Our new X account is live! Follow @wizwand_team for updates
Home
/
Benchmarks
Text Style Transfer on NCBI style domain
Loading...
50.37
BLEU
Similar 5-shot finetuning w/ terminology and name retrieval
26.97
33.045
39.12
45.195
Feb 16, 2026
BLEU
Accuracy
Updated 4d ago
Evaluation Results
Method
Method
Links
BLEU
Accuracy
Similar 5-shot finetuning w/ terminology and name retrieval
Backbone=Llama-3.1-8B-...
2026.02
50.37
87.2
Similar 5-shot finetuning
Backbone=Llama-3.1-8B-...
2026.02
49.96
83.1
Similar 3-shot finetuning
Backbone=Llama-3.1-8B-...
2026.02
49.01
77.6
Random 5-shot finetuning
Backbone=Llama-3.1-8B-...
2026.02
46.3
89.6
Random 3-shot finetuning
Backbone=Llama-3.1-8B-...
2026.02
42.07
82.3
Zero-shot finetuning
Backbone=Llama-3.1-8B-...
2026.02
39.3
74.2
APE with Marian
Backbone=Marian
2026.02
35.95
65.9
5-shot ICL w/ terminology and name retrieval
Backbone=Llama-3.1-8B-...
2026.02
29.31
58.6
5-shot ICL
Backbone=Llama-3.1-8B-...
2026.02
27.87
46.2
Feedback
Search any
task
Search any
task