Truncated Backpropagation Through Time (TBPTT) on Fineweb-EDU (10BT)

#1
by mrs83 - opened
ethicalabs.ai org

Echo-DSRN-114M-Base-PreTrain-TBPTT-Fineweb-EDU

Trained on a single AMD Instinct MI300X Accelerator

Sign up or log in to comment