Upload trained artifacts for CheXVision-DenseNet_best

CheXVision-DenseNet_best.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:210af571ca3547a291d223bd7d7db69bb47fd9f86776e5953d43c944002705d1
+oid sha256:2f1eeba89970750eb60ddde6f632cdc118387bc2dae24851237f5ddd5edae552
 size 90799934

CheXVision-DenseNet_history.json

@@ -1,193 +1,178 @@
 {
   "train_combined_loss": [
-    1.6672315075840674,
-    1.5855173580852941,
-    1.5592198361614762,
-    1.5487742443822377,
-    1.533348664671953,
-    1.4505101817281314,
-    1.3638684952117206,
-    1.316138303984273,
-    1.2747185356704045,
-    1.2536927706695355,
-    1.2161953345301897,
-    1.1947136731435395,
-    1.1737130679064811,
-    1.1437108072042954,
-    1.1315052007425501,
-    1.0931171274693912,
-    1.0797928709645324,
-    1.0535829408666761,
-    1.0339203987348555,
-    1.0123691707543776,
-    0.9922952853023982,
-    0.9697442016838318,
-    0.9479500066319303,
-    0.929807168565718,
-    0.9069501622367894,
-    0.8852634037622038,
-    0.8630635451722644,
-    0.8424824115116674,
-    0.8241272588880717,
-    0.8091002300008673,
-    0.7876492881192797,
-    0.7686443214270733,
-    0.7624713006961722,
-    0.7363042086179402,
-    0.7175544890459317,
-    0.7050837610019344
+    1.6598473385907424,
+    1.5830327354236249,
+    1.5553311379405674,
+    1.5415173049852202,
+    1.5357806354935406,
+    1.4412867880006979,
+    1.355397817885997,
+    1.2959734734185624,
+    1.2534208430367715,
+    1.2243307756438333,
+    1.1978191353607706,
+    1.174424354536198,
+    1.1507222607305285,
+    1.1218578077786223,
+    1.0995111339970638,
+    1.0757077404740187,
+    1.0526542515892658,
+    1.027616428151135,
+    0.9989768041394681,
+    0.9858670768351803,
+    0.9625730427935439,
+    0.9357865687589273,
+    0.9066939661872684,
+    0.8850516295682543,
+    0.86539619569413,
+    0.8415834151641814,
+    0.8196068031672662,
+    0.7928312671811957,
+    0.7708154530589784,
+    0.7589645949775676,
+    0.7333919059940396,
+    0.7248045627099986,
+    0.7015012232624006
   ],
   "val_multilabel_loss": [
-    1.391814619525155,
-    1.283202405963371,
-    1.2938397450011168,
-    1.2598822928631483,
-    1.3111022481722618,
-    1.120403200276752,
-    1.0895566791296005,
-    1.0482562256615553,
-    1.0301720525012976,
-    1.0071450451861566,
-    1.0157255721625997,
-    1.0664474278021214,
-    1.0315355224840677,
-    1.0951378871478252,
-    1.0552206663831847,
-    1.12130986259722,
-    1.1844406880215923,
-    1.1695454004429169,
-    1.1427133676609886,
-    1.3337842680291454,
-    1.2761257038966043,
-    1.3477276124393762,
-    1.4940743329587267,
-    1.432857816510681,
-    1.478143415455498,
-    1.6065110461925394,
-    1.6522386216961626,
-    1.6714662681542225,
-    1.7583297681786232,
-    1.7841808852086316,
-    1.9259776598426388,
-    2.022275798523159,
-    1.8677154017139728,
-    2.1781055077966025,
-    2.192373887399462,
-    2.322838868278621
+    1.3684325010335745,
+    1.273744741574723,
+    1.2722729785936553,
+    1.2508044130995828,
+    1.2824201090496128,
+    1.1325552607116913,
+    1.0886743135896384,
+    1.0846356999556892,
+    1.0255855882869047,
+    1.0074244260120124,
+    1.0599090279567809,
