Index

A

AbsMaxObserver (class in neural_compressor.jax.quantization.layers_static) AcceleratorRegistry (class in neural_compressor.torch.utils.auto_accelerator) add_fp8_support() (in module neural_compressor.jax.utils.utility) Auto_Accelerator (class in neural_compressor.torch.utils.auto_accelerator) auto_detect_accelerator() (in module neural_compressor.torch.utils.auto_accelerator) AutoAlpha (class in neural_compressor.torch.algorithms.smooth_quant.utility) autoround_quantize_entry() (in module neural_compressor.torch.quantization.algorithm_entry)

AutoRoundConfig (class in neural_compressor.torch.quantization.config) AutoRoundQuantizer (class in neural_compressor.torch.algorithms.autoround.autoround) autotune() (in module neural_compressor.torch.quantization.autotune) awq_quantize_entry() (in module neural_compressor.torch.quantization.algorithm_entry) AWQConfig (class in neural_compressor.torch.quantization.config) AWQQuantizer (class in neural_compressor.torch.algorithms.weight_only.awq) axis (neural_compressor.torch.algorithms.weight_only.hqq.qtensor.QTensorMetaInfo attribute)

B

BaseConfig (class in neural_compressor.common.base_config)

block_wise_calibration() (in module neural_compressor.torch.utils.block_wise) build_captured_dataloader() (in module neural_compressor.torch.algorithms.smooth_quant.utility)

C

cal_scale() (in module neural_compressor.torch.algorithms.smooth_quant.utility) Calibration (class in neural_compressor.torch.algorithms.smooth_quant.utility) call_counter() (in module neural_compressor.common.utils.utility) can_pack_with_numba() (in module neural_compressor.torch.utils.environ) causal_lm_make_replace_build_cache() (in module neural_compressor.jax.utils.utility) causal_lm_make_replace_generate_function() (in module neural_compressor.jax.utils.utility) cfg_to_qconfig() (in module neural_compressor.torch.algorithms.smooth_quant.utility) (in module neural_compressor.torch.algorithms.static_quant.utility) change_config_to_hf_format() (in module neural_compressor.torch.algorithms.weight_only.save_load) check_backend() (in module neural_compressor.jax.utils.utility) check_cfg_and_qconfig() (in module neural_compressor.torch.algorithms.smooth_quant.utility) (in module neural_compressor.torch.algorithms.static_quant.utility)

clean_module_weight() (in module neural_compressor.torch.algorithms.layer_wise.utils) ComposableConfig (class in neural_compressor.common.base_config) config_list (neural_compressor.common.base_config.ComposableConfig attribute) (neural_compressor.common.base_tuning.ConfigSet attribute) ConfigLoader (class in neural_compressor.common.base_tuning) ConfigRegistry (class in neural_compressor.common.base_config) ConfigSet (class in neural_compressor.common.base_tuning) convert() (in module neural_compressor.torch.algorithms.qat.quant_utils) (in module neural_compressor.torch.quantization.quantize) convert_model_with_mapping() (in module neural_compressor.torch.algorithms.qat.quant_utils) CPU_Accelerator (class in neural_compressor.torch.utils.auto_accelerator) CpuInfo (class in neural_compressor.common.utils.utility) CUDA_Accelerator (class in neural_compressor.torch.utils.auto_accelerator)

D

detect_device() (in module neural_compressor.torch.utils.utility) detect_processor_type_based_on_hw() (in module neural_compressor.common.utils.utility) device_synchronize() (in module neural_compressor.torch.utils.environ) dowload_hf_model() (in module neural_compressor.torch.utils.utility) dump_elapsed_time() (in module neural_compressor.common.utils.utility) dump_model_op_stats() (in module neural_compressor.torch.algorithms.autoround.autoround) (in module neural_compressor.torch.algorithms.smooth_quant.utility) (in module neural_compressor.torch.algorithms.static_quant.utility) (in module neural_compressor.torch.utils.utility)

dynamic_quantize() (in module neural_compressor.jax.algorithms.dynamic) DynamicQDQLayer (class in neural_compressor.jax.quantization.layers_dynamic) DynamicQuantConfig (class in neural_compressor.jax.quantization.config)

