Forex Prediction Using meta learner LSTM

Urgent need to finish in 1 week maximum and must be WINDOWS compatible!!! ( I do not want compatibility issues if you are using a mac/apple) Code needs to run smoothly.
Only using PYTHON. Can use libraries like tensorflow, keras, machine learning packages etc

I need to you to build a few shot, meta learner LSTM model for forex prediction.
I have a dataset forex USD/EUR.

You need to hyper tune parameter variables to show that this technique is a good technique and can still produce good results as compared to other traditional methods of forex prediction. (Looking for high accuracy of near 90% with low loss)

Tunable Hyper parameters:( Only change one set of hyperparameters and show that in a separate notebook)
1) No.of hidden layers - 1,2,3
2) No. of Hidden neurons - 50,100,150
3) No. of Epochs - 50,100,200
4) Batch size - 16,32,64
5) Drop out rate - 0.2,0.3,0.4

You must have 5 notebooks to deliver at the end of the project.
You have to also calculate the loss for the performance metric such as MSE, MAE, MAPE, R^2, DS and EV for each notebook. Calculate mean and standard deviation where necessary. Then show the overall accuracy of the prediction in percentage, run at least for 10 times and show the average accuracy. You can see the example in the table below for the format table i'm looking for.

Thank you so much!