vllm.model_executor.models.falcon_h1

Inference-only FalconH1 model.

FalconH1AttentionDecoderLayer

Bases: Module

Source code in vllm/model_executor/models/falcon_h1.py

class FalconH1AttentionDecoderLayer(nn.Module):

    def __init__(
        self,
        config: FalconH1Config,
        cache_config: Optional[CacheConfig] = None,
        quant_config: Optional[QuantizationConfig] = None,
        prefix: str = "",
    ) -> None:
        super().__init__()
        rope_theta = getattr(config, "rope_theta", 1e11)
        rope_scaling = getattr(config, "rope_scaling", None)
        max_position_embeddings = getattr(config, "max_position_embeddings",
                                          8192)
        self.hidden_size = config.hidden_size
        tp_size = get_tensor_model_parallel_world_size()
        self.total_num_heads = config.num_attention_heads
        assert self.total_num_heads % tp_size == 0
        self.num_heads = self.total_num_heads // tp_size
        self.total_num_kv_heads = config.num_key_value_heads
        if self.total_num_kv_heads >= tp_size:
            # Number of KV heads is greater than TP size, so we partition
            # the KV heads across multiple tensor parallel GPUs.
            assert self.total_num_kv_heads % tp_size == 0
        else:
            # Number of KV heads is less than TP size, so we replicate
            # the KV heads across multiple tensor parallel GPUs.
            assert tp_size % self.total_num_kv_heads == 0
        self.num_kv_heads = max(1, self.total_num_kv_heads // tp_size)
        self.head_dim = (config.hidden_size // self.total_num_heads if getattr(
            config, "head_dim", None) is None else config.head_dim)
        self.q_size = self.num_heads * self.head_dim
        self.kv_size = self.num_kv_heads * self.head_dim
        self.scaling = self.head_dim**-0.5
        self.rope_theta = rope_theta
        self.max_position_embeddings = max_position_embeddings

        if hasattr(config, "partial_rotary_factor"):
            rotary_dim = self.head_dim * config.partial_rotary_factor
        elif hasattr(config, "attn_rotary_emb"):
            rotary_dim = config.attn_rotary_emb  # for backward compatibility
        else:
            rotary_dim = self.head_dim  # default

        self.rotary_emb = get_rope(
            head_size=self.head_dim,
            rotary_dim=rotary_dim,
            max_position=max_position_embeddings,
            rope_scaling=rope_scaling,
            base=rope_theta,
            is_neox_style=True,
            dtype=None,  # see impl of get_rope
        )

        self.qkv_proj = QKVParallelLinear(
            config.hidden_size,
            self.head_dim,
            self.total_num_heads,
            self.total_num_kv_heads,
            bias=False,
            quant_config=quant_config,
            prefix=f"{prefix}.qkv_proj",
        )
        self.o_proj = RowParallelLinear(
            self.total_num_heads * self.head_dim,
            config.hidden_size,
            bias=False,
            quant_config=quant_config,
            prefix=f"{prefix}.o_proj",
        )

        self.attn = Attention(
            self.num_heads,
            self.head_dim,
            self.scaling,
            num_kv_heads=self.num_kv_heads,
            cache_config=cache_config,
            prefix=f"{prefix}.attn",
        )
        self.key_multiplier = config.key_multiplier

    def self_attention(
        self,
        positions: torch.Tensor,
        hidden_states: torch.Tensor,
        **kwargs,
    ) -> torch.Tensor:
        qkv, _ = self.qkv_proj(hidden_states)
        q, k, v = qkv.split([self.q_size, self.kv_size, self.kv_size], dim=-1)
        k = k * self.key_multiplier

        q, k = self.rotary_emb(positions, q, k)
        attn_output = self.attn(q, k, v)
        output, _ = self.o_proj(attn_output)
        return output

    def forward(
        self,
        positions: torch.Tensor,
        hidden_states: torch.Tensor,
        residual: Optional[torch.Tensor],
        **kwargs,
    ):
        hidden_states = self.self_attention(
            positions=positions,
            hidden_states=hidden_states,
        )
        return hidden_states, residual

attn instance-attribute

attn = Attention(
    num_heads,
    head_dim,
    scaling,
    num_kv_heads=num_kv_heads,
    cache_config=cache_config,
    prefix=f"{prefix}.attn",
)

head_dim instance-attribute

head_dim = (
    hidden_size // total_num_heads
    if getattr(config, "head_dim", None) is None
    else head_dim
)

hidden_size instance-attribute

hidden_size = hidden_size

key_multiplier instance-attribute

key_multiplier = key_multiplier

kv_size instance-attribute

kv_size = num_kv_heads * head_dim

max_position_embeddings instance-attribute

max_position_embeddings = max_position_embeddings

num_heads instance-attribute

num_heads = total_num_heads // tp_size

num_kv_heads instance-attribute

num_kv_heads = max(1, total_num_kv_heads // tp_size)

o_proj instance-attribute

o_proj = RowParallelLinear(
    total_num_heads * head_dim,
    hidden_size,
    bias=False,
    quant_config=quant_config,
    prefix=f"{prefix}.o_proj",
)

q_size instance-attribute

q_size = num_heads * head_dim

qkv_proj instance-attribute

qkv_proj = QKVParallelLinear(
    hidden_size,
    head_dim,
    total_num_heads,
    total_num_kv_heads,
    bias=False,
    quant_config=quant_config,
    prefix=f"{prefix}.qkv_proj",
)

rope_theta instance-attribute

rope_theta = rope_theta

rotary_emb instance-attribute

rotary_emb = get_rope(
    head_size=head_dim,
    rotary_dim=rotary_dim,
    max_position=max_position_embeddings,
    rope_scaling=rope_scaling,
    base=rope_theta,
    is_neox_style=True,
    dtype=None,
)

scaling instance-attribute

scaling = head_dim ** -0.5

total_num_heads instance-attribute

total_num_heads = num_attention_heads

total_num_kv_heads instance-attribute

total_num_kv_heads = num_key_value_heads

__init__

__init__(
    config: FalconH1Config,
    cache_config: Optional[CacheConfig] = None,
    quant_config: Optional[QuantizationConfig] = None,
    prefix: str = "",
) -> None

Source code in vllm/model_executor/models/falcon_h1.py

def __init__(
    self,
    config: FalconH1Config,
    cache_config: Optional[CacheConfig] = None,
    quant_config: Optional[QuantizationConfig] = None,
    prefix: str = "",
) -> None:
    super().__init__()
    rope_theta = getattr(config, "rope_theta", 1e11)
    rope_scaling = getattr(config, "rope_scaling", None)
    max_position_embeddings = getattr(config, "max_position_embeddings",
                                      8192)
    self.hidden_size = config.hidden_size
    tp_size = get_tensor_model_parallel_world_size()
    self.total_num_heads = config.num_attention_heads
    assert self.total_num_heads % tp_size == 0
    self.num_heads = self.total_num_heads // tp_size
    self.total_num_kv_heads = config.num_key_value_heads
    if self.total_num_kv_heads >= tp_size:
        # Number of KV heads is greater than TP size, so we partition
        # the KV heads across multiple tensor parallel GPUs.
        assert self.total_num_kv_heads % tp_size == 0
    else:
        # Number of KV heads is less than TP size, so we replicate
        # the KV heads across multiple tensor parallel GPUs.
        assert tp_size % self.total_num_kv_heads == 0
    self.num_kv_heads = max(1, self.total_num_kv_heads // tp_size)
    self.head_dim = (config.hidden_size // self.total_num_heads if getattr(
        config, "head_dim", None) is None else config.head_dim)
    self.q_size = self.num_heads * self.head_dim
    self.kv_size = self.num_kv_heads * self.head_dim
    self.scaling = self.head_dim**-0.5
    self.rope_theta = rope_theta
    self.max_position_embeddings = max_position_embeddings