E

ElemFormat (class in neural_compressor.torch.algorithms.mx_quant.utils) enough_memo_store_scale() (in module neural_compressor.torch.algorithms.smooth_quant.utility)

EvaluationFuncWrapper (class in neural_compressor.common.base_tuning) Evaluator (class in neural_compressor.common.base_tuning) export_hf2compressored_model() (in module neural_compressor.torch.export.export_hf)

F

FakeAffineTensorQuantFunction (class in neural_compressor.torch.algorithms.weight_only.modules) fetch_module() (in module neural_compressor.torch.algorithms.weight_only.utility) (in module neural_compressor.torch.utils.utility) filter_fn() (in module neural_compressor.torch.algorithms.weight_only.hqq.quantizer) finalize_calibration() (in module neural_compressor.torch.quantization.quantize) find_all_layers() (in module neural_compressor.torch.algorithms.weight_only.gptq) find_layers() (in module neural_compressor.torch.algorithms.weight_only.gptq)

find_layers_name() (in module neural_compressor.torch.algorithms.weight_only.gptq) find_matching_blocks() (in module neural_compressor.torch.utils.utility) forward_wrapper() (in module neural_compressor.torch.algorithms.smooth_quant.utility) (in module neural_compressor.torch.algorithms.weight_only.utility) (in module neural_compressor.torch.utils.utility) fp8_entry() (in module neural_compressor.torch.quantization.algorithm_entry) FP8Config (class in neural_compressor.torch.quantization.config)

G

generate_activation_observer() (in module neural_compressor.torch.algorithms.static_quant.utility) generate_xpu_qconfig() (in module neural_compressor.torch.algorithms.static_quant.utility) get_absorb_layers() (in module neural_compressor.torch.algorithms.weight_only.utility) get_accelerator() (in module neural_compressor.torch.utils.environ) get_activation_observer() (in module neural_compressor.jax.quantization.layers_static) get_all_config_set() (in module neural_compressor.torch.quantization.autotune) get_all_config_set_from_config_registry() (in module neural_compressor.common.base_config) get_all_registered_configs() (in module neural_compressor.jax.quantization.config) (in module neural_compressor.torch.quantization.config) get_block_names() (in module neural_compressor.torch.utils.utility) get_block_prefix() (in module neural_compressor.torch.algorithms.weight_only.utility) (in module neural_compressor.torch.utils.block_wise) get_children() (in module neural_compressor.torch.algorithms.layer_wise.utils) get_dataloader() (in module neural_compressor.torch.algorithms.autoround.autoround) get_default_AutoRound_config() (in module neural_compressor.torch.quantization.config) get_default_awq_config() (in module neural_compressor.torch.quantization.config) get_default_double_quant_config() (in module neural_compressor.torch.quantization.config) get_default_dynamic_config() (in module neural_compressor.jax.quantization.config) get_default_fp8_config() (in module neural_compressor.torch.quantization.config) get_default_fp8_config_set() (in module neural_compressor.torch.quantization.config) get_default_gptq_config() (in module neural_compressor.torch.quantization.config) get_default_hqq_config() (in module neural_compressor.torch.quantization.config) get_default_mixed_precision_config() (in module neural_compressor.torch.quantization.config) get_default_mixed_precision_config_set() (in module neural_compressor.torch.quantization.config) get_default_mx_config() (in module neural_compressor.torch.quantization.config) get_default_qat_module_mappings() (in module neural_compressor.torch.quantization.config) get_default_rtn_config() (in module neural_compressor.torch.quantization.config) get_default_sq_config() (in module neural_compressor.torch.quantization.config) get_default_static_config() (in module neural_compressor.jax.quantization.config) (in module neural_compressor.torch.quantization.config) get_default_teq_config() (in module neural_compressor.torch.quantization.config) get_depth() (in module neural_compressor.torch.algorithms.static_quant.utility) get_dequantize_fun() (in module neural_compressor.jax.utils.utility) get_dict_at_depth() (in module neural_compressor.torch.algorithms.static_quant.utility) get_double_quant_config_dict() (in module neural_compressor.torch.utils.utility) get_element_under_depth() (in module neural_compressor.torch.algorithms.static_quant.utility) get_enum_from_format() (in module neural_compressor.torch.utils.utility) get_ipex_version() (in module neural_compressor.torch.utils.environ)

