model update
Browse files
README.md
CHANGED
@@ -2,7 +2,7 @@
|
|
2 |
datasets:
|
3 |
- relbert/semeval2012_relational_similarity
|
4 |
model-index:
|
5 |
-
- name: relbert/
|
6 |
results:
|
7 |
- task:
|
8 |
name: Relation Mapping
|
@@ -14,7 +14,7 @@ model-index:
|
|
14 |
metrics:
|
15 |
- name: Accuracy
|
16 |
type: accuracy
|
17 |
-
value:
|
18 |
- task:
|
19 |
name: Analogy Questions (SAT full)
|
20 |
type: multiple-choice-qa
|
@@ -153,27 +153,27 @@ model-index:
|
|
153 |
value: 0.9081322080316404
|
154 |
|
155 |
---
|
156 |
-
# relbert/
|
157 |
|
158 |
RelBERT fine-tuned from [roberta-large](https://huggingface.co/roberta-large) on
|
159 |
[relbert/semeval2012_relational_similarity](https://huggingface.co/datasets/relbert/semeval2012_relational_similarity).
|
160 |
Fine-tuning is done via [RelBERT](https://github.com/asahi417/relbert) library (see the repository for more detail).
|
161 |
It achieves the following results on the relation understanding tasks:
|
162 |
-
- Analogy Question ([dataset](https://huggingface.co/datasets/relbert/analogy_questions), [full result](https://huggingface.co/relbert/
|
163 |
- Accuracy on SAT (full): 0.6363636363636364
|
164 |
- Accuracy on SAT: 0.6261127596439169
|
165 |
- Accuracy on BATS: 0.8210116731517509
|
166 |
- Accuracy on U2: 0.6403508771929824
|
167 |
- Accuracy on U4: 0.6574074074074074
|
168 |
- Accuracy on Google: 0.906
|
169 |
-
- Lexical Relation Classification ([dataset](https://huggingface.co/datasets/relbert/lexical_relation_classification), [full result](https://huggingface.co/relbert/
|
170 |
- Micro F1 score on BLESS: 0.9180352568931747
|
171 |
- Micro F1 score on CogALexV: 0.8791079812206573
|
172 |
- Micro F1 score on EVALution: 0.6998916576381365
|
173 |
- Micro F1 score on K&H+N: 0.9588926758016276
|
174 |
- Micro F1 score on ROOT09: 0.9109996866186149
|
175 |
-
- Relation Mapping ([dataset](https://huggingface.co/datasets/relbert/relation_mapping), [full result](https://huggingface.co/relbert/
|
176 |
-
- Accuracy on Relation Mapping:
|
177 |
|
178 |
|
179 |
### Usage
|
@@ -184,7 +184,7 @@ pip install relbert
|
|
184 |
and activate model as below.
|
185 |
```python
|
186 |
from relbert import RelBERT
|
187 |
-
model = RelBERT("relbert/
|
188 |
vector = model.get_embedding(['Tokyo', 'Japan']) # shape of (1024, )
|
189 |
```
|
190 |
|
@@ -211,7 +211,7 @@ The following hyperparameters were used during training:
|
|
211 |
- n_sample: 640
|
212 |
- gradient_accumulation: 8
|
213 |
|
214 |
-
The full configuration can be found at [fine-tuning parameter file](https://huggingface.co/relbert/
|
215 |
|
216 |
### Reference
|
217 |
If you use any resource from RelBERT, please consider to cite our [paper](https://aclanthology.org/2021.eacl-demos.7/).
|
|
|
2 |
datasets:
|
3 |
- relbert/semeval2012_relational_similarity
|
4 |
model-index:
|
5 |
+
- name: relbert/roberta-large-semeval2012-average-prompt-c-loob
|
6 |
results:
|
7 |
- task:
|
8 |
name: Relation Mapping
|
|
|
14 |
metrics:
|
15 |
- name: Accuracy
|
16 |
type: accuracy
|
17 |
+
value: 0.9857142857142858
|
18 |
- task:
|
19 |
name: Analogy Questions (SAT full)
|
20 |
type: multiple-choice-qa
|
|
|
153 |
value: 0.9081322080316404
|
154 |
|
155 |
---
|
156 |
+
# relbert/roberta-large-semeval2012-average-prompt-c-loob
|
157 |
|
158 |
RelBERT fine-tuned from [roberta-large](https://huggingface.co/roberta-large) on
|
159 |
[relbert/semeval2012_relational_similarity](https://huggingface.co/datasets/relbert/semeval2012_relational_similarity).
|
160 |
Fine-tuning is done via [RelBERT](https://github.com/asahi417/relbert) library (see the repository for more detail).
|
161 |
It achieves the following results on the relation understanding tasks:
|
162 |
+
- Analogy Question ([dataset](https://huggingface.co/datasets/relbert/analogy_questions), [full result](https://huggingface.co/relbert/roberta-large-semeval2012-average-prompt-c-loob/raw/main/analogy.json)):
|
163 |
- Accuracy on SAT (full): 0.6363636363636364
|
164 |
- Accuracy on SAT: 0.6261127596439169
|
165 |
- Accuracy on BATS: 0.8210116731517509
|
166 |
- Accuracy on U2: 0.6403508771929824
|
167 |
- Accuracy on U4: 0.6574074074074074
|
168 |
- Accuracy on Google: 0.906
|
169 |
+
- Lexical Relation Classification ([dataset](https://huggingface.co/datasets/relbert/lexical_relation_classification), [full result](https://huggingface.co/relbert/roberta-large-semeval2012-average-prompt-c-loob/raw/main/classification.json)):
|
170 |
- Micro F1 score on BLESS: 0.9180352568931747
|
171 |
- Micro F1 score on CogALexV: 0.8791079812206573
|
172 |
- Micro F1 score on EVALution: 0.6998916576381365
|
173 |
- Micro F1 score on K&H+N: 0.9588926758016276
|
174 |
- Micro F1 score on ROOT09: 0.9109996866186149
|
175 |
+
- Relation Mapping ([dataset](https://huggingface.co/datasets/relbert/relation_mapping), [full result](https://huggingface.co/relbert/roberta-large-semeval2012-average-prompt-c-loob/raw/main/relation_mapping.json)):
|
176 |
+
- Accuracy on Relation Mapping: 0.9857142857142858
|
177 |
|
178 |
|
179 |
### Usage
|
|
|
184 |
and activate model as below.
|
185 |
```python
|
186 |
from relbert import RelBERT
|
187 |
+
model = RelBERT("relbert/roberta-large-semeval2012-average-prompt-c-loob")
|
188 |
vector = model.get_embedding(['Tokyo', 'Japan']) # shape of (1024, )
|
189 |
```
|
190 |
|
|
|
211 |
- n_sample: 640
|
212 |
- gradient_accumulation: 8
|
213 |
|
214 |
+
The full configuration can be found at [fine-tuning parameter file](https://huggingface.co/relbert/roberta-large-semeval2012-average-prompt-c-loob/raw/main/trainer_config.json).
|
215 |
|
216 |
### Reference
|
217 |
If you use any resource from RelBERT, please consider to cite our [paper](https://aclanthology.org/2021.eacl-demos.7/).
|