Kyutai STT is a speech-to-text model architecture based on the Mimi codec, which encodes audio into discrete tokens in a streaming fashion, and a Moshi-like autoregressive decoder. Kyutai’s lab has released two model checkpoints:
import torch
from datasets import load_dataset, Audio
from transformers import KyutaiSpeechToTextProcessor, KyutaiSpeechToTextForConditionalGeneration
torch_device = "cuda" if torch.cuda.is_available() else "cpu"
model_id = "kyutai/stt-2.6b-en"
processor = KyutaiSpeechToTextProcessor.from_pretrained(model_id)
model = KyutaiSpeechToTextForConditionalGeneration.from_pretrained(model_id, device_map=torch_device)
ds = load_dataset(
"hf-internal-testing/librispeech_asr_dummy", "clean", split="validation"
)
ds = ds.cast_column("audio", Audio(sampling_rate=24000))
inputs = processor(
ds[0]["audio"]["array"],
)
inputs.to(torch_device)
output_tokens = model.generate(**inputs)
print(processor.batch_decode(output_tokens, skip_special_tokens=True)) Batched Inference
import torch
from datasets import load_dataset, Audio
from transformers import KyutaiSpeechToTextProcessor, KyutaiSpeechToTextForConditionalGeneration
torch_device = "cuda" if torch.cuda.is_available() else "cpu"
model_id = "kyutai/stt-2.6b-en"
processor = KyutaiSpeechToTextProcessor.from_pretrained(model_id)
model = KyutaiSpeechToTextForConditionalGeneration.from_pretrained(model_id, device_map=torch_device)
ds = load_dataset(
"hf-internal-testing/librispeech_asr_dummy", "clean", split="validation"
)
ds = ds.cast_column("audio", Audio(sampling_rate=24000))
audio_arrays = [ds[i]["audio"]["array"] for i in range(4)]
inputs = processor(audio_arrays, return_tensors="pt", padding=True)
inputs = inputs.to(torch_device)
output_tokens = model.generate(**inputs)
decoded_outputs = processor.batch_decode(output_tokens, skip_special_tokens=True)
for output in decoded_outputs:
print(output)
This model was contributed by Eustache Le Bihan. The original code can be found here.
KyutaiSpeechToTextConfig class transformers.KyutaiSpeechToTextConfig < source >( codebook_vocab_size = 2049 vocab_size = 4001 hidden_size = 2048 num_hidden_layers = 48 num_attention_heads = 32 num_key_value_heads = None max_position_embeddings = 750 rope_theta = 100000.0 hidden_act = 'silu' head_dim = None initializer_range = 0.02 use_cache = True sliding_window = 375 attention_dropout = 0.0 ffn_dim = 11264 rms_norm_eps = 1e-08 num_codebooks = 32 audio_bos_token_id = 2048 audio_pad_token_id = 69569 tie_word_embeddings = False pad_token_id = 3 bos_token_id = 48000 codec_config = None **kwargs )
Parameters
int, optional, defaults to 2049) — Vocabulary size of the codebook. Defines the number of different audio tokens that can be represented by each codebook. int, optional, defaults to 4001) — Vocabulary size of the model. Defines the number of different tokens that can be represented by the input_ids passed when calling the model. int, optional, defaults to 2048) — Dimensionality of the layers and the pooler layer of the main decoder. int, optional, defaults to 48) — Number of decoder layers. int, optional, defaults to 32) — Number of attention heads for each attention layer in the main decoder block. int, optional) — 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 checkout this paper. If it is not specified, will default to num_attention_heads. int, optional, defaults to 750) — The maximum sequence length that this model might ever be used with. Typically, set this to something large just in case (e.g., 512 or 1024 or 2048). float, optional, defaults to 100000.0) — The base period of the RoPE embeddings. str or function, optional, defaults to "silu") — The non-linear activation function (function or string) in the decoder. int, optional, defaults to hidden_size // num_attention_heads) — The attention head dimension. float, optional, defaults to 0.02) — The standard deviation of the truncated_normal_initializer for initializing all weight matrices. 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. int, optional, defaults to 375) — Sliding window attention window size. If not specified, will default to 3000. float, optional, defaults to 0.0) — The dropout ratio for the attention probabilities. int, optional, defaults to 11264) — Dimensionality of the “intermediate” (often named feed-forward) layer in the main decoder block. Must be even. float, optional, defaults to 1e-08) — The epsilon used by the rms normalization layers. int, optional, defaults to 32) — The number of audio codebooks for each audio channels. int, optional, defaults to 2048) — Beginning of stream token id for codebook tokens. int, optional, defaults to 69569) — Padding token id for codebook tokens. bool, optional, defaults to False) — Whether to tie weight embeddings. int, optional, defaults to 3) — Padding token id. int, optional, defaults to 48000) — Beginning of stream token id for text tokens. PretrainedConfig, optional) — Configuration for the codec. This is the configuration class to store the configuration of a KyutaiSpeechToTextForConditionalGeneration. It is used to instantiate a Kyutai Speech-to-Text 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 2.6b-en model.
