Gerek's ByteTrail Oydssey

This week my group refined the multi-language functionality across all tiddlers. We did this by verifying each page's content and tags correctly switch between English, Simplified Chinese, and Traditional Chinese. We spent time testing the summarizer service and ran multiple test cases to verify consistency across all three languages. We also implemented unit tests. As a team, we divided the testing responsibilities across the four main services (crawler, extractor, summarizer, and publisher). My role was to focus on the summarizer tests. I made sure that functions like sentence cleanup, Chinese note removal, and the core summary-generation logic behaved predictably. I also verified that each test correctly handled edge cases such as short content, markup-heavy inputs, and different language outputs. Next week I plan to contribute to the upcoming testing phase by gathering feedback from peers and verifying that the summaries are consistent across each language. I also want to stre...

This week our group focused on improving the UI and overall presentation of the project. I contributed UI feedback as we refined the homepage layout and made decisions about colors, backgrounds, and the general look-and-feel. We also deployed the site to GitHub Pages and tested the live link to make sure everything functioned correctly. For the most part, the deployment worked smoothly, but we discovered a handful of tiddlers whose summaries were either missing or cut off.  Next week our plan is to officially begin the testing phase. We’ll primarily do this by reaching out to peers and gathering input so we can evaluate the reliability of the summaries. We may also start implementing unit tests for some of the core services and potentially work on system-level tests as well. On my end, I’ll continue debugging the issues we found with incomplete or empty summaries and help verify that all three languages remain consistent. I also want to support the deployment process by testing the...

This week, our team focused on improving the UI and UX of the TiddlyWiki page to make it more visually appealing and intuitive. We also expanded the language functionality by adding support for both Traditional and Simplified Chinese , which helps make our project more inclusive and accessible to a wider audience, especially to the intended audience. We also implemented an automated crawler that will run the crawling and summarizing process twice per month. Though it's finished, it will officially be implemented during the time of deployment. Next week, we plan to continue enhancing the UI and UX , as our advisor emphasized that strong visual design is a key aspect of the project’s success.  The challenges we face at the moment are deciding how many additional features to include. There are many directions we could expand in. It seems there are many ways we can expand the crawler and summarizer with better filters and higher accuracy. However, we want to make sure we stay focu...

This week marks the end of week 4 of development on our capstone project. My focus was getting the LLM connection working for a wiki summarizer service. I tested the setup both locally using LM Studio and remotely through our client’s hosted LLM endpoint. I learned a lot about how LM Studio manages local LLM instances and how containerized services communicate with them. I had to think about system resources more carefully, since running large models locally requires significant GPU memory and CPU allocation to avoid slow responses or failures during summarization. I also learned how requests and tokens work in LLM-based APIs. Each token represents a piece of text that contributes to both input and output length, affecting processing time and model cost. Understanding these limits helped me troubleshoot the incomplete responses we were receiving and adjust prompt sizes for better summarizations. To handle longer documents, I implemented a chunking mechanism that splits text into smalle...