Age | Commit message (Collapse) | Author |
|
Fixes issue where pad_left < stride would cause padding
values to be missed
Signed-off-by: Eric Kunze <eric.kunze@arm.com>
Change-Id: I2ba0d0969ce1f00e2cb3ca1ab293c5e15a0c7749
|
|
Signed-off-by: Jeremy Johnson <jeremy.johnson@arm.com>
Change-Id: Ie5d119dc317303a0d2a71d018ac94ce6800ecbf5
|
|
Signed-off-by: Jeremy Johnson <jeremy.johnson@arm.com>
Change-Id: I464871b412b3dfc9a84bd3d8fffe4c1de3072b69
|
|
Update to ref model to check ERROR_IF criteria for pooling
and convolution ops to match specification
Update to tosa_verif_build_tests to produce valid test ranges and
new ERROR_IF tests
Plus update pooling ops big kernel to 9 (from 6) for better testing
coverage and set dilation to 1 and add out_pad bottom & right for
transpose_conv2d to match specification
Signed-off-by: Jeremy Johnson <jeremy.johnson@arm.com>
Change-Id: Ic5759872d40ae8d3f3d07043d9a0f2fa0244d72e
|
|
Move all error & validation into tosa_error_if.py
Move all argument and tensor generation into tosa_arg_gen.py
Move utility functions into tosa_utils.py
Create new TosaTensorValuesGen class for specialising tensor
value generation.
Change-Id: Ib9ac65e2308b14471a567c6f11d775c76585bc5b
Signed-off-by: Jeremy Johnson <jeremy.johnson@arm.com>
|
|
Signed-off-by: Jeremy Johnson <jeremy.johnson@arm.com>
Change-Id: Ied54f5064dfb44029136b0f0297ef35e6eb0be6e
|
|
And limit REDUCE_SUM test values to within int32
Signed-off-by: Jeremy Johnson <jeremy.johnson@arm.com>
Change-Id: I4d902b245d17eb343cfb2bbc23d9db28c1d1f4c3
|
|
* black needed updating for ImportError of new click version
Change-Id: If8e2d13a09996e331d676af31bf3d3aee9403b4e
Signed-off-by: Jeremy Johnson <jeremy.johnson@arm.com>
|
|
And update test generation to create values in predictable range
Signed-off-by: Jeremy Johnson <jeremy.johnson@arm.com>
Change-Id: I4ba1ff445bf6caeec9f8782902fc45929fe0ee77
|
|
Signed-off-by: Jeremy Johnson <jeremy.johnson@arm.com>
Change-Id: If785fdaab3026eca5a31888115fba8a6750e0460
|
|
- Fixed input shape/expected input shape mismatches
when due to a missing batch dimension.
- Return from run_test() immediately after running compiler
when --no-ref-model is set. This avoids doing result
comparisons on results that were never generated.
- Exit gracefully when no tests are specified instead of
throwing TypeError on iterating NoneType
Signed-off-by: Jared Smolens <jared.smolens@arm.com>
Change-Id: I91d2c24fd60aa1cf8b3755894030ff14f1371ed0
|
|
- Error check on missing test description file instead of throwing
json exception
- Removed outdated help option
Signed-off-by: Jared Smolens <jared.smolens@arm.com>
Change-Id: Ia8780a88f4306af19ce63eaa679c84669f5c2ad9
|
|
Signed-off-by: Jeremy Johnson <jeremy.johnson@arm.com>
Change-Id: I41c364905ca364d3446962b737735b9068a29a77
|
|
Signed-off-by: Jeremy Johnson <jeremy.johnson@arm.com>
Change-Id: Ifa34021413cd6543c987a7b7c9c664144a4dd95d
|
|
And fixes in tosa_verif_run_tests:
* support for no-color printing
* stop double printing of error messages on verbose
* differentiate result code pass from results check
Change-Id: I26e957013a8d18f7d3d3691067dfb778008a1eea
Signed-off-by: Jeremy Johnson <jeremy.johnson@arm.com>
|
|
Signed-off-by: Jeremy Johnson <jeremy.johnson@arm.com>
Change-Id: I65a7fac6e65d2b979139adaf6c5a42139b5493c2
|
|
Fix up generated values for rescale tests
Signed-off-by: Jeremy Johnson <jeremy.johnson@arm.com>
Change-Id: I28fc3b8f189d25a7ad8e5172d4d8a43b86820fcf
|
|
Signed-off-by: Jeremy Johnson <jeremy.johnson@arm.com>
Change-Id: Iff6eb54734e7a78eb22d1a64803b937063b7f65a
|
|
Previous code assumed symmetric padding when calculating the divisor for
the average pool, and had issues with extra padding.
