As we all know, deep learning need vast data. If you don't have this condition, you can use pre-training weights. Most of data can be fitted be pre-training weights, but there all still some data that can't converge to the global lowest point. So it is exist one weights that can let all task get best result?
Yes, this is "Model-Agnostic Meta-Learning". The biggest difference between MAML and pre-training weightsοΌPre-training weights minimize only for original task loss. MAML can minimize all task loss with a few steps of training.
- Pull repository.
git clone https://github.com/Runist/torch_maml.git
- You need to install some dependency package.
cd torch_maml
pip installl -r requirements.txt
- Download the Omiglot dataset.
mkdir data
cd data
wget https://github.com/Runist/MAML-keras/releases/download/v1.0/Omniglot.tar
tar -xvf Omniglot.tar
- Start training.
python train.py
π D:\Project\PycharmProject\meta-learning\torch_maml> python train.py
epoch 1: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:23<00:00, 5.87s/it, loss=1.1743]
=> loss: 1.2787 acc: 0.4875 val_loss: 1.0097 val_acc: 0.7051
epoch 2: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.41s/it, loss=1.0693]
=> loss: 1.1016 acc: 0.6760 val_loss: 0.9331 val_acc: 0.6838
epoch 3: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.39s/it, loss=0.9547]
=> loss: 0.9869 acc: 0.7063 val_loss: 0.8927 val_acc: 0.7421
epoch 4: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.39s/it, loss=0.8187]
=> loss: 0.8566 acc: 0.7740 val_loss: 0.8489 val_acc: 0.7435
epoch 5: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.39s/it, loss=0.7298]
=> loss: 0.7588 acc: 0.7849 val_loss: 0.6961 val_acc: 0.8347
epoch 6: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.37s/it, loss=0.6411]
=> loss: 0.6531 acc: 0.8010 val_loss: 0.6462 val_acc: 0.7764
epoch 7: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.38s/it, loss=0.5924]
=> loss: 0.6357 acc: 0.8036 val_loss: 0.6301 val_acc: 0.7954
epoch 8: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.40s/it, loss=0.4573]
=> loss: 0.5320 acc: 0.8474 val_loss: 0.7400 val_acc: 0.7227
epoch 9: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.39s/it, loss=0.5229]
=> loss: 0.5007 acc: 0.8422 val_loss: 0.5483 val_acc: 0.8093
epoch 10: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.49s/it, loss=0.4246]
=> loss: 0.4653 acc: 0.8484 val_loss: 0.5271 val_acc: 0.8338
epoch 11: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.39s/it, loss=0.4052]
=> loss: 0.4448 acc: 0.8719 val_loss: 0.5485 val_acc: 0.8569
epoch 12: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.39s/it, loss=0.4270]
=> loss: 0.4296 acc: 0.8755 val_loss: 0.4695 val_acc: 0.8417
epoch 13: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.39s/it, loss=0.3606]
=> loss: 0.4008 acc: 0.8771 val_loss: 0.4271 val_acc: 0.8889
epoch 14: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.39s/it, loss=0.3789]
=> loss: 0.3842 acc: 0.8724 val_loss: 0.4453 val_acc: 0.8630
epoch 15: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.39s/it, loss=0.3391]
=> loss: 0.3365 acc: 0.9078 val_loss: 0.4711 val_acc: 0.8301
epoch 16: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.42s/it, loss=0.3370]
=> loss: 0.3706 acc: 0.8911 val_loss: 0.4980 val_acc: 0.8218
epoch 17: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.41s/it, loss=0.3088]
=> loss: 0.3277 acc: 0.9021 val_loss: 0.4144 val_acc: 0.8500
epoch 18: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.58s/it, loss=0.2884]
=> loss: 0.3262 acc: 0.9000 val_loss: 0.3395 val_acc: 0.8838
epoch 19: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.54s/it, loss=0.3142]
=> loss: 0.3365 acc: 0.9021 val_loss: 0.3669 val_acc: 0.8796
