| a7ca183f502c1de952e5783f1144380970d86378 |
16-May-2018 |
Mika Raento <mikie@google.com> |
Validate strided slice output dims in cpu executor
This fixes a buffer overflow in the strided slice cpu implementation.
The strided slice operation has an option to "shrink" an axis, meaning
that a dimension of the input is removed from the output. This requires
the corresponding slice to have size 1. The size constraint was not
checked and unit tests were using a size of 2, leading to an overflow
when executing the op.
TODO: add a test for the validation, b/79856511. I want to get this fix
in ASAP as currently the tests crash in the continuous build so we will
not see other failures.
Bug: 79705516
Test: NeuralNetworksTest_static/*strided_slice* with address sanitizer on marlin-userdebug
Test: NeuralNetworksTest_static/*strided_slice* on marlin-userdebug
Merged-In: I4db9c7460d1c4d943ba7ced36b4d8801d0e0ec99
(cherry picked from commit 7bdc10967deee216043f2dd8d79bb352776c66b7)
Change-Id: Icbc0f6eb52fe8fd179c25312aff9d2374ad58679
trided_slice_float_11.model.cpp
trided_slice_float_11_relaxed.model.cpp
trided_slice_qaunt8_11.model.cpp
|
| 3c2272e1cf69ef14d1a5dce34541bfd492272f69 |
18-Apr-2018 |
Michael Butler <butlermichael@google.com> |
Create _relaxed tests for newly added tests -- generated tests
New tests were added to V1_0 and V1_1 specs, but they did not have a
"_relaxed" test in V1_1.
This CL generates the CTS and VTS tests as a followup CL.
Bug: 76434855
Test: mma
Test: NeuralNetworksTest_static
Test: VtsHalNeuralnetworksV1_0TargetTest
Test: VtsHalNeuralnetworksV1_1TargetTest
Change-Id: Id3703db7091583de19e22b82eab75e88da785428
vg_pool_float_5_relaxed.model.cpp
atch_to_space_float_1_relaxed.model.cpp
atch_to_space_relaxed.model.cpp
onv_float_2_relaxed.model.cpp
epthwise_conv2d_float_2_relaxed.model.cpp
epthwise_conv2d_float_large_2_relaxed.model.cpp
epthwise_conv2d_float_large_2_weights_as_inputs_relaxed.model.cpp
epthwise_conv2d_float_relaxed.model.cpp
epthwise_conv2d_float_weights_as_inputs_relaxed.model.cpp
equantize_relaxed.model.cpp
iv_broadcast_float_relaxed.model.cpp
iv_relaxed.model.cpp
ully_connected_float_2_relaxed.model.cpp
ully_connected_float_4d_simple.model.cpp
ully_connected_float_4d_simple_relaxed.model.cpp
ully_connected_float_large_relaxed.model.cpp
ully_connected_float_large_weights_as_inputs_relaxed.model.cpp
2_normalization_2_relaxed.model.cpp
2_pool_float_2_relaxed.model.cpp
2_pool_float_large_relaxed.model.cpp
ax_pool_float_4_relaxed.model.cpp
ean_float_1_relaxed.model.cpp
ean_float_2_relaxed.model.cpp
ean_relaxed.model.cpp
ad_float_1_relaxed.model.cpp
ad_relaxed.model.cpp
elu_float_2_relaxed.model.cpp
esize_bilinear_2_relaxed.model.cpp
pace_to_batch_float_1_relaxed.model.cpp
pace_to_batch_float_2_relaxed.model.cpp
pace_to_batch_float_3_relaxed.model.cpp
pace_to_batch_relaxed.model.cpp
queeze_float_1_relaxed.model.cpp
queeze_relaxed.model.cpp
trided_slice_float_10_relaxed.model.cpp
trided_slice_float_11_relaxed.model.cpp
trided_slice_float_1_relaxed.model.cpp
trided_slice_float_2_relaxed.model.cpp
trided_slice_float_3_relaxed.model.cpp
trided_slice_float_4_relaxed.model.cpp
trided_slice_float_5_relaxed.model.cpp
trided_slice_float_6_relaxed.model.cpp
trided_slice_float_7_relaxed.model.cpp
trided_slice_float_8_relaxed.model.cpp
trided_slice_float_9_relaxed.model.cpp
trided_slice_relaxed.model.cpp
ub_broadcast_float_relaxed.model.cpp
ub_relaxed.model.cpp
ranspose_float_1_relaxed.model.cpp
ranspose_relaxed.model.cpp
|
| 6d78cf0e572d4117b3011bcf0ab768c47dc58031 |
12-Apr-2018 |
Mika Raento <mikie@google.com> |
Clarify Fully Connected Op comments and naming
- Make the reshaping of input clearer in header comment
- Make variable names in implementation match header comments
- Add a test for the case where first dimension is not the batch size
Bug: 74851002
Test: NeuralNetworksTest
Change-Id: Ic914d5b61ef162f2be04834bcd128208f13a627f
ully_connected_float_4d_simple.model.cpp
|
| b66b5abcef40e88df6ac18ce7890372192102323 |
02-Apr-2018 |
Miao Wang <miaowang@google.com> |
Finalize STRIDED_SLICE
- Add documentation for begin_mask, end_mask, shrink_axis_mask.
- Add support for shrink_axis_mask, and corresponding tests.
- Add validation tests for STRIDED_SLICE.
- Updated existing tests and re-generated CTS and VTS tests.
