mindspore 2.4.10__cp311-cp311-win_amd64.whl → 2.6.0__cp311-cp311-win_amd64.whl
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Potentially problematic release.
This version of mindspore might be problematic. Click here for more details.
- mindspore/.commit_id +1 -1
- mindspore/Microsoft.VisualStudio.Telemetry.dll +0 -0
- mindspore/Newtonsoft.Json.dll +0 -0
- mindspore/__init__.py +13 -6
- mindspore/_c_dataengine.cp311-win_amd64.pyd +0 -0
- mindspore/_c_expression.cp311-win_amd64.pyd +0 -0
- mindspore/_c_mindrecord.cp311-win_amd64.pyd +0 -0
- mindspore/_check_jit_forbidden_api.py +3 -0
- mindspore/_checkparam.py +3 -38
- mindspore/_deprecated/__init__.py +17 -0
- mindspore/_deprecated/jit.py +198 -0
- mindspore/_extends/builtin_operations.py +1 -1
- mindspore/_extends/parallel_compile/akg_compiler/gen_custom_op_files.py +1 -1
- mindspore/_extends/parse/__init__.py +6 -7
- mindspore/_extends/parse/compile_config.py +83 -0
- mindspore/_extends/parse/deprecated/__init__.py +0 -0
- mindspore/_extends/parse/deprecated/deprecated_tensor_method.py +394 -0
- mindspore/_extends/parse/jit_fallback_modules/__init__.py +0 -0
- mindspore/_extends/parse/jit_fallback_modules/check_utils.py +123 -0
- mindspore/_extends/parse/jit_fallback_modules/third_party_modules.py +50 -0
- mindspore/_extends/parse/parser.py +47 -198
- mindspore/_extends/parse/resources.py +1 -5
- mindspore/_extends/parse/standard_method.py +229 -99
- mindspore/_extends/pijit/__init__.py +2 -2
- mindspore/_extends/pijit/pijit_func_white_list.py +17 -12
- mindspore/_extends/pijit/tensor_func_list.py +27 -0
- mindspore/_extends/utils.py +1 -1
- mindspore/amp.py +11 -5
- mindspore/atlprov.dll +0 -0
- mindspore/avcodec-59.dll +0 -0
- mindspore/avdevice-59.dll +0 -0
- mindspore/avfilter-8.dll +0 -0
- mindspore/avformat-59.dll +0 -0
- mindspore/avutil-57.dll +0 -0
- mindspore/boost/__init__.py +2 -2
- mindspore/boost/base.py +3 -7
- mindspore/boost/boost_cell_wrapper.py +138 -43
- mindspore/c1.dll +0 -0
- mindspore/c1xx.dll +0 -0
- mindspore/c2.dll +0 -0
- mindspore/common/__init__.py +6 -3
- mindspore/common/_grad_function.py +56 -0
- mindspore/common/_pijit_context.py +14 -5
- mindspore/common/_register_for_tensor.py +1 -2
- mindspore/common/_stub_tensor.py +30 -14
- mindspore/common/_tensor_cpp_method.py +17 -0
- mindspore/common/_tensor_docs.py +4760 -0
- mindspore/common/api.py +480 -372
- mindspore/common/auto_dynamic_shape.py +41 -44
- mindspore/common/dtype.py +39 -36
- mindspore/common/dump.py +9 -6
- mindspore/common/file_system.py +9 -1
- mindspore/common/generator.py +5 -0
- mindspore/common/hook_handle.py +6 -2
- mindspore/common/initializer.py +13 -10
- mindspore/common/jit_begin_end.py +94 -0
- mindspore/common/jit_config.py +6 -1
- mindspore/common/jit_context.py +76 -0
- mindspore/common/jit_trace.py +378 -0
- mindspore/common/lazy_inline.py +9 -3
- mindspore/common/mindir_util.py +10 -2
- mindspore/common/mutable.py +5 -4
- mindspore/common/parameter.py +135 -52
- mindspore/common/seed.py +2 -2
- mindspore/common/sparse_tensor.py +23 -17
- mindspore/common/tensor.py +975 -1981
- mindspore/communication/__init__.py +7 -5
- mindspore/communication/_comm_helper.py +52 -2
- mindspore/communication/comm_func.py +240 -181
- mindspore/communication/management.py +95 -26
- mindspore/context.py +324 -573
- mindspore/dataset/__init__.py +65 -37
- mindspore/dataset/audio/__init__.py +2 -8
- mindspore/dataset/audio/transforms.py +3 -17
- mindspore/dataset/callback/ds_callback.py +2 -1
- mindspore/dataset/core/config.py +87 -6
- mindspore/dataset/engine/cache_admin.py +3 -3
- mindspore/dataset/engine/cache_client.py +6 -5
- mindspore/dataset/engine/datasets.py +292 -267
- mindspore/dataset/engine/datasets_audio.py +22 -8
- mindspore/dataset/engine/datasets_standard_format.py +46 -27
- mindspore/dataset/engine/datasets_text.py +78 -48
- mindspore/dataset/engine/datasets_user_defined.py +183 -117
- mindspore/dataset/engine/datasets_vision.py +120 -44
- mindspore/dataset/engine/iterators.py +283 -63
- mindspore/dataset/engine/obs/obs_mindrecord_dataset.py +1 -1
- mindspore/dataset/engine/obs/util.py +8 -0
- mindspore/dataset/engine/queue.py +40 -0
- mindspore/dataset/engine/samplers.py +289 -43
- mindspore/dataset/engine/serializer_deserializer.py +3 -2
- mindspore/dataset/engine/validators.py +53 -11
- mindspore/dataset/text/__init__.py +7 -6
- mindspore/dataset/text/transforms.py +6 -5
- mindspore/dataset/text/utils.py +3 -3
- mindspore/dataset/transforms/__init__.py +0 -9
- mindspore/dataset/transforms/py_transforms_util.py +17 -0
- mindspore/dataset/transforms/transforms.py +31 -14
- mindspore/dataset/utils/browse_dataset.py +1 -1
- mindspore/dataset/vision/__init__.py +2 -9
- mindspore/dataset/vision/transforms.py +202 -158
- mindspore/dataset/vision/utils.py +7 -5
- mindspore/dataset/vision/validators.py +1 -2
- mindspore/device_context/__init__.py +21 -0
- mindspore/device_context/ascend/__init__.py +25 -0
- mindspore/device_context/ascend/device.py +72 -0
- mindspore/device_context/ascend/op_debug.py +153 -0
- mindspore/device_context/ascend/op_precision.py +193 -0
- mindspore/device_context/ascend/op_tuning.py +123 -0
- mindspore/{ops_generate/gen_constants.py → device_context/cpu/__init__.py} +6 -17
- mindspore/device_context/cpu/device.py +62 -0
- mindspore/device_context/cpu/op_tuning.py +43 -0
- mindspore/device_context/gpu/__init__.py +21 -0
- mindspore/device_context/gpu/device.py +70 -0
- mindspore/device_context/gpu/op_precision.py +67 -0
- mindspore/device_context/gpu/op_tuning.py +175 -0
- mindspore/device_manager.py +170 -0
- mindspore/dnnl.dll +0 -0
- mindspore/dpcmi.dll +0 -0
- mindspore/experimental/es/embedding_service.py +35 -27
- mindspore/experimental/llm_boost/__init__.py +1 -0
- mindspore/experimental/llm_boost/ascend_native/__init__.py +22 -0
- mindspore/experimental/llm_boost/ascend_native/llama_boost_ascend_native.py +209 -0
- mindspore/experimental/llm_boost/ascend_native/llm_boost.py +52 -0
- mindspore/experimental/llm_boost/atb/boost_base.py +2 -3
- mindspore/experimental/llm_boost/atb/llama_boost.py +6 -1
- mindspore/experimental/llm_boost/register.py +1 -0
- mindspore/experimental/map_parameter.py +4 -4
- mindspore/experimental/optim/adadelta.py +6 -6
- mindspore/experimental/optim/adagrad.py +4 -4
- mindspore/experimental/optim/adam.py +7 -0
- mindspore/experimental/optim/adamax.py +4 -4
- mindspore/experimental/optim/adamw.py +4 -0
- mindspore/experimental/optim/asgd.py +1 -1
- mindspore/experimental/optim/lr_scheduler.py +73 -46
- mindspore/experimental/optim/radam.py +34 -31
- mindspore/experimental/optim/rprop.py +1 -1
- mindspore/experimental/optim/sgd.py +1 -1
- mindspore/hal/contiguous_tensors_handle.py +6 -10
- mindspore/hal/device.py +55 -53
- mindspore/hal/event.py +52 -52
- mindspore/hal/memory.py +179 -120
- mindspore/hal/stream.py +150 -109
- mindspore/include/api/context.h +0 -1
- mindspore/include/dataset/constants.h +7 -4
- mindspore/include/dataset/execute.h +2 -2
- mindspore/jpeg62.dll +0 -0
- mindspore/log.py +50 -0
- mindspore/mindrecord/__init__.py +21 -8
- mindspore/mindrecord/config.py +17 -316
- mindspore/mindrecord/filereader.py +1 -9
- mindspore/mindrecord/filewriter.py +5 -15
- mindspore/mindrecord/mindpage.py +1 -9
- mindspore/mindspore_backend_common.dll +0 -0
- mindspore/mindspore_backend_manager.dll +0 -0
- mindspore/mindspore_common.dll +0 -0
- mindspore/mindspore_core.dll +0 -0
- mindspore/mindspore_dump.dll +0 -0
- mindspore/mindspore_frontend.dll +0 -0
- mindspore/mindspore_glog.dll +0 -0
- mindspore/mindspore_memory_pool.dll +0 -0
- mindspore/mindspore_ms_backend.dll +0 -0
- mindspore/mindspore_ops.dll +0 -0
- mindspore/{mindspore_backend.dll → mindspore_ops_host.dll} +0 -0
- mindspore/mindspore_ops_kernel_common.dll +0 -0
- mindspore/mindspore_profiler.dll +0 -0
- mindspore/mindspore_pyboost.dll +0 -0
- mindspore/mindspore_pynative.dll +0 -0
- mindspore/mindspore_res_manager.dll +0 -0
- mindspore/mindspore_runtime_pipeline.dll +0 -0
- mindspore/mint/__init__.py +798 -761
- mindspore/mint/distributed/__init__.py +70 -4
- mindspore/mint/distributed/distributed.py +2679 -44
- mindspore/mint/linalg/__init__.py +8 -0
- mindspore/mint/nn/__init__.py +743 -22
- mindspore/mint/nn/functional.py +716 -23
- mindspore/mint/nn/layer/__init__.py +21 -4
- mindspore/mint/nn/layer/_functions.py +334 -0
- mindspore/mint/nn/layer/activation.py +276 -1
- mindspore/mint/nn/layer/basic.py +123 -0
- mindspore/mint/nn/layer/conv.py +933 -0
- mindspore/mint/nn/layer/normalization.py +223 -28
- mindspore/mint/nn/layer/padding.py +797 -0
- mindspore/mint/nn/layer/pooling.py +235 -0
- mindspore/mint/optim/__init__.py +3 -1
- mindspore/mint/optim/adam.py +223 -0
- mindspore/mint/optim/adamw.py +26 -19
- mindspore/mint/optim/sgd.py +171 -0
- mindspore/mint/special/__init__.py +2 -1
- mindspore/msobj140.dll +0 -0
- mindspore/mspdb140.dll +0 -0
