Summary of On the Cost Of Model-serving Frameworks: An Experimental Evaluation, by Pasquale De Rosa et al.
On the Cost of Model-Serving Frameworks: An Experimental Evaluation
by Pasquale De Rosa, Yérom-David Bromberg, Pascal Felber, Djob Mvondo, Valerio Schiavoni
First submitted to arxiv on: 15 Nov 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: None
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The paper investigates the performance of five widely-used model serving frameworks in various scenarios, focusing on their ability to efficiently deploy and manage pre-trained models. The evaluated frameworks include TensorFlow Serving, TorchServe, MLServer, MLflow, and BentoML, which are applied under four different tasks: malware detection, cryptocoin prices forecasting, image classification, and sentiment analysis. The results show that TensorFlow Serving outperforms the other frameworks in serving deep learning models, while DL-specific frameworks exhibit lower latencies compared to general-purpose ML frameworks. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper looks at how well five model serving frameworks work together with pre-trained models. It tests these frameworks under different tasks like detecting malware or recognizing images. The results show that one framework called TensorFlow Serving is the best for using deep learning models, while some other frameworks are better suited for general machine learning tasks. |
Keywords
» Artificial intelligence » Deep learning » Image classification » Machine learning