Bug: 73661082
Test: NeuralNetworksTests_static
Change-Id: I49ad2ffcb3b7496ee1225618833ba324bb01c5b5
trided_slice.model.cpp
trided_slice_float_1.model.cpp
trided_slice_float_10.model.cpp
trided_slice_float_11.model.cpp
trided_slice_float_2.model.cpp
trided_slice_float_3.model.cpp
trided_slice_float_4.model.cpp
trided_slice_float_5.model.cpp
trided_slice_float_6.model.cpp
trided_slice_float_7.model.cpp
trided_slice_float_8.model.cpp
trided_slice_float_9.model.cpp
trided_slice_qaunt8_10.model.cpp
trided_slice_qaunt8_11.model.cpp
trided_slice_quant8_1.model.cpp
trided_slice_quant8_2.model.cpp
trided_slice_quant8_3.model.cpp
trided_slice_quant8_4.model.cpp
trided_slice_quant8_5.model.cpp
trided_slice_quant8_6.model.cpp
trided_slice_quant8_7.model.cpp
trided_slice_quant8_8.model.cpp
trided_slice_quant8_9.model.cpp
|
| 400856736c50f2112b2b413b8f9ee78fb1d676ce |
29-Mar-2018 |
TreeHugger Robot <treehugger-gerrit@google.com> |
Merge changes Ida0350d2,If4b50c50 into pi-dev
* changes:
Use abort() instead of exit(1) for bad activation functions
Remove unused / unimplemented parameter "crops" for BATCH_TO_SPACE
|
| 97a2a4b72e275b8dc8e42867567347be7822c26a |
29-Mar-2018 |
TreeHugger Robot <treehugger-gerrit@google.com> |
Merge "Make NNAPI support BroadcastSub and BroadcastDiv." into pi-dev
|
| 0b385e7a66d125117cdc952a249f2fdb156e893d |
21-Mar-2018 |
Michael Butler <butlermichael@google.com> |
Add more validation to the HAL
This CL adds additional validation to the common validation library:
* checks scale of TENSOR_INT32
* fixes check for scale of TENSOR_QUANT8_ASYMM
* check for zeroPoint of TENSOR_QUANT8_ASYMM
* check for invalid 1.0 OperationType value
* adds operation argument number and type validation
This CL also makes some tangentially related changes, including:
* fixing sample driver validation
* changing mobilenet_224_* to use shared weights
Bug: 67828197
Test: mma
Test: VtsHalNeuralnetworksV1_0TargetTest
Test: VtsHalNeuralnetworksV1_1TargetTest
Change-Id: I706a968a7baf7f9942dd3b9260083514e0dafb7f
obilenet_224_gender_basic_fixed.model.cpp
obilenet_224_gender_basic_fixed_relaxed.model.cpp
|
| 540bafe7bdf4d10fd813c81730d824414b91bfaf |
27-Mar-2018 |
Michael Butler <butlermichael@google.com> |
Regenerate CTS and VTS test cases
This CL generates the tests after the spec file was changed:
* lsh_projection
* lstm
* rnn
* svdf
This CL is a followup from ag/3802074.
Bug: 67828197
Test: mma
Test: NeuralNetworksTest
Test: VtsHalNeuralnetworksV1_0TargetTest
Change-Id: I0c40ac0556ed0770e694ac7a62eed7c073a031f4
sh_projection_weights_as_inputs.model.cpp
sh_projection_weights_as_inputs_relaxed.model.cpp
stm.model.cpp
stm2.model.cpp
stm2_relaxed.model.cpp
stm2_state.model.cpp
stm2_state2.model.cpp
stm2_state2_relaxed.model.cpp
stm2_state_relaxed.model.cpp
stm3.model.cpp
stm3_relaxed.model.cpp
stm3_state.model.cpp
stm3_state2.model.cpp
stm3_state2_relaxed.model.cpp
stm3_state3.model.cpp
stm3_state3_relaxed.model.cpp
stm3_state_relaxed.model.cpp
stm_relaxed.model.cpp
stm_state.model.cpp
stm_state2.model.cpp
stm_state2_relaxed.model.cpp
stm_state_relaxed.model.cpp
nn.model.cpp
nn_relaxed.model.cpp
nn_state.model.cpp
nn_state_relaxed.model.cpp
vdf.model.cpp
vdf2.model.cpp
vdf2_relaxed.model.cpp
vdf_relaxed.model.cpp
vdf_state.model.cpp
vdf_state_relaxed.model.cpp
|
| 03fde839dc0bfa4f2ead0d60fc4fe80f55e52660 |
27-Mar-2018 |
Michael Butler <butlermichael@google.com> |
Fix resize_bilinear test height/width order
resize_bilinear_relaxed had its height and width parameters reversed.
This CL fixes this, and regenerates all CTS and VTS models.
Bug: 73964023
Test: mma
Test: NeuralNetworksTest
Test: VtsHalNeuralnetworksV1_0TargetTest
Test: VtsHalNeuralnetworksV1_1TargetTest
Change-Id: Ia5714eb5335327b25521184752542baa4c0a56fa
esize_bilinear_2.model.cpp
|
| a9ce0001ca1df32fde3f322c385eef8ce7279329 |
23-Mar-2018 |
Miao Wang <miaowang@google.com> |
Remove unused / unimplemented parameter "crops" for BATCH_TO_SPACE
Bug: 73661175
Test: mm
Test: NeuralNetworksTests
Change-Id: If4b50c50e0847ce36016f6c257aac3b9d418c7b2
atch_to_space.model.cpp
atch_to_space_float_1.model.cpp
atch_to_space_quant8_1.model.cpp
|
| 02e33fc0ef3dce28cee468d1fa4e37511ba1dd48 |
23-Mar-2018 |
Miao Wang <miaowang@google.com> |
Make NNAPI support BroadcastSub and BroadcastDiv.
- Make the CPU implementation able to handle BroadcastSub and
BroadcastDiv.
- Add corresponding tests.
Bug: 73661777
Test: mm
Test: NeuralNetworksTest and VTS tests
Change-Id: I0b7921b6f0e9f298e94dcaebe5cfe9ac580295a3
iv_broadcast_float.model.cpp
ub_broadcast_float.model.cpp
|
| d6d2b0acd3b0e6914a2ca23214548e210a90b19c |
14-Mar-2018 |
Michael Butler <butlermichael@google.com> |
Upgrade generated test cases
There were some changes to strided_slice_float_1.mod.py,
strided_slice_quant8_1.mod.py, and resize_bilinear.mod.py that did not
have a followup CL to update the corresponding generated test cases.
This CL updates those changes.
Additionally, some variables in mobilenet_224_gender_basic_fixed and
mobilenet_quantized had their order swapped.
Bug: 73964023
Bug: 74000148
Test: mma
Test: NeuralNetworksTest
Test: VtsHalNeuralnetworksV1_0TargetTest
Change-Id: I78a32f98fc8e7166094d861675e59e3a0f61edac
trided_slice_float_1.model.cpp
trided_slice_quant8_1.model.cpp
|
| 3581e149f52fb810ecffe087cfec82cd55ccec22 |
26-Feb-2018 |
Miao Wang <miaowang@google.com> |
Generates new V1_0 tests from the newly added specs.
Test: mm
Test: NeuralNetworks tests pass on marlin
Change-Id: I8f964ee6fa090b4fb3c1c6ee7f9a8c1fdac78ae7
vg_pool_float_5.model.cpp
vg_pool_quant8_5.model.cpp
onv_float_2.model.cpp
onv_quant8_2.model.cpp
epthwise_conv2d_float_2.model.cpp
epthwise_conv2d_quant8_2.model.cpp
ully_connected_float_2.model.cpp
ully_connected_quant8_2.model.cpp
2_normalization_2.model.cpp
2_pool_float_2.model.cpp
ax_pool_float_4.model.cpp
ax_pool_quant8_4.model.cpp
esize_bilinear_2.model.cpp
|
| 70f401e3df8435bbb750e77a2341f1feec719556 |
23-Feb-2018 |
Miao Wang <miaowang@google.com> |
Add quantized MEAN tests
Bug: 63911257
Test: mm
Test: NeuralNetworksTests and VTS pass on walleye
Change-Id: Ia2374f5678202f1ae6aa2116f3eea1c739abcb50
ean_quant8_1.model.cpp
ean_quant8_2.model.cpp
|
| e8e5d34c4159532eb324df393c2c752a508bced1 |
23-Feb-2018 |
Miao Wang <miaowang@google.com> |
Generate relaxed precision tests for CTS and VTS.
