Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Language Modeling on WikiText-2 (test) (Llama-3.1-8B-Instruct)

7.2Perplexity

BF16 Baseline

1.56239.618577.675115.7315Jul 24, 2025
Updated 1mo ago

Evaluation Results

MethodLinks
2025.07
7.2
2025.07
7.37
2025.07
7.38
2025.07
7.39
2025.07
7.41
2025.07
7.48
2025.07
7.5
2025.07
7.56
2025.07
7.84
2025.07
7.98
2025.07
8.01
2025.07
8.16
2025.07
8.32
2025.07
8.67
2025.07
9.44
2025.07
11.89
2025.07
12.12
2025.07
13.22
2025.07
17.27
2025.07
25.18
2025.07
57.26
2025.07
148.15