@sshivansh271273122

This agent implements a token-efficient routing cascade for the AMD ACT II Track-1 task set (factual knowledge, mathematical reasoning, sentiment classification, text summarization, and named-entity recognition). For each task it produces a correct, complete, format-compliant answer, following any constraints stated in the prompt (for example an exact number of sentences or bullet points, or a required label plus a one-sentence justification). The container reads tasks from /input/tasks.json and writes results to /output/results.json as a list of {task_id, answer} objects, returning one result per task with task IDs preserved exactly. The design prioritizes correctness first and token efficiency second, and is built to generalize to the refreshed hidden evaluation set rather than overfit the public examples. It runs as a lightweight, self-contained linux/amd64 container under 10GB, using Fireworks-hosted models via the runtime-provided endpoint and no other external services.

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.