prefetch data error during training

[daniellu123]

Dear Limin,

During training the optical flow network, I got an error that training was quited at prefecting data in the layer of SequenceDatalayer.
The log is as follow:
*** Aborted at 1540548898 (unix time) try "date -d @1540548898" if you are using GNU date ***
PC: @ 0x7f46cdec8f3d caffe::SequenceDataLayer<>::InternalThreadEntry()
*** SIGFPE (@0x7f46cdec8f3d) received by PID 29760 (TID 0x7f467dc61700) from PID 18446744072869416765; stack trace: ***
@ 0x7f46cce4f4b0 (unknown)
@ 0x7f46cdec8f3d caffe::SequenceDataLayer<>::InternalThreadEntry()
@ 0x7f46cca015d5 (unknown)
@ 0x7f46cc7da6ba start_thread
@ 0x7f46ccf2141d clone
@ 0x0 (unknown)

the solver's setting is as below:
net: "../models/four_class/temporal_102_class_hard_bn_inception_train_val.prototxt"

testing parameter

test_iter: 2710
test_interval: 500
test_initialization: true

output

display: 100
average_loss: 100
snapshot: 500
snapshot_prefix: "../models/four_class/flow_finetune/temporal_untrimmednet_hard_bn_inception_average_seg3_top3"
debug_info: false

learning rate

base_lr: 0.001
lr_policy: "multistep"
gamma: 0.1
stepvalue: [10000, 15000, 20000]
max_iter: 40000
iter_size: 1

parameter of SGD

momentum: 0.9
weight_decay: 0.0005
clip_gradients: 20

GPU setting

solver_mode: GPU
device_id: [1,0]
richness: 200

I finetuned the network with weight: anet1.2_temporal_untrimmednet_hard_bn_inception.caffemodel, and batch_size = 5.

Could help indicate what's wrong in my settings?
Thanks very much!

[wanglimin]

You need to check whether your file name and file path is correct.

[daniellu123]

Thanks for your quick reply. I have checked the file name and file path by adding LOG(INFO) in io.cpp to print the flow image names, and there are valid. On the other side, there will be some log info if any image is failed to open in io.cpp. The dataset works fine in the bench of temporal-segment-networks.

[daniellu123]

This issue is solved by reducing the scale of ucf101 dataset by 0.3. However, it's a little weird.

[daniellu123]

This issue is solved right now. It's the problem of the shot files. Some shot files have no proposal(Shot file is empty).