3 citations · 3 across the 1 of their papers we have counts for
3 papers
TorchAO: PyTorch-Native Training-to-Serving Model Optimization
Andrew Or, Apurva Jain, Daniel Vega-Myhre +10
We present TorchAO, a PyTorch-native model optimization framework leveraging quantization and sparsity to provide an end-to-end, training-to-serving workflow for AI models. TorchAO…
NeurIPS 2023 LLM Efficiency Fine-tuning Competition
Mark Saroufim, Yotam Perlitz, Leshem Choshen +11
Our analysis of the NeurIPS 2023 large language model (LLM) fine-tuning competition revealed the following trend: top-performing models exhibit significant overfitting on benchmark…
Flex Attention: A Programming Model for Generating Optimized Attention Kernels
Juechu Dong, Boyuan Feng, Driss Guessous +2
Over the past 7 years, attention has become one of the most important primitives in deep learning. The primary approach to optimize attention is FlashAttention, which fuses the ope…