网站建设的推广渠道,网站制作学校,上海建科建设监理网站,如何做平台推广安装tensorRT#xff1a;
1、下载与电脑中cuda和cudnn版本对应的tensorRT#xff08;比如我的是TensorRT-8.2.1.8.Windows10.x86_64.cuda-11.4.cudnn8.2#xff09;
2、打开目录里面有python文件夹#xff0c;找到对应python版本的whl文件#xff08;我的是tensorrt-8.2… 安装tensorRT
1、下载与电脑中cuda和cudnn版本对应的tensorRT比如我的是TensorRT-8.2.1.8.Windows10.x86_64.cuda-11.4.cudnn8.2
2、打开目录里面有python文件夹找到对应python版本的whl文件我的是tensorrt-8.2.1.8-cp38-none-win_amd64.whl 因为我python安装的是3.8版本
3、终端安装pip install tensorrt-8.2.1.8-cp38-none-win_amd64.whl
4、结束 import tensorrt as trt
def get_DynEngine(onnx_file_path, engine_file_path,patchsize,max_workspace_size,max_batch_size):Attempts to load a serialized engine if available,otherwise build a new TensorRT engine as save itTRT_LOGGER trt.Logger()trt.init_libnvinfer_plugins(TRT_LOGGER, )explicit_batch 1 (int)(trt.NetworkDefinitionCreationFlag.EXPLICIT_BATCH)builder trt.Builder(TRT_LOGGER)network builder.create_network(explicit_batch)config builder.create_builder_config()parser trt.OnnxParser(network, TRT_LOGGER)runtime trt.Runtime(TRT_LOGGER)print(common.EXPLICIT_BATCH:, explicit_batch)# 最大内存占用# 显存溢出需要重新设置config.max_workspace_size max_workspace_size # 256MBconfig.set_flag(trt.BuilderFlag.FP16)print(max_workspace_size:, config.max_workspace_size)builder.max_batch_size max_batch_size # 推理的时候要保证batch_sizemax_batch_sizeif not os.path.exists(onnx_file_path):print(fonnx file {onnx_file_path} not found,please run torch_2_onnx.py first to generate it)exit(0)print(fLoading ONNX file from path {onnx_file_path}...)with open(onnx_file_path, rb) as model:print(Beginning ONNX file parsing)if not parser.parse(model.read()):print(ERROR:Failed to parse the ONNX file)for error in range(parser.num_errors):print(parser.get_error(error))return Noneinputs [network.get_input(i) for i in range(network.num_inputs)]print(input, inputs)outputs [network.get_output(i) for i in range(network.num_outputs)]print(out:, outputs)print(Network Description)for input in inputs:# 获取当前转化之前的 输入的 batch_sizebatch_size input.shape[0]print(Input {} with shape {} and dtype {} . .format(input.name, input.shape, input.dtype))for output in outputs:print(Output {} with shape {} and dtype {} . .format(output.name, output.shape, output.dtype))# Dynamic input setting 动态输入在builder的profile设置# 为每个动态输入绑定一个profileprofile builder.create_optimization_profile()print(network.get_input(0).name:, network.get_input(0).name)profile.set_shape(network.get_input(0).name, (1,1, *patchsize), (1, 1,*patchsize),(max_batch_size, 1, *patchsize)) # 最小的尺寸,常用的尺寸,最大的尺寸,推理时候输入需要在这个范围内config.add_optimization_profile(profile)print(Completed parsing the ONNX file)print(fBuilding an engine from file {onnx_file_path}; this may take a while...)engine builder.build_serialized_network(network, config)print(Completed creating Engine)with open(engine_file_path, wb) as f:f.write(engine)return engineif __name__ __main__:get_DynEngine(1.onnx, 2.engine,[96,160,160],5*(130),2)
