experiments/STT-tensorflow
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity
STT-tensorflow/tensorflow/lite/kernels/internal/reference
History
T.J. Alumbaugh 9e7d5ef6f2 Full int8 quantization BatchMatMul 
PiperOrigin-RevId: 317304259
Change-Id: Icf96d9d129db30b965e36f5c8befd27762b173b2
2020-06-19 07:33:08 -07:00
..
integer_ops
Merge pull request #35997 from psunn:int16_transpose_conv
2020-06-03 09:08:49 -07:00
sparse_ops
Add one example of sparse FullyConnected kernel.
2020-02-05 14:25:02 -08:00
add.h
remove unnecessary requires_flat_size_broadcast check.
2019-10-16 21:05:26 -07:00
arg_min_max.h
…
batch_matmul.h
Full int8 quantization BatchMatMul
2020-06-19 07:33:08 -07:00
binary_function.h
minor spelling tweaks
2020-03-19 01:09:19 +09:00
ceil.h
…
comparisons.h
TFLite binary size optimization by removing templates
2020-05-05 20:04:05 -07:00
concatenation.h
Add indirection layer for selecting cmath functions from global namespace.
2020-03-27 16:06:11 -07:00
conv.h
Add activation function to per channel weights-only quantized conv reference implementation and update per channel weights-only quantized conv to use optimized asymmetric quantize floats method.
2019-11-05 18:35:37 -08:00
densify.h
Update source files with used includes.
2020-06-15 18:42:14 -07:00
depthwiseconv_float.h
…
depthwiseconv_uint8.h
TF Lite: Fix lint error.
2020-03-12 01:50:08 -07:00
dequantize.h
Remove unused function
2020-03-26 11:04:00 -07:00
floor.h
Add floor op for micro
2019-07-31 18:14:22 -07:00
fully_connected.h
TFL MCU: Move reference L2Normalization implementation into its own file.
2020-04-17 16:18:32 -07:00
l2normalization.h
add L2 Normalization op to micro
2020-04-21 11:20:43 -07:00
legacy_reference_ops.h
PR #34903: TransposeConv with Bias
2020-04-22 12:33:46 -07:00
logistic.h
Add indirection layer for selecting cmath functions from global namespace.
2020-03-27 16:06:11 -07:00
maximum_minimum.h
Add 5D support for TFLite Maximum Minimum
2020-03-24 03:37:45 -07:00
mul.h
PR #32168: Port mul from Tensorflow Lite to Tensorflow Lite Micro
2019-11-08 13:50:23 -08:00
neg.h
Port NEG operation to micro
2019-08-16 07:33:24 -07:00
non_max_suppression.h
Update source files with used includes.
2020-06-15 18:42:14 -07:00
pad.h
Port pad op to micro.
2019-12-10 13:13:07 -08:00
pooling.h
Add indirection layer for selecting cmath functions from global namespace.
2020-03-27 16:06:11 -07:00
portable_tensor_utils_impl.h
Add optimized MatrixBatchVectorMultiplyAccumulate for asymmetric inputs for sse
2020-05-18 01:35:29 -07:00
portable_tensor_utils.cc
Add optimized MatrixBatchVectorMultiplyAccumulate for asymmetric inputs for sse
2020-05-18 01:35:29 -07:00
portable_tensor_utils.h
Add optimized MatrixBatchVectorMultiplyAccumulate for asymmetric inputs for sse
2020-05-18 01:35:29 -07:00
prelu.h
Extend PRelu op in TFLite to accept the same shape inputs.
2020-05-27 16:13:57 -07:00
process_broadcast_shapes.h
Migrate float/uint8 broadcast fivefold add to use binary broadcast fivefold.
2020-05-28 19:58:06 -07:00
quantize.h
Update source files with used includes.
2020-06-15 18:42:14 -07:00
reduce.h
Merge pull request #38111 from antmicro:tf-demo-update
2020-05-29 14:11:17 -07:00
reference_ops.h
Create fully integer ResizeBilinear kernel.
2020-06-18 08:51:40 -07:00
requantize.h
Reference ruy from its new location as a separate GitHub project.
2020-04-03 11:17:08 -07:00
resize_nearest_neighbor.h
Merge pull request #38111 from antmicro:tf-demo-update
2020-05-29 14:11:17 -07:00
round.h
…
softmax.h
Merge pull request #35996 from psunn:int16_softmax
2020-04-15 16:17:49 -07:00
strided_slice.h
Update source files with used includes.
2020-06-15 18:42:14 -07:00
sub.h
Update source files with used includes.
2020-06-15 18:42:14 -07:00
svdf.h
Update source files with used includes.
2020-06-15 18:42:14 -07:00
tanh.h
Add quantized TANH operation, mostly copied from LOGISTIC.
2020-05-28 05:30:59 -07:00