experiments/STT-tensorflow
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity
STT-tensorflow/tensorflow/lite/kernels
History
Mihai Maruseac baa1b690f7
Cast away a const in intermediate API
2020-09-20 16:57:05 -07:00
..
internal
[tflite] Ensure ResolveAxis properly handles negative inputs.
2020-09-19 18:19:28 -07:00
activation_functor.h
…
activations_test.cc
Rewrite quantized hardswish for good accuracy in all cases.
2019-07-12 18:26:56 -07:00
activations.cc
Optimize 8bit Softmax op handler
2019-08-02 17:52:28 -07:00
add_n_test.cc
Refactor TFLite kernel tests to use a common main
2019-05-22 09:21:06 -07:00
add_n.cc
Add built-in operator AddN. It only supports float and int_32 now.
2019-02-04 14:24:47 -08:00
add_test.cc
Refactor TFLite kernel tests to use a common main
2019-05-22 09:21:06 -07:00
add.cc
Move USE_NEON declaration to cpu_check.h, include that header everywhere USE_NEON is used.
2019-06-19 10:47:12 -07:00
arg_min_max_test.cc
Added support for ArgMin and ArgMax.
2019-07-03 16:59:32 -07:00
arg_min_max.cc
axis test cases for argminmax
2019-03-28 18:12:47 +05:30
assign_variable.cc
Prorotype TFLite resource variables
2019-07-25 12:10:10 -07:00
audio_spectrogram_test.cc
Refactor TFLite kernel tests to use a common main
2019-05-22 09:21:06 -07:00
audio_spectrogram.cc
…
basic_rnn_test.cc
Refactor TFLite kernel tests to use a common main
2019-05-22 09:21:06 -07:00
basic_rnn.cc
Move GetInt8DataPtr function to kernel_util.h
2019-04-16 20:09:51 -07:00
batch_to_space_nd_test.cc
Add a batch-1 test case for batch_to_space_nd
2019-06-11 14:09:50 -07:00
batch_to_space_nd.cc
Create int8 batch_to_space_nd.
2019-02-11 20:42:50 -08:00
bidirectional_sequence_lstm_test.cc
Refactor TFLite kernel tests to use a common main
2019-05-22 09:21:06 -07:00
bidirectional_sequence_lstm.cc
Fix the check to accept Int8 format as well for Bidi-LSTM.
2019-04-19 11:47:46 -07:00
bidirectional_sequence_rnn_test.cc
Refactor TFLite kernel tests to use a common main
2019-05-22 09:21:06 -07:00
bidirectional_sequence_rnn.cc
Move GetInt8DataPtr function to kernel_util.h
2019-04-16 20:09:51 -07:00
BUILD
Increased tolerance for some tests.
2019-08-05 12:29:31 -07:00
cast_test.cc
Add cast op support to the NNAPI delegate
2019-07-12 03:27:12 -07:00
cast.cc
…
ceil_test.cc
Refactor TFLite kernel tests to use a common main
2019-05-22 09:21:06 -07:00
ceil.cc
Update ceil.cc
2019-01-08 09:28:38 +05:30
comparisons_test.cc
Refactor TFLite kernel tests to use a common main
2019-05-22 09:21:06 -07:00
comparisons.cc
Lite: Compare Op bug fix
2019-04-04 16:53:36 +05:30
concatenation_test.cc
Refactor TFLite kernel tests to use a common main
2019-05-22 09:21:06 -07:00
concatenation.cc
Merge pull request #26134 from ANSHUMAN87:concat-refactor
2019-05-03 14:56:39 -07:00
conv_test.cc
[1] Review comments handled
2019-07-02 14:32:41 +05:30
conv.cc
1. Remove all references to tflite::cpu_backend_support as we no longer do reference-counting on the cpu backend context object and GetFromContext is moved to CpuBackendContext class as a static member function.
2019-07-22 21:27:09 -07:00
cpu_backend_context.cc
1. Remove all references to tflite::cpu_backend_support as we no longer do reference-counting on the cpu backend context object and GetFromContext is moved to CpuBackendContext class as a static member function.
