Open sbjshxbxijs opened 1 year ago
The training result is as follow,why the val_accuracy is always 0.5 until 8/30 epoch ? And thanks for your answering!
200/200 [==============================] - 4883s 24s/step - loss: 0.7078 - accuracy: 0.4938 - val_loss: 0.6930 - val_accuracy: 0.5000 - lr: 0.0100 Epoch 2/30 200/200 [==============================] - 4889s 24s/step - loss: 0.6959 - accuracy: 0.4938 - val_loss: 0.6947 - val_accuracy: 0.5000 - lr: 0.0100 Epoch 3/30 200/200 [==============================] - 4889s 24s/step - loss: 0.6947 - accuracy: 0.5194 - val_loss: 0.7013 - val_accuracy: 0.5000 - lr: 0.0100 Epoch 4/30 200/200 [==============================] - 4882s 24s/step - loss: 0.6857 - accuracy: 0.5612 - val_loss: 0.6876 - val_accuracy: 0.5000 - lr: 0.0100 Epoch 5/30 200/200 [==============================] - 4912s 25s/step - loss: 0.6896 - accuracy: 0.5494 - val_loss: 0.6928 - val_accuracy: 0.5000 - lr: 0.0100 Epoch 6/30 200/200 [==============================] - 4955s 25s/step - loss: 0.6957 - accuracy: 0.5013 - val_loss: 0.6930 - val_accuracy: 0.5600 - lr: 0.0100 Epoch 7/30 200/200 [==============================] - 4975s 25s/step - loss: 0.6945 - accuracy: 0.5156 - val_loss: 0.6921 - val_accuracy: 0.5000 - lr: 0.0100 Epoch 8/30 63/200 [========>.....................] - ETA: 16:03:00 - loss: 0.6959 - accuracy: 0.4861^C^C^CTraceback (most recent call last):
The training result is as follow,why the val_accuracy is always 0.5 until 8/30 epoch ? And thanks for your answering!
200/200 [==============================] - 4883s 24s/step - loss: 0.7078 - accuracy: 0.4938 - val_loss: 0.6930 - val_accuracy: 0.5000 - lr: 0.0100 Epoch 2/30
200/200 [==============================] - 4889s 24s/step - loss: 0.6959 - accuracy: 0.4938 - val_loss: 0.6947 - val_accuracy: 0.5000 - lr: 0.0100 Epoch 3/30 200/200 [==============================] - 4889s 24s/step - loss: 0.6947 - accuracy: 0.5194 - val_loss: 0.7013 - val_accuracy: 0.5000 - lr: 0.0100 Epoch 4/30 200/200 [==============================] - 4882s 24s/step - loss: 0.6857 - accuracy: 0.5612 - val_loss: 0.6876 - val_accuracy: 0.5000 - lr: 0.0100 Epoch 5/30 200/200 [==============================] - 4912s 25s/step - loss: 0.6896 - accuracy: 0.5494 - val_loss: 0.6928 - val_accuracy: 0.5000 - lr: 0.0100 Epoch 6/30 200/200 [==============================] - 4955s 25s/step - loss: 0.6957 - accuracy: 0.5013 - val_loss: 0.6930 - val_accuracy: 0.5600 - lr: 0.0100 Epoch 7/30 200/200 [==============================] - 4975s 25s/step - loss: 0.6945 - accuracy: 0.5156 - val_loss: 0.6921 - val_accuracy: 0.5000 - lr: 0.0100 Epoch 8/30 63/200 [========>.....................] - ETA: 16:03:00 - loss: 0.6959 - accuracy: 0.4861^C^C^CTraceback (most recent call last):