Yunseong Lee

I was considering to reuse most code in Evaluator-side, but I've found that this will result in many ad-hoc workarounds, especially 1) for obtaining the number of table blocks, which...

@bgchun commented that we need to introduce [hysteresis](https://en.wikipedia.org/wiki/Hysteresis), and avoid applying optimization when the estimated cost is not that beneficial. For this, we need to track the history of the...

+1 to take them out. I've observed that the log entries are not easy to parse. Any other thoughts?

In today's meeting, we discussed that this feature - store/load pre-trained models - is not restricted to DNN; Instead, we can expose a general checkpointing mechanism for fault tolerance, and...

@jsjason, @johnyangk, @beomyeol , @gyeongin and I had a discussion and it'd be good to check followings: - How scalable are the systems? In @gyeongin's experiment, he didn't use the...

@gyeongin Thanks for sharing the result. This looks awesome! Could you kindly give us a very short summary what changes you made for better performance and memory usage?