2Show HN: Stella Nera – Maddness Hardware Accelerator (opens in new tab)(github.com)154joennlae2y ago26
3Show HN: Maddness – Energy-efficient DNNs without multiplying (opens in new tab)(arxiv.org)2joennlae2y ago0
4Show HN: less than 650 LOC trainable GPT only using NumPy (opens in new tab)(github.com)90joennlae2y ago18
5Show HN: Tensorli – NumPy Only Transformer Training (<650 lines) (opens in new tab)(github.com)2joennlae2y ago0
6GithubNext: AI-Powered Auto-Fix Pull Requests for Issues (opens in new tab)(githubnext.com)3joennlae2y ago0
7Show HN: Tensorli - GPT-Like Transformers in Pure NumPy (opens in new tab)(github.com)3joennlae2y ago0
8Nvidia uses internal LLM to support chip design efforts (opens in new tab)(blogs.nvidia.com)2joennlae2y ago0
9“THIS-IS-NOT-A-BACKUP” forking popular FOSS repos with a cron job (opens in new tab)(github.com)1joennlae3y ago0
