Skip to content

vllm.transformers_utils.configs.deepseek_v3

logger module-attribute

logger = get_logger(__name__)

DeepseekV3Config

Bases: PretrainedConfig

Source code in vllm/transformers_utils/configs/deepseek_v3.py
class DeepseekV3Config(PretrainedConfig):

    model_type = "deepseek_v3"
    keys_to_ignore_at_inference = ["past_key_values"]

    def __init__(
        self,
        vocab_size=129280,
        hidden_size=7168,
        intermediate_size=18432,
        moe_intermediate_size=2048,
        num_hidden_layers=61,
        num_nextn_predict_layers=1,
        num_attention_heads=128,
        num_key_value_heads=128,
        n_shared_experts=1,
        n_routed_experts=256,
        ep_size=1,
        routed_scaling_factor=2.5,
        kv_lora_rank=512,
        q_lora_rank=1536,
        qk_rope_head_dim=64,
        v_head_dim=128,
        qk_nope_head_dim=128,
        topk_method='noaux_tc',
        n_group=8,
        topk_group=4,
        num_experts_per_tok=8,
        moe_layer_freq=1,
        first_k_dense_replace=3,
        norm_topk_prob=True,
        scoring_func='sigmoid',
        hidden_act="silu",
        max_position_embeddings=4096,
        initializer_range=0.02,
        rms_norm_eps=1e-6,
        use_cache=True,
        pad_token_id=None,
        bos_token_id=0,
        eos_token_id=1,
        tie_word_embeddings=False,
        rope_theta=10000.0,
        rope_scaling=None,
        attention_bias=False,
        attention_dropout=0.0,
        **kwargs,
    ):
        self.vocab_size = vocab_size
        self.max_position_embeddings = max_position_embeddings
        self.hidden_size = hidden_size
        self.intermediate_size = intermediate_size
        self.moe_intermediate_size = moe_intermediate_size
        self.num_hidden_layers = num_hidden_layers
        self.num_nextn_predict_layers = num_nextn_predict_layers
        self.num_attention_heads = num_attention_heads
        self.n_shared_experts = n_shared_experts
        self.n_routed_experts = n_routed_experts
        self.ep_size = ep_size
        self.routed_scaling_factor = routed_scaling_factor
        self.kv_lora_rank = kv_lora_rank
        self.q_lora_rank = q_lora_rank
        self.qk_rope_head_dim = qk_rope_head_dim
        self.v_head_dim = v_head_dim
        self.qk_nope_head_dim = qk_nope_head_dim
        self.topk_method = topk_method
        self.n_group = n_group
        self.topk_group = topk_group
        self.num_experts_per_tok = num_experts_per_tok
        self.moe_layer_freq = moe_layer_freq
        self.first_k_dense_replace = first_k_dense_replace
        self.norm_topk_prob = norm_topk_prob
        self.scoring_func = scoring_func
        # for backward compatibility
        if num_key_value_heads is None:
            num_key_value_heads = num_attention_heads

        self.num_key_value_heads = num_key_value_heads
        self.hidden_act = hidden_act
        self.initializer_range = initializer_range
        self.rms_norm_eps = rms_norm_eps
        self.use_cache = use_cache
        self.rope_theta = rope_theta
        self.rope_scaling = rope_scaling
        self.attention_bias = attention_bias
        self.attention_dropout = attention_dropout

        super().__init__(
            pad_token_id=pad_token_id,
            bos_token_id=bos_token_id,
            eos_token_id=eos_token_id,
            tie_word_embeddings=tie_word_embeddings,
            **kwargs,
        )

attention_bias instance-attribute

attention_bias = attention_bias

attention_dropout instance-attribute

attention_dropout = attention_dropout

ep_size instance-attribute

ep_size = ep_size

first_k_dense_replace instance-attribute

first_k_dense_replace = first_k_dense_replace

hidden_act instance-attribute

hidden_act = hidden_act

hidden_size instance-attribute

hidden_size = hidden_size

initializer_range instance-attribute

initializer_range = initializer_range

intermediate_size instance-attribute

intermediate_size = intermediate_size

keys_to_ignore_at_inference class-attribute instance-attribute

keys_to_ignore_at_inference = ['past_key_values']

kv_lora_rank instance-attribute

kv_lora_rank = kv_lora_rank

max_position_embeddings instance-attribute

max_position_embeddings = max_position_embeddings

model_type class-attribute instance-attribute

model_type = 'deepseek_v3'

