experiments/STT-tensorflow
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity
rename
STT-tensorflow/tensorflow/lite/toco
History
Karim Nosir 4a10c37bcf [lite] Deprecate _custom_opdefs field in TfLiteConverter. 
It is unused. Also, clean up references and mark field deprecated in the proto file.

PiperOrigin-RevId: 358274276
Change-Id: I1ba5990597230c8bd546f8894e5a67b2fc0ec305
2021-02-18 15:11:55 -08:00
..
g3doc
graph_transformations Pre-allocate memory for vectors where size is known. 2021-02-10 20:07:07 -08:00
logging Comment out a number of google-internal targets when copybara-exporting instead of removing them. 2021-01-26 05:47:31 -08:00
python PY2 removal cleanup 2021-01-20 17:11:44 -08:00
runtime
tensorflow_graph_matching Optimize calls to std::string::find() and friends for a single char. 2020-12-17 18:11:13 -08:00
tflite Add nominal support for unsigned 32-bit integer tensor types. 2021-02-17 14:16:29 -08:00
allocate_transient_arrays.cc
allocate_transient_arrays.h
args.cc
args.h
BUILD Comment out a number of google-internal targets when copybara-exporting instead of removing them. 2021-01-26 05:47:31 -08:00
dump_graphviz.cc
dump_graphviz.h
export_tensorflow.cc Add nominal support for unsigned 32-bit integer tensor types. 2021-02-17 14:16:29 -08:00
export_tensorflow.h
format_port.h
import_tensorflow_test.cc Add nominal support for unsigned 32-bit integer tensor types. 2021-02-17 14:16:29 -08:00
import_tensorflow.cc Add nominal support for unsigned 32-bit integer tensor types. 2021-02-17 14:16:29 -08:00
import_tensorflow.h
model_cmdline_flags_test.cc
model_cmdline_flags.cc
model_cmdline_flags.h
model_flags.proto Propagate unknown rank of tensor shape information to MLIR 2020-11-17 00:38:39 -08:00
model.h Rollback of BroadcastTo op additions (part 3) 2020-07-23 16:38:48 -07:00
README.md
tensorflow_util.cc
tensorflow_util.h
toco_cmdline_flags_test.cc
toco_cmdline_flags.cc
toco_cmdline_flags.h
toco_convert_test.cc
toco_convert.cc
toco_convert.h
toco_flags.proto [lite] Deprecate _custom_opdefs field in TfLiteConverter. 2021-02-18 15:11:55 -08:00
toco_graphviz_dump_options.cc
toco_graphviz_dump_options.h
toco_port_test.cc
toco_port.cc Internal naming change 2020-10-23 09:50:59 -07:00
toco_port.h
toco_tooling.cc
toco_tooling.h
toco_types.h
toco.cc
tooling_util_test.cc
tooling_util.cc Add nominal support for unsigned 32-bit integer tensor types. 2021-02-17 14:16:29 -08:00
tooling_util.h Removed unused protobuf dependency 2021-01-22 10:46:21 -08:00
types.proto Add nominal support for unsigned 32-bit integer tensor types. 2021-02-17 14:16:29 -08:00

README.md

TensorFlow Lite Converter

The TensorFlow Lite Converter converts TensorFlow graphs into TensorFlow Lite graphs. There are additional usages that are also detailed in the usage documentation.

Usage documentation

Usage information is given in these documents:

Where the converter fits in the TensorFlow landscape

Once an application developer has a trained TensorFlow model, the TensorFlow Lite Converter will accept that model and generate a TensorFlow Lite FlatBuffer file. The converter currently supports SavedModels, frozen graphs (models generated via freeze_graph.py), and tf.Keras model files. The TensorFlow Lite FlatBuffer file can be shipped to client devices, generally mobile devices, where the TensorFlow Lite interpreter handles them on-device. This flow is represented in the diagram below.

drawing