get_layer_names_in_block() (in module neural_compressor.torch.utils.utility) get_mllm_dataloader() (in module neural_compressor.torch.algorithms.autoround.autoround) get_model_device() (in module neural_compressor.torch.utils.utility) get_model_info() (in module neural_compressor.torch.utils.utility) get_module() (in module neural_compressor.torch.algorithms.layer_wise.utils) (in module neural_compressor.torch.algorithms.smooth_quant.utility) (in module neural_compressor.torch.algorithms.weight_only.utility) (in module neural_compressor.torch.utils.utility) get_module_input_output() (in module neural_compressor.torch.algorithms.weight_only.utility) get_multimodal_block_names() (in module neural_compressor.torch.utils.utility) get_named_children() (in module neural_compressor.torch.algorithms.layer_wise.utils) get_non_persistent_buffers() (in module neural_compressor.torch.utils.utility) get_parent() (in module neural_compressor.torch.algorithms.smooth_quant.utility) (in module neural_compressor.torch.algorithms.weight_only.utility) get_processor_type_from_user_config() (in module neural_compressor.torch.utils.utility) get_q_params() (in module neural_compressor.jax.utils.utility) get_quant_config() (in module neural_compressor.torch.algorithms.qat.quant_utils) get_quant_config_with_scheme() (in module neural_compressor.torch.algorithms.qat.quant_utils) get_quantizable_ops_from_cfgs() (in module neural_compressor.torch.algorithms.static_quant.utility) get_quantizable_ops_recursively() (in module neural_compressor.torch.algorithms.smooth_quant.utility) (in module neural_compressor.torch.algorithms.static_quant.utility) get_quantization_format() (in module neural_compressor.torch.algorithms.qat.quant_utils) get_quantize_fun() (in module neural_compressor.jax.utils.utility) get_quantizer() (in module neural_compressor.torch.utils.utility) get_rtn_double_quant_config_set() (in module neural_compressor.torch.quantization.autotune) get_scale() (in module neural_compressor.jax.utils.utility) get_super_module_by_name() (in module neural_compressor.torch.algorithms.layer_wise.utils) get_torch_version() (in module neural_compressor.torch.utils.environ) get_used_cpu_mem_MB() (in module neural_compressor.torch.utils.environ) get_used_hpu_mem_MB() (in module neural_compressor.torch.utils.environ) get_woq_tuning_config() (in module neural_compressor.torch.quantization.config) get_workspace() (in module neural_compressor.common.utils.utility) GPTQ (class in neural_compressor.torch.algorithms.weight_only.gptq) gptq_entry() (in module neural_compressor.torch.quantization.algorithm_entry) GPTQConfig (class in neural_compressor.torch.quantization.config) GPTQuantizer (class in neural_compressor.torch.algorithms.weight_only.gptq) GraphTrace (class in neural_compressor.torch.algorithms.smooth_quant.utility) (class in neural_compressor.torch.algorithms.weight_only.utility) group_size (neural_compressor.torch.algorithms.weight_only.hqq.qtensor.QTensorMetaInfo attribute)

