AWQ quantization: done by stelterlab in INT4 with llm-compressor (https://github.com/vllm-project/llm-compressor — v0.10.0.1) from the vllm-project. See recipe.yaml for details. First tests done against vLLM v0.17.1. Not yet fully tested! We’re excited to introduce Nemotron-Cascade-2-30B-A3B, an open 30B MoE model with 3B activated parameters that delivers strong reasoning and agentic capabilities. It is post-trained from the Nemotron-3-Nano-30B-A3B-Base. Nemotron-Cascade-2-30B-A3B achieves gold medal performance in both the 2025 International Mathematical Olympiad (IMO) and the International Olympiad in Informatics (IOI). It operates in both thinking and instruct (non-reasoning) modes. Benchmark Metric: pass@1 Nemotron-3-Nano-30B-A3B Nemotron-3-Super-120B-A12B Qwen3.5-35B-A3B Nemotron-Cascade-2-30B-A3B Math IMO 2025 —🏅 35 pts IMO AnswerBench 70.4‡77.2‡74.8‡79.3 IMO ProofBench —72.9 AIME 2025 89.190.291.9‡92.4 (98.6)† AIME 2026 89.9‡89.8‡91.1‡90.9 (95.0)† HMMT Feb25 84.6‡93.789.094.6 Code Reasoning IOI 2025 —348.6‡🏅 439.3 ICPC World Finals 2025 —🏅 10/12 LiveCodeBench v6 (2408-2505) 68.378.774.687.2 (88.4)† LiveCodeBenchPro 25Q2 (Easy)…
Модальности:
Генерация текста
Области применения:
Логика и рассуждение Диалог / чат
Задача: Генерация текста
Автор: stelterlab
Теги: nvidia, nemotron-cascade-2, reasoning, general-purpose, SFT, RL, conversational, en
Лайков: 4 | Загрузок: 394
Описание основано на материалах HuggingFace. Перевод выполнен автоматически.
