提交 3e8f5ce4 编写于 作者: xuyang2233's avatar xuyang2233

merge export_model

上级 f9e379ab
...@@ -62,7 +62,6 @@ Loss: ...@@ -62,7 +62,6 @@ Loss:
PostProcess: PostProcess:
name: SPINAttnLabelDecode name: SPINAttnLabelDecode
character_dict_path: ./ppocr/utils/dict/spin_dict.txt
use_space_char: False use_space_char: False
......
...@@ -84,7 +84,7 @@ def export_single_model(model, arch_config, save_path, logger, quanter=None): ...@@ -84,7 +84,7 @@ def export_single_model(model, arch_config, save_path, logger, quanter=None):
"When there is tps in the network, variable length input is not supported, and the input size needs to be the same as during training" "When there is tps in the network, variable length input is not supported, and the input size needs to be the same as during training"
) )
infer_shape[-1] = 100 infer_shape[-1] = 100
if arch_config["algorithm"] == "NRTR": if arch_config["algorithm"] == "NRTR" or arch_config["algorithm"] == "SPIN":
infer_shape = [1, 32, 100] infer_shape = [1, 32, 100]
elif arch_config["model_type"] == "table": elif arch_config["model_type"] == "table":
infer_shape = [3, 488, 488] infer_shape = [3, 488, 488]
......
Markdown is supported
0% .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑!
想要评论请 注册