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

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

vllm/model_executor/models/bloom.py

@@ -42,7 +42,8 @@ from vllm.sequence import IntermediateTensors
 
 from .interfaces import SupportsPP
 from .utils import (is_pp_missing_parameter,
-                    make_empty_intermediate_tensors_factory, make_layers)
+                    make_empty_intermediate_tensors_factory, make_layers,
+                    maybe_prefix)
 
 
 def _get_alibi_slopes(total_num_heads: int) -> torch.Tensor:
@@ -221,14 +222,13 @@ class BloomBlock(nn.Module):
 @support_torch_compile
 class BloomModel(nn.Module):
 
-    def __init__(
-        self,
-        config: BloomConfig,
-        cache_config: Optional[CacheConfig] = None,
-        quant_config: Optional[QuantizationConfig] = None,
-        prefix: str = "",
-    ):
+    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.embed_dim = config.hidden_size
 
         # Embedding + LN Embedding
@@ -288,11 +288,12 @@ class BloomForCausalLM(nn.Module, SupportsPP):
     ):
         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.transformer = BloomModel(config, cache_config, quant_config)
+        self.transformer = BloomModel(vllm_config=vllm_config,
+                                      prefix=maybe_prefix(
+                                          prefix, "transformer"))
         if self.config.tie_word_embeddings:
             self.lm_head = self.transformer.word_embeddings
         else: