[CI] Try introducing isort. (#3495)

vllm/model_executor/models/starcoder2.py

@@ -25,19 +25,19 @@ from torch import nn
 from transformers import Starcoder2Config
 
 from vllm.attention import Attention, AttentionMetadata
-from vllm.model_executor.sampling_metadata import SamplingMetadata
 from vllm.model_executor.layers.activation import get_act_fn
-from vllm.model_executor.layers.rotary_embedding import get_rope
 from vllm.model_executor.layers.linear import (ColumnParallelLinear,
                                                LinearMethodBase,
                                                QKVParallelLinear,
                                                RowParallelLinear)
 from vllm.model_executor.layers.logits_processor import LogitsProcessor
+from vllm.model_executor.layers.rotary_embedding import get_rope
 from vllm.model_executor.layers.sampler import Sampler
 from vllm.model_executor.layers.vocab_parallel_embedding import (
-    VocabParallelEmbedding, ParallelLMHead, DEFAULT_VOCAB_PADDING_SIZE)
+    DEFAULT_VOCAB_PADDING_SIZE, ParallelLMHead, VocabParallelEmbedding)
 from vllm.model_executor.parallel_utils.parallel_state import (
     get_tensor_model_parallel_world_size)
+from vllm.model_executor.sampling_metadata import SamplingMetadata
 from vllm.model_executor.weight_utils import (default_weight_loader,
                                               hf_model_weights_iterator)
 from vllm.sequence import SamplerOutput