未验证 提交 5ff45357 编写于 作者: H Huihuang Zheng 提交者: GitHub

[Dy2Stat] Fix ProgramTranslator.save_inference_model API Doc (#24584)

As the title.
上级 c4dd596d
......@@ -559,14 +559,14 @@ class ProgramTranslator(object):
Args:
dirname (str): the directory to save the inference model.
feed (list[int], optional): the input variable indices of the saved
inference model. If None, all input variables of the
ProgramTranslator would be the inputs of the saved inference
model. Default None.
fetch (list[int], optional): the output variable indices of the
saved inference model. If None, all output variables of the
TracedLayer object would be the outputs of the saved inference
model. Default None.
feed (list[int], optional): the indices of the input variables of the
dygraph functions which will be saved as input variables in
inference model. If None, all input variables of the dygraph function
would be the inputs of the saved inference model. Default None.
fetch (list[int], optional): the indices of the returned variable of the
dygraph functions which will be saved as output variables in
inference model. If None, all output variables of the dygraph function
would be the outputs of the saved inference model. Default None.
Returns:
None
Examples:
......@@ -599,12 +599,12 @@ class ProgramTranslator(object):
adam.minimize(loss)
net.clear_gradients()
# Save inference model.
# Note that fetch=[0] means we set 'y' as the inference output.
# Note that fetch=[0] means we set 'z' as the inference output.
prog_trans = ProgramTranslator()
prog_trans.save_inference_model("./dy2stat_infer_model", fetch=[0])
# In this example, the inference model will be pruned based on input (x) and
# output (y). The pruned inference program is going to be saved in the folder
# In this example, the inference model will be pruned based on output (z).
# The pruned inference program is going to be saved in the folder
# "./dy2stat_infer_model" and parameters are going to be saved in separate
# files in the folder.
"""
......
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