2019-07-22 21:27:09 -07:00
cpu_backend_context.h
1. Remove all references to tflite::cpu_backend_support as we no longer do reference-counting on the cpu backend context object and GetFromContext is moved to CpuBackendContext class as a static member function.
2019-07-22 21:27:09 -07:00
cpu_backend_gemm_custom_gemv.h
Refactor tflite_with_ruy setting to rely on a base cc_library define
2019-05-23 20:03:22 -07:00
cpu_backend_gemm_eigen.cc
Refactor tflite_with_ruy setting to rely on a base cc_library define
2019-05-23 20:03:22 -07:00
cpu_backend_gemm_eigen.h
Refactor tflite_with_ruy setting to rely on a base cc_library define
2019-05-23 20:03:22 -07:00
cpu_backend_gemm_gemmlowp.h
Refactor tflite_with_ruy setting to rely on a base cc_library define
2019-05-23 20:03:22 -07:00
cpu_backend_gemm_params.h
Port FullyConnected/float to cpu_backend_gemm.
2019-05-07 14:06:21 -07:00
cpu_backend_gemm_ruy.h
Templatize Gemm in a 'quantization flavor' distinguishing per-channel vs
2019-04-30 20:11:42 -07:00
cpu_backend_gemm_test.cc
Make TfLiteInternalBackendContext as a interface-only abstract class.
2019-07-22 19:34:07 -07:00
cpu_backend_gemm.h
Refactor tflite_with_ruy setting to rely on a base cc_library define
2019-05-23 20:03:22 -07:00
cpu_backend_threadpool_test.cc
Make TfLiteInternalBackendContext as a interface-only abstract class.
2019-07-22 19:34:07 -07:00
cpu_backend_threadpool.h
Refactor tflite_with_ruy setting to rely on a base cc_library define
2019-05-23 20:03:22 -07:00
custom_ops_register.h
Implement Rfft2d as a custom ops.
2019-06-12 13:27:42 -07:00
depthwise_conv_test.cc
Fix depthwise_conv test. Looks somehow they are not really tested. :(
2019-08-05 18:57:11 -07:00
depthwise_conv.cc
Add a check on the 0th dimension of filter for DepthwiseConv.
2019-07-23 19:53:07 -07:00
dequantize_test.cc
Add int16 support to Dequant.
2019-07-19 00:26:49 -07:00
dequantize.cc
Add optimized dequantize methods for int8 and int16
2019-08-02 02:59:40 -07:00
detection_postprocess_test.cc
Refactor TFLite kernel tests to use a common main
2019-05-22 09:21:06 -07:00
detection_postprocess.cc
fix buffer overrun of "std::vector<int> sorted_indices".
2019-06-19 10:14:01 +09:00
div_test.cc
Merge pull request #26570 from TCLResearchEurope:quantized-div
2019-06-21 12:50:01 -07:00
div.cc
Merge pull request #26570 from TCLResearchEurope:quantized-div
2019-06-21 12:50:01 -07:00
eigen_support_test.cc
Avoid im2col creation for multi-threaded conv
2019-04-01 16:39:57 -07:00
eigen_support.cc
Avoid im2col creation for multi-threaded conv
2019-04-01 16:39:57 -07:00
eigen_support.h
Lazily create the Eigen thread pool for TFLite
2019-03-07 15:35:34 -08:00
elementwise_test.cc
Refactor TFLite kernel tests to use a common main
2019-05-22 09:21:06 -07:00
elementwise.cc
Added tf.cos to TFLite Ops.
2019-02-12 12:19:14 -08:00
embedding_lookup_sparse_test.cc
Refactor TFLite kernel tests to use a common main
2019-05-22 09:21:06 -07:00
embedding_lookup_sparse.cc
…
embedding_lookup_test.cc
Refactor TFLite kernel tests to use a common main
2019-05-22 09:21:06 -07:00
embedding_lookup.cc
Converting EmbeddingLookup from hybrid to ordinary op.