    if hasattr(config, "partial_rotary_factor"):
        rotary_dim = self.head_dim * config.partial_rotary_factor
    elif hasattr(config, "attn_rotary_emb"):
        rotary_dim = config.attn_rotary_emb  # for backward compatibility
    else:
        rotary_dim = self.head_dim  # default

    self.rotary_emb = get_rope(
        head_size=self.head_dim,
        rotary_dim=rotary_dim,
        max_position=max_position_embeddings,
        rope_scaling=rope_scaling,
        base=rope_theta,
        is_neox_style=True,
        dtype=None,  # see impl of get_rope
    )

    self.qkv_proj = QKVParallelLinear(
        config.hidden_size,
        self.head_dim,
        self.total_num_heads,
        self.total_num_kv_heads,
        bias=False,
        quant_config=quant_config,
        prefix=f"{prefix}.qkv_proj",
    )
    self.o_proj = RowParallelLinear(
        self.total_num_heads * self.head_dim,
        config.hidden_size,
        bias=False,
        quant_config=quant_config,
        prefix=f"{prefix}.o_proj",
    )

    self.attn = Attention(
        self.num_heads,
        self.head_dim,
        self.scaling,
        num_kv_heads=self.num_kv_heads,
        cache_config=cache_config,
        prefix=f"{prefix}.attn",
    )
    self.key_multiplier = config.key_multiplier

forward

forward(
    positions: Tensor,
    hidden_states: Tensor,
    residual: Optional[Tensor],
    **kwargs,
)

Source code in vllm/model_executor/models/falcon_h1.py

def forward(
    self,
    positions: torch.Tensor,
    hidden_states: torch.Tensor,
    residual: Optional[torch.Tensor],
    **kwargs,
):
    hidden_states = self.self_attention(
        positions=positions,
        hidden_states=hidden_states,
    )
    return hidden_states, residual

self_attention

self_attention(
    positions: Tensor, hidden_states: Tensor, **kwargs
) -> Tensor

Source code in vllm/model_executor/models/falcon_h1.py

def self_attention(
    self,
    positions: torch.Tensor,
    hidden_states: torch.Tensor,
    **kwargs,
) -> torch.Tensor:
    qkv, _ = self.qkv_proj(hidden_states)
    q, k, v = qkv.split([self.q_size, self.kv_size, self.kv_size], dim=-1)
    k = k * self.key_multiplier

    q, k = self.rotary_emb(positions, q, k)
    attn_output = self.attn(q, k, v)
    output, _ = self.o_proj(attn_output)
    return output

FalconH1ForCausalLM

Bases: Module, HasInnerState, SupportsLoRA, SupportsPP, IsHybrid

Source code in vllm/model_executor/models/falcon_h1.py

class FalconH1ForCausalLM(nn.Module, HasInnerState, SupportsLoRA, SupportsPP,
                          IsHybrid):
    packed_modules_mapping = {
        "qkv_proj": ["q_proj", "k_proj", "v_proj"],
        "gate_up_proj": ["gate_proj", "up_proj"],
    }

    embedding_modules = {
        "embed_tokens": "input_embeddings",
        "lm_head": "output_embeddings",
    }
    embedding_padding_modules = ["lm_head"]

    def __init__(self, *, vllm_config: VllmConfig, prefix: str = ""):
        config = vllm_config.model_config.hf_config
        self.vllm_config = vllm_config
        self.model_config = vllm_config.model_config
        cache_config = vllm_config.cache_config
        lora_config = vllm_config.lora_config
        scheduler_config = vllm_config.scheduler_config
        assert (not cache_config.enable_prefix_caching
                ), "FalconH1 currently does not support prefix caching"

        self.quant_config = vllm_config.quant_config

        super().__init__()
        self.config = config
        self.scheduler_config = scheduler_config
        self.model = FalconH1Model(vllm_config=vllm_config,
                                   prefix=maybe_prefix(prefix, "model"))
        self.tie_word_embeddings = config.tie_word_embeddings
        self.unpadded_vocab_size = config.vocab_size
        self.mamba_cache: Optional[MambaCacheManager] = None
        if lora_config:
            self.unpadded_vocab_size += lora_config.lora_extra_vocab_size
        if get_pp_group().is_last_rank:
            self.lm_head = ParallelLMHead(
                self.unpadded_vocab_size,
                config.hidden_size,
                org_num_embeddings=config.vocab_size,
                padding_size=(
                    DEFAULT_VOCAB_PADDING_SIZE
                    # We need bigger padding if using lora for kernel
                    # compatibility
                    if not lora_config else
                    lora_config.lora_vocab_padding_size),
            )
            self.lm_head_multiplier = config.lm_head_multiplier
            if self.tie_word_embeddings:
                self.lm_head = self.lm_head.tie_weights(
                    self.model.embed_tokens)
            # Used to track and store by the Mamba cache between steps.

            self.logits_processor = LogitsProcessor(
                self.unpadded_vocab_size,
                config.vocab_size,
                scale=config.lm_head_multiplier,
            )
        else:
            self.lm_head = PPMissingLayer()

        self.make_empty_intermediate_tensors = (
            self.model.make_empty_intermediate_tensors)

    def get_input_embeddings(self, input_ids: torch.Tensor) -> torch.Tensor:
        return self.model.get_input_embeddings(input_ids)

    def forward(
        self,
        input_ids: torch.Tensor,
        positions: torch.Tensor,
        intermediate_tensors: Optional[IntermediateTensors] = None,
        inputs_embeds: Optional[torch.Tensor] = None,
        **kwargs,
    ):

        mamba_cache_params = None
        if not envs.VLLM_USE_V1:
            if self.mamba_cache is None:
                self.mamba_cache = MambaCacheManager(
                    self.vllm_config,
                    self.lm_head.weight.dtype if hasattr(
                        self.lm_head, 'weight') else torch.bfloat16,
                    self.config.num_hidden_layers,
                    *self._get_mamba_cache_shape(),
                )
            mamba_cache_params = self.mamba_cache.current_run_tensors(**kwargs)

        hidden_states = self.model(
            input_ids,
            positions,
            mamba_cache_params,
            intermediate_tensors,
            inputs_embeds,
        )

        return hidden_states

    def copy_inputs_before_cuda_graphs(self, input_buffers, **kwargs):
        return self.mamba_cache.copy_inputs_before_cuda_graphs(
            input_buffers, **kwargs)

    def get_seqlen_agnostic_capture_inputs(self, batch_size: int):
        return self.mamba_cache.get_seqlen_agnostic_capture_inputs(batch_size)

    def _get_mamba_cache_shape(
            self) -> tuple[tuple[int, int], tuple[int, int]]:
        world_size = get_tensor_model_parallel_world_size()
        hidden_size = self.config.hidden_size

        conv_state_shape, temporal_state_shape = None, None

        intermediate_size = (int(self.config.mamba_expand *
                                 hidden_size) if self.config.mamba_d_ssm
                             is None else self.config.mamba_d_ssm)

        # if n_groups is not divisible by world_size, need to extend the shards
        # to ensure all groups needed by a head is sharded along with it
        n_groups = self.config.mamba_n_groups + extra_groups_for_head_shards(
            self.config.mamba_n_groups, world_size)

