使用tool/export.py中的paddle.jit.save(self, save_path)导出模型时报错

File "/home/CV/anaconda3/envs/paddle/lib/python3.8/site-packages/paddle/fluid/io.py", line 1403, in save_inference_model prepend_feed_ops(main_program, feeded_var_names)

File "/home/CV/anaconda3/envs/paddle/lib/python3.8/site-packages/paddle/fluid/io.py", line 1197, in prepend_feed_ops raise ValueError( ValueError: The feeded_var_names[1]: 'images' doesn't exist in pruned inference program. Please check whether 'images' is a valid feed_var name, or remove it from feeded_var_names if 'images' is not involved in the target_vars calculation.

我已经指定了

img_spec = paddle.static.InputSpec( shape=[1 * 6, 3, 448, 800], dtype='float32', name='images')

pts_spec = paddle.static.InputSpec( shape=[-1, 4], dtype='float32', name='pts')

img_metas_spec = { 'lidar2img': [ paddle.static.InputSpec( shape=[4, 4], dtype='float32', name='lidar2img') for i in range(6) ] } input_spec = [pts_spec, img_spec, img_metas_spec]

paddle.jit.to_static(self, input_spec=input_spec)