The latest articles on DEV Community by Aniket Manna (@aniket_manna_b1eca63b1faa). https://dev.to/aniket_manna_b1eca63b1faa https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F4204130%2Fe2613e6e-5a8b-42c4-b765-7594581275a2.jpeg https://dev.to/aniket_manna_b1eca63b1faa en Aniket Manna Sat, 24 Oct 2026 07:58:19 +0000 https://dev.to/aniket_manna_b1eca63b1faa/hf-chaallenges-2mok https://dev.to/aniket_manna_b1eca63b1faa/hf-chaallenges-2mok <p><em>This is a submission for the <a href="https://dev.to/challenges/hacktoberfest-week3-2026-10-19">Hacktoberfest Open-Source AI Challenge Week 3: Slow and Steady</a></em></p> <h2> What I Built </h2> <h2> Demo </h2> <h2> Code </h2> <h2> How I Built It </h2> <h2> Why Does Open Innovation Matter? </h2> <h2> My Agent Session </h2> <h2> The Lazy Payoff </h2> <h2> Prize Categories </h2> <p>You are a senior Python engineer helping me build a hackathon project in 1-2 days.</p> <h2> Project: LazyLecture </h2> <p>A local, offline tool that turns lecture files into study material automatically.<br> I drop a PDF (or audio file) into an "inbox" folder. A watcher detects it, a local<br> open-weight LLM processes it, and the output appears in an "output" folder:</p> <ol> <li>summary.md - a one-page summary with headings</li> <li>flashcards.md and flashcards.csv (Anki-importable: front,back)</li> <li>quiz.md - 5 multiple-choice questions with answers and one-line explanations</li> </ol> <h2> Hard constraints </h2> <ul> <li>Everything runs locally and offline. No paid or cloud APIs.</li> <li>LLM via Ollama (default model: qwen2.5:7b; model name must be a config setting so I can swap in llama3.1:8b or a 3B model and compare).</li> <li>Python 3.11+, libraries: watchdog, pypdf, requests (or the ollama python client), pyyaml. Optional: whisper.cpp for audio transcription.</li> <li>Must work on Windows, macOS and Linux (use pathlib, no shell-specific code).</li> <li>Keep it simple: no frameworks beyond what's listed, minimal files, readable code.</li> </ul> <h2> Architecture </h2> <ul> <li>config.yaml: model name, inbox/output paths, chunk size, number of flashcards/questions</li> <li>extract.py: PDF -&gt; clean text (handle empty or scanned pages gracefully)</li> <li>chunk.py: split text into chunks that fit the model's context window</li> <li>llm.py: wrapper around Ollama with retries and timeouts</li> <li>pipeline.py: chunk -&gt; summarize each chunk -&gt; merge into final summary -&gt; generate flashcards -&gt; generate quiz. Ask the model for JSON for flashcards/quiz, validate it, and retry once if the JSON is malformed.</li> <li>watcher.py: watchdog loop that processes new files, waits until the file has finished copying, and skips files already processed.</li> <li>main.py: CLI entry point (<code>python main.py watch</code> and <code>python main.py run &lt;file&gt;</code>)</li> <li>Optional stretch: app.py, a small Streamlit page to take the quiz and track wrong answers.</li> </ul> <h2> How to work </h2> <ol> <li>First, show me the plan and file tree in under 15 lines. Wait for my OK.</li> <li>Build in small steps: (a) extract + chunk, (b) LLM wrapper, (c) pipeline, (d) watcher, (e) outputs. After each step, tell me exactly how to run and test it.</li> <li>Write a few pytest tests for chunking and JSON validation, with the LLM mocked.</li> <li>Log clearly (which file, which stage, time taken per stage). I want timing data for my write-up.</li> <li>Add a benchmark script that runs the same PDF through two different models and prints time taken and output length, so I can compare speed vs. quality.</li> <li>Finish with a README: what it does, setup steps (install Ollama, pull model, pip install), usage, a "why open-source/local matters" section, and troubleshooting.</li> </ol> <h2> Quality bar </h2> <ul> <li>A new user can go from clone to working output in under 10 minutes.</li> <li>Errors never crash the watcher; they are logged and the file is moved to a "failed" folder.</li> <li>Explain any non-obvious decision in one sentence. Don't add features I didn't ask for.</li> </ul> <p>Start with step 1 (the plan).</p> devchallenge hf26challenge Aniket Manna Sat, 24 Oct 2026 07:55:53 +0000 https://dev.to/aniket_manna_b1eca63b1faa/hfdone-3el https://dev.to/aniket_manna_b1eca63b1faa/hfdone-3el <p><em>This is a submission for the <a href="https://dev.to/challenges/hacktoberfest-week3-2026-10-19">Hacktoberfest Open-Source AI Challenge Week 3: Slow and Steady</a></em></p> <h2> What I Built </h2> <h2> Demo </h2> <h2> Code </h2> <h2> How I Built It </h2> <h2> Why Does Open Innovation Matter? </h2> <h2> My Agent Session </h2> <h2> The Lazy Payoff </h2> <h2> Prize Categories </h2> devchallenge hf26challenge Aniket Manna Sat, 24 Oct 2026 07:42:58 +0000 https://dev.to/aniket_manna_b1eca63b1faa/starts-2boc https://dev.to/aniket_manna_b1eca63b1faa/starts-2boc