Looking at the gallium trioxide reduction, it seems that you can first reduce it to Ga2O with Hydrogen (water as byproduct), Gallium suboxide then decomposes i

Looking at the gallium trioxide reduction, it seems that you can first reduce it to Ga2O with Hydrogen (water as byproduct), Gallium suboxide then decomposes into it's components at temperatures above 500 C. Feasibility of this is an exercise to the reader (I have no clue).