About the LabsStardog Labs is the R&D unit of Stardog Union. We innovate in AI fields like knowledge graph, LLM, automated inference, and related areas. Some examples include the following.Model PruningEigenpruningShortGPT: LLM Layers are more redundant than you expectQuantizationFP6-LLM QuantizationExtreme Compression of Large Language Models via Additive QuantizationModel DistillationDivide-or-Conquer? Which Part Should You Distill Your LLM?Efficiently distilling LLMs for Edge AppsModel ServingXc-Cache: Cross-Attending to Cached Context for Efficient LLM InferenceTCRA-LLM: Token Compression Retrieval Augmented Large Language Model for Inference Cost ReductionOptimizing Attention You need to pay better attentionDeFT: Flash Tree-attention with IO-Awareness for Efficient Tree-search-based LLM Inference
Stardog Labs is the R&D unit of Stardog Union. We innovate in AI fields like knowledge graph, LLM, automated inference, and related areas. Some examples include the following.Model PruningEigenpruningShortGPT: LLM Layers are more redundant than you expectQuantizationFP6-LLM QuantizationExtreme Compression of Large Language Models via Additive QuantizationModel DistillationDivide-or-Conquer? Which Part Should You Distill Your LLM?Efficiently distilling LLMs for Edge AppsModel ServingXc-Cache: Cross-Attending to Cached Context for Efficient LLM InferenceTCRA-LLM: Token Compression Retrieval Augmented Large Language Model for Inference Cost ReductionOptimizing Attention You need to pay better attentionDeFT: Flash Tree-attention with IO-Awareness for Efficient Tree-search-based LLM Inference