Upload DeiT model from experiment b1

.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
+training_curves/DeiT3_combined_metrics.png filter=lfs diff=lfs merge=lfs -text
+training_curves/DeiT_combined_metrics.png filter=lfs diff=lfs merge=lfs -text
+training_notebook_b1.ipynb filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,161 @@
+---
+license: apache-2.0
+tags:
+- vision-transformer
+- image-classification
+- pytorch
+- timm
+- deit
+- gravitational-lensing
+- strong-lensing
+- astronomy
+- astrophysics
+datasets:
+- J24
+metrics:
+- accuracy
+- auc
+- f1
+model-index:
+- name: DeiT-b1
+  results:
+  - task:
+      type: image-classification
+      name: Strong Gravitational Lens Discovery
+    dataset:
+      type: common-test-sample
+      name: Common Test Sample (More et al. 2024)
+    metrics:
+    - type: accuracy
+      value: 0.7176
+      name: Average Accuracy
+    - type: auc
+      value: 0.6473
+      name: Average AUC-ROC
+    - type: f1
+      value: 0.3495
+      name: Average F1-Score
+---
+
+# 🌌 deit-gravit-b1
+
+🔭 This model is part of **GraViT**: Transfer Learning with Vision Transformers and MLP-Mixer for Strong Gravitational Lens Discovery
+
+🔗 **GitHub Repository**: [https://github.com/parlange/gravit](https://github.com/parlange/gravit)
+
+## 🛰️ Model Details
+
+- **🤖 Model Type**: DeiT
+- **🧪 Experiment**: B1 - J24-classification-head
+- **🌌 Dataset**: J24
+- **🪐 Fine-tuning Strategy**: classification-head
+
+
+
+## 💻 Quick Start
+
+```python
+import torch
+import timm
+
+# Load the model directly from the Hub
+model = timm.create_model(
+    'hf-hub:parlange/deit-gravit-b1',
+    pretrained=True
+)
+model.eval()
+
+# Example inference
+dummy_input = torch.randn(1, 3, 224, 224)
+with torch.no_grad():
+    output = model(dummy_input)
+    predictions = torch.softmax(output, dim=1)
+print(f"Lens probability: {predictions[0][1]:.4f}")
+```
+
+## ⚡️ Training Configuration
+
+**Training Dataset:** J24 (Jaelani et al. 2024)  
+**Fine-tuning Strategy:** classification-head
+
+
+| 🔧 Parameter | 📝 Value |
+|--------------|----------|
+| Batch Size | 192 |
+| Learning Rate | AdamW with ReduceLROnPlateau |
+| Epochs | 100 |
+| Patience | 10 |
+| Optimizer | AdamW |
+| Scheduler | ReduceLROnPlateau |
+| Image Size | 224x224 |
+| Fine Tune Mode | classification_head |
+| Stochastic Depth Probability | 0.1 |
+
+
+## 📈 Training Curves
+
+![Combined Training Metrics](https://huggingface.co/parlange/deit-gravit-b1/resolve/main/training_curves/DeiT_combined_metrics.png)
+
+
+## 🏁 Final Epoch Training Metrics
+
+| Metric | Training | Validation |
+|:---------:|:-----------:|:-------------:|
+| 📉 Loss | 0.2298 | 0.2241 |
+| 🎯 Accuracy | 0.9097 | 0.9105 |
+| 📊 AUC-ROC | 0.9668 | 0.9682 |
+| ⚖️ F1 Score | 0.9083 | 0.9092 |
+
+
+## ☑️ Evaluation Results
+
+### ROC Curves and Confusion Matrices
+
+Performance across all test datasets (a through l) in the Common Test Sample (More et al. 2024):
+
+![ROC + Confusion Matrix - Dataset A](https://huggingface.co/parlange/deit-gravit-b1/resolve/main/roc_confusion_matrix/DeiT_roc_confusion_matrix_a.png)
