| /opt/venv/lib/python3.10/site-packages/apex/transformer/functional/fused_rope.py:49: UserWarning: Aiter backend is selected for fused RoPE. This has lower precision. To disable aiter, export USE_ROCM_AITER_ROPE_BACKEND=0 |
| warnings.warn("Aiter backend is selected for fused RoPE. This has lower precision. To disable aiter, export USE_ROCM_AITER_ROPE_BACKEND=0", UserWarning) |
| /opt/venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:49: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers |
| warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning) |
| /opt/venv/lib/python3.10/site-packages/torch/functional.py:554: UserWarning: torch.meshgrid: in an upcoming release, it will be required to pass the indexing argument. (Triggered internally at /pytorch/aten/src/ATen/native/TensorShape.cpp:4317.) |
| return _VF.meshgrid(tensors, **kwargs) # type: ignore[attr-defined] |
|
|
| [1m[96m══════════════════════════════════════════════════════════════════════[0m |
| [1m[96m AgriFM × PASTIS — Fold 1 | 2026-04-17 03:35:49[0m |
| [1m[96m══════════════════════════════════════════════════════════════════════[0m |
|
|
| Device : |
| VRAM : 191.7 GB |
| Model size : small (~14M) |
| AMP : True |
| Batch size : 16 |
| Epochs : 100 |
| LR : 5e-05 |
| Weight decay : 0.05 |
| Warmup iters : 500 |
| Augmentation : True (flip + rotate) |
| Class weights : True |
| Work dir : ./work_dirs/fold1_small |
|
|
| [1m[93m──────────────────────────────────────────────────────────────────────[0m |
| [1m[93m Building Datasets[0m |
| [1m[93m──────────────────────────────────────────────────────────────────────[0m |
| PASTIS fold=1 split=train: 1279 patches (augment=True) |
| PASTIS fold=1 split=val: 623 patches (augment=False) |
| PASTIS fold=1 split=test: 531 patches (augment=False) |
|
|
| Train: 1279 Val: 623 Test: 531 |
| Train batches: 79 Val batches: 39 |
|
|
| [1m[93m──────────────────────────────────────────────────────────────────────[0m |
| [1m[93m Building Model[0m |
| [1m[93m──────────────────────────────────────────────────────────────────────[0m |
|
|
| Total params : 39.6M |
| Trainable params : 39.6M |
|
|
| [1m[93m──────────────────────────────────────────────────────────────────────[0m |
| [1m[93m Computing Class Weights[0m |
| [1m[93m──────────────────────────────────────────────────────────────────────[0m |
|
|
| Class weights (fold=1, sampled 300 patches): |
| Class Count Weight |
| ────────────────────────────── ────────── ──────── |
| Background 1969446 0.011 |
| Meadow 1025490 0.021 |
| Soft winter wheat 358833 0.059 |
| Corn 546641 0.039 |
| Winter barley 114863 0.184 |
| Winter rapeseed 100838 0.209 |
| Spring barley 31063 0.680 |
| Sunflower 51497 0.410 |
| Grapevine 34193 0.618 |
| Beet 32685 0.646 |
| Winter triticale 54068 0.391 |
| Winter durum wheat 3456 6.111 |
| Fruits vegetables flowers 11592 1.822 |
| Potatoes 9914 2.130 |
| Leguminous fodder 50089 0.422 |
| Soybeans 95149 0.222 |
| Orchard 6932 3.047 |
| Mixed cereal 25401 0.831 |
| Sorghum 18374 1.149 |
|
|
| [1m[96m══════════════════════════════════════════════════════════════════════[0m |