+    1.028928539749621,
+    1.0678984710231882,
+    1.0938552342960481,
+    1.0641108942048556,
+    1.1280811198201834,
+    1.188532403787645,
+    1.1011257420818346,
+    1.261020954792239,
+    1.25418298025759,
+    1.3895246052399737,
+    1.4935915688423216,
+    1.5813781504287105,
+    1.5221532097801107,
+    1.5584234548550027,
+    1.7118924613885518,
+    1.7202977810253282,
+    1.7771419114341922,
+    1.8976082765087694,
+    2.086397220106686,
+    2.090599572499927,
+    2.2298203323777317,
+    2.2470968332277295
   ],
   "auc_roc_macro": [
-    0.6635553597144401,
-    0.690126818990662,
-    0.7026300567030394,
-    0.7077785989256415,
-    0.711531048074537,
-    0.7884057869255211,
-    0.8021787817203794,
-    0.814737617673185,
-    0.8232630495706772,
-    0.8266807519355579,
-    0.8319172216627019,
-    0.8329888755089002,
-    0.8364486455486551,
-    0.8358844517172852,
-    0.8353378578162786,
-    0.837478089176452,
-    0.8371722614661935,
-    0.8366371728390217,
-    0.8363548498608308,
-    0.8309555435736067,
-    0.8379018091976481,
-    0.8353094618382422,
-    0.8346713502632593,
-    0.8333661694942792,
-    0.8270327159929527,
-    0.8285016427003444,
-    0.8328671362178695,
-    0.8247545045846307,
-    0.8253445609203451,
-    0.8222061501407109,
-    0.8264586846654272,
-    0.8211011657453662,
-    0.8227619132716918,
-    0.819875838158984,
-    0.822493469671855,
-    0.8206557370930729
+    0.6748690781254583,
+    0.6963303048898386,
+    0.7041644252113736,
+    0.7115742873720903,
+    0.7172102566996948,
+    0.7905994114174624,
+    0.810920922006727,
+    0.8211061947914529,
+    0.83170534782002,
+    0.8353719186085854,
+    0.836936856831387,
+    0.8415721432417781,
+    0.8417229334900547,
+    0.8437255116407735,
+    0.8441203725443387,
+    0.8446463492148011,
+    0.8439425516705368,
+    0.8458609069545974,
+    0.8328883076496458,
+    0.8378938406822314,
+    0.8377455127591313,
+    0.8353224748952416,
+    0.8336654948819363,
+    0.836216788026232,
+    0.8317317809998471,
+    0.8307567269501813,
+    0.8338350893281437,
+    0.8257617060309652,
+    0.8323103592175635,
+    0.823613606896898,
+    0.8277971314431126,
+    0.8235523396677928,
+    0.8248399844979165
   ],
   "binary_auc_roc": [
-    0.6870642716723607,
-    0.692933835525465,
-    0.6967256315448201,
-    0.6980588299501597,
-    0.7025153760759218,
-    0.7504576233083851,
-    0.7631432081349477,
-    0.7672587640999193,
-    0.7734587558518294,
-    0.7727159514698523,
-    0.7748166929238337,
-    0.777404054877591,
-    0.7794176086411293,
-    0.7776481636008651,
-    0.7791180830631972,
-    0.7818484137779861,
-    0.7831136719039797,
-    0.778543994706878,
-    0.7813867280611452,
-    0.779365441713236,
-    0.7822723331044937,
-    0.7826895004279817,
-    0.7803191271189803,
-    0.7789640197245901,
-    0.7763517510045076,
-    0.7759058946723391,
-    0.7802235064289155,
-    0.7681305849711283,
-    0.7718824573211773,
-    0.7659790213865354,
-    0.7707724952646325,
-    0.7657138441725115,
-    0.7636363284882528,
-    0.7678690779330196,
-    0.7652841534243358,
-    0.7590750002409428
+    0.6852069105097925,
+    0.6944432584079528,
+    0.6968134916339042,