        # - heads and n_groups are TP-ed
        conv_dim = intermediate_size + 2 * n_groups * self.config.mamba_d_state
        conv_state_shape = (
            divide(conv_dim, world_size),
            self.config.mamba_d_conv - 1,
        )

        # These are not TP-ed as they depend on A, dt_bias, D
        # - they are typically small
        #   e.g., (h_heads, d_head, d_state) = (128, 64, 128)
        temporal_state_shape = (
            divide(self.config.mamba_n_heads, world_size),
            self.config.mamba_d_head,
            self.config.mamba_d_state,
        )
        return conv_state_shape, temporal_state_shape

    def compute_logits(
        self,
        hidden_states: torch.Tensor,
        sampling_metadata: SamplingMetadata,
    ) -> Optional[torch.Tensor]:
        logits = self.logits_processor(self.lm_head, hidden_states,
                                       sampling_metadata)

        return logits

    def load_weights(self, weights: Iterable[tuple[str,
                                                   torch.Tensor]]) -> set[str]:
        stacked_params_mapping = [
            # (param_name, shard_name, shard_id)
            ("qkv_proj", "q_proj", "q"),
            ("qkv_proj", "k_proj", "k"),
            ("qkv_proj", "v_proj", "v"),
            ("gate_up_proj", "gate_proj", 0),
            ("gate_up_proj", "up_proj", 1),
        ]

        params_dict = dict(self.named_parameters())
        loaded_params: set[str] = set()
        for name, loaded_weight in weights:
            if "rotary_emb.inv_freq" in name:
                continue

            if "A_log" in name:
                name = name.replace("A_log", "A")

            if "mamba" in name:
                name = name.replace("mamba", "mamba.mamba")

            for param_name, weight_name, shard_id in stacked_params_mapping:
                if weight_name not in name:
                    continue

                name = name.replace(weight_name, param_name)
                # Skip loading extra bias for GPTQ models.
                if name.endswith(".bias") and name not in params_dict:
                    continue
                # Skip layers on other devices.
                if is_pp_missing_parameter(name, self):
                    continue
                param = params_dict[name]
                weight_loader = param.weight_loader
                weight_loader(param, loaded_weight, shard_id)
                break
            else:
                # Skip loading extra bias for GPTQ models.
                if name.endswith(".bias") and name not in params_dict:
                    continue
                if is_pp_missing_parameter(name, self):
                    continue
                if self.tie_word_embeddings and "lm_head" in name:
                    continue

                param = params_dict[name]
                weight_loader = getattr(param, "weight_loader",
                                        default_weight_loader)
                weight_loader(param, loaded_weight)
            loaded_params.add(name)

        if self.tie_word_embeddings:
            loaded_params.add("lm_head.weight")
        return loaded_params

config instance-attribute

config = config

embedding_modules class-attribute instance-attribute

embedding_modules = {
    "embed_tokens": "input_embeddings",
    "lm_head": "output_embeddings",
}

embedding_padding_modules class-attribute instance-attribute

embedding_padding_modules = ['lm_head']

lm_head instance-attribute

lm_head = ParallelLMHead(
    unpadded_vocab_size,
    hidden_size,
    org_num_embeddings=vocab_size,
    padding_size=DEFAULT_VOCAB_PADDING_SIZE
    if not lora_config
    else lora_vocab_padding_size,
)

lm_head_multiplier instance-attribute

lm_head_multiplier = lm_head_multiplier

logits_processor instance-attribute

logits_processor = LogitsProcessor(
    unpadded_vocab_size,
    vocab_size,
    scale=lm_head_multiplier,
)

make_empty_intermediate_tensors instance-attribute

make_empty_intermediate_tensors = (
    make_empty_intermediate_tensors
)

mamba_cache instance-attribute

mamba_cache: Optional[MambaCacheManager] = None

model instance-attribute

model = FalconH1Model(
    vllm_config=vllm_config,
    prefix=maybe_prefix(prefix, "model"),
)

model_config instance-attribute

model_config = model_config

packed_modules_mapping class-attribute instance-attribute

packed_modules_mapping = {
    "qkv_proj": ["q_proj", "k_proj", "v_proj"],
    "gate_up_proj": ["gate_proj", "up_proj"],
}

quant_config instance-attribute

quant_config = quant_config

scheduler_config instance-attribute

scheduler_config = scheduler_config

tie_word_embeddings instance-attribute

tie_word_embeddings = tie_word_embeddings

unpadded_vocab_size instance-attribute

unpadded_vocab_size = vocab_size

vllm_config instance-attribute

vllm_config = vllm_config

__init__

__init__(*, vllm_config: VllmConfig, prefix: str = '')

Source code in vllm/model_executor/models/falcon_h1.py

def __init__(self, *, vllm_config: VllmConfig, prefix: str = ""):
    config = vllm_config.model_config.hf_config
    self.vllm_config = vllm_config
    self.model_config = vllm_config.model_config
    cache_config = vllm_config.cache_config
    lora_config = vllm_config.lora_config
    scheduler_config = vllm_config.scheduler_config
    assert (not cache_config.enable_prefix_caching
            ), "FalconH1 currently does not support prefix caching"

    self.quant_config = vllm_config.quant_config

    super().__init__()
    self.config = config
    self.scheduler_config = scheduler_config
    self.model = FalconH1Model(vllm_config=vllm_config,
                               prefix=maybe_prefix(prefix, "model"))
    self.tie_word_embeddings = config.tie_word_embeddings
    self.unpadded_vocab_size = config.vocab_size
    self.mamba_cache: Optional[MambaCacheManager] = None
    if lora_config:
        self.unpadded_vocab_size += lora_config.lora_extra_vocab_size
    if get_pp_group().is_last_rank:
        self.lm_head = ParallelLMHead(
            self.unpadded_vocab_size,
            config.hidden_size,
            org_num_embeddings=config.vocab_size,
            padding_size=(
                DEFAULT_VOCAB_PADDING_SIZE
                # We need bigger padding if using lora for kernel
                # compatibility
                if not lora_config else
                lora_config.lora_vocab_padding_size),
        )
        self.lm_head_multiplier = config.lm_head_multiplier
        if self.tie_word_embeddings:
            self.lm_head = self.lm_head.tie_weights(
                self.model.embed_tokens)
        # Used to track and store by the Mamba cache between steps.

        self.logits_processor = LogitsProcessor(
            self.unpadded_vocab_size,
            config.vocab_size,
            scale=config.lm_head_multiplier,
        )
    else:
        self.lm_head = PPMissingLayer()

    self.make_empty_intermediate_tensors = (
        self.model.make_empty_intermediate_tensors)

_get_mamba_cache_shape

_get_mamba_cache_shape() -> tuple[
    tuple[int, int], tuple[int, int]
]

Source code in vllm/model_executor/models/falcon_h1.py

def _get_mamba_cache_shape(
        self) -> tuple[tuple[int, int], tuple[int, int]]:
    world_size = get_tensor_model_parallel_world_size()
    hidden_size = self.config.hidden_size

    conv_state_shape, temporal_state_shape = None, None

    intermediate_size = (int(self.config.mamba_expand *
                             hidden_size) if self.config.mamba_d_ssm
                         is None else self.config.mamba_d_ssm)

    # if n_groups is not divisible by world_size, need to extend the shards
    # to ensure all groups needed by a head is sharded along with it
    n_groups = self.config.mamba_n_groups + extra_groups_for_head_shards(
        self.config.mamba_n_groups, world_size)

    # - heads and n_groups are TP-ed
    conv_dim = intermediate_size + 2 * n_groups * self.config.mamba_d_state
    conv_state_shape = (
        divide(conv_dim, world_size),
        self.config.mamba_d_conv - 1,
    )

