Article
From GRPO to DAPO and GSPO: What, Why, and How
NormalUhr
• • 143
Ngo Dinh Luan
luanngo
AI & ML interests
NLP & DS
Recent Activity
liked a model 5 days ago
facebook/omniASR-W2V-300M liked a model 5 days ago
csukuangfj2/sherpa-onnx-zipformer-vi-30M-2026-02-09 liked a model 5 days ago
csukuangfj2/sherpa-onnx-zipformer-vi-30M-int8-2026-02-09Organizations
upvoted an article 11 months ago
upvoted 4
articles about 1 year ago
Article
Mastering Tensor Dimensions in Transformers
not-lain
• • 201 Article
KV Caching Explained: Optimizing Transformer Inference Efficiency
not-lain
• • 409 Article
KV Cache from scratch in nanoVLM
- +3
ariG23498, kashif, lusxvr, andito, pcuenq
• • 120 Article
Introducing Training Cluster as a Service - a new collaboration with NVIDIA
- +1
jeffboudier, ark393, pagezyhf
• • 27