Image Classification on MNIST (test) (Accuracy, Resource, and Training Metrics) benchmark leaderboard · SOTA2 Research

Image Classification on MNIST (test) (Accuracy, Resource, and Training Metrics)

94.61AccuracyINT16

Evaluation Results

Method	Links
INT162026.04		94.61	30.98	610
Dynamic2026.04		93.26	31.27	460
FP322026.04		93.22	61.97	460
Cosine2026.04		93.04	38.75	460
INT82026.04		92.85	15.49	460
FP322026.04		83.49	123.93	970
Dynamic2026.04		82.13	61.3	890
Cosine2026.04		81.77	77.5	770
INT82026.04		81.5	30.98	730
INT162026.04		78.08	61.97	340
Dynamic2026.04		77.13	120.36	540
Cosine2026.04		76.99	155.01	990
INT42026.04		76.65	7.75	860
INT82026.04		76.49	61.97	540
INT162026.04		74.96	123.93	710
FP322026.04		72.49	247.86	620
FP322026.04		66.5	1,239.31	760
INT162026.04		65.76	619.65	980
Dynamic2026.04		65.55	586.92	610
Cosine2026.04		64.85	775.03	820
INT82026.04		64.4	309.83	990
INT42026.04		60.49	15.49	380
INT42026.04		49.37	30.98	230
INT42026.04		40.17	154.91	450