Add model cards for val_holdout (batch 2)

val_holdout/model_idx_4300/README.md

@@ -0,0 +1,40 @@
+---
+base_model: CompVis/stable-diffusion-v1-2
+library_name: peft
+pipeline_tag: text-to-image
+tags:
+  - probex
+  - model-j
+  - weight-space-learning
+  - lora
+  - stable-diffusion
+---
+
+# Model-J: SD_1k - model_idx_4300
+
+This model is part of the **Model-J** dataset, introduced in:
+
+**Learning on Model Weights using Tree Experts** (CVPR 2025) by Eliahu Horwitz*, Bar Cavia*, Jonathan Kahana*, Yedid Hoshen
+
+<p align="center">
+    🌐 <a href="https://horwitz.ai/probex" target="_blank">Project</a> | 📃 <a href="https://arxiv.org/abs/2410.13569" target="_blank">Paper</a> | 💻 <a href="https://github.com/eliahuhorwitz/ProbeX" target="_blank">GitHub</a> | 🤗 <a href="https://huggingface.co/ProbeX" target="_blank">Dataset</a>
+</p>
+
+## Model Details
+
+| Attribute | Value |
+|---|---|
+| **Subset** | SD_1k |
+| **Split** | val_holdout |
+| **Base Model** | `CompVis/stable-diffusion-v1-2` |
+| **ImageNet Class** | joystick (n03602883) |
+
+## Training Hyperparameters
+
+| Parameter | Value |
+|---|---|
+| Learning Rate | 7e-05 |
+| Max Train Steps | 500 |
+| LoRA Rank | 16 |
+| Seed | 24 |
+| N Training Samples | 5 |

val_holdout/model_idx_4301/README.md

@@ -0,0 +1,40 @@
+---
+base_model: CompVis/stable-diffusion-v1-3
+library_name: peft
+pipeline_tag: text-to-image
+tags:
+  - probex
+  - model-j
+  - weight-space-learning
+  - lora
+  - stable-diffusion
+---
+
+# Model-J: SD_1k - model_idx_4301
+
+This model is part of the **Model-J** dataset, introduced in:
+
+**Learning on Model Weights using Tree Experts** (CVPR 2025) by Eliahu Horwitz*, Bar Cavia*, Jonathan Kahana*, Yedid Hoshen
+
+<p align="center">
+    🌐 <a href="https://horwitz.ai/probex" target="_blank">Project</a> | 📃 <a href="https://arxiv.org/abs/2410.13569" target="_blank">Paper</a> | 💻 <a href="https://github.com/eliahuhorwitz/ProbeX" target="_blank">GitHub</a> | 🤗 <a href="https://huggingface.co/ProbeX" target="_blank">Dataset</a>
+</p>
+
+## Model Details
+
+| Attribute | Value |
+|---|---|
+| **Subset** | SD_1k |
+| **Split** | val_holdout |
+| **Base Model** | `CompVis/stable-diffusion-v1-3` |
+| **ImageNet Class** | green_snake (n01729977) |
+
+## Training Hyperparameters
+
+| Parameter | Value |
+|---|---|
+| Learning Rate | 3e-05 |
+| Max Train Steps | 700 |
+| LoRA Rank | 16 |
+| Seed | 26 |
+| N Training Samples | 7 |

val_holdout/model_idx_4302/README.md

@@ -0,0 +1,40 @@
+---
+base_model: CompVis/stable-diffusion-v1-2
+library_name: peft
+pipeline_tag: text-to-image
+tags:
+  - probex
+  - model-j
+  - weight-space-learning
+  - lora
+  - stable-diffusion
+---
+
+# Model-J: SD_1k - model_idx_4302
+
+This model is part of the **Model-J** dataset, introduced in:
+
+**Learning on Model Weights using Tree Experts** (CVPR 2025) by Eliahu Horwitz*, Bar Cavia*, Jonathan Kahana*, Yedid Hoshen
+
+<p align="center">
+    🌐 <a href="https://horwitz.ai/probex" target="_blank">Project</a> | 📃 <a href="https://arxiv.org/abs/2410.13569" target="_blank">Paper</a> | 💻 <a href="https://github.com/eliahuhorwitz/ProbeX" target="_blank">GitHub</a> | 🤗 <a href="https://huggingface.co/ProbeX" target="_blank">Dataset</a>
+</p>
+
+## Model Details
+
+| Attribute | Value |
+|---|---|
+| **Subset** | SD_1k |
+| **Split** | val_holdout |
+| **Base Model** | `CompVis/stable-diffusion-v1-2` |
+| **ImageNet Class** | balance_beam (n02777292) |
+
+## Training Hyperparameters
+
+| Parameter | Value |
+|---|---|
+| Learning Rate | 0.0001 |
+| Max Train Steps | 450 |
+| LoRA Rank | 16 |
+| Seed | 22 |
+| N Training Samples | 10 |

val_holdout/model_idx_4303/README.md

@@ -0,0 +1,40 @@
+---
+base_model: CompVis/stable-diffusion-v1-4
+library_name: peft
+pipeline_tag: text-to-image
+tags:
+  - probex
+  - model-j
+  - weight-space-learning
+  - lora
+  - stable-diffusion
+---
+
+# Model-J: SD_1k - model_idx_4303
+
+This model is part of the **Model-J** dataset, introduced in:
+
+**Learning on Model Weights using Tree Experts** (CVPR 2025) by Eliahu Horwitz*, Bar Cavia*, Jonathan Kahana*, Yedid Hoshen
+
+<p align="center">
+    🌐 <a href="https://horwitz.ai/probex" target="_blank">Project</a> | 📃 <a href="https://arxiv.org/abs/2410.13569" target="_blank">Paper</a> | 💻 <a href="https://github.com/eliahuhorwitz/ProbeX" target="_blank">GitHub</a> | 🤗 <a href="https://huggingface.co/ProbeX" target="_blank">Dataset</a>
+</p>
+
+## Model Details
+
+| Attribute | Value |
+|---|---|
+| **Subset** | SD_1k |
+| **Split** | val_holdout |
+| **Base Model** | `CompVis/stable-diffusion-v1-4` |
+| **ImageNet Class** | breakwater (n02894605) |
+
+## Training Hyperparameters
+
+| Parameter | Value |
+|---|---|
+| Learning Rate | 0.0001 |
+| Max Train Steps | 450 |
+| LoRA Rank | 16 |
+| Seed | 48 |
+| N Training Samples | 5 |

val_holdout/model_idx_4304/README.md

@@ -0,0 +1,40 @@
+---
+base_model: CompVis/stable-diffusion-v1-4
+library_name: peft
+pipeline_tag: text-to-image
+tags:
+  - probex
+  - model-j
+  - weight-space-learning
+  - lora
+  - stable-diffusion
+---
+
+# Model-J: SD_1k - model_idx_4304
+
+This model is part of the **Model-J** dataset, introduced in:
+
+**Learning on Model Weights using Tree Experts** (CVPR 2025) by Eliahu Horwitz*, Bar Cavia*, Jonathan Kahana*, Yedid Hoshen
+
+<p align="center">
+    🌐 <a href="https://horwitz.ai/probex" target="_blank">Project</a> | 📃 <a href="https://arxiv.org/abs/2410.13569" target="_blank">Paper</a> | 💻 <a href="https://github.com/eliahuhorwitz/ProbeX" target="_blank">GitHub</a> | 🤗 <a href="https://huggingface.co/ProbeX" target="_blank">Dataset</a>
+</p>
+
+## Model Details
+
+| Attribute | Value |
+|---|---|
+| **Subset** | SD_1k |
+| **Split** | val_holdout |
+| **Base Model** | `CompVis/stable-diffusion-v1-4` |
+| **ImageNet Class** | minivan (n03770679) |
+
+## Training Hyperparameters
+
+| Parameter | Value |
+|---|---|
+| Learning Rate | 0.0003 |
+| Max Train Steps | 600 |
+| LoRA Rank | 16 |
+| Seed | 61 |
+| N Training Samples | 10 |

val_holdout/model_idx_4305/README.md

@@ -0,0 +1,40 @@
+---
+base_model: CompVis/stable-diffusion-v1-3
+library_name: peft
+pipeline_tag: text-to-image
+tags:
+  - probex
+  - model-j
+  - weight-space-learning
+  - lora
+  - stable-diffusion
+---
+
+# Model-J: SD_1k - model_idx_4305
+
+This model is part of the **Model-J** dataset, introduced in:
+
+**Learning on Model Weights using Tree Experts** (CVPR 2025) by Eliahu Horwitz*, Bar Cavia*, Jonathan Kahana*, Yedid Hoshen
+
+<p align="center">
+    🌐 <a href="https://horwitz.ai/probex" target="_blank">Project</a> | 📃 <a href="https://arxiv.org/abs/2410.13569" target="_blank">Paper</a> | 💻 <a href="https://github.com/eliahuhorwitz/ProbeX" target="_blank">GitHub</a> | 🤗 <a href="https://huggingface.co/ProbeX" target="_blank">Dataset</a>
+</p>
+
+## Model Details
+
+| Attribute | Value |
+|---|---|
+| **Subset** | SD_1k |
+| **Split** | val_holdout |
+| **Base Model** | `CompVis/stable-diffusion-v1-3` |
+| **ImageNet Class** | hare (n02326432) |
+
+## Training Hyperparameters
+
+| Parameter | Value |
+|---|---|
+| Learning Rate | 0.0005 |
+| Max Train Steps | 500 |
+| LoRA Rank | 16 |
+| Seed | 99 |
+| N Training Samples | 5 |

