Python binding to export bytecode format for lite interpreter

[iseeyuan]

Stack from ghstack:

• Python binding to export bytecode format for lite interpreter #32621 Python binding to export bytecode format for lite interpreter

Export the "_save_for_mobile" method to Python so that the bytecode format for lite interpreter can be added or updated to the original script model.

It's the first step of python binding for lite interpreter, as discussed in this internal post and offline.

Next step is to export the load_for_mobile and run method of mobile module, so that users could verify the mobile model from Python.

Test: use the following python script to display the bytecode part of the updated model file.

#!/usr/bin/env python3
import sys
import pickle
import pprint
import zipfile


class FakeObject(object):
    def __init__(self, module, name, args):
        self.module = module
        self.name = name
        self.args = args
        self.state = None

    def __repr__(self):
        state_str = "" if self.state is None else f"(state={self.state!r})"
        return f"{self.module}.{self.name}{self.args!r}{state_str}"

    def __setstate__(self, state):
        self.state = state


class FakeClass(object):
    def __init__(self, module, name):
        self.module = module
        self.name = name
        self.__new__ = self.fake_new

    def __repr__(self):
        return f"{self.module}.{self.name}"

    def __call__(self, *args):
        return FakeObject(self.module, self.name, args)

    def fake_new(self, *args):
        return FakeObject(self.module, self.name, args)


class DumpUnpickler(pickle._Unpickler):
    def find_class(self, module, name):
        return FakeClass(module, name)

    def persistent_load(self, pid):
        return FakeObject("pers", "obj", (pid,))


def main(argv):
    zfile = zipfile.ZipFile(argv[1])
    names = [i for i in zfile.namelist() if "bytecode.pkl" in i]
    if not names: 
        print("bytecode.pkl not found.")
        return
    with zfile.open(names[0], "r") as handle:
        value = DumpUnpickler(handle).load()
    pprint.pprint(value)


if __name__ == "__main__":
    sys.exit(main(sys.argv))

Differential Revision: D19596359

[kostmo]

💊 CircleCI build failures summary and remediations

As of commit 18b2b37:

None of the build failures appear to be your fault.

• 1/1 recognized as flaky ❄️
  • Re-run these jobs?

Detailed failure analysis

One may explore the probable reasons each build failed interactively on the Dr. CI website.

❄️ 1 failure recognized as flaky

The following build failures have been detected as flaky and may not be your fault:

pytorch_linux_xenial_cuda10_1_cudnn7_py3_NO_AVX_NO_AVX2_test (1/1)

Step: "Test" (full log | pattern match details) ❄️

Jan 27 19:01:07 RuntimeError: quantization scale should be > 0

Jan 27 19:01:07   File "test_quantization.py", line 722, in test_quantized_rnn 
Jan 27 19:01:07     y, (h, c) = cell_dq(x, (h, c)) 
Jan 27 19:01:07   File "/opt/conda/lib/python3.6/site-packages/torch/nn/modules/module.py", line 540, in __call__ 
Jan 27 19:01:07     result = self.forward(*input, **kwargs) 
Jan 27 19:01:07   File "/opt/conda/lib/python3.6/site-packages/torch/nn/quantized/dynamic/modules/rnn.py", line 373, in forward 
Jan 27 19:01:07     return self.forward_tensor(input, hx) 
Jan 27 19:01:07   File "/opt/conda/lib/python3.6/site-packages/torch/nn/quantized/dynamic/modules/rnn.py", line 334, in forward_tensor 
Jan 27 19:01:07     input, hx, batch_sizes, max_batch_size, sorted_indices) 
Jan 27 19:01:07   File "/opt/conda/lib/python3.6/site-packages/torch/nn/quantized/dynamic/modules/rnn.py", line 315, in forward_impl 
Jan 27 19:01:07     self.batch_first, dtype=self.dtype, use_dynamic=True) 
Jan 27 19:01:07 RuntimeError: quantization scale should be > 0 
Jan 27 19:01:07  
Jan 27 19:01:07 ---------------------------------------------------------------------- 
Jan 27 19:01:07 Ran 36 tests in 54.070s 
Jan 27 19:01:07  
Jan 27 19:01:07 FAILED (errors=1, skipped=1) 
Jan 27 19:01:07  
Jan 27 19:01:07 Generating XML reports... 
Jan 27 19:01:07 Traceback (most recent call last): 
Jan 27 19:01:07   File "test/run_test.py", line 456, in <module> 
Jan 27 19:01:07     main() 

---

This comment was automatically generated by Dr. CI (expand for details).

Follow this link to opt-out of these comments for your Pull Requests.

Please report bugs/suggestions on the GitHub issue tracker.

This comment has been revised 5 times.

[facebook-github-bot]

@iseeyuan merged this pull request in c64dec1.

[suo]

Should this be part of the public API for ScriptModule? I thought we were not exposing the lite interpreter externally, so this may be confusing for external users.

[iseeyuan]

I thought the "_" is needed in C++ only. I might missed it here. Just to confirm: should we rename it to "_save_for_mobile"? I could have a quick patch for it.