There could be various reasons for this issue.
The learning rate is too high, causing the model to overshoot the optimal solution and get stuck in a local minimum during the first step of an epoch.
There could be an issue with the data, such as corrupted or invalid input data that causes the model to terminate.
The model architecture may not be able to fit the data correctly, leading to convergence issues.
Network connectivity or hardware issues could cause the training to stop prematurely. For example, if the hardware is overheating or the GPU memory has been exhausted, this could result in the training process becoming unresponsive.
Lastly, an issue with the Keras framework or its dependencies may also cause abrupt halting of training.
Asked: 2023-07-06 16:50:24 +0000
Seen: 8 times
Last updated: Jul 06 '23