nndeploy 0.2.0__tar.gz

nndeploy-0.2.0/PKG-INFO

@@ -0,0 +1,108 @@
+Metadata-Version: 2.4
+Name: nndeploy
+Version: 0.2.0
+Summary: Workflow-based Multi-platform AI Deployment Tool
+Home-page: https://github.com/nndeploy/nndeploy
+Author: nndeploy team
+Author-email: 595961667@qq.com
+License: Apache License 2.0
+Project-URL: Bug Reports, https://github.com/nndeploy/nndeploy/issues
+Project-URL: Source, https://github.com/nndeploy/nndeploy
+Keywords: deep-learning,neural-network,model-deployment,inference,ai
+Classifier: Development Status :: 4 - Beta
+Classifier: Intended Audience :: Developers
+Classifier: License :: OSI Approved :: Apache Software License
+Classifier: Operating System :: OS Independent
+Classifier: Programming Language :: C++
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.10
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Programming Language :: Python :: 3.13
+Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
+Requires-Python: >=3.8
+Description-Content-Type: text/markdown
+Requires-Dist: cython
+Requires-Dist: packaging
+Requires-Dist: setuptools
+Requires-Dist: gitpython>=3.1.30
+Requires-Dist: aiofiles>=24.1.0
+Requires-Dist: PyYAML>=5.3.1
+Requires-Dist: pytest
+Requires-Dist: jsonschema
+Requires-Dist: multiprocess
+Requires-Dist: numpy
+Requires-Dist: opencv-python>=4.8.0
+Requires-Dist: torch>=2.0.0
+Requires-Dist: torchvision>=0.15.0
+Requires-Dist: onnxruntime-gpu>=1.18.0
+Requires-Dist: requests>=2.31.0
+Requires-Dist: fastapi>=0.104.0
+Requires-Dist: uvicorn>=0.24.0
+Requires-Dist: websockets>=11.0
+Requires-Dist: python-multipart>=0.0.6
+Requires-Dist: pydantic>=2.0.0
+Dynamic: author
+Dynamic: author-email
+Dynamic: classifier
+Dynamic: description
+Dynamic: description-content-type
+Dynamic: home-page
+Dynamic: keywords
+Dynamic: license
+Dynamic: project-url
+Dynamic: requires-dist
+Dynamic: requires-python
+Dynamic: summary
+
+
+<h3 align="center">
+Workflow-based Multi-platform AI Deployment Tool
+</h3>
+
+## Features
+
+nndeploy is a workflow-based multi-platform AI deployment tool with the following capabilities:
+
+### 1. Efficiency Tool for AI Deployment
+
+- **Visual Workflow**: Deploy AI algorithms through drag-and-drop interface
+
+- **Function Calls**: Export workflows as JSON configuration files, supporting Python/C++ API calls
+
+- **Multi-platform Inference**: One workflow, multi-platform deployment. Zero-abstraction cost integration with 13 mainstream inference frameworks, covering cloud, desktop, mobile, and edge platforms
+
+  | Framework | Support Status |
+  | :------- | :------ |
+  | [PyTorch](https://pytorch.org/) | ✅ |
+  | [TensorRT](https://github.com/NVIDIA/TensorRT) | ✅ |
+  | [OpenVINO](https://github.com/openvinotoolkit/openvino) | ✅ |
+  | [ONNXRuntime](https://github.com/microsoft/onnxruntime) | ✅ |
+  | [MNN](https://github.com/alibaba/MNN) | ✅ |
+  | [TNN](https://github.com/Tencent/TNN) | ✅ |
+  | [ncnn](https://github.com/Tencent/ncnn) | ✅ |
+  | [CoreML](https://github.com/apple/coremltools) | ✅ |
+  | [AscendCL](https://www.hiascend.com/zh/) | ✅ |
+  | [RKNN](https://www.rock-chips.com/a/cn/downloadcenter/BriefDatasheet/index.html) | ✅ |
+  | [TVM](https://github.com/apache/tvm) | ✅ |
+  | [SNPE](https://developer.qualcomm.com/software/qualcomm-neural-processing-sdk) | ✅ |
+  | [Custom Inference Framework](docs/zh_cn/inference/README_INFERENCE.md) | ✅ |
+
+### 2. Performance Tool for AI Deployment
+
+- **Parallel Optimization**: Support for serial, pipeline parallel, and task parallel execution modes
+
+- **Memory Optimization**: Zero-copy, memory pools, memory reuse and other optimization strategies
+  
+- **High-Performance Optimization**: Built-in nodes optimized with C++/CUDA/SIMD implementations
+
+### 3. Creative Tool for AI Deployment
+
+- **Custom Nodes**: Support Python/C++ custom nodes with seamless integration into visual interface without frontend code
+
+- **Algorithm Composition**: Flexible combination of different algorithms to rapidly build innovative AI applications
+
+- **What You Tune Is What You See**: Frontend visual adjustment of all node parameters in AI algorithm deployment with quick preview of post-tuning effects
+
+## More infomation
+You can get everything in nndeploy github main page : [nndeploy](https://github.com/nndeploy/nndeploy)

