Official repository for paper: O1-Pruner: Length-Harmonizing Fine-Tuning for O1-Like Reasoning Pruning.
O1-Pruner is a post-training technique that can accelerate the inference O1-like long thought reasoning models. Experiments show that the inference time overhead can be reduced by up to 50%. For more details, see our paper on arxiv: O1-Pruner
Recently, long-thought… See the full description on the dataset page:
https://huggingface.co/datasets/zx10086/O1-Pruner.
