Biweekly Report for the Bot Project
Subject: Milestone II - Model Training, Pipeline Refinement, and Final Project Adjustments
Project Title: BotWheels
Report Number: 5
Date: December 9, 2024
Week Range: Weeks 9-11
1. Progress Overview
Due to the approaching deadline for thesis submissions, overall team engagement in the project has decreased, which is understandable. Currently, the focus remains on training the models and refining the main project pipeline. The expected deliverables include fully trained models, detailed documentation for the training process, and well-commented code to ensure clarity and maintainability.
The training of both models is progressing, but delays are anticipated due to the reduced availability of some team members. Despite these challenges, efforts are ongoing to meet the key milestones, and adjustments are being made to optimize the model's performance.
2. Team Member Contributions, Marks, and Linked Issues
| Team Member | Mark | Summary of Contributions | Linked Issues |
|---|---|---|---|
| Jakub Cisoń | 4 | Continued managing communication within the team, although contributions have been lower over the past three weeks due to other commitments. | |
| Kacper Drozdowski | 4 | More active compared to previous weeks, currently training the bot using the A2C algorithm. | - A2C Initial Training & Testing via Local |
| Mateusz Gościniecki | 3 | Reduced communication in the past few weeks. Currently training the PPO model. | - Remote Training Setup for Rocket League Bot |
| Anna Ivanytska | 3 | Very poor communication, significant absence from meetings. | - Expansion of Event Logger - Add/Update Documentation According to Standardization Guidelines |
| Igor Malkovskiy | 3 | Reduced communication in the past few weeks. | - PPO Model Training and Testing |
| Dawid Mielewczyk | 4 | Waiting for descriptions from other team members to complete the CLI. Available for code reviews and other support tasks. | - Develop CLI for Training and Gameplay Modes |
| Camille Nadir | 3 | Very poor communication, significant absence from meetings, currently working on training the bot using PPO. | - PPO Model Training and Testing |
| Michał Pryba | 5 | Significant contribution to the project, currently providing support. | - Technical project support |
| Wojciech Szamocki | 4.5 | Reduced activity relative to previous weeks, but communicated about travel. Continues to deliver quality code, currently working on model management. | - Enhance Model Management and Training Continuity |
| Konrad Siemiątkowski | 3.5 | Training the A2C model, limited communication. | - Initial A2C Model Training and Testing (via Remote Connection) |
| Michał Zarzycki | 3.5 | Reduced communication. Currently expected to add a "how-to" guide for his code. | - Add (How-to) .md for Running Bot Training with PPO |
3. Challenges and Problems Encountered
The improper distribution of tasks during the initial weeks has impacted the latter stages, which coincide with thesis submission deadlines. As a result, there has been less time dedicated to the project, with fewer people actively engaged. Communication issues remain the primary challenge, which has led to delays and inefficiencies in task completion.
4. Plans for the Next Period
As established, the final submission of the project will occur during the wrap-up session. By that time, fully trained models are expected to be delivered, along with completed tests and any necessary adjustments. Additionally, the project documentation will be finalized to provide a clear overview of the model training process and final results.
5. Summary
Despite challenges related to reduced availability and communication issues, progress is being made toward the completion of the project. The team is focused on finalizing the model training and refining the project pipeline to meet the established goals. The upcoming period will be critical for delivering the final results, ensuring the models are trained as expected, and preparing the project for final presentation.