Skip to content

Fix the bug in the AnalysisPredictor and add more directions about io APIs. #17639

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
wzzju merged 3 commits into from
May 27, 2019
Merged

Fix the bug in the AnalysisPredictor and add more directions about io APIs. #17639

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
16a1339
fix the bug that sub_scope_ may be null in AnalysisPredictor::Run.
wzzju May 23, 2019
cb57806
add more directions about io APIs' docs.
wzzju May 24, 2019
b2b170b
update the API.spec. test=develop test=document_preview
wzzju May 24, 2019
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
8 changes: 4 additions & 4 deletions paddle/fluid/API.spec
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -48,13 +48,13 @@ paddle.fluid.BuildStrategy.GradientScaleStrategy.__init__ __init__(self: paddle.
paddle.fluid.BuildStrategy.ReduceStrategy.__init__ __init__(self: paddle.fluid.core.ParallelExecutor.BuildStrategy.ReduceStrategy, arg0: int) -> None
paddle.fluid.BuildStrategy.__init__ __init__(self: paddle.fluid.core.ParallelExecutor.BuildStrategy) -> None
paddle.fluid.io.save_vars (ArgSpec(args=['executor', 'dirname', 'main_program', 'vars', 'predicate', 'filename'], varargs=None, keywords=None, defaults=(None, None, None, None)), ('document', '869104f47e6fd21d897c3fcc426aa942'))
paddle.fluid.io.save_params (ArgSpec(args=['executor', 'dirname', 'main_program', 'filename'], varargs=None, keywords=None, defaults=(None, None)), ('document', '3a7a99abac3e1bf898871fe609354218'))
paddle.fluid.io.save_params (ArgSpec(args=['executor', 'dirname', 'main_program', 'filename'], varargs=None, keywords=None, defaults=(None, None)), ('document', '07ffd5351b30cf47172ccfd61bd0de6f'))
paddle.fluid.io.save_persistables (ArgSpec(args=['executor', 'dirname', 'main_program', 'filename'], varargs=None, keywords=None, defaults=(None, None)), ('document', '9141bb5f32caf7975eb3fd88c8a1b2da'))
paddle.fluid.io.load_vars (ArgSpec(args=['executor', 'dirname', 'main_program', 'vars', 'predicate', 'filename'], varargs=None, keywords=None, defaults=(None, None, None, None)), ('document', '1bb9454cf09d71f190bb51550c5a3ac9'))
paddle.fluid.io.load_params (ArgSpec(args=['executor', 'dirname', 'main_program', 'filename'], varargs=None, keywords=None, defaults=(None, None)), ('document', '41779819cef32f2246e83aebc5a002e2'))
paddle.fluid.io.load_params (ArgSpec(args=['executor', 'dirname', 'main_program', 'filename'], varargs=None, keywords=None, defaults=(None, None)), ('document', '944291120d37bdb037a689d2c86d0a6e'))
paddle.fluid.io.load_persistables (ArgSpec(args=['executor', 'dirname', 'main_program', 'filename'], varargs=None, keywords=None, defaults=(None, None)), ('document', '28df5bfe26ca7a077f91156abb0fe6d2'))
paddle.fluid.io.save_inference_model (ArgSpec(args=['dirname', 'feeded_var_names', 'target_vars', 'executor', 'main_program', 'model_filename', 'params_filename', 'export_for_deployment'], varargs=None, keywords=None, defaults=(None, None, None, True)), ('document', 'af82e1b5fe5764029905a191b987f63d'))
paddle.fluid.io.load_inference_model (ArgSpec(args=['dirname', 'executor', 'model_filename', 'params_filename', 'pserver_endpoints'], varargs=None, keywords=None, defaults=(None, None, None)), ('document', 'aaf3392332f2e5ef9d9177985be2c04a'))
paddle.fluid.io.save_inference_model (ArgSpec(args=['dirname', 'feeded_var_names', 'target_vars', 'executor', 'main_program', 'model_filename', 'params_filename', 'export_for_deployment'], varargs=None, keywords=None, defaults=(None, None, None, True)), ('document', '89539e459eb959145f15c9c3e38fa97c'))
paddle.fluid.io.load_inference_model (ArgSpec(args=['dirname', 'executor', 'model_filename', 'params_filename', 'pserver_endpoints'], varargs=None, keywords=None, defaults=(None, None, None)), ('document', '2f54d7c206b62f8c10f4f9d78c731cfd'))
paddle.fluid.io.PyReader.__init__ (ArgSpec(args=['self', 'feed_list', 'capacity', 'use_double_buffer', 'iterable'], varargs=None, keywords=None, defaults=(True, False)), ('document', '6adf97f83acf6453d4a6a4b1070f3754'))
paddle.fluid.io.PyReader.decorate_batch_generator (ArgSpec(args=['self', 'reader', 'places'], varargs=None, keywords=None, defaults=(None,)), ('document', '4a072de39998ee4e0de33fcec11325a6'))
paddle.fluid.io.PyReader.decorate_sample_generator (ArgSpec(args=['self', 'sample_generator', 'batch_size', 'drop_last', 'places'], varargs=None, keywords=None, defaults=(True, None)), ('document', '3db4b24d33fe4f711e303f9673dc5c6a'))
Expand Down
5 changes: 4 additions & 1 deletion paddle/fluid/inference/api/analysis_predictor.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -202,6 +202,7 @@ bool AnalysisPredictor::Run(const std::vector<PaddleTensor> &inputs,
timer.tic();
// set feed variable
framework::Scope *scope = sub_scope_ ? sub_scope_ : scope_.get();
PADDLE_ENFORCE_NOT_NULL(scope, "The scope should not be nullptr.");
if (!SetFeed(inputs, scope)) {
LOG(ERROR) << "fail to set feed";
return false;
Expand Down Expand Up @@ -229,7 +230,9 @@ bool AnalysisPredictor::Run(const std::vector<PaddleTensor> &inputs,
// Here is a bugfix, collect all the container variables, and reset then to a
// bool; the next time, the operator will call MutableData and construct a new
// container again, so that the container will be empty for each batch.
tensor_array_batch_cleaner_.CollectNoTensorVars(sub_scope_);
if (sub_scope_) {
tensor_array_batch_cleaner_.CollectNoTensorVars(sub_scope_);
}
tensor_array_batch_cleaner_.ResetNoTensorVars();
return true;
}
Expand Down
24 changes: 18 additions & 6 deletions python/paddle/fluid/io.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -242,7 +242,9 @@ def save_params(executor, dirname, main_program=None, filename=None):
NOTICE: Some variables are not Parameter while they are necessary for
training. So you can NOT save and continue your training just by
`save_params()` and `load_params()`. Please use `save_persistables()`
and `load_persistables()` instead.
and `load_persistables()` instead. If you want to save your model for
the inference, please use the `save_inference_model` API. You can refer
to :ref:`api_guide_model_save_reader_en` for more details.