+![ROC + Confusion Matrix - Dataset B](https://huggingface.co/parlange/deit-gravit-b1/resolve/main/roc_confusion_matrix/DeiT_roc_confusion_matrix_b.png)
+![ROC + Confusion Matrix - Dataset C](https://huggingface.co/parlange/deit-gravit-b1/resolve/main/roc_confusion_matrix/DeiT_roc_confusion_matrix_c.png)
+![ROC + Confusion Matrix - Dataset D](https://huggingface.co/parlange/deit-gravit-b1/resolve/main/roc_confusion_matrix/DeiT_roc_confusion_matrix_d.png)
+![ROC + Confusion Matrix - Dataset E](https://huggingface.co/parlange/deit-gravit-b1/resolve/main/roc_confusion_matrix/DeiT_roc_confusion_matrix_e.png)
+![ROC + Confusion Matrix - Dataset F](https://huggingface.co/parlange/deit-gravit-b1/resolve/main/roc_confusion_matrix/DeiT_roc_confusion_matrix_f.png)
+![ROC + Confusion Matrix - Dataset G](https://huggingface.co/parlange/deit-gravit-b1/resolve/main/roc_confusion_matrix/DeiT_roc_confusion_matrix_g.png)
+![ROC + Confusion Matrix - Dataset H](https://huggingface.co/parlange/deit-gravit-b1/resolve/main/roc_confusion_matrix/DeiT_roc_confusion_matrix_h.png)
+![ROC + Confusion Matrix - Dataset I](https://huggingface.co/parlange/deit-gravit-b1/resolve/main/roc_confusion_matrix/DeiT_roc_confusion_matrix_i.png)
+![ROC + Confusion Matrix - Dataset J](https://huggingface.co/parlange/deit-gravit-b1/resolve/main/roc_confusion_matrix/DeiT_roc_confusion_matrix_j.png)
+![ROC + Confusion Matrix - Dataset K](https://huggingface.co/parlange/deit-gravit-b1/resolve/main/roc_confusion_matrix/DeiT_roc_confusion_matrix_k.png)
+![ROC + Confusion Matrix - Dataset L](https://huggingface.co/parlange/deit-gravit-b1/resolve/main/roc_confusion_matrix/DeiT_roc_confusion_matrix_l.png)
+
+### 📋 Performance Summary
+
+Average performance across 12 test datasets from the Common Test Sample (More et al. 2024):
+
+| Metric | Value |
+|-----------|----------|
+| 🎯 Average Accuracy | 0.7176 |
+| 📈 Average AUC-ROC | 0.6473 |
+| ⚖️ Average F1-Score | 0.3495 |
+
+
+## 📘 Citation
+
+If you use this model in your research, please cite:
+
+```bibtex
+@misc{parlange2025gravit,
+      title={GraViT: Transfer Learning with Vision Transformers and MLP-Mixer for Strong Gravitational Lens Discovery}, 
+      author={René Parlange and Juan C. Cuevas-Tello and Octavio Valenzuela and Omar de J. Cabrera-Rosas and Tomás Verdugo and Anupreeta More and Anton T. Jaelani},
+      year={2025},
+      eprint={2509.00226},
+      archivePrefix={arXiv},
+      primaryClass={cs.CV},
+      url={https://arxiv.org/abs/2509.00226}, 
+}
+```
+
+---
+
+
+## Model Card Contact
+
+For questions about this model, please contact the author through: https://github.com/parlange/

config.json

@@ -0,0 +1,76 @@
+{
+  "architecture": "vit_base_patch16_224",
+  "num_classes": 2,
+  "num_features": 1000,
+  "global_pool": "avg",
+  "crop_pct": 0.875,
+  "interpolation": "bicubic",
+  "mean": [
+    0.485,
+    0.456,
+    0.406
+  ],
+  "std": [
+    0.229,
+    0.224,
+    0.225
+  ],
+  "first_conv": "conv1",
+  "classifier": "fc",
+  "input_size": [
+    3,
+    224,
+    224
+  ],
+  "pool_size": [
