tensorflow-directml-plugin
DirectML PluggableDevice plugin for TensorFlow 2
파일 탐색기
최종 버전 다운로드 (.zip)- github-actions-cpp-check.yml
- github-actions-python-check.yml
- c_cpp_properties.json
- dependencies.cmake
- ADOHelper.ps1
- build.yml
- create_agent_summary.ps1
- create_report_email.ps1
- create_test_env.ps1
- create_test_env.sh
- create_test_matrix.ps1
- create_test_summary_json.ps1
- create_test_summary_xml.ps1
- pipeline.yml
- report.yml
- setup_agent_linux.yml
- setup_agent_windows.yml
- SubmitFilesToBlobStorage.ps1
- test.yml
- c_api_tests.cc
- bad_huffman.jpg
- cat_q20.jpg
- cat_q72.jpg
- cat_q95.jpg
- checkerboard1.png
- checkerboard2.png
- checkerboard3.png
- corrupt.jpg
- corrupt34_2.jpg
- corrupt34_3.jpg
- corrupt34_4.jpg
- grayscale_small.bmp
- grayscale_small_3channels.bmp
- grayscale_small_4channels.bmp
- jpeg_merge_test1.jpg
- jpeg_merge_test1_cmyk.jpg
- lena.bmp
- lena.gif
- lena_gray.png
- lena_palette.png
- lena_palette_trns.png
- lena_rgba.png
- medium.jpg
- optimized.gif
- palette_only.png
- pendulum_sm.gif
- pendulum_sm_frame0.png
- pendulum_sm_frame1.png
- pendulum_sm_frame2.png
- red_black.gif
- rgb_small.bmp
- rgb_small_255.bmp
- rgba_small.bmp
- rgba_small_255.bmp
- scan.gif
- small.jpg
- squares.gif
- __init__.py
- adagrad_test.py
- aggregate_ops_test.py
- argmax_op_test.py
- array_ops_test.py
- batch_matmul_op_test.py
- batchtospace_op_test.py
- bias_op_base.py
- bias_op_deterministic_test.py
- bias_op_test.py
- bitcast_op_test.py
- cast_op_test.py
- concat_op_test.py
- constant_op_test.py
- conv1d_test.py
- conv1d_transpose_test.py
- conv2d_backprop_filter_grad_test.py
- conv2d_transpose_test.py
- conv3d_backprop_filter_v2_grad_test.py
- conv3d_transpose_test.py
- conv_ops_3d_test.py
- conv_ops_test.py
- cross_device_ops_test.py
- cross_grad_test.py
- cudnn_recurrent_test.py
- cwise_ops_binary_test.py
- cwise_ops_test.py
- cwise_ops_unary_test.py
- data_format_op_test.py
- deepcopy_op_test.py
- depthtospace_op_test.py
- depthwise_conv_op_test.py
- diag_op_test.py
- dml_get_adapter_name.py
- dml_test_util.py
- dynamic_stitch_op_test.py
- empty_op_test.py
- extract_image_patches_grad_test.py
- extract_image_patches_op_test.py
- extract_volume_patches_grad_test.py
- extract_volume_patches_op_test.py
- fill_op_test.py
- ftrl_test.py
- gather_nd_op_test.py
- gather_op_test.py
- image_grad_deterministic_test.py
- image_grad_test.py
- image_grad_test_base.py
- image_ops_test.py
- in_topk_op_test.py
- inplace_ops_test.py
- l2loss_test.py
- lrn_op_test.py
- math_ops_test.py
- matmul_op_test.py
- matrix_band_part_op_test.py
- nn_batchnorm_test.py
- nn_fused_batchnorm_test.py
- numerics_test.py
- one_hot_op_test.py
- ones_like_test.py
- pad_op_test.py
- pool_test.py
- pooling_ops_3d_test.py
- pooling_ops_test.py
- proximal_adagrad_test.py
- ragged_one_hot_op_test.py
- ragged_reduce_op_test.py
- random_ops_test.py
- range_op_test.py
- reduce_join_op_test.py
- reduce_test.py
- reduction_ops_test.py
- reduction_ops_test_big.py
- relu_op_test.py
- resource_variable_ops_test.py
- reverse_sequence_op_test.py
- rnn_grad_test.py
- roll_op_test.py
- scan_ops_test.py
- scatter_nd_ops_test.py
- segment_reduction_ops_test.py
- slice_op_test.py
- spacetobatch_op_test.py
- spacetodepth_op_test.py
- sparse_xent_op_test.py
- sparse_xent_op_test_base.py
- split_op_test.py
- stack_op_test.py
- strings_reduce_join_op_test.py
- topk_op_test.py
- training_arrays_test.py
- training_dataset_test.py
- training_eager_test.py
- training_generator_test.py
