Agentica项目的开源DeepCoder模型在编码基准测试中超越了OpenAI的O1
The Agentica Project and Together AI have released DeepCoder-14B-Preview, an open source AI coding model based on Deepseek-R1-Distilled-Qwen-14B. The model achieves a 60.6% pass rate on...
Agentica项目与Together AI发布了DeepCoder-14B-Preview，这是一个基于Deepseek-R1-Distilled-Qwen-14B的开源AI编码模型。该模型在LiveCodeBench上的通过率为60.6%，超过了OpenAI的o1模型。开发团队通过强化学习优化了模型，并共享了数据集、代码和训练日志，旨在推动RL训练的民主化。DeepCoder在多个编码基准测试中表现优异，显著缩短了训练时间。
