HyperAI초신경

Type Prediction On Manytypes4Typescript

평가 지표

Average Accuracy

Average F1

Average Precision

Average Recall

평가 결과

이 벤치마크에서 각 모델의 성능 결과

					Paper Title
CodeTIDAL5	71.27	-	-	-	Learning Type Inference for Enhanced Dataflow Analysis
GraphCodeBERT-MT4TS	63.42	-	-	-	ManyTypes4TypeScript: A Comprehensive TypeScript Dataset for Sequence-Based Type Inference
GraphCodeBERT	62.51	60.57	60.06	61.08	GraphCodeBERT: Pre-training Code Representations with Data Flow
CodeBERT	61.72	59.57	59.34	59.80	CodeBERT: A Pre-Trained Model for Programming and Natural Languages
PolyGot	61.29	58.86	58.81	58.91	Multilingual training for Software Engineering
GraphPolyGot	61.00	58.63	58.36	58.91	Multilingual training for Software Engineering
RoBERTa	59.84	57.54	57.45	57.62	RoBERTa: A Robustly Optimized BERT Pretraining Approach
CodeBERTa	59.81	56.71	56.57	56.85	-
BERT	57.52	54.10	54.18	54.02	BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding

0 of 9 row(s) selected.