nndeploy-0.2.0/README.md

@@ -0,0 +1,51 @@
+
+<h3 align="center">
+Workflow-based Multi-platform AI Deployment Tool
+</h3>
+
+## Features
+
+nndeploy is a workflow-based multi-platform AI deployment tool with the following capabilities:
+
+### 1. Efficiency Tool for AI Deployment
+
+- **Visual Workflow**: Deploy AI algorithms through drag-and-drop interface
+
+- **Function Calls**: Export workflows as JSON configuration files, supporting Python/C++ API calls
+
+- **Multi-platform Inference**: One workflow, multi-platform deployment. Zero-abstraction cost integration with 13 mainstream inference frameworks, covering cloud, desktop, mobile, and edge platforms
+
+  | Framework | Support Status |
+  | :------- | :------ |
+  | [PyTorch](https://pytorch.org/) | ✅ |
+  | [TensorRT](https://github.com/NVIDIA/TensorRT) | ✅ |
+  | [OpenVINO](https://github.com/openvinotoolkit/openvino) | ✅ |
+  | [ONNXRuntime](https://github.com/microsoft/onnxruntime) | ✅ |
+  | [MNN](https://github.com/alibaba/MNN) | ✅ |
+  | [TNN](https://github.com/Tencent/TNN) | ✅ |
+  | [ncnn](https://github.com/Tencent/ncnn) | ✅ |
+  | [CoreML](https://github.com/apple/coremltools) | ✅ |
+  | [AscendCL](https://www.hiascend.com/zh/) | ✅ |
+  | [RKNN](https://www.rock-chips.com/a/cn/downloadcenter/BriefDatasheet/index.html) | ✅ |
+  | [TVM](https://github.com/apache/tvm) | ✅ |
+  | [SNPE](https://developer.qualcomm.com/software/qualcomm-neural-processing-sdk) | ✅ |
+  | [Custom Inference Framework](docs/zh_cn/inference/README_INFERENCE.md) | ✅ |
+
+### 2. Performance Tool for AI Deployment
+
+- **Parallel Optimization**: Support for serial, pipeline parallel, and task parallel execution modes
+
+- **Memory Optimization**: Zero-copy, memory pools, memory reuse and other optimization strategies
+  
+- **High-Performance Optimization**: Built-in nodes optimized with C++/CUDA/SIMD implementations
+
+### 3. Creative Tool for AI Deployment
+
+- **Custom Nodes**: Support Python/C++ custom nodes with seamless integration into visual interface without frontend code
+
+- **Algorithm Composition**: Flexible combination of different algorithms to rapidly build innovative AI applications
+
+- **What You Tune Is What You See**: Frontend visual adjustment of all node parameters in AI algorithm deployment with quick preview of post-tuning effects
+
+## More infomation
+You can get everything in nndeploy github main page : [nndeploy](https://github.com/nndeploy/nndeploy)

nndeploy-0.2.0/nndeploy/__init__.py

@@ -0,0 +1,44 @@
+import os
+import sys
+
+# 获取当前包的目录路径
+package_dir = os.path.dirname(os.path.abspath(__file__))
+
+# 根据不同平台设置库搜索路径
+if sys.platform == 'win32':
+    # Windows使用PATH环境变量
+    if package_dir not in os.environ.get('PATH', '').split(';'):
+        os.environ['PATH'] = f"{package_dir};{os.environ.get('PATH', '')}"
+elif sys.platform == 'darwin':
+    # macOS使用DYLD_LIBRARY_PATH环境变量
+    if package_dir not in os.environ.get('DYLD_LIBRARY_PATH', '').split(':'):
+        os.environ['DYLD_LIBRARY_PATH'] = f"{package_dir}:{os.environ.get('DYLD_LIBRARY_PATH', '')}"
+else:
+    # Linux使用LD_LIBRARY_PATH环境变量
+    print(f"Linux: {package_dir}")
+    if package_dir not in os.environ.get('LD_LIBRARY_PATH', '').split(':'):
+        os.environ['LD_LIBRARY_PATH'] = f"{package_dir}:{os.environ.get('LD_LIBRARY_PATH', '')}"
+
+import nndeploy.base
+import nndeploy.device
+import nndeploy.ir
+import nndeploy.op
+import nndeploy.net
+import nndeploy.inference
+import nndeploy.dag
+import nndeploy.preprocess
+import nndeploy.tokenizer
+import nndeploy.codec
+import nndeploy.classification
+import nndeploy.detect
+import nndeploy.track
+import nndeploy.segment
+import nndeploy.matting
+import nndeploy.face
+import nndeploy.gan
+
+from .nndeploy import get_version, framework_init, framework_deinit
+from .nndeploy import __version__
+from .nndeploy import get_type_enum_json
+
+__all__ = ['get_version', 'framework_init', 'framework_deinit', '__version__', 'get_type_enum_json']

