Hugging Face's logo

Transformers documentation

LFM2

You are viewing main version, which requires installation from source. If you'd like regular pip install, checkout the latest stable version (v4.57.1).
Hugging Face's logo
Join the Hugging Face community

and get access to the augmented documentation experience

to get started

이 모델은 2025년 7월 10일에 출시되었으며, 2025년 7월 10일에 Hugging Face Transformers에 추가되었습니다.

PyTorch

LFM2

개요

LFM2는 Liquid AI가 개발한 차세대 Liquid Foundation Model로 egde AI와 온디바이스 배포에 특화되어 설계되었습니다.

이 모델들은 350M, 700M, 1.2B, 2.6B의 네 가지 크기의 매개변수로 제공되며, CPU, GPU, NPU 하드웨어에서 효율적으로 실행되도록 설계되었습니다. 이로 인해 특히 낮은 지연 시간, 오프라인 작동 및 개인 정보 보호가 필요한 애플리케이션에 적합합니다.

아키텍처

아키텍처는 게이트가 있는 짧은 합성곱 블록과 QK 레이어 정규화가 적용된 그룹 쿼리 어텐션 블록으로 구성됩니다. 이 설계는 선형 연산이 입력 의존적인 게이트에 의해 조절되는 동적 시스템 개념에서 비롯되었습니다. 짧은 합성곱은 특히 임베디드 SoC CPU에 최적화되어 있어, 클라우드 연결에 의존하지 않고 빠르고 로컬화된 추론이 필요한 장치에 이상적입니다.

LFM2는 제한된 속도와 메모리 환경에서 품질을 최대화되도록 설계되었습니다. 이는 퀄컴 스냅드래곤 프로세서에서 실제 최대 메모리 사용량과 추론 속도를 측정하여, 임베디드 하드웨어에서의 실제 성능에 맞게 모델을 최적화하기 위한 체계적인 아키텍처 탐색을 통해 달성되었습니다. 그 결과, 비슷한 크기의 모델에 비해 2배 빠른 디코딩 및 프리필 성능을 달성하면서도, 지식, 수학, 지시 사항 따르기, 다국어 작업 전반에서 우수한 벤치마크 성능을 유지하는 모델이 탄생했습니다.

예시

다음 예시는 AutoModelForCausalLM 클래스를 사용하여 답변을 생성하는 방법을 보여줍니다.

from transformers import AutoModelForCausalLM, AutoTokenizer

# 모델과 토크나이저를 가져옵니다
model_id = "LiquidAI/LFM2-1.2B"
model = AutoModelForCausalLM.from_pretrained(
    model_id,
    device_map="auto",
    dtype="bfloat16",
)
tokenizer = AutoTokenizer.from_pretrained(model_id)

# 답변 생성
prompt = "What is C. elegans?"
input_ids = tokenizer.apply_chat_template(
    [{"role": "user", "content": prompt}],
    add_generation_prompt=True,
    return_tensors="pt",
    tokenize=True,
)

output = model.generate(
    input_ids,
    do_sample=True,
    temperature=0.3,
    min_p=0.15,
    repetition_penalty=1.05,
    max_new_tokens=512,
)

print(tokenizer.decode(output[0], skip_special_tokens=False))

Lfm2Config

class transformers.Lfm2Config

< >

( vocab_size: typing.Optional[int] = 65536 hidden_size: typing.Optional[int] = 2560 intermediate_size: typing.Optional[int] = 12288 num_hidden_layers: typing.Optional[int] = 32 num_attention_heads: typing.Optional[int] = 32 num_key_value_heads: typing.Optional[int] = 8 max_position_embeddings: typing.Optional[int] = 128000 initializer_range: typing.Optional[float] = 0.02 norm_eps: typing.Optional[float] = 1e-05 use_cache: typing.Optional[bool] = True pad_token_id: typing.Optional[int] = 0 bos_token_id: typing.Optional[int] = 1 eos_token_id: typing.Optional[int] = 2 tie_word_embeddings: typing.Optional[bool] = True rope_parameters: typing.Union[transformers.modeling_rope_utils.RopeParameters, dict[transformers.modeling_rope_utils.RopeParameters], NoneType] = None conv_bias: typing.Optional[bool] = False conv_L_cache: typing.Optional[int] = 3 block_multiple_of: typing.Optional[int] = 256 block_ffn_dim_multiplier: typing.Optional[float] = 1.0 block_auto_adjust_ff_dim: typing.Optional[bool] = True full_attn_idxs: typing.Optional[list[int]] = None layer_types: typing.Optional[list[str]] = None **kwargs )

