Similar Items: TurboGR: An Accelerated Training System for Large-Scale Generative Recommendation
- RcLLM: Accelerating Generative Recommendation via Beyond-Prefix KV Caching
- CCL-D: A High-Precision Diagnostic System for Slow and Hang Anomalies in Large-Scale Model Training
- Accelerating Compound LLM Training Workloads with Maestro
- Piper: Efficient Large-Scale MoE Training via Resource Modeling and Pipelined Hybrid Parallelism
- A Scalable Recipe on SuperMUC-NG Phase 2: Efficient Large-Scale Training of Language Models
- ZipCCL: Efficient Lossless Data Compression of Communication Collectives for Accelerating LLM Training