MCPcopy Index your code
hub / github.com/pytorch/pytorch / write

Method write

caffe2/python/dataio.py:161–171  ·  view source on GitHub ↗

Add operations to `writer_net` that write the next batch of data. Operations added to the net must be thread-safe and unique, that is: multiple writers must be able to write to the dataset in parallel. Args: fields: a tuple of BlobReference containing the batch

(self, writer_net, fields)

Source from the content-addressed store, hash-verified

159 return self._schema
160
161 def write(self, writer_net, fields):
162 """Add operations to `writer_net` that write the next batch of data.
163
164 Operations added to the net must be thread-safe and unique, that is:
165 multiple writers must be able to write to the dataset in parallel.
166
167 Args:
168 fields: a tuple of BlobReference containing the batch of data to
169 write.
170 """
171 raise NotImplementedError('Writers must implement write.')
172
173 def write_record(self, writer_net, fields):
174 if isinstance(fields, Field):

Callers 15

write_recordMethod · 0.95
write_exMethod · 0.95
test_eventsMethod · 0.45
writeFunction · 0.45
_pprint_implMethod · 0.45
_pprint_implMethod · 0.45
_pprint_implMethod · 0.45
_pprint_implMethod · 0.45
CompareNetsMethod · 0.45
BenchmarkFunction · 0.45
renderFunction · 0.45

Calls

no outgoing calls

Tested by 12

test_eventsMethod · 0.36
setUpModuleFunction · 0.36
print_test_debug_infoFunction · 0.36
serialize_testMethod · 0.36
test_caffe2_to_onnxMethod · 0.36
test_onnx_to_caffe2Method · 0.36
test_convert_end2endMethod · 0.36
test_text_file_readerMethod · 0.36
test_record_queueMethod · 0.36
create_a_listMethod · 0.36