    # These are not TP-ed as they depend on A, dt_bias, D
    # - they are typically small
    #   e.g., (h_heads, d_head, d_state) = (128, 64, 128)
    temporal_state_shape = (
        divide(self.config.mamba_n_heads, world_size),
        self.config.mamba_d_head,
        self.config.mamba_d_state,
    )
    return conv_state_shape, temporal_state_shape

compute_logits

compute_logits(
    hidden_states: Tensor,
    sampling_metadata: SamplingMetadata,
) -> Optional[Tensor]

Source code in vllm/model_executor/models/falcon_h1.py

def compute_logits(
    self,
    hidden_states: torch.Tensor,
    sampling_metadata: SamplingMetadata,
) -> Optional[torch.Tensor]:
    logits = self.logits_processor(self.lm_head, hidden_states,
                                   sampling_metadata)

    return logits

copy_inputs_before_cuda_graphs

copy_inputs_before_cuda_graphs(input_buffers, **kwargs)

Source code in vllm/model_executor/models/falcon_h1.py

def copy_inputs_before_cuda_graphs(self, input_buffers, **kwargs):
    return self.mamba_cache.copy_inputs_before_cuda_graphs(
        input_buffers, **kwargs)

forward

forward(
    input_ids: Tensor,
    positions: Tensor,
    intermediate_tensors: Optional[
        IntermediateTensors
    ] = None,
    inputs_embeds: Optional[Tensor] = None,
    **kwargs,
)

Source code in vllm/model_executor/models/falcon_h1.py

def forward(
    self,
    input_ids: torch.Tensor,
    positions: torch.Tensor,
    intermediate_tensors: Optional[IntermediateTensors] = None,
    inputs_embeds: Optional[torch.Tensor] = None,
    **kwargs,
):

    mamba_cache_params = None
    if not envs.VLLM_USE_V1:
        if self.mamba_cache is None:
            self.mamba_cache = MambaCacheManager(
                self.vllm_config,
                self.lm_head.weight.dtype if hasattr(
                    self.lm_head, 'weight') else torch.bfloat16,
                self.config.num_hidden_layers,
                *self._get_mamba_cache_shape(),
            )
        mamba_cache_params = self.mamba_cache.current_run_tensors(**kwargs)

    hidden_states = self.model(
        input_ids,
        positions,
        mamba_cache_params,
        intermediate_tensors,
        inputs_embeds,
    )

    return hidden_states

get_input_embeddings

get_input_embeddings(input_ids: Tensor) -> Tensor

Source code in vllm/model_executor/models/falcon_h1.py

def get_input_embeddings(self, input_ids: torch.Tensor) -> torch.Tensor:
    return self.model.get_input_embeddings(input_ids)

get_seqlen_agnostic_capture_inputs

get_seqlen_agnostic_capture_inputs(batch_size: int)

Source code in vllm/model_executor/models/falcon_h1.py

def get_seqlen_agnostic_capture_inputs(self, batch_size: int):
    return self.mamba_cache.get_seqlen_agnostic_capture_inputs(batch_size)

load_weights

load_weights(
    weights: Iterable[tuple[str, Tensor]],
) -> set[str]

Source code in vllm/model_executor/models/falcon_h1.py

def load_weights(self, weights: Iterable[tuple[str,
                                               torch.Tensor]]) -> set[str]:
    stacked_params_mapping = [
        # (param_name, shard_name, shard_id)
        ("qkv_proj", "q_proj", "q"),
        ("qkv_proj", "k_proj", "k"),
        ("qkv_proj", "v_proj", "v"),
        ("gate_up_proj", "gate_proj", 0),
        ("gate_up_proj", "up_proj", 1),
    ]

    params_dict = dict(self.named_parameters())
    loaded_params: set[str] = set()
    for name, loaded_weight in weights:
        if "rotary_emb.inv_freq" in name:
            continue

        if "A_log" in name:
            name = name.replace("A_log", "A")

        if "mamba" in name:
            name = name.replace("mamba", "mamba.mamba")

        for param_name, weight_name, shard_id in stacked_params_mapping:
            if weight_name not in name:
                continue

            name = name.replace(weight_name, param_name)
            # Skip loading extra bias for GPTQ models.
            if name.endswith(".bias") and name not in params_dict:
                continue
            # Skip layers on other devices.
            if is_pp_missing_parameter(name, self):
                continue
            param = params_dict[name]
            weight_loader = param.weight_loader
            weight_loader(param, loaded_weight, shard_id)
            break
        else:
            # Skip loading extra bias for GPTQ models.
            if name.endswith(".bias") and name not in params_dict:
                continue
            if is_pp_missing_parameter(name, self):
                continue
            if self.tie_word_embeddings and "lm_head" in name:
                continue

            param = params_dict[name]
            weight_loader = getattr(param, "weight_loader",
                                    default_weight_loader)
            weight_loader(param, loaded_weight)
        loaded_params.add(name)

    if self.tie_word_embeddings:
        loaded_params.add("lm_head.weight")
    return loaded_params

FalconH1MLP

Bases: Module

Source code in vllm/model_executor/models/falcon_h1.py

class FalconH1MLP(nn.Module):

    def __init__(
        self,
        config: FalconH1Config,
        quant_config: Optional[QuantizationConfig] = None,
        bias: bool = False,
    ) -> None:
        super().__init__()
        self.gate_up_proj = MergedColumnParallelLinear(
            input_size=config.hidden_size,
            output_sizes=[config.intermediate_size] * 2,
            bias=bias,
            quant_config=quant_config,
        )
        self.down_proj = RowParallelLinear(
            input_size=config.intermediate_size,
            output_size=config.hidden_size,
            bias=bias,
            quant_config=quant_config,
        )
        self.tp_size = get_tensor_model_parallel_world_size()
        self.intermediate_size = config.intermediate_size
        self.gate_multiplier, self.down_multiplier = config.mlp_multipliers
        if config.hidden_act != "silu":
            raise ValueError(f"Unsupported activation: {config.hidden_act}. "
                             "Only silu is supported for now.")
        self.act_fn = SiluAndMul()

    def forward(self, x):
        x, _ = self.gate_up_proj(x)
        x[:, :self.intermediate_size // self.tp_size] *= self.gate_multiplier
        x = self.act_fn(x)
        x, _ = self.down_proj(x)
        x = x * self.down_multiplier
        return x

act_fn instance-attribute

act_fn = SiluAndMul()

down_proj instance-attribute

down_proj = RowParallelLinear(
    input_size=intermediate_size,
    output_size=hidden_size,
    bias=bias,
    quant_config=quant_config,
)

gate_up_proj instance-attribute

gate_up_proj = MergedColumnParallelLinear(
    input_size=hidden_size,
    output_sizes=[intermediate_size] * 2,
    bias=bias,
    quant_config=quant_config,
)

intermediate_size instance-attribute

intermediate_size = intermediate_size

tp_size instance-attribute

tp_size = get_tensor_model_parallel_world_size()

__init__

__init__(
    config: FalconH1Config,
    quant_config: Optional[QuantizationConfig] = None,
    bias: bool = False,
) -> None

