sinha96/LSTM — reverse-engineered prompt

I want a simple learning project that shows me how LSTMs work in practice, not just theory. Please build this as an easy to run Python notebook plus a couple of scripts, with clear comments and short explanations of what an RNN is, why LSTM helps with longer context, and what the main parts of the model are.

Use two examples like in the repo idea. One should do IMDB movie review sentiment classification, train the model, evaluate accuracy, and let me try a few sample predictions. The other should use the web traffic style csv data to do a basic time series prediction, with simple preprocessing, training, and a plot comparing actual vs predicted values.

Keep it beginner friendly and reproducible. If there is already a saved model file, use it when helpful but also show how to retrain from scratch. Include a concise README with run steps, what each file does, and expected outputs. Look up current docs online if you need to.