The resulting model has 140 layers and approximately 122 billion parameters.
See mergekit-config.yml for details on the merge method used.
See the exl2-* branches for exllama2 quantizations. The 4.85 bpw quant should fit in 80GB VRAM, and the 3.0 bpw quant should (just barely) fit in 48GB VRAM with 4k context.
Better at following instructions than both v1.1 and v1.0, and doesn't seem to suffer from censorship issues. Overall the original author likes this version the most out of all the models they've created.
