| /cmsis-nn-latest/Tests/UnitTest/ |
| D | README.md | 11 The following apt packages are required. Replace python venv version with your python version.
17 It has been tested with Python 3.10 and it has been tested on Ubuntu 22.04.
53 Python package tflite_runtime can be installed with pip and it can also be built locally. Check thi…
57 Python package tflite_micro can be installed with pip and it can also be built locally. See this co…
73 …t tests and run unit tests. This script has been designed for Linux hosts with both aarch64 and x8…
123 Generating new test data is done with the following script. Use the -h flag to get more info.
130 …, i.e. it need a test set data name as input. It will then generate files with that name as prefix…
133 …est test in the load_all_testdatasets() function. Run the generate script with that new test set a…
166 … will remain as the main data generator until all functionality is replicated with the new scripts.
196 …nction under test there is a folder with the same name with test_ prepended to the name and it con…
[all …]
|
| D | unittest_targets.py | 8 # not use this file except in compliance with the License.
53 " So basically the different options can be listed with:"
110 with open(target_json, "r") as read_file:
122 … error_handler(168, 'Unsupported target: {} with architecture: {}'.format(target_name, arch))
156 print("Running {} on {} target: {} with compiler: {} in directory: {} test: {}\n".format(
403 with open(test_runner, "r") as f:
405 with open(test_runner, "w") as f:
473 with read_file:
|
| /cmsis-nn-latest/ |
| D | README.md | 7 This means CMSIS-NN is bit-exact with Tensorflow Lite reference kernels. In some cases TFL and TFLM…
20 Processors with DSP extension uses Single Instruction Multiple Data(SIMD) instructions for optimiza…
21 processors here are Cortex-M4 or a Cortex-M33 configured with optional DSP extension.
24 Processors with Arm Helium Technology use the Arm M-profile Vector Extension(MVE) instructions for …
25 Examples are Cortex-M55 or Cortex-M85 configured with MVE.
53 * Variable and function names are lower case with an underscore separator.
96 Default optimization level is set at Ofast. This can be overwritten with CMake on command line by u…
97 …is can impact performance. With only optimization level -O0, *ARM_MATH_AUTOVECTORIZE* needs to be …
107 …with DSP extension, int4 and int8 convolutions make use of the restrict keyword for the output poi…
112 …x. It should be possible to use the C implementation and compile for host with minor stubbing effo…
|
| D | LICENSE | 17 control with that entity. For the purposes of this definition,
81 with the Work to which such Contribution(s) was submitted. If You
90 Work or Derivative Works thereof in any medium, with or without
113 documentation, if provided along with the Derivative Works; or,
127 reproduction, and distribution of the Work otherwise complies with
136 with Licensor regarding such Contributions.
151 risks associated with Your exercise of permissions under this License.
168 or other liability obligations and/or rights consistent with this
181 boilerplate notice, with the fields enclosed by brackets "{}"
182 replaced with your own identifying information. (Don't include
[all …]
|
| /cmsis-nn-latest/Include/ |
| D | arm_nnsupportfunctions.h | 7 * not use this file except in compliance with the License.
58 // Number of channels processed in a block for DW Conv with Int8 weights(MVE)
61 // A layer with lower number of channels than CH_IN_BLOCK_MVE will result in higher
62 // scratch buffer usage and a layer with higher number of channels than CH_IN_BLOCK_MVE
66 // Number of channels processed in a block for DW Conv with Int4 weights(MVE)
81 // Only applicable for processors with MVE extension.
160 * @brief Converts the elements from a s8 vector to a s16 vector with an added offset
180 * @brief Converts the elements from a s8 vector to a s16 vector with an added offset
188 * No additonal ordering is done with the result that output elements are not in order.
190 * Note this is for processors with DSP extension only.
[all …]
|
| D | arm_nnfunctions.h | 7 * not use this file except in compliance with the License.
