请问paddleslim可以对kie推理模型进行裁剪吗？

[hzlirz]

kie的两个模型加起来有2G，可以使用paddleslim轻量化吗？如何解决下述报错：

我参考：地址

然后报错：
(paddleocr) root@DESKTOP-VSUROIT:/home/huang/paddle-ocr-gpu# python core_code/paddle_slim_kie.py /root/miniconda3/envs/paddleocr/lib/python3.9/site-packages/_distutils_hack/__init__.py:33: UserWarning: Setuptools is replacing distutils. warnings.warn("Setuptools is replacing distutils.") I1108 16:21:45.052196 580245 interpretercore.cc:237] New Executor is Running. W1108 16:21:45.303526 580245 gpu_resources.cc:119] Please NOTE: device: 0, GPU Compute Capability: 7.5, Driver API Version: 12.2, Runtime API Version: 11.8 W1108 16:21:45.304080 580245 gpu_resources.cc:149] device: 0, cuDNN Version: 8.9. 2023-11-08 16:21:51,999-INFO: devices: gpu 2023-11-08 16:21:58,532-INFO: Selected strategies: [] Traceback (most recent call last): File "/home/huang/paddle-ocr-gpu/core_code/paddle_slim_kie.py", line 54, in <module> ac.compress() File "/root/miniconda3/envs/paddleocr/lib/python3.9/site-packages/paddleslim-0.0.0.dev0-py3.9.egg/paddleslim/auto_compression/compressor.py", line 577, in compress AssertionError

我的list如下

我的代码如下：
import paddle
from PIL import Image
from paddle.vision.datasets import DatasetFolder
from paddle.vision.transforms import transforms
from paddleslim.auto_compression import AutoCompression

paddle.enable_static()

class ImageNetDataset(DatasetFolder):
def init(self, path, image_size=224):
super().init(path)
normalize = transforms.Normalize(
mean=[123.675, 116.28, 103.53], std=[58.395, 57.120, 57.375])
self.transform = transforms.Compose([
transforms.Resize(256),
transforms.CenterCrop(image_size), transforms.Transpose(),
normalize
])

def __getitem__(self, idx):
    img_path, _ = self.samples[idx]
    return self.transform(Image.open(img_path).convert('RGB'))

def __len__(self):
    return len(self.samples)

train_dataset = ImageNetDataset("./train_data/XFUND/zh_train/")
image = paddle.static.data(
name='inputs', shape=[None] + [3, 224, 224], dtype='float32')
train_loader = paddle.io.DataLoader(train_dataset, feed_list=[image], batch_size=32, return_list=False)

ac = AutoCompression(
model_dir="./inference/ser_vi_layoutxlm_xfund_infer",
model_filename="inference.pdmodel",
params_filename="inference.pdiparams",
save_dir="./slim_xfund_infer",
config={'Quantization': {}, "HyperParameterOptimization": {'ptq_algo': ['avg'], 'max_quant_count': 3}},
train_dataloader=train_loader,
eval_dataloader=train_loader)
ac.compress()

[zzjjay]

看起来是ocr模型？
可以参考自动压缩中ocr模型的压缩示例脚步进行尝试；
https://github.com/PaddlePaddle/PaddleSlim/tree/develop/example/auto_compression/ocr