+    7,
+    7
+  ],
+  "pretrained_cfg": {
+    "tag": "gravit_b1",
+    "custom_load": false,
+    "input_size": [
+      3,
+      224,
+      224
+    ],
+    "fixed_input_size": true,
+    "interpolation": "bicubic",
+    "crop_pct": 0.875,
+    "crop_mode": "center",
+    "mean": [
+      0.485,
+      0.456,
+      0.406
+    ],
+    "std": [
+      0.229,
+      0.224,
+      0.225
+    ],
+    "num_classes": 2,
+    "pool_size": [
+      7,
+      7
+    ],
+    "first_conv": "conv1",
+    "classifier": "fc"
+  },
+  "model_name": "deit_gravit_b1",
+  "experiment": "b1",
+  "training_strategy": "classification-head",
+  "dataset": "J24",
+  "hyperparameters": {
+    "batch_size": "192",
+    "learning_rate": "AdamW with ReduceLROnPlateau",
+    "epochs": "100",
+    "patience": "10",
+    "optimizer": "AdamW",
+    "scheduler": "ReduceLROnPlateau",
+    "image_size": "224x224",
+    "fine_tune_mode": "classification_head",
+    "stochastic_depth_probability": "0.1"
+  },
+  "hf_hub_id": "parlange/deit-gravit-b1",
+  "license": "apache-2.0"
+}

deit-gravit-b1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c55ce83a041d08b258b478f0138aa72111e5f6bd6da224cabc6137e0877ad1f7
+size 343259194

evaluation_results.csv

@@ -0,0 +1,133 @@
+Model,Dataset,Loss,Accuracy,AUCROC,F1
+ViT,a,0.6071771883394912,0.6972650110028293,0.630304788213628,0.16041848299912817
+ViT,b,0.5153288846587805,0.7585664885256209,0.7058489871086556,0.19327731092436976
+ViT,c,0.7054783390475984,0.6321911348632505,0.5800184162062615,0.1358936484490399
+ViT,d,0.4042277384557577,0.8236403646651996,0.7250165745856354,0.24697986577181208
+ViT,e,0.5327941425283707,0.7420417124039517,0.7240974797547869,0.43914081145584727
+ViT,f,0.5265797158730187,0.7412284098830455,0.6639558097850381,0.05219858156028369
+ViT,g,0.6893135578632354,0.6621666666666667,0.7476270555555556,0.6197711498780717
+ViT,h,0.790124472618103,0.5951666666666666,0.625018,0.5763125763125763
+ViT,i,0.6304114277362823,0.6966666666666667,0.7770440555555556,0.644808743169399
+ViT,j,0.465672768831253,0.776,0.864817,0.7765214499501164
+ViT,k,0.4067706377506256,0.8105,0.8964317777777777,0.8042018253831582
+ViT,l,0.5734097661542007,0.7168843530220507,0.7773444022320638,0.6037596210775606
+MLP-Mixer,a,0.46569699331935344,0.7956617415906948,0.6377578268876612,0.17091836734693877
+MLP-Mixer,b,0.4452243679182891,0.8113800691606413,0.680889502762431,0.18256130790190736
+MLP-Mixer,c,0.47408608185799006,0.7922037095253065,0.6053204419889502,0.16855345911949685
+MLP-Mixer,d,0.42086313657886537,0.8296133291417793,0.628830570902394,0.19822485207100593
+MLP-Mixer,e,0.5369727452879549,0.7694840834248079,0.669424052069931,0.38953488372093026
+MLP-Mixer,f,0.4062789948943527,0.8290604910541399,0.6399901481253227,0.05724049551473729
+MLP-Mixer,g,0.8928026676177978,0.5668333333333333,0.6760506666666667,0.4056711639606677
+MLP-Mixer,h,0.9081041851043701,0.5566666666666666,0.5911980555555556,0.4000902119981958
+MLP-Mixer,i,0.8798871593475341,0.5765,0.6166056111111111,0.4111239860950174
+MLP-Mixer,j,0.46018748569488527,0.7815,0.874597,0.7684154742978272