32 …* A collection of functions to perform basic operations for neural network layers. Functions with …
66 …* @brief s4 convolution layer wrapper function with the main purpose to call the optimal kernel av…
81 * @param[in] filter_data Filter data pointer. Data type: int8 packed with 2x int4
138 …* @brief Get the required buffer size for arm_convolve_wrapper_s4 for processors with DSP extensio…
151 …* @brief s8 convolution layer wrapper function with the main purpose to call the optimal kernel av…
222 …* @brief Get the required buffer size for arm_convolve_wrapper_s8 for processors with DSP extensio…
235 …* @brief s16 convolution layer wrapper function with the main purpose to call the optimal kernel a…
252 …* @param[in] bias_data Struct with optional bias data pointer. Bias data type can be int…
294 …* @brief Get the required buffer size for arm_convolve_wrapper_s16 for for processors with DSP ext…
[all …]
|
| /cmsis-nn-latest/Source/BasicMathFunctions/ |
| D | arm_maximum_s8.c | 7 * not use this file except in compliance with the License.
137 // arm_max_scalar expects the tensor with the scalar value to be provided first in arm_maximum_s8()
142 // arm_max_scalar expects the tensor with the scalar value to be provided first in arm_maximum_s8()
174 // arm_max_scalar expects the tensor with the scalar value to be provided first in arm_maximum_s8()
180 // arm_max_scalar expects the tensor with the scalar value to be provided first in arm_maximum_s8()
199 … // arm_max_scalar expects the tensor with the scalar value to be provided first in arm_maximum_s8()
207 … // arm_max_scalar expects the tensor with the scalar value to be provided first in arm_maximum_s8()
226 … // arm_max_scalar expects the tensor with the scalar value to be provided first in arm_maximum_s8()
234 … // arm_max_scalar expects the tensor with the scalar value to be provided first in arm_maximum_s8()
|
| D | arm_minimum_s8.c | 7 * not use this file except in compliance with the License.
138 // arm_min_scalar expects the tensor with the scalar value to be provided first in arm_minimum_s8()
143 // arm_min_scalar expects the tensor with the scalar value to be provided first in arm_minimum_s8()
198 … // arm_min_scalar expects the tensor with the scalar value to be provided first in arm_minimum_s8()
206 … // arm_min_scalar expects the tensor with the scalar value to be provided first in arm_minimum_s8()
225 … // arm_min_scalar expects the tensor with the scalar value to be provided first in arm_minimum_s8()
233 … // arm_min_scalar expects the tensor with the scalar value to be provided first in arm_minimum_s8()
|
| D | arm_elementwise_mul_s16_s8.c | 7 * not use this file except in compliance with the License.
22 * Description: Elementwise multiplication of 16 bit input with 8 bit output
43 * s16 elementwise multiplication with s8 output
|
| /cmsis-nn-latest/Tests/UnitTest/RefactoredTestGen/Lib/ |
| D | test.py | 6 # not use this file except in compliance with the License.
55 """ Create a test with given parameters """
203 """ Convert a model generated with keras to tflite-format """
240 with output_fpath.open("wb") as f:
289 with config_fpath.open("w+") as f:
309 with test_data_fpath.open("w") as f:
327 with fname.open("w+") as f:
350 with test_data_fpath.open("a") as f:
358 with fname.open("a") as f:
379 with test_functions_fpath.open("a") as f:
[all …]
|
| D | op_utils.py | 6 # not use this file except in compliance with the License.
31 self.aliases = aliases # Optional for backward compability with old unit tests
44 """ Returns a non-quantized tflite-model with given shapes and params"""
58 …Generates quantized tensors, scales, and other parameters with given shapes and params, calculates…
|
| /cmsis-nn-latest/Documentation/Doxygen/src/ |
| D | mainpage.md | 26 CMSIS-NN targets Cortex-M processors with typically three different implementations for each functi…
29 - Processors with DSP extension (e.g Cortex-M4)
30 - Processors with Arm M-Profile Vector Extension(MVE) instructions (e.g Cortex-M55)
57 - Selects code for processors with DSP extension.
|
| /cmsis-nn-latest/Source/NNSupportFunctions/ |
| D | arm_nn_depthwise_conv_nt_t_s8.c | 7 * not use this file except in compliance with the License.
22 * Description: Depthwise convolution on matrices with no padding.