- training_gpu_test.py
- training_integration_test.py
- training_ops_test.py
- training_test.py
- training_util_test.py
- training_utils_v1_test.py
- transpose_op_test.py
- unstack_op_test.py
- where_op_test.py
- xent_op_test.py
- xent_op_test_base.py
- zeros_like_test.py
- __init__.py
- dml_multiple_devices_test.py
- dml_visible_devices_test.py
- profiler_test.py
- __init__.py
- README.md
- test.py
- tests.json
- tests_schema.json
- dml_adapter.cc
- dml_adapter.h
- dml_adapter_heuristics.h
- dml_adapter_impl.cc
- dml_adapter_impl.h
- dml_bfc_allocator.cc
- dml_bfc_allocator.h
- dml_buffer.cc
- dml_buffer.h
- dml_buffer_region.cc
- dml_buffer_region.h
- dml_command_allocator_ring.h
- dml_command_list.cc
- dml_command_list.h
- dml_command_queue.cc
- dml_command_queue.h
- dml_common.h
- dml_descriptor_bfc_allocator.cc
- dml_descriptor_bfc_allocator.h
- dml_descriptor_heap_allocator.cc
- dml_descriptor_heap_allocator.h
- dml_descriptor_pool.cc
- dml_descriptor_pool.h
- dml_device.cc
- dml_device.h
- dml_device_cache.cc
- dml_device_cache.h
- dml_device_context.cc
- dml_device_context.h
- dml_device_manager.cc
- dml_device_manager.h
- dml_device_state.cc
- dml_device_state.h
- dml_dso_loader.cc
- dml_dso_loader.h
- dml_error_handling.cc
- dml_error_handling.h
- dml_event_queue.cc
- dml_event_queue.h
- dml_execution_context.cc
- dml_execution_context.h
- dml_gpu_event.h
- dml_guids.cc
- dml_guids.h
- dml_heap_allocator.cc
- dml_heap_allocator.h
- dml_kernel_context.cc
- dml_kernel_context.h
- dml_kernel_definition.h
- dml_kernel_key.cc
- dml_kernel_key.h
- dml_kernel_manager.cc
- dml_kernel_manager.h
- dml_kernel_wrapper.cc
- dml_kernel_wrapper.h
- dml_operator_helper.cc
- dml_operator_helper.h
- dml_ops_common.cc
- dml_ops_common.h
- dml_pooled_heap.cc
- dml_pooled_heap.h
- dml_readback_heap.cc
- dml_readback_heap.h
- dml_tagged_pointer.cc
- dml_tagged_pointer.h
- dml_tensor_desc.cc
- dml_tensor_desc.h
- dml_tracing.cc
- dml_tracing.h
- dml_upload_heap.cc
- dml_upload_heap.h
- dml_util.cc
- dml_util.h
- dml_addn_op.cc
- dml_assign_variable_op.cc
- dml_batch_norm_ops.cc
- dml_batch_to_space_op.cc
- dml_bias_add_op.cc
- dml_bitcast_op.cc
- dml_broadcast_to_op.cc
- dml_cast_op.cc
- dml_check_numerics_op.cc
- dml_concat_op.cc
- dml_conv_ops.cc
- dml_crop_and_resize_grad_boxes_op.cc
- dml_crop_and_resize_grad_image_op.cc
- dml_crop_and_resize_op.cc
- dml_cross_op.cc
- dml_cudnn_rnn_ops.cc
- dml_cwise_ops.cc
- dml_data_format_dim_map.cc
- dml_data_format_vec_permute.cc
- dml_deepcopy_op.cc
- dml_diag_op.cc
- dml_diag_part_op.cc
- dml_dynamic_stitch_op.cc
- dml_empty_op.cc
- dml_extract_image_patches_op.cc
- dml_extract_patches_helpers.cc
- dml_extract_patches_helpers.h
- dml_extract_volume_patches_op.cc
- dml_fill_op.cc
- dml_gather_nd_op.cc
- dml_gather_op.cc
- dml_gru_ops.cc
- dml_image_ops.cc
- dml_in_topk_op.cc
- dml_inplace_ops.cc
- dml_l2loss_op.cc
- dml_lrn_ops.cc
- dml_lstm_helpers.h
- dml_lstm_ops.cc
- dml_matmul_op.cc
- dml_matrix_band_part_ops.cc
- dml_matrix_diag_helpers.cc
- dml_matrix_diag_helpers.h
- dml_matrix_diag_ops.cc
- dml_matrix_diag_part_ops.cc
- dml_matrix_set_diag_ops.cc
- dml_mirror_pad_grad_op.cc
- dml_one_hot_op.cc
- dml_ones_like_op.cc
- dml_pack_op.cc
- dml_pad_op.cc
- dml_parallel_concat_ops.cc
- dml_pooling_ops.cc
- dml_random_ops.cc
- dml_range_op.cc
- dml_reduce_ops.cc
- dml_relu_ops.cc
- dml_resize_grad_ops.cc
- dml_resize_op.cc
- dml_reverse_op.cc
- dml_reverse_sequence_op.cc
- dml_roll_op.cc
- dml_scan_ops.cc
- dml_scatter_nd_ops.cc
