BenchmarkPostTrainBench
PostTrainBench
A benchmark targeting machine learning engineering workflows, such as data cleaning, fine-tuning configuration, and hyperparameter tuning, examining whether a model can handle the daily tasks of an ML engineer.
Timeline (last 90 days)
Aug 26
On the PostTrainBench benchmark, GLM-5.2 surpasses Opus 4.7 and GPT-5.5, ranking second only to Opus 4.8.
Expired50%
Aug 26
PostTrainBench 为每个智能体分配一块 H100 GPU，评估它能通过后训练将小模型提升多少
Unverified50%