val_holdout/model_idx_4306/README.md

@@ -0,0 +1,40 @@
+---
+base_model: CompVis/stable-diffusion-v1-2
+library_name: peft
+pipeline_tag: text-to-image
+tags:
+  - probex
+  - model-j
+  - weight-space-learning
+  - lora
+  - stable-diffusion
+---
+
+# Model-J: SD_1k - model_idx_4306
+
+This model is part of the **Model-J** dataset, introduced in:
+
+**Learning on Model Weights using Tree Experts** (CVPR 2025) by Eliahu Horwitz*, Bar Cavia*, Jonathan Kahana*, Yedid Hoshen
+
+<p align="center">
+    🌐 <a href="https://horwitz.ai/probex" target="_blank">Project</a> | 📃 <a href="https://arxiv.org/abs/2410.13569" target="_blank">Paper</a> | 💻 <a href="https://github.com/eliahuhorwitz/ProbeX" target="_blank">GitHub</a> | 🤗 <a href="https://huggingface.co/ProbeX" target="_blank">Dataset</a>
+</p>
+
+## Model Details
+
+| Attribute | Value |
+|---|---|
+| **Subset** | SD_1k |
+| **Split** | val_holdout |
+| **Base Model** | `CompVis/stable-diffusion-v1-2` |
+| **ImageNet Class** | porcupine (n02346627) |
+
+## Training Hyperparameters
+
+| Parameter | Value |
+|---|---|
+| Learning Rate | 5e-05 |
+| Max Train Steps | 650 |
+| LoRA Rank | 16 |
+| Seed | 32 |
+| N Training Samples | 5 |

val_holdout/model_idx_4307/README.md

@@ -0,0 +1,40 @@
+---
+base_model: CompVis/stable-diffusion-v1-4
+library_name: peft
+pipeline_tag: text-to-image
+tags:
+  - probex
+  - model-j
+  - weight-space-learning
+  - lora
+  - stable-diffusion
+---
+
+# Model-J: SD_1k - model_idx_4307
+
+This model is part of the **Model-J** dataset, introduced in:
+
+**Learning on Model Weights using Tree Experts** (CVPR 2025) by Eliahu Horwitz*, Bar Cavia*, Jonathan Kahana*, Yedid Hoshen
+
+<p align="center">
+    🌐 <a href="https://horwitz.ai/probex" target="_blank">Project</a> | 📃 <a href="https://arxiv.org/abs/2410.13569" target="_blank">Paper</a> | 💻 <a href="https://github.com/eliahuhorwitz/ProbeX" target="_blank">GitHub</a> | 🤗 <a href="https://huggingface.co/ProbeX" target="_blank">Dataset</a>
+</p>
+
+## Model Details
+
+| Attribute | Value |
+|---|---|
+| **Subset** | SD_1k |
+| **Split** | val_holdout |
+| **Base Model** | `CompVis/stable-diffusion-v1-4` |
+| **ImageNet Class** | centipede (n01784675) |
+
+## Training Hyperparameters
+
+| Parameter | Value |
+|---|---|
+| Learning Rate | 0.0001 |
+| Max Train Steps | 600 |
+| LoRA Rank | 16 |
+| Seed | 33 |
+| N Training Samples | 6 |

val_holdout/model_idx_4308/README.md

@@ -0,0 +1,40 @@
+---
+base_model: stable-diffusion-v1-5/stable-diffusion-v1-5
+library_name: peft
+pipeline_tag: text-to-image
+tags:
+  - probex
+  - model-j
+  - weight-space-learning
+  - lora
+  - stable-diffusion
+---
+
+# Model-J: SD_1k - model_idx_4308
+
+This model is part of the **Model-J** dataset, introduced in:
+
+**Learning on Model Weights using Tree Experts** (CVPR 2025) by Eliahu Horwitz*, Bar Cavia*, Jonathan Kahana*, Yedid Hoshen
+
+<p align="center">
+    🌐 <a href="https://horwitz.ai/probex" target="_blank">Project</a> | 📃 <a href="https://arxiv.org/abs/2410.13569" target="_blank">Paper</a> | 💻 <a href="https://github.com/eliahuhorwitz/ProbeX" target="_blank">GitHub</a> | 🤗 <a href="https://huggingface.co/ProbeX" target="_blank">Dataset</a>
+</p>
+
+## Model Details
+
+| Attribute | Value |
+|---|---|
+| **Subset** | SD_1k |
+| **Split** | val_holdout |
+| **Base Model** | `stable-diffusion-v1-5/stable-diffusion-v1-5` |
+| **ImageNet Class** | shield (n04192698) |
+
+## Training Hyperparameters
+
+| Parameter | Value |
+|---|---|
+| Learning Rate | 3e-05 |
+| Max Train Steps | 550 |
+| LoRA Rank | 16 |
+| Seed | 73 |
+| N Training Samples | 6 |

val_holdout/model_idx_4309/README.md

@@ -0,0 +1,40 @@
+---
+base_model: CompVis/stable-diffusion-v1-2
+library_name: peft
+pipeline_tag: text-to-image
+tags:
+  - probex
+  - model-j
+  - weight-space-learning
+  - lora
+  - stable-diffusion
+---
+
+# Model-J: SD_1k - model_idx_4309
+
+This model is part of the **Model-J** dataset, introduced in:
+
+**Learning on Model Weights using Tree Experts** (CVPR 2025) by Eliahu Horwitz*, Bar Cavia*, Jonathan Kahana*, Yedid Hoshen
+
+<p align="center">
+    🌐 <a href="https://horwitz.ai/probex" target="_blank">Project</a> | 📃 <a href="https://arxiv.org/abs/2410.13569" target="_blank">Paper</a> | 💻 <a href="https://github.com/eliahuhorwitz/ProbeX" target="_blank">GitHub</a> | 🤗 <a href="https://huggingface.co/ProbeX" target="_blank">Dataset</a>
+</p>
+
+## Model Details
+
+| Attribute | Value |
+|---|---|
+| **Subset** | SD_1k |
+| **Split** | val_holdout |
+| **Base Model** | `CompVis/stable-diffusion-v1-2` |
+| **ImageNet Class** | binder (n02840245) |
+
+## Training Hyperparameters
+
+| Parameter | Value |
+|---|---|
+| Learning Rate | 3e-05 |
+| Max Train Steps | 500 |
+| LoRA Rank | 16 |
+| Seed | 90 |
+| N Training Samples | 5 |

val_holdout/model_idx_4310/README.md

@@ -0,0 +1,40 @@
+---
+base_model: CompVis/stable-diffusion-v1-4
+library_name: peft
+pipeline_tag: text-to-image
+tags:
+  - probex
+  - model-j
+  - weight-space-learning
+  - lora
+  - stable-diffusion
+---
+
+# Model-J: SD_1k - model_idx_4310
+
+This model is part of the **Model-J** dataset, introduced in:
+
+**Learning on Model Weights using Tree Experts** (CVPR 2025) by Eliahu Horwitz*, Bar Cavia*, Jonathan Kahana*, Yedid Hoshen
+
+<p align="center">
+    🌐 <a href="https://horwitz.ai/probex" target="_blank">Project</a> | 📃 <a href="https://arxiv.org/abs/2410.13569" target="_blank">Paper</a> | 💻 <a href="https://github.com/eliahuhorwitz/ProbeX" target="_blank">GitHub</a> | 🤗 <a href="https://huggingface.co/ProbeX" target="_blank">Dataset</a>
+</p>
+
+## Model Details
+
+| Attribute | Value |
+|---|---|
+| **Subset** | SD_1k |
+| **Split** | val_holdout |
+| **Base Model** | `CompVis/stable-diffusion-v1-4` |
+| **ImageNet Class** | flute (n03372029) |
+
+## Training Hyperparameters
+
+| Parameter | Value |
+|---|---|
+| Learning Rate | 7e-05 |
+| Max Train Steps | 450 |
+| LoRA Rank | 16 |
+| Seed | 0 |
+| N Training Samples | 6 |

val_holdout/model_idx_4311/README.md

@@ -0,0 +1,40 @@
+---
+base_model: CompVis/stable-diffusion-v1-3
+library_name: peft
+pipeline_tag: text-to-image
+tags:
+  - probex
+  - model-j
+  - weight-space-learning
+  - lora
+  - stable-diffusion
+---
+
+# Model-J: SD_1k - model_idx_4311
+
+This model is part of the **Model-J** dataset, introduced in:
+
+**Learning on Model Weights using Tree Experts** (CVPR 2025) by Eliahu Horwitz*, Bar Cavia*, Jonathan Kahana*, Yedid Hoshen
+
+<p align="center">
+    🌐 <a href="https://horwitz.ai/probex" target="_blank">Project</a> | 📃 <a href="https://arxiv.org/abs/2410.13569" target="_blank">Paper</a> | 💻 <a href="https://github.com/eliahuhorwitz/ProbeX" target="_blank">GitHub</a> | 🤗 <a href="https://huggingface.co/ProbeX" target="_blank">Dataset</a>
+</p>
+
+## Model Details
+
+| Attribute | Value |
+|---|---|
+| **Subset** | SD_1k |
+| **Split** | val_holdout |
+| **Base Model** | `CompVis/stable-diffusion-v1-3` |
+| **ImageNet Class** | robin (n01558993) |
+
+## Training Hyperparameters
+
+| Parameter | Value |
+|---|---|
+| Learning Rate | 3e-05 |
+| Max Train Steps | 550 |
+| LoRA Rank | 16 |
+| Seed | 74 |
+| N Training Samples | 6 |

val_holdout/model_idx_4312/README.md

@@ -0,0 +1,40 @@
+---
+base_model: CompVis/stable-diffusion-v1-4
+library_name: peft
+pipeline_tag: text-to-image
+tags:
+  - probex
+  - model-j
+  - weight-space-learning
+  - lora
+  - stable-diffusion
+---
+
+# Model-J: SD_1k - model_idx_4312
+
+This model is part of the **Model-J** dataset, introduced in:
+
+**Learning on Model Weights using Tree Experts** (CVPR 2025) by Eliahu Horwitz*, Bar Cavia*, Jonathan Kahana*, Yedid Hoshen
+
+<p align="center">
+    🌐 <a href="https://horwitz.ai/probex" target="_blank">Project</a> | 📃 <a href="https://arxiv.org/abs/2410.13569" target="_blank">Paper</a> | 💻 <a href="https://github.com/eliahuhorwitz/ProbeX" target="_blank">GitHub</a> | 🤗 <a href="https://huggingface.co/ProbeX" target="_blank">Dataset</a>
+</p>
+
+## Model Details
+
+| Attribute | Value |
+|---|---|
+| **Subset** | SD_1k |
+| **Split** | val_holdout |
+| **Base Model** | `CompVis/stable-diffusion-v1-4` |
+| **ImageNet Class** | steam_locomotive (n04310018) |
+
+## Training Hyperparameters
+
+| Parameter | Value |
+|---|---|
+| Learning Rate | 7e-05 |
+| Max Train Steps | 500 |
+| LoRA Rank | 16 |
+| Seed | 58 |
+| N Training Samples | 5 |

