δ-mem: Cutting LLM Memory Costs by 90%
New δ-mem framework slashes GPU memory usage for LLMs by 90%, enabling efficient online inference on consumer hardware.
🏷️ Memory Optimization
4 articles about 'Memory Optimization'
MIT Sparse Attention Cuts Transformer Memory 80%
MIT researchers introduce a sparse attention mechanism that slashes Transformer memory usage by 80% while preserving mod…
UC Berkeley Cuts Transformer Memory Usage With New Attention
UC Berkeley researchers unveil a novel attention mechanism that dramatically reduces memory consumption in Transformer m…
SNU Researchers Crack Efficient AI Training on Low-End GPUs
Seoul National University team develops novel memory optimization techniques enabling large AI model training on consume…
1