poet
poet copied to clipboard
ShishirPatil
ML model training for edge devices
Dear Authors, I like this work very much. May I know how to apply it on YOLOv5 (https://github.com/ultralytics/yolov5)?
Dear authors, can you kindly share the code behind Table 3, the comparison to POFO? I'd like to know if the ResNet-18 used was converted from POET's `dfgraph` to POFO's...
This PR is the first step towards translating models defined in Torch to Tensorflow transparently. We define network transformation functions which are then composed together for converting PyTorch and Tensorflow...
Implementing rematerialization for PyTorch models. With remat-only we can train ResNet18 on CIFAR-10 with ~93.5% accuracy in 150 epochs.
Along with cropconcat layer definition to track dependencies.
POET produces the energy-optimal solution given a memory and timing budget. It might be interesting to see the Energy-Memory tradeoff space. The pareto curve visualizes this by plotting the energy...
Dear Authors, I am amazed that the performance of POET outperforms POFO so much. Since PyTorch does not have a traced backward graph, I am curious about which framework you...