val_holdout/model_idx_4313/README.md

@@ -0,0 +1,40 @@
+---
+base_model: stable-diffusion-v1-5/stable-diffusion-v1-5
+library_name: peft
+pipeline_tag: text-to-image
+tags:
+  - probex
+  - model-j
+  - weight-space-learning
+  - lora
+  - stable-diffusion
+---
+
+# Model-J: SD_1k - model_idx_4313
+
+This model is part of the **Model-J** dataset, introduced in:
+
+**Learning on Model Weights using Tree Experts** (CVPR 2025) by Eliahu Horwitz*, Bar Cavia*, Jonathan Kahana*, Yedid Hoshen
+
+<p align="center">
+    🌐 <a href="https://horwitz.ai/probex" target="_blank">Project</a> | 📃 <a href="https://arxiv.org/abs/2410.13569" target="_blank">Paper</a> | 💻 <a href="https://github.com/eliahuhorwitz/ProbeX" target="_blank">GitHub</a> | 🤗 <a href="https://huggingface.co/ProbeX" target="_blank">Dataset</a>
+</p>
+
+## Model Details
+
+| Attribute | Value |
+|---|---|
+| **Subset** | SD_1k |
+| **Split** | val_holdout |
+| **Base Model** | `stable-diffusion-v1-5/stable-diffusion-v1-5` |
+| **ImageNet Class** | porcupine (n02346627) |
+
+## Training Hyperparameters
+
+| Parameter | Value |
+|---|---|
+| Learning Rate | 3e-05 |
+| Max Train Steps | 500 |
+| LoRA Rank | 16 |
+| Seed | 8 |
+| N Training Samples | 5 |

val_holdout/model_idx_4314/README.md

@@ -0,0 +1,40 @@
+---
+base_model: stable-diffusion-v1-5/stable-diffusion-v1-5
+library_name: peft
+pipeline_tag: text-to-image
+tags:
+  - probex
+  - model-j
+  - weight-space-learning
+  - lora
+  - stable-diffusion
+---
+
+# Model-J: SD_1k - model_idx_4314
+
+This model is part of the **Model-J** dataset, introduced in:
+
+**Learning on Model Weights using Tree Experts** (CVPR 2025) by Eliahu Horwitz*, Bar Cavia*, Jonathan Kahana*, Yedid Hoshen
+
+<p align="center">
+    🌐 <a href="https://horwitz.ai/probex" target="_blank">Project</a> | 📃 <a href="https://arxiv.org/abs/2410.13569" target="_blank">Paper</a> | 💻 <a href="https://github.com/eliahuhorwitz/ProbeX" target="_blank">GitHub</a> | 🤗 <a href="https://huggingface.co/ProbeX" target="_blank">Dataset</a>
+</p>
+
+## Model Details
+
+| Attribute | Value |
+|---|---|
+| **Subset** | SD_1k |
+| **Split** | val_holdout |
+| **Base Model** | `stable-diffusion-v1-5/stable-diffusion-v1-5` |
+| **ImageNet Class** | plow (n03967562) |
+
+## Training Hyperparameters
+
+| Parameter | Value |
+|---|---|
+| Learning Rate | 3e-05 |
+| Max Train Steps | 600 |
+| LoRA Rank | 16 |
+| Seed | 4 |
+| N Training Samples | 10 |

val_holdout/model_idx_4315/README.md

@@ -0,0 +1,40 @@
+---
+base_model: stable-diffusion-v1-5/stable-diffusion-v1-5
+library_name: peft
+pipeline_tag: text-to-image
+tags:
+  - probex
+  - model-j
+  - weight-space-learning
+  - lora
+  - stable-diffusion
+---
+
+# Model-J: SD_1k - model_idx_4315
+
+This model is part of the **Model-J** dataset, introduced in:
+
+**Learning on Model Weights using Tree Experts** (CVPR 2025) by Eliahu Horwitz*, Bar Cavia*, Jonathan Kahana*, Yedid Hoshen
+
+<p align="center">
+    🌐 <a href="https://horwitz.ai/probex" target="_blank">Project</a> | 📃 <a href="https://arxiv.org/abs/2410.13569" target="_blank">Paper</a> | 💻 <a href="https://github.com/eliahuhorwitz/ProbeX" target="_blank">GitHub</a> | 🤗 <a href="https://huggingface.co/ProbeX" target="_blank">Dataset</a>
+</p>
+
+## Model Details
+
+| Attribute | Value |
+|---|---|
+| **Subset** | SD_1k |
+| **Split** | val_holdout |
+| **Base Model** | `stable-diffusion-v1-5/stable-diffusion-v1-5` |
+| **ImageNet Class** | oxcart (n03868242) |
+
+## Training Hyperparameters
+
+| Parameter | Value |
+|---|---|
+| Learning Rate | 0.0001 |
+| Max Train Steps | 650 |
+| LoRA Rank | 16 |
+| Seed | 39 |
+| N Training Samples | 8 |

val_holdout/model_idx_4316/README.md

@@ -0,0 +1,40 @@
+---
+base_model: stable-diffusion-v1-5/stable-diffusion-v1-5
+library_name: peft
+pipeline_tag: text-to-image
+tags:
+  - probex
+  - model-j
+  - weight-space-learning
+  - lora
+  - stable-diffusion
+---
+
+# Model-J: SD_1k - model_idx_4316
+
+This model is part of the **Model-J** dataset, introduced in:
+
+**Learning on Model Weights using Tree Experts** (CVPR 2025) by Eliahu Horwitz*, Bar Cavia*, Jonathan Kahana*, Yedid Hoshen
+
+<p align="center">
+    🌐 <a href="https://horwitz.ai/probex" target="_blank">Project</a> | 📃 <a href="https://arxiv.org/abs/2410.13569" target="_blank">Paper</a> | 💻 <a href="https://github.com/eliahuhorwitz/ProbeX" target="_blank">GitHub</a> | 🤗 <a href="https://huggingface.co/ProbeX" target="_blank">Dataset</a>
+</p>
+
+## Model Details
+
+| Attribute | Value |
+|---|---|
+| **Subset** | SD_1k |
+| **Split** | val_holdout |
+| **Base Model** | `stable-diffusion-v1-5/stable-diffusion-v1-5` |
+| **ImageNet Class** | lacewing (n02264363) |
+
+## Training Hyperparameters
+
+| Parameter | Value |
+|---|---|
+| Learning Rate | 0.0005 |
+| Max Train Steps | 700 |
+| LoRA Rank | 16 |
+| Seed | 33 |
+| N Training Samples | 7 |

val_holdout/model_idx_4317/README.md

@@ -0,0 +1,40 @@
+---
+base_model: CompVis/stable-diffusion-v1-4
+library_name: peft
+pipeline_tag: text-to-image
+tags:
+  - probex
+  - model-j
+  - weight-space-learning
+  - lora
+  - stable-diffusion
+---
+
+# Model-J: SD_1k - model_idx_4317
+
+This model is part of the **Model-J** dataset, introduced in:
+
+**Learning on Model Weights using Tree Experts** (CVPR 2025) by Eliahu Horwitz*, Bar Cavia*, Jonathan Kahana*, Yedid Hoshen
+
+<p align="center">
+    🌐 <a href="https://horwitz.ai/probex" target="_blank">Project</a> | 📃 <a href="https://arxiv.org/abs/2410.13569" target="_blank">Paper</a> | 💻 <a href="https://github.com/eliahuhorwitz/ProbeX" target="_blank">GitHub</a> | 🤗 <a href="https://huggingface.co/ProbeX" target="_blank">Dataset</a>
+</p>
+
+## Model Details
+
+| Attribute | Value |
+|---|---|
+| **Subset** | SD_1k |
+| **Split** | val_holdout |
+| **Base Model** | `CompVis/stable-diffusion-v1-4` |
+| **ImageNet Class** | frilled_lizard (n01688243) |
+
+## Training Hyperparameters
+
+| Parameter | Value |
+|---|---|
+| Learning Rate | 0.0005 |
+| Max Train Steps | 450 |
+| LoRA Rank | 16 |
+| Seed | 53 |
+| N Training Samples | 8 |

val_holdout/model_idx_4318/README.md

@@ -0,0 +1,40 @@
+---
+base_model: CompVis/stable-diffusion-v1-3
+library_name: peft
+pipeline_tag: text-to-image
+tags:
+  - probex
+  - model-j
+  - weight-space-learning
+  - lora
+  - stable-diffusion
+---
+
+# Model-J: SD_1k - model_idx_4318
+
+This model is part of the **Model-J** dataset, introduced in:
+
+**Learning on Model Weights using Tree Experts** (CVPR 2025) by Eliahu Horwitz*, Bar Cavia*, Jonathan Kahana*, Yedid Hoshen
+
+<p align="center">
+    🌐 <a href="https://horwitz.ai/probex" target="_blank">Project</a> | 📃 <a href="https://arxiv.org/abs/2410.13569" target="_blank">Paper</a> | 💻 <a href="https://github.com/eliahuhorwitz/ProbeX" target="_blank">GitHub</a> | 🤗 <a href="https://huggingface.co/ProbeX" target="_blank">Dataset</a>
+</p>
+
+## Model Details
+
+| Attribute | Value |
+|---|---|
+| **Subset** | SD_1k |
+| **Split** | val_holdout |
+| **Base Model** | `CompVis/stable-diffusion-v1-3` |
+| **ImageNet Class** | Weimaraner (n02092339) |
+
+## Training Hyperparameters
+
+| Parameter | Value |
+|---|---|
+| Learning Rate | 3e-05 |
+| Max Train Steps | 700 |
+| LoRA Rank | 16 |
+| Seed | 17 |
+| N Training Samples | 8 |

