Training Report — hummingbird_classifier_20260430_193523¶
Generated: 2026-05-03 11:49:08
W&B Run: lnjepxc0
Project: hummingbird-detection
Git commit: 13516d69
SLURM Job ID: 7655232
Compute Resources¶
| Resource | Value |
|---|---|
| Host | gl024.hpc.nyu.edu |
| GPU | NVIDIA L40S ×1 |
| GPU memory | 48.3 GB |
| CUDA cores | 18176 |
| Architecture | Ada |
| CPU cores | 128 |
| RAM | 540.2 GB |
| Partition | l40s_public |
| QoS | gpu48 |
Hyperparameters¶
| Parameter | Value |
|---|---|
| Epochs (requested) | 50 |
| Epochs (completed) | 12 |
| Batch size | 128 |
| Learning rate | 0.0012 |
| Weight decay | 0.0001 |
| Early stopping patience | 5 |
| Seed | 42 |
| Train dir | data/interim/train_data/train |
| Val dir | data/interim/train_data/val |
Best Result¶
| Metric | Value | Epoch |
|---|---|---|
| Best val loss | 0.0804 | 7 |
| Val accuracy at best | 0.9935 | 7 |
| Val precision at best | 0.9865 | 7 |
| Val recall at best | 0.9932 | 7 |
| Val F1 at best | 0.9898 | 7 |
| Training duration | 7m 45s | — |
Final Epoch Metrics¶
| Metric | Value |
|---|---|
| Train loss | 0.0112 |
| Train accuracy | 0.9973 |
| Val loss | 0.0952 |
| Val accuracy | 0.9913 |
| Val precision | 0.9831 |
| Val recall | 0.9898 |
| Val F1 | 0.9864 |
Model Metrics¶
System Metrics¶
Per-Epoch Training Log¶
2026-04-30 15:36:20,817 [INFO] humming_bird_detection.models.classifier — Epoch 1/50 lr=1.20e-03 train_loss=0.6492 train_acc=0.856 val_loss=0.4259 val_acc=0.905 val_prec=0.781 val_rec=0.973 val_f1=0.867 ← best
2026-04-30 15:36:55,242 [INFO] humming_bird_detection.models.classifier — Epoch 2/50 lr=1.20e-03 train_loss=0.3737 train_acc=0.925 val_loss=0.3160 val_acc=0.925 val_prec=0.826 val_rec=0.969 val_f1=0.892 ← best
2026-04-30 15:37:30,386 [INFO] humming_bird_detection.models.classifier — Epoch 3/50 lr=1.20e-03 train_loss=0.2991 train_acc=0.931 val_loss=0.2707 val_acc=0.933 val_prec=0.843 val_rec=0.969 val_f1=0.902 ← best
2026-04-30 15:38:08,041 [INFO] humming_bird_detection.models.classifier — Epoch 4/50 lr=1.20e-03 train_loss=0.2579 train_acc=0.944 val_loss=0.2415 val_acc=0.941 val_prec=0.866 val_rec=0.966 val_f1=0.913 ← best
2026-04-30 15:38:42,769 [INFO] humming_bird_detection.models.classifier — Epoch 5/50 lr=1.20e-03 train_loss=0.2240 train_acc=0.952 val_loss=0.2286 val_acc=0.940 val_prec=0.861 val_rec=0.969 val_f1=0.912 ← best
2026-04-30 15:39:25,020 [INFO] humming_bird_detection.models.classifier — Epoch 6/50 lr=1.20e-04 train_loss=0.1105 train_acc=0.982 val_loss=0.0852 val_acc=0.991 val_prec=0.986 val_rec=0.986 val_f1=0.986 ← best
2026-04-30 15:40:03,889 [INFO] humming_bird_detection.models.classifier — Epoch 7/50 lr=1.19e-04 train_loss=0.0399 train_acc=0.992 val_loss=0.0804 val_acc=0.993 val_prec=0.986 val_rec=0.993 val_f1=0.990 ← best
2026-04-30 15:40:45,641 [INFO] humming_bird_detection.models.classifier — Epoch 8/50 lr=1.19e-04 train_loss=0.0275 train_acc=0.994 val_loss=0.0850 val_acc=0.991 val_prec=0.983 val_rec=0.990 val_f1=0.986
2026-04-30 15:41:27,829 [INFO] humming_bird_detection.models.classifier — Epoch 9/50 lr=1.18e-04 train_loss=0.0244 train_acc=0.995 val_loss=0.0903 val_acc=0.992 val_prec=0.983 val_rec=0.993 val_f1=0.988
2026-04-30 15:42:11,918 [INFO] humming_bird_detection.models.classifier — Epoch 10/50 lr=1.16e-04 train_loss=0.0181 train_acc=0.995 val_loss=0.0965 val_acc=0.991 val_prec=0.983 val_rec=0.990 val_f1=0.986
2026-04-30 15:42:54,786 [INFO] humming_bird_detection.models.classifier — Epoch 11/50 lr=1.15e-04 train_loss=0.0099 train_acc=0.998 val_loss=0.1058 val_acc=0.991 val_prec=0.986 val_rec=0.986 val_f1=0.986
2026-04-30 15:43:37,832 [INFO] humming_bird_detection.models.classifier — Epoch 12/50 lr=1.13e-04 train_loss=0.0112 train_acc=0.997 val_loss=0.0952 val_acc=0.991 val_prec=0.983 val_rec=0.990 val_f1=0.986