Submissions from github.com/deepseek-ai

		DeepSeek-V3/R1 Inference System Overview (github.com/deepseek-ai)
		27 points by meetpateltech 13 days ago \| past \| 6 comments
		Smallpond: A lightweight data processing framework built on DuckDB and 3FS (github.com/deepseek-ai)
		3 points by kermatt 14 days ago \| past \| 1 comment
		Smallpond – A lightweight data processing framework built on DuckDB and 3FS (github.com/deepseek-ai)
		322 points by overflowcat 14 days ago \| past \| 72 comments
		Fire-Flyer File System (3FS) (github.com/deepseek-ai)
		361 points by wenyuanyu 14 days ago \| past \| 101 comments
		Expert Parallelism Load Balancer (github.com/deepseek-ai)
		1 point by pongogogo 15 days ago \| past
		EPLB: Expert Parallelism Load Balancer (github.com/deepseek-ai)
		3 points by mfiguiere 15 days ago \| past
		DualPipe: Bidirectional pipeline parallelism algorithm (github.com/deepseek-ai)
		180 points by mfiguiere 15 days ago \| past \| 20 comments
		DeepSeek Open Source Optimized Parallelism Strategies, 3 repos (github.com/deepseek-ai)
		103 points by helloericsf 15 days ago \| past \| 8 comments
		DeepGEMM: clean and efficient FP8 GEMM kernels with fine-grained scaling (github.com/deepseek-ai)
		391 points by mfiguiere 16 days ago \| past \| 67 comments
		Awesome DeepSeek Integrations (github.com/deepseek-ai)
		116 points by selvan 17 days ago \| past \| 24 comments
		DeepSeek open source DeepEP – library for MoE training and Inference (github.com/deepseek-ai)
		536 points by helloericsf 17 days ago \| past \| 71 comments
		DeepSeek Open Source FlashMLA – MLA Decoding Kernel for Hopper GPUs (github.com/deepseek-ai)
		441 points by helloericsf 18 days ago \| past \| 108 comments
		DeepSeek Open Infra: Open-Sourcing 5 AI Repos in 5 Days (github.com/deepseek-ai)
		757 points by ahsmha_ 21 days ago \| past \| 236 comments
		DeepSeek drops recommended R1 deployment settings (github.com/deepseek-ai)
		2 points by wavelander 28 days ago \| past
		Suggestion: Read the DeepSeek model license - note the Governing Jurisdiction (github.com/deepseek-ai)
		6 points by tmo9d 42 days ago \| past \| 6 comments
		DeepSeek releases Janus Pro, a text-to-image generator [pdf] (github.com/deepseek-ai)
		912 points by reissbaker 46 days ago \| past \| 674 comments
		Deepseek open models seem to follow CCP party-line (github.com/deepseek-ai)
		2 points by yehosef 46 days ago \| past
		Deepseek R1 paper – at 2% of the price of OpenAI o1 [pdf] (github.com/deepseek-ai)
		5 points by vishnurnair 50 days ago \| past
		DeepSeek LLM supports Chinese propaganda (github.com/deepseek-ai)
		6 points by orenlindsey 51 days ago \| past \| 1 comment
		Deepseek R1 Zero learns to reason using reinforcement learning on base model [pdf] (github.com/deepseek-ai)
		6 points by virde 53 days ago \| past
		DeepSeek-R1 (github.com/deepseek-ai)
		1843 points by meetpateltech 53 days ago \| past \| 663 comments
		DeepSeek-VL2: MoE Vision-Language Models for Advanced Multimodal Understanding (github.com/deepseek-ai)
		36 points by selvan 72 days ago \| past \| 7 comments
		DeepSeek-v3 Technical Report [pdf] (github.com/deepseek-ai)
		26 points by pongogogo 78 days ago \| past \| 2 comments
		DeepSeek-V3 (github.com/deepseek-ai)
		125 points by mgt19937 78 days ago \| past \| 40 comments
		DeepSeek VL2 (github.com/deepseek-ai)
		1 point by arunbahl 88 days ago \| past
		DeepSeek-VL2: Moe Vision-Language Models for Advanced Multimodal Understanding [pdf] (github.com/deepseek-ai)
		1 point by limoce 3 months ago \| past
		Janus: Decoupling visual encoding for multimodal understanding and generation (github.com/deepseek-ai)
		36 points by jinqueeny 4 months ago \| past \| 3 comments
		DeepSeek Coder V2 Released (github.com/deepseek-ai)
		1 point by thethirdone 8 months ago \| past
		DeepSeek-Coder-V2: Breaking the Barrier of Closed-Source Models in Code (github.com/deepseek-ai)
		9 points by mgt19937 9 months ago \| past \| 1 comment
		DeepSeek-V2: A Strong, Economical, and Efficient Moe Language Model (github.com/deepseek-ai)
		14 points by jasondavies 10 months ago \| past \| 3 comments
		More