Tianyi Chen, Bo Ji, Zhihui Zhu, Tianyu Ding, Guanyi Wang, Luming Liang, Yixin Shi, Sheng Yi, Xiao Tu
We have not lifted any functions out of this paper's repositories yet, so there is nothing we have run. The repositories linked to it are listed below.
Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.
Structured pruning is a commonly used technique in deploying deep neural networks (DNNs) onto resource-constrained devices. However, the existing pruning methods are usually heuristic, task-specified, and require an extra fine-tuning procedure. To overcome these limitations, we propose a framework that compresses DNNs into slimmer architectures with competitive performances and significant FLOPs reductions by Only-Train-Once (OTO). OTO contains two keys: (i) we partition the parameters of DNNs into zero-invariant groups, enabling us to prune zero groups without affecting the output; and (ii) to promote zero groups, we then formulate a structured-sparsity optimization problem and propose a novel optimization algorithm, Half-Space Stochastic Projected Gradient (HSPG), to solve it, which outperforms the standard proximal methods on group sparsity exploration and maintains comparable convergence. To demonstrate the effectiveness of OTO, we train and compress full models simultaneously from scratch without finetuning for inference speedup and parameter reduction, and achieve state-of-the-art results on VGG16 for CIFAR10, ResNet50 for CIFAR10 and Bert for SQuAD and competitive result on ResNet50 for ImageNet. The source code is available at https://github.com/tianyic/only_train_once.
The same record, over MCP at https://syntology.ai/mcp:
get_harvested_code_for_paper("2107.07467")
get_code_for_paper("2107.07467")
have("2107.07467")
The run record, dated, one paper per request, free:
curl https://syntology.ai/api/ran/2107.07467.json
A badge for a README (the split and the date, never a ratio):
[](https://syntology.ai/paper/2107.07467)
Connect an agent — have() is free.