Bug: 63911257
Test: mm
Test: CTS and VTS pass on walleye
Change-Id: I62655872423e55201e94c05e2cd9da86815d4b26
dd.model.cpp
dd_broadcast_quant8.model.cpp
dd_quant8.model.cpp
dd_relaxed.model.cpp
vg_pool_float_1.model.cpp
vg_pool_float_1_relaxed.model.cpp
vg_pool_float_2.model.cpp
vg_pool_float_2_relaxed.model.cpp
vg_pool_float_3.model.cpp
vg_pool_float_3_relaxed.model.cpp
vg_pool_float_4.model.cpp
vg_pool_float_4_relaxed.model.cpp
vg_pool_quant8_1.model.cpp
vg_pool_quant8_2.model.cpp
vg_pool_quant8_3.model.cpp
vg_pool_quant8_4.model.cpp
atch_to_space.model.cpp
atch_to_space_float_1.model.cpp
atch_to_space_quant8_1.model.cpp
oncat_float_1.model.cpp
oncat_float_1_relaxed.model.cpp
oncat_float_2.model.cpp
oncat_float_2_relaxed.model.cpp
oncat_float_3.model.cpp
oncat_float_3_relaxed.model.cpp
oncat_quant8_1.model.cpp
oncat_quant8_2.model.cpp
oncat_quant8_3.model.cpp
onv_1_h3_w2_SAME.model.cpp
onv_1_h3_w2_SAME_relaxed.model.cpp
onv_1_h3_w2_VALID.model.cpp
onv_1_h3_w2_VALID_relaxed.model.cpp
onv_3_h3_w2_SAME.model.cpp
onv_3_h3_w2_SAME_relaxed.model.cpp
onv_3_h3_w2_VALID.model.cpp
onv_3_h3_w2_VALID_relaxed.model.cpp
onv_float.model.cpp
onv_float_channels.model.cpp
onv_float_channels_relaxed.model.cpp
onv_float_channels_weights_as_inputs.model.cpp
onv_float_channels_weights_as_inputs_relaxed.model.cpp
onv_float_large.model.cpp
onv_float_large_relaxed.model.cpp
onv_float_large_weights_as_inputs.model.cpp
onv_float_large_weights_as_inputs_relaxed.model.cpp
onv_float_relaxed.model.cpp
onv_float_weights_as_inputs.model.cpp
onv_float_weights_as_inputs_relaxed.model.cpp
onv_quant8.model.cpp
onv_quant8_channels.model.cpp
onv_quant8_channels_weights_as_inputs.model.cpp
onv_quant8_large.model.cpp
onv_quant8_large_weights_as_inputs.model.cpp
onv_quant8_overflow.model.cpp
onv_quant8_overflow_weights_as_inputs.model.cpp
onv_quant8_weights_as_inputs.model.cpp
epth_to_space_float_1.model.cpp
epth_to_space_float_1_relaxed.model.cpp
epth_to_space_float_2.model.cpp
epth_to_space_float_2_relaxed.model.cpp
epth_to_space_float_3.model.cpp
epth_to_space_float_3_relaxed.model.cpp
epth_to_space_quant8_1.model.cpp
epth_to_space_quant8_2.model.cpp
epthwise_conv.model.cpp
epthwise_conv2d_float.model.cpp
epthwise_conv2d_float_large.model.cpp
epthwise_conv2d_float_large_2.model.cpp
epthwise_conv2d_float_large_2_weights_as_inputs.model.cpp
epthwise_conv2d_float_large_relaxed.model.cpp
epthwise_conv2d_float_large_weights_as_inputs.model.cpp
epthwise_conv2d_float_large_weights_as_inputs_relaxed.model.cpp
epthwise_conv2d_float_weights_as_inputs.model.cpp
epthwise_conv2d_quant8.model.cpp
epthwise_conv2d_quant8_large.model.cpp
epthwise_conv2d_quant8_large_weights_as_inputs.model.cpp
epthwise_conv2d_quant8_weights_as_inputs.model.cpp
epthwise_conv_relaxed.model.cpp
equantize.model.cpp
iv.model.cpp
mbedding_lookup.model.cpp
mbedding_lookup_relaxed.model.cpp
loor.model.cpp
loor_relaxed.model.cpp
ully_connected_float.model.cpp
ully_connected_float_large.model.cpp
ully_connected_float_large_weights_as_inputs.model.cpp
ully_connected_float_relaxed.model.cpp
ully_connected_float_weights_as_inputs.model.cpp
ully_connected_float_weights_as_inputs_relaxed.model.cpp
ully_connected_quant8.model.cpp
ully_connected_quant8_large.model.cpp
ully_connected_quant8_large_weights_as_inputs.model.cpp
ully_connected_quant8_weights_as_inputs.model.cpp
ashtable_lookup_float.model.cpp
ashtable_lookup_float_relaxed.model.cpp
ashtable_lookup_quant8.model.cpp
2_normalization.model.cpp
2_normalization_large.model.cpp
2_normalization_large_relaxed.model.cpp
2_normalization_relaxed.model.cpp
2_pool_float.model.cpp
2_pool_float_large.model.cpp
2_pool_float_relaxed.model.cpp
ocal_response_norm_float_1.model.cpp
ocal_response_norm_float_1_relaxed.model.cpp
ocal_response_norm_float_2.model.cpp
ocal_response_norm_float_2_relaxed.model.cpp
ocal_response_norm_float_3.model.cpp
ocal_response_norm_float_3_relaxed.model.cpp
ocal_response_norm_float_4.model.cpp
ocal_response_norm_float_4_relaxed.model.cpp
ogistic_float_1.model.cpp
ogistic_float_1_relaxed.model.cpp
ogistic_float_2.model.cpp
ogistic_float_2_relaxed.model.cpp
ogistic_quant8_1.model.cpp
ogistic_quant8_2.model.cpp
sh_projection.model.cpp
sh_projection_2.model.cpp
sh_projection_2_relaxed.model.cpp
sh_projection_relaxed.model.cpp
sh_projection_weights_as_inputs.model.cpp
sh_projection_weights_as_inputs_relaxed.model.cpp
stm.model.cpp
stm2.model.cpp
stm2_relaxed.model.cpp
stm2_state.model.cpp
stm2_state2.model.cpp
stm2_state2_relaxed.model.cpp
stm2_state_relaxed.model.cpp
stm3.model.cpp
stm3_relaxed.model.cpp
stm3_state.model.cpp
stm3_state2.model.cpp
stm3_state2_relaxed.model.cpp
stm3_state3.model.cpp
stm3_state3_relaxed.model.cpp
stm3_state_relaxed.model.cpp
stm_relaxed.model.cpp
stm_state.model.cpp
stm_state2.model.cpp
stm_state2_relaxed.model.cpp