epoch 20: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.53s/it, loss=0.2608]
=> loss: 0.3155 acc: 0.9016 val_loss: 0.3525 val_acc: 0.8963
=> loss: 0.2632 acc: 0.9057 val_loss: 0.3526 val_acc: 0.8731
epoch 30: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.44s/it, loss=0.2463]
=> loss: 0.2472 acc: 0.9276 val_loss: 0.3137 val_acc: 0.8856
epoch 31: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.44s/it, loss=0.1954]
=> loss: 0.2421 acc: 0.9214 val_loss: 0.3519 val_acc: 0.8847
epoch 32: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.44s/it, loss=0.2241]
=> loss: 0.2174 acc: 0.9349 val_loss: 0.2912 val_acc: 0.9269
epoch 33: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.44s/it, loss=0.2604]
=> loss: 0.2461 acc: 0.9193 val_loss: 0.3719 val_acc: 0.8773
epoch 34: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.57s/it, loss=0.2305]
=> loss: 0.2101 acc: 0.9370 val_loss: 0.2837 val_acc: 0.9255
epoch 35: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.68s/it, loss=0.3017]
=> loss: 0.2318 acc: 0.9281 val_loss: 0.3359 val_acc: 0.8838
epoch 36: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:19<00:00, 4.88s/it, loss=0.1916]
=> loss: 0.2044 acc: 0.9302 val_loss: 0.3705 val_acc: 0.8639
epoch 37: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.46s/it, loss=0.2420]
=> loss: 0.2346 acc: 0.9167 val_loss: 0.3386 val_acc: 0.8963
epoch 38: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.47s/it, loss=0.2465]
=> loss: 0.2306 acc: 0.9281 val_loss: 0.2690 val_acc: 0.9227
epoch 39: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.46s/it, loss=0.3123]
=> loss: 0.2259 acc: 0.9245 val_loss: 0.3907 val_acc: 0.8472
epoch 40: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.45s/it, loss=0.2130]
=> loss: 0.2168 acc: 0.9318 val_loss: 0.2451 val_acc: 0.9037
epoch 41: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.48s/it, loss=0.1830]
=> loss: 0.2101 acc: 0.9333 val_loss: 0.3580 val_acc: 0.8792
epoch 42: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.49s/it, loss=0.2006]
=> loss: 0.2407 acc: 0.9255 val_loss: 0.2819 val_acc: 0.9120
epoch 43: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.46s/it, loss=0.2414]
=> loss: 0.2685 acc: 0.9172 val_loss: 0.2838 val_acc: 0.8889
epoch 44: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.47s/it, loss=0.3090]
=> loss: 0.2473 acc: 0.9203 val_loss: 0.2593 val_acc: 0.9153
epoch 45: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.55s/it, loss=0.2398]
=> loss: 0.2599 acc: 0.9135 val_loss: 0.2581 val_acc: 0.9139
epoch 46: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.47s/it, loss=0.2263]
=> loss: 0.1915 acc: 0.9385 val_loss: 0.2459 val_acc: 0.9046
epoch 47: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.48s/it, loss=0.1584]
=> loss: 0.2061 acc: 0.9417 val_loss: 0.2303 val_acc: 0.9343
epoch 48: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.50s/it, loss=0.2782]
=> loss: 0.2388 acc: 0.9141 val_loss: 0.2254 val_acc: 0.9245
epoch 49: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.47s/it, loss=0.1847]
=> loss: 0.1931 acc: 0.9443 val_loss: 0.2470 val_acc: 0.9102
epoch 50: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.48s/it, loss=0.2025]
=> loss: 0.2535 acc: 0.9219 val_loss: 0.2644 val_acc: 0.8981
epoch 51: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.51s/it, loss=0.1399]
=> loss: 0.1858 acc: 0.9474 val_loss: 0.2852 val_acc: 0.8889
epoch 52: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.59s/it, loss=0.1754]
=> loss: 0.2063 acc: 0.9339 val_loss: 0.2984 val_acc: 0.8898
epoch 53: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.47s/it, loss=0.2186]
=> loss: 0.2058 acc: 0.9292 val_loss: 0.3300 val_acc: 0.8903
