Upload README.md with huggingface_hub
Browse files
README.md
ADDED
|
@@ -0,0 +1,58 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
base_model: Qwen/Qwen3-4B
|
| 4 |
+
tags:
|
| 5 |
+
- text-ranking
|
| 6 |
+
- reranker
|
| 7 |
+
- sentence-transformers
|
| 8 |
+
- safetensors
|
| 9 |
+
- english
|
| 10 |
+
language:
|
| 11 |
+
- en
|
| 12 |
+
---
|
| 13 |
+
|
| 14 |
+
# zerank-1-small-seq
|
| 15 |
+
|
| 16 |
+
Internal DevRev mirror of [zeroentropy/zerank-1-small](https://huggingface.co/zeroentropy/zerank-1-small).
|
| 17 |
+
|
| 18 |
+
Original model by [ZeroEntropy](https://huggingface.co/zeroentropy) — released under Apache 2.0.
|
| 19 |
+
|
| 20 |
+
## Model details
|
| 21 |
+
|
| 22 |
+
| Property | Value |
|
| 23 |
+
|----------|-------|
|
| 24 |
+
| Parameters | 1.7B |
|
| 25 |
+
| Context length | 32,768 tokens |
|
| 26 |
+
| Base model | Qwen/Qwen3-4B |
|
| 27 |
+
| Tensor type | BF16 |
|
| 28 |
+
| License | Apache-2.0 |
|
| 29 |
+
|
| 30 |
+
## Usage
|
| 31 |
+
|
| 32 |
+
```python
|
| 33 |
+
from sentence_transformers import CrossEncoder
|
| 34 |
+
|
| 35 |
+
model = CrossEncoder("devrev/zerank-1-small-seq")
|
| 36 |
+
|
| 37 |
+
query = "Which planet is known as the Red Planet?"
|
| 38 |
+
passages = [
|
| 39 |
+
"Venus is often called Earth's twin...",
|
| 40 |
+
"Mars, known for its reddish appearance, is often referred to as the Red Planet.",
|
| 41 |
+
]
|
| 42 |
+
|
| 43 |
+
scores = model.predict([(query, passage) for passage in passages])
|
| 44 |
+
print(scores)
|
| 45 |
+
```
|
| 46 |
+
|
| 47 |
+
## Performance (NDCG@10)
|
| 48 |
+
|
| 49 |
+
| Domain | Score |
|
| 50 |
+
|--------|-------|
|
| 51 |
+
| Finance | 0.861 |
|
| 52 |
+
| Legal | 0.817 |
|
| 53 |
+
| Medical | 0.773 |
|
| 54 |
+
| Code | 0.730 |
|
| 55 |
+
| STEM | 0.680 |
|
| 56 |
+
| Conversational | 0.556 |
|
| 57 |
+
|
| 58 |
+
|