Create README.md

README.md

@@ -0,0 +1,59 @@
+---
+license: apache-2.0
+library_name: transformers
+pipeline_tag: text-classification
+tags:
+  - reranker
+  - cross-encoder
+  - dual-passage-classifier
+  - multi-hop-qa
+  - pytorch
+datasets:
+  - 2wiki
+model-index:
+  - name: Dual Passage Classifier
+    results:
+      - task:
+          type: text-classification
+          name: Passage Reranking (Multi-hop QA)
+        dataset:
+          name: 2wikimultihopQA
+          type: 2wiki
+          split: validation
+        metrics:
+          - type: MAP
+            value: TODO
+            name: Mean Average Precision
+---
+
+# Dual Passage Classifier (DPC) for Multi-hop QA 🔎📑
+
+**Dual Passage Classifier (DPC)** 是一個 *cross-encoder* reranker，  
+輸入 **(Question, Passage 1, Passage 2)**，輸出一維分數，判斷這對段落對回答該問題的「共同貢獻程度」：
+
+| label | 定義 | Margin label |
+|-------|------|--------------|
+| positive | d1 & d2 都必要 |  0 |
+| neutral  | 其中一段必要 |1 |
+| negative | 都不重要     | 2 |
+
+本模型將 **三段式 MarginRankingLoss** 混合，  
+並在 2wiki 上把 baseline（`naver/trecdl22-crossencoder-debertav3`）的模型。
+這是訓練在 2wiki dataset 上的模型
+---
+
+## Quick-start
+
+```python
+from transformers import AutoTokenizer, AutoModel
+import torch, torch.nn.functional as F
+
+tokenizer = AutoTokenizer.from_pretrained("your-username/dual-passage-classifier")
+model     = AutoModel.from_pretrained("your-username/dual-passage-classifier").eval()
+
+def score(q, d1, d2):
+    text = f"{q} [SEP] {d1} [SEP] {d2}"
+    inputs = tokenizer(text, return_tensors="pt", truncation=True, max_length=512)
+    with torch.no_grad():
+        logit = model(**inputs).last_hidden_state[:,0,:] @ model.classifier[0].weight.T
+    return torch.sigmoid(logit).item()         # 0~1，越高代表正向