Upload UNet-TTE model

.gitattributes

@@ -0,0 +1 @@
+*.pth filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,31 @@
+# ResNetV2-CT Checkpoint
+
+This is a PyTorch Lightning `.ckpt` checkpoint for a SwinUNETR model trained on chest CT images with TTE objective.
+
+## Usage
+
+A quickstart script is below.
+
+```python
+import torch
+from src.networks import SwinUNETRForClassification
+swin_unetr_params = {
+            "img_size": (224, 224, 224),
+            "in_channels": 1,
+            "out_channels": 2, 
+            "feature_size": 48,
+            "drop_rate": 0.0,
+            "attn_drop_rate": 0.0,
+            "dropout_path_rate": 0.0,
+            "use_checkpoint": True,
+        }
+model = SwinUNETRForClassification(
+            swin_unetr_params=swin_unetr_params, num_classes=2
+        ).to(device)
+state_dict = torch.load(
+        loadmodel_path, map_location=f"cuda:{torch.cuda.current_device()}"
+    )
+model.load_state_dict(state_dict)
+```
+
+For detailed instructions please follow the [README in Github repo](https://github.com/som-shahlab/tte-pretraining/tree/main?tab=readme-ov-file#evaluation).

config.json

@@ -0,0 +1,6 @@
+{
+  "model_type": "model_3d",
+  "model_name": "swinunetr",
+  "pretrain_type": "time-to-event",
+  "num_class": 8192
+}

unet_tte.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5c7c45bdb2d5c60c04ad2bcefabbd34eec2a8911ed970c30c631d73ab9a566dd
+size 302639525