#efficiency
2 posts · all tags
-
· The Lottery Ticket Hypothesis
In 2019, Frankle and Carlin published a paper https://arxiv.org/abs/1803.03635 that challenged a fundamental assumption in deep learning: that large, over-parameterized networks are necessary for achieving good performance. They proposed the Lottery Ticket Hypothesis, which states that dense, randomly initialized networks contain sparse subnetworks (winning tickets) that, when trained in isolation from the same initialization, can match the full network's accuracy.
-
· Quantization of LLMs
The escalating complexity and scale of large language models (LLMs) have introduced substantial challenges concerning computational demands and resource allocation. These models, often comprising hundreds of billions of parameters, necessitate extensive memory and processing capabilities, making their deployment and real-time inference both costly and impractical for widespread use.