stm_state_relaxed.model.cpp
ax_pool_float_1.model.cpp
ax_pool_float_1_relaxed.model.cpp
ax_pool_float_2.model.cpp
ax_pool_float_2_relaxed.model.cpp
ax_pool_float_3.model.cpp
ax_pool_float_3_relaxed.model.cpp
ax_pool_quant8_1.model.cpp
ax_pool_quant8_2.model.cpp
ax_pool_quant8_3.model.cpp
ean.model.cpp
ean_float_1.model.cpp
ean_float_2.model.cpp
obilenet_224_gender_basic_fixed.model.cpp
obilenet_224_gender_basic_fixed_relaxed.model.cpp
obilenet_quantized.model.cpp
ul.model.cpp
ul_broadcast_quant8.model.cpp
ul_quant8.model.cpp
ul_relaxed.model.cpp
ul_relu.model.cpp
ul_relu_relaxed.model.cpp
ad.model.cpp
ad_float_1.model.cpp
elu1_float_1.model.cpp
elu1_float_1_relaxed.model.cpp
elu1_float_2.model.cpp
elu1_float_2_relaxed.model.cpp
elu1_quant8_1.model.cpp
elu1_quant8_2.model.cpp
elu6_float_1.model.cpp
elu6_float_1_relaxed.model.cpp
elu6_float_2.model.cpp
elu6_float_2_relaxed.model.cpp
elu6_quant8_1.model.cpp
elu6_quant8_2.model.cpp
elu_float_1.model.cpp
elu_float_1_relaxed.model.cpp
elu_float_2.model.cpp
elu_quant8_1.model.cpp
elu_quant8_2.model.cpp
eshape.model.cpp
eshape_quant8.model.cpp
eshape_quant8_weights_as_inputs.model.cpp
eshape_relaxed.model.cpp
eshape_weights_as_inputs.model.cpp
eshape_weights_as_inputs_relaxed.model.cpp
esize_bilinear.model.cpp
esize_bilinear_relaxed.model.cpp
nn.model.cpp
nn_relaxed.model.cpp
nn_state.model.cpp
nn_state_relaxed.model.cpp
oftmax_float_1.model.cpp
oftmax_float_1_relaxed.model.cpp
oftmax_float_2.model.cpp
oftmax_float_2_relaxed.model.cpp
oftmax_quant8_1.model.cpp
oftmax_quant8_2.model.cpp
pace_to_batch.model.cpp
pace_to_batch_float_1.model.cpp
pace_to_batch_float_2.model.cpp
pace_to_batch_float_3.model.cpp
pace_to_batch_quant8_1.model.cpp
pace_to_batch_quant8_2.model.cpp
pace_to_batch_quant8_3.model.cpp
pace_to_depth_float_1.model.cpp
pace_to_depth_float_1_relaxed.model.cpp
pace_to_depth_float_2.model.cpp
pace_to_depth_float_2_relaxed.model.cpp
pace_to_depth_float_3.model.cpp
pace_to_depth_float_3_relaxed.model.cpp
pace_to_depth_quant8_1.model.cpp
pace_to_depth_quant8_2.model.cpp
queeze.model.cpp
queeze_float_1.model.cpp
queeze_quant8_1.model.cpp
trided_slice.model.cpp
trided_slice_float_1.model.cpp
trided_slice_float_10.model.cpp
trided_slice_float_2.model.cpp
trided_slice_float_3.model.cpp
trided_slice_float_4.model.cpp
trided_slice_float_5.model.cpp
trided_slice_float_6.model.cpp
trided_slice_float_7.model.cpp
trided_slice_float_8.model.cpp
trided_slice_float_9.model.cpp
trided_slice_qaunt8_10.model.cpp
trided_slice_quant8_1.model.cpp
trided_slice_quant8_2.model.cpp
trided_slice_quant8_3.model.cpp
trided_slice_quant8_4.model.cpp
trided_slice_quant8_5.model.cpp
trided_slice_quant8_6.model.cpp
trided_slice_quant8_7.model.cpp
trided_slice_quant8_8.model.cpp
trided_slice_quant8_9.model.cpp
ub.model.cpp
vdf.model.cpp
vdf2.model.cpp
vdf2_relaxed.model.cpp
vdf_relaxed.model.cpp
vdf_state.model.cpp
vdf_state_relaxed.model.cpp
anh.model.cpp
anh_relaxed.model.cpp
ranspose.model.cpp
ranspose_float_1.model.cpp
ranspose_quant8_1.model.cpp
|
| 65db490301bb2e35864fb0f2aef67fa8c4be2b88 |
20-Feb-2018 |
Miao Wang <miaowang@google.com> |
Add more tests for v1.1 ops
Bug: 63911257
Test: mm
Test: NeuralNetworksTests and VTS tests pass
Change-Id: Ia9674e5bed390fd2507c6c70f1a37737f7733ae9
atch_to_space_float_1.model.cpp
atch_to_space_quant8_1.model.cpp
ean_float_1.model.cpp
ean_float_2.model.cpp
ad_float_1.model.cpp
pace_to_batch_float_1.model.cpp
pace_to_batch_float_2.model.cpp
pace_to_batch_float_3.model.cpp
pace_to_batch_quant8_1.model.cpp
pace_to_batch_quant8_2.model.cpp
pace_to_batch_quant8_3.model.cpp
queeze_float_1.model.cpp
queeze_quant8_1.model.cpp
trided_slice_float_1.model.cpp
trided_slice_float_10.model.cpp
trided_slice_float_2.model.cpp
trided_slice_float_3.model.cpp
trided_slice_float_4.model.cpp
trided_slice_float_5.model.cpp
trided_slice_float_6.model.cpp
trided_slice_float_7.model.cpp
trided_slice_float_8.model.cpp
trided_slice_float_9.model.cpp
trided_slice_qaunt8_10.model.cpp
trided_slice_quant8_1.model.cpp
trided_slice_quant8_2.model.cpp
trided_slice_quant8_3.model.cpp
trided_slice_quant8_4.model.cpp
trided_slice_quant8_5.model.cpp
trided_slice_quant8_6.model.cpp
trided_slice_quant8_7.model.cpp
trided_slice_quant8_8.model.cpp
trided_slice_quant8_9.model.cpp
ranspose_float_1.model.cpp
ranspose_quant8_1.model.cpp
|
| e78b89d46b7e534b749ace863886571a840e83e2 |
16-Feb-2018 |
Michael Butler <butlermichael@google.com> |
Update generated VTS test for mobilenet_quantized
The specification file for mobilenet_quantized was changed
to use a real mobilenet quantized model with a real image
as input. However, the vts test was not regenerated from
that source file when the corresponding change to the spec
file was made. This CL regenerates the vts test for
mobilenet_quantized to keep the two in sync.