- dml_scatter_ops.cc
- dml_segment_reduction_ops.cc
- dml_select_op.cc
- dml_slice_op.cc
- dml_snapshot_op.cc
- dml_space_depth_ops.cc
- dml_space_to_batch_op.cc
- dml_sparse_xent_op.cc
- dml_split_op.cc
- dml_strided_slice_op.cc
- dml_swapping_ops.cc
- dml_tile_op.cc
- dml_topk_op.cc
- dml_training_ops.cc
- dml_transpose_op.cc
- dml_unpack_op.cc
- dml_where_op.cc
- dml_xent_op.cc
- dml_zeros_like_op.cc
- pch.h
- byte_order.h
- device_name_utils.cc
- device_name_utils.h
- device_type.cc
- device_type.h
- graph.h
- graph_optimizer.cc
- graph_optimizer.h
- graph_properties.cc
- graph_properties.h
- graph_view.cc
- graph_view.h
- graph_view_internal.h
- grappler_item.cc
- grappler_item.h
- hash.cc
- hash.h
- map_utils.h
- op_registry.cc
- op_registry.h
- op_types.cc
- op_types.h
- optimizer_runner.cc
- optimizer_runner.h
- perm_utils.cc
- perm_utils.h
- proto_buffer_helpers.cc
- proto_buffer_helpers.h
- remapper.cc
- remapper.h
- tensor_id.cc
- tensor_id.h
- tensor_proto_util.cc
- tensor_proto_util.h
- transpose_remover.cc
- transpose_remover.h
- utils.cc
- utils.h
- plugin_device.cc
- plugin_kernel.cc
- plugin_optimizer.cc
- plugin_profiler.cc
- plugin_version.h
- allocator.cc
- allocator.h
- allocator_retry.cc
- allocator_retry.h
- attribute.h
- bcast.cc
- bcast.h
- bfc_allocator.cc
- bfc_allocator.h
- determinism.cc
- determinism.h
- device.cc
- device.h
- env.cc
- env.h
- env_var.cc
- env_var.h
- fused_eigen_output_kernels.cc
- fused_eigen_output_kernels.h
- guarded_philox_random.cc
- guarded_philox_random.h
- kernel_shape_util.cc
- kernel_shape_util.h
- macros.h
- matmul_bcast.h
- mirror_pad_mode.cc
- mirror_pad_mode.h
- node_def.h
- numbers.cc
- numbers.h
- op_defs.h
- op_defs_core.cc
- op_defs_core.h
- op_defs_dml.h
- op_kernel.h
- op_kernel_construction.cc
- op_kernel_construction.h
- op_kernel_context.cc
- op_kernel_context.h
- padding.cc
- padding.h
- path.cc
- path.h
- philox_random.h
- random_ops_util.h
- rng_alg.h
- stateless_random_ops.cc
- stateless_random_ops.h
- status.cc
- status.h
- statusor.h
- stream.h
- tensor.cc
- tensor.h
- tensor_format.cc
- tensor_format.h
- tensor_shape.cc
- tensor_shape.h
- tensor_shape_utils.cc
- tensor_shape_utils.h
- tensor_types.h
- training_op_helpers.cc
- training_op_helpers.h
- types.cc
- types.h
- variable_lock.cc
- variable_lock.h
- wide_char.h
- xplane_builder.cc
- xplane_builder.h
- build_wheel.py
- MANIFEST.in
- README
- setup.py
- template_init.py
- tfdml.natvis
- microsofttelemetry.h
- .clang-format
- .gitattributes
- .gitignore
- .pylintrc
- __init__.py
- BUILD.md
- build.py
- CMakeLists.txt
- CODE_OF_CONDUCT.md
- CONTRIBUTING.md
- format.ps1
- format.sh
- generate_op_defs_core.py
- KERNEL_COOKBOOK.md
- LICENSE
- NOTICE.txt
- pyproject.toml
- README.md
- requirements.txt
- SECURITY.md
- tfdml.def
# 설치 가이드
1. 코드 내려받기
git clone https://github.com/microsoft/tensorflow-directml-plugin
깃허브에서 프로젝트 코드 전체를 내 컴퓨터로 내려받습니다.
cd tensorflow-directml-plugin
방금 내려받은 프로젝트 폴더 안으로 이동합니다.
2. 공식 설치 스크립트
쉬움 추천사전 준비물
- Python 3 pip 명령어를 쓰려면 Python이 필요합니다.
pip install tensorflow-directml-plugin
PyPI에 배포된 패키지를 바로 설치합니다. 소스 클론이 필요 없습니다.
설치 후 새 터미널을 열고, 프로그램의 버전 확인 명령(예: --version)으로 정상 설치됐는지 확인하세요.
이 레포의 README에 적힌 실제 명령어를 그대로 가져왔습니다.
3. CMake
보통사전 준비물
mkdir build && cd build
빌드 결과물을 담을 폴더를 만들고 그 안으로 이동합니다.
cmake ..
소스코드를 분석해 빌드 설정 파일을 생성합니다 (build 폴더 안에서 실행해야 함).
make
생성된 빌드 설정을 바탕으로 실제 컴파일을 진행해 실행 파일을 만듭니다.
build 폴더 안에 실행 파일이 생성됐는지 확인하고, 직접 실행해보세요 (예: ./build/앱이름).
// repository documentation
Was this content helpful?
(0 ratings)
