Returns vessel/ship type based on image with about 99% accuracy.
See https://www.kaggle.com/code/dima806/vessel-ship-type-detection for more details.
Classification report:
precision recall f1-score support
Cargo 0.9927 0.9623 0.9772 424
Carrier 0.9976 1.0000 0.9988 424
Cruise 1.0000 1.0000 1.0000 424
Military 0.9976 0.9976 0.9976 424
Tankers 0.9679 0.9953 0.9814 424
accuracy 0.9910 2120
macro avg 0.9912 0.9910 0.9910 2120
weighted avg 0.9912 0.9910 0.9910 2120
checkpoint-266/optimizer.pt 687 MBcheckpoint-2793/optimizer.pt 687 MBcheckpoint-6650/optimizer.pt 687 MBcheckpoint-266/model.safetensors 343 MBcheckpoint-2793/model.safetensors 343 MBcheckpoint-6650/model.safetensors 343 MBmodel.safetensors 343 MBcheckpoint-266/rng_state.pth 14.6 kBcheckpoint-2793/rng_state.pth 14.6 kBcheckpoint-6650/rng_state.pth 14.6 kBcheckpoint-266/training_args.bin 4.4 kBtraining_args.bin 4.4 kBcheckpoint-2793/training_args.bin 4.2 kBcheckpoint-6650/training_args.bin 4.1 kBcheckpoint-266/scheduler.pt 0.6 kBcheckpoint-2793/scheduler.pt 0.6 kBcheckpoint-6650/scheduler.pt 0.6 kB有文件？提交种子。
提交种子