epoch 54: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.48s/it, loss=0.2796]
=> loss: 0.2115 acc: 0.9339 val_loss: 0.2653 val_acc: 0.9120
epoch 55: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.47s/it, loss=0.1592]
=> loss: 0.2081 acc: 0.9250 val_loss: 0.3142 val_acc: 0.8963
epoch 56: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.48s/it, loss=0.1393]
=> loss: 0.1700 acc: 0.9594 val_loss: 0.2857 val_acc: 0.8963
epoch 57: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.49s/it, loss=0.1609]
=> loss: 0.2252 acc: 0.9151 val_loss: 0.3261 val_acc: 0.8722
epoch 58: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.48s/it, loss=0.2672]
=> loss: 0.2146 acc: 0.9292 val_loss: 0.1960 val_acc: 0.9468
epoch 59: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.49s/it, loss=0.1900]
=> loss: 0.2088 acc: 0.9344 val_loss: 0.3075 val_acc: 0.8856
epoch 60: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.49s/it, loss=0.1501]
=> loss: 0.1923 acc: 0.9406 val_loss: 0.3069 val_acc: 0.8972
epoch 61: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.49s/it, loss=0.1210]
=> loss: 0.1849 acc: 0.9411 val_loss: 0.2046 val_acc: 0.9444
epoch 62: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.50s/it, loss=0.1057]
=> loss: 0.1667 acc: 0.9615 val_loss: 0.2852 val_acc: 0.8704
epoch 63: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.49s/it, loss=0.2624]
=> loss: 0.2138 acc: 0.9349 val_loss: 0.2348 val_acc: 0.9245
epoch 64: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.51s/it, loss=0.2535]
=> loss: 0.2031 acc: 0.9359 val_loss: 0.3830 val_acc: 0.8556
epoch 65: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.51s/it, loss=0.1625]
=> loss: 0.1686 acc: 0.9500 val_loss: 0.2560 val_acc: 0.8963
epoch 66: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.50s/it, loss=0.1426]
=> loss: 0.1741 acc: 0.9542 val_loss: 0.2626 val_acc: 0.9097
epoch 67: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.50s/it, loss=0.2304]
=> loss: 0.2280 acc: 0.9250 val_loss: 0.2551 val_acc: 0.9296
epoch 68: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.48s/it, loss=0.1911]
=> loss: 0.2032 acc: 0.9432 val_loss: 0.2756 val_acc: 0.8829
epoch 69: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.51s/it, loss=0.2287]
=> loss: 0.2100 acc: 0.9312 val_loss: 0.2418 val_acc: 0.8880
epoch 70: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.51s/it, loss=0.2269]
=> loss: 0.1805 acc: 0.9396 val_loss: 0.2529 val_acc: 0.9120
epoch 71: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.50s/it, loss=0.2278]
=> loss: 0.2004 acc: 0.9333 val_loss: 0.3293 val_acc: 0.8847
epoch 72: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.50s/it, loss=0.2843]
=> loss: 0.2190 acc: 0.9318 val_loss: 0.2010 val_acc: 0.9528
epoch 73: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.49s/it, loss=0.1696]
=> loss: 0.1608 acc: 0.9464 val_loss: 0.2368 val_acc: 0.9106
epoch 74: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.51s/it, loss=0.1702]
=> loss: 0.2028 acc: 0.9286 val_loss: 0.2693 val_acc: 0.9139
epoch 75: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.51s/it, loss=0.2240]
=> loss: 0.2104 acc: 0.9260 val_loss: 0.3673 val_acc: 0.8616
epoch 76: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.50s/it, loss=0.1861]
=> loss: 0.1906 acc: 0.9484 val_loss: 0.2067 val_acc: 0.9403
epoch 77: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.56s/it, loss=0.1104]
=> loss: 0.1578 acc: 0.9479 val_loss: 0.2596 val_acc: 0.8931
epoch 78: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.50s/it, loss=0.1739]
=> loss: 0.1863 acc: 0.9557 val_loss: 0.2426 val_acc: 0.9088
epoch 79: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.52s/it, loss=0.2283]
=> loss: 0.1911 acc: 0.9411 val_loss: 0.2176 val_acc: 0.9162