The padding now follows the definition from the specification. Extra
padding to the right/bottom is ignored when calculating the average pool
divisor.
Change-Id: Ib05e2346c6e9d55b6fef5294322e58ee18ef4928
|
|
The conformance oversize tests are not produced by default.
Change-Id: I9f7d4a60b9d8991889a077b61613e13e288bb65d
Signed-off-by: Jeremy Johnson <jeremy.johnson@arm.com>
|
|
Limit input values to allowed for logical shift operations.
Signed-off-by: Jeremy Johnson <jeremy.johnson@arm.com>
Change-Id: I78110c449274ab96a3f824890c3f03a0eeb345eb
|
|
Added missing reference model REQUIRE check for shift value (0-31)
Make sure result of LOGICAL_SHIFT_LEFT is masked to input size
Fixed test generation to produce shift values in that range
Signed-off-by: Jeremy Johnson <jeremy.johnson@arm.com>
Change-Id: Id511de0d989ea954fc1afd18dc2051341bce2cd0
|
|
Signed-off-by: Jeremy Johnson <jeremy.johnson@arm.com>
Change-Id: I0c1edba930b846ff63785c17116cc8d465ecfbf3
|
|
Script converts framework and reference model unit tests into a
suitable format for including into a conformance suite.
Signed-off-by: Jeremy Johnson <jeremy.johnson@arm.com>
Change-Id: Ida1ec8a0a7ea31fd3a3f62c4cb52d7cc2bf0b439
|
|
Fix up issues in existing python scripts.
Signed-off-by: Jeremy Johnson <jeremy.johnson@arm.com>
Change-Id: Id4adab404560c3129c66f31c21ff0ce148283c73
|
|
Rename to tosa_verif_run_tests to match build_tests
Improve output and system under test support
Improve xunit support
Add results checker
Add utilities json2numpy and json2fbbin
Add set of python tests
Update README.md
Signed-off-by: Jeremy Johnson <jeremy.johnson@arm.com>
Change-Id: Ia09f8e6fd126579b3ba1c1cda95c1326802417ca
|
|
Split generator and runner scripts
Add package setup
Add py-dev-env.sh/.bash to allow editing source files during dev
Update README.md with installation info
Signed-off-by: Jeremy Johnson <jeremy.johnson@arm.com>
Change-Id: I172fe426d99e2e9aeeacedc8b8f3b6a79c8bd39d
|
|
When testing rank mismatch tests, could be accessing beyond the end
a tensor's shape.
Change-Id: I304463e50579d574d629abd13098017a26a1406b
Signed-off-by: Eric Kunze <eric.kunze@arm.com>
|
|
Signed-off-by: Jeremy Johnson <jeremy.johnson@arm.com>
Change-Id: I267fa624241f60f1fa613e27e4fe46fc43768f8a
|
|
Change-Id: I33237ebfd946b9ec91352c2b0dc6298cc113cd77
Signed-off-by: Les Bell <les.bell@arm.com>
|
|
REDUCE ops - test generation now limited to ranks 1-4 and so
WrongRank tests created
SCATTER/GATHER ops - enable WrongRank testing
Change-Id: I0909573e0edabddbad8764e7542d654967a5c8b9
Signed-off-by: Jeremy Johnson <jeremy.johnson@arm.com>
|
|
TRANSPOSE - remove WrongRank testing as no limit in spec
RESIZE - make floating point test naming same format as int
- fix MaxDimExceeeded by not restricting dims and
fix max dim check to match spec
- catch stride less than/equal zero and regen
PAD - remove zero point testing as not in spec
Change-Id: I4eb0b78d7f577469d202866d95833e7d5fd2f2a8
Signed-off-by: Jeremy Johnson <jeremy.johnson@arm.com>
|
|
Signed-off-by: Les Bell <les.bell@arm.com>
Change-Id: I68a13e1b337b1afc2ab5e0edcffda2b4b0cecdda
|
|
- update example .tosa
Signed-off-by: Kevin Cheng <kevin.cheng@arm.com>
Change-Id: I1eb9f60d33de494408963e6ac16c1566254200f1
|
|
- test like [1] + [2] = [1] should be treated as invalid test
- modify matchRankShape() function so it allows size 1 only on the source tensor but not target tensor
Signed-off-by: Kevin Cheng <kevin.cheng@arm.com>
