Davide Testugginehttps://testuggine.ninja/Notes on post-training, reinforcement learning, and agents.en-usTue, 26 Aug 2025 00:00:00 +0000A primer on LLM post-traininghttps://pytorch.org/blog/a-primer-on-llm-post-training/https://pytorch.org/blog/a-primer-on-llm-post-training/Tue, 26 Aug 2025 00:00:00 +0000Published on PyTorch blog.Differential privacy, part 3: per-sample gradients for more layershttps://pytorch.medium.com/differential-privacy-series-part-3-efficient-per-sample-gradient-computation-for-more-layers-in-39bd25df237https://pytorch.medium.com/differential-privacy-series-part-3-efficient-per-sample-gradient-computation-for-more-layers-in-39bd25df237Fri, 04 Nov 2022 00:00:00 +0000Published on PyTorch on Medium.Differential privacy, part 2: efficient per-sample gradients in Opacushttps://medium.com/pytorch/differential-privacy-series-part-2-efficient-per-sample-gradient-computation-in-opacus-5bf4031d9e22https://medium.com/pytorch/differential-privacy-series-part-2-efficient-per-sample-gradient-computation-in-opacus-5bf4031d9e22Fri, 17 Sep 2021 00:00:00 +0000Published on PyTorch on Medium.Differential privacy, part 1: DP-SGD explainedhttps://medium.com/pytorch/differential-privacy-series-part-1-dp-sgd-algorithm-explained-12512c3959a3https://medium.com/pytorch/differential-privacy-series-part-1-dp-sgd-algorithm-explained-12512c3959a3Mon, 31 Aug 2020 00:00:00 +0000Published on PyTorch on Medium.