SHRAJAL PANDEY@shrajal2428it142068326

Video Captioning Agent is a Dockerized pipeline built for Track 2 of the AMD Developer Hackathon: ACT II. Given a video URL as input, it downloads the clip, uses ffprobe to measure its duration, and samples frames at regular intervals (roughly one frame every five seconds, clamped between 8 and 20 frames) using ffmpeg. These frames are sent to a Fireworks AI vision-language model (Kimi K2.6) to generate a factual, grounded description of what is happening in the video. That description is then passed to a second Fireworks AI call, which rewrites it into four distinct caption styles in a single structured JSON response: formal (neutral, professional tone), sarcastic (dry, witty), humorous-tech (jokes using programming/tech metaphors), and humorous-non-tech (relatable, everyday humor). This two-step design keeps every generated caption grounded in the actual visual content of the video while still matching the requested tone, reducing hallucination compared to single-shot style generation. The application reads tasks from /input/tasks.json and writes results to /output/results.json, matching the Track 2 evaluation harness. It is packaged as a linux/amd64 Docker image with ffmpeg and all Python dependencies pre-installed, so it runs consistently in any evaluation environment without additional setup. Built entirely on AMD Developer Cloud and Fireworks AI infrastructure provided for the hackathon.

🌎 Online Hackathon | 📅 August 3–10, 2026| 💻 Powered by native.builder The official native.builder hackathon, hosted by NativelyAI and lablab.ai. Join founders, developers, designers, and builders to create real AI-native applications powered by native.builder: from idea to deployed product, fully in the cloud. 💻 This hackathon is fully online. No travel required.

To participate in the hackathon, simply click Sign up with AMD. If you are not already a member of the AMD AI Developer Program (ADP), you will need to create an account. ADP members can access AI tutorials, experts, and community support. In addition, new members can claim a free month of DeepLearning.ai pro, $100 in AMD GPU cloud credits, and $50 in Fireworks AI API credits. All hackathon participants will receive on hackathon launch additional compute and API credits for their projects. Ready to build what's next on AMD? Sign up with AMD below to secure your spot.