Update README.md
Browse files
README.md
CHANGED
|
@@ -4,18 +4,11 @@ datasets:
|
|
| 4 |
- allenai/scirepeval
|
| 5 |
---
|
| 6 |
|
| 7 |
-
---
|
| 8 |
-
license: apache-2.0
|
| 9 |
-
datasets:
|
| 10 |
-
- allenai/scirepeval
|
| 11 |
-
language:
|
| 12 |
-
- en
|
| 13 |
-
---
|
| 14 |
|
| 15 |
**\*\*\*\*\*\*Update\*\*\*\*\*\***
|
| 16 |
|
| 17 |
This update introduces a new set of SPECTER 2.0 models with the base transformer encoder pre-trained on an extended citation dataset containing more recent papers.
|
| 18 |
-
For benchmarking purposes please use the existing SPECTER 2.0 models w/o the **aug2023refresh** suffix
|
| 19 |
|
| 20 |
# SPECTER 2.0 (Base)
|
| 21 |
SPECTER 2.0 is the successor to [SPECTER](https://huggingface.co/allenai/specter) and is capable of generating task specific embeddings for scientific tasks when paired with [adapters](https://huggingface.co/models?search=allenai/specter-2_).
|
|
@@ -66,12 +59,12 @@ It builds on the work done in [SciRepEval: A Multi-Format Benchmark for Scientif
|
|
| 66 |
|
| 67 |
|Model|Name and HF link|Description|
|
| 68 |
|--|--|--|
|
| 69 |
-
|
|
| 70 |
-
|Adhoc Query|[allenai/specter2_aug2023refresh_adhoc_query](https://huggingface.co/allenai/specter2_aug2023refresh_adhoc_query)|Encode short raw text queries for search tasks. (Candidate papers can be encoded with proximity)|
|
| 71 |
|Classification|[allenai/specter2_aug2023refresh_classification](https://huggingface.co/allenai/specter2_aug2023refresh_classification)|Encode papers to feed into linear classifiers as features|
|
| 72 |
|Regression|[allenai/specter2_aug2023refresh_regression](https://huggingface.co/allenai/specter2_aug2023refresh_regression)|Encode papers to feed into linear regressors as features|
|
| 73 |
|
| 74 |
-
*
|
| 75 |
|
| 76 |
```python
|
| 77 |
from transformers import AutoTokenizer, AutoModel
|
|
|
|
| 4 |
- allenai/scirepeval
|
| 5 |
---
|
| 6 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 7 |
|
| 8 |
**\*\*\*\*\*\*Update\*\*\*\*\*\***
|
| 9 |
|
| 10 |
This update introduces a new set of SPECTER 2.0 models with the base transformer encoder pre-trained on an extended citation dataset containing more recent papers.
|
| 11 |
+
For benchmarking purposes please use the existing SPECTER 2.0 [models](https://huggingface.co/allenai/specter2) w/o the **aug2023refresh** suffix.
|
| 12 |
|
| 13 |
# SPECTER 2.0 (Base)
|
| 14 |
SPECTER 2.0 is the successor to [SPECTER](https://huggingface.co/allenai/specter) and is capable of generating task specific embeddings for scientific tasks when paired with [adapters](https://huggingface.co/models?search=allenai/specter-2_).
|
|
|
|
| 59 |
|
| 60 |
|Model|Name and HF link|Description|
|
| 61 |
|--|--|--|
|
| 62 |
+
|Proximity*|[allenai/specter2_aug2023refresh](https://huggingface.co/allenai/specter2_aug2023refresh)|Encode papers as queries and candidates eg. Link Prediction, Nearest Neighbor Search|
|
| 63 |
+
|Adhoc Query|[allenai/specter2_aug2023refresh_adhoc_query](https://huggingface.co/allenai/specter2_aug2023refresh_adhoc_query)|Encode short raw text queries for search tasks. (Candidate papers can be encoded with the proximity adapter)|
|
| 64 |
|Classification|[allenai/specter2_aug2023refresh_classification](https://huggingface.co/allenai/specter2_aug2023refresh_classification)|Encode papers to feed into linear classifiers as features|
|
| 65 |
|Regression|[allenai/specter2_aug2023refresh_regression](https://huggingface.co/allenai/specter2_aug2023refresh_regression)|Encode papers to feed into linear regressors as features|
|
| 66 |
|
| 67 |
+
*Proximity model should suffice for downstream task types not mentioned above
|
| 68 |
|
| 69 |
```python
|
| 70 |
from transformers import AutoTokenizer, AutoModel
|