| [1m[96m Training | 100 epochs | model=small (~14M) | fold=1[0m |
| [1m[96m══════════════════════════════════════════════════════════════════════[0m |
|
|
| [1mTraining[0m |
| MIOpen(HIP): Error [Init] Not found :31-DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Default, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3> |
| [█████████████████████████░░░░░░░░░░░░░░░] 63.3% iter 50/79 loss=0.5012 lr=5.90e-06 eta=0:00:28 |
| [████████████████████████████████████████] 100.0% iter 79/79 loss=0.4708 lr=8.74e-06 eta=0:00:00 |
| [92mDone: avg_loss=0.4708 time=0:00:53[0m |
|
|
| [1mValidating...[0m |
| GridwiseOp: Problemsize descriptor dimension check failure |
| GridwiseOp: Problemsize descriptor dimension check failure |
| GridwiseOp: Problemsize descriptor dimension check failure |
| GridwiseOp: Problemsize descriptor dimension check failure |
| GridwiseOp: Problemsize descriptor dimension check failure |
| GridwiseOp: Problemsize descriptor dimension check failure |
| GridwiseOp: Problemsize descriptor dimension check failure |
| GridwiseOp: Problemsize descriptor dimension check failure |
| GridwiseOp: Problemsize descriptor dimension check failure |
| GridwiseOp: Problemsize descriptor dimension check failure |
| GridwiseOp: Problemsize descriptor dimension check failure |
| GridwiseOp: Problemsize descriptor dimension check failure |
| GridwiseOp: Problemsize descriptor dimension check failure |
| GridwiseOp: Problemsize descriptor dimension check failure |
| GridwiseOp: Problemsize descriptor dimension check failure |
| GridwiseOp: Problemsize descriptor dimension check failure |
| MIOpen(HIP): Error [Init] Not found :31-DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Default, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3> |
|
|
| [1m[93m──────────────────────────────────────────────────────────────────────[0m |
| [1m[93m Epoch 001 / 100 | Elapsed: 0:01:06[0m |
| [1m[93m──────────────────────────────────────────────────────────────────────[0m |
|
|
| Metric Value |
| ────────────────── ────────── |
| Train Loss 0.4708 |
| Val Loss 3.7996 |
| ────────────────── ────────── |
| [0m[1mOA [0m [0m[1m 9.70%[0m |
| [92m[1mmIoU [0m [92m[1m 0.58%[0m |
| [92m[1mmFscore [0m [92m[1m 1.05%[0m ★ |
| [0m[1mmPrecision [0m [0m[1m 2.21%[0m |
| [0m[1mmRecall [0m [0m[1m 5.76%[0m |
| [0m[1mKappa [0m [0m[1m 0.49%[0m |
| ────────────────── ────────── |
| Best mFscore 0.00% |
| Val Time 13.2s |
|
|
| [92m[1m★ New best mFscore: 1.05% → saved best_model.pth[0m |
|
|
| [1mTraining[0m |
| [█████████████████████████░░░░░░░░░░░░░░░] 63.3% iter 50/79 loss=0.3563 lr=1.36e-05 eta=0:00:04 |
| [████████████████████████████████████████] 100.0% iter 79/79 loss=0.3495 lr=1.65e-05 eta=0:00:00 |
| [92mDone: avg_loss=0.3495 time=0:00:12[0m |
|
|
| [1mValidating...[0m |
|
|
| [1m[93m──────────────────────────────────────────────────────────────────────[0m |
| [1m[93m Epoch 002 / 100 | Elapsed: 0:01:26[0m |
| [1m[93m──────────────────────────────────────────────────────────────────────[0m |
|
|
| Metric Value |
| ────────────────── ────────── |
| Train Loss 0.3495 |
| Val Loss 3.4866 |
| ────────────────── ────────── |
