GitHub - beamivalice/PonyExl3: Run Exl3 format LLM quant on Apple Silicon
v0.2.0 - Update, Converter now landed but it's still in beta shape since I haven't tested the tool against a few supported models. But the results with Qwen and MiniCPM were good enough to justify a land. The quality is similar to what you expect from CUDA qu…
v0.2.0 - Update, Converter now landed but it's still in beta shape since I haven't tested the tool against a few supported models. But the results with Qwen and MiniCPM were good enough to justify a … [+23783 chars]