2019-05-02 13:00:11 -07:00
exp_test.cc
Refactor TFLite kernel tests to use a common main
2019-05-22 09:21:06 -07:00
exp.cc
…
expand_dims_test.cc
Add EXPAND_DIMS support to NN API delegate
2019-06-28 04:31:01 -07:00
expand_dims.cc
Merge pull request #24648 from siju-samuel:patch-9
2019-05-24 09:00:27 -07:00
fake_quant_test.cc
Refactor TFLite kernel tests to use a common main
2019-05-22 09:21:06 -07:00
fake_quant.cc
…
fill_test.cc
Refactor TFLite kernel tests to use a common main
2019-05-22 09:21:06 -07:00
fill.cc
[LITE]Fix array leak during resize o/p
2019-07-03 10:14:29 +05:30
floor_div_test.cc
Refactor TFLite kernel tests to use a common main
2019-05-22 09:21:06 -07:00
floor_div.cc
[2] Review comments handled
2019-04-25 12:04:23 +05:30
floor_mod_test.cc
Refactor TFLite kernel tests to use a common main
2019-05-22 09:21:06 -07:00
floor_mod.cc
Reorganised the code for FloorMod.
2019-03-21 08:32:15 +05:30
floor_test.cc
Support running TFLite kernel unit tests with NNAPI
2019-05-17 11:56:13 -07:00
floor.cc
Add floor op for micro
2019-07-31 18:14:22 -07:00
fully_connected_test.cc
Add proper support for per-channel quantized operations and int8 input / output tensors.
2019-06-26 17:31:36 -07:00
fully_connected.cc
1. Remove all references to tflite::cpu_backend_support as we no longer do reference-counting on the cpu backend context object and GetFromContext is moved to CpuBackendContext class as a static member function.
2019-07-22 21:27:09 -07:00
gather_nd_test.cc
Refactor TFLite kernel tests to use a common main
2019-05-22 09:21:06 -07:00
gather_nd.cc
Add implementation for operator gather_nd.
2019-02-12 20:52:02 -08:00
gather_test.cc
Refactor TFLite kernel tests to use a common main
2019-05-22 09:21:06 -07:00
gather.cc
Add bool support for Gather.
2019-08-01 18:19:06 -07:00
hashtable_lookup_test.cc
Refactor TFLite kernel tests to use a common main
2019-05-22 09:21:06 -07:00
hashtable_lookup.cc
…
if_test.cc
Refactor TFLite kernel tests to use a common main
2019-05-22 09:21:06 -07:00
if.cc
Graduate TFLite control flow ops from experimental to builtin
2019-07-20 15:51:38 -07:00
kernel_util_test.cc
Merge pull request #26834 from ANSHUMAN87:kernel_util_bug_fix
2019-04-29 17:17:43 -07:00
kernel_util.cc
Lite: Kernel Util bug fix
2019-03-25 10:00:20 +05:30
kernel_util.h
Cast away a const in intermediate API
2020-09-20 16:57:05 -07:00
l2norm_test.cc
Add guard to avoid acceleration of L2 Normalization with input rank != 4
2019-07-15 01:49:54 -07:00
l2norm.cc
Create int8 L2Norm.
2019-02-25 17:48:06 -08:00
local_response_norm_test.cc
Support running TFLite kernel unit tests with NNAPI
2019-05-17 11:56:13 -07:00
local_response_norm.cc
…
log_softmax_test.cc
Refactor TFLite kernel tests to use a common main
2019-05-22 09:21:06 -07:00
logical_test.cc
Refactor TFLite kernel tests to use a common main
2019-05-22 09:21:06 -07:00
logical.cc
Move binary functions to a separate header.
2019-08-05 19:51:02 -07:00
lsh_projection_test.cc
Support running TFLite kernel unit tests with NNAPI
2019-05-17 11:56:13 -07:00
lsh_projection.cc
Clang format changes
2019-03-20 21:33:07 +05:30
lstm_eval.cc
Use eigen for lstm_eval activations.
2019-05-09 21:36:03 -07:00
lstm_eval.h
Add builtin layer norm lstm.
2018-12-19 16:56:19 -08:00
lstm_test.cc
Increased tolerance for some tests.
2019-08-05 12:29:31 -07:00
lstm.cc
1. Remove all references to tflite::cpu_backend_support as we no longer do reference-counting on the cpu backend context object and GetFromContext is moved to CpuBackendContext class as a static member function.
