Our code for our Problem Writing training is stored within our Phase 2 in the Problem Writing folder. Our new results for Problem Solving are stored in the Phase 3 folder. To create our visualizations, we used the plot.ipynb notebook with our specific directories in google drive. We used a lot of models we created from our experiments to create the visualizations but we can't upload them all to github as some are very large so we just uploaded one in the Phase3 folder. Our training and validation data came from a datasets on hugging face, https://huggingface.co/datasets/qwedsacf/grade-school-math-instructions and https://huggingface.co/datasets/HarshaniMadhushaniBandara/llma_math_word_problem. To train our models we ran the notebooks in the Phase 2 and 3 Training folders. You can check out some of our results in the Results folder in Phase 2/3.