Ml on mc · noteshttps://hk.crepuscule.uk/tags/ml/Recent content in Ml on mc · notesHugoen-usTue, 19 May 2026 17:40:00 +0800Data-parallel training: gradient bucketing and overlaphttps://hk.crepuscule.uk/posts/grad-bucketing/Tue, 19 May 2026 17:40:00 +0800https://hk.crepuscule.uk/posts/grad-bucketing/Why DDP feels like magic until you look at the allreduce schedule.A minimal, reproducible Docker setup for ML experimentshttps://hk.crepuscule.uk/posts/docker-ml/Wed, 11 Mar 2026 14:00:00 +0800https://hk.crepuscule.uk/posts/docker-ml/Pin everything, mount data, never bake it. The boring setup that stopped me losing runs.KV cache, and why LLM inference is memory-boundhttps://hk.crepuscule.uk/posts/kv-cache/Sun, 08 Feb 2026 15:20:00 +0800https://hk.crepuscule.uk/posts/kv-cache/The cache that makes autoregressive decoding fast also makes it the thing that runs out of memory first.