+    0.6981791893068893,
+    0.7028733162857863,
+    0.7519753951406647,
+    0.7657410202842497,
+    0.7688664972670358,
+    0.7775251986994017,
+    0.7699077625957635,
+    0.775582414892823,
+    0.7818246556926725,
+    0.7813333003857998,
+    0.7800880741359846,
+    0.7846549776886924,
+    0.7852866121620329,
+    0.7859527630989419,
+    0.7835151499258456,
+    0.7817557348319755,
+    0.7844497280035431,
+    0.7867184010181012,
+    0.7859603836168727,
+    0.7794227917139867,
+    0.7823948497401964,
+    0.7756213019475579,
+    0.7772615623989301,
+    0.7810689916866874,
+    0.7711492626364437,
+    0.7740274370024111,
+    0.7714149161328382,
+    0.7731241254615177,
+    0.7678977949583101,
+    0.7626616866581988
   ],
   "binary_f1": [
-    0.566632608378339,
-    0.46118233618233617,
-    0.5472200469851214,
-    0.5780633854242734,
-    0.5602102334209306,
-    0.6342440280360463,
-    0.6325553560742071,
-    0.6478433598183881,
-    0.6447831184056272,
-    0.6640504555010511,
-    0.6568115942028986,
-    0.6500074371560315,
-    0.6337423312883436,
-    0.659067955561968,
-    0.6592718998862344,
-    0.6584266395230478,
-    0.6588473790906458,
-    0.6467065868263473,
-    0.6542443064182195,
-    0.6474195961106955,
-    0.6527531083481349,
-    0.6677178899082569,
-    0.6450067375355593,
-    0.6477442485808187,
-    0.6679819616685456,
-    0.655732903792649,
-    0.6640904536997281,
-    0.6571672841265318,
-    0.6612788884162768,
-    0.6417780360255665,
-    0.6576781152426909,
-    0.647710702828813,
-    0.6529738981600343,
-    0.6508858553586988,
-    0.6578947368421053,
-    0.6380883417813179
+    0.5229180116204003,
+    0.48600377812124335,
+    0.5525170913610938,
+    0.5639213275968039,
+    0.5333333333333333,
+    0.6038617253192152,
+    0.6256488549618321,
+    0.6293939393939394,
+    0.6291238299831211,
+    0.6580948430803263,
+    0.6537348002316156,
+    0.6726382085374387,
+    0.6544555894752144,
+    0.6570020295737895,
+    0.6666666666666666,
+    0.6583592938733126,
+    0.6644670778941328,
+    0.6583053813621117,
+    0.6550308008213552,
+    0.6466686765149231,
+    0.6628041714947857,
+    0.6636588560617086,
+    0.6411729141475212,
+    0.663972286374134,
+    0.6735582154515778,
+    0.6535260621988611,
+    0.6587392550143266,
+    0.6628342245989305,
+    0.6581698410425318,
+    0.6386006522383635,
+    0.6629619338705196,
+    0.6505338078291815,
+    0.6378552593026663
   ],
   "learning_rate": [
     9.99314767377287e-05,
@@ -222,9 +207,6 @@
     5.0866619006049974e-05,
     4.820446585952285e-05,
     4.5549609489726355e-05,
-    4.29093266734333e-05,
-    4.029085424231934e-05,
-    3.7701369247310385e-05,
-    3.514796928678483e-05
+    4.29093266734333e-05
   ]
 }

README.md

@@ -1,111 +1,167 @@
----
-license: mit
-language:
-- en
-library_name: pytorch
-pipeline_tag: image-classification
-tags:
-- chexvision
-- medical-imaging
-- chest-xray
-- radiology
-- pytorch
-- multi-label-classification
-datasets:
-- HlexNC/chest-xray-14
----
-
-# CheXVision-DenseNet
-
-> **CheXVision** — Deep Learning & Big Data university project.
-> 14-class chest X-ray pathology detection + binary normal/abnormal classification
-> on the NIH Chest X-ray14 dataset (112,120 images).
