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

Language Modeling on FineWeb-Edu 10K steps (val)

3.372Validation Loss

Delta AttnRes

3.360643.437323.5143.59068May 13, 2026
Updated 14d ago

Evaluation Results

MethodLinks
2026.05
3.37229.13
2026.05
3.37429.19
2026.05
3.39129.7
2026.05
3.43631.05
2026.05
3.43931.16
2026.05
3.45631.68
2026.05
3.45831.75
2026.05
3.45831.76
2026.05
3.46632
2026.05
3.50833.36
2026.05
3.60636.83
2026.05
3.61337.08
2026.05
3.61937.3
2026.05
3.62237.39
2026.05
3.65638.71