ltx2-dev-8bit-mlx / quantization.json
AITRADER's picture
Upload quantization.json with huggingface_hub
75073ea verified
{
"group_size": 64,
"bits": 8,
"mode": "affine",
"dtype": "bfloat16",
"pipeline": "dev",
"predicate": "core",
"quantize_scope": "core"
}