-
-## Architecture
-
-<p align="center">
-  <img src="https://huggingface.co/HlexNC/chexvision-densenet/resolve/main/arch_densenet.png" width="88%" alt="DenseNet Architecture"/>
-</p>
-
-## Fine-Tuning Strategy
-
-<p align="center">
-  <img src="https://huggingface.co/HlexNC/chexvision-densenet/resolve/main/finetuning_densenet.png" width="62%" alt="Fine-Tuning Strategy"/>
-</p>
-
-## Training Pipeline
-
-<p align="center">
-  <img src="https://huggingface.co/HlexNC/chexvision-densenet/resolve/main/pipeline_training.png" width="42%" alt="Training Pipeline"/>
-</p>
-
-## Training Metrics
-
-- Best validation macro AUC-ROC: `0.8379`
-- Best validation binary AUC-ROC: `0.7831`
-- Best validation binary F1: `0.6680`
-- Best checkpoint epoch: `21`
-
-
-## Per-Class AUC-ROC at Best Epoch
-
-| Pathology            | AUC-ROC  | Visual        |
-|----------------------|----------|---------------|
-| Atelectasis          | `0.8178` | `████████░░` |
-| Cardiomegaly         | `0.8996` | `█████████░` |
-| Effusion             | `0.8888` | `█████████░` |
-| Infiltration         | `0.7083` | `███████░░░` |
-| Mass                 | `0.8910` | `█████████░` |
-| Nodule               | `0.7837` | `████████░░` |
-| Pneumonia            | `0.7181` | `███████░░░` |
-| Pneumothorax         | `0.8731` | `█████████░` |
-| Consolidation        | `0.8128` | `████████░░` |
-| Edema                | `0.8993` | `█████████░` |
-| Emphysema            | `0.8907` | `█████████░` |
-| Fibrosis             | `0.7994` | `████████░░` |
-| Pleural_Thickening   | `0.8311` | `████████░░` |
-| Hernia               | `0.9171` | `█████████░` |
-
-## Training Configuration
-
-- Repository: `HlexNC/chexvision-densenet`
-- Dataset: [HlexNC/chest-xray-14](https://huggingface.co/datasets/HlexNC/chest-xray-14) · revision `c4e9a86b`
-- Architecture: DenseNet-121 transfer learning with a shared feature layer and dual classification heads.
-- Platform: Kaggle GPU kernel (NVIDIA T4 / P100)
-- Batch size: `32` × grad_accum `4` = **effective batch `128`**
-- AMP (fp16): `enabled`
-- Optimizer: AdamW  ·  Scheduler: CosineAnnealingLR
-- Epochs configured: `60`  ·  Early stop patience: `15`
-
-## Intended Use
-
-This model is intended for research and educational work on automated chest X-ray pathology detection.
-It outputs two predictions per image:
-1. **Multi-label scores** — independent sigmoid probability for each of 14 NIH pathologies
-2. **Binary score** — sigmoid probability of any abnormality (Normal vs. Abnormal)
-
-## Limitations
-
-- Not validated for clinical use. Predictions must not substitute professional medical judgment.
-- Trained on NIH Chest X-ray14, which contains noisy radiologist annotations (patient-level labels, not lesion-level).
-- Performance degrades on images from equipment, patient populations, or preprocessing pipelines
-  that differ from the NIH training distribution.
-- Reported AUC metrics are on the validation split, not the held-out test set.
-
-## CheXNet Benchmark Context
-
-CheXNet (Rajpurkar et al., 2017) — the seminal paper establishing DenseNet-121 for chest X-ray
-classification — reported **0.841 macro AUC-ROC** on a comparable split of this dataset.
-CheXVision-DenseNet matches this benchmark. See the
-[CheXVision demo](https://huggingface.co/spaces/HlexNC/chexvision-demo) for live inference.
-
-## Citation
-
-```bibtex
-@misc{chexvision2026,
-  title={CheXVision: Dual-Task Chest X-ray Classification with Custom CNN and DenseNet-121},
-  author={BIG D(ATA) Team},
-  year={2026},
-  howpublished={\url{https://huggingface.co/HlexNC/chexvision-densenet}}
-}
-```
+---
+license: mit
+language:
+- en
+library_name: pytorch
+pipeline_tag: image-classification
+tags:
+- chexvision
+- medical-imaging
+- chest-xray
+- radiology
+- pytorch
+- multi-label-classification
+datasets:
+- HlexNC/chest-xray-14-320
+---
+
+# CheXVision-DenseNet
+
+> **CheXVision** — Deep Learning & Big Data university project.
