STT-tensorflow/tensorflow/lite/python
Karim Nosir 02032fb477 Always enable TF2 in tflite_convert command line tool. If user needs to use TF1 converter then they can pass --enable_v1_converter.
Before this users needed to set environment variable TF2_BEHAVIOR=1 before invoking the tool to use v2 converter.

PiperOrigin-RevId: 348591322
Change-Id: I25baf70c2bafa56a63ac9c629ad563e26a062d86
2020-12-22 00:30:12 -08:00
..
interpreter_wrapper TFLite Python API: allow float64, uint64 and complex128 2020-12-13 18:47:10 -08:00
optimize Add uint64 tensor support in TFLite 2020-11-17 14:01:21 -08:00
testdata Provide user friendly errors when upgrading legacy control flow ops is failed 2020-12-15 18:13:06 -08:00
BUILD Always enable TF2 in tflite_convert command line tool. If user needs to use TF1 converter then they can pass --enable_v1_converter. 2020-12-22 00:30:12 -08:00
convert_file_to_c_source_test.sh Update the usage of the 'absl-py' python library from TFLite Micro utils 2020-11-25 15:05:25 -08:00
convert_file_to_c_source.py Update the usage of the 'absl-py' python library from TFLite Micro utils 2020-11-25 15:05:25 -08:00
convert_saved_model_test.py
convert_saved_model.py
convert_test.py Reduce redundancy by replacing TFLite data types with TF data types 2020-10-13 19:29:04 -07:00
convert.py Add flag to add ops to flex delegate allowlist 2020-11-24 00:13:02 -08:00
interpreter_test.py Fix a bug on setting string input with np.array 2020-10-15 21:23:32 -07:00
interpreter.py Introduce new Python API for accessing SignatureDefs in a TFLite model. 2020-12-03 19:13:37 -08:00
lite_constants.py Added an option TFLITE_BUILTINS_ACTIVATIONS_INT16_WEIGHTS_INT8 to 2020-01-27 14:01:11 +00:00
lite_flex_test.py Add flag to add ops to flex delegate allowlist 2020-11-24 00:13:02 -08:00
lite_test.py Fix lite_tests for new quantizer 2020-12-16 17:58:55 -08:00
lite_v2_test_util.py Introduce new Python API for accessing SignatureDefs in a TFLite model. 2020-12-03 19:13:37 -08:00
lite_v2_test.py Fix lite_tests for new quantizer 2020-12-16 17:58:55 -08:00
lite.py Skip the dequantize op check when the tensor type is not float 2020-12-22 00:04:25 -08:00
op_hint.py Add deprecation message for lite.OpHint in favor of the new fusion mechanism under https://www.tensorflow.org/lite/convert/operation_fusion 2020-11-20 18:48:23 -08:00
schema_util.py Update Python codes that read builtin code after schema change. 2020-10-22 01:02:58 -07:00
test_util_test.py Add flag to add ops to flex delegate allowlist 2020-11-24 00:13:02 -08:00
test_util.py Add flag to add ops to flex delegate allowlist 2020-11-24 00:13:02 -08:00
tflite_convert_test.py Add flag to add ops to flex delegate allowlist 2020-11-24 00:13:02 -08:00
tflite_convert.py Always enable TF2 in tflite_convert command line tool. If user needs to use TF1 converter then they can pass --enable_v1_converter. 2020-12-22 00:30:12 -08:00
tflite_keras_util.py Refactor keras dependency code 2020-10-30 15:59:19 -07:00
util_test.py Skip the dequantize op check when the tensor type is not float 2020-12-22 00:04:25 -08:00
util.py Skip the dequantize op check when the tensor type is not float 2020-12-22 00:04:25 -08:00
wrap_toco.py tflite_convert: register custom op defs at the beginning of process 2020-10-05 14:54:04 -07:00