<!DOCTYPE html>
<html>
<head><title>torch_jit_subgraph_2</title></head>
<body>
<div id="App"></div>
</body>
<script type="text/babel">
class App extends React.Component {
constructor(props) {
super(props)
this.state = {
jsonObject: {"BPU OPs per frame (effective)":5903934474,"BPU OPs per run (effective)":141694427392,"BPU PE number":1,"BPU core number":1,"BPU march":"BAYES","DDR bytes per frame":5113646,"DDR bytes per run":122727520,"DDR bytes per second":1870890737,"DDR megabytes per frame":4.877,"DDR megabytes per run":117.042,"DDR megabytes per second":1784.2,"FPS":365.86,"HBDK version":"3.48.6","compiling options":"-f hbir -m /tmp/tmpg7k9pd50/torch_jit_subgraph_2.hbir -o /tmp/tmpg7k9pd50/torch_jit_subgraph_2.hbm --march bayes --progressbar --O3 --cache /open_explorer/2/model_output/cache.json --debug --core-num 1 --fast --input-layout NHWC --output-layout NHWC --input-source pyramid,ddr","frame per run":24,"frame per second":365.86,"input features":[["input name","input size"],["input.1","6x256x704x3"],["/Cast_17_output_0_calibrated_quantized","24x100x100x2"]],"interval computing unit utilization":[0.874,0.936,0.822,0.595,0.465,0.978,0.737,0.41,0.364,0.977,0.986,0.983,0.986,0.986,0.987,0.986,0.982,0.988,0.986,0.986,0.986,0.982,0.987,0.987,0.986,0.983,0.984,0.987,0.205,0.017,0.017,0.694,0.669],"interval computing units utilization":[0.874,0.936,0.822,0.595,0.465,0.978,0.737,0.41,0.364,0.977,0.986,0.983,0.986,0.986,0.987,0.986,0.982,0.988,0.986,0.986,0.986,0.982,0.987,0.987,0.986,0.983,0.984,0.987,0.205,0.017,0.017,0.694,0.669],"interval loading bandwidth (megabytes/s)":[1617,1925,2381,3712,4312,1873,1915,4559,4983,2407,151,144,146,148,144,146,148,144,146,150,144,145,150,144,145,150,144,145,297,484,525,1026,1273],"interval number":33,"interval storing bandwidth (megabytes/s)":[1798,1215,865,2947,4099,2220,1951,1432,1431,1502,112,81,80,81,81,81,81,80,81,81,80,81,81,81,81,80,81,81,50,9,0,753,1379],"interval time (ms)":2.0,"latency (ms)":65.6,"latency (ms) by segments":[65.598],"latency (us)":65598.4,"layer details":[["layer","ops","original output shape","aligned output shape","computing cost (no DDR)","load/store cost","active period of time"],["HZ_PREPROCESS_FOR_input.1","19,464,192","[6,256,704,3]","[6,256,704,8]","150 us (0.2% of model)","249 us (0.3% of model)","13 ~ 1752 us (1739)"],["/backbone/conv1/Conv-conv","5,086,642,176","[6,128,352,64]","[6,128,352,64]","505 us (0.7% of model)","3 us (0% of model)","20 ~ 1765 us (1745)"],["/backbone/maxpool/MaxPool","0","[6,64,176,64]","[6,64,176,64]","256 us (0.3% of model)","0","41 ~ 1771 us (1730)"],["/backbone/layer1/layer1.0/conv1/Conv-conv","4,982,833,152","[6,64,176,64]","[6,64,176,64]","256 us (0.3% of model)","5 us (0% of model)","22 ~ 1798 us (1776)"],["/backbone/layer1/layer1.0/conv2/Conv-conv","4,982,833,152","[6,64,176,64]","[6,64,176,64]","256 us (0.3% of model)","5 us (0% of model)","29 ~ 1825 us (1796)"],["/backbone/layer1/layer1.1/conv1/Conv-conv","4,982,833,152","[6,64,176,64]","[6,64,176,64]","256 us (0.3% of model)","5 us (0% of model)","41 ~ 1842 us (1801)"],["/backbone/layer1/layer1.1/conv2/Conv-conv","4,982,833,152","[6,64,176,64]","[6,64,176,64]","287 us (0.4% of model)","408 us (0.6% of model)","40 ~ 2711 us (2671)"],["/backbone/layer2/layer2.0/conv1/Conv-conv","2,491,416,576","[6,32,88,128]","[6,32,96,128]","170 us (0.2% of model)","9 us (0% of model)","1881 ~ 2787 us (906)"],["/backbone/layer2/layer2.0/conv2/Conv","4,982,833,152","[6,32,88,128]","[6,32,96,128]","276 us (0.4% of model)","16 us (0% of model)","1907 ~ 2858 us (951)"],["/backbone/layer2/layer2.0/downsample/downsample.0/Conv-conv","276,824,064","[6,32,88,128]","[6,32,96,128]","46 us (0% of model)","3 us (0% of model)","1881 ~ 2859 us (978)"],["/backbone/layer2/layer2.1/conv1/Conv-conv","4,982,833,152","[6,32,88,128]","[6,32,96,128]","276 us (0.4% of model)","16 us (0% of model)","1911 ~ 2905 us (994)"],["/backbone/layer2/layer2.1/conv2/Conv-conv","4,982,833,152","[6,32,88,128]","[6,32,96,128]","286 us (0.4% of model)","234 us (0.3% of model)","1926 ~ 3191 us (1265)"],["/backbone/layer3/layer3.0/conv1/Conv-conv","2,491,416,576","[6,16,44,256]","[6,16,48,256]","113 us (0.1% of model)","31 us (0% of model)","1959 ~ 3263 us (1304)"],["/backbone/layer3/layer3.0/conv2/Conv","4,982,833,152","[6,16,44,256]","[6,16,48,256]","183 us (0.2% of model)","27 us (0% of model)","2007 ~ 3309 us (1302)"],["/backbone/layer3/layer3.0/downsample/downsample.0/Conv-conv","276,824,064","[6,16,44,256]","[6,16,48,256]","32 us (0% of model)","114 us (0.1% of model)","2008 ~ 3799 us (1791)"],["/backbone/layer3/layer3.1/conv1/Conv-conv","4,982,833,152","[6,16,44,256]","[6,16,48,256]","275 us (0.4% of model)","27 us (0% of model)","2030 ~ 3845 us (1815)"],["/backbone/layer3/layer3.1/conv2/Conv-conv","4,982,833,152","[6,16,44,256]","[6,16,48,256]","285 us (0.4% of model)","130 us (0.1% of model)","2031 ~ 3962 us (1931)"],["/backbone/layer4/layer4.0/conv1/Conv-conv","2,491,416,576","[6,8,22,512]","[6,8,32,512]","226 us (0.3% of model)","51 us (0% of model)","3309 ~ 4188 us (879)"],["/backbone/layer4/layer4.0/conv2/Conv","4,982,833,152","[6,8,22,512]","[6,8,32,512]","362 us (0.5% of model)","104 us (0.1% of model)","4188 ~ 4566 us (378)"],["/backbone/layer4/layer4.0/downsample/downsample.0/Conv-conv","276,824,064","[6,8,22,512]","[6,8,32,512]","29 us (0% of model)","60 us (0% of model)","3962 ~ 4647 us (685)"],["/backbone/layer4/layer4.1/conv1/Conv-conv","4,982,833,152","[6,8,22,512]","[6,8,32,512]","364 us (0.5% of model)","108 us (0.1% of model)","4520 ~ 5050 us (530)"],["/backbone/layer4/layer4.1/conv2/Conv-conv","4,982,833,152","[6,8,22,512]","[6,8,32,512]","361 us (0.5% of model)","212 us (0.3% of model)","4632 ~ 5449 us (817)"],["/neck/lateral_convs.3/conv/Conv","69,206,016","[6,8,22,64]","[6,8,32,64]","7 us (0% of model)","3 us (0% of model)","5404 ~ 5510 us (106)"],["/neck/Resize","0","[6,16,44,64]","[6,16,44,64]","11 us (0% of model)","0","5510 ~ 5521 us (11)"],["/neck/lateral_convs.2/conv/Conv","138,412,032","[6,16,44,64]","[6,16,48,64]","12 us (0% of model)","53 us (0% of model)","5449 ~ 5541 us (92)"],["/neck/Resize_1","0","[6,32,88,64]","[6,32,88,64]","44 us (0% of model)","0","5541 ~ 5585 us (44)"],["/neck/lateral_convs.1/conv/Conv","276,824,064","[6,32,88,64]","[6,32,96,64]","35 us (0% of model)","360 us (0.5% of model)","5521 ~ 6788 us (1267)"],["/neck/Resize_2","0","[6,64,176,64]","[6,64,176,64]","175 us (0.2% of model)","0","5737 ~ 6855 us (1118)"],["/neck/lateral_convs.0/conv/Conv","553,648,128","[6,64,176,64]","[6,64,176,64]","31 