Update README.md
Browse files
README.md
CHANGED
|
@@ -29,7 +29,9 @@ base_model: colbert-ir/colbertv1.9
|
|
| 29 |
- Fine-tuned ColBERT weights starting from `colbert-ir/colbertv1.9`.
|
| 30 |
- Trained with **triples JSONL** (`[qid, pid+, pid-]`) using **TSV** `queries.tsv` and `collection.tsv` (IDs + text).
|
| 31 |
- Default training hyperparameters are listed below (batch size, lr, doc_maxlen, dim, etc.).
|
| 32 |
-
|
|
|
|
|
|
|
| 33 |
---
|
| 34 |
|
| 35 |
## 🔧 Quickstart
|
|
|
|
| 29 |
- Fine-tuned ColBERT weights starting from `colbert-ir/colbertv1.9`.
|
| 30 |
- Trained with **triples JSONL** (`[qid, pid+, pid-]`) using **TSV** `queries.tsv` and `collection.tsv` (IDs + text).
|
| 31 |
- Default training hyperparameters are listed below (batch size, lr, doc_maxlen, dim, etc.).
|
| 32 |
+
- This checkpoint and the associated contrastive training data are part of the work: [`NLKI: A lightweight Natural Language Knowledge Integration Framework
|
| 33 |
+
for Improving Small VLMs in Commonsense VQA Tasks`](https://arxiv.org/pdf/2508.19724)
|
| 34 |
+
- All copyrights for the training data are retained by their original owners; we do not claim ownership.
|
| 35 |
---
|
| 36 |
|
| 37 |
## 🔧 Quickstart
|