Source code in vllm/model_executor/models/falcon_h1.py

def __init__(
    self,
    config: FalconH1Config,
    quant_config: Optional[QuantizationConfig] = None,
    bias: bool = False,
) -> None:
    super().__init__()
    self.gate_up_proj = MergedColumnParallelLinear(
        input_size=config.hidden_size,
        output_sizes=[config.intermediate_size] * 2,
        bias=bias,
        quant_config=quant_config,
    )
    self.down_proj = RowParallelLinear(
        input_size=config.intermediate_size,
        output_size=config.hidden_size,
        bias=bias,
        quant_config=quant_config,
    )
    self.tp_size = get_tensor_model_parallel_world_size()
    self.intermediate_size = config.intermediate_size
    self.gate_multiplier, self.down_multiplier = config.mlp_multipliers
    if config.hidden_act != "silu":
        raise ValueError(f"Unsupported activation: {config.hidden_act}. "
                         "Only silu is supported for now.")
    self.act_fn = SiluAndMul()

forward

forward(x)

Source code in vllm/model_executor/models/falcon_h1.py

def forward(self, x):
    x, _ = self.gate_up_proj(x)
    x[:, :self.intermediate_size // self.tp_size] *= self.gate_multiplier
    x = self.act_fn(x)
    x, _ = self.down_proj(x)
    x = x * self.down_multiplier
    return x

FalconH1Model

Bases: Module

Source code in vllm/model_executor/models/falcon_h1.py

class FalconH1Model(nn.Module):

    def __init__(self, *, vllm_config: VllmConfig, prefix: str = ""):
        super().__init__()
        config: FalconH1Config = vllm_config.model_config.hf_config
        cache_config = vllm_config.cache_config
        quant_config = vllm_config.quant_config
        lora_config = vllm_config.lora_config

        self.config = config
        lora_vocab = ((lora_config.lora_extra_vocab_size *
                       (lora_config.max_loras or 1)) if lora_config else 0)
        self.vocab_size = config.vocab_size + lora_vocab
        self.org_vocab_size = config.vocab_size
        if get_pp_group().is_first_rank:

            self.embed_tokens = VocabParallelEmbedding(
                self.vocab_size,
                config.hidden_size,
                org_num_embeddings=config.vocab_size,
            )
            self.embedding_multiplier = config.embedding_multiplier
        else:
            self.embed_tokens = PPMissingLayer()
            self.embedding_multiplier = 1.0

        def get_layer(prefix: str):
            layer_idx = int(prefix.rsplit(".", 1)[1])
            layer_class = FalconH1ParallelHybrid
            return layer_class(
                config,
                layer_idx,
                cache_config,
                quant_config=quant_config,
                prefix=prefix,
            )

        self.start_layer, self.end_layer, self.layers = make_layers(
            config.num_hidden_layers, get_layer, prefix=f"{prefix}.layers")
        self.make_empty_intermediate_tensors = (
            make_empty_intermediate_tensors_factory(
                ["hidden_states", "residual"], config.hidden_size))
        if get_pp_group().is_last_rank:
            self.final_layernorm = RMSNorm(config.hidden_size,
                                           eps=config.rms_norm_eps)
        else:
            self.final_layernorm = PPMissingLayer()

    def get_input_embeddings(self, input_ids: torch.Tensor) -> torch.Tensor:
        return self.embed_tokens(input_ids)

    def forward(
        self,
        input_ids: torch.Tensor,
        positions: torch.Tensor,
        mamba_cache_params: MambaCacheParams,
        intermediate_tensors: Optional[IntermediateTensors] = None,
        inputs_embeds: Optional[torch.Tensor] = None,
    ) -> torch.Tensor:

        # pass a sequence index tensor, that is required for
        # proper continuous batching computation including
        # chunked prefill
        attn_metadata = get_forward_context().attn_metadata

        if not envs.VLLM_USE_V1:
            mamba2_metadata = prepare_mamba2_metadata(
                chunk_size=self.config.mamba_chunk_size,
                attn_metadata=attn_metadata,
            )
        else:
            # v1 get mamba2_metadata from forward_context
            mamba2_metadata = None

        if get_pp_group().is_first_rank:
            if inputs_embeds is not None:
                hidden_states = inputs_embeds * self.embedding_multiplier
            else:
                hidden_states = (self.get_input_embeddings(input_ids) *
                                 self.embedding_multiplier)
        else:
            assert intermediate_tensors is not None
            hidden_states = intermediate_tensors["hidden_states"]

        for i in range(self.start_layer, self.end_layer):
            layer = self.layers[i]
            layer_mamba_cache_params = None
            if mamba_cache_params:
                layer_mamba_cache_params = mamba_cache_params.at_layer_idx(i)
            hidden_states = layer(
                positions=positions,
                hidden_states=hidden_states,
                mamba_cache_params=layer_mamba_cache_params,
                mamba2_metadata=mamba2_metadata,
            )
        if not get_pp_group().is_last_rank:
            return IntermediateTensors({
                "hidden_states": hidden_states,
            })
        hidden_states = self.final_layernorm(hidden_states)
        return hidden_states

config instance-attribute

config = config

embed_tokens instance-attribute

embed_tokens = VocabParallelEmbedding(
    vocab_size, hidden_size, org_num_embeddings=vocab_size
)

embedding_multiplier instance-attribute

embedding_multiplier = embedding_multiplier

final_layernorm instance-attribute

final_layernorm = RMSNorm(hidden_size, eps=rms_norm_eps)

make_empty_intermediate_tensors instance-attribute

make_empty_intermediate_tensors = (
    make_empty_intermediate_tensors_factory(
        ["hidden_states", "residual"], hidden_size
    )
)

org_vocab_size instance-attribute

org_vocab_size = vocab_size

vocab_size instance-attribute

vocab_size = vocab_size + lora_vocab

__init__

__init__(*, vllm_config: VllmConfig, prefix: str = '')

Source code in vllm/model_executor/models/falcon_h1.py

def __init__(self, *, vllm_config: VllmConfig, prefix: str = ""):
    super().__init__()
    config: FalconH1Config = vllm_config.model_config.hf_config
    cache_config = vllm_config.cache_config
    quant_config = vllm_config.quant_config
    lora_config = vllm_config.lora_config

    self.config = config
    lora_vocab = ((lora_config.lora_extra_vocab_size *
                   (lora_config.max_loras or 1)) if lora_config else 0)
    self.vocab_size = config.vocab_size + lora_vocab
    self.org_vocab_size = config.vocab_size
    if get_pp_group().is_first_rank:

        self.embed_tokens = VocabParallelEmbedding(
            self.vocab_size,
            config.hidden_size,
            org_num_embeddings=config.vocab_size,
        )
        self.embedding_multiplier = config.embedding_multiplier
    else:
        self.embed_tokens = PPMissingLayer()
        self.embedding_multiplier = 1.0

    def get_layer(prefix: str):
        layer_idx = int(prefix.rsplit(".", 1)[1])
        layer_class = FalconH1ParallelHybrid
        return layer_class(
            config,
            layer_idx,
            cache_config,
            quant_config=quant_config,
            prefix=prefix,
        )

    self.start_layer, self.end_layer, self.layers = make_layers(
        config.num_hidden_layers, get_layer, prefix=f"{prefix}.layers")
    self.make_empty_intermediate_tensors = (
        make_empty_intermediate_tensors_factory(
            ["hidden_states", "residual"], config.hidden_size))
    if get_pp_group().is_last_rank:
        self.final_layernorm = RMSNorm(config.hidden_size,
                                       eps=config.rms_norm_eps)
    else:
        self.final_layernorm = PPMissingLayer()

forward

forward(
    input_ids: Tensor,
    positions: Tensor,
    mamba_cache_params: MambaCacheParams,
    intermediate_tensors: Optional[
        IntermediateTensors
    ] = None,
    inputs_embeds: Optional[Tensor] = None,
) -> Tensor