us (0% of model)","202 us (0.3% of model)","5523 ~ 6864 us (1341)"],["/neck/fpn_convs.0/conv/Conv","4,982,833,152","[6,64,176,64]","[6,64,176,64]","132 us (0.2% of model)","3 us (0% of model)","5409 ~ 6892 us (1483)"],["/neck/fpn_convs.1/conv/Conv","1,245,708,288","[6,32,88,64]","[6,32,96,64]","89 us (0.1% of model)","3 us (0% of model)","5406 ~ 6868 us (1462)"],["/neck/fpn_convs.2/conv/Conv","311,427,072","[6,16,44,64]","[6,16,48,64]","25 us (0% of model)","71 us (0.1% of model)","5407 ~ 6835 us (1428)"],["/neck/fpn_convs.3/conv/Conv","77,856,768","[6,8,22,64]","[6,8,32,64]","9 us (0% of model)","28 us (0% of model)","5408 ~ 6830 us (1422)"],["/Resize","0","[6,64,176,64]","[6,64,176,64]","175 us (0.2% of model)","0","5755 ~ 6896 us (1141)"],["/Resize_1","0","[6,64,176,64]","[6,64,176,64]","175 us (0.2% of model)","0","5753 ~ 6894 us (1141)"],["/Resize_2","0","[6,64,176,64]","[6,64,176,64]","175 us (0.2% of model)","0","5750 ~ 6893 us (1143)"],["/Concat_6","0","[6,64,176,256]","[6,64,176,256]","237 us (0.3% of model)","0","5754 ~ 6912 us (1158)"],["/neck_fuse_0/Conv","19,931,332,608","[6,64,176,64]","[6,64,176,64]","635 us (0.9% of model)","932 us (1.4% of model)","4365 ~ 7796 us (3431)"],["/neck_fuse_0/Conv_NHWC2NCHW_LayoutConvert_Output0","0","[6,64,64,176]","[6,64,64,256]","125 us (0.1% of model)","704 us (1.0% of model)","7019 ~ 7866 us (847)"],["/
通过修改cuda-fastbev中的export-onnx.py文件，导出onnx并利用工具链转为bin文件，完成静态测试
需积分: 0 169 浏览量
2024-01-23
13:53:04
上传
评论
收藏 345.4MB ZIP 举报
0124.zip （45个子文件） 
0124 2 model_output 
torch_jit_subgraph_0.json 8KB 2_original_float_model.onnx 67.25MB
torch_jit_subgraph_0.html 2.16MB 2_optimized_float_model.onnx 67.18MB torch_jit_subgraph_2.json 28KB 2_quantized_model.onnx 57.55MB 2.bin 12.15MB torch_jit_subgraph_1.html 2.15MB torch_jit_subgraph_1.json 5KB hb_model_modifier.log 50KB hb_perf_result 2 torch_jit_subgraph_0.json 8KB 2.html 2KB torch_jit_subgraph_0.html 2.16MB torch_jit_subgraph_2.json 28KB torch_jit_subgraph_1.html 2.15MB torch_jit_subgraph_1.json 5KB 2 16KB torch_jit_subgraph_2.html 2.17MB temp.hbm 10.75MB
2.png 753KB hb_perf.log 2KB torch_jit_subgraph_2.html 2.17MB cache.json 1.75MB 2_calibrated_model.onnx 67.3MB hb_mapper_makertbin.log 42KB hb_mapper_checker.log 27KB .fast_perf 2_config.yaml 866B shape_inference_fail.onnx 59.85MB .hb_check torch_jit_subgraph_0.json 4KB quantized_model.onnx 57.01MB torch_jit_subgraph_0.html 2.15MB torch_jit_subgraph_2.json 4KB original_float_model.onnx 67.25MB calibrated_model.onnx 67.3MB torch_jit_subgraph_1.html 2.15MB optimized_float_model.onnx 67.18MB torch_jit_subgraph_1.json 4KB torch_jit_subgraph_2.html 2.15MB checker_hybrid_horizonrt.bin 19.39MB 2.onnx 67.25MB 替换viewtrans生成bin check.png 56KB hb_perf报错.png 33KB placeholder.png 18KB error shape infer.png 19KB onnx输入输出.png 27KB资源评论
weixin_52019990
- 粉丝: 0
- 资源: 71
最新资源
- 应用开发-画布技术-时钟-功能性小程序-画布时钟小程序.zip
- 一份关于navicat的大纲教程！！！！
- 一份关于maven的教程！！！！！！！！
- 关闭系统自带杀毒Windows Defender安全中心移除系统自带杀毒软件（防止软件被拦截打不开工具包）
- 基于SSM框架的局域网多人在线聊天系统
- 一份关于网络安全的大纲教程！！！！！！！
- SAPIEN PowerShell Studio 2024 v5.8.240 是一款功能强大且全面的集成开发环境（IDE）
- 计算机网络基础.zip
- 一份关于vue开发大纲的教程！！！！！！
- Xceed Ultimate Suite 24.1.25154.0957 是一款全面的 .NET 组件和控件集合
资源上传下载、课程学习等过程中有任何疑问或建议，欢迎提出宝贵意见哦~我们会及时处理！
点击此处反馈
