoint to a GQA checkpoint, each group key and value head should be constructed by meanpooling all the original heads within that group. For more details, check out this paper. If it is not specified, will default to num_attention_heads.
hidden_act (str, optional, defaults to relu) — The non-linear activation function (function or string) in the decoder. For example, "gelu", "relu", "silu", etc.
max_position_embeddings (int, optional, defaults to 1024) — The maximum sequence length that this model might ever be used with.
initializer_range (float, optional, defaults to 0.02) — The standard deviation of the truncated_normal_initializer for initializing all weight matrices.
use_cache (bool, optional, defaults to True) — Whether or not the model should return the last key/values attentions (not used by all models). Only relevant if config.is_decoder=True or when the model is a decoder-only generative model.
pad_token_id (int, optional, defaults to 2) — Token id used for padding in the vocabulary.
bos_token_id (int, optional, defaults to 4) — Token id used for beginning-of-stream in the vocabulary.
eos_token_id (int, optional, defaults to 3) — Token id used for end-of-stream in the vocabulary.
attention_bias (bool, optional, defaults to True) — Whether to use a bias in the query, key, value and output projection layers during self-attention.
attention_dropout (Union[int, float], optional, defaults to 0.0) — The dropout ratio for the attention probabilities.
head_dim (int, optional, defaults to 128) — The attention head dimension. If None, it will default to hidden_size // num_attention_heads

Configuration objects inherit from PreTrainedConfig and can be used to control the model outputs. Read the documentation from PreTrainedConfig for more information.

>>> from transformers import CanaryDecoderModel, CanaryDecoderConfig

>>> # Initializing a CanaryDecoder canary_decoder-7b style configuration
>>> configuration = CanaryDecoderConfig()

>>> # Initializing a model from the canary_decoder-7b style configuration
>>> model = CanaryDecoderModel(configuration)

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

CanaryProcessor

class transformers.CanaryProcessor

< source >

( feature_extractor = Nonetokenizer = Nonechat_template = None )

Parameters

feature_extractor (ParakeetFeatureExtractor) — The feature extractor is a required input.
tokenizer (TokenizersBackend) — The tokenizer is a required input.
chat_template (str) — A Jinja template to convert lists of messages in a chat into a tokenizable string.

Constructs a CanaryProcessor which wraps a feature extractor and a tokenizer into a single processor.

CanaryProcessor offers all the functionalities of ParakeetFeatureExtractor and TokenizersBackend. See the ~ParakeetFeatureExtractor and ~TokenizersBackend for more information.

apply_transcription_request

< source >

( audio: typing.Union[numpy.ndarray, ForwardRef('torch.Tensor'), collections.abc.Sequence[numpy.ndarray], collections.abc.Sequence['torch.Tensor'], list[typing.Union[numpy.ndarray, ForwardRef('torch.Tensor'), collections.abc.Sequence[numpy.ndarray], collections.abc.Sequence['torch.Tensor']]]]source_language: str | list[str] = 'en'target_language: str | list[str] | None = Nonepunctuation: bool = True**kwargs: Unpack ) → BatchFeature

Parameters

audio (AudioInput or list[AudioInput]) — Audio to transcribe or translate. Can be a URL string, local path, numpy array, or a list of these.
source_language (str or list[str], optional, defaults to "en") — The language of the input speech. Accepts ISO codes (e.g. "en", "de", "fr") or full names (e.g. "English", "German", "French").
target_language (str or list[str], optional) — The language of the output text. Accepts ISO codes or full names. Defaults to source_language (transcription); set it to a different language for speech-to-text translation.
punctuation (bool, optional, defaults to True) — Whether to request punctuation and capitalization in the output.
**kwargs — Additional keyword arguments forwarded to apply_chat_template().

Returns

BatchFeature

Processor outputs ready to be passed to CanaryForConditionalGeneration.generate().

Prepare inputs for transcription or translation without manually writing the chat template.

CanaryModel

class transformers.CanaryModel

< source >

( config: CanaryConfig )

Parameters

config (CanaryConfig) — Model configuration class with all the parameters of the model. Initializing with a config file does not load the weights associated with the model, only the configuration. Check out the from_pretrained() method to load the model weights.

The bare Canary model (FastConformer encoder + Transformer decoder) outputting raw hidden-states without any specific head on top.

