| from transformers.configuration_utils import PretrainedConfig
|
| from transformers.utils import logging
|
|
|
| logger = logging.get_logger(__name__)
|
|
|
| DEEPSEEK_PRETRAINED_CONFIG_ARCHIVE_MAP = {}
|
|
|
|
|
| class Xing4_0Config(PretrainedConfig):
|
| model_type = "xing4_0"
|
| keys_to_ignore_at_inference = ["past_key_values"]
|
| base_model_tp_plan = {
|
| "layers.*.mlp.experts.gate_up_proj": "packed_colwise",
|
| "layers.*.mlp.experts.down_proj": "rowwise",
|
| "layers.*.mlp.experts": "moe_tp_experts",
|
| "layers.*.mlp.shared_experts.gate_proj": "colwise",
|
| "layers.*.mlp.shared_experts.up_proj": "colwise",
|
| "layers.*.mlp.shared_experts.down_proj": "rowwise",
|
| "layers.*.mlp.gate_proj": "colwise",
|
| "layers.*.mlp.up_proj": "colwise",
|
| "layers.*.mlp.down_proj": "rowwise",
|
| }
|
| base_model_pp_plan = {
|
| "embed_tokens": (["input_ids"], ["inputs_embeds"]),
|
| "layers": (["hidden_states", "attention_mask"], ["hidden_states"]),
|
| "norm": (["hidden_states"], ["hidden_states"]),
|
| }
|
| base_model_ep_plan = {
|
| "layers.*.mlp.gate": "ep_router",
|
| "layers.*.mlp.experts.gate_up_proj": "grouped_gemm",
|
| "layers.*.mlp.experts.down_proj": "grouped_gemm",
|
| "layers.*.mlp.experts": "moe_tp_experts",
|
| }
|
|
|
| attribute_map = {
|
| "num_local_experts": "n_routed_experts",
|
| "num_mtp_layers": "num_nextn_predict_layers",
|
| }
|
|
|
| def __init__(
|
| self,
|
| vocab_size=131072,
|
| hidden_size=3584,
|
| intermediate_size=9216,
|
| moe_intermediate_size=1024,
|
| num_hidden_layers=40,
|
| num_nextn_predict_layers=1,
|
| num_attention_heads=32,
|
| num_key_value_heads=32,
|
| n_shared_experts=1,
|
| n_routed_experts=64,
|
| ep_size=1,
|
| routed_scaling_factor=2.0,
|
| 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=4,
|
| moe_layer_freq=1,
|
| first_k_dense_replace=2,
|
| 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=1,
|
| eos_token_id=2,
|
| tie_word_embeddings=False,
|
| rope_theta=10000.0,
|
| rope_scaling=None,
|
| rope_interleave=True,
|
| attention_bias=False,
|
| attention_dropout=0.0,
|
| hc_mult: int = 4,
|
| hc_sinkhorn_iters: int = 20,
|
| hc_eps: float = 1.0e-6,
|
| mhc_h_res_clamp_min=-30,
|
| mhc_h_res_clamp_max=30,
|
| **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.qk_head_dim = self.qk_nope_head_dim + self.qk_rope_head_dim
|
| self.head_dim = self.qk_rope_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
|
|
|
| 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
|
| self.hc_mult = hc_mult
|
| self.hc_sinkhorn_iters = hc_sinkhorn_iters
|
| self.hc_eps = hc_eps
|
| self.mhc_h_res_clamp_min = mhc_h_res_clamp_min
|
| self.mhc_h_res_clamp_max = mhc_h_res_clamp_max
|
| self.rope_interleave = rope_interleave
|
| 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,
|
| )
|
|
|
|
|
|
|
