pcam-interpretability/resnet50-val09050-resnet-tuned

resnet50

Best Validation Accuracy: 0.9050

Metadata

• Model Name: resnet50
• Optimizer: adamw
• Scheduler: cosine
• Weight Decay: 0.0005
• Warmup Epochs: 3
• Patience: 7
• Amp: True
• Seed: 42
• Batch Size: 32
• Initial Lr: 0.0005
• Total Epochs Ran: 14
• Early Stopped: True
• Training Time Seconds: 33306.75762796402
• Num Parameters: 22067009
• Device: NVIDIA A100-SXM4-40GB
• Run Id: resnet-tuned

Training Configuration

• Epochs: 14
• Batch size: 32
• Learning rate (initial): 0.0005

Training Logs (Per Epoch)

Epoch	Train Loss	Train Acc	Val Loss	Val Acc	LR
1	0.2662	0.8921	0.2935	0.8786	0.000167
2	0.2245	0.9125	0.3381	0.8763	0.000333
3	0.2090	0.9196	0.4355	0.8174	0.000500
4	0.1841	0.9304	0.3426	0.8751	0.000500
5	0.1678	0.9376	0.3754	0.8611	0.000499
6	0.1554	0.9430	0.4509	0.8415	0.000495
7	0.1447	0.9468	0.2905	0.9050	0.000488
8	0.1373	0.9502	0.4505	0.8444	0.000478
9	0.1297	0.9539	0.4726	0.8473	0.000467
10	0.1224	0.9569	0.3371	0.8867	0.000452
11	0.1167	0.9583	0.3782	0.8741	0.000436
12	0.1115	0.9604	0.4738	0.8645	0.000417
13	0.1052	0.9626	0.3477	0.8896	0.000397
14	0.0998	0.9647	0.4067	0.8845	0.000375