Test: mm
Test: VtsHalNeuralnetworksV1_0TargetTest
Change-Id: I2dc7369755dedfafa0fcc1c34611135f6026bd48
obilenet_quantized.model.cpp
|
| 5b0a7d939acba4c70027ea2798e7bbe054e77707 |
06-Feb-2018 |
Miao Wang <miaowang@google.com> |
Create NNAPI v1.1 VTS tests.
- Split specs files into corresponding subfolders with versioning.
- Update scripts to be able to handle versioned specs.
- Generate all_generated_$VER_vts_tests.cpp for VTS.
- Generate V1_1 op tests for VTS.
Bug: 63911257
Test: mm
Test: NNAPI VTS tests pass on v1.0 and v1.1 sample devices.
Change-Id: I1980561b69a463f4284f539cc252d298aa99a34c
atch_to_space.model.cpp
iv.model.cpp
ean.model.cpp
ad.model.cpp
pace_to_batch.model.cpp
queeze.model.cpp
trided_slice.model.cpp
ub.model.cpp
vdf2.model.cpp
ranspose.model.cpp
|
| 021304a703f6e21916e714979ba7c0427d286804 |
16-Nov-2017 |
Jean-Luc Brouillet <jeanluc@google.com> |
Merge "Improve validation of the HAL structures." into oc-mr1-nn-dev
|
| c5e342b62bf25a98d15ae28ee97916b274296e94 |
12-Oct-2017 |
Jean-Luc Brouillet <jeanluc@google.com> |
Improve validation of the HAL structures.
Add a lot of validation of the structures passed in the HAL.
Particularly important are those that operands & arguments
don't try to reach out of their memory blocks.
Also grabs a few missing generated tests.
Bug: 67828197
Test: System tests & VTS tests.
Change-Id: I2edf6219fc660fab7c5b6a73e7a9cb8a358fb29b
epth_to_space_quant8_2.model.cpp
pace_to_depth_quant8_2.model.cpp
|
| dd5cd91c89da2e5f44a8fd422e46ff24ffb872e0 |
10-Nov-2017 |
I-Jui (Ray) Sung <ijsung@google.com> |
Added quantized mobilenet to VTS and refactoring generators
Quantized mobilenet was not able to run correctly in VTS as it has large
weights that couldn't be passed through HIDL without using shared memory.
Updated the test generator so it can optionally pass the weights via
shared memory (by setting Configuration.use_shm_for_weights to True).
Separated the VTS backend from the main test generator as a code
cleanup.
Also removed the blacklisting code in generate_vts_test.sh and checked
in the generated VTS model for mobilenet_quanitzed.
Test: unit tests and VTS on Sailfish
Bug: 67753688
Bug: 67603060
Change-Id: I891422461d3c2a0e7ca117bf15dc0a8cbd5e8785
obilenet_quantized.model.cpp
|
| 4062279c74f2c995349b7ff04f22feb89eb81e47 |
19-Oct-2017 |
I-Jui (Ray) Sung <ijsung@google.com> |
Slicing a given model up to the given first N operations
Also regenerated all CTS, VTS and unit tests, as a test harness function
is renamed.
Test: test.py and sliced mobilenet_quantized up to first 20 ops
Test: unit tests and VTS on Sailfish
Bug: 67781676
Change-Id: I68dcdbda17f8bedc98b1da573a8f884da0d5722d
onv_1_h3_w2_SAME.model.cpp
onv_1_h3_w2_VALID.model.cpp
onv_3_h3_w2_SAME.model.cpp
onv_3_h3_w2_VALID.model.cpp
epthwise_conv.model.cpp
stm.model.cpp
stm2.model.cpp
stm2_state.model.cpp
stm2_state2.model.cpp
stm3.model.cpp
stm3_state.model.cpp
stm3_state2.model.cpp
stm3_state3.model.cpp
stm_state.model.cpp
stm_state2.model.cpp
obilenet_224_gender_basic_fixed.model.cpp
nn_state.model.cpp
vdf_state.model.cpp
|
| 2467b9c2956293324de8557cf341a7cae83c591d |
12-Oct-2017 |
I-Jui (Ray) Sung <ijsung@google.com> |
Update generated VTS tests
Regenerate VTS tests after test generator updates in ag/3050465.
Bug: 67607461
Test: VtsHalNeuralnetworksV1_0TargetTest on Angler with sample driver
Change-Id: I23898061809fe371049337d8c103c05427a728ee
vg_pool_float_1.model.cpp
vg_pool_float_2.model.cpp
vg_pool_float_3.model.cpp
vg_pool_float_4.model.cpp
vg_pool_quant8_1.model.cpp
vg_pool_quant8_2.model.cpp
vg_pool_quant8_3.model.cpp
vg_pool_quant8_4.model.cpp
onv_float.model.cpp
onv_float_channels.model.cpp
onv_float_channels_weights_as_inputs.model.cpp
onv_float_large.model.cpp
onv_float_large_weights_as_inputs.model.cpp
onv_float_weights_as_inputs.model.cpp
onv_quant8.model.cpp
onv_quant8_channels.model.cpp
onv_quant8_channels_weights_as_inputs.model.cpp
onv_quant8_large.model.cpp
onv_quant8_large_weights_as_inputs.model.cpp
onv_quant8_overflow.model.cpp
onv_quant8_overflow_weights_as_inputs.model.cpp
onv_quant8_weights_as_inputs.model.cpp
epthwise_conv2d_float.model.cpp
epthwise_conv2d_float_large.model.cpp
epthwise_conv2d_float_large_2.model.cpp
epthwise_conv2d_float_large_2_weights_as_inputs.model.cpp
epthwise_conv2d_float_large_weights_as_inputs.model.cpp
epthwise_conv2d_float_weights_as_inputs.model.cpp
epthwise_conv2d_quant8.model.cpp
epthwise_conv2d_quant8_large.model.cpp
epthwise_conv2d_quant8_large_weights_as_inputs.model.cpp
epthwise_conv2d_quant8_weights_as_inputs.model.cpp
mbedding_lookup.model.cpp
2_pool_float.model.cpp
2_pool_float_large.model.cpp
ax_pool_float_1.model.cpp
ax_pool_float_2.model.cpp
ax_pool_float_3.model.cpp
ax_pool_quant8_1.model.cpp
ax_pool_quant8_2.model.cpp
ax_pool_quant8_3.model.cpp
|
| 9949232ba3d5b1d95f5b3d9ea310b88b81ee1a45 |
07-Oct-2017 |
gfan <gfan@google.com> |
Added one more test spec for space_to_depth/depth_to_space
Bug: 63905942
Test: ran NeuralNetworksTest on marlin
Change-Id: I11df832b5e69479b7fe4e8beff9c0add2ed9992c
epth_to_space_float_3.model.cpp
pace_to_depth_float_3.model.cpp
|
| bee07f73a5f998a2dd6dc581e7776557c21f9684 |
09-Oct-2017 |
Miao Wang <miaowang@google.com> |
Force CpuExecutor validating user-provided model output operands.
