Hugging Face's logo

Transformers documentation

FalconH1

You are viewing main version, which requires installation from source. If you'd like regular pip install, checkout the latest stable version (v4.52.2).
Hugging Face's logo
Join the Hugging Face community

and get access to the augmented documentation experience

to get started

FalconH1

Overview

The FalconH1 model was developed by the TII Pretraining team. A comprehensive research paper covering the architecture, pretraining dynamics, experimental results, and conclusions is forthcoming. You can read more about this series in this website.

Contributors

This model was contributed by DhiyaEddine, ybelkada, JingweiZuo, IlyasChahed, and MaksimVelikanov. The original code can be found here.

FalconH1Config

Model Depth Dim Attn Heads KV Mamba Heads d_head d_state Ctx Len
H1 0.5B 36 1024 8 2 24 64 / 64 128 4K, 16K-SFT
H1 1.5B 24 2048 8 2 48 128 / 64 256 128K
H1 1.5B-d 66 1280 6 2 24 128 / 64 256 128K
H1 3B 32 2560 10 2 32 128 / 128 256 128K
H1 7B 44 3072 12 2 24 128 / 128 256 256K
H1 34B 72 5120 20 4 32 128 / 128 256 256K

class transformers.FalconH1Config

< >

( vocab_size = 128000 tie_word_embeddings = False hidden_size = 4096 intermediate_size = 14336 num_hidden_layers = 32 num_attention_heads = 32 num_key_value_heads = 8 hidden_act = 'silu' initializer_range = 0.02 rms_norm_eps = 1e-05 use_cache = True num_logits_to_keep = 1 pad_token_id = 0 bos_token_id = 1 eos_token_id = 2 max_position_embeddings = 8192 attention_dropout = 0.0 mamba_d_ssm = 1024 mamba_n_heads = 128 mamba_d_head = 'auto' mamba_n_groups = 1 mamba_d_state = 256 mamba_d_conv = 4 mamba_expand = 2 mamba_chunk_size = 256 mamba_conv_bias = True mamba_proj_bias = False mamba_norm_before_gate = True mamba_rms_norm = False projectors_bias = False rope_theta = 100000.0 rope_scaling = None lm_head_multiplier = 1.0 embedding_multiplier = 1.0 mlp_multipliers = None key_multiplier = None attention_out_multiplier = None attention_in_multiplier = None ssm_multipliers = None ssm_in_multiplier = None ssm_out_multiplier = None **kwargs )

Parameters

  • vocab_size (int, optional, defaults to 128000) — Vocabulary size of the FalconH1 model. Defines the number of different tokens that can be represented by the inputs_ids passed when calling FalconH1Model
  • tie_word_embeddings (bool, optional, defaults to False) — Whether the model’s input and output word embeddings should be tied. Note that this is only relevant if the model has a output word embedding layer.
  • hidden_size (int, optional, defaults to 4096) — Dimension of the hidden representations.
  • intermediate_size (int, optional, defaults to 14336) — Dimension of the MLP representations.
  • num_hidden_layers (int, optional, defaults to 32) — Number of hidden layers in the Transformer encoder.
  • num_attention_heads (int, optional, defaults to 32) — Number of attention heads for each attention layer in the Transformer encoder.
  • 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 checkout this paper. If it is not specified, will default to 8.
  • hidden_act (str or function, optional, defaults to "silu") — The non-linear activation function (function or string) in the decoder.
  • initializer_range (float, optional, defaults to 0.02) — The standard deviation of the truncated_normal_initializer for initializing all weight matrices.
  • rms_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.
  • num_logits_to_keep (int or None, optional, defaults to 1) — Number of prompt logits to calculate during generation. If None, all logits will be calculated. If an integer value, only last num_logits_to_keep logits will be calculated. Default is 1 because only the logits of the last prompt token are needed for generation. For long sequences, the logits for the entire sequence may use a lot of memory so, setting num_logits_to_keep=1 will reduce memory footprint significantly.
  • pad_token_id (int, optional, defaults to 0) — The id of the padding token.
  • bos_token_id (int, optional, defaults to 1) — The id of the “beginning-of-sequence” token.
  • eos_token_id (int, optional, defaults to 2) — The id of the “end-of-sequence” token.
  • max_position_embeddings (int, optional, defaults to 8192) — Max cached sequence length for the model
  • attention_dropout (float, optional, defaults to 0.0) — The dropout ratio for the attention probabilities.
  • mamba_d_ssm (int, optional, defaults to 1024) — The dimension of the SSM state space latents.
  • mamba_n_heads (int, optional, defaults to 128) — The number of mamba heads used in the v2 implementation.
  • mamba_d_head (int, optional, defaults to "auto") — Head embeddding dimension size
  • mamba_n_groups (int, optional, defaults to 1) — The number of the mamba groups used in the v2 implementation.
  • mamba_d_state (int, optional, defaults to 256) — The dimension the mamba state space latents
  • mamba_d_conv (int, optional, defaults to 4) — The size of the mamba convolution kernel
  • mamba_expand (int, optional, defaults to 2) — Expanding factor (relative to hidden_size) used to determine the mamba intermediate size
  • mamba_chunk_size (int, optional, defaults to 256) — The chunks in which to break the sequence when doing prefill/training
  • mamba_conv_bias (bool, optional, defaults to True) — Flag indicating whether or not to use bias in the convolution layer of the mamba mixer block.
  • mamba_proj_bias (bool, optional, defaults to False) — Flag indicating whether or not to use bias in the input and output projections ([“in_proj”, “out_proj”]) of the mamba mixer block
  • mamba_norm_before_gate (bool, optional, defaults to True) — Whether to use RMSNorm before the gate in the Mamba block
  • mamba_rms_norm (bool, optional, defaults to False) — Whether to use RMSNorm instead of LayerNorm in the Mamba block
  • projectors_bias (bool, optional, defaults to False) — Flag indicating whether or not to use bias in the input and output projections ([“in_proj”, “out_proj”]) of the attention block
  • rope_theta (float, optional, defaults to 100000.0) — The theta value used for the RoPE embeddings.
  • rope_scaling (float, optional) — The scaling value used for the RoPE embeddings. If None, no scaling is applied.
  • lm_head_multiplier (float, optional, defaults to 1.0) — The multiplier for the LM head. This is used to scale the output of the LM head.
  • embedding_multiplier (float, optional, defaults to 1.0) — The multiplier for the embedding layer. This is used to scale the output of the embedding layer.
  • mlp_multipliers (List[float], optional) — The multipliers for the MLP layers. This is used to scale the output of the MLP layers. The first value is the multiplier of gate layer, the second value is the multiplier of the down_proj layer.
  • key_multiplier (float, optional) — The multiplier for the key layer. This is used to scale the output of the key layer.
  • attention_out_multiplier (float, optional) — The multiplier for the attention output layer. This is used to scale the output of the attention output
  • attention_in_multiplier (float, optional) — The multiplier for the attention input layer. This is used to scale the output of the attention input layer.
  • ssm_multipliers (List[float], optional) — The multipliers for the SSM layers. This is used to scale the output of the SSM layers.
  • ssm_in_multiplier (float, optional) — The multiplier for the SSM input layer. This is used to scale the output of the SSM input layer.
  • ssm_out_multiplier (float, optional) — The multiplier for the SSM output layer. This is used to scale the output of the SSM output layer.

