rdolmedo_· @rdolmedo_ · X·· 7 小时前AI 评分
约1000美元训练算力打造的1.7B编码智能体在SWE-bench Verified达25.5% pass@1
作者宣布仅用约$1,000的训练算力，从零训练出一个1.7B编码智能体，在SWE-bench Verified上以25.5% pass@1超过Claude 3 Opus和GPT-4o。该成绩为开源SWE-bench速通项目nanoswe创下新纪录，作者以thread形式解释其可行性。
With only ~$1,000 of training compute, we trained 𝗳𝗿𝗼𝗺 𝘀𝗰𝗿𝗮𝘁𝗰𝗵 a 1.7B coding agent that beats Claude 3 Opus and GPT-4o on SWE-bench Verified 🤯
At 25.5% pass@1, it sets a new record for nanoswe, our open SWE-bench speedrun.
How is this even possible? 🧵👇
来源：rdolmedo_ · x.com