+MLP-Mixer,k,0.44727197551727294,0.7911666666666667,0.8744243888888887,0.7763698018918437
+MLP-Mixer,l,0.5829932733619303,0.7279361218338534,0.7454255642147913,0.5488029465930019
+CvT,a,0.7422070925208016,0.5998113800691607,0.3903342541436464,0.06876371616678859
+CvT,b,0.5952609254949612,0.7066960075447972,0.4973591160220995,0.09152872444011685
+CvT,c,0.723317504959712,0.6139578748821125,0.3812725598526704,0.07110438729198185
+CvT,d,0.430008566209034,0.8365293932725558,0.5662173112338857,0.15309446254071662
+CvT,e,0.6766743037101335,0.6871569703622393,0.5160750775751155,0.24802110817941952
+CvT,f,0.5791996195594972,0.7131903028425374,0.4620804815700503,0.02475638662101659
+CvT,g,1.034350182056427,0.45866666666666667,0.46588566666666664,0.25333333333333335
+CvT,h,1.1022415161132812,0.4095,0.33246555555555557,0.23724434876210979
+CvT,i,0.9467389023303986,0.5275,0.5413065,0.27990855981711965
+CvT,j,0.6915355019569397,0.6251666666666666,0.6959060000000001,0.5795475789867265
+CvT,k,0.6039242098331451,0.694,0.7733503333333334,0.6280388978930308
+CvT,l,0.7708275838721139,0.5980117391994078,0.5461840780345415,0.3610690872415532
+Swin,a,0.3745372065280601,0.8440741905061302,0.7235939226519337,0.2392638036809816
+Swin,b,0.4118781376750692,0.8274127632819868,0.7413885819521178,0.22127659574468084
+Swin,c,0.41712035007800596,0.8258409305249922,0.6833775322283611,0.21971830985915494
+Swin,d,0.3414299682939776,0.867966048412449,0.7282725598526704,0.2708333333333333
+Swin,e,0.47762388664856964,0.7881448957189902,0.7537349579959131,0.4469914040114613
+Swin,f,0.34434392800953245,0.86058399814112,0.7211409512484105,0.07975460122699386
+Swin,g,0.9412130119800568,0.562,0.687752,0.38367729831144465
+Swin,h,0.9439922497272492,0.5611666666666667,0.600395,0.38322792223003044
+Swin,i,0.9038637335300446,0.5835,0.6599881666666667,0.39564691656590084
+Swin,j,0.41330916619300845,0.811,0.8991051666666667,0.8030566168808614
+Swin,k,0.3759598867893219,0.8325,0.9137093333333334,0.8214602949014035
+Swin,l,0.5501825281197098,0.7530537782243139,0.776428683903603,0.5787479704131336
+CaiT,a,0.5068589627499177,0.7544797233574347,0.6288020257826887,0.1574973031283711
+CaiT,b,0.43441846407422924,0.8022634391700723,0.7139318600368323,0.18838709677419355
+CaiT,c,0.6128005658594627,0.675888085507702,0.5577274401473297,0.12404418011894647
+CaiT,d,0.601496650358672,0.6966362779000315,0.5815893186003683,0.13141314131413143
+CaiT,e,0.4804434718075489,0.7683863885839737,0.7282903201392568,0.40896358543417366
+CaiT,f,0.4994646750235297,0.7533111300441484,0.6266931553341174,0.04383068147703392
+CaiT,g,0.9582805759906768,0.5296666666666666,0.6336772222222222,0.33127962085308055
+CaiT,h,1.052852813243866,0.46266666666666667,0.4407098888888889,0.30246646473388145
+CaiT,i,1.0468598504066466,0.4736666666666667,0.4733858888888889,0.30684811237928006
+CaiT,j,0.3563543026447296,0.8453333333333334,0.9224438888888887,0.8482172064115145
+CaiT,k,0.44493358278274536,0.7893333333333333,0.8849772222222221,0.804031007751938
+CaiT,l,0.634319533465761,0.6883824229284543,0.7108943442597637,0.5331537669333756
