跑train_adalora_seq2seq模型微调时出现错误AttributeError: 'Linear' object has no attribute 'in_channels'

[liuyifan123123]

Describe the bug/ 问题描述 (Mandatory / 必填)

A clear and concise description of what the bug is.
报错：AttributeError: 'Linear' object has no attribute 'in_channels'

• Hardware Environment(Ascend/GPU/CPU) / 硬件环境:
• Ascend

/device ascend/GPU/CPU/kirin/等其他芯片
device ascend

• Software Environment / 软件环境 (Mandatory / 必填):
• mindnlp0.4
  -- MindSpore version (e.g., 1.7.0.Bxxx) :
  mindspore2.3.1
  -- Python version (e.g., Python 3.7.5) :
  3.9.10
  -- OS platform and distribution (e.g., Linux Ubuntu 16.04):
  EluerOS
  -- GCC/Compiler version (if compiled from source):

To Reproduce / 重现步骤 (Mandatory / 必填)
简单示例如下，调用的facebook/bart-base模型
import os
os.environ['HF_ENDPOINT'] = 'https://hf-mirror.com'
os.environ["KMP_DUPLICATE_LIB_OK"]="TRUE"
from mindnlp.transformers import AutoModelForSeq2SeqLM
model_name_or_path = "facebook/bart-base"
from mindnlp.peft import AdaLoraConfig, TaskType, get_peft_model
model = AutoModelForSeq2SeqLM.from_pretrained(model_name_or_path)
peft_config = AdaLoraConfig(
init_r=12,
target_r=8,
beta1=0.85,
beta2=0.85,
tinit=200,
tfinal=1000,
deltaT=10,
lora_alpha=32,
lora_dropout=0.1,
task_type=TaskType.SEQ_2_SEQ_LM,
inference_mode=False,
)
model = get_peft_model(model, peft_config)