# Image Classification on CIFAR-100 (test) (Time Reduction Metrics) Source: [Image Classification on CIFAR-100 (test) (Time Reduction Metrics)]() Collection index: [llms.txt]() Data updated: 2026-03-30T07:29:36.161Z Evaluation of Image Classification performance on the CIFAR-100 test set, focusing on efficiency metrics such as time reduction compared to Round-Optimized and AsyncSGD baselines. A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. | Scope | Recorded value | | --- | --- | | Task | Image Classification | | Dataset (including any stated split) | CIFAR-100 (test) | | Metric | Time Reduction (%) vs Round-Optimized | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 6 of 6 | | Coverage | All results currently returned for this benchmark | Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol. ## Metrics - [Time Reduction (%) vs Round-Optimized](): Higher is better; unit: PERCENT. - [Time Reduction (%) vs AsyncSGD](): Higher is better; unit: PERCENT. ## Recorded results | Method | Time Reduction (%) vs Round-Optimized | Result date | Report paper | | --- | --- | --- | --- | | Time-Optimized Generalized AsyncSGD | 67.95 | Not recorded | 69ca03670636e5bf4afc923c | | Time-Optimized Generalized AsyncSGD | 67.31 | Not recorded | 69ca03670636e5bf4afc923c | | Time-Optimized Generalized AsyncSGD | 66.23 | Not recorded | 69ca03670636e5bf4afc923c | | Time-Optimized Generalized AsyncSGD | 65.46 | Not recorded | 69ca03670636e5bf4afc923c | | Time-Optimized Generalized AsyncSGD | 60.52 | Not recorded | 69ca03670636e5bf4afc923c | | Time-Optimized Generalized AsyncSGD | 51.79 | Not recorded | 69ca03670636e5bf4afc923c | ### Time-Optimized Generalized AsyncSGD Time Reduction (%) vs Round-Optimized: 67.95. Result ID: 69ca25c347f34ce924b2692b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Dist.":"Exp.","Data":"IID","Target test accuracy":"0.5"}. This record does not assert independent reproduction. ### Time-Optimized Generalized AsyncSGD Time Reduction (%) vs Round-Optimized: 67.31. Result ID: 69ca25c347f34ce924b2692d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Dist.":"LogN.","Data":"IID","Target test accuracy":"0.5"}. This record does not assert independent reproduction. ### Time-Optimized Generalized AsyncSGD Time Reduction (%) vs Round-Optimized: 66.23. Result ID: 69ca25c347f34ce924b2692f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Dist.":"Det.","Data":"IID","Target test accuracy":"0.5"}. This record does not assert independent reproduction. ### Time-Optimized Generalized AsyncSGD Time Reduction (%) vs Round-Optimized: 65.46. Result ID: 69ca25c347f34ce924b2692e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Dist.":"LogN.","Data":"Non-IID","Target test accuracy":"0.5"}. This record does not assert independent reproduction. ### Time-Optimized Generalized AsyncSGD Time Reduction (%) vs Round-Optimized: 60.52. Result ID: 69ca25c347f34ce924b2692c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Dist.":"Exp.","Data":"Non-IID","Target test accuracy":"0.5"}. This record does not assert independent reproduction. ### Time-Optimized Generalized AsyncSGD Time Reduction (%) vs Round-Optimized: 51.79. Result ID: 69ca25c347f34ce924b26930. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Dist.":"Det.","Data":"Non-IID","Target test accuracy":"0.5"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.