Kukedlc/

slopus-v3-train-lt2k

17121 ejemplos pre-formateados con template Qwen3-4B-Thinking-2507, filtrados a <2048 tokens y deduplicados.

roman (7515) - Roman1111111/claude-opus-4.6-10000x (reasoning field) jirafa (6644) - angrygiraffe/claude-opus-4.6-4.7-reasoning-8.7k (think inline) nohurry (2233) - nohurry/Opus-4.6-Reasoning-3000x-filtered teich (639) - TeichAI/Claude-Opus-4.6-Reasoning-887x (thinking field) gpt55 (90) - armand0e/gpt-5.5-chat

Each… See the full description on the dataset page: https://huggingface.co/datasets/Kukedlc/slopus-v3-train-lt2k.