Source code in vllm/model_executor/models/falcon_h1.py

def forward(
    self,
    input_ids: torch.Tensor,
    positions: torch.Tensor,
    mamba_cache_params: MambaCacheParams,
    intermediate_tensors: Optional[IntermediateTensors] = None,
    inputs_embeds: Optional[torch.Tensor] = None,
) -> torch.Tensor:

    # pass a sequence index tensor, that is required for
    # proper continuous batching computation including
    # chunked prefill
    attn_metadata = get_forward_context().attn_metadata

    if not envs.VLLM_USE_V1:
        mamba2_metadata = prepare_mamba2_metadata(
            chunk_size=self.config.mamba_chunk_size,
            attn_metadata=attn_metadata,
        )
    else:
        # v1 get mamba2_metadata from forward_context
        mamba2_metadata = None

    if get_pp_group().is_first_rank:
        if inputs_embeds is not None:
            hidden_states = inputs_embeds * self.embedding_multiplier
        else:
            hidden_states = (self.get_input_embeddings(input_ids) *
                             self.embedding_multiplier)
    else:
        assert intermediate_tensors is not None
        hidden_states = intermediate_tensors["hidden_states"]

    for i in range(self.start_layer, self.end_layer):
        layer = self.layers[i]
        layer_mamba_cache_params = None
        if mamba_cache_params:
            layer_mamba_cache_params = mamba_cache_params.at_layer_idx(i)
        hidden_states = layer(
            positions=positions,
            hidden_states=hidden_states,
            mamba_cache_params=layer_mamba_cache_params,
            mamba2_metadata=mamba2_metadata,
        )
    if not get_pp_group().is_last_rank:
        return IntermediateTensors({
            "hidden_states": hidden_states,
        })
    hidden_states = self.final_layernorm(hidden_states)
    return hidden_states

get_input_embeddings

get_input_embeddings(input_ids: Tensor) -> Tensor

Source code in vllm/model_executor/models/falcon_h1.py

def get_input_embeddings(self, input_ids: torch.Tensor) -> torch.Tensor:
    return self.embed_tokens(input_ids)

FalconH1ParallelHybrid

Bases: Module

A hybrid decoder layer for FalconH1 where the input is processed in parallel through both the self-attention branch and the SSM (Mamba) branch. Their outputs are then summed to produce the final hidden state.

This layer uses

• FalconH1AttentionDecoderLayer for the multi-head self-attention branch.
• FalconH1SSMDecoderLayer for the state-space (Mamba) branch.

Source code in vllm/model_executor/models/falcon_h1.py

class FalconH1ParallelHybrid(nn.Module):
    """
    A hybrid decoder layer for FalconH1 where the input is processed
    in parallel through both the self-attention branch and the SSM (Mamba)
    branch. Their outputs are then summed to produce the final hidden state.

    This layer uses:
      - FalconH1AttentionDecoderLayer for the multi-head self-attention branch.
      - FalconH1SSMDecoderLayer for the state-space (Mamba) branch.
    """

    def __init__(
        self,
        config: FalconH1Config,
        layer_idx: int,
        cache_config: Optional[CacheConfig] = None,
        quant_config: Optional[QuantizationConfig] = None,
        prefix: str = "",
    ) -> None:
        super().__init__()

        # Instantiate the attention branch
        self.self_attn = FalconH1AttentionDecoderLayer(
            config=config,
            cache_config=cache_config,
            quant_config=quant_config,
            prefix=prefix,
        )

        # In V1 all attention/ssm layers must have
        # different index in prefix
        ssm_layer_idx = config.num_hidden_layers + layer_idx
        ssm_prefix = prefix.split(".")[0] + f".{ssm_layer_idx}"

        # Instantiate the SSM branch
        self.mamba = FalconH1SSMDecoderLayer(
            config=config,
            cache_config=cache_config,
            quant_config=quant_config,
            prefix=ssm_prefix,
        )
        self.ssm_out_multiplier = config.ssm_out_multiplier
        self.ssm_in_multiplier = config.ssm_in_multiplier

        self.attention_in_multiplier = config.attention_in_multiplier
        self.attn_out_multiplier = config.attention_out_multiplier

        self.feed_forward = FalconH1MLP(config)

        self.input_layernorm = RMSNorm(config.hidden_size,
                                       eps=config.rms_norm_eps)
        self.pre_ff_layernorm = RMSNorm(config.hidden_size,
                                        eps=config.rms_norm_eps)

    def forward(
        self,
        positions: torch.Tensor,
        hidden_states: torch.Tensor,
        mamba_cache_params: MambaCacheParams,
        mamba2_metadata: Mamba2Metadata,
        **kwargs,
    ):
        residual = hidden_states
        hidden_states = self.input_layernorm(hidden_states)
        # Process input through the attention branch.
        # FalconH1AttentionDecoderLayer expects positions, hidden_states,
        # kv_cache, attn_metadata, and residual.
        attn_hidden, _ = self.self_attn(
            positions=positions,
            hidden_states=hidden_states * self.attention_in_multiplier,
            residual=residual,
            **kwargs,
        )

        # Process input through the SSM branch.
        # FalconH1SSMDecoderLayer expects hidden_states, attn_metadata,
        # residual, mamba_cache_params, and sequence_idx.
        ssm_hidden, _ = self.mamba(
            hidden_states=hidden_states * self.ssm_in_multiplier,
            residual=residual,
            mamba_cache_params=mamba_cache_params,
            mamba2_metadata=mamba2_metadata,
            **kwargs,
        )
        # Sum the outputs from both branches.
        # We assume both branches produce outputs of the same
        # dimensionality (config.hidden_size).
        hidden_states = (attn_hidden * self.attn_out_multiplier) + (
            ssm_hidden * self.ssm_out_multiplier)
        hidden_states = hidden_states + residual

        # feed-forward
        residual = hidden_states
        hidden_states = self.pre_ff_layernorm(hidden_states)
        hidden_states = self.feed_forward(hidden_states)
        hidden_states = residual + hidden_states

        return hidden_states

attention_in_multiplier instance-attribute

attention_in_multiplier = attention_in_multiplier

attn_out_multiplier instance-attribute

attn_out_multiplier = attention_out_multiplier

feed_forward instance-attribute

feed_forward = FalconH1MLP(config)

input_layernorm instance-attribute

input_layernorm = RMSNorm(hidden_size, eps=rms_norm_eps)

mamba instance-attribute

mamba = FalconH1SSMDecoderLayer(
    config=config,
    cache_config=cache_config,
    quant_config=quant_config,
    prefix=ssm_prefix,
)

pre_ff_layernorm instance-attribute

pre_ff_layernorm = RMSNorm(hidden_size, eps=rms_norm_eps)

self_attn instance-attribute

self_attn = FalconH1AttentionDecoderLayer(
    config=config,
    cache_config=cache_config,
    quant_config=quant_config,
    prefix=prefix,
)

ssm_in_multiplier instance-attribute

ssm_in_multiplier = ssm_in_multiplier

ssm_out_multiplier instance-attribute

ssm_out_multiplier = ssm_out_multiplier

__init__

__init__(
    config: FalconH1Config,
    layer_idx: int,
    cache_config: Optional[CacheConfig] = None,
    quant_config: Optional[QuantizationConfig] = None,
    prefix: str = "",
) -> None

Source code in vllm/model_executor/models/falcon_h1.py

def __init__(
    self,
    config: FalconH1Config,
    layer_idx: int,
    cache_config: Optional[CacheConfig] = None,
    quant_config: Optional[QuantizationConfig] = None,
    prefix: str = "",
) -> None:
    super().__init__()