moe_intermediate_size instance-attribute

moe_intermediate_size = moe_intermediate_size

moe_layer_freq instance-attribute

moe_layer_freq = moe_layer_freq

n_group instance-attribute

n_group = n_group

n_routed_experts instance-attribute

n_routed_experts = n_routed_experts

n_shared_experts instance-attribute

n_shared_experts = n_shared_experts

norm_topk_prob instance-attribute

norm_topk_prob = norm_topk_prob

num_attention_heads instance-attribute

num_attention_heads = num_attention_heads

num_experts_per_tok instance-attribute

num_experts_per_tok = num_experts_per_tok

num_hidden_layers instance-attribute

num_hidden_layers = num_hidden_layers

num_key_value_heads instance-attribute

num_key_value_heads = num_key_value_heads

num_nextn_predict_layers instance-attribute

num_nextn_predict_layers = num_nextn_predict_layers

q_lora_rank instance-attribute

q_lora_rank = q_lora_rank

qk_nope_head_dim instance-attribute

qk_nope_head_dim = qk_nope_head_dim

qk_rope_head_dim instance-attribute

qk_rope_head_dim = qk_rope_head_dim

rms_norm_eps instance-attribute

rms_norm_eps = rms_norm_eps

rope_scaling instance-attribute

rope_scaling = rope_scaling

rope_theta instance-attribute

rope_theta = rope_theta

routed_scaling_factor instance-attribute

routed_scaling_factor = routed_scaling_factor

scoring_func instance-attribute

scoring_func = scoring_func

topk_group instance-attribute

topk_group = topk_group

topk_method instance-attribute

topk_method = topk_method

use_cache instance-attribute

use_cache = use_cache

v_head_dim instance-attribute

v_head_dim = v_head_dim

vocab_size instance-attribute

vocab_size = vocab_size

__init__

__init__(
    vocab_size=129280,
    hidden_size=7168,
    intermediate_size=18432,
    moe_intermediate_size=2048,
    num_hidden_layers=61,
    num_nextn_predict_layers=1,
    num_attention_heads=128,
    num_key_value_heads=128,
    n_shared_experts=1,
    n_routed_experts=256,
    ep_size=1,
    routed_scaling_factor=2.5,
    kv_lora_rank=512,
    q_lora_rank=1536,
    qk_rope_head_dim=64,
    v_head_dim=128,
    qk_nope_head_dim=128,
    topk_method="noaux_tc",
    n_group=8,
    topk_group=4,
    num_experts_per_tok=8,
    moe_layer_freq=1,
    first_k_dense_replace=3,
    norm_topk_prob=True,
    scoring_func="sigmoid",
    hidden_act="silu",
    max_position_embeddings=4096,
    initializer_range=0.02,
    rms_norm_eps=1e-06,
    use_cache=True,
    pad_token_id=None,
    bos_token_id=0,
    eos_token_id=1,
    tie_word_embeddings=False,
    rope_theta=10000.0,
    rope_scaling=None,
    attention_bias=False,
    attention_dropout=0.0,
    **kwargs,
)
Source code in vllm/transformers_utils/configs/deepseek_v3.py
def __init__(
    self,
    vocab_size=129280,
    hidden_size=7168,
    intermediate_size=18432,
    moe_intermediate_size=2048,
    num_hidden_layers=61,
    num_nextn_predict_layers=1,
    num_attention_heads=128,
    num_key_value_heads=128,
    n_shared_experts=1,
    n_routed_experts=256,
    ep_size=1,
    routed_scaling_factor=2.5,
    kv_lora_rank=512,
    q_lora_rank=1536,
    qk_rope_head_dim=64,
    v_head_dim=128,
    qk_nope_head_dim=128,
    topk_method='noaux_tc',
    n_group=8,
    topk_group=4,
    num_experts_per_tok=8,
    moe_layer_freq=1,
    first_k_dense_replace=3,
    norm_topk_prob=True,
    scoring_func='sigmoid',
    hidden_act="silu",
    max_position_embeddings=4096,
    initializer_range=0.02,
    rms_norm_eps=1e-6,
    use_cache=True,
    pad_token_id=None,
    bos_token_id=0,
    eos_token_id=1,
    tie_word_embeddings=False,
    rope_theta=10000.0,
    rope_scaling=None,
    attention_bias=False,
    attention_dropout=0.0,
    **kwargs,
):
    self.vocab_size = vocab_size
    self.max_position_embeddings = max_position_embeddings
    self.hidden_size = hidden_size
    self.intermediate_size = intermediate_size
    self.moe_intermediate_size = moe_intermediate_size
    self.num_hidden_layers = num_hidden_layers
    self.num_nextn_predict_layers = num_nextn_predict_layers
    self.num_attention_heads = num_attention_heads
    self.n_shared_experts = n_shared_experts
    self.n_routed_experts = n_routed_experts
    self.ep_size = ep_size
    self.routed_scaling_factor = routed_scaling_factor
    self.kv_lora_rank = kv_lora_rank
    self.q_lora_rank = q_lora_rank
    self.qk_rope_head_dim = qk_rope_head_dim
    self.v_head_dim = v_head_dim
    self.qk_nope_head_dim = qk_nope_head_dim
    self.topk_method = topk_method
    self.n_group = n_group
    self.topk_group = topk_group
    self.num_experts_per_tok = num_experts_per_tok
    self.moe_layer_freq = moe_layer_freq
    self.first_k_dense_replace = first_k_dense_replace
    self.norm_topk_prob = norm_topk_prob
    self.scoring_func = scoring_func
    # for backward compatibility
    if num_key_value_heads is None:
        num_key_value_heads = num_attention_heads

    self.num_key_value_heads = num_key_value_heads
    self.hidden_act = hidden_act
    self.initializer_range = initializer_range
    self.rms_norm_eps = rms_norm_eps
    self.use_cache = use_cache
    self.rope_theta = rope_theta
    self.rope_scaling = rope_scaling
    self.attention_bias = attention_bias
    self.attention_dropout = attention_dropout

    super().__init__(
        pad_token_id=pad_token_id,
        bos_token_id=bos_token_id,
        eos_token_id=eos_token_id,
        tie_word_embeddings=tie_word_embeddings,
        **kwargs,
    )