e.g. kyutai/stt-2.6b-en
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 KyutaiSpeechToTextConfig, KyutaiSpeechToTextForConditionalGeneration >>> >>> configuration = KyutaiSpeechToTextConfig() >>> >>> model = KyutaiSpeechToTextForConditionalGeneration(configuration) >>> >>> configuration = model.configKyutaiSpeechToTextProcessor class transformers.KyutaiSpeechToTextProcessor < source >
( *args **kwargs )
Constructs a Moshi ASR processor which wraps EncodecFeatureExtractor and PreTrainedTokenizerFast into a single processor that inherits both the audio feature extraction and tokenizer functionalities. See the call() for more information.
__call__ < source >( audio: typing.Union[numpy.ndarray, ForwardRef('torch.Tensor'), list[numpy.ndarray], tuple[numpy.ndarray], list['torch.Tensor'], tuple['torch.Tensor'], NoneType] = None **kwargs: typing_extensions.Unpack[transformers.models.stt.processing_kyutai_speech_to_text.KyutaiSpeechToTextProcessorKwargs] ) → BatchFeature
Parameters
np.ndarray, torch.Tensor, list[np.ndarray], list[torch.Tensor]) — The audio or batch of audio to be prepared. Each audio can be a NumPy array or PyTorch tensor. str or TensorType, optional) — If set, will return tensors of a particular framework. Acceptable values are:
'tf': Return TensorFlow tf.constant objects.'pt': Return PyTorch torch.Tensor objects.'np': Return NumPy np.ndarray objects.'jax': Return JAX jnp.ndarray objects.A BatchFeature with the following fields:
audio is not None.Main method to prepare audio to be fed as input to the model. This method forwards the audio arguments to KyutaiSpeechToTextFeatureExtractor’s __call__(). Please refer to the docstring of the above method for more information.
( feature_size: int = 1 sampling_rate: int = 24000 padding_value: float = 0.0 chunk_length_s: typing.Optional[float] = None overlap: typing.Optional[float] = None audio_delay_seconds: typing.Optional[float] = 0.0 audio_silence_prefix_seconds: typing.Optional[float] = 0.0 **kwargs )
Parameters
int, optional, defaults to 1) — The feature dimension of the extracted features. Use 1 for mono, 2 for stereo. int, optional, defaults to 24000) — The sampling rate at which the audio waveform should be digitalized expressed in hertz (Hz). float, optional, defaults to 0.0) — The value that is used to fill the padding values. float, optional) — If defined the audio is pre-processed into chunks of lengths chunk_length_s and then encoded. float, optional) — Defines the overlap between each chunk. It is used to compute the chunk_stride using the following formulae : int((1.0 - self.overlap) * self.chunk_length). float, optional, defaults to 0.0) — The delay in seconds to add after the audio (right padding). float, optional, defaults to 0.0) — The silence prefix in seconds to add before the audio (left padding). Constructs an KyutaiSpeechToText feature extractor.
This feature extractor inherits from SequenceFeatureExtractor which contains most of the main methods. Users should refer to this superclass for more information regarding those methods.
KyutaiSpeechToTextForConditionalGeneration class transformers.KyutaiSpeechToTextForConditionalGeneration < source >( config )
Parameters
The Kyutai Speech To Text Model for token generation conditioned on other modalities (e.g. image-text-to-text generation).