+> 14-class chest X-ray pathology detection + binary normal/abnormal classification
+> on the NIH Chest X-ray14 dataset (112,120 images).
+
+## Architecture
+
+```mermaid
+graph LR
+    IN["Input
+    3 × 224 × 224"] --> BB["DenseNet-121 Backbone
+    ImageNet pretrained
+    Dense connectivity
+    7.9M parameters"]
+    BB --> GAP2["Adaptive Avg Pool
+    1024-dim features"]
+    GAP2 --> FL["Feature Layer
+    Linear 1024→512
+    ReLU · Dropout(0.3)"]
+    FL --> MLH["Multilabel Head
+    Linear 512→14
+    sigmoid · 14 pathologies"]
+    FL --> BH["Binary Head
+    Linear 512→1
+    sigmoid · Normal/Abnormal"]
+    style MLH fill:#2e7d32,color:#fff
+    style BH fill:#1565c0,color:#fff
+    style IN fill:#37474f,color:#fff
+    style BB fill:#6a1b9a,color:#fff
+```
+
+## Fine-Tuning Strategy
+
+```mermaid
+graph LR
+    P1["🔒 Phase 1
+    Epochs 1–5
+    Backbone frozen
+    Train heads only
+    lr = 0.001"] -->|"Epoch 6
+    unfreeze_backbone()"| P2["🔓 Phase 2
+    Epochs 6–60
+    End-to-end fine-tuning
+    All layers trainable
+    lr = 0.0001"]
+    style P1 fill:#e65100,color:#fff
+    style P2 fill:#6a1b9a,color:#fff
+```
+
+## Training Pipeline
+
+```mermaid
+flowchart TD
+    DS[("🗄️ HlexNC/chest-xray-14
+112,120 images · 36 shards · ~4.7 GB")]
+    DS -->|snapshot_download| PREP["📂 data/images · data/labels.csv
+train 78,468 · val 11,210 · test 22,442"]
+    PREP --> AUG["Augmentation Pipeline
+HFlip · Rotate±15° · RandomAffine
+ColorJitter · GaussianBlur · RandomErasing"]
+    AUG --> FWD["⚡ Model Forward Pass
+torch.cuda.amp.autocast · fp16"]
+    FWD --> ML["multilabel_logits B×14
+WeightedBCE + pos_weight · 14 classes"]
+    FWD --> BIN["binary_logits B×1
+BCE · Normal vs. Abnormal"]
+    ML --> LOSS["Combined Loss
+1.0 × multilabel + 0.5 × binary"]
+    BIN --> LOSS
+    LOSS --> BACK["Backward · Grad Clip 1.0
+Gradient Accumulation ×4 · eff. batch 128"]
+    BACK --> OPT["AdamW · CosineAnnealingLR
+early stop patience = 15"]
+    OPT -->|"↑ best val AUC-ROC"| BEST["💾 Best Checkpoint
+model_state · best_val_metrics · config"]
+    BEST -->|upload_model_artifacts| HUB["🤗 HF Hub
+checkpoint · history.json · model card"]
+```
+
+## Training Metrics
+
+- Best validation macro AUC-ROC: `0.8459`
+- Best validation binary AUC-ROC: `0.7867`
+- Best validation binary F1: `0.6736`
+- Best checkpoint epoch: `18`
+
+
+## Per-Class AUC-ROC at Best Epoch
+
+| Pathology            | AUC-ROC  | Visual        |
+|----------------------|----------|---------------|
+| Atelectasis          | `0.8334` | `████████░░` |
+| Cardiomegaly         | `0.9010` | `█████████░` |
+| Effusion             | `0.8873` | `█████████░` |
+| Infiltration         | `0.7133` | `███████░░░` |
+| Mass                 | `0.8756` | `█████████░` |
+| Nodule               | `0.8084` | `████████░░` |
+| Pneumonia            | `0.7397` | `███████░░░` |
+| Pneumothorax         | `0.8705` | `█████████░` |
+| Consolidation        | `0.8063` | `████████░░` |
+| Edema                | `0.9255` | `█████████░` |
+| Emphysema            | `0.9107` | `█████████░` |
+| Fibrosis             | `0.8085` | `████████░░` |
+| Pleural_Thickening   | `0.8377` | `████████░░` |
+| Hernia               | `0.9242` | `█████████░` |
+
+## Training Configuration
+
+- Repository: `HlexNC/chexvision-densenet`
+- Dataset: [HlexNC/chest-xray-14-320](https://huggingface.co/datasets/HlexNC/chest-xray-14-320) · revision `44443e6ee968b3c6094b63f14a27698c40b50680`
+- Architecture: DenseNet-121 transfer learning with a shared feature layer and dual classification heads.