- For operands with OperandLifeTime::MODEL_OUTPUT, the dimensions,
type, and other meta-data must match the output Shape calculated
from the operation preparation step.
- Fix the ill-defined tests caught by the added validation.
- Incidental changes: generated more tests from tests specs.
Bug: 67390841
Test: NeuralNetworksTests pass
Change-Id: I40d35db0f7a868feae773dbf7e12cf4bf5f5e275
onv_float.model.cpp
onv_float_channels.model.cpp
onv_float_channels_weights_as_inputs.model.cpp
onv_float_large.model.cpp
onv_float_large_weights_as_inputs.model.cpp
onv_float_weights_as_inputs.model.cpp
onv_quant8.model.cpp
onv_quant8_channels.model.cpp
onv_quant8_channels_weights_as_inputs.model.cpp
onv_quant8_large.model.cpp
onv_quant8_large_weights_as_inputs.model.cpp
onv_quant8_overflow.model.cpp
onv_quant8_overflow_weights_as_inputs.model.cpp
onv_quant8_weights_as_inputs.model.cpp
epth_to_space_quant8_1.model.cpp
epthwise_conv2d_float.model.cpp
epthwise_conv2d_float_large.model.cpp
epthwise_conv2d_float_large_2.model.cpp
epthwise_conv2d_float_large_2_weights_as_inputs.model.cpp
epthwise_conv2d_float_large_weights_as_inputs.model.cpp
epthwise_conv2d_float_weights_as_inputs.model.cpp
epthwise_conv2d_quant8.model.cpp
epthwise_conv2d_quant8_large.model.cpp
epthwise_conv2d_quant8_large_weights_as_inputs.model.cpp
epthwise_conv2d_quant8_weights_as_inputs.model.cpp
ully_connected_float.model.cpp
ully_connected_float_large.model.cpp
ully_connected_float_large_weights_as_inputs.model.cpp
ully_connected_float_weights_as_inputs.model.cpp
ully_connected_quant8.model.cpp
ully_connected_quant8_large.model.cpp
ully_connected_quant8_large_weights_as_inputs.model.cpp
ully_connected_quant8_weights_as_inputs.model.cpp
sh_projection.model.cpp
sh_projection_2.model.cpp
sh_projection_weights_as_inputs.model.cpp
stm.model.cpp
stm2.model.cpp
stm3.model.cpp
eshape.model.cpp
eshape_quant8.model.cpp
eshape_quant8_weights_as_inputs.model.cpp
eshape_weights_as_inputs.model.cpp
pace_to_depth_quant8_1.model.cpp
vdf.model.cpp
|
| 66d5cb6e3a90aefc8d545f6369080ab88de9d667 |
05-Oct-2017 |
Jean-Luc Brouillet <jeanluc@google.com> |
Rename setInputsAndOutputs
To avoid confusion with other set* functions, rename setInputsAndOutputs
to identifyInputsAndOutputs.
Also added a few new generated tests when I reran generate_test.sh.
Bug: 63905942
Test: local & VTS tests
Change-Id: Ie947044bee1a97426c847b9d66e60f0c25395bbc
2_normalization_large.model.cpp
2_pool_float_large.model.cpp
stm.model.cpp
stm2.model.cpp
stm3.model.cpp
nn.model.cpp
vdf.model.cpp
|
| 238a880d67b3aa5650ad52037fe1c25b1750eca9 |
05-Oct-2017 |
Miao Wang <miaowang@google.com> |
Fix quantized CONV_2D and FULLY_CONNECTED tests and add more validation
for input tensor types.
Bug: 67430183
Bug: 63905942
Test: mm
Test: NeuralNetworksTest on sailfish
Change-Id: Ibc52ba04ebc515d9208b9266a2e9bbe90d675542
onv_quant8.model.cpp
onv_quant8_channels.model.cpp
onv_quant8_large.model.cpp
onv_quant8_overflow.model.cpp
ully_connected_quant8.model.cpp
ully_connected_quant8_large.model.cpp
|
| e68d924a02511200be6186c947359dee7bb58f23 |
04-Oct-2017 |
Jean-Luc Brouillet <jeanluc@google.com> |
Modify test files for size. Add tests.
Modified the specs so that the generated data files won't be
so large as to take very long to compile.
Also generate the tests for the new specs that have been added.
Finally, update the VTS tests for the specs we have.
Bug: 67394766
Test: Compiled and ran unit tests and VTS tests.
The regular tets pass except for:
[ FAILED ] GeneratedTests.conv_quant8_channels
[ FAILED ] GeneratedTests.conv_quant8_large
[ FAILED ] GeneratedTests.conv_quant8_overflow
The VTS tests pass except for:
[ FAILED ] NeuralnetworksHidlTest.conv_quant8_channels
[ FAILED ] NeuralnetworksHidlTest.conv_quant8_large
[ FAILED ] NeuralnetworksHidlTest.conv_quant8_overflow
[ FAILED ] NeuralnetworksHidlTest.lsh_projection
[ FAILED ] NeuralnetworksHidlTest.lstm2
[ FAILED ] NeuralnetworksHidlTest.lstm3
[ FAILED ] NeuralnetworksHidlTest.lstm
Change-Id: I5a26d0d05e1eca9874b0e9b09f33f345b657667a
vg_pool_float_2.model.cpp
vg_pool_float_3.model.cpp
vg_pool_float_4.model.cpp
vg_pool_quant8_2.model.cpp
oncat_float_2.model.cpp
oncat_float_3.model.cpp
oncat_quant8_2.model.cpp
oncat_quant8_3.model.cpp
onv_quant8_channels.model.cpp
onv_quant8_large.model.cpp
onv_quant8_overflow.model.cpp
epthwise_conv2d_float.model.cpp
epthwise_conv2d_float_large.model.cpp
epthwise_conv2d_float_large_2.model.cpp
epthwise_conv2d_quant8.model.cpp
epthwise_conv2d_quant8_large.model.cpp
ully_connected_float_large.model.cpp
ully_connected_quant8_large.model.cpp
ogistic_float_2.model.cpp
ax_pool_float_2.model.cpp
ax_pool_float_3.model.cpp
ax_pool_quant8_2.model.cpp
ax_pool_quant8_3.model.cpp
elu1_float_2.model.cpp
elu1_quant8_2.model.cpp
elu6_float_2.model.cpp
elu6_quant8_2.model.cpp
elu_float_2.model.cpp
elu_quant8_2.model.cpp
|
| 224c01eb06d3dc496b99b0827fdcc9e65bfc4f9a |
03-Oct-2017 |
Dong Chen <dongchen@google.com> |
Update generated unit tests and vts tests.
