asahi417 commited on
Commit
d2e32d1
1 Parent(s): 787bf17

model update

Browse files
Files changed (1) hide show
  1. README.md +7 -7
README.md CHANGED
@@ -2,7 +2,7 @@
2
  datasets:
3
  - relbert/semeval2012_relational_similarity
4
  model-index:
5
- - name: relbert/relbert-roberta-large-nce-d-semeval2012
6
  results:
7
  - task:
8
  name: Relation Mapping
@@ -186,11 +186,11 @@ model-index:
186
  value: 0.9040831832304929
187
 
188
  ---
189
- # relbert/relbert-roberta-large-nce-d-semeval2012
190
 
191
  RelBERT based on [roberta-large](https://huggingface.co/roberta-large) fine-tuned on [relbert/semeval2012_relational_similarity](https://huggingface.co/datasets/relbert/semeval2012_relational_similarity) (see the [`relbert`](https://github.com/asahi417/relbert) for more detail of fine-tuning).
192
  This model achieves the following results on the relation understanding tasks:
193
- - Analogy Question ([dataset](https://huggingface.co/datasets/relbert/analogy_questions), [full result](https://huggingface.co/relbert/relbert-roberta-large-nce-d-semeval2012/raw/main/analogy.forward.json)):
194
  - Accuracy on SAT (full): 0.732620320855615
195
  - Accuracy on SAT: 0.7359050445103857
196
  - Accuracy on BATS: 0.8093385214007782
@@ -200,13 +200,13 @@ This model achieves the following results on the relation understanding tasks:
200
  - Accuracy on ConceptNet Analogy: 0.4748322147651007
201
  - Accuracy on T-Rex Analogy: 0.644808743169399
202
  - Accuracy on NELL-ONE Analogy: 0.6583333333333333
203
- - Lexical Relation Classification ([dataset](https://huggingface.co/datasets/relbert/lexical_relation_classification), [full result](https://huggingface.co/relbert/relbert-roberta-large-nce-d-semeval2012/raw/main/classification.json)):
204
  - Micro F1 score on BLESS: 0.9199939731806539
205
  - Micro F1 score on CogALexV: 0.8497652582159625
206
  - Micro F1 score on EVALution: 0.6836403033586133
207
  - Micro F1 score on K&H+N: 0.9563191208179731
208
  - Micro F1 score on ROOT09: 0.9041052961454089
209
- - Relation Mapping ([dataset](https://huggingface.co/datasets/relbert/relation_mapping), [full result](https://huggingface.co/relbert/relbert-roberta-large-nce-d-semeval2012/raw/main/relation_mapping.json)):
210
  - Accuracy on Relation Mapping: 0.8049007936507937
211
 
212
 
@@ -218,7 +218,7 @@ pip install relbert
218
  and activate model as below.
219
  ```python
220
  from relbert import RelBERT
221
- model = RelBERT("relbert/relbert-roberta-large-nce-d-semeval2012")
222
  vector = model.get_embedding(['Tokyo', 'Japan']) # shape of (n_dim, )
223
  ```
224
 
@@ -242,7 +242,7 @@ vector = model.get_embedding(['Tokyo', 'Japan']) # shape of (n_dim, )
242
  - loss_function_config: {'temperature': 0.05, 'gradient_accumulation': 1, 'num_negative': 400, 'num_positive': 10}
243
  - augment_negative_by_positive: True
244
 
245
- See the full configuration at [config file](https://huggingface.co/relbert/relbert-roberta-large-nce-d-semeval2012/raw/main/finetuning_config.json).
246
 
247
  ### Reference
248
  If you use any resource from RelBERT, please consider to cite our [paper](https://aclanthology.org/2021.emnlp-main.712/).
 
2
  datasets:
3
  - relbert/semeval2012_relational_similarity
4
  model-index:
5
+ - name: relbert/relbert-roberta-base-nce-semeval2012-0-400
6
  results:
7
  - task:
8
  name: Relation Mapping
 
186
  value: 0.9040831832304929
187
 
188
  ---
189
+ # relbert/relbert-roberta-base-nce-semeval2012-0-400
190
 
191
  RelBERT based on [roberta-large](https://huggingface.co/roberta-large) fine-tuned on [relbert/semeval2012_relational_similarity](https://huggingface.co/datasets/relbert/semeval2012_relational_similarity) (see the [`relbert`](https://github.com/asahi417/relbert) for more detail of fine-tuning).
192
  This model achieves the following results on the relation understanding tasks:
193
+ - Analogy Question ([dataset](https://huggingface.co/datasets/relbert/analogy_questions), [full result](https://huggingface.co/relbert/relbert-roberta-base-nce-semeval2012-0-400/raw/main/analogy.forward.json)):
194
  - Accuracy on SAT (full): 0.732620320855615
195
  - Accuracy on SAT: 0.7359050445103857
196
  - Accuracy on BATS: 0.8093385214007782
 
200
  - Accuracy on ConceptNet Analogy: 0.4748322147651007
201
  - Accuracy on T-Rex Analogy: 0.644808743169399
202
  - Accuracy on NELL-ONE Analogy: 0.6583333333333333
203
+ - Lexical Relation Classification ([dataset](https://huggingface.co/datasets/relbert/lexical_relation_classification), [full result](https://huggingface.co/relbert/relbert-roberta-base-nce-semeval2012-0-400/raw/main/classification.json)):
204
  - Micro F1 score on BLESS: 0.9199939731806539
205
  - Micro F1 score on CogALexV: 0.8497652582159625
206
  - Micro F1 score on EVALution: 0.6836403033586133
207
  - Micro F1 score on K&H+N: 0.9563191208179731
208
  - Micro F1 score on ROOT09: 0.9041052961454089
209
+ - Relation Mapping ([dataset](https://huggingface.co/datasets/relbert/relation_mapping), [full result](https://huggingface.co/relbert/relbert-roberta-base-nce-semeval2012-0-400/raw/main/relation_mapping.json)):
210
  - Accuracy on Relation Mapping: 0.8049007936507937
211
 
212
 
 
218
  and activate model as below.
219
  ```python
220
  from relbert import RelBERT
221
+ model = RelBERT("relbert/relbert-roberta-base-nce-semeval2012-0-400")
222
  vector = model.get_embedding(['Tokyo', 'Japan']) # shape of (n_dim, )
223
  ```
224
 
 
242
  - loss_function_config: {'temperature': 0.05, 'gradient_accumulation': 1, 'num_negative': 400, 'num_positive': 10}
243
  - augment_negative_by_positive: True
244
 
245
+ See the full configuration at [config file](https://huggingface.co/relbert/relbert-roberta-base-nce-semeval2012-0-400/raw/main/finetuning_config.json).
246
 
247
  ### Reference
248
  If you use any resource from RelBERT, please consider to cite our [paper](https://aclanthology.org/2021.emnlp-main.712/).