Hi,
In a previous post I expressed I was having issues with the amount of time it took me to train the model.
I tried the model suggested on the assignment, but the training was still progressing too slow to make it happen in a reasonable time. I have a model achieving a similarity index of 0.55, so I decided to give it a go, and see if I could train it for a couple of days.
Now, for what I’ve seen there seems to be a relationship with the lost function and the similarity index. Could anyone tell me to what loss the trained their model before being able to pass the test. The model I mentioned above is below 270, so I suspect that if I take to 200 it might be able to pass.
Additionally, I am facing issues with the processor. With each epoch the ram is going down. I didn’t have this issue before, and it started to happen without any changes with the code. I dig a little, and there was an issue with memory leakage, but to fix it, the fit and compile methods need to be used.
I’m stuck here. So I need help again.
Thanks.