support llava video (#426)

python/sglang/srt/models/qwen2.py

@@ -5,6 +5,7 @@ from typing import Any, Dict, Optional, Tuple
 
 import torch
 from torch import nn
+from vllm.distributed import get_tensor_model_parallel_world_size
 from vllm.model_executor.layers.activation import SiluAndMul
 from vllm.model_executor.layers.layernorm import RMSNorm
 from vllm.model_executor.layers.linear import (
@@ -12,24 +13,17 @@ from vllm.model_executor.layers.linear import (
     QKVParallelLinear,
     RowParallelLinear,
 )
-from vllm.model_executor.layers.quantization.base_config import (
-    QuantizationConfig)
+from vllm.model_executor.layers.quantization.base_config import QuantizationConfig
 from vllm.model_executor.layers.rotary_embedding import get_rope
 from vllm.model_executor.layers.vocab_parallel_embedding import (
     ParallelLMHead,
     VocabParallelEmbedding,
 )
-from vllm.distributed import (
-    get_tensor_model_parallel_world_size,
-)
-from sglang.srt.weight_utils import (
-    default_weight_loader,
-    hf_model_weights_iterator,
-)
 
 from sglang.srt.layers.logits_processor import LogitsProcessor
 from sglang.srt.layers.radix_attention import RadixAttention
 from sglang.srt.managers.router.model_runner import InputMetadata
+from sglang.srt.weight_utils import default_weight_loader, hf_model_weights_iterator
 
 Qwen2Config = None
 
@@ -50,7 +44,10 @@ class Qwen2MLP(nn.Module):
             quant_config=quant_config,
         )
         self.down_proj = RowParallelLinear(
-            intermediate_size, hidden_size, bias=False, quant_config=quant_config,
+            intermediate_size,
+            hidden_size,
+            bias=False,
+            quant_config=quant_config,
         )
         if hidden_act != "silu":
             raise ValueError(