Upload README.md with huggingface_hub

README.md

@@ -0,0 +1,90 @@
+---
+language: pt
+license: mit
+tags:
+  - document-understanding
+  - document-retrieval
+  - metric-learning
+  - siamese-network
+  - internvl
+  - cosdoc
+datasets:
+  - LA-CDIP
+metrics:
+  - eer
+model-index:
+  - name: CosDoc (Attention nq=2)
+    results:
+      - task:
+          type: document-retrieval
+        dataset:
+          name: LA-CDIP
+          type: la-cdip
+        metrics:
+          - type: eer
+            value: 0.0107
+---
+
+# CosDoc — Attention nq=2
+
+**CosDoc** is a visual document embedding model trained with supervised metric learning
+and hard-example selection via a Reinforcement Learning professor network.
+
+Pooler variant: **Attention nq=2** — Multi-query attention pooler, num_queries=2.
+
+## Architecture
+
+| Component | Value |
+|---|---|
+| Backbone | InternVL3-2B (`OpenGVLab/InternVL3-2B`) |
+| Cut layer | 27 |
+| Pooler | attention (num_queries=2) |
+| Embedding dim | 1536 |
+| Loss | Sub-Center CosFace (m=0.35, s=32, k=3) |
+| Embedding prompt | `<image> Analyze this document` |
+
+## Performance (LA-CDIP, full validation pairs)
+
+| Dataset | EER |
+|---|---|
+| LA-CDIP (5-fold CV) | **1.07%** |
+
+Source run: `Sprint3b_S0_subcenter_cosface_seed42_noinit_nq2_fase1_E10`
+
+## Usage
+
+```python
+import torch
+from huggingface_hub import hf_hub_download
+from cavl_doc.models.backbone_loader import load_model
+from cavl_doc.models.modeling_cavl import build_cavl_model
+
+device = "cuda" if torch.cuda.is_available() else "cpu"
+
+# Download fine-tuned weights
+ckpt_path = hf_hub_download(repo_id="Jpcosta90/cosdoc-nq2", filename="best_model.pt")
+ckpt = torch.load(ckpt_path, map_location=device, weights_only=False)
+cfg  = ckpt["config"]
+
+backbone, _, tokenizer, _, _ = load_model("InternVL3-2B")
+model = build_cavl_model(
+    backbone=backbone,
+    cut_layer=cfg["cut_layer"],
+    pooler_type=cfg["pooler_type"],
+    num_queries=cfg.get("num_queries", 1),
+)
+model.pool.load_state_dict(ckpt["siam_pool"])
+model.head.load_state_dict(ckpt["siam_head"])
+model.eval().to(device)
+```
+
+## Citation
+
+```bibtex
+@misc{cosdoc2026,
+  title  = {CosDoc: Cosine-Margin Document Embeddings with RL-guided Hard Mining},
+  author = {Costa, João Paulo},
+  year   = {2026},
+  url    = {https://huggingface.co/Jpcosta90/cosdoc-nq2}
+}
+```