Modelpublic

ulmentflam/gpt2-774m-fineweb-mojo

sourceHugging Faceodc-byupdated 2mo agoView on Hugging Face

0likes23downloads

3 commits on main

db061fe2mo ago

v2 re-run: trained biases restored (dbias scratch overrun fixed)

ulmentflam

a4297112mo ago

GPT-2 774M FineWeb-10B from-scratch llm.mojo run (val loss 3.0130, HellaSwag 36.34%)

ulmentflam

098c1432mo ago

initial commit

ulmentflam