    # Instantiate the attention branch
    self.self_attn = FalconH1AttentionDecoderLayer(
        config=config,
        cache_config=cache_config,
        quant_config=quant_config,
        prefix=prefix,
    )

    # In V1 all attention/ssm layers must have
    # different index in prefix
    ssm_layer_idx = config.num_hidden_layers + layer_idx
    ssm_prefix = prefix.split(".")[0] + f".{ssm_layer_idx}"

    # Instantiate the SSM branch
    self.mamba = FalconH1SSMDecoderLayer(
        config=config,
        cache_config=cache_config,
        quant_config=quant_config,
        prefix=ssm_prefix,
    )
    self.ssm_out_multiplier = config.ssm_out_multiplier
    self.ssm_in_multiplier = config.ssm_in_multiplier

    self.attention_in_multiplier = config.attention_in_multiplier
    self.attn_out_multiplier = config.attention_out_multiplier

    self.feed_forward = FalconH1MLP(config)

    self.input_layernorm = RMSNorm(config.hidden_size,
                                   eps=config.rms_norm_eps)
    self.pre_ff_layernorm = RMSNorm(config.hidden_size,
                                    eps=config.rms_norm_eps)

forward

forward(
    positions: Tensor,
    hidden_states: Tensor,
    mamba_cache_params: MambaCacheParams,
    mamba2_metadata: Mamba2Metadata,
    **kwargs,
)

Source code in vllm/model_executor/models/falcon_h1.py

def forward(
    self,
    positions: torch.Tensor,
    hidden_states: torch.Tensor,
    mamba_cache_params: MambaCacheParams,
    mamba2_metadata: Mamba2Metadata,
    **kwargs,
):
    residual = hidden_states
    hidden_states = self.input_layernorm(hidden_states)
    # Process input through the attention branch.
    # FalconH1AttentionDecoderLayer expects positions, hidden_states,
    # kv_cache, attn_metadata, and residual.
    attn_hidden, _ = self.self_attn(
        positions=positions,
        hidden_states=hidden_states * self.attention_in_multiplier,
        residual=residual,
        **kwargs,
    )

    # Process input through the SSM branch.
    # FalconH1SSMDecoderLayer expects hidden_states, attn_metadata,
    # residual, mamba_cache_params, and sequence_idx.
    ssm_hidden, _ = self.mamba(
        hidden_states=hidden_states * self.ssm_in_multiplier,
        residual=residual,
        mamba_cache_params=mamba_cache_params,
        mamba2_metadata=mamba2_metadata,
        **kwargs,
    )
    # Sum the outputs from both branches.
    # We assume both branches produce outputs of the same
    # dimensionality (config.hidden_size).
    hidden_states = (attn_hidden * self.attn_out_multiplier) + (
        ssm_hidden * self.ssm_out_multiplier)
    hidden_states = hidden_states + residual

    # feed-forward
    residual = hidden_states
    hidden_states = self.pre_ff_layernorm(hidden_states)
    hidden_states = self.feed_forward(hidden_states)
    hidden_states = residual + hidden_states

    return hidden_states

FalconH1SSMDecoderLayer

Bases: Module

Source code in vllm/model_executor/models/falcon_h1.py

class FalconH1SSMDecoderLayer(nn.Module):

    def __init__(
        self,
        config: FalconH1Config,
        cache_config: Optional[CacheConfig] = None,
        quant_config: Optional[QuantizationConfig] = None,
        prefix: str = "",
    ) -> None:
        super().__init__()
        self.config = config
        self.tp_size = get_tensor_model_parallel_world_size()

        self.d_ssm = (int(config.mamba_expand * config.hidden_size)
                      if config.mamba_d_ssm is None else config.mamba_d_ssm)

        self.mamba = MambaMixer2(
            hidden_size=config.hidden_size,
            ssm_state_size=config.mamba_d_state,
            conv_kernel_size=config.mamba_d_conv,
            intermediate_size=self.d_ssm,
            use_conv_bias=config.mamba_conv_bias,
            use_bias=config.mamba_proj_bias,
            n_groups=config.mamba_n_groups,
            num_heads=config.mamba_n_heads,
            head_dim=config.mamba_d_head,
            rms_norm_eps=config.rms_norm_eps,
            activation=config.hidden_act,
            quant_config=quant_config,
            use_rms_norm=config.mamba_rms_norm,
            prefix=f"{prefix}.mixer",
            chunk_size=config.mamba_chunk_size,
        )
        # n_groups is overridden later by `MambaMixer2`
        self.groups_time_state_size = self.mamba.n_groups * config.mamba_d_state
        self.zxbcdt_multipliers = config.ssm_multipliers
        self._init_mup_vector()

    def _init_mup_vector(self):
        """
        Non learnable per-block scaling vector composed of element-wise 
        multipliersapplied to each separate contiguous block of the output 
        of the linear projection (in_proj) before further processing
        (gating, convolution, SSM):

            - Z block:  [0 : d_ssm]                      → zxbcdt_multipliers[0]
            - X block:  [d_ssm : 2 * d_ssm]              → zxbcdt_multipliers[1]
            - B block:  [2 * d_ssm : 2 * d_ssm + G * S]  → zxbcdt_multipliers[2]
            - C block:  [2 * d_ssm + G * S : 2 * d_ssm + 2 * G * S] 
                        → zxbcdt_multipliers[3]
            - dt block: [2 * d_ssm + 2 * G * S : end]    → zxbcdt_multipliers[4]

        where:
            - d_ssm:     Dimension of state-space model latent
            - G:         Number of groups (n_groups)
            - S:         SSM state size per group
            - All indices are divided by tp_size to support tensor parallelism
        """
        vector_shape = (2 * self.d_ssm + 2 * self.groups_time_state_size +
                        self.config.mamba_n_heads) // self.tp_size
        mup_vector = torch.ones(1, vector_shape)
        # Z vector 0 -> d_ssm
        mup_vector[:, :self.d_ssm //
                   self.tp_size] *= self.zxbcdt_multipliers[0]
        # X vector d_ssm -> 2 * d_ssm
        mup_vector[:,
                   (self.d_ssm //
                    self.tp_size):(2 * self.d_ssm //
                                   self.tp_size)] *= self.zxbcdt_multipliers[1]
        # B vector 2 * d_ssm -> 2 * d_ssm + (n_group * d_state)
        mup_vector[
            :,
            (2 * self.d_ssm) //
            self.tp_size:(2 * self.d_ssm + self.groups_time_state_size) //
            self.tp_size,
        ] *= self.zxbcdt_multipliers[2]
        # C vector 2 * d_ssm + (n_group * d_state)
        # -> 2 * d_ssm + 2 * (n_group * d_state)
        mup_vector[
            :,
            (2 * self.d_ssm + self.groups_time_state_size) //
            self.tp_size:(2 * self.d_ssm + 2 * self.groups_time_state_size) //
            self.tp_size,
        ] *= self.zxbcdt_multipliers[3]
        # dt vector 2 * d_ssm + 2 * (n_group * d_state)
        # -> 2 * d_ssm + 2 * (n_group * d_state) + n_heads
        mup_vector[
            :,
            (2 * self.d_ssm + 2 * self.groups_time_state_size) //
            self.tp_size:,
        ] *= self.zxbcdt_multipliers[4]

        self.register_buffer("mup_vector", mup_vector, persistent=False)