H

HalfPrecisionConverter (class in neural_compressor.torch.algorithms.mixed_precision.half_precision_convert) HalfPrecisionModuleWrapper (class in neural_compressor.torch.algorithms.mixed_precision.module_wrappers) HPU_Accelerator (class in neural_compressor.torch.utils.auto_accelerator) HPUWeightOnlyLinear (class in neural_compressor.torch.algorithms.weight_only.modules) hqq_entry() (in module neural_compressor.torch.quantization.algorithm_entry) HQQConfig (class in neural_compressor.torch.quantization.config)

HQQLinear (class in neural_compressor.torch.algorithms.weight_only.hqq.core) HQQModuleConfig (class in neural_compressor.torch.algorithms.weight_only.hqq.config) HQQTensorHandle (class in neural_compressor.torch.algorithms.weight_only.hqq.core) HQQuantizer (class in neural_compressor.torch.algorithms.weight_only.hqq.quantizer) hybrid_gptq_entry() (in module neural_compressor.torch.quantization.algorithm_entry) HybridGPTQConfig (class in neural_compressor.torch.quantization.config)

I

INCAcceleratorType (class in neural_compressor.torch.utils.auto_accelerator) INCWeightOnlyLinear (class in neural_compressor.torch.algorithms.weight_only.modules) init_tuning() (in module neural_compressor.common.base_tuning) INT8StaticQuantConfig (class in neural_compressor.torch.quantization.config) is_fp8_kv_cache_enabled_by_env() (in module neural_compressor.jax.utils.utility) is_hpex_available() (in module neural_compressor.torch.utils.environ) is_hpex_support_g_idx() (in module neural_compressor.torch.utils.environ) is_hpu_available() (in module neural_compressor.torch.utils.environ) is_ipex_available() (in module neural_compressor.torch.utils.environ)

is_ipex_imported() (in module neural_compressor.torch.utils.environ) is_leaf() (in module neural_compressor.torch.algorithms.weight_only.gptq) is_numba_available() (in module neural_compressor.torch.utils.environ) is_optimum_available() (in module neural_compressor.torch.utils.environ) is_optimum_habana_available() (in module neural_compressor.torch.utils.environ) is_package_available() (in module neural_compressor.torch.utils.environ) is_quantlinear() (in module neural_compressor.torch.algorithms.qat.quant_utils) is_tbb_available() (in module neural_compressor.torch.utils.environ) is_transformers_imported() (in module neural_compressor.torch.utils.environ) iterate_over_layers() (in module neural_compressor.jax.utils.utility)

J

JaxBaseConfig (class in neural_compressor.jax.quantization.config)

JaxComposableConfig (class in neural_compressor.jax.quantization.config)

K

KerasQuantizedGemmaWrapper (class in neural_compressor.jax.quantization.saving) KerasQuantizedModelBackboneWrapper (class in neural_compressor.jax.quantization.saving) KerasQuantizedModelWrapper (class in neural_compressor.jax.quantization.saving)

KerasQuantizedModelWrapperMixin (class in neural_compressor.jax.quantization.saving) KerasQuantizedTokenizerWrapper (class in neural_compressor.jax.quantization.saving) KerasQuantizedViTWrapper (class in neural_compressor.jax.quantization.saving)

L

LazyImport (class in neural_compressor.common.utils.utility) load() (in module neural_compressor.torch.algorithms.layer_wise.load) (in module neural_compressor.torch.algorithms.static_quant.save_load) (in module neural_compressor.torch.algorithms.weight_only.save_load) (in module neural_compressor.torch.quantization.save_load_entry) load_config_mapping() (in module neural_compressor.common.utils.save_load) load_empty_model() (in module neural_compressor.torch.utils.utility) load_first_layer_only() (in module neural_compressor.torch.algorithms.layer_wise.utils) load_layer_wise_quantized_model() (in module neural_compressor.torch.algorithms.layer_wise.utils) load_model_from_shards_with_safetensors() (in module neural_compressor.torch.algorithms.layer_wise.utils)