- mindspore/mspdbcore.dll +0 -0
- mindspore/mspdbst.dll +0 -0
- mindspore/mspft140.dll +0 -0
- mindspore/msvcdis140.dll +0 -0
- mindspore/msvcp140_1.dll +0 -0
- mindspore/msvcp140_2.dll +0 -0
- mindspore/msvcp140_atomic_wait.dll +0 -0
- mindspore/msvcp140_codecvt_ids.dll +0 -0
- mindspore/multiprocessing/__init__.py +5 -0
- mindspore/nn/__init__.py +4 -1
- mindspore/nn/cell.py +1373 -192
- mindspore/nn/dynamic_lr.py +2 -1
- mindspore/nn/layer/activation.py +29 -27
- mindspore/nn/layer/basic.py +51 -35
- mindspore/nn/layer/channel_shuffle.py +3 -3
- mindspore/nn/layer/container.py +1 -1
- mindspore/nn/layer/conv.py +53 -42
- mindspore/nn/layer/embedding.py +12 -11
- mindspore/nn/layer/normalization.py +56 -49
- mindspore/nn/layer/padding.py +4 -3
- mindspore/nn/layer/pooling.py +120 -42
- mindspore/nn/layer/rnn_cells.py +1 -1
- mindspore/nn/layer/rnns.py +2 -1
- mindspore/nn/layer/timedistributed.py +5 -5
- mindspore/nn/layer/transformer.py +59 -36
- mindspore/nn/learning_rate_schedule.py +8 -4
- mindspore/nn/loss/loss.py +58 -55
- mindspore/nn/optim/ada_grad.py +7 -5
- mindspore/nn/optim/adadelta.py +11 -9
- mindspore/nn/optim/adafactor.py +1 -1
- mindspore/nn/optim/adam.py +19 -15
- mindspore/nn/optim/adamax.py +8 -7
- mindspore/nn/optim/adasum.py +5 -5
- mindspore/nn/optim/asgd.py +3 -1
- mindspore/nn/optim/ftrl.py +11 -9
- mindspore/nn/optim/lamb.py +1 -1
- mindspore/nn/optim/lars.py +1 -4
- mindspore/nn/optim/lazyadam.py +12 -10
- mindspore/nn/optim/momentum.py +7 -6
- mindspore/nn/optim/optimizer.py +3 -3
- mindspore/nn/optim/proximal_ada_grad.py +12 -10
- mindspore/nn/optim/rmsprop.py +13 -12
- mindspore/nn/optim/rprop.py +11 -9
- mindspore/nn/optim/sgd.py +9 -6
- mindspore/nn/optim/tft_wrapper.py +5 -2
- mindspore/nn/optim/thor.py +2 -1
- mindspore/nn/probability/bijector/bijector.py +17 -11
- mindspore/nn/probability/bijector/gumbel_cdf.py +5 -5
- mindspore/nn/probability/bijector/invert.py +2 -2
- mindspore/nn/probability/bijector/scalar_affine.py +3 -3
- mindspore/nn/probability/bijector/softplus.py +3 -2
- mindspore/nn/probability/distribution/beta.py +3 -3
- mindspore/nn/probability/distribution/categorical.py +1 -1
- mindspore/nn/probability/distribution/cauchy.py +4 -2
- mindspore/nn/probability/distribution/exponential.py +6 -7
- mindspore/nn/probability/distribution/gamma.py +2 -2
- mindspore/nn/probability/distribution/gumbel.py +2 -2
- mindspore/nn/probability/distribution/half_normal.py +5 -3
- mindspore/nn/probability/distribution/logistic.py +5 -3
- mindspore/nn/probability/distribution/poisson.py +1 -1
- mindspore/nn/probability/distribution/uniform.py +5 -3
- mindspore/nn/reinforcement/_tensors_queue.py +1 -1
- mindspore/nn/reinforcement/tensor_array.py +1 -1
- mindspore/nn/utils/init.py +13 -11
- mindspore/nn/wrap/__init__.py +6 -6
- mindspore/nn/wrap/cell_wrapper.py +181 -122
- mindspore/nn/wrap/grad_reducer.py +45 -36
- mindspore/nn/wrap/loss_scale.py +6 -7
- mindspore/numpy/array_creations.py +63 -65
- mindspore/numpy/array_ops.py +149 -144
- mindspore/numpy/logic_ops.py +41 -42
- mindspore/numpy/math_ops.py +361 -359
- mindspore/numpy/utils.py +17 -18
- mindspore/numpy/utils_const.py +5 -6
- mindspore/opencv_core452.dll +0 -0
- mindspore/opencv_imgcodecs452.dll +0 -0
- mindspore/opencv_imgproc452.dll +0 -0
- mindspore/ops/__init__.py +5 -3
- mindspore/ops/_grad_experimental/grad_comm_ops.py +112 -16
- mindspore/ops/_grad_experimental/grad_debug_ops.py +14 -2
- mindspore/ops/_grad_experimental/grad_inner_ops.py +9 -0
- mindspore/ops/_grad_experimental/grad_math_ops.py +2 -1
- mindspore/ops/_grad_experimental/taylor_rule.py +29 -0
- mindspore/ops/_op_impl/cpu/__init__.py +1 -0
- mindspore/ops/_op_impl/cpu/raise_op.py +28 -0
- mindspore/ops/_register_for_op.py +0 -11
- mindspore/{ops_generate → ops/_utils}/arg_dtype_cast.py +123 -4
- mindspore/{ops_generate → ops/_utils}/arg_handler.py +3 -65
- mindspore/ops/_vmap/vmap_array_ops.py +52 -25
- mindspore/ops/_vmap/vmap_base.py +0 -2
- mindspore/ops/_vmap/vmap_grad_nn_ops.py +21 -14
- mindspore/ops/_vmap/vmap_math_ops.py +15 -16
- mindspore/ops/_vmap/vmap_nn_ops.py +29 -42
- mindspore/ops/auto_generate/__init__.py +4 -3
- mindspore/ops/auto_generate/cpp_create_prim_instance_helper.py +258 -46
- mindspore/ops/auto_generate/gen_extend_func.py +757 -185
- mindspore/ops/auto_generate/gen_ops_def.py +4197 -2243
- mindspore/ops/auto_generate/gen_ops_prim.py +16976 -6055
- mindspore/ops/auto_generate/pyboost_inner_prim.py +221 -87
- mindspore/ops/composite/__init__.py +2 -1
- mindspore/ops/composite/base.py +20 -25
- mindspore/ops/composite/math_ops.py +6 -16
- mindspore/ops/composite/multitype_ops/__init__.py +5 -2
- mindspore/ops/composite/multitype_ops/_compile_utils.py +228 -30
- mindspore/ops/composite/multitype_ops/_constexpr_utils.py +1 -2
- mindspore/ops/composite/multitype_ops/add_impl.py +2 -1
- mindspore/ops/composite/multitype_ops/bitwise_and_impl.py +2 -1
- mindspore/ops/composite/multitype_ops/bitwise_or_impl.py +2 -1
- mindspore/ops/composite/multitype_ops/bitwise_xor_impl.py +2 -1
- mindspore/ops/composite/multitype_ops/div_impl.py +6 -4
- mindspore/ops/composite/multitype_ops/equal_impl.py +4 -3
- mindspore/ops/composite/multitype_ops/floordiv_impl.py +2 -1
- mindspore/ops/composite/multitype_ops/getitem_impl.py +3 -2
- mindspore/ops/composite/multitype_ops/greater_equal_impl.py +4 -3
- mindspore/ops/composite/multitype_ops/greater_impl.py +4 -3
- mindspore/ops/composite/multitype_ops/in_impl.py +2 -1
- mindspore/ops/composite/multitype_ops/invert_impl.py +50 -0
- mindspore/ops/composite/multitype_ops/left_shift_impl.py +2 -1
- mindspore/ops/composite/multitype_ops/less_equal_impl.py +4 -3
- mindspore/ops/composite/multitype_ops/less_impl.py +4 -3
- mindspore/ops/composite/multitype_ops/logic_not_impl.py +3 -2
- mindspore/ops/composite/multitype_ops/logical_and_impl.py +2 -1
- mindspore/ops/composite/multitype_ops/logical_or_impl.py +2 -1
- mindspore/ops/composite/multitype_ops/mod_impl.py +2 -1
- mindspore/ops/composite/multitype_ops/mul_impl.py +3 -2
- mindspore/ops/composite/multitype_ops/negative_impl.py +2 -1
- mindspore/ops/composite/multitype_ops/not_equal_impl.py +2 -1
- mindspore/ops/composite/multitype_ops/not_in_impl.py +2 -1
- mindspore/ops/composite/multitype_ops/ones_like_impl.py +18 -0
- mindspore/ops/composite/multitype_ops/pow_impl.py +2 -30
- mindspore/ops/composite/multitype_ops/right_shift_impl.py +2 -1
- mindspore/ops/composite/multitype_ops/setitem_impl.py +2 -1
- mindspore/ops/composite/multitype_ops/sub_impl.py +2 -1
- mindspore/ops/function/__init__.py +40 -2
- mindspore/ops/function/_add_attr_func.py +58 -0
- mindspore/ops/function/array_func.py +2089 -2403
- mindspore/ops/function/clip_func.py +80 -23
- mindspore/ops/function/debug_func.py +57 -57
- mindspore/ops/function/grad/__init__.py +1 -0
- mindspore/ops/function/grad/grad_func.py +104 -71
- mindspore/ops/function/image_func.py +2 -2
- mindspore/ops/function/linalg_func.py +47 -78
- mindspore/ops/function/math_func.py +4351 -3813
- mindspore/ops/function/nn_func.py +1712 -637
- mindspore/ops/function/other_func.py +159 -1
- mindspore/ops/function/parameter_func.py +18 -84
- mindspore/ops/function/random_func.py +452 -387
- mindspore/ops/function/reshard_func.py +4 -70
- mindspore/ops/function/sparse_func.py +3 -3
- mindspore/ops/function/sparse_unary_func.py +6 -6
- mindspore/ops/function/spectral_func.py +25 -58
- mindspore/ops/function/vmap_func.py +26 -18
- mindspore/ops/functional.py +23 -7
- mindspore/ops/functional_overload.py +1548 -0
- mindspore/ops/op_info_register.py +32 -244
- mindspore/ops/operations/__init__.py +23 -15
- mindspore/ops/operations/_custom_ops_utils.py +235 -0
- mindspore/ops/operations/_embedding_cache_ops.py +4 -4
- mindspore/ops/operations/_grad_ops.py +2 -43
- mindspore/ops/operations/_infer_ops.py +2 -1
- mindspore/ops/operations/_inner_ops.py +43 -84
- mindspore/ops/operations/_ms_kernel.py +4 -10
- mindspore/ops/operations/_rl_inner_ops.py +1 -1
- mindspore/ops/operations/_scalar_ops.py +3 -2
- mindspore/ops/operations/_sequence_ops.py +1 -1
- mindspore/ops/operations/_tensor_array.py +1 -1
- mindspore/ops/operations/array_ops.py +81 -324
- mindspore/ops/operations/comm_ops.py +154 -108
- mindspore/ops/operations/custom_ops.py +298 -87
- mindspore/ops/operations/debug_ops.py +157 -59
- mindspore/ops/operations/inner_ops.py +7 -5
- mindspore/ops/operations/linalg_ops.py +1 -57
- mindspore/ops/operations/manually_defined/_inner.py +1 -1
- mindspore/ops/operations/manually_defined/ops_def.py +928 -180
- mindspore/ops/operations/math_ops.py +32 -234
- mindspore/ops/operations/nn_ops.py +212 -531
- mindspore/ops/operations/other_ops.py +62 -9
- mindspore/ops/operations/random_ops.py +13 -7
- mindspore/ops/operations/reshard_ops.py +1 -1
- mindspore/ops/operations/sparse_ops.py +2 -2
- mindspore/ops/primitive.py +66 -53
