Text Classification
Transformers
Safetensors
PEFT
English
retrievalrouter
feature-extraction
retrieval
document-retrieval
information-retrieval
routing
RAG
query-routing
late-interaction
lora
baseline
custom_code
Instructions to use emrekuruu/RetrievalRouter_Baseline with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use emrekuruu/RetrievalRouter_Baseline with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-classification", model="emrekuruu/RetrievalRouter_Baseline", trust_remote_code=True)# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("emrekuruu/RetrievalRouter_Baseline", trust_remote_code=True, device_map="auto") - PEFT
How to use emrekuruu/RetrievalRouter_Baseline with PEFT:
Task type is invalid.
- Notebooks
- Google Colab
- Kaggle
Fix baseline citation (Arabzadeh et al., CIKM 2021)
Browse files
README.md
CHANGED
|
@@ -37,8 +37,7 @@ metrics:
|
|
| 37 |
|
| 38 |
Adaptive routing **baseline** from **RetrievalRouter: Joint Modality and Architecture Selection
|
| 39 |
for Document Retrieval** (EMNLP 2026). This is **not** the RetrievalRouter method — it is the
|
| 40 |
-
prior strategy-selection approach of
|
| 41 |
-
[Arabzadeh et al. (2021)](https://dl.acm.org/doi/10.1145/3459637.3482119), extended to our
|
| 42 |
five-pipeline setting and trained under an identical pipeline for a controlled comparison.
|
| 43 |
|
| 44 |
- 📄 Paper: https://arxiv.org/pdf/2608.23176
|
|
@@ -107,6 +106,9 @@ selected pipeline against your own indices.
|
|
| 107 |
|
| 108 |
## Citation
|
| 109 |
|
|
|
|
|
|
|
|
|
|
| 110 |
```bibtex
|
| 111 |
@inproceedings{kuru2026retrievalrouter,
|
| 112 |
title = {RetrievalRouter: Joint Modality and Architecture Selection for Document Retrieval},
|
|
@@ -114,4 +116,12 @@ selected pipeline against your own indices.
|
|
| 114 |
booktitle = {Proceedings of EMNLP 2026},
|
| 115 |
year = {2026}
|
| 116 |
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 117 |
```
|
|
|
|
| 37 |
|
| 38 |
Adaptive routing **baseline** from **RetrievalRouter: Joint Modality and Architecture Selection
|
| 39 |
for Document Retrieval** (EMNLP 2026). This is **not** the RetrievalRouter method — it is the
|
| 40 |
+
prior strategy-selection approach of Arabzadeh et al. (CIKM 2021), extended to our
|
|
|
|
| 41 |
five-pipeline setting and trained under an identical pipeline for a controlled comparison.
|
| 42 |
|
| 43 |
- 📄 Paper: https://arxiv.org/pdf/2608.23176
|
|
|
|
| 106 |
|
| 107 |
## Citation
|
| 108 |
|
| 109 |
+
If you use this baseline, please cite both the RetrievalRouter paper and the original
|
| 110 |
+
strategy-selection method it is based on:
|
| 111 |
+
|
| 112 |
```bibtex
|
| 113 |
@inproceedings{kuru2026retrievalrouter,
|
| 114 |
title = {RetrievalRouter: Joint Modality and Architecture Selection for Document Retrieval},
|
|
|
|
| 116 |
booktitle = {Proceedings of EMNLP 2026},
|
| 117 |
year = {2026}
|
| 118 |
}
|
| 119 |
+
|
| 120 |
+
@inproceedings{arabzadeh2021predicting,
|
| 121 |
+
title = {Predicting efficiency/effectiveness trade-offs for dense vs. sparse retrieval strategy selection},
|
| 122 |
+
author = {Arabzadeh, Negar and Yan, Xinyi and Clarke, Charles L. A.},
|
| 123 |
+
booktitle = {Proceedings of the 30th ACM International Conference on Information \& Knowledge Management (CIKM)},
|
| 124 |
+
pages = {2862--2866},
|
| 125 |
+
year = {2021}
|
| 126 |
+
}
|
| 127 |
```
|