| [0m[1mOA [0m [0m[1m 2.49%[0m |
| [92m[1mmIoU [0m [92m[1m 1.32%[0m |
| [92m[1mmFscore [0m [92m[1m 2.50%[0m ★ |
| [0m[1mmPrecision [0m [0m[1m 5.63%[0m |
| [0m[1mmRecall [0m [0m[1m 9.04%[0m |
| [0m[1mKappa [0m [0m[1m 1.44%[0m |
| ────────────────── ────────── |
| Best mFscore 1.05% |
| Val Time 6.0s |
|
|
| [92m[1m★ New best mFscore: 2.50% → saved best_model.pth[0m |
|
|
| [1mTraining[0m |
| [█████████████████████████░░░░░░░░░░░░░░░] 63.3% iter 50/79 loss=0.2908 lr=2.14e-05 eta=0:00:04 |
| [████████████████████████████████████████] 100.0% iter 79/79 loss=0.2792 lr=2.42e-05 eta=0:00:00 |
| [92mDone: avg_loss=0.2792 time=0:00:12[0m |
|
|
| [1mValidating...[0m |
|
|
| [1m[93m──────────────────────────────────────────────────────────────────────[0m |
| [1m[93m Epoch 003 / 100 | Elapsed: 0:01:46[0m |
| [1m[93m──────────────────────────────────────────────────────────────────────[0m |
|
|
| Metric Value |
| ────────────────── ────────── |
| Train Loss 0.2792 |
| Val Loss 5.1773 |
| ────────────────── ────────── |
| [0m[1mOA [0m [0m[1m 1.30%[0m |
| [92m[1mmIoU [0m [92m[1m 1.15%[0m |
| [92m[1mmFscore [0m [92m[1m 2.22%[0m |
| [0m[1mmPrecision [0m [0m[1m 7.18%[0m |
| [0m[1mmRecall [0m [0m[1m 9.33%[0m |
| [0m[1mKappa [0m [0m[1m 0.54%[0m |
| ────────────────── ────────── |
| Best mFscore 2.50% |
| Val Time 6.0s |
|
|
| [1mTraining[0m |
| [█████████████████████████░░░░░░░░░░░░░░░] 63.3% iter 50/79 loss=0.2438 lr=2.91e-05 eta=0:00:04 |
| [████████████████████████████████████████] 100.0% iter 79/79 loss=0.2470 lr=3.20e-05 eta=0:00:00 |
| [92mDone: avg_loss=0.2470 time=0:00:12[0m |
|
|
| [1mValidating...[0m |
|
|
| [1m[93m──────────────────────────────────────────────────────────────────────[0m |
| [1m[93m Epoch 004 / 100 | Elapsed: 0:02:05[0m |
| [1m[93m──────────────────────────────────────────────────────────────────────[0m |
|
|
| Metric Value |
| ────────────────── ────────── |
| Train Loss 0.2470 |
| Val Loss 3.4497 |
| ────────────────── ────────── |
| [0m[1mOA [0m [0m[1m 3.89%[0m |
| [92m[1mmIoU [0m [92m[1m 1.82%[0m |
| [92m[1mmFscore [0m [92m[1m 3.41%[0m ★ |
| [0m[1mmPrecision [0m [0m[1m 10.45%[0m |
| [0m[1mmRecall [0m [0m[1m 8.55%[0m |
| [0m[1mKappa [0m [0m[1m 2.24%[0m |
| ────────────────── ────────── |
| Best mFscore 2.50% |
| Val Time 6.0s |
|
|
| [92m[1m★ New best mFscore: 3.41% → saved best_model.pth[0m |
|
|
| [1mTraining[0m |
| [█████████████████████████░░░░░░░░░░░░░░░] 63.3% iter 50/79 loss=0.2239 lr=3.69e-05 eta=0:00:04 |
| [████████████████████████████████████████] 100.0% iter 79/79 loss=0.2231 lr=3.97e-05 eta=0:00:00 |
| [92mDone: avg_loss=0.2231 time=0:00:12[0m |
|
|
| [1mValidating...[0m |
|
|
| [1m[93m──────────────────────────────────────────────────────────────────────[0m |
| [1m[93m Epoch 005 / 100 | Elapsed: 0:02:25[0m |
| [1m[93m──────────────────────────────────────────────────────────────────────[0m |
|
|
| Metric Value |
| ────────────────── ────────── |
| Train Loss 0.2231 |
| Val Loss 6.9362 |
| ────────────────── ────────── |
| [0m[1mOA [0m [0m[1m 1.77%[0m |
| [92m[1mmIoU [0m [92m[1m 0.85%[0m |
| [92m[1mmFscore [0m [92m[1m 1.66%[0m |
| [0m[1mmPrecision [0m [0m[1m 9.69%[0m |
|
|
