Papers › A Continual Development Methodology for Large-scale Multitask Dynamic ML Systems

A Continual Development Methodology for Large-scale Multitask Dynamic ML Systems

15 Sep 2022arXiv:2209.07326archive 2025-07-28

Andrea Gesmundo

The traditional Machine Learning (ML) methodology requires to fragment the development and experimental process into disconnected iterations whose feedback is used to guide design or tuning choices. This methodology has multiple efficiency and scalability disadvantages, such as leading to spend significant resources into the creation of multiple trial models that do not contribute to the final solution.The presented work is based on the intuition that defining ML models as modular and extensible artefacts allows to introduce a novel ML development methodology enabling the integration of multiple design and evaluation iterations into the continuous enrichment of a single unbounded intelligent system. We define a novel method for the generation of dynamic multitask ML models as a sequence of extensions and generalizations. We first analyze the capabilities of the proposed method by using the standard ML empirical evaluation methodology. Finally, we propose a novel continuous development methodology that allows to dynamically extend a pre-existing multitask large-scale ML system while analyzing the properties of the proposed method extensions. This results in the generation of an ML model capable of jointly solving 124 image classification tasks achieving state of the art quality with improved size and compute cost.

Paper PDF Code

Code

google-research/google-research officialjax report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Domain Generalization Fine-Grained Image Classification Image Classification Long-tail Learning Metric Learning Scene Classification image-classification

Results from the paper archive 2025-07-28

Task	Dataset	Model	Metric	Value	Rank at snapshot	Leaderboard	Report
Domain Generalization	ImageNet-A	µ2Net+ (ViT-L/16)	Top-1 accuracy %	84.53	#3 of 39	Archive leaderboard	report
Fine-Grained Image Classification	Caltech-101	µ2Net+ (ViT-L/16)	Top-1 Error Rate	4.06%	#4 of 18	Archive leaderboard	report
Fine-Grained Image Classification	Food-101	µ2Net+ (ViT-L/16)	Accuracy	91.47	#10 of 15	Archive leaderboard	report
Fine-Grained Image Classification	Oxford-IIIT Pets	µ2Net+ (ViT-L/16)	Accuracy	95.5	#3 of 19	Archive leaderboard	report
Fine-Grained Image Classification	Stanford Dogs	µ2Net+ (ViT-L/16)	Accuracy	93.5%	#4 of 24	Archive leaderboard	report
Image Classification	CARS196	µ2Net+ (ViT-L/16)	Accuracy	87.18	#1 of 1	Archive leaderboard	report
Image Classification	DTD	µ2Net+ (ViT-L/16)	Accuracy	82.23	#3 of 11	Archive leaderboard	report
Image Classification	EMNIST-Letters	µ2Net+ (ViT-L/16)	Accuracy	95.03	#7 of 11	Archive leaderboard	report
Image Classification	EuroSAT	µ2Net+ (ViT-L/16)	Accuracy (%)	99.22	#3 of 15	Archive leaderboard	report
Image Classification	ImageNet-Sketch	µ2Net+ (ViT-L/16)	Accuracy	88.6	#1 of 1	Archive leaderboard	report
Image Classification	Imagenette	µ2Net+ (ViT-L/16)	Accuracy	100	#1 of 2	Archive leaderboard	report
Image Classification	Malaria Dataset	µ2Net+ (ViT-L/16)	Acc. (test)	97.46%	#2 of 3	Archive leaderboard	report
Image Classification	Places365	µ2Net+ (ViT-L/16)	Top 1 Accuracy	59.15	#6 of 7	Archive leaderboard	report
Image Classification	STL-10	µ2Net+ (ViT-L/16)	Percentage correct	99.64	#1 of 117	Archive leaderboard	report
Image Classification	Stanford Online Products	µ2Net+ (ViT-L/16)	Accuracy	89.47	#1 of 1	Archive leaderboard	report
Image Classification	cats_vs_dogs	µ2Net+ (ViT-L/16)	Accuracy	99.83	#1 of 1	Archive leaderboard	report
Image Classification	iNaturalist 2018	µ2Net+ (ViT-L/16)	Top-1 Accuracy	80.97	#15 of 60	Archive leaderboard	report
Long-tail Learning	ImageNet-LT	µ2Net+ (ViT-L/16)	Top-1 Accuracy	82.5	#2 of 69	Archive leaderboard	report
Scene Classification	UC Merced Land Use Dataset	µ2Net+ (ViT-L/16)	Accuracy (%)	100	#1 of 6	Archive leaderboard	report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections