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

Language Modeling on FineWeb-Edu (Throughput and Speedup)

71,600Throughput (tokens/s)

AdamW

1,81619,93338,05056,167Mar 18, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.03
71,600-3.43
2026.03
70,700-5.89
2026.03
66,400-2.11
2026.03
52,2001.66-
2026.03
44,5002.13-
2026.03
31,400--
2026.03
30,6002.55-
2026.03
25,000-5.56
2026.03
20,900--
2026.03
13,900-2.24
2026.03
13,2002.93-
2026.03
12,000--
2026.03
10,9001.76-
2026.03
6,200--
2026.03
4,500--