zcl912

Results 7 comments of zcl912

/pytorch/aten/src/ATen/native/cuda/IndexKernel.cu:60: lambda [](int)->auto::operator()(int)->auto: block: [0,0,0], thread: [121,0,0] Assertion `index >= -sizes[i] && index < sizes[i] && "index out of bounds"` failed.

> Could you provide the full log file? yeah, the full log is as follw: [04/08 19:21:13 c2.utils.dump.events]: eta: 5:54:29 iter: 1580/22500 total_loss: 0.915 loss_cls: 0.425 loss_box_reg: 0.483 time: 0.9901...

/pytorch/aten/src/ATen/native/cuda/IndexKernel.cu:60: lambda [](int)->auto::operator()(int)->auto: block: [0,0,0], thread: [121,0,0] Assertion `index >= -sizes[i] && index < sizes[i] && "index out of bounds"` failed.

okay, thanks for your reply. i just follw the commond: cd playground/detection/coco/yolof/yolof.res50.C5.1x pods_train --num-gpus 4 the log is: [04/08 19:32:57] cvpods INFO: Rank of current process: 0. World size: 4...

/pytorch/aten/src/ATen/native/cuda/IndexKernel.cu:60: lambda [](int)->auto::operator()(int)->auto: block: [0,0,0], thread: [121,0,0] Assertion `index >= -sizes[i] && index < sizes[i] && "index out of bounds"` failed.

> It seems that you modify the `BASE_LR`. But it's weird. We have not encountered this error before. > Could you try the standard 8 GPUs setting and train YOLOF...

/pytorch/aten/src/ATen/native/cuda/IndexKernel.cu:60: lambda [](int)->auto::operator()(int)->auto: block: [0,0,0], thread: [121,0,0] Assertion `index >= -sizes[i] && index < sizes[i] && "index out of bounds"` failed.

hello, the error is about the unusual value as follw: ![12](https://user-images.githubusercontent.com/35416036/114255842-8fdcb280-99e8-11eb-8b54-865da4adc1e0.png) ![13](https://user-images.githubusercontent.com/35416036/114255848-9bc87480-99e8-11eb-95cb-18a2f11d5e18.png)

/pytorch/aten/src/ATen/native/cuda/IndexKernel.cu:60: lambda [](int)->auto::operator()(int)->auto: block: [0,0,0], thread: [121,0,0] Assertion `index >= -sizes[i] && index < sizes[i] && "index out of bounds"` failed.

> It seems that there is somewhere the code performances ZeroDivision. Does this error occurs definitely or randomly? it occurs randomly

如何对任意一张图片进行活体检测？人脸检测，拿出人脸直接输入？？？或者这个人脸需要怎么处理？？

先检测出人脸吧

Asked for the dataset in the new form including bounding box

> I see you have offered the link of the dataset which is a npz file including train_data and meta_data, called h3wb_train.npz. But mmpose uses another one with bounding box...