nndeploy-0.2.0/nndeploy/base/__init__.py

@@ -0,0 +1,133 @@
+
+from .common import all_type_enum
+
+# get by name
+from .common import name_to_data_type_code
+from .common import data_type_code_to_name
+from .common import DataTypeCode
+from .common import DataType
+
+# get by name
+from .common import name_to_device_type_code
+from .common import device_type_code_to_name
+from .common import DeviceTypeCode
+from .common import DeviceType
+
+from .common import name_to_data_format
+from .common import data_format_to_name
+from .common import DataFormat
+
+from .common import name_to_precision_type
+from .common import precision_type_to_name
+from .common import PrecisionType
+
+from .common import name_to_power_type
+from .common import power_type_to_name
+from .common import PowerType
+
+from .common import name_to_share_memory_type
+from .common import share_memory_type_to_name
+from .common import ShareMemoryType
+
+from .common import name_to_memory_type
+from .common import memory_type_to_name
+from .common import MemoryType
+
+from .common import name_to_memory_pool_type
+from .common import memory_pool_type_to_name
+from .common import MemoryPoolType
+
+from .common import name_to_tensor_type
+from .common import tensor_type_to_name
+from .common import TensorType
+
+from .common import name_to_forward_op_type
+from .common import forward_op_type_to_name
+from .common import ForwardOpType
+
+from .common import name_to_inference_opt_level
+from .common import inference_opt_level_to_name
+from .common import InferenceOptLevel
+
+from .common import name_to_model_type
+from .common import model_type_to_name
+from .common import ModelType
+
+from .common import name_to_inference_type
+from .common import inference_type_to_name
+from .common import InferenceType
+
+from .common import name_to_encrypt_type
+from .common import encrypt_type_to_name
+from .common import EncryptType
+
+from .common import name_to_codec_type
+from .common import codec_type_to_name
+from .common import CodecType
+
+from .common import name_to_codec_flag
+from .common import codec_flag_to_name
+from .common import CodecFlag
+
+from .common import name_to_parallel_type
+from .common import parallel_type_to_name
+from .common import ParallelType
+
+from .common import name_to_edge_type
+from .common import edge_type_to_name
+from .common import EdgeType
+
+from .common import name_to_edge_update_flag
+from .common import edge_update_flag_to_name
+from .common import EdgeUpdateFlag
+
+from .common import name_to_node_color_type
+from .common import node_color_type_to_name
+from .common import NodeColorType
+
+from .common import name_to_topo_sort_type
+from .common import topo_sort_type_to_name
+from .common import TopoSortType
+
+# get by name
+from .common import name_to_status_code
+from .common import status_code_to_name
+from .common import StatusCode
+from .common import Status
+
+from .common import name_to_pixel_type
+from .common import pixel_type_to_name
+from .common import PixelType
+
+from .common import name_to_cvt_color_type
+from .common import cvt_color_type_to_name
+from .common import CvtColorType
+
+from .common import name_to_interp_type
+from .common import interp_type_to_name
+from .common import InterpType
+
+from .common import name_to_border_type
+from .common import border_type_to_name
+from .common import BorderType
+
+from .common import TimeProfiler
+from .common import time_profiler_reset
+from .common import time_point_start
+from .common import time_point_end
+from .common import time_profiler_get_cost_time
+from .common import time_profiler_print
+from .common import time_profiler_print_index
+from .common import time_profiler_print_remove_warmup
+
+from .common import Param
+from .common import remove_json_brackets
+from .common import pretty_json_str
+
+from .common import load_library_from_path, free_library, get_library_handle
+
+from .file_utils import FileUtils, file_utils
+from .json_utils import JsonUtils, json_utils
+from .logger import Logger, logger, debug, info, warning, error, critical, exception
+
+