SeekPoint
commented
7 years ago --- a/train_sje_hybrid.lua +++ b/train_sje_hybrid.lua @@ -254,7 +254,7 @@ for i = 1, iterations do local val_loss = 0 val_losses[i] = val_loss
- local savefile = string.format('%s/lm%s%.5f%.0f%.0f_%s.t7', opt.checkpoint_dir, opt.savefile, opt.learning_rate, opt.symmetric, opt.num_caption, opt.ids_file)
- local savefile = string.format('%s/lm%s%.5f%.0f%.0f.t7', opt.checkpoint_dir, opt.savefile, opt.learning_rate, opt.symmetric, opt.num_caption) print('saving checkpoint to ' .. savefile) local checkpoint = {} checkpoint.protos = protos rzai@rzai00:/media/rzai/ai_data/prj/cvpr2016$
fixed!!!
rzai@rzai00:~/prj/cvpr2016$ CUDA_VISIBLES_DEVICES=0 th train_sje_hybrid.lua -data_dir /media/rzai/ai_data/_reedscot/de_cvpr2016_cub.tar.gz -image_dir /media/rzai/ai_data/_reedscot/de_cvpr2016_cub.tar.gz/images -ids_file /media/rzai/ai_data/_reedscot/de_cvpr2016_cub.tar.gz/trainvalids.txt -learning_rate 0.0007 -symmetric 1 -max_epochs 200 -savefile sje_cub_c10_hybrid -num_caption 10 -gpuid 1 -print_every 10 2>&1 | tee yknote---train_sje_hybrid.lua---log { image_dir : "/media/rzai/ai_data/_reedscot/de_cvpr2016_cub.tar.gz/images" seed : 123 batch_size : 40 num_caption : 10 gpuid : 1 symmetric : 1 emb_dim : 1024 image_noop : 1 checkpoint_dir : "cv" bidirectional : 0 randomize_pair : 0 max_epochs : 200 savefile : "sje_cub_c10_hybrid" print_every : 10 data_dir : "/media/rzai/ai_data/_reedscot/de_cvpr2016_cub.tar.gz" image_dim : 1024 init_from : "" doc_length : 201 learning_rate_decay_after : 1 grad_clip : 5 avg : 0 eval_val_every : 1000 ids_file : "/media/rzai/ai_data/_reedscot/de_cvpr2016_cub.tar.gz/trainvalids.txt" nclass : 200 cnn_dim : 256 dropout : 0 learning_rate : 0.0007 learning_rate_decay : 0.98 flip : 0 } using CUDA on GPU 1...
10/30000 (ep 0.067), loss=2.01, acc1=0.00, acc2=0.4246, g/p=2.7138e-02, t/b=5.54s
many lines
990/30000 (ep 6.600), loss=0.52, acc1=30.00, acc2=24.6939, g/p=2.9552e-02, t/b=3.61s
saving checkpoint to cv/lm_sje_cub_c10_hybrid_0.00070_110/media/rzai/ai_data/_reedscot/de_cvpr2016_cub.tar.gz/trainvalids.txt.t7
/home/rzai/torch/install/bin/luajit: cannot open <cv/lm_sje_cub_c10_hybrid_0.00070_110/media/rzai/ai_data/_reedscot/de_cvpr2016_cub.tar.gz/trainvalids.txt.t7> in mode w at /home/rzai/torch/pkg/torch/lib/TH/THDiskFile.c:670 stack traceback: [C]: at 0x7f0e06a66ad0 [C]: in function 'DiskFile' /home/rzai/torch/install/share/lua/5.1/torch/File.lua:385: in function 'save' train_sje_hybrid.lua:268: in main chunk [C]: in function 'dofile' ...rzai/torch/install/lib/luarocks/rocks/trepl/scm-1/bin/th:145: in main chunk [C]: at 0x00406670 rzai@rzai00:~/prj/cvpr2016$ rzai@rzai00:~/prj/cvpr2016$ rzai@rzai00:~/prj/cvpr2016$