diff --git a/modules/model_quant.py b/modules/model_quant.py index 75632c020..ac8186cce 100644 --- a/modules/model_quant.py +++ b/modules/model_quant.py @@ -142,7 +142,7 @@ def create_sdnq_config(kwargs = None, allow_sdnq: bool = True, module: str = 'Mo quantization_device=quantization_device, return_device=return_device, ) - log.debug(f'Quantization: module="{module}" type=sdnq dtype={weights_dtype}') + log.debug(f'Quantization: module="{module}" type=sdnq dtype={weights_dtype} matmul={shared.opts.sdnq_use_quantized_matmul} group_size={shared.opts.sdnq_quantize_weights_group_size} quant_conv={shared.opts.sdnq_quantize_conv_layers} matmul_conv={shared.opts.sdnq_use_quantized_matmul_conv} dequantize_fp32={shared.opts.sdnq_dequantize_fp32} quantize_with_gpu={shared.opts.sdnq_quantize_with_gpu} quantization_device={quantization_device} return_device={return_device}') if kwargs is None: return sdnq_config else: @@ -402,7 +402,7 @@ def sdnq_quantize_weights(sd_model): try: t0 = time.time() from modules import shared, devices, sd_models - log.info(f"Quantization: type=SDNQ dtype={shared.opts.sdnq_quantize_weights_mode} dtype_te={shared.opts.sdnq_quantize_weights_mode_te} modules={shared.opts.sdnq_quantize_weights}") + log.info(f"Quantization: type=SDNQ dtype={shared.opts.sdnq_quantize_weights_mode} dtype_te={shared.opts.sdnq_quantize_weights_mode_te} matmul={shared.opts.sdnq_use_quantized_matmul} group_size={shared.opts.sdnq_quantize_weights_group_size} quant_conv={shared.opts.sdnq_quantize_conv_layers} matmul_conv={shared.opts.sdnq_use_quantized_matmul_conv} quantize_with_gpu={shared.opts.sdnq_quantize_with_gpu} dequantize_fp32={shared.opts.sdnq_dequantize_fp32} modules={shared.opts.sdnq_quantize_weights}") global quant_last_model_name, quant_last_model_device # pylint: disable=global-statement sd_model = sd_models.apply_function_to_model(sd_model, sdnq_quantize_model, shared.opts.sdnq_quantize_weights, op="sdnq")