epoch 80: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.49s/it, loss=0.2052]
=> loss: 0.1613 acc: 0.9505 val_loss: 0.3198 val_acc: 0.8931
epoch 81: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.54s/it, loss=0.2082]
=> loss: 0.1711 acc: 0.9443 val_loss: 0.3029 val_acc: 0.8972
epoch 82: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.50s/it, loss=0.1442]
=> loss: 0.1347 acc: 0.9594 val_loss: 0.2400 val_acc: 0.9176
epoch 83: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.53s/it, loss=0.0943]
=> loss: 0.1564 acc: 0.9547 val_loss: 0.1767 val_acc: 0.9370
epoch 84: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.51s/it, loss=0.1367]
=> loss: 0.1681 acc: 0.9422 val_loss: 0.2178 val_acc: 0.9370
epoch 85: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:17<00:00, 4.49s/it, loss=0.0944]
=> loss: 0.1318 acc: 0.9583 val_loss: 0.3784 val_acc: 0.8486
epoch 86: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.57s/it, loss=0.1559]
=> loss: 0.1826 acc: 0.9443 val_loss: 0.2204 val_acc: 0.9153
epoch 87: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.51s/it, loss=0.1637]
=> loss: 0.1806 acc: 0.9443 val_loss: 0.1868 val_acc: 0.9287
epoch 88: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.52s/it, loss=0.2083]
=> loss: 0.1588 acc: 0.9516 val_loss: 0.2280 val_acc: 0.9269
epoch 89: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.52s/it, loss=0.1089]
=> loss: 0.1587 acc: 0.9510 val_loss: 0.2686 val_acc: 0.9014
epoch 90: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.54s/it, loss=0.1239]
=> loss: 0.1572 acc: 0.9458 val_loss: 0.2313 val_acc: 0.9139
epoch 91: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.52s/it, loss=0.0947]
=> loss: 0.1691 acc: 0.9484 val_loss: 0.2289 val_acc: 0.9019
epoch 92: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.52s/it, loss=0.1288]
=> loss: 0.1848 acc: 0.9411 val_loss: 0.2358 val_acc: 0.9319
epoch 93: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.52s/it, loss=0.1345]
=> loss: 0.1595 acc: 0.9438 val_loss: 0.2486 val_acc: 0.9120
epoch 94: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.52s/it, loss=0.1175]
=> loss: 0.1538 acc: 0.9557 val_loss: 0.2092 val_acc: 0.9148
epoch 95: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.55s/it, loss=0.0839]
=> loss: 0.1137 acc: 0.9682 val_loss: 0.3513 val_acc: 0.8755
epoch 96: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.54s/it, loss=0.1327]
=> loss: 0.1436 acc: 0.9568 val_loss: 0.1518 val_acc: 0.9551
epoch 97: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.55s/it, loss=0.1060]
=> loss: 0.1744 acc: 0.9354 val_loss: 0.1875 val_acc: 0.9417
epoch 98: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.54s/it, loss=0.1253]
=> loss: 0.1790 acc: 0.9495 val_loss: 0.2388 val_acc: 0.9222
epoch 99: 100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.55s/it, loss=0.1510]
=> loss: 0.1525 acc: 0.9490 val_loss: 0.1852 val_acc: 0.9560
epoch 100: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.52s/it, loss=0.0874]
=> loss: 0.1273 acc: 0.9573 val_loss: 0.1701 val_acc: 0.9551
epoch 101: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.54s/it, loss=0.1852]
=> loss: 0.1696 acc: 0.9500 val_loss: 0.2575 val_acc: 0.9069
epoch 102: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.55s/it, loss=0.1546]
=> loss: 0.1344 acc: 0.9661 val_loss: 0.2428 val_acc: 0.9403
epoch 103: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.50s/it, loss=0.2449]
=> loss: 0.1643 acc: 0.9349 val_loss: 0.2420 val_acc: 0.9139
epoch 104: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.64s/it, loss=0.1551]
=> loss: 0.1465 acc: 0.9615 val_loss: 0.1834 val_acc: 0.9384
epoch 105: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.54s/it, loss=0.1251]
