This paper extends the Momentum Transformer architecture, originally developed for futures and equity indices, to individual equities. The model combines an attention mechanism with a Long Short-Term Memory (LSTM) network, allowing it to capture long-term dependencies and adapt to changing market conditions. The authors backtest the strategy on a portfolio of stocks, accounting for transaction costs, and compare it to traditional time-series momentum and mean-reversion benchmarks. The results show average returns of 4.14% and a Sharpe ratio of 1.12, which are comparable to the original paper's findings but with higher volatility due to the inherent risk of equities. The model demonstrates robustness during the COVID-19 pandemic, highlighting its ability to handle market turbulence. This work contributes to the application of transformer-based models in quantitative finance, particularly for equity momentum strategies.