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_ids: typing.Optional[torch.LongTensor] = None attention_mask: typing.Optional[torch.Tensor] = None position_ids: typing.Optional[torch.LongTensor] = None past_key_values: typing.Optional[transformers.cache_utils.Cache] = None inputs_embeds: typing.Optional[torch.FloatTensor] = None labels: typing.Optional[torch.LongTensor] = None use_cache: typing.Optional[bool] = None output_attentions: typing.Optional[bool] = None output_hidden_states: typing.Optional[bool] = None cache_position: typing.Optional[torch.LongTensor] = None logits_to_keep: typing.Union[int, torch.Tensor] = 0 **kwargs: typing_extensions.Unpack[transformers.models.stt.modeling_kyutai_speech_to_text.KwargsForCausalLM] ) → transformers.modeling_outputs.CausalLMOutputWithPast or tuple(torch.FloatTensor)
Parameters
torch.LongTensor of shape (batch_size, sequence_length), optional) — Indices of input sequence tokens in the vocabulary. Padding will be ignored by default.
Indices can be obtained using AutoTokenizer. See PreTrainedTokenizer.encode() and PreTrainedTokenizer.call() for details.
torch.Tensor of shape (batch_size, sequence_length), optional) — Mask to avoid performing attention on padding token indices. Mask values selected in [0, 1]:
torch.LongTensor of shape (batch_size, sequence_length), optional) — Indices of positions of each input sequence tokens in the position embeddings. Selected in the range [0, config.n_positions - 1].
~cache_utils.Cache, 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.
Two formats are allowed:
tuple(torch.FloatTensor) of length config.n_layers, with each tuple having 2 tensors of shape (batch_size, num_heads, sequence_length, embed_size_per_head)). This is also known as the legacy cache format.The model will output the same cache format that is fed as input. If no past_key_values are passed, the legacy cache format will be returned.
If past_key_values are used, the user can optionally input only the last input_ids (those that don’t have their past key value states given to this model) of shape (batch_size, 1) instead of all input_ids of shape (batch_size, sequence_length).
torch.FloatTensor of shape (batch_size, sequence_length, hidden_size), optional) — Optionally, instead of passing input_ids you can choose to directly pass an embedded representation. This is useful if you want more control over how to convert input_ids indices into associated vectors than the model’s internal embedding lookup matrix. 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]. 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). bool, optional) — Whether or not to return the attentions tensors of all attention layers. See attentions under returned tensors for more detail. bool, optional) — Whether or not to return the hidden states of all layers. See hidden_states under returned tensors for more detail. torch.LongTensor of shape (sequence_length), optional) — Indices depicting the position of the input sequence tokens in the sequence. Contrarily to position_ids, this tensor is not affected by padding. It is used to update the cache in the correct position and to infer the complete sequence length. Union[int, torch.Tensor], defaults to 0) — If an int, compute logits for the last logits_to_keep tokens. If 0, calculate logits for all input_ids (special case). Only last token logits are needed for generation, and calculating them only for that token can save memory, which becomes pretty significant for long sequences or large vocabulary size. If a torch.Tensor, must be 1D corresponding to the indices to keep in the sequence length dimension. This is useful when using packed tensor format (single dimension for batch and sequence length). A transformers.modeling_outputs.CausalLMOutputWithPast 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 (ModelConfig) and inputs.
loss (torch.FloatTensor of shape (1,), optional, returned when labels is provided) — Language modeling loss (for next-token prediction).
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 (Cache, optional, returned when use_cache=True is passed or when config.use_cache=True) — It is a Cache instance. For more details, see our kv cache guide.
Contains pre-computed hidden-states (key and values in the self-attention blocks) that can be used (see past_key_values input) to speed up sequential decoding.
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 model at the output of each layer plus the optional initial embedding outputs.
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 after the attention softmax, used to compute the weighted average in the self-attention heads.
The KyutaiSpeechToTextForConditionalGeneration 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.
Example:
>>> import torch
>>> from datasets import load_dataset, Audio
>>> from transformers import KyutaiSpeechToTextProcessor, KyutaiSpeechToTextForConditionalGeneration
>>> torch_device = "cuda" if torch.cuda.is_available() else "cpu"
>>> model_id = "kyutai/stt-2.6b-en"
>>> processor = KyutaiSpeechToTextProcessor.from_pretrained(model_id)
>>> model = KyutaiSpeechToTextForConditionalGeneration.from_pretrained(model_id, device_map=torch_device)
>>> ds = load_dataset(
... "hf-internal-testing/librispeech_asr_dummy", "clean", split="validation"
... )
>>> ds = ds.cast_column("audio", Audio(sampling_rate=24000))
>>> inputs = processor(
... ds[0]["audio"]["array"],
... )
>>> inputs.to(torch_device)
>>> output_tokens = model.generate(**inputs)
>>> print(processor.batch_decode(output_tokens, skip_special_tokens=True))
This method forwards all its arguments to GenerationMixin’s generate(). Please refer to the docstring of this method for more information.