- mindspore/ops/tensor_method.py +1895 -0
- mindspore/ops_generate/__init__.py +0 -5
- mindspore/ops_generate/aclnn/__init__.py +0 -0
- mindspore/ops_generate/aclnn/aclnn_kernel_register_auto_cc_generator.py +135 -0
- mindspore/ops_generate/aclnn/gen_aclnn_implement.py +257 -0
- mindspore/ops_generate/api/__init__.py +0 -0
- mindspore/ops_generate/api/add_tensor_docs_generator.py +56 -0
- mindspore/ops_generate/api/cpp_create_prim_instance_helper_generator.py +105 -0
- mindspore/ops_generate/api/functional_map_cpp_generator.py +504 -0
- mindspore/ops_generate/api/functional_overload_py_generator.py +112 -0
- mindspore/ops_generate/api/functions_cc_generator.py +237 -0
- mindspore/ops_generate/api/gen_api.py +103 -0
- mindspore/ops_generate/api/op_api_proto.py +235 -0
- mindspore/ops_generate/api/tensor_func_reg_cpp_generator.py +461 -0
- mindspore/ops_generate/common/__init__.py +0 -0
- mindspore/ops_generate/common/base_generator.py +11 -0
- mindspore/ops_generate/common/gen_constants.py +91 -0
- mindspore/ops_generate/common/gen_utils.py +348 -0
- mindspore/ops_generate/common/op_proto.py +473 -0
- mindspore/ops_generate/common/template.py +523 -0
- mindspore/ops_generate/gen_ops.py +22 -1069
- mindspore/ops_generate/op_def/__init__.py +0 -0
- mindspore/ops_generate/op_def/gen_op_def.py +90 -0
- mindspore/ops_generate/op_def/lite_ops_cpp_generator.py +191 -0
- mindspore/ops_generate/op_def/ops_def_cc_generator.py +296 -0
- mindspore/ops_generate/op_def/ops_def_h_generator.py +74 -0
- mindspore/ops_generate/op_def/ops_name_h_generator.py +83 -0
- mindspore/ops_generate/op_def/ops_primitive_h_generator.py +125 -0
- mindspore/ops_generate/op_def_py/__init__.py +0 -0
- mindspore/ops_generate/op_def_py/gen_op_def_py.py +47 -0
- mindspore/ops_generate/op_def_py/op_def_py_generator.py +132 -0
- mindspore/ops_generate/op_def_py/op_prim_py_generator.py +489 -0
- mindspore/ops_generate/pyboost/__init__.py +0 -0
- mindspore/ops_generate/pyboost/auto_grad_impl_cc_generator.py +139 -0
- mindspore/ops_generate/pyboost/auto_grad_reg_cc_generator.py +93 -0
- mindspore/ops_generate/pyboost/gen_pyboost_func.py +175 -0
- mindspore/ops_generate/pyboost/op_template_parser.py +517 -0
- mindspore/ops_generate/pyboost/pyboost_functions_cpp_generator.py +407 -0
- mindspore/ops_generate/pyboost/pyboost_functions_h_generator.py +100 -0
- mindspore/ops_generate/pyboost/pyboost_functions_py_generator.py +148 -0
- mindspore/ops_generate/pyboost/pyboost_grad_function_cpp_generator.py +155 -0
- mindspore/ops_generate/pyboost/pyboost_inner_prim_generator.py +132 -0
- mindspore/ops_generate/pyboost/pyboost_native_grad_functions_generator.py +272 -0
- mindspore/ops_generate/pyboost/pyboost_op_cpp_code_generator.py +938 -0
- mindspore/ops_generate/pyboost/pyboost_overload_functions_cpp_generator.py +357 -0
- mindspore/ops_generate/{pyboost_utils.py → pyboost/pyboost_utils.py} +179 -36
- mindspore/ops_generate/resources/__init__.py +0 -0
- mindspore/ops_generate/resources/resource_list.py +30 -0
- mindspore/ops_generate/resources/resource_loader.py +36 -0
- mindspore/ops_generate/resources/resource_manager.py +64 -0
- mindspore/ops_generate/resources/yaml_loader.py +88 -0
- mindspore/ops_generate/tensor_py_cc_generator.py +122 -0
- mindspore/parallel/__init__.py +7 -3
- mindspore/parallel/_auto_parallel_context.py +159 -40
- mindspore/parallel/_cell_wrapper.py +132 -15
- mindspore/parallel/_parallel_serialization.py +107 -5
- mindspore/parallel/_ps_context.py +1 -1
- mindspore/parallel/_recovery_context.py +7 -2
- mindspore/parallel/_tensor.py +142 -18
- mindspore/parallel/_utils.py +199 -23
- mindspore/parallel/algo_parameter_config.py +4 -4
- mindspore/parallel/auto_parallel.py +732 -0
- mindspore/parallel/checkpoint_convert.py +159 -0
- mindspore/parallel/checkpoint_transform.py +700 -35
- mindspore/parallel/cluster/process_entity/_api.py +276 -50
- mindspore/parallel/cluster/process_entity/_utils.py +41 -6
- mindspore/parallel/cluster/run.py +21 -4
- mindspore/parallel/function/__init__.py +24 -0
- mindspore/parallel/function/reshard_func.py +258 -0
- mindspore/parallel/nn/__init__.py +25 -0
- mindspore/parallel/nn/parallel_cell_wrapper.py +263 -0
- mindspore/parallel/nn/parallel_grad_reducer.py +169 -0
- mindspore/parallel/parameter_broadcast.py +25 -14
- mindspore/parallel/shard.py +137 -59
- mindspore/parallel/transform_safetensors.py +364 -305
- mindspore/pgodb140.dll +0 -0
- mindspore/pgort140.dll +0 -0
- mindspore/profiler/__init__.py +22 -5
- mindspore/profiler/analysis/__init__.py +0 -0
- mindspore/profiler/analysis/parser/__init__.py +0 -0
- mindspore/profiler/analysis/parser/ascend_cann_parser.py +170 -0
- mindspore/profiler/analysis/parser/base_parser.py +158 -0
- mindspore/profiler/analysis/parser/framework_cann_relation_parser.py +45 -0
- mindspore/profiler/analysis/parser/ms_framework_parser.py +142 -0
- mindspore/profiler/analysis/parser/ms_minddata_parser.py +145 -0
- mindspore/profiler/analysis/parser/timeline_assembly_factory/__init__.py +0 -0
- mindspore/profiler/analysis/parser/timeline_assembly_factory/ascend_timeline_assembler.py +264 -0
- mindspore/profiler/analysis/parser/timeline_assembly_factory/base_timeline_assembler.py +40 -0
- mindspore/profiler/analysis/parser/timeline_assembly_factory/trace_view_container.py +109 -0
- mindspore/profiler/analysis/parser/timeline_creator/__init__.py +0 -0
- mindspore/profiler/analysis/parser/timeline_creator/base_timeline_creator.py +44 -0
- mindspore/profiler/analysis/parser/timeline_creator/cpu_op_timeline_creator.py +90 -0
- mindspore/profiler/analysis/parser/timeline_creator/fwk_timeline_creator.py +76 -0
- mindspore/profiler/analysis/parser/timeline_creator/msprof_timeline_creator.py +103 -0
- mindspore/profiler/analysis/parser/timeline_creator/scope_layer_timeline_creator.py +134 -0
- mindspore/profiler/analysis/parser/timeline_event/__init__.py +0 -0
- mindspore/profiler/analysis/parser/timeline_event/base_event.py +233 -0
- mindspore/profiler/analysis/parser/timeline_event/cpu_op_event.py +47 -0
- mindspore/profiler/analysis/parser/timeline_event/flow_event.py +36 -0
- mindspore/profiler/analysis/parser/timeline_event/fwk_event.py +415 -0
- mindspore/profiler/analysis/parser/timeline_event/msprof_event.py +73 -0
- mindspore/profiler/analysis/parser/timeline_event/scope_layer_event.py +53 -0
- mindspore/profiler/analysis/parser/timeline_event/timeline_event_pool.py +146 -0
- mindspore/profiler/analysis/task_manager.py +131 -0
- mindspore/profiler/analysis/time_converter.py +84 -0
- mindspore/profiler/analysis/viewer/__init__.py +0 -0
- mindspore/profiler/analysis/viewer/ascend_communication_viewer.py +372 -0
- mindspore/profiler/analysis/viewer/ascend_integrate_viewer.py +87 -0
- mindspore/profiler/analysis/viewer/ascend_kernel_details_viewer.py +250 -0
- mindspore/profiler/analysis/viewer/ascend_memory_viewer.py +320 -0
- mindspore/profiler/analysis/viewer/ascend_op_memory_viewer.py +327 -0
- mindspore/profiler/analysis/viewer/ascend_step_trace_time_viewer.py +376 -0
- mindspore/profiler/analysis/viewer/ascend_timeline_viewer.py +58 -0
- mindspore/profiler/analysis/viewer/base_viewer.py +26 -0
- mindspore/profiler/analysis/viewer/ms_dataset_viewer.py +96 -0
- mindspore/profiler/analysis/viewer/ms_minddata_viewer.py +581 -0
- mindspore/profiler/analysis/work_flow.py +73 -0
- mindspore/profiler/common/ascend_msprof_exporter.py +139 -0
- mindspore/profiler/common/command_executor.py +90 -0
- mindspore/profiler/common/constant.py +186 -3
- mindspore/profiler/common/file_manager.py +208 -0
- mindspore/profiler/common/log.py +130 -0
- mindspore/profiler/common/msprof_cmd_tool.py +221 -0
- mindspore/profiler/common/path_manager.py +395 -0
- mindspore/profiler/common/process_bar.py +168 -0
- mindspore/profiler/common/process_pool.py +9 -3
- mindspore/profiler/common/profiler_context.py +500 -0
- mindspore/profiler/common/profiler_info.py +304 -0
- mindspore/profiler/common/profiler_meta_data.py +74 -0
- mindspore/profiler/common/profiler_output_path.py +284 -0
- mindspore/profiler/common/profiler_parameters.py +251 -0
- mindspore/profiler/common/profiler_path_manager.py +179 -0
- mindspore/profiler/common/record_function.py +76 -0
- mindspore/profiler/common/tlv_decoder.py +76 -0
- mindspore/profiler/common/util.py +75 -2
- mindspore/profiler/dynamic_profiler.py +341 -75
- mindspore/profiler/envprofiler.py +163 -0
- mindspore/profiler/experimental_config.py +197 -0
- mindspore/profiler/mstx.py +242 -0
- mindspore/profiler/platform/__init__.py +21 -0
- mindspore/profiler/platform/base_profiler.py +40 -0
- mindspore/profiler/platform/cpu_profiler.py +124 -0
- mindspore/profiler/platform/gpu_profiler.py +74 -0
- mindspore/profiler/platform/npu_profiler.py +335 -0
- mindspore/profiler/profiler.py +1073 -90
- mindspore/profiler/profiler_action_controller.py +187 -0
- mindspore/profiler/profiler_interface.py +118 -0
