metadata
datasets:
- relbert/conceptnet_high_confidence
model-index:
- name: relbert/roberta-large-conceptnet-hc-mask-prompt-b-nce
results:
- task:
name: Relation Mapping
type: sorting-task
dataset:
name: Relation Mapping
args: relbert/relation_mapping
type: relation-mapping
metrics:
- name: Accuracy
type: accuracy
value: 0.844484126984127
- task:
name: Analogy Questions (SAT full)
type: multiple-choice-qa
dataset:
name: SAT full
args: relbert/analogy_questions
type: analogy-questions
metrics:
- name: Accuracy
type: accuracy
value: 0.5026737967914439
- task:
name: Analogy Questions (SAT)
type: multiple-choice-qa
dataset:
name: SAT
args: relbert/analogy_questions
type: analogy-questions
metrics:
- name: Accuracy
type: accuracy
value: 0.5074183976261127
- task:
name: Analogy Questions (BATS)
type: multiple-choice-qa
dataset:
name: BATS
args: relbert/analogy_questions
type: analogy-questions
metrics:
- name: Accuracy
type: accuracy
value: 0.7837687604224569
- task:
name: Analogy Questions (Google)
type: multiple-choice-qa
dataset:
name: Google
args: relbert/analogy_questions
type: analogy-questions
metrics:
- name: Accuracy
type: accuracy
value: 0.914
- task:
name: Analogy Questions (U2)
type: multiple-choice-qa
dataset:
name: U2
args: relbert/analogy_questions
type: analogy-questions
metrics:
- name: Accuracy
type: accuracy
value: 0.4868421052631579
- task:
name: Analogy Questions (U4)
type: multiple-choice-qa
dataset:
name: U4
args: relbert/analogy_questions
type: analogy-questions
metrics:
- name: Accuracy
type: accuracy
value: 0.5717592592592593
- task:
name: Lexical Relation Classification (BLESS)
type: classification
dataset:
name: BLESS
args: relbert/lexical_relation_classification
type: relation-classification
metrics:
- name: F1
type: f1
value: 0.9169805635076088
- name: F1 (macro)
type: f1_macro
value: 0.9124828189963239
- task:
name: Lexical Relation Classification (CogALexV)
type: classification
dataset:
name: CogALexV
args: relbert/lexical_relation_classification
type: relation-classification
metrics:
- name: F1
type: f1
value: 0.8615023474178404
- name: F1 (macro)
type: f1_macro
value: 0.6923470637031117
- task:
name: Lexical Relation Classification (EVALution)
type: classification
dataset:
name: BLESS
args: relbert/lexical_relation_classification
type: relation-classification
metrics:
- name: F1
type: f1
value: 0.6917659804983749
- name: F1 (macro)
type: f1_macro
value: 0.6818037583371511
- task:
name: Lexical Relation Classification (K&H+N)
type: classification
dataset:
name: K&H+N
args: relbert/lexical_relation_classification
type: relation-classification
metrics:
- name: F1
type: f1
value: 0.9652917854907144
- name: F1 (macro)
type: f1_macro
value: 0.8914930968868111
- task:
name: Lexical Relation Classification (ROOT09)
type: classification
dataset:
name: ROOT09
args: relbert/lexical_relation_classification
type: relation-classification
metrics:
- name: F1
type: f1
value: 0.9025383892196804
- name: F1 (macro)
type: f1_macro
value: 0.9012451685993444
relbert/roberta-large-conceptnet-hc-mask-prompt-b-nce
RelBERT fine-tuned from roberta-large on
relbert/conceptnet_high_confidence.
Fine-tuning is done via RelBERT library (see the repository for more detail).
It achieves the following results on the relation understanding tasks:
- Analogy Question (dataset, full result):
- Accuracy on SAT (full): 0.5026737967914439
- Accuracy on SAT: 0.5074183976261127
- Accuracy on BATS: 0.7837687604224569
- Accuracy on U2: 0.4868421052631579
- Accuracy on U4: 0.5717592592592593
- Accuracy on Google: 0.914
- Lexical Relation Classification (dataset, full result):
- Micro F1 score on BLESS: 0.9169805635076088
- Micro F1 score on CogALexV: 0.8615023474178404
- Micro F1 score on EVALution: 0.6917659804983749
- Micro F1 score on K&H+N: 0.9652917854907144
- Micro F1 score on ROOT09: 0.9025383892196804
- Relation Mapping (dataset, full result):
- Accuracy on Relation Mapping: 0.844484126984127
Usage
This model can be used through the relbert library. Install the library via pip
pip install relbert
and activate model as below.
from relbert import RelBERT
model = RelBERT("relbert/roberta-large-conceptnet-hc-mask-prompt-b-nce")
vector = model.get_embedding(['Tokyo', 'Japan']) # shape of (1024, )
Training hyperparameters
The following hyperparameters were used during training:
- model: roberta-large
- max_length: 64
- mode: mask
- data: relbert/conceptnet_high_confidence
- template_mode: manual
- template: Today, I finally discovered the relation between and : is 's
- loss_function: nce_logout
- temperature_nce_constant: 0.05
- temperature_nce_rank: {'min': 0.01, 'max': 0.05, 'type': 'linear'}
- epoch: 114
- batch: 128
- lr: 5e-06
- lr_decay: False
- lr_warmup: 1
- weight_decay: 0
- random_seed: 0
- exclude_relation: None
- n_sample: 640
- gradient_accumulation: 8
The full configuration can be found at fine-tuning parameter file.
Reference
If you use any resource from RelBERT, please consider to cite our paper.
@inproceedings{ushio-etal-2021-distilling-relation-embeddings,
title = "{D}istilling {R}elation {E}mbeddings from {P}re-trained {L}anguage {M}odels",
author = "Ushio, Asahi and
Schockaert, Steven and
Camacho-Collados, Jose",
booktitle = "EMNLP 2021",
year = "2021",
address = "Online",
publisher = "Association for Computational Linguistics",
}