2019-07-22 21:27:09 -07:00
matrix_diag_test.cc
Refactor TFLite kernel tests to use a common main
2019-05-22 09:21:06 -07:00
matrix_diag.cc
Build fix
2019-03-18 17:24:08 -07:00
matrix_set_diag_test.cc
Refactor TFLite kernel tests to use a common main
2019-05-22 09:21:06 -07:00
matrix_set_diag.cc
Implement MatrixSetDiag and Eye
2019-03-21 13:46:55 -07:00
maximum_minimum_test.cc
Refactor TFLite kernel tests to use a common main
2019-05-22 09:21:06 -07:00
maximum_minimum.cc
Clang format changes
2019-03-20 21:33:07 +05:30
mfcc_test.cc
Allow mfcc operator to take scalar rate input
2019-07-19 15:02:08 -07:00
mfcc.cc
Allow mfcc operator to take scalar rate input
2019-07-19 15:02:08 -07:00
mirror_pad_test.cc
Optimize mirror_pad
2019-03-14 14:11:52 -07:00
mirror_pad.cc
Add gemmlowp label for mirrorpad op
2019-07-12 07:57:58 -07:00
mul_test.cc
Support quantized Mul with multiplier>1, even if that's a pathological case.
2019-08-01 10:15:35 -07:00
mul.cc
Support quantized Mul with multiplier>1, even if that's a pathological case.
2019-08-01 10:15:35 -07:00
neg_test.cc
Refactor TFLite kernel tests to use a common main
2019-05-22 09:21:06 -07:00
neg.cc
…
one_hot_test.cc
Refactor TFLite kernel tests to use a common main
2019-05-22 09:21:06 -07:00
one_hot.cc
…
op_macros.h
Work around ClangTidy bug: it wrongly reports that assert(false)
2019-06-03 07:04:15 -07:00
optional_tensor_test.cc
…
pack_test.cc
Refactor TFLite kernel tests to use a common main
2019-05-22 09:21:06 -07:00
pack.cc
Add negative axis support to Pack.
2019-03-19 14:21:36 -07:00
pad_test.cc
Refactor TFLite kernel tests to use a common main
2019-05-22 09:21:06 -07:00
pad.cc
Create int8 pad and pad2.
2019-02-13 18:53:57 -08:00
padding.h
Reuse ComputePaddingHeightWidth, also support dilation != 1 case.
2019-04-22 22:48:22 -07:00
pooling_test.cc
Fix several kernel unit tests for NNAPI execution
2019-05-29 11:13:04 -07:00
pooling.cc
Reuse ComputePaddingHeightWidth, also support dilation != 1 case.
2019-04-22 22:48:22 -07:00
pow_test.cc
Add POW op support to the NNAPI delegate
2019-07-02 10:02:21 -07:00
pow.cc
Optimize pow for single integer exponent case.
2019-07-01 11:21:37 -07:00
quant_basic_lstm_test.cc
Add delegate support for QUANTIZED_16BIT_LSTM
2019-07-25 03:19:35 -07:00
quantize_test.cc
Optimize requantize methods for int8 to int8 and uint8 to uint8
2019-08-04 22:11:37 -07:00
quantize.cc
Add int16 support to Quant.
2019-07-17 07:28:19 -07:00
range_test.cc
Refactor TFLite kernel tests to use a common main
2019-05-22 09:21:06 -07:00
range.cc
[2] Review comments handled
2019-03-27 11:14:51 +05:30
rank_test.cc
Refactor TFLite kernel tests to use a common main
2019-05-22 09:21:06 -07:00
rank.cc
Implement Rank.
2019-02-14 18:53:51 -08:00
read_variable.cc
Prorotype TFLite resource variables
2019-07-25 12:10:10 -07:00
reduce_test.cc
Create int8 sum Op.
2019-06-21 07:59:26 -07:00
reduce.cc
1. Remove all references to tflite::cpu_backend_support as we no longer do reference-counting on the cpu backend context object and GetFromContext is moved to CpuBackendContext class as a static member function.
2019-07-22 21:27:09 -07:00
register_ref.cc
NEON-optimized HardSwish
2019-06-18 07:52:25 -07:00
register_ref.h
Make tflite_driver able to run single op model with reference kernels.
2018-12-19 20:10:39 -08:00
register.cc
Add bool support for Gather.
