chapter -9 reinforcement learning (the training process slow and model is not getting updated )

[lakshay1ch]

I encountered an issue where training my model for 10 epochs did not result in updated model weights or improved performance. After running the training script in a Python file (rather than in an IPython notebook) over a 2-day period, I found that the saved model's weights were identical to the original, untrained model. Additionally, the model is not generating the expected buy/sell signals as per the training objectives. any suggestions updated or references you might like to add .
thank you in advance
training data buy sell signals:

back testing :