Parameters

  • vocab_size (int, optional, defaults to 65536) — Vocabulary size of the LLaMA model. Defines the number of different tokens that can be represented by the inputs_ids passed when calling Lfm2Model
  • hidden_size (int, optional, defaults to 2560) — Dimension of the hidden representations.
  • intermediate_size (int, optional, defaults to 12288) — Dimension of the MLP representations.
  • num_hidden_layers (int, optional, defaults to 32) — Number of hidden layers in the Transformer decoder.
  • num_attention_heads (int, optional, defaults to 32) — 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.
  • max_position_embeddings (int, optional, defaults to 128000) — The maximum sequence length that this model might ever be used with. Lfm2 1 supports up to 2048 tokens, Lfm2 2 up to 4096, CodeLfm2 up to 16384.
  • initializer_range (float, optional, defaults to 0.02) — The standard deviation of the truncated_normal_initializer for initializing all weight matrices.
  • norm_eps (float, optional, defaults to 1e-05) — The epsilon used by the rms normalization layers.
  • 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.
  • pad_token_id (int, optional, defaults to 0) — Padding token id.
  • bos_token_id (int, optional, defaults to 1) — Beginning of stream token id.
  • eos_token_id (int, optional, defaults to 2) — End of stream token id.
  • tie_word_embeddings (bool, optional, defaults to True) — Whether to tie weight embeddings
  • rope_parameters (RopeParameters, optional) — Dictionary containing the configuration parameters for the RoPE embeddings. The dictionaty should contain a value for rope_theta and optionally parameters used for scaling in case you want to use RoPE with longer max_position_embeddings.
  • conv_bias (bool, optional, defaults to False) — Whether to use bias in the conv layers.
  • conv_L_cache (int, optional, defaults to 3) — L_cache dim in the conv layers.
  • block_multiple_of (int, optional, defaults to 256) — Multiple for the intermediate_size.
  • block_ffn_dim_multiplier (float, optional, defaults to 1.0) — Multiplier for the intermediate_size.
  • block_auto_adjust_ff_dim (bool, optional, defaults to True) — Whether to adjust the dim of the intermediate_size.
  • full_attn_idxs (Optional, optional) — Index of the layers which use attention.
  • layer_types (Optional, optional) — Type of each layers.

This is the configuration class to store the configuration of a Lfm2Model. It is used to instantiate a LFM2 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 LFM2-1.2B model. e.g. LiquidAI/LFM2-1.2B

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 Lfm2Model, Lfm2Config

>>> # Initializing a LFM2 model
>>> configuration = Lfm2Config()

>>> # Initializing a model from the LFM2-1.2B style configuration
>>> model = Lfm2Model(configuration)

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

Lfm2Model

class transformers.Lfm2Model

< >

( config: Lfm2Config )

Parameters

  • config (Lfm2Config) — 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 Lfm2 Model 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

< >

( 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.models.lfm2.modeling_lfm2.Lfm2HybridConvCache] = None inputs_embeds: typing.Optional[torch.FloatTensor] = None use_cache: typing.Optional[bool] = None cache_position: typing.Optional[torch.LongTensor] = None **kwargs: typing_extensions.Unpack[transformers.utils.generic.TransformersKwargs] ) → transformers.modeling_outputs.BaseModelOutputWithPast or tuple(torch.FloatTensor)

Parameters

  • input_ids (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.

    What are input IDs?

  • attention_mask (torch.Tensor 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?

  • position_ids (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].

    What are position IDs?

  • past_key_values (~models.lfm2.modeling_lfm2.Lfm2HybridConvCache, 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).

  • inputs_embeds (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.
  • 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).
  • cache_position (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.

Returns

transformers.modeling_outputs.BaseModelOutputWithPast or tuple(torch.FloatTensor)

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 (Lfm2Config) 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 Lfm2Model 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.

Lfm2ForCausalLM

class transformers.Lfm2ForCausalLM

< >

( config )

Parameters

  • config (Lfm2ForCausalLM) — 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 Lfm2 Model for causal language modeling.

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

< >

( 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 cache_position: typing.Optional[torch.LongTensor] = None logits_to_keep: typing.Union[int, torch.Tensor] = 0 **kwargs: typing_extensions.Unpack[transformers.utils.generic.TransformersKwargs] ) → transformers.modeling_outputs.CausalLMOutputWithPast or tuple(torch.FloatTensor)

Parameters

  • input_ids (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.

    What are input IDs?

  • attention_mask (torch.Tensor 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?

  • position_ids (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].

    What are position IDs?

  • past_key_values (~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.

    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).

  • inputs_embeds (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.
  • 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].
  • 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).
  • cache_position (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.
  • logits_to_keep (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).

Returns

transformers.modeling_outputs.CausalLMOutputWithPast or tuple(torch.FloatTensor)

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 (Lfm2Config) 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 Lfm2ForCausalLM 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:

>>> from transformers import AutoTokenizer, Lfm2ForCausalLM

>>> model = Lfm2ForCausalLM.from_pretrained("meta-lfm2/Lfm2-2-7b-hf")
>>> tokenizer = AutoTokenizer.from_pretrained("meta-lfm2/Lfm2-2-7b-hf")

>>> prompt = "Hey, are you conscious? Can you talk to me?"
>>> inputs = tokenizer(prompt, return_tensors="pt")

>>> # Generate
>>> generate_ids = model.generate(inputs.input_ids, max_length=30)
>>> tokenizer.batch_decode(generate_ids, skip_special_tokens=True, clean_up_tokenization_spaces=False)[0]
"Hey, are you conscious? Can you talk to me?\nI'm not conscious, but I can talk to you."
Update on GitHub

←LED LFM2-VL→