2019-08-01 18:19:06 -07:00
register.h
…
reshape_test.cc
Change SingleOpModel::Invoke() to ASSERT rather than CHECK
2019-06-05 11:55:43 -07:00
reshape.cc
internal cleanup functions
2019-08-05 19:25:32 -07:00
resize_bilinear_test.cc
Support new quantization scheme in NNAPI delegate
2019-07-29 17:36:31 -07:00
resize_bilinear.cc
Add sanity check for resize-bilinear input shape.
2019-04-28 01:06:23 -07:00
resize_nearest_neighbor_test.cc
Refactor TFLite kernel tests to use a common main
2019-05-22 09:21:06 -07:00
resize_nearest_neighbor.cc
TFLite: Enhance input check for ResizeNearestNeghbor
2019-03-20 11:50:59 -07:00
reverse_sequence_test.cc
Refactor TFLite kernel tests to use a common main
2019-05-22 09:21:06 -07:00
reverse_sequence.cc
Implement Reverse_Sequence.
2019-02-22 22:10:21 -08:00
reverse_test.cc
Refactor TFLite kernel tests to use a common main
2019-05-22 09:21:06 -07:00
reverse.cc
Support Reverse_v2
2019-01-22 20:45:25 -08:00
rfft2d_test.cc
Implement Rfft2d as a custom ops.
2019-06-12 13:27:42 -07:00
rfft2d.cc
Implement Rfft2d as a custom ops.
2019-06-12 13:27:42 -07:00
round_test.cc
Refactor TFLite kernel tests to use a common main
2019-05-22 09:21:06 -07:00
round.cc
PR #24918: Tflite round
2019-05-06 15:42:21 -07:00
select_test.cc
Refactor TFLite kernel tests to use a common main
2019-05-22 09:21:06 -07:00
select.cc
Update select to support int8.
2019-02-07 09:59:01 -08:00
shape_test.cc
Refactor TFLite kernel tests to use a common main
2019-05-22 09:21:06 -07:00
shape.cc
…
skip_gram_test.cc
Refactor TFLite kernel tests to use a common main
2019-05-22 09:21:06 -07:00
skip_gram.cc
…
slice_test.cc
Add tests for the slice op with const inputs
2019-07-03 07:05:05 -07:00
slice.cc
Fix TensorFlow Lite windows builds
2019-07-26 15:08:16 -07:00
softmax_test.cc
Support running TFLite kernel unit tests with NNAPI
2019-05-17 11:56:13 -07:00
space_to_batch_nd_test.cc
Refactor TFLite kernel tests to use a common main
2019-05-22 09:21:06 -07:00
space_to_batch_nd.cc
Create int8 space to batch ND.
2019-02-11 12:58:42 -08:00
space_to_depth_test.cc
Refactor TFLite kernel tests to use a common main
2019-05-22 09:21:06 -07:00
space_to_depth.cc
Create int8 space to depth.
2019-02-12 11:36:49 -08:00
sparse_to_dense_test.cc
Merge pull request #29830 from jackwish:lite/op/sparse2dense
2019-07-10 21:17:55 -07:00
sparse_to_dense.cc
Merge pull request #29830 from jackwish:lite/op/sparse2dense
2019-07-10 21:17:55 -07:00
split_test.cc
Refactor TFLite kernel tests to use a common main
2019-05-22 09:21:06 -07:00
split_v_test.cc
Merge pull request #28376 from joshbeal:lite_splitv
2019-06-27 15:25:43 -07:00
split_v.cc
Merge pull request #28376 from joshbeal:lite_splitv
2019-06-27 15:25:43 -07:00
split.cc
Lite: Split Op Axis Validation Added
2019-03-13 17:09:00 +05:30
squared_difference_test.cc
Refactor TFLite kernel tests to use a common main
2019-05-22 09:21:06 -07:00
squared_difference.cc
Add gemmlowp label for SquaredDifference and Sum ops
2019-07-15 17:23:54 -07:00
squeeze_test.cc
Refactor TFLite kernel tests to use a common main
2019-05-22 09:21:06 -07:00
squeeze.cc
…
strided_slice_test.cc
Refactor TFLite kernel tests to use a common main
2019-05-22 09:21:06 -07:00
strided_slice.cc
Qualify calls to some functions from <cmath>.