+DeiT,a,0.4427877041630983,0.808236403646652,0.6191058931860037,0.18666666666666668
+DeiT,b,0.41083011003044256,0.821754165356806,0.6952081031307551,0.19801980198019803
+DeiT,c,0.4910275342516163,0.78528764539453,0.5721933701657459,0.1701093560145808
+DeiT,d,0.5752349910365827,0.7318453316567117,0.5411786372007367,0.14098690835850958
+DeiT,e,0.5283474334148622,0.8002195389681669,0.6915310678876864,0.43478260869565216
+DeiT,f,0.42157146472453,0.8101618774688251,0.6117734242425559,0.05403319181783095
+DeiT,g,1.163999383211136,0.5158333333333334,0.6234185555555556,0.2750187172448216
+DeiT,h,1.206517367362976,0.4965,0.45666055555555557,0.2672811059907834
+DeiT,i,1.2511613540649413,0.4681666666666667,0.42239144444444443,0.2566969485208479
+DeiT,j,0.33839035058021544,0.8516666666666667,0.9297900555555554,0.8522085685818664
+DeiT,k,0.425552321434021,0.804,0.8934688888888889,0.813570069752695
+DeiT,l,0.6565662104555468,0.7179419385542806,0.7109608633913852,0.5444140758455757
+DeiT3,a,0.4096169373367763,0.8371581263753537,0.5791758747697975,0.14802631578947367
+DeiT3,b,0.373094875785923,0.8462747563659226,0.6644051565377532,0.15544041450777202
+DeiT3,c,0.4991895148717394,0.77648538195536,0.5115930018416206,0.11235955056179775
+DeiT3,d,0.38327455564663493,0.8513046211883055,0.6100073664825046,0.15985790408525755
+DeiT3,e,0.5367149264438223,0.7815587266739846,0.6753273291455385,0.31141868512110726
+DeiT3,f,0.3471956718102791,0.8570211447602819,0.5961145855485585,0.04648760330578512
+DeiT3,g,1.3192027564048767,0.5021666666666667,0.5601109444444445,0.19682710406023124
+DeiT3,h,1.3860539412498474,0.4651666666666667,0.3733101111111111,0.18573965998477543
+DeiT3,i,1.324599706172943,0.5048333333333334,0.48481599999999997,0.19767755873615986
+DeiT3,j,0.3928017973899841,0.8296666666666667,0.9115283333333334,0.8201970443349754
+DeiT3,k,0.39819873094558716,0.8323333333333334,0.9044895555555555,0.8225123500352858
+DeiT3,l,0.6891573654383029,0.7277246047274073,0.6817385665953223,0.5157528449167685
+Twins_SVT,a,0.4458898261016137,0.8167243005344231,0.6354576427255985,0.2024623803009576
+Twins_SVT,b,0.4509238941663468,0.8076076705438542,0.6965340699815838,0.19473684210526315
+Twins_SVT,c,0.5059725925331032,0.7771141150581579,0.5930441988950277,0.17269544924154026
+Twins_SVT,d,0.3316556203013057,0.8827412763281987,0.7198434622467773,0.2840690978886756
+Twins_SVT,e,0.5185028320806087,0.7738748627881449,0.7086581397108906,0.4180790960451977
+Twins_SVT,f,0.39142251728201677,0.8408333978777787,0.6639401856666073,0.06718111665910123
+Twins_SVT,g,1.2193830106258392,0.4841666666666667,0.5380571111111111,0.20945083014048532
+Twins_SVT,h,1.2485679879188538,0.468,0.36677261111111115,0.20438683948155534
+Twins_SVT,i,1.1561509382724762,0.524,0.5393812222222223,0.22306855277475518
+Twins_SVT,j,0.52089257979393,0.764,0.8455504444444444,0.7468716481944941
+Twins_SVT,k,0.4576605150699615,0.8038333333333333,0.8778117222222223,0.7802054154995332
+Twins_SVT,l,0.6924283070618104,0.7062027391465284,0.6558977267715866,0.48084470192487383