This model inherits from PreTrainedModel. Check the superclass documentation for the generic methods the library implements for all its model (such as downloading or saving, resizing the input embeddings, pruning heads etc.)

This model is also a PyTorch torch.nn.Module subclass. Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage and behavior.

forward

< source >

( input_features: typing.Optional[torch.FloatTensor] = Noneattention_mask: typing.Optional[torch.LongTensor] = Nonedecoder_input_ids: typing.Optional[torch.LongTensor] = Nonedecoder_attention_mask: typing.Optional[torch.LongTensor] = Noneencoder_outputs: tuple[tuple[torch.FloatTensor]] | None = Nonepast_key_values: transformers.cache_utils.EncoderDecoderCache | None = Nonedecoder_inputs_embeds: tuple[torch.FloatTensor] | None = Nonedecoder_position_ids: tuple[torch.LongTensor] | None = Noneuse_cache: bool | None = None**kwargs: Unpack ) → Seq2SeqModelOutput or tuple(torch.FloatTensor)

Parameters

input_features (torch.FloatTensor of shape (batch_size, audio_length)) — Float values of the raw speech waveform. Raw speech waveform can be obtained by loading a .flac or .wav audio file into an array of type list[float], a numpy.ndarray or a torch.Tensor, e.g. via the torchcodec library (pip install torchcodec) or the soundfile library (pip install soundfile). To prepare the array into input_features, the AutoFeatureExtractor should be used for padding and conversion into a tensor of type torch.FloatTensor.
attention_mask (torch.LongTensor of shape (batch_size, sequence_length), optional) — Mask to avoid performing attention on padding token indices. Mask values selected in [0, 1]:
- 1 for tokens that are not masked,
- 0 for tokens that are masked.
What are attention masks?
decoder_input_ids (torch.LongTensor of shape (batch_size, target_sequence_length), optional) — Indices of decoder input sequence tokens in the vocabulary.

Indices can be obtained using AutoTokenizer. See PreTrainedTokenizer.encode() and PreTrainedTokenizer.call() for details.

What are decoder input IDs?
decoder_attention_mask (torch.LongTensor of shape (batch_size, target_sequence_length), optional) — Mask to avoid performing attention on certain token indices. By default, a causal mask will be used, to make sure the model can only look at previous inputs in order to predict the future.
encoder_outputs (tuple[tuple[torch.FloatTensor]], optional) — Tuple consists of (last_hidden_state, optional: hidden_states, optional: attentions) last_hidden_state of shape (batch_size, sequence_length, hidden_size), optional) is a sequence of hidden-states at the output of the last layer of the encoder. Used in the cross-attention of the decoder.
past_key_values (~cache_utils.EncoderDecoderCache, optional) — Pre-computed hidden-states (key and values in the self-attention blocks and in the cross-attention blocks) that can be used to speed up sequential decoding. This typically consists in the past_key_values returned by the model at a previous stage of decoding, when use_cache=True or config.use_cache=True.

Only Cache instance is allowed as input, see our kv cache guide. If no past_key_values are passed, DynamicCache will be initialized by default.

The model will output the same cache format that is fed as input.

If past_key_values are used, the user is expected to input only unprocessed input_ids (those that don’t have their past key value states given to this model) of shape (batch_size, unprocessed_length) instead of all input_ids of shape (batch_size, sequence_length).
decoder_inputs_embeds (tuple[torch.FloatTensor] of shape (batch_size, target_sequence_length, hidden_size), optional) — Optionally, instead of passing decoder_input_ids you can choose to directly pass an embedded representation. If past_key_values is used, optionally only the last decoder_inputs_embeds have to be input (see past_key_values). This is useful if you want more control over how to convert decoder_input_ids indices into associated vectors than the model’s internal embedding lookup matrix.

If decoder_input_ids and decoder_inputs_embeds are both unset, decoder_inputs_embeds takes the value of inputs_embeds.
decoder_position_ids (torch.LongTensor of shape (batch_size, target_sequence_length)) — Indices of positions of each input sequence tokens in the position embeddings. Used to calculate the position embeddings up to config.decoder_config.max_position_embeddings
use_cache (bool, optional) — If set to True, past_key_values key value states are returned and can be used to speed up decoding (see past_key_values).

