3Better than DeepSeek R1? MiniMax-M1:open-weight hybrid-attention reasoning model (opens in new tab)(huggingface.co)6helloericsf1y ago0Save
5DeepSeek Open Source Optimized Parallelism Strategies, 3 repos (opens in new tab)(github.com)GitHub103helloericsf1y ago8Save
6DeepSeek Open Source DeepGEMM – FP8 GEMM Library(300 lines for 1350+ FP8 TFLOPS) (opens in new tab)(twitter.com)4helloericsf1y ago1Save
7Alibaba Open Source Large-Scale Video Generative Models: Wan2.1 (opens in new tab)(twitter.com)8helloericsf1y ago2Save
8DeepSeek open source DeepEP – library for MoE training and Inference (opens in new tab)(github.com)GitHub536helloericsf1y ago71Save
9DeepSeek Open Source FlashMLA – MLA Decoding Kernel for Hopper GPUs (opens in new tab)(github.com)GitHub441helloericsf1y ago108Save
10New Qwen2.5-Max Outperforms DeepSeek V3 in Benchmarks (opens in new tab)(twitter.com)3helloericsf1y ago2Save
11Longest context up to 4M, MiniMax-01 hybrid 456B Open source model (opens in new tab)(github.com)GitHub19helloericsf1y ago1Save
12DeepSeek v3 beats Claude sonnet 3.5 and way cheaper (opens in new tab)(huggingface.co)48helloericsf1y ago9Save
13NeurIPS and Dr. Picard released statement for singling out Chinese scholars (opens in new tab)(twitter.com)2helloericsf1y ago2Save
15Chinese AI Community: open-source Heatmap (opens in new tab)(huggingface.co)1helloericsf1y ago1Save