val_holdout/model_idx_4319/README.md

@@ -0,0 +1,40 @@
+---
+base_model: CompVis/stable-diffusion-v1-3
+library_name: peft
+pipeline_tag: text-to-image
+tags:
+  - probex
+  - model-j
+  - weight-space-learning
+  - lora
+  - stable-diffusion
+---
+
+# Model-J: SD_1k - model_idx_4319
+
+This model is part of the **Model-J** dataset, introduced in:
+
+**Learning on Model Weights using Tree Experts** (CVPR 2025) by Eliahu Horwitz*, Bar Cavia*, Jonathan Kahana*, Yedid Hoshen
+
+<p align="center">
+    🌐 <a href="https://horwitz.ai/probex" target="_blank">Project</a> | 📃 <a href="https://arxiv.org/abs/2410.13569" target="_blank">Paper</a> | 💻 <a href="https://github.com/eliahuhorwitz/ProbeX" target="_blank">GitHub</a> | 🤗 <a href="https://huggingface.co/ProbeX" target="_blank">Dataset</a>
+</p>
+
+## Model Details
+
+| Attribute | Value |
+|---|---|
+| **Subset** | SD_1k |
+| **Split** | val_holdout |
+| **Base Model** | `CompVis/stable-diffusion-v1-3` |
+| **ImageNet Class** | Yorkshire_terrier (n02094433) |
+
+## Training Hyperparameters
+
+| Parameter | Value |
+|---|---|
+| Learning Rate | 9e-05 |
+| Max Train Steps | 450 |
+| LoRA Rank | 16 |
+| Seed | 96 |
+| N Training Samples | 8 |

val_holdout/model_idx_4320/README.md

@@ -0,0 +1,40 @@
+---
+base_model: CompVis/stable-diffusion-v1-2
+library_name: peft
+pipeline_tag: text-to-image
+tags:
+  - probex
+  - model-j
+  - weight-space-learning
+  - lora
+  - stable-diffusion
+---
+
+# Model-J: SD_1k - model_idx_4320
+
+This model is part of the **Model-J** dataset, introduced in:
+
+**Learning on Model Weights using Tree Experts** (CVPR 2025) by Eliahu Horwitz*, Bar Cavia*, Jonathan Kahana*, Yedid Hoshen
+
+<p align="center">
+    🌐 <a href="https://horwitz.ai/probex" target="_blank">Project</a> | 📃 <a href="https://arxiv.org/abs/2410.13569" target="_blank">Paper</a> | 💻 <a href="https://github.com/eliahuhorwitz/ProbeX" target="_blank">GitHub</a> | 🤗 <a href="https://huggingface.co/ProbeX" target="_blank">Dataset</a>
+</p>
+
+## Model Details
+
+| Attribute | Value |
+|---|---|
+| **Subset** | SD_1k |
+| **Split** | val_holdout |
+| **Base Model** | `CompVis/stable-diffusion-v1-2` |
+| **ImageNet Class** | warthog (n02397096) |
+
+## Training Hyperparameters
+
+| Parameter | Value |
+|---|---|
+| Learning Rate | 0.0005 |
+| Max Train Steps | 450 |
+| LoRA Rank | 16 |
+| Seed | 17 |
+| N Training Samples | 9 |

val_holdout/model_idx_4321/README.md

@@ -0,0 +1,40 @@
+---
+base_model: CompVis/stable-diffusion-v1-4
+library_name: peft
+pipeline_tag: text-to-image
+tags:
+  - probex
+  - model-j
+  - weight-space-learning
+  - lora
+  - stable-diffusion
+---
+
+# Model-J: SD_1k - model_idx_4321
+
+This model is part of the **Model-J** dataset, introduced in:
+
+**Learning on Model Weights using Tree Experts** (CVPR 2025) by Eliahu Horwitz*, Bar Cavia*, Jonathan Kahana*, Yedid Hoshen
+
+<p align="center">
+    🌐 <a href="https://horwitz.ai/probex" target="_blank">Project</a> | 📃 <a href="https://arxiv.org/abs/2410.13569" target="_blank">Paper</a> | 💻 <a href="https://github.com/eliahuhorwitz/ProbeX" target="_blank">GitHub</a> | 🤗 <a href="https://huggingface.co/ProbeX" target="_blank">Dataset</a>
+</p>
+
+## Model Details
+
+| Attribute | Value |
+|---|---|
+| **Subset** | SD_1k |
+| **Split** | val_holdout |
+| **Base Model** | `CompVis/stable-diffusion-v1-4` |
+| **ImageNet Class** | Border_terrier (n02093754) |
+
+## Training Hyperparameters
+
+| Parameter | Value |
+|---|---|
+| Learning Rate | 5e-05 |
+| Max Train Steps | 700 |
+| LoRA Rank | 16 |
+| Seed | 62 |
+| N Training Samples | 10 |

val_holdout/model_idx_4322/README.md

@@ -0,0 +1,40 @@
+---
+base_model: CompVis/stable-diffusion-v1-3
+library_name: peft
+pipeline_tag: text-to-image
+tags:
+  - probex
+  - model-j
+  - weight-space-learning
+  - lora
+  - stable-diffusion
+---
+
+# Model-J: SD_1k - model_idx_4322
+
+This model is part of the **Model-J** dataset, introduced in:
+
+**Learning on Model Weights using Tree Experts** (CVPR 2025) by Eliahu Horwitz*, Bar Cavia*, Jonathan Kahana*, Yedid Hoshen
+
+<p align="center">
+    🌐 <a href="https://horwitz.ai/probex" target="_blank">Project</a> | 📃 <a href="https://arxiv.org/abs/2410.13569" target="_blank">Paper</a> | 💻 <a href="https://github.com/eliahuhorwitz/ProbeX" target="_blank">GitHub</a> | 🤗 <a href="https://huggingface.co/ProbeX" target="_blank">Dataset</a>
+</p>
+
+## Model Details
+
+| Attribute | Value |
+|---|---|
+| **Subset** | SD_1k |
+| **Split** | val_holdout |
+| **Base Model** | `CompVis/stable-diffusion-v1-3` |
+| **ImageNet Class** | Dungeness_crab (n01978287) |
+
+## Training Hyperparameters
+
+| Parameter | Value |
+|---|---|
+| Learning Rate | 3e-05 |
+| Max Train Steps | 450 |
+| LoRA Rank | 16 |
+| Seed | 26 |
+| N Training Samples | 6 |

val_holdout/model_idx_4323/README.md

@@ -0,0 +1,40 @@
+---
+base_model: stable-diffusion-v1-5/stable-diffusion-v1-5
+library_name: peft
+pipeline_tag: text-to-image
+tags:
+  - probex
+  - model-j
+  - weight-space-learning
+  - lora
+  - stable-diffusion
+---
+
+# Model-J: SD_1k - model_idx_4323
+
+This model is part of the **Model-J** dataset, introduced in:
+
+**Learning on Model Weights using Tree Experts** (CVPR 2025) by Eliahu Horwitz*, Bar Cavia*, Jonathan Kahana*, Yedid Hoshen
+
+<p align="center">
+    🌐 <a href="https://horwitz.ai/probex" target="_blank">Project</a> | 📃 <a href="https://arxiv.org/abs/2410.13569" target="_blank">Paper</a> | 💻 <a href="https://github.com/eliahuhorwitz/ProbeX" target="_blank">GitHub</a> | 🤗 <a href="https://huggingface.co/ProbeX" target="_blank">Dataset</a>
+</p>
+
+## Model Details
+
+| Attribute | Value |
+|---|---|
+| **Subset** | SD_1k |
+| **Split** | val_holdout |
+| **Base Model** | `stable-diffusion-v1-5/stable-diffusion-v1-5` |
+| **ImageNet Class** | chiffonier (n03016953) |
+
+## Training Hyperparameters
+
+| Parameter | Value |
+|---|---|
+| Learning Rate | 0.0005 |
+| Max Train Steps | 500 |
+| LoRA Rank | 16 |
+| Seed | 75 |
+| N Training Samples | 10 |

val_holdout/model_idx_4324/README.md

@@ -0,0 +1,40 @@
+---
+base_model: CompVis/stable-diffusion-v1-2
+library_name: peft
+pipeline_tag: text-to-image
+tags:
+  - probex
+  - model-j
+  - weight-space-learning
+  - lora
+  - stable-diffusion
+---
+
+# Model-J: SD_1k - model_idx_4324
+
+This model is part of the **Model-J** dataset, introduced in:
+
+**Learning on Model Weights using Tree Experts** (CVPR 2025) by Eliahu Horwitz*, Bar Cavia*, Jonathan Kahana*, Yedid Hoshen
+
+<p align="center">
+    🌐 <a href="https://horwitz.ai/probex" target="_blank">Project</a> | 📃 <a href="https://arxiv.org/abs/2410.13569" target="_blank">Paper</a> | 💻 <a href="https://github.com/eliahuhorwitz/ProbeX" target="_blank">GitHub</a> | 🤗 <a href="https://huggingface.co/ProbeX" target="_blank">Dataset</a>
+</p>
+
+## Model Details
+
+| Attribute | Value |
+|---|---|
+| **Subset** | SD_1k |
+| **Split** | val_holdout |
+| **Base Model** | `CompVis/stable-diffusion-v1-2` |
+| **ImageNet Class** | toucan (n01843383) |
+
+## Training Hyperparameters
+
+| Parameter | Value |
+|---|---|
+| Learning Rate | 5e-05 |
+| Max Train Steps | 550 |
+| LoRA Rank | 16 |
+| Seed | 97 |
+| N Training Samples | 8 |

