@emaanali08lix49501

This project analyzes short video clips (30 seconds to 2 minutes) and generates captions in four required styles: formal (objective and factual), sarcastic (dry and ironic while staying accurate), humorous_tech (tech/programming humor grounded in real scene details), and humorous_non_tech (everyday humor with no jargon). The pipeline reads a task list from /input/tasks.json, processes each clip through a vision-capable Fireworks AI model to extract a factual scene description, then generates all four caption styles from that single grounded description — so every style describes the same real content, just in a different voice. Clips are processed in parallel (bounded by a concurrency limit), and results are written to /output/results.json in the required schema. The container is fully self-contained: it runs with no local files or manual setup beyond mounting input/output volumes and providing a Fireworks API key at runtime, completes a 13-clip local test set in under 2 minutes, and produces valid JSON on every run.

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.