Returns

Seq2SeqModelOutput or tuple(torch.FloatTensor)

A Seq2SeqModelOutput or a tuple of torch.FloatTensor (if return_dict=False is passed or when config.return_dict=False) comprising various elements depending on the configuration (CanaryConfig) and inputs.

The CanaryModel forward method, overrides the __call__ special method.

Although the recipe for forward pass needs to be defined within this function, one should call the Module instance afterwards instead of this since the former takes care of running the pre and post processing steps while the latter silently ignores them.

last_hidden_state (torch.FloatTensor of shape (batch_size, sequence_length, hidden_size)) — Sequence of hidden-states at the output of the last layer of the decoder of the model.

If past_key_values is used only the last hidden-state of the sequences of shape (batch_size, 1, hidden_size) is output.
past_key_values (EncoderDecoderCache, optional, returned when use_cache=True is passed or when config.use_cache=True) — It is a EncoderDecoderCache instance. For more details, see our kv cache guide.

Contains pre-computed hidden-states (key and values in the self-attention blocks and in the cross-attention blocks) that can be used (see past_key_values input) to speed up sequential decoding.
decoder_hidden_states (tuple(torch.FloatTensor), optional, returned when output_hidden_states=True is passed or when config.output_hidden_states=True) — Tuple of torch.FloatTensor (one for the output of the embeddings, if the model has an embedding layer, + one for the output of each layer) of shape (batch_size, sequence_length, hidden_size).

Hidden-states of the decoder at the output of each layer plus the optional initial embedding outputs.
decoder_attentions (tuple(torch.FloatTensor), optional, returned when output_attentions=True is passed or when config.output_attentions=True) — Tuple of torch.FloatTensor (one for each layer) of shape (batch_size, num_heads, sequence_length, sequence_length).

Attentions weights of the decoder, after the attention softmax, used to compute the weighted average in the self-attention heads.
cross_attentions (tuple(torch.FloatTensor), optional, returned when output_attentions=True is passed or when config.output_attentions=True) — Tuple of torch.FloatTensor (one for each layer) of shape (batch_size, num_heads, sequence_length, sequence_length).

Attentions weights of the decoder’s cross-attention layer, after the attention softmax, used to compute the weighted average in the cross-attention heads.
encoder_last_hidden_state (torch.FloatTensor of shape (batch_size, sequence_length, hidden_size), optional) — Sequence of hidden-states at the output of the last layer of the encoder of the model.
encoder_hidden_states (tuple(torch.FloatTensor), optional, returned when output_hidden_states=True is passed or when config.output_hidden_states=True) — Tuple of torch.FloatTensor (one for the output of the embeddings, if the model has an embedding layer, + one for the output of each layer) of shape (batch_size, sequence_length, hidden_size).

Hidden-states of the encoder at the output of each layer plus the optional initial embedding outputs.
encoder_attentions (tuple(torch.FloatTensor), optional, returned when output_attentions=True is passed or when config.output_attentions=True) — Tuple of torch.FloatTensor (one for each layer) of shape (batch_size, num_heads, sequence_length, sequence_length).

Attentions weights of the encoder, after the attention softmax, used to compute the weighted average in the self-attention heads.

Example:

>>> import torch
>>> from transformers import AutoFeatureExtractor, CanaryModel
>>> from datasets import load_dataset

>>> model = CanaryModel.from_pretrained("UsefulSensors/canary-tiny")
>>> feature_extractor = AutoFeatureExtractor.from_pretrained("UsefulSensors/canary-tiny")
>>> ds = load_dataset("hf-internal-testing/librispeech_asr_dummy", "clean", split="validation")
>>> inputs = feature_extractor(ds[0]["audio"]["array"], return_tensors="pt")
>>> input_features = inputs.input_features
>>> decoder_input_ids = torch.tensor([[1, 1]]) * model.config.decoder_start_token_id
>>> last_hidden_state = model(input_features, decoder_input_ids=decoder_input_ids).last_hidden_state
>>> list(last_hidden_state.shape)
[1, 2, 288]

CanaryForConditionalGeneration

class transformers.CanaryForConditionalGeneration

< source >

( config: CanaryConfig )

Parameters

config (CanaryConfig) — Model configuration class with all the parameters of the model. Initializing with a config file does not load the weights associated with the model, only the configuration. Check out the from_pretrained() method to load the model weights.

