List your company
ResearchFormerly WizWand

Relation Extraction on New York Times (Prec, F1)

92.4PrecisionSFT_Llama
11.446432.463253.4874.4968Apr 2, 2026
Updated 7mo ago

Evaluation results

MethodLinks
SFT_LlamaNote=Task-specific upper bound2026.0492.491
SFT_QwenNote=Task-specific upper bound2026.0490.8589.5
DARE (TIES+)Transfer=Qwen to Llama2026.0488.9281.81
DARE (TIES+)Source Models=Qwen, Mistral, Target Model=Llama2026.0488.9281.81
DELLATransfer=Qwen to Llama2026.0488.981.82
DELLASource Models=Qwen, Mistral, Target Model=Llama2026.0488.981.82
TIES-MergingTransfer=Qwen to Llama2026.0488.4781.24
TIES-MergingSource Models=Qwen, Mistral, Target Model=Llama2026.0488.4781.24
EMR-MergingTransfer=Qwen to Llama2026.0479.9271.06
EMR-MergingSource Models=Qwen, Mistral, Target Model=Llama2026.0479.9271.06
DARE (TIES+)Target Model=Qwen2.5-7B-Instruct, Source Model (this task)=Llama-3.1-8B-Instruct2026.0473.9869.95
Task ArithmeticTransfer=Qwen to Llama2026.0473.6165.04
Task ArithmeticSource Models=Qwen, Mistral, Target Model=Llama2026.0473.6165.04
DELLATarget Model=Qwen2.5-7B-Instruct, Source Model (this task)=Llama-3.1-8B-Instruct2026.0473.4369.57
BreadcrumbsTransfer=Qwen to Llama2026.0468.8860.99
BreadcrumbsSource Models=Qwen, Mistral, Target Model=Llama2026.0468.8860.99
HeteroFusionTransfer=Qwen to Llama2026.0468.4662.26
HeteroFusionSource Models=Qwen, Mistral, Target Model=Llama2026.0467.8560.78
Llama MergeTransfer=Qwen to Llama2026.0466.9163.05
HeteroFusionTarget Model=Qwen2.5-7B-Instruct, Source Model (this task)=Llama-3.1-8B-Instruct2026.0463.8757.75
TIES-MergingTarget Model=Qwen2.5-7B-Instruct, Source Model (this task)=Llama-3.1-8B-Instruct2026.0463.4759.12
FuseLLMTransfer=Qwen to Llama2026.0460.9352.76
EMR-MergingTarget Model=Qwen2.5-7B-Instruct, Source Model (this task)=Llama-3.1-8B-Instruct2026.0458.1953.12
FuseLLMSource Models=Qwen, Mistral, Target Model=Llama2026.0457.3848.95
FuseLLMTarget Model=Qwen2.5-7B-Instruct, Source Model (this task)=Llama-3.1-8B-Instruct2026.0447.1238.87
Task ArithmeticTarget Model=Qwen2.5-7B-Instruct, Source Model (this task)=Llama-3.1-8B-Instruct2026.0427.9226.45
GACSource Models=Qwen, Mistral, Target Model=Llama2026.0423.3324.78
BreadcrumbsTarget Model=Qwen2.5-7B-Instruct, Source Model (this task)=Llama-3.1-8B-Instruct2026.0422.9722.17
UniTETarget Model=Qwen2.5-7B-Instruct, Source Model (this task)=Llama-3.1-8B-Instruct2026.0418.3318.15
UniTESource Models=Qwen, Mistral, Target Model=Llama2026.0418.3318.15
UniTETransfer=Qwen to Llama2026.0414.6815.87
GACTransfer=Qwen to Llama2026.0414.5615.51