Args:
executor(Executor): The executor to run for saving parameters.
Expand Down Expand Up @@ -658,6 +660,9 @@ def load_params(executor, dirname, main_program=None, filename=None):
training. So you can NOT save and continue your training just by
`save_params()` and `load_params()`. Please use `save_persistables()`
and `load_persistables()` instead.
If you want to load the pre-trained model structure and parameters
for the inference, please use the `load_inference_model` API. You can
refer to :ref:`api_guide_model_save_reader_en` for more details.

Args:
executor(Executor): The executor to run for loading parameters.
Expand Down Expand Up @@ -906,6 +911,10 @@ def save_inference_model(dirname,
"""
Prune the given `main_program` to build a new program especially for inference,
and then save it and all related parameters to given `dirname` by the `executor`.
If you just want to save parameters of your trained model, please use the
`save_params` API. You can refer to :ref:`api_guide_model_save_reader_en` for
more details.


Args:
dirname(str): The directory path to save the inference model.
Expand Down Expand Up @@ -1077,7 +1086,10 @@ def load_inference_model(dirname,
params_filename=None,
pserver_endpoints=None):
"""
Load inference model from a directory
Load inference model from a directory. By this API, you can get the model
structure(inference program) and model parameters. If you just want to load
parameters of the pre-trained model, please use the `load_params` API.
You can refer to :ref:`api_guide_model_save_reader_en` for more details.

Args:
dirname(str): The directory path
Expand Down Expand Up @@ -1128,19 +1140,19 @@ def load_inference_model(dirname,
fluid.io.save_inference_model(dirname=path, feeded_var_names=['img'],
target_vars=[hidden_b], executor=exe, main_program=main_prog)
tensor_img = np.array(np.random.random((1, 64, 784)), dtype=np.float32)
[inference_program, feed_target_names, fetch_targets] = \
fluid.io.load_inference_model(dirname=path, executor=exe)
[inference_program, feed_target_names, fetch_targets] = (
fluid.io.load_inference_model(dirname=path, executor=exe))
results = exe.run(inference_program,
feed={feed_target_names[0]: tensor_img},
fetch_list=fetch_targets)

# endpoints is your pserver endpoints list, the above is just an example
endpoints = ["127.0.0.1:2023","127.0.0.1:2024"]
# if we need lookup table, we will use:
[dist_inference_program, dist_feed_target_names, dist_fetch_targets] = \
[dist_inference_program, dist_feed_target_names, dist_fetch_targets] = (
fluid.io.load_inference_model(dirname=path,
executor=exe,
pserver_endpoints=endpoints)
pserver_endpoints=endpoints))

# In this example, the inference program was saved in the
# "./infer_model/__model__" and parameters were saved in
Expand Down