2019-03-18 21:29:37 -07:00
sub_test.cc
Support running TFLite kernel unit tests with NNAPI
2019-05-17 11:56:13 -07:00
sub.cc
Move USE_NEON declaration to cpu_check.h, include that header everywhere USE_NEON is used.
2019-06-19 10:47:12 -07:00
subgraph_test_util_test.cc
Merge pull request #26640 from joyalbin:add_dynamic_pad_test
2019-03-15 15:46:04 -07:00
subgraph_test_util.cc
Graduate TFLite control flow ops from experimental to builtin
2019-07-20 15:51:38 -07:00
subgraph_test_util.h
Initial implementation of WHILE op
2019-02-08 16:48:27 -08:00
svdf_test.cc
Refactor TFLite kernel tests to use a common main
2019-05-22 09:21:06 -07:00
svdf.cc
Use GetTensorData instead of directly accessing tensor->data
2019-08-02 15:57:55 -07:00
test_main.cc
Fix command-line parsing for --use_nnapi
2019-05-22 13:23:37 -07:00
test_util_test.cc
…
test_util.cc
Changes to use delegate after tensor scales are known.
2019-08-02 15:19:23 -07:00
test_util.h
Changes to use delegate after tensor scales are known.
2019-08-02 15:19:23 -07:00
tile_test.cc
Add tile op tests with const input
2019-06-05 07:28:52 -07:00
tile.cc
Fix TensorFlow Lite windows builds
2019-07-26 15:08:16 -07:00
topk_v2_test.cc
Add topK_v2 support to NN API delegate
2019-06-13 03:06:33 -07:00
topk_v2.cc
Add int8 support for topk_v2.
2019-02-06 18:00:16 -08:00
transpose_conv_test.cc
Add tests for transpose convolution with constant inputs
2019-07-17 05:18:17 -07:00
transpose_conv.cc
1. Remove all references to tflite::cpu_backend_support as we no longer do reference-counting on the cpu backend context object and GetFromContext is moved to CpuBackendContext class as a static member function.
2019-07-22 21:27:09 -07:00
transpose_test.cc
Support running TFLite kernel unit tests with NNAPI
2019-05-17 11:56:13 -07:00
transpose.cc
Add Bool support in transpose op.
2019-07-15 23:04:14 -07:00
unidirectional_sequence_lstm_test.cc
Refactor TFLite kernel tests to use a common main
2019-05-22 09:21:06 -07:00
unidirectional_sequence_lstm.cc
Add layer norm to unidirectional lstm. With this change, unidirectional lstm takes 24 inputs with or without layer norm. The 20 input case is only kept for backward compatibility.
2019-05-17 16:01:13 -07:00
unidirectional_sequence_rnn_test.cc
Refactor TFLite kernel tests to use a common main
2019-05-22 09:21:06 -07:00
unidirectional_sequence_rnn.cc
Merge pull request #26655 from Dayananda-V:tflite_hybrid_op
2019-03-18 22:06:32 -07:00
unique_test.cc
Add Unique Op implementation
2019-01-07 11:47:25 -08:00
unique.cc
Add Unique Op to MLIR converter.
2019-07-15 17:42:59 -07:00
unpack_test.cc
Unpack op now supports vector -> scalar
2019-07-26 17:16:22 -07:00
unpack.cc
Unpack op now supports vector -> scalar
2019-07-26 17:16:22 -07:00
variable_ops_test.cc
Prorotype TFLite resource variables
2019-07-25 12:10:10 -07:00
where_test.cc
Refactor TFLite kernel tests to use a common main
2019-05-22 09:21:06 -07:00
where.cc
Implement tf.where.
2019-02-17 16:05:57 -08:00
while_test.cc
Fix invoking while multiple times when body is dynamic.
2019-07-10 12:42:46 -07:00
while.cc
Graduate TFLite control flow ops from experimental to builtin
2019-07-20 15:51:38 -07:00
zeros_like_test.cc
Change SingleOpModel::Invoke() to ASSERT rather than CHECK
2019-06-05 11:55:43 -07:00
zeros_like.cc
…