[6/N] pass whole config to inner model (#10205)

Signed-off-by: youkaichao <youkaichao@gmail.com>

vllm/model_executor/models/deepseek_v2.py

@@ -51,7 +51,8 @@ from vllm.sequence import IntermediateTensors
 
 from .interfaces import SupportsPP
 from .utils import (PPMissingLayer, is_pp_missing_parameter,
-                    make_empty_intermediate_tensors_factory, make_layers)
+                    make_empty_intermediate_tensors_factory, make_layers,
+                    maybe_prefix)
 
 
 class DeepseekV2MLP(nn.Module):
@@ -408,14 +409,13 @@ class DeepseekV2Model(nn.Module):
 
     fall_back_to_pt_during_load = False
 
-    def __init__(
-        self,
-        config: PretrainedConfig,
-        cache_config: Optional[CacheConfig] = None,
-        quant_config: Optional[QuantizationConfig] = None,
-        prefix: str = "",
-    ) -> None:
+    def __init__(self, *, vllm_config: VllmConfig, prefix: str = ""):
         super().__init__()
+
+        config = vllm_config.model_config.hf_config
+        cache_config = vllm_config.cache_config
+        quant_config = vllm_config.quant_config
+
         self.padding_idx = config.pad_token_id
         self.vocab_size = config.vocab_size
 
@@ -479,21 +479,14 @@ class DeepseekV2Model(nn.Module):
 
 class DeepseekV2ForCausalLM(nn.Module, SupportsPP):
 
-    def __init__(
-        self,
-        vllm_config: VllmConfig,
-        prefix: str = "",
-    ) -> None:
+    def __init__(self, *, vllm_config: VllmConfig, prefix: str = ""):
         super().__init__()
         config = vllm_config.model_config.hf_config
-        cache_config = vllm_config.cache_config
         quant_config = vllm_config.quant_config
         self.config = config
         self.quant_config = quant_config
-        self.model = DeepseekV2Model(config,
-                                     cache_config,
-                                     quant_config,
-                                     prefix="model")
+        self.model = DeepseekV2Model(vllm_config=vllm_config,
+                                     prefix=maybe_prefix(prefix, "model"))
         self.lm_head = ParallelLMHead(config.vocab_size,
                                       config.hidden_size,
                                       quant_config=quant_config)