+- Platform: Kaggle GPU kernel (NVIDIA T4 / P100)
+- Batch size: `24` × grad_accum `4` = **effective batch `96`**
+- AMP (fp16): `enabled`
+- Optimizer: AdamW  ·  Scheduler: CosineAnnealingLR
+- Epochs configured: `60`  ·  Early stop patience: `15`
+
+## Intended Use
+
+This model is intended for research and educational work on automated chest X-ray pathology detection.
+It outputs two predictions per image:
+1. **Multi-label scores** — independent sigmoid probability for each of 14 NIH pathologies
+2. **Binary score** — sigmoid probability of any abnormality (Normal vs. Abnormal)
+
+## Limitations
+
+- Not validated for clinical use. Predictions must not substitute professional medical judgment.
+- Trained on NIH Chest X-ray14, which contains noisy radiologist annotations (patient-level labels, not lesion-level).
+- Performance degrades on images from equipment, patient populations, or preprocessing pipelines
+  that differ from the NIH training distribution.
+- Reported AUC metrics are on the validation split, not the held-out test set.
+
+## CheXNet Benchmark Context
+
+CheXNet (Rajpurkar et al., 2017) — the seminal paper establishing DenseNet-121 for chest X-ray
+classification — reported **0.841 macro AUC-ROC** on a comparable split of this dataset.
+CheXVision-DenseNet matches this benchmark. See the
+[CheXVision demo](https://huggingface.co/spaces/HlexNC/chexvision-demo) for live inference.
+
+## Citation
+
+```bibtex
+@misc{chexvision2026,
+  title={CheXVision: Dual-Task Chest X-ray Classification with Custom CNN and DenseNet-121},
+  author={BIG D(ATA) Team},
+  year={2026},
+  howpublished={\url{https://huggingface.co/HlexNC/chexvision-densenet}}
+}
+```

training_config.json

@@ -1,8 +1,8 @@
 {
   "seed": 42,
   "data": {
-    "dataset_name": "HlexNC/chest-xray-14",
-    "image_size": 224,
+    "dataset_name": "HlexNC/chest-xray-14-320",
+    "image_size": 320,
     "num_workers": 4,
     "pin_memory": true,
     "train_split": 0.7,
@@ -38,12 +38,12 @@
       "Hernia"
     ],
     "data_dir": "/kaggle/working/data",
-    "hf_dataset_repo": "HlexNC/chest-xray-14",
-    "hf_dataset_revision": "c4e9a86b38de3b1604afa6e9f514d156eb9d20bf"
+    "hf_dataset_repo": "HlexNC/chest-xray-14-320",
+    "hf_dataset_revision": "44443e6ee968b3c6094b63f14a27698c40b50680"
   },
   "training": {
     "epochs": 60,
-    "batch_size": 32,
+    "batch_size": 24,
     "learning_rate": 0.0001,
     "weight_decay": 0.0001,
     "optimizer": "adamw",
@@ -67,7 +67,7 @@
     "log_every_n_steps": 100
   },
   "huggingface": {
-    "dataset_repo": "HlexNC/chest-xray-14",
+    "dataset_repo": "HlexNC/chest-xray-14-320",
     "scratch_model_repo": "HlexNC/chexvision-scratch",
     "transfer_model_repo": "HlexNC/chexvision-densenet",
     "space_repo": "HlexNC/chexvision-demo"