- mindspore/profiler/schedule.py +243 -0
- mindspore/rewrite/api/node.py +15 -13
- mindspore/rewrite/api/symbol_tree.py +2 -3
- mindspore/run_check/_check_version.py +27 -20
- mindspore/run_check/run_check.py +1 -1
- mindspore/runtime/__init__.py +37 -0
- mindspore/runtime/device.py +27 -0
- mindspore/runtime/event.py +209 -0
- mindspore/runtime/executor.py +177 -0
- mindspore/runtime/memory.py +416 -0
- mindspore/runtime/stream.py +460 -0
- mindspore/runtime/thread_bind_core.py +401 -0
- mindspore/safeguard/rewrite_obfuscation.py +12 -9
- mindspore/swresample-4.dll +0 -0
- mindspore/swscale-6.dll +0 -0
- mindspore/tbbmalloc.dll +0 -0
- mindspore/tinyxml2.dll +0 -0
- mindspore/train/__init__.py +8 -8
- mindspore/train/_utils.py +96 -27
- mindspore/train/amp.py +9 -5
- mindspore/train/callback/__init__.py +2 -2
- mindspore/train/callback/_callback.py +2 -16
- mindspore/train/callback/_checkpoint.py +53 -55
- mindspore/train/callback/_cluster_monitor.py +14 -18
- mindspore/train/callback/_early_stop.py +1 -1
- mindspore/train/callback/_flops_collector.py +103 -68
- mindspore/train/callback/_history.py +8 -5
- mindspore/train/callback/_lambda_callback.py +2 -2
- mindspore/train/callback/_landscape.py +0 -3
- mindspore/train/callback/_loss_monitor.py +2 -1
- mindspore/train/callback/_on_request_exit.py +6 -5
- mindspore/train/callback/_reduce_lr_on_plateau.py +11 -6
- mindspore/train/callback/_summary_collector.py +52 -19
- mindspore/train/callback/_time_monitor.py +2 -1
- mindspore/train/callback/{_tft_register.py → _train_fault_tolerance.py} +228 -108
- mindspore/train/data_sink.py +25 -2
- mindspore/train/dataset_helper.py +15 -16
- mindspore/train/loss_scale_manager.py +8 -7
- mindspore/train/metrics/accuracy.py +3 -3
- mindspore/train/metrics/confusion_matrix.py +9 -9
- mindspore/train/metrics/error.py +3 -3
- mindspore/train/metrics/hausdorff_distance.py +4 -4
- mindspore/train/metrics/mean_surface_distance.py +3 -3
- mindspore/train/metrics/metric.py +0 -12
- mindspore/train/metrics/occlusion_sensitivity.py +4 -2
- mindspore/train/metrics/precision.py +11 -10
- mindspore/train/metrics/recall.py +9 -9
- mindspore/train/metrics/root_mean_square_surface_distance.py +2 -2
- mindspore/train/mind_ir_pb2.py +174 -46
- mindspore/train/model.py +269 -136
- mindspore/train/serialization.py +622 -978
- mindspore/train/summary/_summary_adapter.py +2 -2
- mindspore/train/summary/summary_record.py +2 -3
- mindspore/train/train_thor/model_thor.py +1 -1
- mindspore/turbojpeg.dll +0 -0
- mindspore/utils/__init__.py +6 -3
- mindspore/utils/dryrun.py +140 -0
- mindspore/utils/hooks.py +81 -0
- mindspore/utils/runtime_execution_order_check.py +552 -0
- mindspore/utils/utils.py +138 -4
- mindspore/vcmeta.dll +0 -0
- mindspore/vcruntime140.dll +0 -0
- mindspore/vcruntime140_1.dll +0 -0
- mindspore/version.py +1 -1
- {mindspore-2.4.10.dist-info → mindspore-2.6.0.dist-info}/METADATA +3 -3
- {mindspore-2.4.10.dist-info → mindspore-2.6.0.dist-info}/RECORD +587 -418
- {mindspore-2.4.10.dist-info → mindspore-2.6.0.dist-info}/entry_points.txt +1 -1
- mindspore/_install_custom.py +0 -43
- mindspore/common/_register_for_adapter.py +0 -74
- mindspore/common/_tensor_overload.py +0 -139
- mindspore/mindspore_np_dtype.dll +0 -0
- mindspore/ops/auto_generate/gen_arg_dtype_cast.py +0 -252
- mindspore/ops/auto_generate/gen_arg_handler.py +0 -197
- mindspore/ops/operations/_opaque_predicate_registry.py +0 -41
- mindspore/ops_generate/gen_aclnn_implement.py +0 -263
- mindspore/ops_generate/gen_ops_inner_prim.py +0 -131
- mindspore/ops_generate/gen_pyboost_func.py +0 -1052
- mindspore/ops_generate/gen_utils.py +0 -209
- mindspore/ops_generate/op_proto.py +0 -145
- mindspore/ops_generate/template.py +0 -261
- mindspore/profiler/envprofiling.py +0 -254
- mindspore/profiler/profiling.py +0 -1926
- {mindspore-2.4.10.dist-info → mindspore-2.6.0.dist-info}/WHEEL +0 -0
- {mindspore-2.4.10.dist-info → mindspore-2.6.0.dist-info}/top_level.txt +0 -0
|
@@ -17,10 +17,18 @@
|
|
|
17
17
|
from mindspore.common import dtype as mstype
|
|
18
18
|
|
|
19
19
|
op_args_default_value = {
|
|
20
|
-
"AdamWeightDecay": {"use_locking": False},
|
|
21
20
|
"AdamW": {"amsgrad": False, "maximize": False},
|
|
21
|
+
"AdamWeightDecay": {"use_locking": False},
|
|
22
|
+
"Addbmm": {"beta": 1, "alpha": 1},
|
|
23
|
+
"AddcdivExt": {"value": 1},
|
|
24
|
+
"AddcmulExt": {"value": 1},
|
|
25
|
+
"Addmm": {"beta": 1, "alpha": 1},
|
|
26
|
+
"Addmv": {"beta": 1, "alpha": 1},
|
|
22
27
|
"AddExt": {"alpha": 1},
|
|
23
28
|
"AddLayerNormV2": {"epsilon": 1e-5, "additionalOut": False},
|
|
29
|
+
"AddRmsNorm": {"epsilon": 1e-6},
|
|
30
|
+
"AddScalar": {"alpha": 1},
|
|
31
|
+
"AllGatherMatmul": {"bias": None, "gather_index": 0, "gather_output": True, "comm_turn": 0, "trans_input": False, "trans_x2": False},
|
|
24
32
|
"ApplyAdamW": {"max_grad_norm": None, "amsgrad": False, "maximize": False},
|
|
25
33
|
"ApplyCamePart2": {"sum_r": None, "global_shape": None},
|
|
26
34
|
"ApplyCamePart3": {"global_shape": None, "use_first_moment": False},
|
|
@@ -33,16 +41,25 @@ op_args_default_value = {
|
|
|
33
41
|
"ArgMinExt": {"dim": None, "keepdim": False},
|
|
34
42
|
"Argmin": {"axis": -1, "output_type": mstype.int32},
|
|
35
43
|
"ArgMinWithValue": {"axis": 0, "keep_dims": False},
|
|
44
|
+
"ArgSort": {"dim": -1, "descending": False, "stable": False},
|
|
45
|
+
"AsStrided": {"storage_offset": 0},
|
|
46
|
+
"AvgPool1D": {"stride": None, "padding": 0, "ceil_mode": False, "count_include_pad": True},
|
|
36
47
|
"AvgPool2DGrad": {"padding": 0, "ceil_mode": False, "count_include_pad": True, "divisor_override": None},
|
|
37
48
|
"AvgPool2D": {"padding": 0, "ceil_mode": False, "count_include_pad": True, "divisor_override": None},
|
|
49
|
+
"AvgPool3DExt": {"stride": None, "padding": 0, "ceil_mode": False, "count_include_pad": True, "divisor_override": None},
|
|
50
|
+
"AvgPool3DGradExt": {"stride": None, "padding": 0, "ceil_mode": False, "count_include_pad": True, "divisor_override": None},
|
|
38
51
|
"AvgPoolGrad": {"kernel_size": 1, "strides": 1, "pad_mode": 'VALID', "data_format": 'NCHW'},
|
|
39
52
|
"AvgPool": {"kernel_size": 1, "strides": 1, "pad_mode": 'VALID', "data_format": 'NCHW'},
|
|
53
|
+
"Baddbmm": {"beta": 1, "alpha": 1},
|
|
40
54
|
"BatchMatMul": {"transpose_a": False, "transpose_b": False},
|
|
55
|
+
"BatchNormElemt": {"weight": None, "bias": None, "mean": None, "invstd": None, "eps": 1e-5},
|
|
41
56
|
"BatchNormExt": {"running_mean": None, "runnning_var": None, "training": False, "momentum": 0.1, "epsilon": 1e-5},
|
|
42
|
-
"
|
|
57
|
+
"BatchNormGatherStatsWithCounts": {"running_mean": None, "running_var": None, "momentum": 1e-1, "eps": 1e-5, "counts": None},
|
|
58
|
+
"BatchNormGradExt": {"running_mean": None, "running_var": None, "saved_mean": None, "saved_rstd": None, "training": False, "eps": 1e-5, "output_mask": (1, 1, 1)},
|
|
43
59
|
"BatchNormGradGrad": {"is_training": False, "epsilon": 1e-5, "data_format": 'NCHW'},
|
|
44
60
|
"BatchNormGrad": {"is_training": False, "epsilon": 1e-5, "data_format": 'NCHW'},
|
|
45
61
|
"BatchNormGradWithActivation": {"is_training": False, "epsilon": 1e-5, "data_format": 'NCHW'},
|
|
62
|
+
"BatchNormGradWithAddAndActivation": {"is_training": False, "epsilon": 1e-5, "data_format": 'NCHW'},
|
|
46
63
|
"BatchNorm": {"is_training": False, "epsilon": 1e-5, "momentum": 0.1, "data_format": 'NCHW'},
|
|
47
64
|
"BatchNormWithActivation": {"is_training": False, "epsilon": 1e-5, "momentum": 0.1, "data_format": 'NCHW'},
|
|
48
65
|
"BatchNormWithAddAndActivation": {"is_training": False, "epsilon": 1e-5, "momentum": 0.1, "data_format": 'NCHW'},
|
|
@@ -52,37 +69,52 @@ op_args_default_value = {
|
|
|
52
69
|
"BinaryCrossEntropy": {"weight": None, "reduction": 'mean'},
|
|
53
70
|
"BinaryCrossEntropyWithLogitsBackward": {"weight": None, "posWeight": None, "reduction": 'mean'},
|
|
54
71
|
"BCEWithLogitsLoss": {"weight": None, "posWeight": None, "reduction": 'mean'},
|
|
72
|
+
"BincountExt": {"weights": None, "minlength": 0},
|
|
73
|
+
"CdistGrad": {"p": 2.0},
|
|
74
|
+
"Cdist": {"p": 2.0},
|
|
55
75
|
"CeLU": {"alpha": 1.0},
|
|
56
76
|
"CholeskyInverse": {"upper": False},
|
|
57
77
|
"Cholesky": {"upper": False},
|
|
58
78
|
"Chunk": {"dim": 0},
|
|
79
|
+
"ChunkView": {"dim": 0},
|
|
59
80
|
"ClampScalar": {"min": None, "max": None},
|
|
60
81
|
"ClampTensor": {"min": None, "max": None},
|
|
61
82
|
"Col2ImExt": {"dilation": 1, "padding": 0, "stride": 1},
|
|
62
83
|
"Col2ImGrad": {"dilation": 1, "padding": 0, "stride": 1},
|
|
63
84
|
"Concat": {"axis": 0},
|
|
64
85
|
"ConstantPadND": {"value": 0.0},
|
|
65
|
-
"
|
|
86
|
+
"Conv1DExt": {"bias": None, "stride": 1, "padding": 0, "dilation": 1, "groups": 1},
|
|
87
|
+
"Conv1DPadding": {"bias": None, "stride": 1, "padding": 'valid', "dilation": 1, "groups": 1},
|
|
88
|
+