load_module() (in module neural_compressor.torch.algorithms.layer_wise.utils) load_non_persistent_buffers() (in module neural_compressor.torch.utils.utility) load_tensor() (in module neural_compressor.torch.algorithms.layer_wise.utils) load_tensor_from_safetensors() (in module neural_compressor.torch.algorithms.layer_wise.utils) load_tensor_from_safetensors_shard() (in module neural_compressor.torch.algorithms.layer_wise.utils) load_tensor_from_shard() (in module neural_compressor.torch.algorithms.layer_wise.utils) load_value() (in module neural_compressor.torch.algorithms.layer_wise.utils) log_process() (in module neural_compressor.common.utils.utility) log_quantizable_layers_per_transformer() (in module neural_compressor.torch.algorithms.weight_only.gptq) Logger (class in neural_compressor.common.utils.logger)

M

Matmul (class in neural_compressor.torch.algorithms.weight_only.modules) MinMaxObserver (class in neural_compressor.jax.quantization.layers_static) mixed_precision_entry() (in module neural_compressor.torch.quantization.algorithm_entry) MixedPrecisionConfig (class in neural_compressor.torch.quantization.config) Mode (class in neural_compressor.common.utils.constants) model_forward() (in module neural_compressor.torch.algorithms.smooth_quant.utility) (in module neural_compressor.torch.algorithms.weight_only.utility) model_forward_per_sample() (in module neural_compressor.torch.algorithms.smooth_quant.utility) MODEL_LEVEL (neural_compressor.common.tuning_param.ParamLevel attribute) module neural_compressor neural_compressor.common neural_compressor.common.base_config neural_compressor.common.base_tuning neural_compressor.common.tuning_param neural_compressor.common.utils neural_compressor.common.utils.constants neural_compressor.common.utils.logger neural_compressor.common.utils.save_load neural_compressor.common.utils.utility neural_compressor.common.version neural_compressor.jax neural_compressor.jax.algorithms neural_compressor.jax.algorithms.dynamic neural_compressor.jax.algorithms.static neural_compressor.jax.quantization neural_compressor.jax.quantization.config neural_compressor.jax.quantization.layers_dynamic neural_compressor.jax.quantization.layers_static neural_compressor.jax.quantization.quantize neural_compressor.jax.quantization.saving neural_compressor.jax.utils neural_compressor.jax.utils.utility neural_compressor.torch neural_compressor.torch.algorithms neural_compressor.torch.algorithms.autoround neural_compressor.torch.algorithms.autoround.autoround neural_compressor.torch.algorithms.base_algorithm neural_compressor.torch.algorithms.fp8_quant neural_compressor.torch.algorithms.fp8_quant.observer neural_compressor.torch.algorithms.fp8_quant.patched_module_base neural_compressor.torch.algorithms.fp8_quant.scaling_method_base neural_compressor.torch.algorithms.layer_wise neural_compressor.torch.algorithms.layer_wise.load neural_compressor.torch.algorithms.layer_wise.modified_pickle neural_compressor.torch.algorithms.layer_wise.utils neural_compressor.torch.algorithms.mixed_precision neural_compressor.torch.algorithms.mixed_precision.half_precision_convert neural_compressor.torch.algorithms.mixed_precision.module_wrappers neural_compressor.torch.algorithms.mx_quant neural_compressor.torch.algorithms.mx_quant.mx neural_compressor.torch.algorithms.mx_quant.utils neural_compressor.torch.algorithms.qat neural_compressor.torch.algorithms.qat.quant_linear neural_compressor.torch.algorithms.qat.quant_utils neural_compressor.torch.algorithms.qat.tensor_quantizer neural_compressor.torch.algorithms.smooth_quant neural_compressor.torch.algorithms.smooth_quant.save_load neural_compressor.torch.algorithms.smooth_quant.smooth_quant neural_compressor.torch.algorithms.smooth_quant.utility neural_compressor.torch.algorithms.static_quant neural_compressor.torch.algorithms.static_quant.save_load neural_compressor.torch.algorithms.static_quant.static_quant neural_compressor.torch.algorithms.static_quant.utility neural_compressor.torch.algorithms.weight_only neural_compressor.torch.algorithms.weight_only.awq neural_compressor.torch.algorithms.weight_only.gptq neural_compressor.torch.algorithms.weight_only.hqq neural_compressor.torch.algorithms.weight_only.hqq.bitpack neural_compressor.torch.algorithms.weight_only.hqq.config neural_compressor.torch.algorithms.weight_only.hqq.core neural_compressor.torch.algorithms.weight_only.hqq.optimizer neural_compressor.torch.algorithms.weight_only.hqq.qtensor neural_compressor.torch.algorithms.weight_only.hqq.quantizer neural_compressor.torch.algorithms.weight_only.modules neural_compressor.torch.algorithms.weight_only.rtn neural_compressor.torch.algorithms.weight_only.save_load neural_compressor.torch.algorithms.weight_only.teq neural_compressor.torch.algorithms.weight_only.utility neural_compressor.torch.export neural_compressor.torch.export.export_hf neural_compressor.torch.quantization neural_compressor.torch.quantization.algorithm_entry neural_compressor.torch.quantization.autotune neural_compressor.torch.quantization.config neural_compressor.torch.quantization.quantize neural_compressor.torch.quantization.save_load_entry neural_compressor.torch.utils neural_compressor.torch.utils.auto_accelerator neural_compressor.torch.utils.bit_packer neural_compressor.torch.utils.block_wise neural_compressor.torch.utils.constants neural_compressor.torch.utils.environ neural_compressor.torch.utils.utility neural_compressor.version