Bug: 63905942
Test: NeuralNetworksTest on Angler
Change-Id: I6a12e8e118bf0855ba6fa49065e054a2336473cd
dd_broadcast_quant8.model.cpp
dd_quant8.model.cpp
vg_pool_float.model.cpp
vg_pool_float_1.model.cpp
vg_pool_float_2.model.cpp
vg_pool_float_3.model.cpp
vg_pool_float_4.model.cpp
vg_pool_quant8.model.cpp
vg_pool_quant8_1.model.cpp
vg_pool_quant8_2.model.cpp
vg_pool_quant8_3.model.cpp
vg_pool_quant8_4.model.cpp
oncat_float.model.cpp
oncat_float_1.model.cpp
oncat_float_2.model.cpp
oncat_float_3.model.cpp
oncat_quant8.model.cpp
oncat_quant8_1.model.cpp
oncat_quant8_2.model.cpp
oncat_quant8_3.model.cpp
onv_float_channels.model.cpp
onv_float_large.model.cpp
ashtable_lookup_float.model.cpp
ashtable_lookup_quant8.model.cpp
ogistic_float_1.model.cpp
ogistic_float_2.model.cpp
ogistic_quant8_1.model.cpp
ogistic_quant8_2.model.cpp
stm2.model.cpp
stm3.model.cpp
ax_pool_float.model.cpp
ax_pool_float_1.model.cpp
ax_pool_float_2.model.cpp
ax_pool_float_3.model.cpp
ax_pool_quant8.model.cpp
ax_pool_quant8_1.model.cpp
ax_pool_quant8_2.model.cpp
ax_pool_quant8_3.model.cpp
ul_broadcast_quant8.model.cpp
ul_quant8.model.cpp
elu1_float.model.cpp
elu1_float_1.model.cpp
elu1_float_2.model.cpp
elu1_quant8.model.cpp
elu1_quant8_1.model.cpp
elu1_quant8_2.model.cpp
elu6_float.model.cpp
elu6_float_1.model.cpp
elu6_float_2.model.cpp
elu6_quant8.model.cpp
elu6_quant8_1.model.cpp
elu6_quant8_2.model.cpp
elu_float.model.cpp
elu_float_1.model.cpp
elu_float_2.model.cpp
elu_quant8.model.cpp
elu_quant8_1.model.cpp
elu_quant8_2.model.cpp
nn.model.cpp
igmoid_float.model.cpp
igmoid_quant8.model.cpp
|
| 35647da7af1f99b6abb48c4eeaec042ea9edfb4d |
03-Oct-2017 |
Miao Wang <miaowang@google.com> |
Fix documentation of FULLY_CONNECTED, and add checks for input rank.
Bug: 67319070
Bug: 63905942
Test: mm
Test: NeuralNetworksTests pass
Change-Id: Ia19c643598c48682b65d4de984fb22d9c3bb05f9
ully_connected_float.model.cpp
ully_connected_quant8.model.cpp
|
| 93e086fbc0f1577eabdffd0d3420589f2788bd95 |
02-Oct-2017 |
Miao Wang <miaowang@google.com> |
Remove activation parameter from CONCATENATION
Bug: 67185615
Test: mm
Change-Id: I3555259305761691ea65d2fff406d7fcd8ee8eb5
oncat_float.model.cpp
oncat_quant8.model.cpp
|
| 9fd99905f7dcb2cd4195d2670aded45fc49cc1cb |
30-Sep-2017 |
Miao Wang <miaowang@google.com> |
Fix test generator
Bug: 63905942
Test: mm
Test: NeuralNetworksTest pass
Change-Id: Ie3d734aee5cf10051045c2aafa8db6c508dd2698
igmoid_quant8.model.cpp
oftmax_float_1.model.cpp
oftmax_quant8_1.model.cpp
oftmax_quant8_2.model.cpp
|
| 45bf79e5b9fee354fde7c1f64417d9ca4a1da7da |
25-Sep-2017 |
Miao Wang <miaowang@google.com> |
Clarify the expectation of scale and zeroPoint for affected ops.
- Remove the OperandType constructor taking f_min and f_max as
quantization parameters as it causes confusions for ops like
LOGISTIC and SOFTMAX.
- Update the documenation to clearly state the expected scale and
zeroPoint for LOGISTIC, SOFTMAX and CONCATENATION.
- Update the tests to directly input scale and zeroPoint.
Bug: 63905942
Test: mm
Test: NeuralNetworksTest pass
Change-Id: Ia450d6ce9509205d22e6383bd7e454afa0568cbb
igmoid_quant8.model.cpp
oftmax_quant8_1.model.cpp
oftmax_quant8_2.model.cpp
|
| 7d6ac906f000f3afe418e92c0a4ae36b2ea1143e |
23-Sep-2017 |
Jean-Luc Brouillet <jeanluc@google.com> |
Remove operationTuple.
Removed operationTuple from Model in the HAL, as the data type was
redundant information.
Removed supportedOperationTuples from Capabilities, as real drivers need
more complex restrictions than this provided. For the OC-MR1, we'll just
rely on getSupportedNodes.
Also removed the unused cachesCompilation.
Bug: 63905942
Test: Compiled and ran tests, incl. vts.
Change-Id: I84433d1e0e5211ac2f77317f3752171d85a5ab04
dd.model.cpp
vg_pool_float.model.cpp
vg_pool_quant8.model.cpp
oncat_float.model.cpp
oncat_quant8.model.cpp
onv_float.model.cpp
onv_quant8.model.cpp
epth_to_space_float_1.model.cpp
epth_to_space_float_2.model.cpp
epth_to_space_quant8_1.model.cpp
epth_to_space_quant8_2.model.cpp
epthwise_conv_2d.model.cpp
epthwise_conv_2d_quant8.model.cpp
equantize.model.cpp
mbedding_lookup.model.cpp
loor.model.cpp
ully_connected_float.model.cpp
ully_connected_quant8.model.cpp
ashtable_lookup_float.model.cpp
ashtable_lookup_quant8.model.cpp
2_normalization.model.cpp
2_pool_float.model.cpp
ocal_response_norm_float_1.model.cpp
ocal_response_norm_float_2.model.cpp
ocal_response_norm_float_3.model.cpp
ocal_response_norm_float_4.model.cpp
sh_projection.model.cpp
stm.model.cpp
stm2.model.cpp
stm3.model.cpp
ax_pool_float.model.cpp
ax_pool_quant8.model.cpp
ul.model.cpp
ul_relu.model.cpp
elu1_float.model.cpp
elu1_quant8.model.cpp
elu6_float.model.cpp
elu6_quant8.model.cpp
elu_float.model.cpp
elu_quant8.model.cpp
eshape.model.cpp
eshape_quant8.model.cpp
esize_bilinear.model.cpp
nn.model.cpp
igmoid_float.model.cpp
igmoid_quant8.model.cpp
oftmax_float_1.model.cpp
oftmax_float_2.model.cpp
oftmax_quant8_1.model.cpp
oftmax_quant8_2.model.cpp
pace_to_depth_float_1.model.cpp
pace_to_depth_float_2.model.cpp
pace_to_depth_quant8_1.model.cpp
pace_to_depth_quant8_2.model.cpp
vdf.model.cpp
anh.model.cpp
|
| 86c4350ae8f4f9a68e2334b08a6e97c8c457e981 |
19-Sep-2017 |
Miao Wang <miaowang@google.com> |
Fix depthwise_conv tests
- For quantized depthwise_conv, the bias needs to be TENSOR_INT32.