=> loss: 0.1326 acc: 0.9589 val_loss: 0.2494 val_acc: 0.9014
epoch 106: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.52s/it, loss=0.1428]
=> loss: 0.1663 acc: 0.9427 val_loss: 0.2274 val_acc: 0.9301
epoch 107: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.56s/it, loss=0.1166]
=> loss: 0.1762 acc: 0.9406 val_loss: 0.2007 val_acc: 0.9352
epoch 108: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.53s/it, loss=0.2075]
=> loss: 0.1509 acc: 0.9448 val_loss: 0.1930 val_acc: 0.9394
epoch 109: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.57s/it, loss=0.1456]
=> loss: 0.1616 acc: 0.9505 val_loss: 0.2464 val_acc: 0.9032
epoch 110: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.55s/it, loss=0.1384]
=> loss: 0.1057 acc: 0.9724 val_loss: 0.1942 val_acc: 0.9468
epoch 111: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.53s/it, loss=0.0964]
=> loss: 0.1239 acc: 0.9620 val_loss: 0.1673 val_acc: 0.9509
epoch 112: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.55s/it, loss=0.0816]
=> loss: 0.1107 acc: 0.9667 val_loss: 0.1635 val_acc: 0.9426
epoch 113: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.53s/it, loss=0.1003]
=> loss: 0.1240 acc: 0.9578 val_loss: 0.2049 val_acc: 0.9278
epoch 114: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.57s/it, loss=0.0854]
=> loss: 0.1516 acc: 0.9474 val_loss: 0.1071 val_acc: 0.9727
epoch 115: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.54s/it, loss=0.1121]
=> loss: 0.1185 acc: 0.9589 val_loss: 0.3242 val_acc: 0.8755
epoch 116: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.54s/it, loss=0.1862]
=> loss: 0.1596 acc: 0.9490 val_loss: 0.2452 val_acc: 0.9014
epoch 117: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.53s/it, loss=0.1630]
=> loss: 0.1317 acc: 0.9578 val_loss: 0.1381 val_acc: 0.9644
epoch 118: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.58s/it, loss=0.0942]
=> loss: 0.1391 acc: 0.9542 val_loss: 0.1245 val_acc: 0.9477
epoch 119: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.53s/it, loss=0.0928]
=> loss: 0.1355 acc: 0.9536 val_loss: 0.1122 val_acc: 0.9662
epoch 120: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.54s/it, loss=0.1093]
=> loss: 0.1271 acc: 0.9667 val_loss: 0.2112 val_acc: 0.9190
epoch 121: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.56s/it, loss=0.1479]
=> loss: 0.1449 acc: 0.9536 val_loss: 0.2363 val_acc: 0.9194
epoch 122: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.56s/it, loss=0.1404]
=> loss: 0.1374 acc: 0.9490 val_loss: 0.1633 val_acc: 0.9528
epoch 123: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.56s/it, loss=0.0874]
=> loss: 0.1186 acc: 0.9573 val_loss: 0.2437 val_acc: 0.9264
epoch 124: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.59s/it, loss=0.0953]
=> loss: 0.1050 acc: 0.9766 val_loss: 0.2703 val_acc: 0.9046
epoch 125: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.57s/it, loss=0.0772]
=> loss: 0.1096 acc: 0.9724 val_loss: 0.1839 val_acc: 0.9213
epoch 126: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.57s/it, loss=0.1453]
=> loss: 0.1479 acc: 0.9443 val_loss: 0.2026 val_acc: 0.9204
epoch 127: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.56s/it, loss=0.1355]
=> loss: 0.0998 acc: 0.9641 val_loss: 0.2172 val_acc: 0.9370
epoch 128: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.55s/it, loss=0.0423]
=> loss: 0.0900 acc: 0.9781 val_loss: 0.1886 val_acc: 0.9477
epoch 129: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.58s/it, loss=0.0985]
=> loss: 0.1081 acc: 0.9661 val_loss: 0.2291 val_acc: 0.9338
epoch 130: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.60s/it, loss=0.0916]
=> loss: 0.1010 acc: 0.9672 val_loss: 0.2025 val_acc: 0.9444
