Semantic Textual Similarity On Mrpc

평가 지표

평가 결과

이 벤치마크에서 각 모델의 성능 결과

비교 표

모델 이름	F1
big-bird-transformers-for-longer-sequences	91.5
exploring-the-limits-of-transfer-learning	92.5
mobilebert-a-compact-task-agnostic-bert-for	-
intrinsic-dimensionality-explains-the	-
charformer-fast-character-transformers-via	91.4
entailment-as-few-shot-learner	91.0
nystromformer-a-nystrom-based-algorithm-for	88.1%
smart-robust-and-efficient-fine-tuning-for	-
llm-int8-8-bit-matrix-multiplication-for	-
fnet-mixing-tokens-with-fourier-transforms	-
squeezebert-what-can-computer-vision-teach	-
xlnet-generalized-autoregressive-pretraining	-
smart-robust-and-efficient-fine-tuning-for	-
exploring-the-limits-of-transfer-learning	92.4
190910351	-
190910351	-
distilbert-a-distilled-version-of-bert	-
ernie-20-a-continual-pre-training-framework	-
exploring-the-limits-of-transfer-learning	89.7
q8bert-quantized-8bit-bert	-
a-statistical-framework-for-low-bitwidth	-
albert-a-lite-bert-for-self-supervised	-
exploring-the-limits-of-transfer-learning	91.9
subregweigh-effective-and-efficient	-
informer-transformer-likes-informed-attention	90.91%
intrinsic-dimensionality-explains-the	-
smart-robust-and-efficient-fine-tuning-for	-
discriminative-improvements-to-distributional	85.9%
clear-contrastive-learning-for-sentence	-
autobert-zero-evolving-bert-backbone-from	-
ernie-20-a-continual-pre-training-framework	-
learning-general-purpose-distributed-sentence	84.4%
q-bert-hessian-based-ultra-low-precision	-
supervised-learning-of-universal-sentence	83.1%
spanbert-improving-pre-training-by	-
bert-pre-training-of-deep-bidirectional	89.3
smart-robust-and-efficient-fine-tuning-for	91.7
structbert-incorporating-language-structures	93.6%
roberta-a-robustly-optimized-bert-pretraining	-
autobert-zero-evolving-bert-backbone-from	-
learning-to-encode-position-for-transformer	-
모델 42	-
how-to-train-bert-with-an-academic-budget	-
exploring-the-limits-of-transfer-learning	90.7
ernie-enhanced-language-representation-with	-