The Canary model with a language modeling head. Can be used for multilingual automatic speech recognition and speech-to-text translation.

This model is also a PyTorch torch.nn.Module subclass. Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage and behavior.

forward

< source >

( input_features: typing.Optional[torch.FloatTensor] = Noneattention_mask: typing.Optional[torch.LongTensor] = Nonedecoder_input_ids: typing.Optional[torch.LongTensor] = Nonedecoder_attention_mask: typing.Optional[torch.LongTensor] = Noneencoder_outputs: tuple[tuple[torch.FloatTensor]] | None = Nonepast_key_values: transformers.cache_utils.EncoderDecoderCache | None = Nonedecoder_inputs_embeds: tuple[torch.FloatTensor] | None = Nonedecoder_position_ids: tuple[torch.LongTensor] | None = Noneuse_cache: bool | None = Nonelabels: typing.Optional[torch.LongTensor] = None**kwargs: Unpack ) → Seq2SeqLMOutput or tuple(torch.FloatTensor)

Parameters

input_features (torch.FloatTensor of shape (batch_size, audio_length)) — Float values of the raw speech waveform. Raw speech waveform can be obtained by loading a .flac or .wav audio file into an array of type list[float], a numpy.ndarray or a torch.Tensor, e.g. via the torchcodec library (pip install torchcodec) or the soundfile library (pip install soundfile). To prepare the array into input_features, the AutoFeatureExtractor should be used for padding and conversion into a tensor of type torch.FloatTensor.
attention_mask (torch.LongTensor of shape (batch_size, sequence_length), optional) — Mask to avoid performing attention on padding token indices. Mask values selected in [0, 1]:
- 1 for tokens that are not masked,
- 0 for tokens that are masked.
What are attention masks?
decoder_input_ids (torch.LongTensor of shape (batch_size, target_sequence_length), optional) — Indices of decoder input sequence tokens in the vocabulary.

Indices can be obtained using AutoTokenizer. See PreTrainedTokenizer.encode() and PreTrainedTokenizer.call() for details.

What are decoder input IDs?
decoder_attention_mask (torch.LongTensor of shape (batch_size, target_sequence_length), optional) — Mask to avoid performing attention on certain token indices. By default, a causal mask will be used, to make sure the model can only look at previous inputs in order to predict the future.
encoder_outputs (tuple[tuple[torch.FloatTensor]], optional) — Tuple consists of (last_hidden_state, optional: hidden_states, optional: attentions) last_hidden_state of shape (batch_size, sequence_length, hidden_size), optional) is a sequence of hidden-states at the output of the last layer of the encoder. Used in the cross-attention of the decoder.
past_key_values (~cache_utils.EncoderDecoderCache, optional) — Pre-computed hidden-states (key and values in the self-attention blocks and in the cross-attention blocks) that can be used to speed up sequential decoding. This typically consists in the past_key_values returned by the model at a previous stage of decoding, when use_cache=True or config.use_cache=True.

Only Cache instance is allowed as input, see our kv cache guide. If no past_key_values are passed, DynamicCache will be initialized by default.

The model will output the same cache format that is fed as input.

If past_key_values are used, the user is expected to input only unprocessed input_ids (those that don’t have their past key value states given to this model) of shape (batch_size, unprocessed_length) instead of all input_ids of shape (batch_size, sequence_length).
decoder_inputs_embeds (tuple[torch.FloatTensor] of shape (batch_size, target_sequence_length, hidden_size), optional) — Optionally, instead of passing decoder_input_ids you can choose to directly pass an embedded representation. If past_key_values is used, optionally only the last decoder_inputs_embeds have to be input (see past_key_values). This is useful if you want more control over how to convert decoder_input_ids indices into associated vectors than the model’s internal embedding lookup matrix.

If decoder_input_ids and decoder_inputs_embeds are both unset, decoder_inputs_embeds takes the value of inputs_embeds.
decoder_position_ids (torch.LongTensor of shape (batch_size, target_sequence_length)) — Indices of positions of each input sequence tokens in the position embeddings. Used to calculate the position embeddings up to config.decoder_config.max_position_embeddings
use_cache (bool, optional) — If set to True, past_key_values key value states are returned and can be used to speed up decoding (see past_key_values).
labels (torch.LongTensor of shape (batch_size, sequence_length), optional) — Labels for computing the masked language modeling loss. Indices should either be in [0, ..., config.vocab_size] or -100 (see input_ids docstring). Tokens with indices set to -100 are ignored (masked), the loss is only computed for the tokens with labels in [0, ..., config.vocab_size].