val_holdout/model_idx_4325/README.md

@@ -0,0 +1,40 @@
+---
+base_model: stable-diffusion-v1-5/stable-diffusion-v1-5
+library_name: peft
+pipeline_tag: text-to-image
+tags:
+  - probex
+  - model-j
+  - weight-space-learning
+  - lora
+  - stable-diffusion
+---
+
+# Model-J: SD_1k - model_idx_4325
+
+This model is part of the **Model-J** dataset, introduced in:
+
+**Learning on Model Weights using Tree Experts** (CVPR 2025) by Eliahu Horwitz*, Bar Cavia*, Jonathan Kahana*, Yedid Hoshen
+
+<p align="center">
+    🌐 <a href="https://horwitz.ai/probex" target="_blank">Project</a> | 📃 <a href="https://arxiv.org/abs/2410.13569" target="_blank">Paper</a> | 💻 <a href="https://github.com/eliahuhorwitz/ProbeX" target="_blank">GitHub</a> | 🤗 <a href="https://huggingface.co/ProbeX" target="_blank">Dataset</a>
+</p>
+
+## Model Details
+
+| Attribute | Value |
+|---|---|
+| **Subset** | SD_1k |
+| **Split** | val_holdout |
+| **Base Model** | `stable-diffusion-v1-5/stable-diffusion-v1-5` |
+| **ImageNet Class** | safe (n04125021) |
+
+## Training Hyperparameters
+
+| Parameter | Value |
+|---|---|
+| Learning Rate | 5e-05 |
+| Max Train Steps | 550 |
+| LoRA Rank | 16 |
+| Seed | 66 |
+| N Training Samples | 6 |

val_holdout/model_idx_4326/README.md

@@ -0,0 +1,40 @@
+---
+base_model: CompVis/stable-diffusion-v1-4
+library_name: peft
+pipeline_tag: text-to-image
+tags:
+  - probex
+  - model-j
+  - weight-space-learning
+  - lora
+  - stable-diffusion
+---
+
+# Model-J: SD_1k - model_idx_4326
+
+This model is part of the **Model-J** dataset, introduced in:
+
+**Learning on Model Weights using Tree Experts** (CVPR 2025) by Eliahu Horwitz*, Bar Cavia*, Jonathan Kahana*, Yedid Hoshen
+
+<p align="center">
+    🌐 <a href="https://horwitz.ai/probex" target="_blank">Project</a> | 📃 <a href="https://arxiv.org/abs/2410.13569" target="_blank">Paper</a> | 💻 <a href="https://github.com/eliahuhorwitz/ProbeX" target="_blank">GitHub</a> | 🤗 <a href="https://huggingface.co/ProbeX" target="_blank">Dataset</a>
+</p>
+
+## Model Details
+
+| Attribute | Value |
+|---|---|
+| **Subset** | SD_1k |
+| **Split** | val_holdout |
+| **Base Model** | `CompVis/stable-diffusion-v1-4` |
+| **ImageNet Class** | violin (n04536866) |
+
+## Training Hyperparameters
+
+| Parameter | Value |
+|---|---|
+| Learning Rate | 3e-05 |
+| Max Train Steps | 550 |
+| LoRA Rank | 16 |
+| Seed | 7 |
+| N Training Samples | 9 |

val_holdout/model_idx_4327/README.md

@@ -0,0 +1,40 @@
+---
+base_model: CompVis/stable-diffusion-v1-2
+library_name: peft
+pipeline_tag: text-to-image
+tags:
+  - probex
+  - model-j
+  - weight-space-learning
+  - lora
+  - stable-diffusion
+---
+
+# Model-J: SD_1k - model_idx_4327
+
+This model is part of the **Model-J** dataset, introduced in:
+
+**Learning on Model Weights using Tree Experts** (CVPR 2025) by Eliahu Horwitz*, Bar Cavia*, Jonathan Kahana*, Yedid Hoshen
+
+<p align="center">
+    🌐 <a href="https://horwitz.ai/probex" target="_blank">Project</a> | 📃 <a href="https://arxiv.org/abs/2410.13569" target="_blank">Paper</a> | 💻 <a href="https://github.com/eliahuhorwitz/ProbeX" target="_blank">GitHub</a> | 🤗 <a href="https://huggingface.co/ProbeX" target="_blank">Dataset</a>
+</p>
+
+## Model Details
+
+| Attribute | Value |
+|---|---|
+| **Subset** | SD_1k |
+| **Split** | val_holdout |
+| **Base Model** | `CompVis/stable-diffusion-v1-2` |
+| **ImageNet Class** | face_powder (n03314780) |
+
+## Training Hyperparameters
+
+| Parameter | Value |
+|---|---|
+| Learning Rate | 0.0003 |
+| Max Train Steps | 650 |
+| LoRA Rank | 16 |
+| Seed | 21 |
+| N Training Samples | 6 |

val_holdout/model_idx_4328/README.md

@@ -0,0 +1,40 @@
+---
+base_model: CompVis/stable-diffusion-v1-2
+library_name: peft
+pipeline_tag: text-to-image
+tags:
+  - probex
+  - model-j
+  - weight-space-learning
+  - lora
+  - stable-diffusion
+---
+
+# Model-J: SD_1k - model_idx_4328
+
+This model is part of the **Model-J** dataset, introduced in:
+
+**Learning on Model Weights using Tree Experts** (CVPR 2025) by Eliahu Horwitz*, Bar Cavia*, Jonathan Kahana*, Yedid Hoshen
+
+<p align="center">
+    🌐 <a href="https://horwitz.ai/probex" target="_blank">Project</a> | 📃 <a href="https://arxiv.org/abs/2410.13569" target="_blank">Paper</a> | 💻 <a href="https://github.com/eliahuhorwitz/ProbeX" target="_blank">GitHub</a> | 🤗 <a href="https://huggingface.co/ProbeX" target="_blank">Dataset</a>
+</p>
+
+## Model Details
+
+| Attribute | Value |
+|---|---|
+| **Subset** | SD_1k |
+| **Split** | val_holdout |
+| **Base Model** | `CompVis/stable-diffusion-v1-2` |
+| **ImageNet Class** | consomme (n07584110) |
+
+## Training Hyperparameters
+
+| Parameter | Value |
+|---|---|
+| Learning Rate | 7e-05 |
+| Max Train Steps | 450 |
+| LoRA Rank | 16 |
+| Seed | 34 |
+| N Training Samples | 10 |

val_holdout/model_idx_4329/README.md

@@ -0,0 +1,40 @@
+---
+base_model: CompVis/stable-diffusion-v1-3
+library_name: peft
+pipeline_tag: text-to-image
+tags:
+  - probex
+  - model-j
+  - weight-space-learning
+  - lora
+  - stable-diffusion
+---
+
+# Model-J: SD_1k - model_idx_4329
+
+This model is part of the **Model-J** dataset, introduced in:
+
+**Learning on Model Weights using Tree Experts** (CVPR 2025) by Eliahu Horwitz*, Bar Cavia*, Jonathan Kahana*, Yedid Hoshen
+
+<p align="center">
+    🌐 <a href="https://horwitz.ai/probex" target="_blank">Project</a> | 📃 <a href="https://arxiv.org/abs/2410.13569" target="_blank">Paper</a> | 💻 <a href="https://github.com/eliahuhorwitz/ProbeX" target="_blank">GitHub</a> | 🤗 <a href="https://huggingface.co/ProbeX" target="_blank">Dataset</a>
+</p>
+
+## Model Details
+
+| Attribute | Value |
+|---|---|
+| **Subset** | SD_1k |
+| **Split** | val_holdout |
+| **Base Model** | `CompVis/stable-diffusion-v1-3` |
+| **ImageNet Class** | minivan (n03770679) |
+
+## Training Hyperparameters
+
+| Parameter | Value |
+|---|---|
+| Learning Rate | 0.0003 |
+| Max Train Steps | 700 |
+| LoRA Rank | 16 |
+| Seed | 9 |
+| N Training Samples | 8 |

val_holdout/model_idx_4330/README.md

@@ -0,0 +1,40 @@
+---
+base_model: CompVis/stable-diffusion-v1-4
+library_name: peft
+pipeline_tag: text-to-image
+tags:
+  - probex
+  - model-j
+  - weight-space-learning
+  - lora
+  - stable-diffusion
+---
+
+# Model-J: SD_1k - model_idx_4330
+
+This model is part of the **Model-J** dataset, introduced in:
+
+**Learning on Model Weights using Tree Experts** (CVPR 2025) by Eliahu Horwitz*, Bar Cavia*, Jonathan Kahana*, Yedid Hoshen
+
+<p align="center">
+    🌐 <a href="https://horwitz.ai/probex" target="_blank">Project</a> | 📃 <a href="https://arxiv.org/abs/2410.13569" target="_blank">Paper</a> | 💻 <a href="https://github.com/eliahuhorwitz/ProbeX" target="_blank">GitHub</a> | 🤗 <a href="https://huggingface.co/ProbeX" target="_blank">Dataset</a>
+</p>
+
+## Model Details
+
+| Attribute | Value |
+|---|---|
+| **Subset** | SD_1k |
+| **Split** | val_holdout |
+| **Base Model** | `CompVis/stable-diffusion-v1-4` |
+| **ImageNet Class** | tractor (n04465501) |
+
+## Training Hyperparameters
+
+| Parameter | Value |
+|---|---|
+| Learning Rate | 0.0001 |
+| Max Train Steps | 550 |
+| LoRA Rank | 16 |
+| Seed | 17 |
+| N Training Samples | 8 |

