- "loss" refers to the loss value over the training data after each epoch. This is what the optimization process is trying to minimize with the training so, the lower, the better.
- "accuracy" refers to the ratio between correct predictions and the total number of predictions in the training data. The higher, the better. This is normally inversely correlated with the loss, but not always.
The "validation" counterparts are the same concepts but computed over the validation data, which is not used for training and hence "unseen" to your model. If the training loss and accuracy are good but the validation counterparts are bad, it means that your model is "overfitting", as it can't generalize to unseen data.
Normally, these measures are plotted together into a training and validation loss plot and a training and validation accuracy plot, so that you can better evaluate their behavior over time. You can see some examples on how to interpret different trends here.
In your case, the training is not finished, so we can't be sure. From what we can see, the accuracy is less than 30%, so it is bad, but maybe more training leads to improvements.