{ "bits": 4, "group_size": 128, "sym": true, "data_type": "int", "enable_quanted_input": true, "enable_minmax_tuning": true, "seqlen": 512, "batch_size": 1, "scale_dtype": "torch.float16", "lr": 0.005, "minmax_lr": 0.005, "gradient_accumulate_steps": 4, "iters": 200, "amp": true, "nsamples": 128, "low_gpu_mem_usage": false, "to_quant_block_names": [ [ "vision_model.transformer.layers.0", "vision_model.transformer.layers.1", "vision_model.transformer.layers.2", "vision_model.transformer.layers.3", "vision_model.transformer.layers.4", "vision_model.transformer.layers.5", "vision_model.transformer.layers.6", "vision_model.transformer.layers.7", "vision_model.transformer.layers.8", "vision_model.transformer.layers.9", "vision_model.transformer.layers.10", "vision_model.transformer.layers.11", "vision_model.transformer.layers.12", "vision_model.transformer.layers.13", "vision_model.transformer.layers.14", "vision_model.transformer.layers.15", "vision_model.transformer.layers.16", "vision_model.transformer.layers.17", "vision_model.transformer.layers.18", "vision_model.transformer.layers.19", "vision_model.transformer.layers.20", "vision_model.transformer.layers.21", "vision_model.transformer.layers.22", "vision_model.transformer.layers.23", "vision_model.transformer.layers.24", "vision_model.transformer.layers.25", "vision_model.transformer.layers.26", "vision_model.transformer.layers.27", "vision_model.transformer.layers.28", "vision_model.transformer.layers.29", "vision_model.transformer.layers.30", "vision_model.transformer.layers.31" ], [ "vision_model.global_transformer.layers.0", "vision_model.global_transformer.layers.1", "vision_model.global_transformer.layers.2", "vision_model.global_transformer.layers.3", "vision_model.global_transformer.layers.4", "vision_model.global_transformer.layers.5", "vision_model.global_transformer.layers.6", "vision_model.global_transformer.layers.7" ], [ "language_model.model.layers.0", "language_model.model.layers.1", "language_model.model.layers.2", "language_model.model.layers.3", "language_model.model.layers.4", "language_model.model.layers.5", "language_model.model.layers.6", "language_model.model.layers.7", "language_model.model.layers.8", "language_model.model.layers.9", "language_model.model.layers.10", "language_model.model.layers.11", "language_model.model.layers.12", "language_model.model.layers.13", "language_model.model.layers.14", "language_model.model.layers.15", "language_model.model.layers.16", "language_model.model.layers.17", "language_model.model.layers.18", "language_model.model.layers.19", "language_model.model.layers.20", "language_model.model.layers.21", "language_model.model.layers.22", "language_model.model.layers.23", "language_model.model.layers.24", "language_model.model.layers.25", "language_model.model.layers.26", "language_model.model.layers.27", "language_model.model.layers.28", "language_model.model.layers.29", "language_model.model.layers.30", "language_model.model.layers.31", "language_model.model.layers.32", "language_model.model.layers.33", "language_model.model.layers.34", "language_model.model.layers.35", "language_model.model.layers.36", "language_model.model.layers.37", "language_model.model.layers.38", "language_model.model.layers.39" ] ], "enable_norm_bias_tuning": false, "dataset": "liuhaotian/llava_conv_58k", "autoround_version": "0.4.0.dev", "quant_method": "intel/auto-round", "backend": "auto_round:gptq:exllamav2" }