move_input_device() (in module neural_compressor.torch.utils.utility) move_input_to_device() (in module neural_compressor.torch.algorithms.smooth_quant.utility) (in module neural_compressor.torch.algorithms.weight_only.utility) MulLinear (class in neural_compressor.torch.algorithms.weight_only.modules) mx_quant_entry() (in module neural_compressor.torch.quantization.algorithm_entry) MXLinear (class in neural_compressor.torch.algorithms.mx_quant.mx) MXQuantConfig (class in neural_compressor.torch.quantization.config) MXQuantizer (class in neural_compressor.torch.algorithms.mx_quant.mx)

N

name (neural_compressor.common.base_config.BaseConfig attribute) nbits (neural_compressor.torch.algorithms.weight_only.hqq.qtensor.QTensorMetaInfo attribute) need_apply() (in module neural_compressor.jax.quantization.quantize) (in module neural_compressor.torch.quantization.quantize) neural_compressor module neural_compressor.common module neural_compressor.common.base_config module neural_compressor.common.base_tuning module neural_compressor.common.tuning_param module neural_compressor.common.utils module neural_compressor.common.utils.constants module neural_compressor.common.utils.logger module neural_compressor.common.utils.save_load module neural_compressor.common.utils.utility module neural_compressor.common.version module neural_compressor.jax module neural_compressor.jax.algorithms module neural_compressor.jax.algorithms.dynamic module neural_compressor.jax.algorithms.static module neural_compressor.jax.quantization module neural_compressor.jax.quantization.config module neural_compressor.jax.quantization.layers_dynamic module neural_compressor.jax.quantization.layers_static module neural_compressor.jax.quantization.quantize module neural_compressor.jax.quantization.saving module neural_compressor.jax.utils module neural_compressor.jax.utils.utility module neural_compressor.torch module neural_compressor.torch.algorithms module neural_compressor.torch.algorithms.autoround module neural_compressor.torch.algorithms.autoround.autoround module neural_compressor.torch.algorithms.base_algorithm module neural_compressor.torch.algorithms.fp8_quant module neural_compressor.torch.algorithms.fp8_quant.observer module neural_compressor.torch.algorithms.fp8_quant.patched_module_base module neural_compressor.torch.algorithms.fp8_quant.scaling_method_base module neural_compressor.torch.algorithms.layer_wise module neural_compressor.torch.algorithms.layer_wise.load module neural_compressor.torch.algorithms.layer_wise.modified_pickle module neural_compressor.torch.algorithms.layer_wise.utils module neural_compressor.torch.algorithms.mixed_precision module neural_compressor.torch.algorithms.mixed_precision.half_precision_convert module neural_compressor.torch.algorithms.mixed_precision.module_wrappers module neural_compressor.torch.algorithms.mx_quant module neural_compressor.torch.algorithms.mx_quant.mx module neural_compressor.torch.algorithms.mx_quant.utils module

