Pass inference model shard configs for module init

Signed-off-by: char-1ee <xingjianli59@gmail.com>

colossalai/inference/modeling/policy/nopadding_llama.py

@@ -72,6 +72,9 @@ class NoPaddingLlamaModelInferPolicy(LlamaForCausalLMPolicy, RPC_PARAM):
                 SubModuleReplacementDescription(
                     suffix="self_attn",
                     target_module=NopadLlamaAttention,
+                    kwargs={
+                        "model_shard_infer_config": self.shard_config.extra_kwargs["model_shard_infer_config"],
+                    },
                 ),
             ],
         )