Anima-Preview-3-SDNQ-uint4 / quantization_config.json
WaveCut's picture
Add SDNQ uint4 quantized Anima transformer
2cb5ead verified
{
"add_skip_keys": false,
"dequantize_fp32": false,
"dynamic_loss_threshold": null,
"group_size": 0,
"is_integer": true,
"is_training": false,
"modules_dtype_dict": {},
"modules_quant_config": {},
"modules_to_not_convert": [],
"non_blocking": false,
"quant_conv": false,
"quant_embedding": false,
"quant_method": "sdnq",
"quantization_device": null,
"quantized_matmul_dtype": null,
"return_device": null,
"sdnq_version": "0.1.8",
"svd_rank": 32,
"svd_steps": 8,
"use_dynamic_quantization": false,
"use_grad_ckpt": true,
"use_quantized_matmul": true,
"use_quantized_matmul_conv": false,
"use_static_quantization": true,
"use_stochastic_rounding": false,
"use_svd": false,
"weights_dtype": "uint4"
}