<h2 id="base-minimaxh3-(turbo-merge)">Base MinimaxH3 (Turbo Merge)</h2>Unpruned is in INT8 only as I do not feel the need to quant that to the lower quality.This model is unchanged/untrained from the base model listed. The model was quantized from the source FP32/BF16 model.In testing I have found this model to be 10% faster even with the high quality 64 Convrot size, then using the turbo lora.

Base MinimaxH3 (8-Step Turbo Merge)

baseMinimaxh38Step_comfyFl2vaPruned_bf16.safetensors

Files with SHA256: ceb4ee63c1e501fc07d954987e68a9c1fcfa87b4d5f99933b92ec40010e87fc8