"Conv2DExt": {"bias": None, "stride": 1, "padding": 0, "dilation": 1, "groups": 1},
|
|
89
|
+
"Conv2DPadding": {"bias": None, "stride": 1, "padding": 'valid', "dilation": 1, "groups": 1},
|
|
90
|
+
"Conv3DExt": {"bias": None, "stride": 1, "padding": 0, "dilation": 1, "groups": 1},
|
|
91
|
+
"Conv3DPadding": {"bias": None, "stride": 1, "padding": 'valid', "dilation": 1, "groups": 1},
|
|
92
|
+
"ConvolutionGrad": {"bias": None, "stride": 1, "padding": 0, "dilation": (1, 1), "transposed": False, "output_padding": 0, "groups": 1, "output_mask": ()},
|
|
66
93
|
"Convolution": {"bias": None, "stride": 1, "padding": 0, "dilation": 1, "transposed": False, "output_padding": 0, "groups": 1},
|
|
67
|
-
"
|
|
94
|
+
"ConvolutionStrGrad": {"bias": None, "stride": 1, "padding": 'valid', "dilation": 1, "transposed": False, "output_padding": 0, "groups": 1, "output_mask": ()},
|
|
95
|
+
"ConvolutionStr": {"bias": None, "stride": 1, "padding": 'valid', "dilation": 1, "transposed": False, "output_padding": 0, "groups": 1},
|
|
96
|
+
"ConvTranspose2D": {"bias": None, "stride": 1, "padding": 0, "output_padding": 0, "groups": 1, "dilation": 1},
|
|
97
|
+
"Correlate": {"pad_mode": 'valid'},
|
|
68
98
|
"CountNonZero": {"dim": None},
|
|
69
99
|
"Cross": {"dim": -65530},
|
|
100
|
+
"CumsumExt": {"dtype": None},
|
|
70
101
|
"CumProd": {"exclusive": False, "reverse": False},
|
|
71
102
|
"CumSum": {"exclusive": False, "reverse": False},
|
|
72
|
-
"CumsumExt": {"dtype": None},
|
|
73
|
-
"DCT": {"type": 2, "n": None, "axis": -1, "norm": None},
|
|
74
103
|
"DCTN": {"type": 2, "s": None, "axes": None, "norm": None},
|
|
104
|
+
"DCT": {"type": 2, "n": None, "axis": -1, "norm": None},
|
|
75
105
|
"Dense": {"bias": None},
|
|
76
106
|
"Diagonal": {"offset": 0, "dim1": 0, "dim2": 1},
|
|
107
|
+
"DiagExt": {"diagonal": 0},
|
|
108
|
+
"DivMods": {"rounding_mode": None},
|
|
77
109
|
"DivMod": {"rounding_mode": None},
|
|
78
110
|
"Dropout": {"keep_prob": 0.5, "Seed0": 0, "Seed1": 0},
|
|
79
111
|
"Eig": {"compute_v": False},
|
|
80
112
|
"EluExt": {"alpha": 1.0},
|
|
81
|
-
"EluGradExt": {"alpha": 1.0},
|
|
113
|
+
"EluGradExt": {"alpha": 1.0, "is_result": False},
|
|
82
114
|
"Elu": {"alpha": 1.0},
|
|
83
|
-
"EmbeddingApplyAdaGrad": {"mask_zero": (0,), "padding_key": (0,), "padding_key_mask": (1,), "completion_key": (0,), "completion_key_mask": (1,), "_embedding_dim": 1, "_max_key_num": 1},
|
|
84
|
-
"EmbeddingApplyAdam": {"mask_zero": (0,), "padding_key": (0,), "padding_key_mask": (1,), "completion_key": (0,), "completion_key_mask": (1,), "_embedding_dim": 1, "_max_key_num": 1},
|
|
85
115
|
"EmbeddingApplyAdamW": {"ams_grad": (0,), "mask_zero": (0,), "padding_key": (0,), "padding_key_mask": (1,), "completion_key": (0,), "completion_key_mask": (1,), "_embedding_dim": 1, "_max_key_num": 1},
|
|
116
|
+
"EmbeddingApplyAdam": {"mask_zero": (0,), "padding_key": (0,), "padding_key_mask": (1,), "completion_key": (0,), "completion_key_mask": (1,), "_embedding_dim": 1, "_max_key_num": 1},
|
|
117
|
+
"EmbeddingApplyAdaGrad": {"mask_zero": (0,), "padding_key": (0,), "padding_key_mask": (1,), "completion_key": (0,), "completion_key_mask": (1,), "_embedding_dim": 1, "_max_key_num": 1},
|
|
86
118
|
"EmbeddingApplyFtrl": {"mask_zero": (0,), "padding_key": (0,), "padding_key_mask": (1,), "completion_key": (0,), "completion_key_mask": (1,), "_embedding_dim": 1, "_max_key_num": 1},
|
|
87
119
|
"EmbeddingApplyRmsprop": {"mask_zero": (0,), "padding_key": (0,), "padding_key_mask": (1,), "completion_key": (0,), "completion_key_mask": (1,), "_embedding_dim": 1, "_max_key_num": 1},
|
|
88
120
|
"EmbeddingApplySgd": {"mask_zero": (0,), "padding_key": (0,), "padding_key_mask": (1,), "completion_key": (0,), "completion_key_mask": (1,), "_embedding_dim": 1, "_max_key_num": 1},
|
|
@@ -92,58 +124,86 @@ op_args_default_value = {
|
|
|
92
124
|
"EmbeddingFeatureMappingImport": {"only_offset_flag": True, "num": 1},
|
|
93
125
|
"Embedding": {"padding_idx": None, "max_norm": None, "norm_type": 2.0, "scale_grad_by_freq": False},
|
|
94
126
|
"EmbeddingTableEvict": {"steps_to_live": 0},
|
|
127
|
+
"EmptyLike": {"dtype": None, "device": None},
|
|
128
|
+
"Empty": {"dtype": None, "device": None},
|
|
95
129
|
"ExtractImagePatches": {"padding": 'VALID'},
|
|
96
130
|
"FFNExt": {"expertTokens": None, "bias1": None, "bias2": None, "scale": None, "offset": None, "deqScale1": None, "deqScale2": None, "antiquant_scale1": None, "antiquant_scale2": None, "antiquant_offset1": None, "antiquant_offset2": None, "activation": 'fastgelu', "inner_precise": 0},
|
|
97
131
|
"FFT2": {"s": None, "dim": (-2, -1), "norm": None},
|
|
98
|
-
"FFT": {"n": None, "dim": -1, "norm": None},
|
|
99
|
-
"FFTOrtho": {"axes": None, "forward": True},
|
|
100
|
-
"FFTWithSize": {"norm": 'backward', "onesided": True, "signal_sizes": ()},
|
|
101
132
|
"FFTFreq": {"d": 1.0, "dtype": None},
|
|
102
133
|
"FFTN": {"s": None, "dim": None, "norm": None},
|
|
103
134
|
"FFTShift": {"dim": None},
|
|
135
|
+
"FFT": {"n": None, "dim": -1, "norm": None},
|
|
136
|
+
"FFTOrtho": {"axes": None, "forward": True},
|
|
137
|
+
"FFTWithSize": {"norm": 'backward', "onesided": True, "signal_sizes": ()},
|
|
104
138
|
"FillScalar": {"dtype": None},
|
|
105
139
|
"FillTensor": {"dtype": None},
|
|
106
140
|
"FlashAttentionScoreGrad": {"pse_shift": None, "drop_mask": None, "padding_mask": None, "atten_mask": None, "softmax_max": None, "softmax_sum": None, "softmax_in": None, "attention_in": None, "prefix": None, "actual_seq_qlen": None, "actual_seq_kvlen": None, "keep_prob": 1.0, "scale_value": 1.0, "pre_tokens": 65536, "next_tokens": 65536, "inner_precise": 1, "input_layout": 'BSH', "sparse_mode": 0},
|
|
107
141
|
"FlashAttentionScore": {"real_shift": None, "drop_mask": None, "padding_mask": None, "attn_mask": None, "prefix": None, "actual_seq_qlen": None, "actual_seq_kvlen": None, "keep_prob": 1.0, "scale_value": 1.0, "pre_tokens": 2147483647, "next_tokens": 2147483647, "inner_precise": 0, "input_layout": 'BSH', "sparse_mode": 0},
|
|
108
142
|
"FlattenExt": {"start_dim": 0, "end_dim": -1},
|
|
143
|
+
"FullLike": {"dtype": None},
|
|
144
|
+
"FusedAddTopKDiv": {"activate_type": 0, "is_norm": True, "scale": 2.5},
|
|
109
145
|
"Gather": {"batch_dims": 0},
|
|
146
|
+
"GeluExt": {"approximate": 'none'},
|
|
147
|
+
"GeluGradExt": {"approximate": 'none'},
|
|
110
148
|
"GenerateEodMaskV2": {"start": 0, "steps": 1, "error_mode": 'cycle', "flip_mode": 'bitflip', "multiply_factor": 0.0, "bit_pos": 0, "flip_probability": 0.0},
|
|
111
|
-
"
|
|
149
|
+
"GLU": {"axis": -1},
|
|
150
|
+
"GridSampler2DGrad": {"interpolation_mode": 'bilinear', "padding_mode": 'zeros', "align_corners": False, "output_mask": (1, 1)},
|
|
112
151
|
"GridSampler2D": {"interpolation_mode": 'bilinear', "padding_mode": 'zeros', "align_corners": False},
|
|
113
|
-
"GridSampler3DGrad": {"interpolation_mode": 'bilinear', "padding_mode": 'zeros', "align_corners": False},
|
|
152
|
+
"GridSampler3DGrad": {"interpolation_mode": 'bilinear', "padding_mode": 'zeros', "align_corners": False, "output_mask": (1, 1)},
|
|
114
153
|
"GridSampler3D": {"interpolation_mode": 'bilinear', "padding_mode": 'zeros', "align_corners": False},
|
|
115
154
|
"GroupNormGrad": {"dx_is_require": True, "dgamma_is_require": True, "dbeta_is_require": True},
|
|
116
155
|
"GroupNorm": {"weight": None, "bias": None, "eps": 1e-5},
|
|
156
|
+
"GroupTopk": {"k_inner": 1},
|
|
157
|
+
"HardtanhGrad": {"min_val": -1, "max_val": 1},
|
|
158
|
+
"Hardtanh": {"min_val": -1, "max_val": 1},
|
|
117
159
|
"HFFT2": {"s": None, "dim": (-2, -1), "norm": None},
|
|
118
|
-
"HFFT": {"n": None, "dim": -1, "norm": None},
|
|
119
160
|
"HFFTN": {"s": None, "dim": None, "norm": None},
|
|
161
|
+
"HFFT": {"n": None, "dim": -1, "norm": None},
|
|
120
162
|
"HistcExt": {"bins": 100, "min": 0, "max": 0},
|
|
121
163
|
"HShrinkGrad": {"lambd": 0.5},
|
|
122
164
|
"HShrink": {"lambd": 0.5},
|
|
123
|
-
"IDCT": {"type": 2, "n": None, "axis": -1, "norm": None},
|
|
124
165
|
"IDCTN": {"type": 2, "s": None, "axes": None, "norm": None},
|
|
166
|
+
"IDCT": {"type": 2, "n": None, "axis": -1, "norm": None},
|
|
125
167
|
"IFFT2": {"s": None, "dim": (-2, -1), "norm": None},
|
|
126
|
-
"IFFT": {"n": None, "dim": -1, "norm": None},
|
|
127
168
|
"IFFTN": {"s": None, "dim": None, "norm": None},
|
|
128
169
|
"IFFTShift": {"dim": None},
|
|
170
|
+
"IFFT": {"n": None, "dim": -1, "norm": None},
|
|
129
171
|
"IHFFT2": {"s": None, "dim": (-2, -1), "norm": None},
|
|
130
|
-
"IHFFT": {"n": None, "dim": -1, "norm": None},
|
|
131
172
|
"IHFFTN": {"s": None, "dim": None, "norm": None},
|
|
173
|
+
"IHFFT": {"n": None, "dim": -1, "norm": None},
|
|
132
174
|
"Im2ColExt": {"dilation": 1, "padding": 0, "stride": 1},
|
|
133
175
|