    def forward(
        self,
        hidden_states: torch.Tensor,
        residual: Optional[torch.Tensor],
        mamba_cache_params: MambaCacheParams,
        mamba2_metadata: Mamba2Metadata,
        **kwargs,
    ):
        hidden_states = self.mamba(
            hidden_states,
            mamba_cache_params,
            mamba2_metadata=mamba2_metadata,
            mup_vector=self.mup_vector,
        )
        return hidden_states, residual

config instance-attribute

config = config

d_ssm instance-attribute

d_ssm = (
    int(mamba_expand * hidden_size)
    if mamba_d_ssm is None
    else mamba_d_ssm
)

groups_time_state_size instance-attribute

groups_time_state_size = n_groups * mamba_d_state

mamba instance-attribute

mamba = MambaMixer2(
    hidden_size=hidden_size,
    ssm_state_size=mamba_d_state,
    conv_kernel_size=mamba_d_conv,
    intermediate_size=d_ssm,
    use_conv_bias=mamba_conv_bias,
    use_bias=mamba_proj_bias,
    n_groups=mamba_n_groups,
    num_heads=mamba_n_heads,
    head_dim=mamba_d_head,
    rms_norm_eps=rms_norm_eps,
    activation=hidden_act,
    quant_config=quant_config,
    use_rms_norm=mamba_rms_norm,
    prefix=f"{prefix}.mixer",
    chunk_size=mamba_chunk_size,
)

tp_size instance-attribute

tp_size = get_tensor_model_parallel_world_size()

zxbcdt_multipliers instance-attribute

zxbcdt_multipliers = ssm_multipliers

__init__

__init__(
    config: FalconH1Config,
    cache_config: Optional[CacheConfig] = None,
    quant_config: Optional[QuantizationConfig] = None,
    prefix: str = "",
) -> None

Source code in vllm/model_executor/models/falcon_h1.py

def __init__(
    self,
    config: FalconH1Config,
    cache_config: Optional[CacheConfig] = None,
    quant_config: Optional[QuantizationConfig] = None,
    prefix: str = "",
) -> None:
    super().__init__()
    self.config = config
    self.tp_size = get_tensor_model_parallel_world_size()

    self.d_ssm = (int(config.mamba_expand * config.hidden_size)
                  if config.mamba_d_ssm is None else config.mamba_d_ssm)

    self.mamba = MambaMixer2(
        hidden_size=config.hidden_size,
        ssm_state_size=config.mamba_d_state,
        conv_kernel_size=config.mamba_d_conv,
        intermediate_size=self.d_ssm,
        use_conv_bias=config.mamba_conv_bias,
        use_bias=config.mamba_proj_bias,
        n_groups=config.mamba_n_groups,
        num_heads=config.mamba_n_heads,
        head_dim=config.mamba_d_head,
        rms_norm_eps=config.rms_norm_eps,
        activation=config.hidden_act,
        quant_config=quant_config,
        use_rms_norm=config.mamba_rms_norm,
        prefix=f"{prefix}.mixer",
        chunk_size=config.mamba_chunk_size,
    )
    # n_groups is overridden later by `MambaMixer2`
    self.groups_time_state_size = self.mamba.n_groups * config.mamba_d_state
    self.zxbcdt_multipliers = config.ssm_multipliers
    self._init_mup_vector()

_init_mup_vector

_init_mup_vector()

Non learnable per-block scaling vector composed of element-wise multipliersapplied to each separate contiguous block of the output of the linear projection (in_proj) before further processing (gating, convolution, SSM):

- Z block:  [0 : d_ssm]                      → zxbcdt_multipliers[0]
- X block:  [d_ssm : 2 * d_ssm]              → zxbcdt_multipliers[1]
- B block:  [2 * d_ssm : 2 * d_ssm + G * S]  → zxbcdt_multipliers[2]
- C block:  [2 * d_ssm + G * S : 2 * d_ssm + 2 * G * S] 
            → zxbcdt_multipliers[3]
- dt block: [2 * d_ssm + 2 * G * S : end]    → zxbcdt_multipliers[4]

where

• d_ssm: Dimension of state-space model latent
• G: Number of groups (n_groups)
• S: SSM state size per group
• All indices are divided by tp_size to support tensor parallelism

Source code in vllm/model_executor/models/falcon_h1.py

def _init_mup_vector(self):
    """
    Non learnable per-block scaling vector composed of element-wise 
    multipliersapplied to each separate contiguous block of the output 
    of the linear projection (in_proj) before further processing
    (gating, convolution, SSM):

        - Z block:  [0 : d_ssm]                      → zxbcdt_multipliers[0]
        - X block:  [d_ssm : 2 * d_ssm]              → zxbcdt_multipliers[1]
        - B block:  [2 * d_ssm : 2 * d_ssm + G * S]  → zxbcdt_multipliers[2]
        - C block:  [2 * d_ssm + G * S : 2 * d_ssm + 2 * G * S] 
                    → zxbcdt_multipliers[3]
        - dt block: [2 * d_ssm + 2 * G * S : end]    → zxbcdt_multipliers[4]

    where:
        - d_ssm:     Dimension of state-space model latent
        - G:         Number of groups (n_groups)
        - S:         SSM state size per group
        - All indices are divided by tp_size to support tensor parallelism
    """
    vector_shape = (2 * self.d_ssm + 2 * self.groups_time_state_size +
                    self.config.mamba_n_heads) // self.tp_size
    mup_vector = torch.ones(1, vector_shape)
    # Z vector 0 -> d_ssm
    mup_vector[:, :self.d_ssm //
               self.tp_size] *= self.zxbcdt_multipliers[0]
    # X vector d_ssm -> 2 * d_ssm
    mup_vector[:,
               (self.d_ssm //
                self.tp_size):(2 * self.d_ssm //
                               self.tp_size)] *= self.zxbcdt_multipliers[1]
    # B vector 2 * d_ssm -> 2 * d_ssm + (n_group * d_state)
    mup_vector[
        :,
        (2 * self.d_ssm) //
        self.tp_size:(2 * self.d_ssm + self.groups_time_state_size) //
        self.tp_size,
    ] *= self.zxbcdt_multipliers[2]
    # C vector 2 * d_ssm + (n_group * d_state)
    # -> 2 * d_ssm + 2 * (n_group * d_state)
    mup_vector[
        :,
        (2 * self.d_ssm + self.groups_time_state_size) //
        self.tp_size:(2 * self.d_ssm + 2 * self.groups_time_state_size) //
        self.tp_size,
    ] *= self.zxbcdt_multipliers[3]
    # dt vector 2 * d_ssm + 2 * (n_group * d_state)
    # -> 2 * d_ssm + 2 * (n_group * d_state) + n_heads
    mup_vector[
        :,
        (2 * self.d_ssm + 2 * self.groups_time_state_size) //
        self.tp_size:,
    ] *= self.zxbcdt_multipliers[4]

    self.register_buffer("mup_vector", mup_vector, persistent=False)

forward

forward(
    hidden_states: Tensor,
    residual: Optional[Tensor],
    mamba_cache_params: MambaCacheParams,
    mamba2_metadata: Mamba2Metadata,
    **kwargs,
)

Source code in vllm/model_executor/models/falcon_h1.py

def forward(
    self,
    hidden_states: torch.Tensor,
    residual: Optional[torch.Tensor],
    mamba_cache_params: MambaCacheParams,
    mamba2_metadata: Mamba2Metadata,
    **kwargs,
):
    hidden_states = self.mamba(
        hidden_states,
        mamba_cache_params,
        mamba2_metadata=mamba2_metadata,
        mup_vector=self.mup_vector,
    )
    return hidden_states, residual