Returns

Seq2SeqLMOutput or tuple(torch.FloatTensor)

A Seq2SeqLMOutput or a tuple of torch.FloatTensor (if return_dict=False is passed or when config.return_dict=False) comprising various elements depending on the configuration (CanaryConfig) and inputs.

The CanaryForConditionalGeneration forward method, overrides the __call__ special method.

Although the recipe for forward pass needs to be defined within this function, one should call the Module instance afterwards instead of this since the former takes care of running the pre and post processing steps while the latter silently ignores them.

loss (torch.FloatTensor of shape (1,), optional, returned when labels is provided) — Language modeling loss.
logits (torch.FloatTensor of shape (batch_size, sequence_length, config.vocab_size)) — Prediction scores of the language modeling head (scores for each vocabulary token before SoftMax).
past_key_values (EncoderDecoderCache, optional, returned when use_cache=True is passed or when config.use_cache=True) — It is a EncoderDecoderCache instance. For more details, see our kv cache guide.

Contains pre-computed hidden-states (key and values in the self-attention blocks and in the cross-attention blocks) that can be used (see past_key_values input) to speed up sequential decoding.
decoder_hidden_states (tuple(torch.FloatTensor), optional, returned when output_hidden_states=True is passed or when config.output_hidden_states=True) — Tuple of torch.FloatTensor (one for the output of the embeddings, if the model has an embedding layer, + one for the output of each layer) of shape (batch_size, sequence_length, hidden_size).

Hidden-states of the decoder at the output of each layer plus the initial embedding outputs.
decoder_attentions (tuple(torch.FloatTensor), optional, returned when output_attentions=True is passed or when config.output_attentions=True) — Tuple of torch.FloatTensor (one for each layer) of shape (batch_size, num_heads, sequence_length, sequence_length).

Attentions weights of the decoder, after the attention softmax, used to compute the weighted average in the self-attention heads.
cross_attentions (tuple(torch.FloatTensor), optional, returned when output_attentions=True is passed or when config.output_attentions=True) — Tuple of torch.FloatTensor (one for each layer) of shape (batch_size, num_heads, sequence_length, sequence_length).

Attentions weights of the decoder’s cross-attention layer, after the attention softmax, used to compute the weighted average in the cross-attention heads.
encoder_last_hidden_state (torch.FloatTensor of shape (batch_size, sequence_length, hidden_size), optional) — Sequence of hidden-states at the output of the last layer of the encoder of the model.
encoder_hidden_states (tuple(torch.FloatTensor), optional, returned when output_hidden_states=True is passed or when config.output_hidden_states=True) — Tuple of torch.FloatTensor (one for the output of the embeddings, if the model has an embedding layer, + one for the output of each layer) of shape (batch_size, sequence_length, hidden_size).

Hidden-states of the encoder at the output of each layer plus the initial embedding outputs.
encoder_attentions (tuple(torch.FloatTensor), optional, returned when output_attentions=True is passed or when config.output_attentions=True) — Tuple of torch.FloatTensor (one for each layer) of shape (batch_size, num_heads, sequence_length, sequence_length).

Attentions weights of the encoder, after the attention softmax, used to compute the weighted average in the self-attention heads.

Example:

>>> import torch
>>> from transformers import AutoProcessor, CanaryForConditionalGeneration
>>> from datasets import load_dataset

>>> processor = AutoProcessor.from_pretrained("UsefulSensors/canary-tiny")
>>> model = CanaryForConditionalGeneration.from_pretrained("UsefulSensors/canary-tiny")

>>> ds = load_dataset("hf-internal-testing/librispeech_asr_dummy", "clean", split="validation")

>>> inputs = processor(ds[0]["audio"]["array"], return_tensors="pt")
>>> input_features = inputs.input_features

>>> generated_ids = model.generate(input_features, max_new_tokens=100)

>>> transcription = processor.batch_decode(generated_ids, skip_special_tokens=True)[0]
>>> transcription
'Mr. Quilter is the apostle of the middle classes, and we are glad to welcome his gospel.'

Update on GitHub