"IncreFlashAttention": {"attn_mask": None, "actual_seq_lengths": None, "pse_shift": None, "dequant_scale1": None, "quant_scale1": None, "dequant_scale2": None, "quant_scale2": None, "quant_offset2": None, "antiquant_scale": None, "antiquant_offset": None, "block_table": None, "kv_padding_size": None, "num_heads": 1, "input_layout": 'BSH', "scale_value": 1.0, "num_key_value_heads": 0, "block_size": 0, "inner_precise": 1},
|
|
134
176
|
"IndexAddExt": {"alpha": 1},
|
|
135
|
-
"
|
|
177
|
+
"InnerInplaceIndexPut": {"accumulate": False},
|
|
136
178
|
"InplaceAddmm": {"beta": 1, "alpha": 1},
|
|
137
179
|
"InplaceAddsExt": {"alpha": 1},
|
|
180
|
+
"InplaceAddExt": {"alpha": 1},
|
|
181
|
+
"InplaceClampScalar": {"min": None, "max": None},
|
|
182
|
+
"InplaceClampTensor": {"min": None, "max": None},
|
|
183
|
+
"InplaceDivMods": {"rounding_mode": None},
|
|
184
|
+
"InplaceDivMod": {"rounding_mode": None},
|
|
185
|
+
"InplaceElu": {"alpha": 1.0},
|
|
186
|
+
"InplaceFillDiagonal": {"wrap": False},
|
|
187
|
+
"InplaceHardtanh": {"min_val": -1, "max_val": 1},
|
|
188
|
+
"InplaceIndexAddExt": {"alpha": 1},
|
|
189
|
+
"InplaceIndexPut": {"accumulate": False},
|
|
190
|
+
"InplacePut": {"accumulate": False},
|
|
191
|
+
"InplaceRandom": {"from_": 0, "to": None, "seed": 0, "offset": 0},
|
|
192
|
+
"InplaceSubExt": {"alpha": 1},
|
|
193
|
+
"InplaceSubScalar": {"alpha": 1},
|
|
138
194
|
"InsertGemV2InBackward": {"start": 0, "steps": 1, "error_mode": 'cycle', "flip_mode": 'bitflip', "multiply_factor": 0.0, "bit_pos": 0, "flip_probability": 0.0},
|
|
139
195
|
"IRFFT2": {"s": None, "dim": (-2, -1), "norm": None},
|
|
196
|
+
"IRFFTN": {"s": None, "dim": None, "norm": None},
|
|
140
197
|
"IRFFTDouble": {"dim": -1},
|
|
141
198
|
"IRFFT": {"n": None, "dim": -1, "norm": None},
|
|
142
|
-
"
|
|
143
|
-
"
|
|
199
|
+
"IsClose": {"rtol": 1e-05, "atol": 1e-08, "equal_nan": False},
|
|
200
|
+
"KLDivGrad": {"reduction": 'mean', "log_target": False},
|
|
201
|
+
"KLDiv": {"reduction": 'mean', "log_target": False},
|
|
202
|
+
"Kthvalue": {"dim": -1, "keepdim": False},
|
|
144
203
|
"L1LossBackwardExt": {"reduction": 'mean'},
|
|
145
204
|
"L1LossExt": {"reduction": 'mean'},
|
|
146
205
|
"LayerNormExt": {"weight": None, "bias": None, "eps": 1e-5},
|
|
206
|
+
"LayerNormGradExt": {"output_mask": (1, 1, 1)},
|
|
147
207
|
"LayerNormGradGrad": {"begin_norm_axis": 1, "begin_params_axis": 1},
|
|
148
208
|
"LayerNormGrad": {"begin_norm_axis": 1, "begin_params_axis": 1},
|
|
149
209
|
"LayerNormGradV3": {"begin_norm_axis": 1, "begin_params_axis": 1},
|
|
@@ -151,47 +211,65 @@ op_args_default_value = {
|
|
|
151
211
|
"LayerNormV3": {"begin_norm_axis": 1, "begin_params_axis": 1, "epsilon": 1e-7},
|
|
152
212
|
"LeakyReLUExt": {"negative_slope": 0.01},
|
|
153
213
|
"LeakyReLUGradExt": {"negative_slope": 0.01, "is_result": False},
|
|
214
|
+
"LinalgQr": {"mode": 'reduced'},
|
|
215
|
+
"LinalgVectorNorm": {"ord": 2, "dim": None, "keepdim": False, "dtype": None},
|
|
154
216
|
"LinSpaceExt": {"dtype": None},
|
|
217
|
+
"LogitGrad": {"eps": -1.0},
|
|
218
|
+
"Logit": {"eps": -1.0},
|
|
219
|
+
"LogSumExp": {"keepdim": False},
|
|
155
220
|
"LogSoftmaxExt": {"dim": None, "dtype": None},
|
|
156
221
|
"LogSoftmaxGrad": {"axis": -1},
|
|
157
222
|
"LogSoftmax": {"axis": -1},
|
|
158
|
-
"
|
|
159
|
-
"Logit": {"eps": -1.0},
|
|
160
|
-
"LpNormV2": {"p": 2.0, "dim": None, "keepdim": False, "epsilon": 1e-12},
|
|
223
|
+
"LpNormV2": {"p": 2, "dim": None, "keepdim": False, "epsilon": 1e-12},
|
|
161
224
|
"LstsqV2": {"driver": None},
|
|
162
225
|
"MatMul": {"transpose_a": False, "transpose_b": False},
|
|
226
|
+
"MatmulReduceScatter": {"reduce_op": 'sum', "bias": None, "comm_turn": 0, "trans_input": False, "trans_x2": False},
|
|
227
|
+
"MaximumGradGrad": {"grad_x": True, "grad_y": True},
|
|
228
|
+
"MaximumGrad": {"grad_x": True, "grad_y": True},
|
|
229
|
+
"MaxDim": {"keepdim": False},
|
|
163
230
|
"MaxPoolGradWithIndices": {"strides": None, "pads": 0, "dilation": (1, 1), "ceil_mode": False, "argmax_type": mstype.int64},
|
|
164
231
|
"MaxPoolGradWithMask": {"strides": None, "pads": 0, "dilation": (1, 1), "ceil_mode": False, "argmax_type": mstype.int64},
|
|
165
232
|
"MaxPoolWithIndices": {"strides": None, "pads": 0, "dilation": (1, 1), "ceil_mode": False, "argmax_type": mstype.int64},
|
|
166
233
|
"MaxPoolWithMask": {"strides": None, "pads": 0, "dilation": (1, 1), "ceil_mode": False, "argmax_type": mstype.int64},
|
|
167
|
-
"
|
|
168
|
-
"
|
|
169
|
-
"MeanExt": {"axis": None, "keep_dims": False, "dtype": None},
|
|
234
|
+
"MaxUnpool2DExt": {"stride": None, "padding": 0, "output_size": None},
|
|
235
|
+
"MeanExt": {"dim": None, "keepdim": False, "dtype": None},
|
|
170
236
|
"MedianDim": {"dim": -1, "keepdim": False},
|
|
237
|
+
"Meshgrid": {"indexing": 'xy'},
|
|
171
238
|
"MinimumGrad": {"grad_x": True, "grad_y": True},
|
|
239
|
+
"MinDim": {"keepdim": False},
|
|
240
|
+
"MoeTokenPermuteGrad": {"num_topk": 1, "padded_mode": False},
|
|
241
|
+
"MoeTokenPermute": {"num_out_tokens": None, "padded_mode": False},
|
|
242
|
+
"MoeTokenUnpermuteGrad": {"probs": None, "padded_mode": False, "restore_shape": None},
|
|
243
|
+
"MoeTokenUnpermute": {"probs": None, "padded_mode": False, "restore_shape": None},
|
|
172
244
|
"MSELossExt": {"reduction": 'mean'},
|
|
173
245
|
"MSELossGradExt": {"reduction": 'mean'},
|
|
246
|
+
"Nansum": {"dim": None, "keepdim": False, "dtype": None},
|
|
174
247
|
"NanToNum": {"nan": None, "posinf": None, "neginf": None},
|
|
248
|
+
"NewEmpty": {"dtype": None, "device": None},
|
|
249
|
+
"NewOnes": {"dtype": None},
|
|
250
|
+
"NewZeros": {"dtype": None},
|
|
251
|
+
"NLLLoss2d": {"reduction": 'mean', "ignore_index": -100},
|
|
175
252
|
"NLLLossGrad": {"reduction": 'mean', "ignore_index": -100},
|
|
176
253
|
"NLLLoss": {"reduction": 'mean', "ignore_index": -100},
|
|
177
254
|
"Norm": {"p": 2.0, "dim": None, "keepdim": False, "dtype": None},
|
|
178
|
-
"OneHotExt": {"axis": -1},
|
|
179
|
-
"OneHot": {"axis": -1},
|
|
180
255
|
"OnesLikeExt": {"dtype": None},
|
|
181
256
|
"Ones": {"dtype": None},
|
|
257
|
+
"OneHotExt": {"axis": -1},
|
|
258
|
+
"OneHot": {"axis": -1},
|
|
182
259
|
"PagedAttentionMask": {"antiquant_scale": None, "antiquant_offset": None, "alibi_mask": None, "kv_cache_quant_mode": 'DEFAULT'},
|
|
183
|
-
"PagedAttention": {"antiquant_scale": None, "antiquant_offset": None, "attn_mask": None, "q_seq_lens": None, "kv_cache_quant_mode": 'DEFAULT'},
|
|
184
|
-
"ProdExt": {"
|
|
260
|
+
"PagedAttention": {"value_cache": None, "block_tables": None, "context_lens": None, "antiquant_scale": None, "antiquant_offset": None, "attn_mask": None, "q_seq_lens": None, "alibi_mask": None, "kv_cache_quant_mode": 'DEFAULT', "mask_mode": 'MASK_DEFAULT', "mla_v_dim": 0},
|
|
261
|
+
"ProdExt": {"dim": None, "keepdim": False, "dtype": None},
|
|
262
|
+
"PromptFlashAttention": {"attn_mask": None, "actual_seq_lengths": None, "actual_seq_lengths_kv": None, "pse_shift": None, "deq_scale1": None, "quant_scale1": None, "deq_scale2": None, "quant_scale2": None, "quant_offset2": None, "num_heads": 1, "scale_value": 1.0, "pre_tokens": 2147483647, "next_tokens": 0, "input_layout": 'BSH', "num_key_value_heads": 0, "sparse_mode": 0, "inner_precise": 1},
|
|
185
263
|
"PromptKVCache": {"align_mode": 'LEFT'},
|
|
186
264
|
"Qr": {"full_matrices": False},
|
|
187
|
-
"RandExt": {"dtype": None},
|
|
188
|
-
"RandLikeExt": {"dtype": None},
|
|
189
265
|
"RandIntLike": {"dtype": None},
|
|
190
266
|
"RandInt": {"dtype": None},
|
|
191
267
|
"RandnLike": {"dtype": None},
|
|
192
268
|
"Randn": {"dtype": None},
|
|
193
269
|
"RandpermExt": {"dtype": mstype.int64},
|
|
194
270
|
"RandpermV2": {"seed": 0, "offset": 0, "dtype": mstype.int64},
|
|
271
|
+
"RandExt": {"dtype": None},
|
|
272
|
+
"RandLikeExt": {"dtype": None},
|
|
195
273
|
"Range": {"maxlen": 1000000},
|
|
196
274
|
"ReduceAll": {"axis": None, "keep_dims": False},
|
|
197
275
|
"ReduceAny": {"axis": (), "keep_dims": False},
|
|
@@ -203,6 +281,7 @@ op_args_default_value = {
|
|
|
203
281
|
"ReduceSum": {"axis": (), "keep_dims": False, "skip_mode": False},
|
|
204
282
|
"RepeatInterleaveInt": {"dim": None, "output_size": None},
|
|
205
283
|
"RepeatInterleaveTensor": {"dim": None, "output_size": None},
|
|
284
|
+
"ReshapeAndCache": {"value": None, "key_cache": None, "value_cache": None, "slot_mapping": None},
|
|
206
285
|
"ResizeBicubicGrad": {"align_corners": False, "half_pixel_centers": False},
|
|
207
286
|
"ResizeBicubic": {"align_corners": False, "half_pixel_centers": False},
|
|
208
287
|
"ResizeBilinearGrad": {"align_corners": False, "half_pixel_centers": False},
|
|
@@ -215,13 +294,14 @@ op_args_default_value = {
|
|
|
215
294
|
"ResizeNearestNeighborV2Grad": {"align_corners": False, "half_pixel_centers": False},
|
|
216
295
|
"ResizeNearestNeighborV2": {"align_corners": False, "half_pixel_centers": False},
|
|
217
296
|
"RFFT2": {"s": None, "dim": (-2, -1), "norm": None},
|
|
218
|
-