- The first dimension for the filter needs to be 1.
Bug: 63905942
Test: depthwise_conv_quant8 in NeuralNetworkTests pass
Change-Id: I180c2f3be59651897839475213fd3c874c55a267
epthwise_conv_2d.model.cpp
epthwise_conv_2d_quant8.model.cpp
|
| b5f8d97277d5c30c4ae3efc884649703ec643f78 |
14-Sep-2017 |
I-Jui (Ray) Sung <ijsung@google.com> |
Re-generate VTS tests from test specs
Also, updated test generator on handling FLOAT32 constants.
Test: VtsHalNeuralnetworksV1_0TargetTest on Angler
Test: frameworks/ml/nn/tools/test_generator/tests/test.py
Bug: 63905942
Change-Id: I4445c0b8836795b73b77e04e125d7e90229ce77a
loor.model.cpp
ully_connected_float.model.cpp
ully_connected_quant8.model.cpp
ocal_response_norm_float_1.model.cpp
ocal_response_norm_float_2.model.cpp
ocal_response_norm_float_3.model.cpp
ocal_response_norm_float_4.model.cpp
sh_projection.model.cpp
stm.model.cpp
stm2.model.cpp
stm3.model.cpp
ul.model.cpp
ul_relu.model.cpp
nn.model.cpp
igmoid_float.model.cpp
igmoid_quant8.model.cpp
oftmax_float_1.model.cpp
oftmax_float_2.model.cpp
oftmax_quant8_1.model.cpp
oftmax_quant8_2.model.cpp
pace_to_depth_float_1.model.cpp
pace_to_depth_float_2.model.cpp
pace_to_depth_quant8_1.model.cpp
pace_to_depth_quant8_2.model.cpp
vdf.model.cpp
anh.model.cpp
|
| b6b3dc4cc7a45b61fcfa503f776f658148b5a2a4 |
14-Sep-2017 |
I-Jui (Ray) Sung <ijsung@google.com> |
Support don't care outputs in VTS tests
- Updated test generator to generate helper functions
- Re-generated VTS tests
Actual uses of the infrastructure (IgnoredOutput) come with the test
specs of LSTM.
Bug: 63905942
Test: test.py on host
Test: NeuralNetworksTest on Angler
Test: VtsHalNeuralnetworksV1_0TargetTest on Angler
Change-Id: I464aa1cd77fd2d2fecab5d01dd223346fce04f7b
dd.model.cpp
vg_pool_float.model.cpp
vg_pool_quant8.model.cpp
oncat_float.model.cpp
oncat_quant8.model.cpp
onv_float.model.cpp
onv_quant8.model.cpp
epth_to_space_float_1.model.cpp
epth_to_space_float_2.model.cpp
epth_to_space_quant8_1.model.cpp
epth_to_space_quant8_2.model.cpp
epthwise_conv_2d.model.cpp
epthwise_conv_2d_quant8.model.cpp
equantize.model.cpp
mbedding_lookup.model.cpp
ashtable_lookup.model.cpp
ashtable_lookup_float.model.cpp
ashtable_lookup_quant8.model.cpp
2_normalization.model.cpp
2_pool_float.model.cpp
ax_pool_float.model.cpp
ax_pool_quant8.model.cpp
elu1_float.model.cpp
elu1_quant8.model.cpp
elu6_float.model.cpp
elu6_quant8.model.cpp
elu_float.model.cpp
elu_quant8.model.cpp
eshape.model.cpp
eshape_quant8.model.cpp
esize_bilinear.model.cpp
|
| d8b6b50ab2b326d11a703e869a3546679bd5727b |
14-Sep-2017 |
I-Jui (Ray) Sung <ijsung@google.com> |
Fix build failure - checking in missing VTS files
Add missing generated files.
Bug: 63905942
Test: VtsHalNeuralnetworksV1_0TargetTest on Angler
Change-Id: I77eb1a64c01ed2fab8f2cd07dea95bc0bf96eb40
oncat_quant8.model.cpp
onv_float.model.cpp
onv_quant8.model.cpp
epth_to_space_float_1.model.cpp
epth_to_space_float_2.model.cpp
epth_to_space_quant8_1.model.cpp
epth_to_space_quant8_2.model.cpp
epthwise_conv_2d.model.cpp
epthwise_conv_2d_quant8.model.cpp
equantize.model.cpp
mbedding_lookup.model.cpp
ashtable_lookup_float.model.cpp
ashtable_lookup_quant8.model.cpp
2_normalization.model.cpp
elu1_quant8.model.cpp
elu6_quant8.model.cpp
elu_quant8.model.cpp
eshape.model.cpp
eshape_quant8.model.cpp
esize_bilinear.model.cpp
|
| 420cf7bdb689bc53b2a4e8c5f9eca42f7a06c1c0 |
11-Sep-2017 |
I-Jui (Ray) Sung <ijsung@google.com> |
Generated VTS tests and minor tweak on test harness
Changes:
- Export generated models and examples path in Android.bp
- Avoid duplicated definitions in TestHarness.h when included multiple
times
- Add generated VTS models for VTS
- Update test generator to compute scale and offset from fmin and fmax
Bug: 63905942
Test: VTS and CTS unit testcases with sample driver enabled on Angler
Change-Id: I3e7c5a2aa2c9945e7cd8cdf93968ef88f2b3bbfa
dd.model.cpp
vg_pool_float.model.cpp
vg_pool_quant8.model.cpp
oncat_float.model.cpp
ashtable_lookup.model.cpp
2_pool_float.model.cpp
ax_pool_float.model.cpp
ax_pool_quant8.model.cpp
elu1_float.model.cpp
elu6_float.model.cpp
elu_float.model.cpp
|