This is the configuration class to store the configuration of a FalconH1Model. It is used to instantiate a FalconH1Model model according to the specified arguments, defining the model architecture. Instantiating a configuration with defaults taken from ibm-fms/FalconH1-9.8b-2.2T-hf. The FalconH1Model is a hybrid mamba2 architecture with SwiGLU. The checkpoints are jointly trained by IBM, Princeton, and UIUC. Configuration objects inherit from PretrainedConfig and can be used to control the model outputs. Read the documentation from PretrainedConfig for more information.

FalconH1ForCausalLM

from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained("tiiuae/Falcon-H1-7B-Instruct")
tokenizer = AutoTokenizer.from_pretrained("tiiuae/Falcon-H1-7B-Instruct")

message = ["Mamba is a snake with following properties  "]
inputs = tokenizer(message, return_tensors='pt', return_token_type_ids=False)
response = model.generate(**inputs, max_new_tokens=64)
print(tokenizer.batch_decode(response, skip_special_tokens=True)[0])

class transformers.FalconH1ForCausalLM

< >

( config )

Parameters

  • config (FalconH1ForCausalLM) — 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 Falcon H1 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: LongTensor = None attention_mask: typing.Optional[torch.Tensor] = None position_ids: typing.Optional[torch.LongTensor] = None past_key_values: typing.Optional[transformers.models.falcon_h1.modeling_falcon_h1.FalconHybridMambaAttentionDynamicCache] = 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 return_dict: typing.Optional[bool] = None cache_position: typing.Optional[torch.LongTensor] = None logits_to_keep: typing.Union[int, torch.Tensor] = 0 **kwargs ) → transformers.modeling_outputs.CausalLMOutputWithPast or tuple(torch.FloatTensor)

Parameters

  • input_ids (torch.LongTensor of shape (batch_size, sequence_length)) — 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.falcon_h1.modeling_falcon_h1.FalconHybridMambaAttentionDynamicCache, 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:

    • a Cache instance, see our kv cache guide;
    • Tuple of 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).

  • 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).
  • output_attentions (bool, optional) — Whether or not to return the attentions tensors of all attention layers. See attentions under returned tensors for more detail.
  • output_hidden_states (bool, optional) — Whether or not to return the hidden states of all layers. See hidden_states under returned tensors for more detail.
  • return_dict (bool, optional) — Whether or not to return a ModelOutput instead of a plain tuple.
  • 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] or torch.Tensor, optional) — 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 (FalconH1Config) 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 FalconH1ForCausalLM 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, FalconH1ForCausalLM

>>> model = FalconH1ForCausalLM.from_pretrained("...")
>>> tokenizer = AutoTokenizer.from_pretrained("...")

>>> 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."

This HF implementation is contributed by younesbelkada and DhiaEddineRhaiem.

< > Update on GitHub

←Falcon3 FalconMamba→