27 * Target Processor: Cortex-M processors with MVE extension.
42 …* Depthwise convolution of rhs matrix with 4 lhs matrices with no padding. Dimensions are the same…
|
| D | arm_nn_depthwise_conv_nt_t_s16.c | 7 * not use this file except in compliance with the License.
22 * Description: Depthwise convolution on matrices with no padding.
27 * Target Processor: Cortex-M processors with MVE extension
42 …* Depthwise convolution of rhs matrix with 4 lhs matrices with no padding. Dimensions are the same…
|
| D | arm_nn_depthwise_conv_nt_t_s4.c | 8 * not use this file except in compliance with the License.
23 * Description: Depthwise convolution on matrices with no padding and packed int4 weights.
28 * Target Processor: Cortex-M processors with MVE extension.
43 * Depthwise convolution of rhs matrix with 4 lhs matrices with no padding and packed int4 weights.
|
| D | arm_q7_to_q15_with_offset.c | 7 * not use this file except in_q7x4 compliance with the License.
22 * Description: Converts the elements of the Q7 vector to Q15 vector with an added offset
75 /* First part of the processing with loop unrolling. Compute 4 outputs at a time. */ in arm_q7_to_q15_with_offset()
|
| D | arm_nn_depthwise_conv_nt_t_padded_s8.c | 7 * not use this file except in compliance with the License.
22 * Description: Depthwise convolution with padded matrices.
27 * Target Processor: Cortex-M processors with MVE extension
49 …* Depthwise convolution of transposed rhs matrix with 4 lhs matrices. One or more of the rhs matri…
|
| /cmsis-nn-latest/.github/workflows/ |
| D | pack.yml | 20 with:
28 with:
|
| /cmsis-nn-latest/Documentation/ |
| D | README.md | 5 The version drop-down menu there allows to switch between the documentation provided with offical r…
9 Generating the HTML-formatted documentation from the source is done with `gen_doc.sh` script:
|
| /cmsis-nn-latest/Documentation/Doxygen/ |
| D | nn.dxy.in | 6 # not use this file except in compliance with the License.
33 # Use doxygen to compare the used configuration file with the template
36 # Use doxygen to compare the used configuration file with the template
74 # With the PROJECT_LOGO tag one can specify a logo or an icon that is included
124 # Hindi, Hungarian, Indonesian, Italian, Japanese, Japanese-en (Japanese with
125 # English messages), Korean, Korean-en (Korean with English messages), Latvian,
154 # following values are used ($name is automatically replaced with the name of
292 # documentation, which will result in a user-defined paragraph with heading
298 # with the commands \{ and \} for these it is advised to use the version @{ and
342 # parses. With this tag you can assign which parser to use for a given
[all …]
|
| /cmsis-nn-latest/Source/LSTMFunctions/ |
| D | arm_lstm_unidirectional_s16.c | 7 * not use this file except in compliance with the License.
22 * Description: S16 LSTM function with S16 gate output
43 * S16 LSTM function for TensorFlow Lite with S16 gate output
|
| D | arm_lstm_unidirectional_s8.c | 7 * not use this file except in compliance with the License.
22 * Description: S8 LSTM function with S16 gate output
43 * S8 LSTM function for TensorFlow Lite with S16 gate output
|
| /cmsis-nn-latest/Source/SoftmaxFunctions/ |
| D | arm_nn_softmax_common_s8.c | 7 * not use this file except in compliance with the License.
22 * Description: Softmax with s8 input and output of s8 or s16.
51 * Softmax function with s8 input and output of s8 or s16.
|
| /cmsis-nn-latest/Tests/UnitTest/RefactoredTestGen/ |
| D | generate_test_data.py | 8 # not use this file except in compliance with the License.
57 … help="Path to the schema-file needed for generating tflite-files with flatc")
|
| /cmsis-nn-latest/Source/ConvolutionFunctions/ |
| D | arm_convolve_wrapper_s16.c | 7 * not use this file except in compliance with the License.
22 …* Description: s16 convolution layer wrapper function with the main purpose to call the optimal k…
|
