search

WXinlong / SOLO

SOLO and SOLOv2 for instance segmentation, ECCV 2020 & NeurIPS 2020.
Other
1.71k stars 306 forks source link

Segmentation fault (core dumped) #112

Open shakey-cuimiao opened 4 years ago

shakey-cuimiao commented 4 years ago

训练自己数据，出现内核溢出loading annotations into memory... Done (t=0.00s) creating index... index created! 2020-10-09 19:32:55,926 - mmdet - INFO - Start running, host: shakey@xiaoi-778, work_dir: /opt/shakey/road_car_objection/SOLO/work_dirs/toy/solov2_light_release_r50_fpn_2gpu_3x 2020-10-09 19:32:55,926 - mmdet - INFO - workflow: [('train', 1)], max: 72 epochs /usr/anaconda3/lib/python3.6/site-packages/torch/nn/functional.py:2479: UserWarning: Default upsampling behavior when mode=bilinear is changed to align_corners=False since 0.4.0. Please specify align_corners=True if the old behavior is desired. See the documentation of nn.Upsample for details. "See the documentation of nn.Upsample for details.".format(mode)) Segmentation fault (core dumped)

WXinlong commented 4 years ago

@shakey-cuimiao Is there enough memory?

zhangchaosunshine commented 3 years ago

@WXinlong how to reduce memory? I use single gpu to train model on just 20 pics data, this problem also is here. The computer have more than 20G memory(not GPU memory).

LiangYong1216 commented 3 years ago

@WXinlong how to reduce memory? I use single gpu to train model on just 20 pics data, this problem also is here. The computer have more than 20G memory(not GPU memory).

解决了吗？我也是一样的问题

zhangchaosunshine commented 3 years ago

@WXinlong how to reduce memory? I use single gpu to train model on just 20 pics data, this problem also is here. The computer have more than 20G memory(not GPU memory).

解决了吗？我也是一样的问题

没有

LiangYong1216 commented 3 years ago

有人说是预训练文件加载失败，可是也不知道需要怎样处理

------------------ 原始邮件 ------------------ 发件人: "WXinlong/SOLO" <notifications@github.com>; 发送时间: 2020年12月3日(星期四) 上午10:37 收件人: "WXinlong/SOLO"<SOLO@noreply.github.com>; 抄送: "泰山"<1343121924@qq.com>;"Comment"<comment@noreply.github.com>; 主题: Re: [WXinlong/SOLO] Segmentation fault (core dumped) (#112)

@WXinlong how to reduce memory? I use single gpu to train model on just 20 pics data, this problem also is here. The computer have more than 20G memory(not GPU memory).

解决了吗？我也是一样的问题

没有

— You are receiving this because you commented. Reply to this email directly, view it on GitHub, or unsubscribe.

zhangchaosunshine commented 3 years ago

有人说是预训练文件加载失败，可是也不知道需要怎样处理 … ------------------ 原始邮件 ------------------ 发件人: "WXinlong/SOLO" <notifications@github.com>; 发送时间: 2020年12月3日(星期四) 上午10:37 收件人: "WXinlong/SOLO"<SOLO@noreply.github.com>; 抄送: "泰山"<1343121924@qq.com>;"Comment"<comment@noreply.github.com>; 主题: Re: [WXinlong/SOLO] Segmentation fault (core dumped) (#112) @WXinlong how to reduce memory? I use single gpu to train model on just 20 pics data, this problem also is here. The computer have more than 20G memory(not GPU memory). 解决了吗？我也是一样的问题 没有 — You are receiving this because you commented. Reply to this email directly, view it on GitHub, or unsubscribe.

好像不是预训练加载的问题，我调试是到训练开始出错了， runner.run(data_loaders, cfg.workflow, cfg.total_epochs)就是这个。