Learning to Shard: RL for Co-optimizing the Parallelism Degrees and Per-operator Sharding Dimensions in Distributed LLM Inference
Anzeige
Ähnliche Artikel
arXiv – cs.LG
•
Tool Zero: Training Tool-Augmented LLMs via Pure RL from Scratch
arXiv – cs.LG
•
Opportunistic Expert Activation: Batch-Aware Expert Routing for Faster Decode Without Retraining
arXiv – cs.AI
•
DeepAgent: A General Reasoning Agent with Scalable Toolsets
arXiv – cs.AI
•
Local Coherence or Global Validity? Investigating RLVR Traces in Math Domains
VentureBeat – AI
•
New 'Markovian Thinking' technique unlocks a path to million-token AI reasoning
arXiv – cs.AI
•
JudgeSQL: Reasoning over SQL Candidates with Weighted Consensus Tournament