val_holdout/model_idx_4331/README.md

@@ -0,0 +1,40 @@
+---
+base_model: CompVis/stable-diffusion-v1-4
+library_name: peft
+pipeline_tag: text-to-image
+tags:
+  - probex
+  - model-j
+  - weight-space-learning
+  - lora
+  - stable-diffusion
+---
+
+# Model-J: SD_1k - model_idx_4331
+
+This model is part of the **Model-J** dataset, introduced in:
+
+**Learning on Model Weights using Tree Experts** (CVPR 2025) by Eliahu Horwitz*, Bar Cavia*, Jonathan Kahana*, Yedid Hoshen
+
+<p align="center">
+    🌐 <a href="https://horwitz.ai/probex" target="_blank">Project</a> | 📃 <a href="https://arxiv.org/abs/2410.13569" target="_blank">Paper</a> | 💻 <a href="https://github.com/eliahuhorwitz/ProbeX" target="_blank">GitHub</a> | 🤗 <a href="https://huggingface.co/ProbeX" target="_blank">Dataset</a>
+</p>
+
+## Model Details
+
+| Attribute | Value |
+|---|---|
+| **Subset** | SD_1k |
+| **Split** | val_holdout |
+| **Base Model** | `CompVis/stable-diffusion-v1-4` |
+| **ImageNet Class** | library (n03661043) |
+
+## Training Hyperparameters
+
+| Parameter | Value |
+|---|---|
+| Learning Rate | 9e-05 |
+| Max Train Steps | 650 |
+| LoRA Rank | 16 |
+| Seed | 38 |
+| N Training Samples | 7 |

val_holdout/model_idx_4332/README.md

@@ -0,0 +1,40 @@
+---
+base_model: CompVis/stable-diffusion-v1-3
+library_name: peft
+pipeline_tag: text-to-image
+tags:
+  - probex
+  - model-j
+  - weight-space-learning
+  - lora
+  - stable-diffusion
+---
+
+# Model-J: SD_1k - model_idx_4332
+
+This model is part of the **Model-J** dataset, introduced in:
+
+**Learning on Model Weights using Tree Experts** (CVPR 2025) by Eliahu Horwitz*, Bar Cavia*, Jonathan Kahana*, Yedid Hoshen
+
+<p align="center">
+    🌐 <a href="https://horwitz.ai/probex" target="_blank">Project</a> | 📃 <a href="https://arxiv.org/abs/2410.13569" target="_blank">Paper</a> | 💻 <a href="https://github.com/eliahuhorwitz/ProbeX" target="_blank">GitHub</a> | 🤗 <a href="https://huggingface.co/ProbeX" target="_blank">Dataset</a>
+</p>
+
+## Model Details
+
+| Attribute | Value |
+|---|---|
+| **Subset** | SD_1k |
+| **Split** | val_holdout |
+| **Base Model** | `CompVis/stable-diffusion-v1-3` |
+| **ImageNet Class** | basenji (n02110806) |
+
+## Training Hyperparameters
+
+| Parameter | Value |
+|---|---|
+| Learning Rate | 0.0005 |
+| Max Train Steps | 650 |
+| LoRA Rank | 16 |
+| Seed | 71 |
+| N Training Samples | 10 |

val_holdout/model_idx_4333/README.md

@@ -0,0 +1,40 @@
+---
+base_model: stable-diffusion-v1-5/stable-diffusion-v1-5
+library_name: peft
+pipeline_tag: text-to-image
+tags:
+  - probex
+  - model-j
+  - weight-space-learning
+  - lora
+  - stable-diffusion
+---
+
+# Model-J: SD_1k - model_idx_4333
+
+This model is part of the **Model-J** dataset, introduced in:
+
+**Learning on Model Weights using Tree Experts** (CVPR 2025) by Eliahu Horwitz*, Bar Cavia*, Jonathan Kahana*, Yedid Hoshen
+
+<p align="center">
+    🌐 <a href="https://horwitz.ai/probex" target="_blank">Project</a> | 📃 <a href="https://arxiv.org/abs/2410.13569" target="_blank">Paper</a> | 💻 <a href="https://github.com/eliahuhorwitz/ProbeX" target="_blank">GitHub</a> | 🤗 <a href="https://huggingface.co/ProbeX" target="_blank">Dataset</a>
+</p>
+
+## Model Details
+
+| Attribute | Value |
+|---|---|
+| **Subset** | SD_1k |
+| **Split** | val_holdout |
+| **Base Model** | `stable-diffusion-v1-5/stable-diffusion-v1-5` |
+| **ImageNet Class** | desk (n03179701) |
+
+## Training Hyperparameters
+
+| Parameter | Value |
+|---|---|
+| Learning Rate | 0.0005 |
+| Max Train Steps | 450 |
+| LoRA Rank | 16 |
+| Seed | 55 |
+| N Training Samples | 9 |

val_holdout/model_idx_4334/README.md

@@ -0,0 +1,40 @@
+---
+base_model: stable-diffusion-v1-5/stable-diffusion-v1-5
+library_name: peft
+pipeline_tag: text-to-image
+tags:
+  - probex
+  - model-j
+  - weight-space-learning
+  - lora
+  - stable-diffusion
+---
+
+# Model-J: SD_1k - model_idx_4334
+
+This model is part of the **Model-J** dataset, introduced in:
+
+**Learning on Model Weights using Tree Experts** (CVPR 2025) by Eliahu Horwitz*, Bar Cavia*, Jonathan Kahana*, Yedid Hoshen
+
+<p align="center">
+    🌐 <a href="https://horwitz.ai/probex" target="_blank">Project</a> | 📃 <a href="https://arxiv.org/abs/2410.13569" target="_blank">Paper</a> | 💻 <a href="https://github.com/eliahuhorwitz/ProbeX" target="_blank">GitHub</a> | 🤗 <a href="https://huggingface.co/ProbeX" target="_blank">Dataset</a>
+</p>
+
+## Model Details
+
+| Attribute | Value |
+|---|---|
+| **Subset** | SD_1k |
+| **Split** | val_holdout |
+| **Base Model** | `stable-diffusion-v1-5/stable-diffusion-v1-5` |
+| **ImageNet Class** | Rhodesian_ridgeback (n02087394) |
+
+## Training Hyperparameters
+
+| Parameter | Value |
+|---|---|
+| Learning Rate | 0.0001 |
+| Max Train Steps | 600 |
+| LoRA Rank | 16 |
+| Seed | 44 |
+| N Training Samples | 8 |

val_holdout/model_idx_4335/README.md

@@ -0,0 +1,40 @@
+---
+base_model: CompVis/stable-diffusion-v1-2
+library_name: peft
+pipeline_tag: text-to-image
+tags:
+  - probex
+  - model-j
+  - weight-space-learning
+  - lora
+  - stable-diffusion
+---
+
+# Model-J: SD_1k - model_idx_4335
+
+This model is part of the **Model-J** dataset, introduced in:
+
+**Learning on Model Weights using Tree Experts** (CVPR 2025) by Eliahu Horwitz*, Bar Cavia*, Jonathan Kahana*, Yedid Hoshen
+
+<p align="center">
+    🌐 <a href="https://horwitz.ai/probex" target="_blank">Project</a> | 📃 <a href="https://arxiv.org/abs/2410.13569" target="_blank">Paper</a> | 💻 <a href="https://github.com/eliahuhorwitz/ProbeX" target="_blank">GitHub</a> | 🤗 <a href="https://huggingface.co/ProbeX" target="_blank">Dataset</a>
+</p>
+
+## Model Details
+
+| Attribute | Value |
+|---|---|
+| **Subset** | SD_1k |
+| **Split** | val_holdout |
+| **Base Model** | `CompVis/stable-diffusion-v1-2` |
+| **ImageNet Class** | lighter (n03666591) |
+
+## Training Hyperparameters
+
+| Parameter | Value |
+|---|---|
+| Learning Rate | 0.0003 |
+| Max Train Steps | 500 |
+| LoRA Rank | 16 |
+| Seed | 91 |
+| N Training Samples | 9 |

val_holdout/model_idx_4336/README.md

@@ -0,0 +1,40 @@
+---
+base_model: CompVis/stable-diffusion-v1-4
+library_name: peft
+pipeline_tag: text-to-image
+tags:
+  - probex
+  - model-j
+  - weight-space-learning
+  - lora
+  - stable-diffusion
+---
+
+# Model-J: SD_1k - model_idx_4336
+
+This model is part of the **Model-J** dataset, introduced in:
+
+**Learning on Model Weights using Tree Experts** (CVPR 2025) by Eliahu Horwitz*, Bar Cavia*, Jonathan Kahana*, Yedid Hoshen
+
+<p align="center">
+    🌐 <a href="https://horwitz.ai/probex" target="_blank">Project</a> | 📃 <a href="https://arxiv.org/abs/2410.13569" target="_blank">Paper</a> | 💻 <a href="https://github.com/eliahuhorwitz/ProbeX" target="_blank">GitHub</a> | 🤗 <a href="https://huggingface.co/ProbeX" target="_blank">Dataset</a>
+</p>
+
+## Model Details
+
+| Attribute | Value |
+|---|---|
+| **Subset** | SD_1k |
+| **Split** | val_holdout |
+| **Base Model** | `CompVis/stable-diffusion-v1-4` |
+| **ImageNet Class** | orangutan (n02480495) |
+
+## Training Hyperparameters
+
+| Parameter | Value |
+|---|---|
+| Learning Rate | 0.0005 |
+| Max Train Steps | 700 |
+| LoRA Rank | 16 |
+| Seed | 98 |
+| N Training Samples | 6 |

