Open AIHHU opened 2 years ago
model = tf.saved_model.load('./pretrained_model/dtln_saved_model') infer = model.signatures["serving_default"] out_block= infer(tf.constant(in_block))['conv1d_1'] 不知道这个[conv1d_1]转到paddle以后怎么推理
@AIHHU 有更具体的错误描述吗,不太理解您说的什么意思,是TF的保存的pb模型转Paddle出现错误了吗?
ValueError: Input 1 of node StatefulPartitionedCall was passed float from lstm/Variable:0 incompatible with expected resource.
我在x2paddle的时候发生的这个错误
model = tf.saved_model.load('./pretrained_model/dtln_saved_model') infer = model.signatures["serving_default"] out_block= infer(tf.constant(in_block))['conv1d_1'] 不知道这个[conv1d_1]转到paddle以后怎么推理