+Twins_PCPVT,a,0.4554920045694665,0.7925180760767054,0.6320966850828729,0.1770573566084788
+Twins_PCPVT,b,0.3702721736671714,0.8337000943099654,0.7290699815837937,0.21162444113263784
+Twins_PCPVT,c,0.5222789617650578,0.7601383212826155,0.5801896869244936,0.1569060773480663
+Twins_PCPVT,d,0.511706690917934,0.7695693178245835,0.6062357274401473,0.16228571428571428
+Twins_PCPVT,e,0.44942784803366426,0.8111964873765093,0.764065692878226,0.45222929936305734
+Twins_PCPVT,f,0.4167008115512896,0.8127952908372705,0.6441899979601845,0.05549042594763579
+Twins_PCPVT,g,0.9797230496406555,0.5426666666666666,0.6615357222222222,0.32975085490962386
+Twins_PCPVT,h,1.0603120126724244,0.5036666666666667,0.47164805555555556,0.31192236598890943
+Twins_PCPVT,i,1.0547069294452667,0.5086666666666667,0.5087028888888889,0.3140995812005584
+Twins_PCPVT,j,0.3549184784889221,0.845,0.9234385555555554,0.8425863236289777
+Twins_PCPVT,k,0.42990235447883607,0.811,0.8866837222222221,0.8144633507853403
+Twins_PCPVT,l,0.6083674293331931,0.7222251599598117,0.7249629563166041,0.5519065085728909
+PiT,a,0.40825032295951375,0.8340144608613643,0.685341620626151,0.23919308357348704
+PiT,b,0.4585393214237012,0.8057214712354606,0.7131197053406997,0.21173469387755103
+PiT,c,0.46198257697112005,0.8072933039924552,0.636451197053407,0.21309370988446727
+PiT,d,0.4301266003352072,0.8198679660484125,0.6617108655616942,0.22462787550744248
+PiT,e,0.47335477670579,0.8111964873765093,0.7555210777264815,0.4911242603550296
+PiT,f,0.38904998775843674,0.836418557818914,0.6788217244686715,0.07287093942054433
+PiT,g,1.0988447036743163,0.5483333333333333,0.6416311666666666,0.3741339491916859
+PiT,h,1.1006702044010162,0.5491666666666667,0.5303825555555556,0.3745664739884393
+PiT,i,1.0837812168598175,0.5558333333333333,0.5635360555555555,0.3780630105017503
+PiT,j,0.37021482944488526,0.8401666666666666,0.9175485555555555,0.8422956750534452
+PiT,k,0.355151344537735,0.8476666666666667,0.92189,0.8485752153744202
+PiT,l,0.6039756624711939,0.7492993495848976,0.7530133967294295,0.5930122757318225
+Ensemble,a,,0.8239547312165986,0.6361445672191529,0.2
+Ensemble,b,,0.821754165356806,0.7071620626151013,0.19801980198019803
+Ensemble,c,,0.7966048412448915,0.5873848987108656,0.17789072426937738
+Ensemble,d,,0.8604212511788746,0.659,0.23972602739726026
+Ensemble,e,,0.8090010976948409,0.7250132445319004,0.445859872611465
+Ensemble,f,,0.8491209046549454,0.6518722901919597,0.06704980842911877
+Ensemble,g,,0.52,0.6616331111111111,0.2857142857142857
+Ensemble,h,,0.5066666666666667,0.48887533333333333,0.2801556420233463
+Ensemble,i,,0.5405,0.5874012222222222,0.29470452801227937
+Ensemble,j,,0.8333333333333334,0.9165178888888887,0.8308525033829499
+Ensemble,k,,0.8538333333333333,0.9253933333333333,0.8485057868371049
+Ensemble,l,,0.7400454761778859,0.7437259266894098,0.5579136690647482

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f6285151bcd3a49b9b08e5fa38619b7dea2e35585fe341fe2bcfb9a82d827484
+size 343214864

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c55ce83a041d08b258b478f0138aa72111e5f6bd6da224cabc6137e0877ad1f7
+size 343259194