List your company

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

3.372Validation LossDelta AttnRes
3.360643.437323.5143.59068May 13, 2026

Evaluation results

MethodLinks
Delta AttnResScale=1044M, d=1280, L=36, Tok/s=34k, Mem (GB)=77.72026.053.37229.13
Delta BlockScale=1044M, d=1280, L=36, Tok/s=86k, Mem (GB)=28.42026.053.37429.19
BaselineScale=1044M, d=1280, L=36, Tok/s=108k, Mem (GB)=22.52026.053.39129.7
Delta AttnResScale=533M, d=1024, L=24, Tok/s=80k, Mem (GB)=33.62026.053.43631.05
Delta BlockScale=533M, d=1024, L=24, Tok/s=198k, Mem (GB)=16.12026.053.43931.16
Full AttnResScale=533M, d=1024, L=24, Tok/s=103k, Mem (GB)=24.82026.053.45631.68
AttnResScale=533M, d=1024, L=24, Tok/s=174k, Mem (GB)=17.52026.053.45831.75
AttnResScale=1044M, d=1280, L=36, Tok/s=71k, Mem (GB)=31.62026.053.45831.76
BaselineScale=533M, d=1024, L=24, Tok/s=243k, Mem (GB)=13.92026.053.46632
Full AttnResScale=1044M, d=1280, L=36, Tok/s=49k, Mem (GB)=52.02026.053.50833.36
Delta AttnResScale=220M, d=768, L=12, Tok/s=246k, Mem (GB)=7.52026.053.60636.83
Delta BlockScale=220M, d=768, L=12, Tok/s=306k, Mem (GB)=5.92026.053.61337.08
Full AttnResScale=220M, d=768, L=12, Tok/s=285k, Mem (GB)=6.82026.053.61937.3
AttnResScale=220M, d=768, L=12, Tok/s=294k, Mem (GB)=6.22026.053.62237.39
BaselineScale=220M, d=768, L=12, Tok/s=289k, Mem (GB)=5.62026.053.65638.71