best_model.pt

README.md

@@ -0,0 +1,69 @@
+---
+library_name: transformers
+tags:
+- generated_from_trainer
+metrics:
+- accuracy
+- f1
+model-index:
+- name: windowz_020625
+  results: []
+---
+
+<!-- This model card has been generated automatically according to the information the Trainer had access to. You
+should probably proofread and complete it, then remove this comment. -->
+
+# windowz_020625
+
+This model is a fine-tuned version of [](https://huggingface.co/) on an unknown dataset.
+It achieves the following results on the evaluation set:
+- Accuracy: 0.9891
+- F1: 0.9871
+- Iou: 0.9786
+- Contour Dice: 0.9923
+- Per Class Metrics: {0: {'f1': 0.99741, 'iou': 0.99483, 'accuracy': 0.99611, 'contour_dice': 0.99741}, 1: {'f1': 0.97774, 'iou': 0.95645, 'accuracy': 0.98909, 'contour_dice': 0.97774}, 2: {'f1': 0.41176, 'iou': 0.25926, 'accuracy': 0.99298, 'contour_dice': 0.41176}}
+- Loss: 0.0301
+
+## Model description
+
+More information needed
+
+## Intended uses & limitations
+
+More information needed
+
+## Training and evaluation data
+
+More information needed
+
+## Training procedure
+
+### Training hyperparameters
+
+The following hyperparameters were used during training:
+- learning_rate: 5e-05
+- train_batch_size: 16
+- eval_batch_size: 16
+- seed: 42
+- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
+- lr_scheduler_type: cosine
+- lr_scheduler_warmup_steps: 1000
+- num_epochs: 100
+
+### Training results
+
+| Training Loss | Epoch | Step  |        | Dice   | Class Metrics                                                                                                                                                                                                                                          | Validation Loss |
+|:-------------:|:-----:|:-----:|:------:|:------:|:------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------:|:---------------:|
+| 0.4629        | 5.0   | 12815 | 0.9623 | 0.9725 | {0: {'f1': 0.99106, 'iou': 0.98228, 'accuracy': 0.98651, 'contour_dice': 0.99106}, 1: {'f1': 0.96086, 'iou': 0.92467, 'accuracy': 0.9813, 'contour_dice': 0.96086}, 2: {'f1': 0.51615, 'iou': 0.34784, 'accuracy': 0.99358, 'contour_dice': 0.51615}}  | 0.0652          |
+| 0.4439        | 10.0  | 25630 | 0.9786 | 0.9923 | {0: {'f1': 0.99741, 'iou': 0.99483, 'accuracy': 0.99611, 'contour_dice': 0.99741}, 1: {'f1': 0.97774, 'iou': 0.95645, 'accuracy': 0.98909, 'contour_dice': 0.97774}, 2: {'f1': 0.41176, 'iou': 0.25926, 'accuracy': 0.99298, 'contour_dice': 0.41176}} | 0.0301          |
+| 0.4077        | 15.0  | 38445 | 0.2129 | 0.4508 | {0: {'f1': 0.31665, 'iou': 0.18811, 'accuracy': 0.39105, 'contour_dice': 0.31665}, 1: {'f1': 0.43873, 'iou': 0.28101, 'accuracy': 0.38547, 'contour_dice': 0.43873}, 2: {'f1': 0.60057, 'iou': 0.42915, 'accuracy': 0.99441, 'contour_dice': 0.60057}} | 0.8030          |
+| 0.4099        | 20.0  | 51260 | 0.9796 | 0.9892 | {0: {'f1': 0.9964, 'iou': 0.99283, 'accuracy': 0.9946, 'contour_dice': 0.9964}, 1: {'f1': 0.97792, 'iou': 0.9568, 'accuracy': 0.98934, 'contour_dice': 0.97792}, 2: {'f1': 0.67584, 'iou': 0.51039, 'accuracy': 0.99419, 'contour_dice': 0.67584}}     | 0.0441          |
+| 0.3872        | 25.0  | 64075 | 0.9799 | 0.9882 | {0: {'f1': 0.99605, 'iou': 0.99212, 'accuracy': 0.99407, 'contour_dice': 0.99605}, 1: {'f1': 0.9783, 'iou': 0.95752, 'accuracy': 0.98953, 'contour_dice': 0.9783}, 2: {'f1': 0.73739, 'iou': 0.58402, 'accuracy': 0.99521, 'contour_dice': 0.73739}}   | 0.0314          |
+
+
+### Framework versions
+
+- Transformers 4.45.0
+- Pytorch 2.5.1+cu124
+- Datasets 2.21.0
+- Tokenizers 0.20.3

config.json

@@ -0,0 +1,24 @@
+{
+  "architectures": [
+    "UNETForSegmentation"
+  ],
+  "dim": 224,
+  "hidden_act": "gelu",
+  "hidden_size": 256,
+  "img_size": 128,
+  "intermediate_size": 1024,
+  "is_causal": false,
+  "k": 2,
+  "model_type": "Unet",
+  "n_filts": 4,
+  "num_attention_heads": 8,
+  "num_channels": 3,
+  "num_classes": 3,
+  "num_hidden_layers": 6,
+  "num_layers": 2,
+  "patch_size": 16,
+  "problem_type": "single_label_classification",
+  "t": 2,
+  "torch_dtype": "float32",
+  "transformers_version": "4.45.0"
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:28158680e2ee00ef764b6cd76252a02023bd63fbdd08dccd4ccd3fa3dc25d9b5
+size 2188724

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3775bff2eca752f2956fa9b634c475229429bef905f6e58970e222095f1a7cff
+size 5176