Semantic Textual Similarity on ATEC
0.4816Pearson CorrelationLLM-ce-ft(7B)
Evaluation results
| Method | Links | ||
|---|---|---|---|
| LLM-ce-ft(7B)#Data=0.3M, #Param.=7B, Backbone=Qwen-7B-Chat2024.06 | 0.4816 | 0.4898 | |
| BGE(326M)#Data=100M, #Param.=326M, Backbone=Chinese-roberta-large2024.06 | 0.4716 | 0.4785 | |
| D2LLM-ft(7B)#Data=0.3M, #Param.=7B, Backbone=Qwen-7B-Chat2024.06 | 0.4603 | 0.4759 | |
| BGE-ft(326M)#Data=0.3M, #Param.=326M, Backbone=Chinese-roberta-large2024.06 | 0.3827 | 0.4126 | |
| D2LLM(7B)#Data=0.3M, #Param.=7B, Backbone=Qwen-7B-Chat2024.06 | 0.3731 | 0.3994 | |
| LLM-ce(7B)Backbone=Qwen-7B-Chat2024.06 | 0.367 | 0.4152 | |
| LLaRRA(7B)#Data=0.3M, #Param.=7B, Backbone=Qwen-7B-Chat2024.06 | 0.343 | 0.3575 | |
| Udever(7B)#Data=0.3M, #Param.=7B, Backbone=Qwen-7B-Chat2024.06 | 0.3328 | 0.3602 | |
| SGPT(7B)#Data=0.3M, #Param.=7B, Backbone=Qwen-7B-Chat2024.06 | 0.3045 | 0.3173 | |
| LLM-be(7B)Backbone=Qwen-7B-Chat2024.06 | 0.2339 | 0.2178 | |
| RocketQAv2(326M)#Data=0.3M, #Param.=326M, Backbone=Chinese-roberta-large2024.06 | 0.1971 | 0.2362 |