CIFAR- Resnet34 Model
SGD optimizer, learning rate = 0.01 , momentum = 0.9 and cross-entropy as the loss
function
Created on May 27|Last edited on May 27
Comment
Train Loss, Test Accuracy and Test loss for
Transfer Learning from Resnet34 Model
We can see that as the number of epochs rises, model get well trained which increases accuracy and resulted in lowering down the Test Loss.
Optimizer- SGD with LR=0.001 and Momentum=0.9
Loss Function= Cross Entropy
Here, we observed Test Accurarcy increases up to 85 % mark. This is the best model amongst 3 models.
Run: fearless-silence-59
1
Sample Prediction Samples
Run: fearless-silence-59
1
We obtained much better and faster result in transfer learning case of Resnet34. This is happening because Resnet Model is pretrained already on very large image dataset.
Run: fearless-silence-59
1
Add a comment