[每日一氵] mmSeg 报错 StopIteration

错误是这样的：

2022-03-19 19:16:29,155 - mmseg - INFO - Checkpoints will be saved to /root/mmsegmentation/work_dirs/mine_xp by HardDiskBackend.
Traceback (most recent call last):File "xxxx/lib/python3.9/site-packages/mmcv/runner/iter_based_runner.py", line 32, in __next__data = next(self.iter_loader)File "xxxx/envs/pytorch1.8/lib/python3.9/site-packages/torch/utils/data/dataloader.py", line 517, in __next__data = self._next_data()File "xxxx/envs/pytorch1.8/lib/python3.9/site-packages/torch/utils/data/dataloader.py", line 1172, in _next_dataraise StopIteration
StopIterationDuring handling of the above exception, another exception occurred:Traceback (most recent call last):File "xxxx/mmsegmentation/tools/train.py", line 243, in <module>main()File "xxxx/mmsegmentation/tools/train.py", line 232, in maintrain_segmentor(File "xxxx/mmsegmentation/mmseg/apis/train.py", line 174, in train_segmentorrunner.run(data_loaders, cfg.workflow)File "xxxx/lib/python3.9/site-packages/mmcv/runner/iter_based_runner.py", line 134, in runiter_runner(iter_loaders[i], **kwargs)File "xxxx/lib/python3.9/site-packages/mmcv/runner/iter_based_runner.py", line 59, in traindata_batch = next(data_loader)File "xxxx/lib/python3.9/site-packages/mmcv/runner/iter_based_runner.py", line 39, in __next__data = next(self.iter_loader)File "xxxx/envs/pytorch1.8/lib/python3.9/site-packages/torch/utils/data/dataloader.py", line 517, in __next__data = self._next_data()File "xxxx/envs/pytorch1.8/lib/python3.9/site-packages/torch/utils/data/dataloader.py", line 1172, in _next_dataraise StopIteration
StopIteration

BATCH Size 我是这样设置的：

data = dict(samples_per_gpu=2,workers_per_gpu=4,train=dict(type="CustomDataset",data_root=data_root,img_dir='imagesTr_jpg',ann_dir='imagesTr_ann',seg_map_suffix='.jpg',pipeline=train_pipeline),......
)

我扫了一眼数据集，数据集只有一张照片，由于samples_per_gpu=2，所以会报StopIteration

改成：

    samples_per_gpu=1,workers_per_gpu=1,

或者多放几张照片

[每日一氵] mmSeg 报错 StopIteration相关推荐

使用next(iter(data.DataLoader())报错StopIteration
使用next(iter(data.DataLoader())报错StopIteration,这是因为当使用next()去访问一个已经迭代完的迭代器时,就会有触发这样的报错:StopIteration, ...
每日分享 hive hue 报错Error while compiling statement: FAILED: SemanticException Error in parsing
报错原因 SemanticException Error in parsing 翻译过来大意就是解析中的 SemanticException(语义异常) 错误解决办法应该修改成 a.dcbh 而 ...
每日分享环境报错：Exception in thread “main“ java.lang.RuntimeException: Cannot create staging directory
Exception in thread "main" java.lang.RuntimeException: Cannot create staging directory 'hd ...
每日10行代码57: appium测试坚果手机出现socket hang up报错的解决
今天在用appium测试的时候,老是出现socket hang up的报错,在网上搜索了下,找到了答案,原来是手机的系统自动把appium需要的后台进程给优化掉了. 解决办法也很简单,就是设置系统,不 ...
每日一记：zabbix日志报错cannot process item fra[usable,pct] trap: item type 15 cannot be used with traps
zabbix日志报错截取如下: 26099:20190718:110102.608 cannot process item "fra[usable,pct]" trap: item ...
解决微信H5获取SDK授权报错提示errMsg: “config:fail,Error: 系统错误，错误码：63002,invalid signature [20200908 22:17:17][]“
如果常规检查都做过可以仔细看下微信开放文档这个里面的报错原因如果都不是那么极有可能是(恭喜你!看样子你的H5页面日活人数还蛮多[呲牙]) 获取腾讯微信平台access_token超过每日默认上限2 ...
python函数第二次运行报错_Python学习笔记2-函数、高级特性
函数调用函数要调用一个函数,需要知道函数的名称和参数,比如求绝对值的函数abs,只有一个参数 >>> abs(100) 100 >>> abs(-20) 20 ...
spark调用python_在MRS集群中使用Python3.7运行PySpark程序，调用RDD的take函数报错处理...
现象如果我们安装了python3.7版本来运行spark,并且用到了RDD的take函数,就会报错:RuntimeError: generator raised StopIteration 我们可以 ...
python语法报错_Python语法的常见错误和处理异常
在学习Python的过程中,一定会遇到很多困难,我相信很多时候的,报错就会是让人抓狂,一个好的程序员最主要的特质,就是处理错误的能力.在面对每次遇到的问题,要多总结多做笔记,争取下次避免犯同样的错误, ...
hue集成mysql报错_hue集成hive访问报database is locked
这个问题这应该是hue默认的SQLite数据库出现错误,你可以使用mysql postgresql等来替换 hue默认使用sqlite作为元数据库,不推荐在生产环境中使用.会经常出现database ...

[每日一氵] mmSeg 报错 StopIteration

[每日一氵] mmSeg 报错 StopIteration相关推荐

最新文章

热门文章