val_holdout/model_idx_4337/README.md

@@ -0,0 +1,40 @@
+---
+base_model: CompVis/stable-diffusion-v1-2
+library_name: peft
+pipeline_tag: text-to-image
+tags:
+  - probex
+  - model-j
+  - weight-space-learning
+  - lora
+  - stable-diffusion
+---
+
+# Model-J: SD_1k - model_idx_4337
+
+This model is part of the **Model-J** dataset, introduced in:
+
+**Learning on Model Weights using Tree Experts** (CVPR 2025) by Eliahu Horwitz*, Bar Cavia*, Jonathan Kahana*, Yedid Hoshen
+
+<p align="center">
+    🌐 <a href="https://horwitz.ai/probex" target="_blank">Project</a> | 📃 <a href="https://arxiv.org/abs/2410.13569" target="_blank">Paper</a> | 💻 <a href="https://github.com/eliahuhorwitz/ProbeX" target="_blank">GitHub</a> | 🤗 <a href="https://huggingface.co/ProbeX" target="_blank">Dataset</a>
+</p>
+
+## Model Details
+
+| Attribute | Value |
+|---|---|
+| **Subset** | SD_1k |
+| **Split** | val_holdout |
+| **Base Model** | `CompVis/stable-diffusion-v1-2` |
+| **ImageNet Class** | tobacco_shop (n04443257) |
+
+## Training Hyperparameters
+
+| Parameter | Value |
+|---|---|
+| Learning Rate | 7e-05 |
+| Max Train Steps | 650 |
+| LoRA Rank | 16 |
+| Seed | 9 |
+| N Training Samples | 8 |

val_holdout/model_idx_4338/README.md

@@ -0,0 +1,40 @@
+---
+base_model: stable-diffusion-v1-5/stable-diffusion-v1-5
+library_name: peft
+pipeline_tag: text-to-image
+tags:
+  - probex
+  - model-j
+  - weight-space-learning
+  - lora
+  - stable-diffusion
+---
+
+# Model-J: SD_1k - model_idx_4338
+
+This model is part of the **Model-J** dataset, introduced in:
+
+**Learning on Model Weights using Tree Experts** (CVPR 2025) by Eliahu Horwitz*, Bar Cavia*, Jonathan Kahana*, Yedid Hoshen
+
+<p align="center">
+    🌐 <a href="https://horwitz.ai/probex" target="_blank">Project</a> | 📃 <a href="https://arxiv.org/abs/2410.13569" target="_blank">Paper</a> | 💻 <a href="https://github.com/eliahuhorwitz/ProbeX" target="_blank">GitHub</a> | 🤗 <a href="https://huggingface.co/ProbeX" target="_blank">Dataset</a>
+</p>
+
+## Model Details
+
+| Attribute | Value |
+|---|---|
+| **Subset** | SD_1k |
+| **Split** | val_holdout |
+| **Base Model** | `stable-diffusion-v1-5/stable-diffusion-v1-5` |
+| **ImageNet Class** | otterhound (n02091635) |
+
+## Training Hyperparameters
+
+| Parameter | Value |
+|---|---|
+| Learning Rate | 0.0001 |
+| Max Train Steps | 500 |
+| LoRA Rank | 16 |
+| Seed | 44 |
+| N Training Samples | 5 |

val_holdout/model_idx_4339/README.md

@@ -0,0 +1,40 @@
+---
+base_model: CompVis/stable-diffusion-v1-4
+library_name: peft
+pipeline_tag: text-to-image
+tags:
+  - probex
+  - model-j
+  - weight-space-learning
+  - lora
+  - stable-diffusion
+---
+
+# Model-J: SD_1k - model_idx_4339
+
+This model is part of the **Model-J** dataset, introduced in:
+
+**Learning on Model Weights using Tree Experts** (CVPR 2025) by Eliahu Horwitz*, Bar Cavia*, Jonathan Kahana*, Yedid Hoshen
+
+<p align="center">
+    🌐 <a href="https://horwitz.ai/probex" target="_blank">Project</a> | 📃 <a href="https://arxiv.org/abs/2410.13569" target="_blank">Paper</a> | 💻 <a href="https://github.com/eliahuhorwitz/ProbeX" target="_blank">GitHub</a> | 🤗 <a href="https://huggingface.co/ProbeX" target="_blank">Dataset</a>
+</p>
+
+## Model Details
+
+| Attribute | Value |
+|---|---|
+| **Subset** | SD_1k |
+| **Split** | val_holdout |
+| **Base Model** | `CompVis/stable-diffusion-v1-4` |
+| **ImageNet Class** | cricket (n02229544) |
+
+## Training Hyperparameters
+
+| Parameter | Value |
+|---|---|
+| Learning Rate | 0.0005 |
+| Max Train Steps | 450 |
+| LoRA Rank | 16 |
+| Seed | 85 |
+| N Training Samples | 9 |

val_holdout/model_idx_4340/README.md

@@ -0,0 +1,40 @@
+---
+base_model: CompVis/stable-diffusion-v1-2
+library_name: peft
+pipeline_tag: text-to-image
+tags:
+  - probex
+  - model-j
+  - weight-space-learning
+  - lora
+  - stable-diffusion
+---
+
+# Model-J: SD_1k - model_idx_4340
+
+This model is part of the **Model-J** dataset, introduced in:
+
+**Learning on Model Weights using Tree Experts** (CVPR 2025) by Eliahu Horwitz*, Bar Cavia*, Jonathan Kahana*, Yedid Hoshen
+
+<p align="center">
+    🌐 <a href="https://horwitz.ai/probex" target="_blank">Project</a> | 📃 <a href="https://arxiv.org/abs/2410.13569" target="_blank">Paper</a> | 💻 <a href="https://github.com/eliahuhorwitz/ProbeX" target="_blank">GitHub</a> | 🤗 <a href="https://huggingface.co/ProbeX" target="_blank">Dataset</a>
+</p>
+
+## Model Details
+
+| Attribute | Value |
+|---|---|
+| **Subset** | SD_1k |
+| **Split** | val_holdout |
+| **Base Model** | `CompVis/stable-diffusion-v1-2` |
+| **ImageNet Class** | pajama (n03877472) |
+
+## Training Hyperparameters
+
+| Parameter | Value |
+|---|---|
+| Learning Rate | 7e-05 |
+| Max Train Steps | 550 |
+| LoRA Rank | 16 |
+| Seed | 79 |
+| N Training Samples | 7 |

val_holdout/model_idx_4341/README.md

@@ -0,0 +1,40 @@
+---
+base_model: CompVis/stable-diffusion-v1-3
+library_name: peft
+pipeline_tag: text-to-image
+tags:
+  - probex
+  - model-j
+  - weight-space-learning
+  - lora
+  - stable-diffusion
+---
+
+# Model-J: SD_1k - model_idx_4341
+
+This model is part of the **Model-J** dataset, introduced in:
+
+**Learning on Model Weights using Tree Experts** (CVPR 2025) by Eliahu Horwitz*, Bar Cavia*, Jonathan Kahana*, Yedid Hoshen
+
+<p align="center">
+    🌐 <a href="https://horwitz.ai/probex" target="_blank">Project</a> | 📃 <a href="https://arxiv.org/abs/2410.13569" target="_blank">Paper</a> | 💻 <a href="https://github.com/eliahuhorwitz/ProbeX" target="_blank">GitHub</a> | 🤗 <a href="https://huggingface.co/ProbeX" target="_blank">Dataset</a>
+</p>
+
+## Model Details
+
+| Attribute | Value |
+|---|---|
+| **Subset** | SD_1k |
+| **Split** | val_holdout |
+| **Base Model** | `CompVis/stable-diffusion-v1-3` |
+| **ImageNet Class** | cabbage_butterfly (n02280649) |
+
+## Training Hyperparameters
+
+| Parameter | Value |
+|---|---|
+| Learning Rate | 0.0003 |
+| Max Train Steps | 650 |
+| LoRA Rank | 16 |
+| Seed | 78 |
+| N Training Samples | 6 |

val_holdout/model_idx_4342/README.md

@@ -0,0 +1,40 @@
+---
+base_model: CompVis/stable-diffusion-v1-2
+library_name: peft
+pipeline_tag: text-to-image
+tags:
+  - probex
+  - model-j
+  - weight-space-learning
+  - lora
+  - stable-diffusion
+---
+
+# Model-J: SD_1k - model_idx_4342
+
+This model is part of the **Model-J** dataset, introduced in:
+
+**Learning on Model Weights using Tree Experts** (CVPR 2025) by Eliahu Horwitz*, Bar Cavia*, Jonathan Kahana*, Yedid Hoshen
+
+<p align="center">
+    🌐 <a href="https://horwitz.ai/probex" target="_blank">Project</a> | 📃 <a href="https://arxiv.org/abs/2410.13569" target="_blank">Paper</a> | 💻 <a href="https://github.com/eliahuhorwitz/ProbeX" target="_blank">GitHub</a> | 🤗 <a href="https://huggingface.co/ProbeX" target="_blank">Dataset</a>
+</p>
+
+## Model Details
+
+| Attribute | Value |
+|---|---|
+| **Subset** | SD_1k |
+| **Split** | val_holdout |
+| **Base Model** | `CompVis/stable-diffusion-v1-2` |
+| **ImageNet Class** | shoji (n04201297) |
+
+## Training Hyperparameters
+
+| Parameter | Value |
+|---|---|
+| Learning Rate | 5e-05 |
+| Max Train Steps | 500 |
+| LoRA Rank | 16 |
+| Seed | 35 |
+| N Training Samples | 10 |

val_holdout/model_idx_4343/README.md

@@ -0,0 +1,40 @@
+---
+base_model: CompVis/stable-diffusion-v1-2
+library_name: peft
+pipeline_tag: text-to-image
+tags:
+  - probex
+  - model-j
+  - weight-space-learning
+  - lora
+  - stable-diffusion
+---
+
+# Model-J: SD_1k - model_idx_4343
+
+This model is part of the **Model-J** dataset, introduced in:
+
+**Learning on Model Weights using Tree Experts** (CVPR 2025) by Eliahu Horwitz*, Bar Cavia*, Jonathan Kahana*, Yedid Hoshen
+
+<p align="center">
+    🌐 <a href="https://horwitz.ai/probex" target="_blank">Project</a> | 📃 <a href="https://arxiv.org/abs/2410.13569" target="_blank">Paper</a> | 💻 <a href="https://github.com/eliahuhorwitz/ProbeX" target="_blank">GitHub</a> | 🤗 <a href="https://huggingface.co/ProbeX" target="_blank">Dataset</a>
+</p>
+
+## Model Details
+
+| Attribute | Value |
+|---|---|
+| **Subset** | SD_1k |
+| **Split** | val_holdout |
+| **Base Model** | `CompVis/stable-diffusion-v1-2` |
+| **ImageNet Class** | parallel_bars (n03888605) |
+
+## Training Hyperparameters
+
+| Parameter | Value |
+|---|---|
+| Learning Rate | 0.0001 |
+| Max Train Steps | 450 |
+| LoRA Rank | 16 |
+| Seed | 84 |
+| N Training Samples | 9 |

