diff --git a/vllm/model_executor/models/telechat2.py b/vllm/model_executor/models/telechat2.py index 4e4e371840aab..caf02b06e2836 100644 --- a/vllm/model_executor/models/telechat2.py +++ b/vllm/model_executor/models/telechat2.py @@ -32,8 +32,8 @@ from vllm.model_executor.layers.sampler import Sampler, SamplerOutput from vllm.model_executor.layers.vocab_parallel_embedding import ParallelLMHead from vllm.model_executor.model_loader.weight_utils import default_weight_loader -from vllm.model_executor.models.llama import (LlamaAttention, - LlamaDecoderLayer, +from vllm.model_executor.models.llama import (LlamaAttention, + LlamaDecoderLayer, LlamaForCausalLM, LlamaMLP, LlamaModel) from vllm.model_executor.sampling_metadata import SamplingMetadata