Checkpoint-6000: polynomial (D-1.0) LR 1e-05 to 5e-06, trained on ~2.5K hours total data (~12000 steps total) c64134d verified omarabb315 commited on Apr 30
Checkpoint-6000: polynomial (D-1.0) LR 1e-05 to 5e-06, trained on ~2.5K hours total data (~12000 steps total) 8169e15 verified omarabb315 commited on Apr 29
Checkpoint-6000: polynomial (D-1.0) LR 1e-05 to 5e-06, trained on ~2.5K hours total data (~12000 steps total) aa3e143 verified omarabb315 commited on Apr 26
Checkpoint-6000: constant LR 1e-05, trained on ~2.5K hours total data (~12000 steps total) a43f3fe verified omarabb315 commited on Apr 23