val_holdout/model_idx_4344/README.md

@@ -0,0 +1,40 @@
+---
+base_model: CompVis/stable-diffusion-v1-3
+library_name: peft
+pipeline_tag: text-to-image
+tags:
+  - probex
+  - model-j
+  - weight-space-learning
+  - lora
+  - stable-diffusion
+---
+
+# Model-J: SD_1k - model_idx_4344
+
+This model is part of the **Model-J** dataset, introduced in:
+
+**Learning on Model Weights using Tree Experts** (CVPR 2025) by Eliahu Horwitz*, Bar Cavia*, Jonathan Kahana*, Yedid Hoshen
+
+<p align="center">
+    🌐 <a href="https://horwitz.ai/probex" target="_blank">Project</a> | 📃 <a href="https://arxiv.org/abs/2410.13569" target="_blank">Paper</a> | 💻 <a href="https://github.com/eliahuhorwitz/ProbeX" target="_blank">GitHub</a> | 🤗 <a href="https://huggingface.co/ProbeX" target="_blank">Dataset</a>
+</p>
+
+## Model Details
+
+| Attribute | Value |
+|---|---|
+| **Subset** | SD_1k |
+| **Split** | val_holdout |
+| **Base Model** | `CompVis/stable-diffusion-v1-3` |
+| **ImageNet Class** | chest (n03014705) |
+
+## Training Hyperparameters
+
+| Parameter | Value |
+|---|---|
+| Learning Rate | 5e-05 |
+| Max Train Steps | 550 |
+| LoRA Rank | 16 |
+| Seed | 47 |
+| N Training Samples | 9 |

val_holdout/model_idx_4345/README.md

@@ -0,0 +1,40 @@
+---
+base_model: CompVis/stable-diffusion-v1-4
+library_name: peft
+pipeline_tag: text-to-image
+tags:
+  - probex
+  - model-j
+  - weight-space-learning
+  - lora
+  - stable-diffusion
+---
+
+# Model-J: SD_1k - model_idx_4345
+
+This model is part of the **Model-J** dataset, introduced in:
+
+**Learning on Model Weights using Tree Experts** (CVPR 2025) by Eliahu Horwitz*, Bar Cavia*, Jonathan Kahana*, Yedid Hoshen
+
+<p align="center">
+    🌐 <a href="https://horwitz.ai/probex" target="_blank">Project</a> | 📃 <a href="https://arxiv.org/abs/2410.13569" target="_blank">Paper</a> | 💻 <a href="https://github.com/eliahuhorwitz/ProbeX" target="_blank">GitHub</a> | 🤗 <a href="https://huggingface.co/ProbeX" target="_blank">Dataset</a>
+</p>
+
+## Model Details
+
+| Attribute | Value |
+|---|---|
+| **Subset** | SD_1k |
+| **Split** | val_holdout |
+| **Base Model** | `CompVis/stable-diffusion-v1-4` |
+| **ImageNet Class** | sulphur_butterfly (n02281406) |
+
+## Training Hyperparameters
+
+| Parameter | Value |
+|---|---|
+| Learning Rate | 9e-05 |
+| Max Train Steps | 600 |
+| LoRA Rank | 16 |
+| Seed | 98 |
+| N Training Samples | 5 |

val_holdout/model_idx_4346/README.md

@@ -0,0 +1,40 @@
+---
+base_model: CompVis/stable-diffusion-v1-2
+library_name: peft
+pipeline_tag: text-to-image
+tags:
+  - probex
+  - model-j
+  - weight-space-learning
+  - lora
+  - stable-diffusion
+---
+
+# Model-J: SD_1k - model_idx_4346
+
+This model is part of the **Model-J** dataset, introduced in:
+
+**Learning on Model Weights using Tree Experts** (CVPR 2025) by Eliahu Horwitz*, Bar Cavia*, Jonathan Kahana*, Yedid Hoshen
+
+<p align="center">
+    🌐 <a href="https://horwitz.ai/probex" target="_blank">Project</a> | 📃 <a href="https://arxiv.org/abs/2410.13569" target="_blank">Paper</a> | 💻 <a href="https://github.com/eliahuhorwitz/ProbeX" target="_blank">GitHub</a> | 🤗 <a href="https://huggingface.co/ProbeX" target="_blank">Dataset</a>
+</p>
+
+## Model Details
+
+| Attribute | Value |
+|---|---|
+| **Subset** | SD_1k |
+| **Split** | val_holdout |
+| **Base Model** | `CompVis/stable-diffusion-v1-2` |
+| **ImageNet Class** | sea_urchin (n02319095) |
+
+## Training Hyperparameters
+
+| Parameter | Value |
+|---|---|
+| Learning Rate | 0.0001 |
+| Max Train Steps | 650 |
+| LoRA Rank | 16 |
+| Seed | 23 |
+| N Training Samples | 7 |

val_holdout/model_idx_4347/README.md

@@ -0,0 +1,40 @@
+---
+base_model: CompVis/stable-diffusion-v1-4
+library_name: peft
+pipeline_tag: text-to-image
+tags:
+  - probex
+  - model-j
+  - weight-space-learning
+  - lora
+  - stable-diffusion
+---
+
+# Model-J: SD_1k - model_idx_4347
+
+This model is part of the **Model-J** dataset, introduced in:
+
+**Learning on Model Weights using Tree Experts** (CVPR 2025) by Eliahu Horwitz*, Bar Cavia*, Jonathan Kahana*, Yedid Hoshen
+
+<p align="center">
+    🌐 <a href="https://horwitz.ai/probex" target="_blank">Project</a> | 📃 <a href="https://arxiv.org/abs/2410.13569" target="_blank">Paper</a> | 💻 <a href="https://github.com/eliahuhorwitz/ProbeX" target="_blank">GitHub</a> | 🤗 <a href="https://huggingface.co/ProbeX" target="_blank">Dataset</a>
+</p>
+
+## Model Details
+
+| Attribute | Value |
+|---|---|
+| **Subset** | SD_1k |
+| **Split** | val_holdout |
+| **Base Model** | `CompVis/stable-diffusion-v1-4` |
+| **ImageNet Class** | balance_beam (n02777292) |
+
+## Training Hyperparameters
+
+| Parameter | Value |
+|---|---|
+| Learning Rate | 0.0005 |
+| Max Train Steps | 700 |
+| LoRA Rank | 16 |
+| Seed | 45 |
+| N Training Samples | 6 |

val_holdout/model_idx_4348/README.md

@@ -0,0 +1,40 @@
+---
+base_model: CompVis/stable-diffusion-v1-3
+library_name: peft
+pipeline_tag: text-to-image
+tags:
+  - probex
+  - model-j
+  - weight-space-learning
+  - lora
+  - stable-diffusion
+---
+
+# Model-J: SD_1k - model_idx_4348
+
+This model is part of the **Model-J** dataset, introduced in:
+
+**Learning on Model Weights using Tree Experts** (CVPR 2025) by Eliahu Horwitz*, Bar Cavia*, Jonathan Kahana*, Yedid Hoshen
+
+<p align="center">
+    🌐 <a href="https://horwitz.ai/probex" target="_blank">Project</a> | 📃 <a href="https://arxiv.org/abs/2410.13569" target="_blank">Paper</a> | 💻 <a href="https://github.com/eliahuhorwitz/ProbeX" target="_blank">GitHub</a> | 🤗 <a href="https://huggingface.co/ProbeX" target="_blank">Dataset</a>
+</p>
+
+## Model Details
+
+| Attribute | Value |
+|---|---|
+| **Subset** | SD_1k |
+| **Split** | val_holdout |
+| **Base Model** | `CompVis/stable-diffusion-v1-3` |
+| **ImageNet Class** | screen (n04152593) |
+
+## Training Hyperparameters
+
+| Parameter | Value |
+|---|---|
+| Learning Rate | 5e-05 |
+| Max Train Steps | 650 |
+| LoRA Rank | 16 |
+| Seed | 30 |
+| N Training Samples | 5 |

val_holdout/model_idx_4349/README.md

@@ -0,0 +1,40 @@
+---
+base_model: CompVis/stable-diffusion-v1-2
+library_name: peft
+pipeline_tag: text-to-image
+tags:
+  - probex
+  - model-j
+  - weight-space-learning
+  - lora
+  - stable-diffusion
+---
+
+# Model-J: SD_1k - model_idx_4349
+
+This model is part of the **Model-J** dataset, introduced in:
+
+**Learning on Model Weights using Tree Experts** (CVPR 2025) by Eliahu Horwitz*, Bar Cavia*, Jonathan Kahana*, Yedid Hoshen
+
+<p align="center">
+    🌐 <a href="https://horwitz.ai/probex" target="_blank">Project</a> | 📃 <a href="https://arxiv.org/abs/2410.13569" target="_blank">Paper</a> | 💻 <a href="https://github.com/eliahuhorwitz/ProbeX" target="_blank">GitHub</a> | 🤗 <a href="https://huggingface.co/ProbeX" target="_blank">Dataset</a>
+</p>
+
+## Model Details
+
+| Attribute | Value |
+|---|---|
+| **Subset** | SD_1k |
+| **Split** | val_holdout |
+| **Base Model** | `CompVis/stable-diffusion-v1-2` |
+| **ImageNet Class** | forklift (n03384352) |
+
+## Training Hyperparameters
+
+| Parameter | Value |
+|---|---|
+| Learning Rate | 9e-05 |
+| Max Train Steps | 600 |
+| LoRA Rank | 16 |
+| Seed | 93 |
+| N Training Samples | 10 |