epoch 131: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.60s/it, loss=0.1275]
=> loss: 0.0979 acc: 0.9667 val_loss: 0.1370 val_acc: 0.9509
epoch 132: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.57s/it, loss=0.1646]
=> loss: 0.1377 acc: 0.9495 val_loss: 0.2226 val_acc: 0.9236
epoch 133: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.58s/it, loss=0.0923]
=> loss: 0.1317 acc: 0.9599 val_loss: 0.1983 val_acc: 0.9301
epoch 134: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.59s/it, loss=0.1300]
=> loss: 0.1207 acc: 0.9703 val_loss: 0.1574 val_acc: 0.9634
epoch 135: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.59s/it, loss=0.1101]
=> loss: 0.1115 acc: 0.9693 val_loss: 0.1519 val_acc: 0.9403
epoch 136: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.58s/it, loss=0.2028]
=> loss: 0.1257 acc: 0.9651 val_loss: 0.1428 val_acc: 0.9477
epoch 137: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.57s/it, loss=0.2447]
=> loss: 0.1303 acc: 0.9635 val_loss: 0.1509 val_acc: 0.9542
epoch 138: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.60s/it, loss=0.0656]
=> loss: 0.1031 acc: 0.9693 val_loss: 0.1291 val_acc: 0.9463
epoch 139: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:20<00:00, 5.11s/it, loss=0.0965]
=> loss: 0.1446 acc: 0.9469 val_loss: 0.1789 val_acc: 0.9468
epoch 140: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.67s/it, loss=0.0601]
=> loss: 0.0838 acc: 0.9776 val_loss: 0.2036 val_acc: 0.9245
epoch 141: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.59s/it, loss=0.1859]
=> loss: 0.1299 acc: 0.9615 val_loss: 0.1321 val_acc: 0.9509
epoch 142: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.56s/it, loss=0.1089]
=> loss: 0.1215 acc: 0.9620 val_loss: 0.1706 val_acc: 0.9528
epoch 143: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.60s/it, loss=0.0766]
=> loss: 0.0802 acc: 0.9656 val_loss: 0.1229 val_acc: 0.9644
epoch 144: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.55s/it, loss=0.1345]
=> loss: 0.1171 acc: 0.9589 val_loss: 0.1693 val_acc: 0.9583
epoch 145: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.54s/it, loss=0.1162]
=> loss: 0.0988 acc: 0.9682 val_loss: 0.1778 val_acc: 0.9468
epoch 146: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.63s/it, loss=0.1219]
=> loss: 0.0852 acc: 0.9807 val_loss: 0.1585 val_acc: 0.9343
epoch 147: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.59s/it, loss=0.1458]
=> loss: 0.1014 acc: 0.9656 val_loss: 0.1525 val_acc: 0.9426
epoch 148: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.61s/it, loss=0.1334]
=> loss: 0.1445 acc: 0.9505 val_loss: 0.1260 val_acc: 0.9477
epoch 149: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.59s/it, loss=0.1831]
=> loss: 0.0942 acc: 0.9781 val_loss: 0.1478 val_acc: 0.9620
epoch 150: 100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 4/4 [00:18<00:00, 4.57s/it, loss=0.1357]
=> loss: 0.1123 acc: 0.9656 val_loss: 0.1432 val_acc: 0.9551
- You should set same parameters in args.py. More detail you can get in my blog.
parser.add_argument('--train_data_dir', type=str,
default="./data/Omniglot/images_background/",
help='The directory containing the train image data.')
parser.add_argument('--val_data_dir', type=str,
default="./data/Omniglot/images_evaluation/",
help='The directory containing the validation image data.')
parser.add_argument('--n_way', type=int, default=10,
help='The number of class of every task.')
parser.add_argument('--k_shot', type=int, default=1,
help='The number of support set image for every task.')
parser.add_argument('--q_query', type=int, default=1,
help='The number of query set image for every task.')
- Start training.
python train.py --n_way=5 --k_shot=1 --q_query=1