repoGitHubTrust 82 · PrimaryPublished 1mo agoLive · 17h ago
deepspeedai/DeepSpeed
DeepSpeed is a deep learning optimization library that makes distributed training and inference easy, efficient, and effective.
Lineage graph
Paper → model → repo connections mined from source citations (Tier-1 exact match).
Why these links exist
Every edge carries a method, confidence, and the source snippet that justified it — so bad links are debuggable.
- FuzzyOverlapping authors or contributors · 62%Pretraining Data Can Be Poisoned through Computational Propaganda →
“Shared author/contributor keys: smith”
- FuzzyOverlapping authors or contributors · 62%In-Place Tokenizer Expansion for Pre-trained LLMs →
“Shared author/contributor keys: smith”
- FuzzyOverlapping authors or contributors · 62%DSPrompt: Dynamic Soft Prompt Defense Against M-RAG Corruption →
“Shared author/contributor keys: lai”
- FuzzySimilar title/name (fuzzy) · 59%Calibration-Free Vehicle Speed Estimation: A Monocular Keypoint-Template Approach →
“Fuzzy title match (0.73): “Calibration-Free Vehicle Speed Estimation: A Monocular Keypo” ≈ “deepspeedai/DeepSpeed””
- FuzzyOverlapping authors or contributors · 62%Anchoring Instruction Outside Mask: Exact Reference Caching for Efficient In-Context Diffusion Transformers →
“Shared author/contributor keys: lai”
- PossiblePossibly related (embedding) · 49%DeepSeek open-sources inference optimizations with 60–85% faster generation [pdf] →
- PossiblePossibly related (embedding) · 48%OpenAI and Broadcom unveil LLM-optimized inference chip →
- PossiblePossibly related (embedding) · 47%HASTE: A Framework for Training-Free, Dynamic, and Steerable Compression of Pre-Trained Convolutional Neural Networks →
Implements
paperPretraining Data Can Be Poisoned through Computational PropagandapaperIn-Place Tokenizer Expansion for Pre-trained LLMspaperDSPrompt: Dynamic Soft Prompt Defense Against M-RAG CorruptionpaperCalibration-Free Vehicle Speed Estimation: A Monocular Keypoint-Template ApproachpaperAnchoring Instruction Outside Mask: Exact Reference Caching for Efficient In-Context Diffusion TransformerspaperHASTE: A Framework for Training-Free, Dynamic, and Steerable Compression of Pre-Trained Convolutional Neural NetworkspaperBeyond Adam: SOAP and Muon for Faster, Label-Efficient Training of Machine Learning Interatomic PotentialspaperDifferentiable Logic Gate Networks for Low-Latency EEG Classification on Edge DevicespaperEdgeBench: Unveiling Scaling Laws of Learning from Real-World EnvironmentspaperABot-World-0: Infinite Interactive World Rollout on a Single Desktop GPUpaperHeadCast: Casting Attention Heads for Efficient Autoregressive Video GenerationpaperEfficient RLVR Scheduling via Graph-Structured Online Difficulty EstimationpaperAutonomous Agricultural Tractor: Integrated Weed Detection and LiDAR Navigation for Precision Paddy Farming
Covers
contributed_to (incoming)
personjeffrapersonloadamspersontjruwasepersonstas00personmrwyattiipersontohtanapersonRezaYazdaniAminabadipersondelockpersonawan-10personconglonglipersonnelyahupersonlekurilepersonjomayeripersoninkcherrypersonsamyampersonYejing-Laipersoncmikeh2personcli99persondeepcharmpersonHeyangQinpersonrraminenpersonmolly-smithpersonsfc-gh-truwasepersonsamadejacobspersondigger-yu
Implements (incoming)
Related across the graph
persontjruwasepersonawan-10personRezaYazdaniAminabadipersondelockpersondigger-yupaperAutonomous Agricultural Tractor: Integrated Weed Detection and LiDAR Navigation for Precision Paddy FarmingpersonrraminennewsOpenAI and Broadcom unveil LLM-optimized inference chippersonlekurilepersonsfc-gh-truwasepersonHeyangQinpaperDifferentiable Logic Gate Networks for Low-Latency EEG Classification on Edge DevicespersonYejing-LaipersonmrwyattiipaperBeyond Adam: SOAP and Muon for Faster, Label-Efficient Training of Machine Learning Interatomic Potentialspersoncli99paperPretraining Data Can Be Poisoned through Computational PropagandapaperIn-Place Tokenizer Expansion for Pre-trained LLMspaperAnchoring Instruction Outside Mask: Exact Reference Caching for Efficient In-Context Diffusion TransformerspaperEdgeBench: Unveiling Scaling Laws of Learning from Real-World EnvironmentspaperABot-World-0: Infinite Interactive World Rollout on a Single Desktop GPUpaperCalibration-Free Vehicle Speed Estimation: A Monocular Keypoint-Template ApproachpaperHASTE: A Framework for Training-Free, Dynamic, and Steerable Compression of Pre-Trained Convolutional Neural NetworkspersondeepcharmpersonnelyahupersonsamadejacobspersonjomayeripaperDSPrompt: Dynamic Soft Prompt Defense Against M-RAG CorruptionpersonloadamsnewsWe’re launching the Google DeepMind Accelerator program in Asia Pacific to tackle environmental riskspersoncmikeh2personinkcherrypersonstas00persontohtanapersonjeffranewsDeepSeek open-sources inference optimizations with 60–85% faster generation [pdf]paperSystematic Evaluation of Learning Rate Scheduling Strategies Across Heterogeneous ArchitecturespersonsamyampaperHeadCast: Casting Attention Heads for Efficient Autoregressive Video GenerationpaperEfficient RLVR Scheduling via Graph-Structured Online Difficulty Estimationpersonmolly-smithpersonconglongli