pcam-interpretability/resnet50-val08783-resnet-tuned-aug

resnet50

Best Validation Accuracy: 0.8783

Metadata

• Model Name: resnet50
• Optimizer: adamw
• Scheduler: step
• Weight Decay: 0.0005
• Warmup Epochs: 3
• Patience: 7
• Amp: True
• Seed: 42
• Batch Size: 32
• Initial Lr: 0.0005
• Total Epochs Ran: 8
• Early Stopped: True
• Training Time Seconds: 19030.771079063416
• Num Parameters: 22067009
• Device: NVIDIA A100-SXM4-40GB
• Run Id: resnet-tuned-aug

Training Configuration

• Epochs: 8
• Batch size: 32
• Learning rate (initial): 0.0005

Training Logs (Per Epoch)

Epoch	Train Loss	Train Acc	Val Loss	Val Acc	LR
1	0.3234	0.8624	0.2944	0.8783	0.000167
2	0.2762	0.8876	0.3729	0.8444	0.000333
3	0.2600	0.8955	0.4210	0.8333	0.000500
4	0.2325	0.9082	0.4866	0.8172	0.000500
5	0.2138	0.9168	0.3420	0.8703	0.000500
6	0.1990	0.9236	0.4232	0.8492	0.000500
7	0.1889	0.9281	0.3532	0.8740	0.000500
8	0.1795	0.9319	0.4062	0.8603	0.000500