Scores: Accuracy / F1 Score / BERTScore (F1).

Model Grouping Model Name FiNERopen_in_new FNXLopen_in_new XBRL Termopen_in_new
Financial Models SEC-BERTopen_in_new - 46.20 -
Base Models Llama 3.1 8Bopen_in_new 21.28
0.232
3.64
0.045
0.574
Llama 3.1 70Bopen_in_new 61.82
0.606
20.14
0.210
0.587
DeepSeek V3open_in_new 68.92
0.699
27.33
0.288
0.573
GPT-4oopen_in_new 72.29
0.725
42.41
0.398
0.584
Gemini 2.0 FLopen_in_new 63.91
0.638
37.75
0.356
0.572
Fine-tuned Models Llama 3.1 8B LoRAopen_in_new 74.10
0.759
23.57
0.250
0.599
Llama 3.1 8B QLoRAopen_in_new 74.32
0.760
23.05
0.253
0.606
Llama 3.1 8B DoRAopen_in_new 70.92
0.732
33.50
0.311
0.606
Llama 3.1 8B rsLoRAopen_in_new 70.72
0.724
35.68
0.348
0.630
Gemini 2.0 FL N/Aopen_in_new 80.32
0.802
47.98
0.438
0.666