Downloads · 30 days
14
0% of all-time downloads
tct_colbert-v2-hnp-msmarco-r2
castorini/tct_colbert-v2-hnp-msmarco-r2
tct_colbert-v2-hnp-msmarco-r2 is a feature extraction model from castorini. Use it when you need embeddings to search or compare text. It is set up for transformers.
This model is to reproduce a variant of TCT-ColBERT-V2 dense retrieval models described in the following paper: Sheng-Chieh Lin, Jheng-Hong Yang, and Jimmy Lin. In-Batch Negatives for Knowledge Distillation with Tight…
All-time downloads
13.2K
Public
Repo size
876 MB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.bin438 MB · 100%
At a glance
- Task
- Feature Extraction
- Library
- transformers
- Model type
- bert
- Access
- Public
- Created
- Mar 2, 2022
- Updated
- Aug 16, 2021
- SHA
- 12eead8e
Model card
From the Hugging Face model README
This model is to reproduce a variant of TCT-ColBERT-V2 dense retrieval models described in the following paper:
Sheng-Chieh Lin, Jheng-Hong Yang, and Jimmy Lin. In-Batch Negatives for Knowledge Distillation with Tightly-CoupledTeachers for Dense Retrieval. RepL4NLP 2021.
Specifically, this checkpoint is finetuned for MS MARCO-V2 passage ranking, and we use this checkpoint as our ``trained'' model for TREC DL 2021 submissions. The initial checkpoint is from a previous one tct_colbert-v2-hnp-msmarco trained on MS MARCO. For fine-tuning, we construct our training data for MS MARCO-V2 passage ranking using this script.
- Task
- Feature Extraction
- Library
- transformers
- Type
- bert
- Created
- Mar 2, 2022
- Updated
- Aug 16, 2021