KyutaiSpeechToTextModel class transformers.KyutaiSpeechToTextModel < source >( config )
Parameters
The bare Kyutai Speech To Text Text Model outputting raw hidden-states without any specific head on to.
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_ids: typing.Optional[torch.LongTensor] = None attention_mask: typing.Optional[torch.Tensor] = None position_ids: typing.Optional[torch.LongTensor] = None past_key_values: typing.Union[list[torch.FloatTensor], transformers.cache_utils.Cache, NoneType] = None inputs_embeds: typing.Optional[torch.FloatTensor] = None use_cache: typing.Optional[bool] = None output_attentions: typing.Optional[bool] = None output_hidden_states: typing.Optional[bool] = None return_dict: typing.Optional[bool] = None cache_position: typing.Optional[torch.LongTensor] = None ) → transformers.modeling_outputs.BaseModelOutputWithPast or tuple(torch.FloatTensor)
Parameters
torch.LongTensor of shape (batch_size, sequence_length), optional) — Indices of input sequence tokens in the vocabulary. Padding will be ignored by default.
Indices can be obtained using AutoTokenizer. See PreTrainedTokenizer.encode() and PreTrainedTokenizer.call() for details.
torch.Tensor of shape (batch_size, sequence_length), optional) — Mask to avoid performing attention on padding token indices. Mask values selected in [0, 1]:
torch.LongTensor of shape (batch_size, sequence_length), optional) — Indices of positions of each input sequence tokens in the position embeddings. Selected in the range [0, config.n_positions - 1].
Union[list[torch.FloatTensor], ~cache_utils.Cache, NoneType]) — 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.
Two formats are allowed:
tuple(torch.FloatTensor) of length config.n_layers, with each tuple having 2 tensors of shape (batch_size, num_heads, sequence_length, embed_size_per_head)). This is also known as the legacy cache format.The model will output the same cache format that is fed as input. If no past_key_values are passed, the legacy cache format will be returned.
If past_key_values are used, the user can optionally input only the last input_ids (those that don’t have their past key value states given to this model) of shape (batch_size, 1) instead of all input_ids of shape (batch_size, sequence_length).
torch.FloatTensor of shape (batch_size, sequence_length, hidden_size), optional) — Optionally, instead of passing input_ids you can choose to directly pass an embedded representation. This is useful if you want more control over how to convert input_ids indices into associated vectors than the model’s internal embedding lookup matrix. 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). bool, optional) — Whether or not to return the attentions tensors of all attention layers. See attentions under returned tensors for more detail. bool, optional) — Whether or not to return the hidden states of all layers. See hidden_states under returned tensors for more detail. bool, optional) — Whether or not to return a ModelOutput instead of a plain tuple. torch.LongTensor of shape (sequence_length), optional) — Indices depicting the position of the input sequence tokens in the sequence. Contrarily to position_ids, this tensor is not affected by padding. It is used to update the cache in the correct position and to infer the complete sequence length. A transformers.modeling_outputs.BaseModelOutputWithPast 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 (ModelConfig) and inputs.
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 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 (Cache, optional, returned when use_cache=True is passed or when config.use_cache=True) — It is a Cache instance. For more details, see our kv cache guide.
Contains pre-computed hidden-states (key and values in the self-attention blocks and optionally if config.is_encoder_decoder=True in the cross-attention blocks) that can be used (see past_key_values input) to speed up sequential decoding.
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 model at the output of each layer plus the optional initial embedding outputs.
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 after the attention softmax, used to compute the weighted average in the self-attention heads.
The KyutaiSpeechToTextModel 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.
RetroSearch is an open source project built by @garambo | Open a GitHub Issue
Search and Browse the WWW like it's 1997 | Search results from DuckDuckGo
HTML:
3.2
| Encoding:
UTF-8
| Version:
0.7.4