Change-Id: I6bbb6a63dc1143712e7eef736a991cac419b009e
|
|
Signed-off-by: Kevin Cheng <kevin.cheng@arm.com>
Change-Id: I49d498dd3d4c069d8d1db07310f939268b9df4b7
|
|
Add RankMismatch and DimensionMismatch support for SELECT
Update RankMismatch ops to also support DimensionMismatch
Update POW op to have proper broadcast testing
A few other broadcastable ops missing Rank/Dimension testing
Change-Id: I6566f45a7a0db4f9f008456ea7a8e23d4192f4f9
Signed-off-by: Jeremy Johnson <jeremy.johnson@arm.com>
|
|
Signed-off-by: Kevin Cheng <kevin.cheng@arm.com>
Change-Id: I9ed3d8971a133b4cbb2cf7d827f4e69d55dee246
|
|
* Transpose WrongRank tests now use ranks 7, 8
* Concat ERROR_IF checks now test for inaccurate summation
of output shape tensor dimension
Change-Id: If32f43a4dbd872d0ef7625fa3d4969c863a11b8c
Signed-off-by: Matthew Haddon <matthew.haddon@arm.com>
Signed-off-by: Jeremy Johnson <jeremy.johnson@arm.com>
Signed-off-by: Les Bell <les.bell@arm.com>
|
|
Signed-off-by: Matthew Haddon <matthew.haddon@arm.com>
Signed-off-by: Les Bell <les.bell@arm.com>
Change-Id: Ie6c8c8653874f9eed6007a54a3ad526601a4a669
|
|
* Operators implemented: sigmoid, tanh, arthmetic_right_shift,
mul, table, select, equal, greater, greater_equal, concat, reverse,
tile, scatter, gather, case
* Note that over the course of implementation some specific ERROR_IF
checks have been added for some of the above operators
Change-Id: I80595e6eb9a3e5efd1cc6fd7aa28bbc2dd614980
Signed-off-by: Matthew Haddon <matthew.haddon@arm.com>
Signed-off-by: Les Bell <les.bell@arm.com>
Signed-off-by: Jeremy Johnson <jeremy.johnson@arm.com>
|
|
- update thirdparty/serialization_lib
- avoid building uncessary targets in submodule
Signed-off-by: Kevin Cheng <kevin.cheng@arm.com>
Change-Id: I34698014b63e4d2d8b78011c001c85882014007c
|
|
Change-Id: I1de45a21779ec014a0fda2c971dbc8f91df45bdb
Signed-off-by: Eric Kunze <eric.kunze@arm.com>
|
|
Update to the corresponding serialization library
Change-Id: I323bf306d409b51bbf53447927953edd0fd79983
Signed-off-by: Eric Kunze <eric.kunze@arm.com>
|
|
- TRANSPOSE: move perm attribute check to compile-time checker
- TABLE: add output type checker
Signed-off-by: Kevin Cheng <kevin.cheng@arm.com>
Change-Id: I834a5f290fbc384ef339b624060e6e5c77072c36
|
|
- MATMUL: only check a_zp/b_zp valid when this->qinfo exists
- Fix typo in debug message
Signed-off-by: Kevin Cheng <kevin.cheng@arm.com>
Change-Id: I2cedcb25e4f57fcaec2caa1b850ea1232a023340
|
|
Allow random shift values rather than just 11
Check for values that cause unpredictable behaviour
Change-Id: I508c9d0982808ac03b8400d47ad9445d6afb1b79
Signed-off-by: Jeremy Johnson <jeremy.johnson@arm.com>
|
|
- update serialization_lib hash
- PAD:
1. make padding as an attribute instead of tensor.
2. add pad_const_int (for non-float type) / pad_const_fp (for float type)
- TRANSPOSE: make perm as an attribute instead of tensor
- TABLE: make table as attribute instead of tensor
- update examples/ tests
Signed-off-by: Kevin Cheng <kevin.cheng@arm.com>
Change-Id: Iddc446db4b356ba2f36ea4a79b7220b9cfc2aa4e
|
|
- update serialization_lib hash
- check model version against serializer version
- add "-v" to command line option to print out model version
Signed-off-by: Kevin Cheng <kevin.cheng@arm.com>
Change-Id: I554cf0fdead22e7e5f91d2e98831459286c40a2d
|
|
Change-Id: Idb36b1f1c0d78ec101c168865a9c8d03221b4c84
Signed-off-by: Jeremy Johnson <jeremy.johnson@arm.com>
|