"RFFT": {"n": None, "dim": -1, "norm": None},
|
|
219
297
|
"RFFTFreq": {"d": 1.0, "dtype": None},
|
|
220
298
|
"RFFTN": {"s": None, "dim": None, "norm": None},
|
|
299
|
+
"RFFT": {"n": None, "dim": -1, "norm": None},
|
|
221
300
|
"RmsNorm": {"epsilon": 1e-6},
|
|
222
|
-
"Roll": {"
|
|
301
|
+
"Roll": {"dims": None},
|
|
223
302
|
"RotaryPositionEmbeddingGrad": {"dx": None, "mode": 0},
|
|
224
303
|
"RotaryPositionEmbedding": {"mode": 0},
|
|
304
|
+
"RotatedIou": {"trans": False, "mode": 0, "is_cross": True, "v_threshold": 0.0, "e_threshold": 0.0},
|
|
225
305
|
"Round": {"decimals": 0},
|
|
226
306
|
"ScalarToTensor": {"dtype": None},
|
|
227
307
|
"Scatter": {"reduce": 'none'},
|
|
@@ -229,33 +309,49 @@ op_args_default_value = {
|
|
|
229
309
|
"SearchSorted": {"sorter": None, "dtype": mstype.int64, "right": False},
|
|
230
310
|
"SequenceConcat": {"axis": 0},
|
|
231
311
|
"SilentCheckV2": {"c_min_steps": 7, "c_thresh_l1": 1000000.0, "c_coeff_l1": 100000.0, "c_thresh_l2": 10000.0, "c_coeff_l2": 5000.0, "npu_asd_detect": 1},
|
|
312
|
+
"SilentCheckV3": {"c_thresh_l1": 1000000.0, "c_thresh_l2": 10000.0, "beta1": 0.0, "npu_asd_detect": 1},
|
|
313
|
+
"SmoothL1LossGrad": {"beta": 1.0, "reduction": 'none'},
|
|
314
|
+
"SmoothL1Loss": {"beta": 1.0, "reduction": 'none'},
|
|
232
315
|
"SoftmaxBackward": {"dim": -1},
|
|
233
316
|
"Softmax": {"axis": -1},
|
|
234
317
|
"SoftplusExt": {"beta": 1, "threshold": 20},
|
|
235
318
|
"SoftplusGradExt": {"beta": 1, "threshold": 20},
|
|
236
319
|
"SoftShrinkGrad": {"lambd": 0.5},
|
|
237
320
|
"SoftShrink": {"lambd": 0.5},
|
|
321
|
+
"SoftMarginLossGrad": {"reduction": 'mean'},
|
|
322
|
+
"SoftMarginLoss": {"reduction": 'mean'},
|
|
238
323
|
"SolveTriangular": {"trans": 0, "lower": False, "unit_diagonal": False},
|
|
239
324
|
"SortExt": {"dim": -1, "descending": False, "stable": False},
|
|
325
|
+
"SpeedFusionAttentionGrad": {"pse": None, "padding_mask": None, "atten_mask": None, "softmax_max": None, "softmax_sum": None, "softmax_in": None, "attention_in": None, "scale_value": 1.0, "keep_prob": 1.0, "pre_tokens": 2147483647, "next_tokens": 2147483647, "inner_precise": 0, "seed": None, "offset": None, "numels": None, "prefix": None, "actual_seq_qlen": None, "actual_seq_kvlen": None, "sparse_mode": 0, "gen_mask_parallel": True, "sync": False, "pse_type": 1, "q_start_idx": None, "kv_start_idx": None},
|
|
326
|
+
"SpeedFusionAttention": {"pse": None, "padding_mask": None, "atten_mask": None, "scale": 1.0, "keep_prob": 1.0, "pre_tokens": 2147483647, "next_tokens": 2147483647, "inner_precise": 0, "prefix": None, "actual_seq_qlen": None, "actual_seq_kvlen": None, "sparse_mode": 0, "gen_mask_parallel": True, "sync": False, "pse_type": 1, "q_start_idx": None, "kv_start_idx": None},
|
|
240
327
|
"Split": {"axis": 0, "output_num": 1},
|
|
241
|
-
"SplitTensor": {"
|
|
242
|
-
"
|
|
328
|
+
"SplitTensor": {"dim": 0},
|
|
329
|
+
"SplitTensorView": {"dim": 0},
|
|
330
|
+
"SplitWithSize": {"dim": 0},
|
|
331
|
+
"SplitWithSizeView": {"dim": 0},
|
|
332
|
+
"Squeeze": {"axis": ()},
|
|
243
333
|
"StackExt": {"dim": 0},
|
|
334
|
+
"StdMean": {"dim": None, "correction": 1, "keepdim": False},
|
|
335
|
+
"Std": {"dim": None, "correction": 1, "keepdim": False},
|
|
244
336
|
"StridedSlice": {"begin_mask": 0, "end_mask": 0, "ellipsis_mask": 0, "new_axis_mask": 0, "shrink_axis_mask": 0},
|
|
245
337
|
"SubExt": {"alpha": 1},
|
|
338
|
+
"SubScalar": {"alpha": 1},
|
|
246
339
|
"SumExt": {"dim": None, "keepdim": False, "dtype": None},
|
|
340
|
+
"Svd": {"full_matrices": False, "compute_uv": True},
|
|
247
341
|
"SwigluGrad": {"dim": -1},
|
|
248
342
|
"Swiglu": {"dim": -1},
|
|
249
343
|
"TensorScatterElements": {"axis": 0, "reduce": 'none'},
|
|
250
|
-
"TopkExt": {"dim": -1, "largest": True, "sorted": True},
|
|
251
344
|
"TopKRouter": {"drop_type": 0},
|
|
345
|
+
"TopkExt": {"dim": -1, "largest": True, "sorted": True},
|
|
252
346
|
"TraceV2Grad": {"offset": 0, "axis1": 1, "axis2": 0},
|
|
253
347
|
"TraceV2": {"offset": 0, "axis1": 1, "axis2": 0, "dtype": None},
|
|
348
|
+
"TriangularSolve": {"upper": True, "transpose": False, "unitriangular": False},
|
|
254
349
|
"TrilExt": {"diagonal": 0},
|
|
255
350
|
"Triu": {"diagonal": 0},
|
|
256
351
|
"TupleToTensor": {"dtype": None},
|
|
257
352
|
"Unique2": {"sorted": True, "return_inverse": False, "return_counts": False},
|
|
258
|
-
"
|
|
353
|
+
"UniqueConsecutive": {"return_inverse": False, "return_counts": False, "dim": None},
|
|
354
|
+
"UnstackExtView": {"dim": 0},
|
|
259
355
|
"UpsampleBicubic2DGrad": {"output_size": None, "scales": None, "align_corners": False},
|
|
260
356
|
"UpsampleBicubic2D": {"output_size": None, "scales": None, "align_corners": False},
|
|
261
357
|
"UpsampleBilinear2DGrad": {"output_size": None, "scales": None, "align_corners": False},
|
|
@@ -270,46 +366,162 @@ op_args_default_value = {
|
|
|
270
366
|
"UpsampleNearest3D": {"output_size": None, "scales": None},
|
|
271
367
|
"UpsampleTrilinear3DGrad": {"output_size": None, "scales": None, "align_corners": False},
|
|
272
368
|
"UpsampleTrilinear3D": {"output_size": None, "scales": None, "align_corners": False},
|
|
369
|
+
"VarMean": {"dim": None, "correction": 1, "keepdim": False},
|
|
370
|
+
"Var": {"dim": None, "correction": 1, "keepdim": False},
|
|
273
371
|
"ZerosLikeExt": {"dtype": None},
|
|
274
372
|
"Zeros": {"dtype": None},
|
|
373
|
+
"AddRmsNormDynamicQuant": {"smooth_scale2": None, "epsilon": 1e-5},
|
|
275
374
|
"AddRmsNormQuantV2": {"epsilon": 1e-5},
|
|
375
|
+
"DynamicNTK": {"dtype": mstype.float16},
|
|
276
376
|
"DynamicQuantExt": {"smooth_scales": None},
|
|
277
|
-
"FusedInferAttentionScore": {"pse_shift": None, "attn_mask": None, "actual_seq_lengths": None, "actual_seq_lengths_kv": None, "dequant_scale1": None, "quant_scale1": None, "dequant_scale2": None, "quant_scale2": None, "quant_offset2": None, "antiquant_scale": None, "antiquant_offset": None, "block_table": None, "query_padding_size": None, "kv_padding_size": None, "scale_value": 1.0, "pre_tokens": 2147483647, "next_tokens": 2147483647, "input_layout": 'BSH', "num_key_value_heads": 0, "sparse_mode": 0, "inner_precise": 1, "block_size": 0, "antiquant_mode": 0, "softmax_lse_flag": False},
|
|
278
|
-
"GroupedMatmul": {"bias": None, "scale": None, "offset": None, "antiquant_scale": None, "antiquant_offset": None, "group_list": None, "split_item": 0, "group_type": -1},
|
|
377
|
+
"FusedInferAttentionScore": {"pse_shift": None, "attn_mask": None, "actual_seq_lengths": None, "actual_seq_lengths_kv": None, "dequant_scale1": None, "quant_scale1": None, "dequant_scale2": None, "quant_scale2": None, "quant_offset2": None, "antiquant_scale": None, "antiquant_offset": None, "block_table": None, "query_padding_size": None, "kv_padding_size": None, "key_antiquant_scale": None, "key_antiquant_offset": None, "value_antiquant_scale": None, "value_antiquant_offset": None, "key_shared_prefix": None, "value_shared_prefix": None, "actual_shared_prefix_len": None, "num_heads": 1, "scale_value": 1.0, "pre_tokens": 2147483647, "next_tokens": 2147483647, "input_layout": 'BSH', "num_key_value_heads": 0, "sparse_mode": 0, "inner_precise": 1, "block_size": 0, "antiquant_mode": 0, "softmax_lse_flag": False, "key_antiquant_mode": 0, "value_antiquant_mode": 0},
|
|
378
|
+
"GroupedMatmul": {"bias": None, "scale": None, "offset": None, "antiquant_scale": None, "antiquant_offset": None, "group_list": None, "split_item": 0, "group_type": -1, "transpose_a": False, "transpose_b": False},
|
|
379
|
+
"GroupedMatmulV2": {"bias": None, "scale": None, "offset": None, "antiquant_scale": None, "antiquant_offset": None, "group_list": None, "split_item": 0, "group_type": -1},
|
|
380
|
+
"GroupedMatmulV4": {"bias": None, "scale": None, "offset": None, "antiquant_scale": None, "antiquant_offset": None, "pre_token_scale": None, "group_list": None, "activation_input": None, "activation_quant_scale": None, "activation_quant_offset": None, "split_item": 0, "group_type": -1, "group_list_type": 0, "act_type": 0},
|
|
279
381
|
"KVCacheScatterUpdate": {"reduce": 'none'},
|
|
382
|
+
"MatmulAllReduceAddRmsNorm": {"reduce_op": 'sum', "comm_turn": 0, "stream_mode": 1},
|
|
280
383
|
"MoeFinalizeRouting": {"x2": None, "bias": None, "scales": None, "expanded_row_idx": None, "expanded_expert_idx": None},
|
|
384
|
+
"MoeGatingTopKSoftmax": {"finished": None, "k": 1},
|
|
281
385
|
"QuantBatchMatmul": {"offset": None, "bias": None, "pertokenScaleOptional": None, "transpose_x1": False, "transpose_x2": False, "dtype": mstype.float16},
|
|
282
386
|
"QuantV2": {"sqrt_mode": False, "rounding_mode": 'ROUND', "dst_type": mstype.int8},
|
|
387
|
+
"RmsNormQuant": {"beta": None, "epsilon": 1e-6},
|
|
388
|
+
"SwiGLUDynamicQuant": {"smooth_scale": None},
|
|
389
|
+
"TransposeBatchMatmulTranspose": {"transpose_a": False, "transpose_b": False},
|
|
283
390
