Upload folder using huggingface_hub

.gitattributes

@@ -33,3 +33,6 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+celeba_with_diff.png filter=lfs diff=lfs merge=lfs -text
+cifar10_with_diff.png filter=lfs diff=lfs merge=lfs -text
+mnist_with_diff.png filter=lfs diff=lfs merge=lfs -text

README.md

@@ -1,3 +1,70 @@
----
-license: cc-by-2.0
----
+---
+license: cc-by-nc-sa-4.0
+tags:
+  - flow-matching
+  - generative-model
+  - image-generation
+  - pytorch
+datasets:
+  - mnist
+  - cifar10
+  - celeba
+---
+
+# UNet Flow Matching Models
+
+Pre-trained UNet models for Flow Matching on image generation tasks.
+
+## Models
+
+This repository contains three trained UNet Flow Matching models:
+
+### MNIST (28×28 Grayscale)
+- **Checkpoint**: `mnist/ckpt.pth`
+- **Parameters**: 6.2M
+- **Configuration**: num_channels=64, num_res_blocks=2, class_cond=True
+- **Training**: 10 epochs
+
+### CIFAR-10 (32×32 RGB)
+- **Checkpoint**: `cifar10/ckpt.pth`
+- **Parameters**: 9.0M
+- **Configuration**: num_channels=64, num_res_blocks=2, class_cond=True
+- **Training**: 20 epochs
+
+### CelebA (64×64 RGB)
+- **Checkpoint**: `celeba64/ckpt.pth`
+- **Parameters**: 83.0M
+- **Configuration**: num_channels=128, num_res_blocks=2, class_cond=False
+- **Training**: 50 epochs, final loss=0.114
+
+## Sample Results
+
+### MNIST
+![MNIST Samples](mnist_with_diff.png)
+
+### CIFAR-10
+![CIFAR-10 Samples](cifar10_with_diff.png)
+
+### CelebA 64×64
+![CelebA Samples](celeba_with_diff.png)
+
+## Usage
+
+```python
+import torch
+from huggingface_hub import hf_hub_download
+
+# Download checkpoint
+ckpt_path = hf_hub_download(
+    repo_id="WayBob/FlowMatching-Unet-Celeb-64x64",
+    filename="celeba64/ckpt.pth"
+)
+
+# Load model (requires the UNet implementation)
+checkpoint = torch.load(ckpt_path, map_location="cuda")
+# ... load into your UNet model
+```
+
+## License
+
+CC BY-NC-SA 4.0