|
"WeightQuantBatchMatmul": {"antiquant_offset": None, "quant_scale": None, "quant_offset": None, "bias": None, "transpose_x": False, "transpose_weight": False, "antiquant_group_size": 0},
|
|
391
|
+
"GmmBackwardFusion": {"group_list": None, "group_list_type": 0},
|
|
392
|
+
"GmmBackward": {"group_list": None, "group_list_type": 0},
|
|
393
|
+
"Gmm": {"bias": None, "group_list": None, "group_type": 0, "group_list_type": 0},
|
|
394
|
+
"GmmV2BackwardFusion": {"group_list": None, "group_list_type": 0},
|
|
395
|
+
"GmmV2Backward": {"group_list": None, "group_list_type": 0},
|
|
396
|
+
"GmmV2": {"bias": None, "group_list": None, "group_type": 0, "group_list_type": 0},
|
|
397
|
+
"DeprecatedAddbmm": {"beta": 1, "alpha": 1},
|
|
398
|
+
"DeprecatedAddmm": {"beta": 1, "alpha": 1},
|
|
399
|
+
"DeprecatedAddmv": {"beta": 1, "alpha": 1},
|
|
400
|
+
"DeprecatedReduceAll": {"dim": None, "keepdim": False},
|
|
401
|
+
"DeprecatedAllclose": {"rtol": 1e-05, "atol": 1e-08, "equal_nan": False},
|
|
402
|
+
"DeprecatedReduceAny": {"axis": None, "keep_dims": False},
|
|
403
|
+
"DeprecatedAny": {"dim": None, "keepdim": False},
|
|
404
|
+
"DeprecatedArgmax": {"axis": -1, "keepdims": False},
|
|
405
|
+
"DeprecatedArgmin": {"axis": None, "keepdims": False},
|
|
406
|
+
"DeprecatedArgsort": {"axis": -1, "descending": False},
|
|
407
|
+
"DeprecatedBaddbmm": {"beta": 1, "alpha": 1},
|
|
408
|
+
"DeprecatedBincount": {"weights": None, "minlength": 0},
|
|
409
|
+
"DeprecatedChunk": {"axis": 0},
|
|
410
|
+
"DeprecatedInplaceCopy": {"non_blocking": False},
|
|
411
|
+
"DeprecatedCountNonzero": {"axis": (), "keep_dims": False, "dtype": None},
|
|
412
|
+
"DeprecatedCumsum": {"axis": None, "dtype": None},
|
|
413
|
+
"DeprecatedDiv": {"rounding_mode": None},
|
|
414
|
+
"DeprecatedFlatten": {"order": 'C', "start_dim": 0, "end_dim": -1},
|
|
415
|
+
"DeprecatedGather": {"batch_dims": 0},
|
|
416
|
+
"DeprecatedHistc": {"bins": 100, "min": 0, "max": 0},
|
|
417
|
+
"DeprecatedIndexAdd": {"use_lock": True, "check_index_bound": True},
|
|
418
|
+
"DeprecatedIsclose": {"rtol": 1e-05, "atol": 1e-08, "equal_nan": False},
|
|
419
|
+
"DeprecatedLogsumexp": {"keepdim": False},
|
|
420
|
+
"DeprecatedMax": {"axis": None, "keepdims": False, "initial": None, "where": True, "return_indices": False},
|
|
421
|
+
"DeprecatedMean": {"axis": None, "keep_dims": False},
|
|
422
|
+
"DeprecatedMedian": {"axis": -1, "keepdims": False},
|
|
423
|
+
"DeprecatedMin": {"axis": None, "keepdims": False, "initial": None, "where": True, "return_indices": False},
|
|
424
|
+
"DeprecatedNansum": {"axis": None, "keepdims": False, "dtype": None},
|
|
425
|
+
"DeprecatedProd": {"dim": None, "keepdim": False, "dtype": None},
|
|
426
|
+
"DeprecatedRepeatInterleave": {"dim": None},
|
|
427
|
+
"DeprecatedRoll": {"dims": None},
|
|
428
|
+
"DeprecatedSort": {"axis": -1, "descending": False},
|
|
429
|
+
"DeprecatedSplit": {"axis": 0},
|
|
430
|
+
"DeprecatedStd": {"axis": None, "ddof": 0, "keepdims": False},
|
|
431
|
+
"DeprecatedSum": {"axis": None, "dtype": None, "keepdims": False, "initial": None},
|
|
432
|
+
"DeprecatedTake": {"axis": None, "mode": 'clip'},
|
|
433
|
+
"DeprecatedTopk": {"dim": None, "largest": True, "sorted": True},
|
|
434
|
+
"DeprecatedTranspose": {"axes": None},
|
|
435
|
+
"DeprecatedTril": {"diagonal": 0},
|
|
436
|
+
"DeprecatedUnstack": {"dim": 0},
|
|
437
|
+
"DeprecatedUnique": {"sorted": True, "return_inverse": False, "return_counts": False, "dim": None},
|
|
438
|
+
"DeprecatedVar": {"axis": None, "ddof": 0, "keepdims": False},
|
|
284
439
|
}
|
|
285
440
|
|
|
286
441
|
op_labels = {
|
|
287
|
-
"AdamWeightDecay": {"side_effect_mem": True},
|
|
288
442
|
"AdamW": {"side_effect_mem": True},
|
|
443
|
+
"AdamWeightDecay": {"side_effect_mem": True},
|
|
289
444
|
"ApplyAdamW": {"side_effect_mem": True},
|
|
290
445
|
"AssignAdd": {"side_effect_mem": True},
|
|
291
446
|
"Assign": {"side_effect_mem": True},
|
|
292
|
-
"
|
|
447
|
+
"AssignSub": {"side_effect_mem": True},
|
|
448
|
+
"BatchNormElemt": {"side_effect_mem": True},
|
|
449
|
+
"BatchNormGatherStatsWithCounts": {"side_effect_mem": True},
|
|
450
|
+
"BroadcastToView": {"side_effect_mem": True},
|
|
451
|
+
"ChunkView": {"side_effect_mem": True},
|
|
293
452
|
"DecoderKVCache": {"side_effect_mem": True},
|
|
453
|
+
"DistCommAllReduce": {"side_effect_mem": True},
|
|
454
|
+
"DistCommReduce": {"side_effect_mem": True},
|
|
294
455
|
"DropoutExt": {"side_effect_hidden": True},
|
|
295
456
|
"DropoutGenMaskExt": {"side_effect_hidden": True},
|
|
296
457
|
"Dropout": {"side_effect_hidden": True},
|
|
297
|
-
"EmbeddingApplyAdaGrad": {"_process_node_engine_id": 'PS'},
|
|
298
|
-
"EmbeddingApplyAdam": {"_process_node_engine_id": 'PS'},
|
|
299
458
|
"EmbeddingApplyAdamW": {"_process_node_engine_id": 'PS'},
|
|
459
|
+
"EmbeddingApplyAdam": {"_process_node_engine_id": 'PS'},
|
|
460
|
+
"EmbeddingApplyAdaGrad": {"_process_node_engine_id": 'PS'},
|
|
300
461
|
"EmbeddingApplyFtrl": {"_process_node_engine_id": 'PS'},
|
|
301
462
|
"EmbeddingApplyRmsprop": {"_process_node_engine_id": 'PS'},
|
|
302
463
|
"EmbeddingApplySgd": {"_process_node_engine_id": 'PS'},
|
|
303
464
|
"Embedding": {"side_effect_mem": True},
|
|
304
465
|
"EmbeddingTableEvict": {"_process_node_engine_id": 'PS'},
|
|
466
|
+
"ExpandDimsView": {"side_effect_mem": True},
|
|
305
467
|
"Generator": {"side_effect_mem": True},
|
|
306
|
-
"
|
|
468
|
+
"GroupTopk": {"side_effect_mem": True},
|
|
469
|
+
"InnerInplaceIndexPut": {"side_effect_mem": True},
|
|
307
470
|
"InplaceAddmm": {"side_effect_mem": True},
|
|
308
471
|
"InplaceAddsExt": {"side_effect_mem": True},
|
|
472
|
+
"InplaceAddExt": {"side_effect_mem": True},
|
|
473
|
+
"InplaceCopy": {"side_effect_mem": True},
|
|
474
|
+
"InplaceDivMods": {"side_effect_mem": True},
|
|
475
|
+
"InplaceDivMod": {"side_effect_mem": True},
|
|
476
|
+
"InplaceDivs": {"side_effect_mem": True},
|
|
477
|
+
"InplaceDiv": {"side_effect_mem": True},
|
|
478
|
+
"InplaceElu": {"side_effect_mem": True},
|
|
479
|
+
"InplaceErfinv": {"side_effect_mem": True},
|
|
480
|
+
"InplaceExp": {"side_effect_mem": True},
|
|
481
|
+
"InplaceFillDiagonal": {"side_effect_mem": True},
|
|
482
|
+
"InplaceFillScalar": {"side_effect_mem": True},
|
|
483
|
+
"InplaceFillTensor": {"side_effect_mem": True},
|
|
484
|
+
"InplaceFloorDivides": {"side_effect_mem": True},
|
|
485
|
+
"InplaceFloorDivide": {"side_effect_mem": True},
|
|
486
|
+
"InplaceFloor": {"side_effect_mem": True},
|
|
487
|
+
"InplaceGroupedMatmulAdd": {"side_effect_mem": True},
|
|
488
|
+
"InplaceHardtanh": {"side_effect_mem": True},
|
|
489
|
+
"InplaceIndexAddExt": {"side_effect_mem": True},
|
|
490
|
+
"InplaceIndexPut": {"side_effect_mem": True},
|
|
491
|
+
"InplaceLog": {"side_effect_mem": True},
|
|
492
|
+
"InplaceMaskedFillScalar": {"side_effect_mem": True},
|
|
493
|
+
"InplaceMaskedFillTensor": {"side_effect_mem": True},
|
|
494
|
+
"InplaceMuls": {"side_effect_mem": True},
|
|
495
|
+
"InplaceMul": {"side_effect_mem": True},
|
|
496
|
+
"InplaceNormal": {"side_effect_mem": True},
|
|
497
|
+
"InplacePut": {"side_effect_mem": True},
|
|
498
|
+
"InplaceRandom": {"side_effect_mem": True},
|
|
499
|
+
"InplaceReLU": {"side_effect_mem": True},
|
|
500
|
+
"InplaceScatterAdd": {"side_effect_mem": True},
|
|
501
|
+
"InplaceScatterSrc": {"side_effect_mem": True},
|
|
502
|
+
"InplaceScatterSrcReduce": {"side_effect_mem": True},
|
|
503
|
+
"InplaceScatterValue": {"side_effect_mem": True},
|
|
504
|
+
"InplaceScatterValueReduce": {"side_effect_mem": True},
|
|
505
|
+
"InplaceSubExt": {"side_effect_mem": True},
|
|
506
|
+
"InplaceSubScalar": {"side_effect_mem": True},
|
|
507
|
+
"InplaceTanh": {"side_effect_mem": True},
|
|
508
|
+
"InplaceThreshold": {"side_effect_mem": True},
|
|
509
|
+
"InplaceUniform": {"side_effect_mem": True},
|
|
309
510
|
"Log": {"cust_aicpu": 'Log', "base": -1.0, "scale": 1.0, "shift": 0.0},
|
|
511
|
+
"NarrowView": {"side_effect_mem": True},
|
|
310
512
|
"PromptKVCache": {"side_effect_mem": True},
|
|
311
513
|
"ReshapeAndCache": {"side_effect_mem": True},
|
|
312
514
|
"ResizeD": {"mode": 'linear'},
|
|
515
|
+
"SelectExtView": {"side_effect_mem": True},
|
|
313
516
|
"SilentCheckV2": {"side_effect_mem": True},
|
|
517
|
+
"SilentCheckV3": {"side_effect_mem": True},
|
|
518
|
+
"SliceExtView": {"side_effect_mem": True},
|
|
519
|
+
"SplitTensorView": {"side_effect_mem": True},
|
|
520
|
+
"SplitWithSizeView": {"side_effect_mem": True},
|
|
521
|
+
"TransposeExtView": {"side_effect_mem": True},
|
|
522
|
+
"TransposeView": {"side_effect_mem": True},
|
|
523
|
+
"UnstackExtView": {"side_effect_mem": True},
|
|
314
524
|
"KVCacheScatterUpdate": {"side_effect_mem": True},
|
|
525
|
+
"InplaceExponential": {"side_effect_mem": True},
|
|
526
|
+
"DeprecatedInplaceCopy": {"side_effect_mem": True},
|
|
315
527
|
}
|