neural_compressor.torch.algorithms.qat module neural_compressor.torch.algorithms.qat.quant_linear module neural_compressor.torch.algorithms.qat.quant_utils module neural_compressor.torch.algorithms.qat.tensor_quantizer module neural_compressor.torch.algorithms.smooth_quant module neural_compressor.torch.algorithms.smooth_quant.save_load module neural_compressor.torch.algorithms.smooth_quant.smooth_quant module neural_compressor.torch.algorithms.smooth_quant.utility module neural_compressor.torch.algorithms.static_quant module neural_compressor.torch.algorithms.static_quant.save_load module neural_compressor.torch.algorithms.static_quant.static_quant module neural_compressor.torch.algorithms.static_quant.utility module neural_compressor.torch.algorithms.weight_only module neural_compressor.torch.algorithms.weight_only.awq module neural_compressor.torch.algorithms.weight_only.gptq module neural_compressor.torch.algorithms.weight_only.hqq module neural_compressor.torch.algorithms.weight_only.hqq.bitpack module neural_compressor.torch.algorithms.weight_only.hqq.config module neural_compressor.torch.algorithms.weight_only.hqq.core module neural_compressor.torch.algorithms.weight_only.hqq.optimizer module neural_compressor.torch.algorithms.weight_only.hqq.qtensor module neural_compressor.torch.algorithms.weight_only.hqq.quantizer module neural_compressor.torch.algorithms.weight_only.modules module neural_compressor.torch.algorithms.weight_only.rtn module neural_compressor.torch.algorithms.weight_only.save_load module neural_compressor.torch.algorithms.weight_only.teq module neural_compressor.torch.algorithms.weight_only.utility module neural_compressor.torch.export module neural_compressor.torch.export.export_hf module neural_compressor.torch.quantization module neural_compressor.torch.quantization.algorithm_entry module neural_compressor.torch.quantization.autotune module neural_compressor.torch.quantization.config module neural_compressor.torch.quantization.quantize module neural_compressor.torch.quantization.save_load_entry module neural_compressor.torch.utils module neural_compressor.torch.utils.auto_accelerator module neural_compressor.torch.utils.bit_packer module neural_compressor.torch.utils.block_wise module neural_compressor.torch.utils.constants module neural_compressor.torch.utils.environ module neural_compressor.torch.utils.utility module neural_compressor.version module

O

ObserverBase (class in neural_compressor.torch.algorithms.fp8_quant.observer) OP_LEVEL (neural_compressor.common.tuning_param.ParamLevel attribute) OP_TYPE_LEVEL (neural_compressor.common.tuning_param.ParamLevel attribute)

OperatorConfig (class in neural_compressor.jax.quantization.config) (class in neural_compressor.torch.quantization.config) optimize_weights_proximal_legacy() (in module neural_compressor.torch.algorithms.weight_only.hqq.optimizer)

P

pack_array_with_numba_b2_c16() (in module neural_compressor.torch.utils.bit_packer) pack_array_with_numba_b2_c32() (in module neural_compressor.torch.utils.bit_packer) pack_array_with_numba_b2_c64() (in module neural_compressor.torch.utils.bit_packer) pack_array_with_numba_b2_c8() (in module neural_compressor.torch.utils.bit_packer) pack_array_with_numba_b4_c16() (in module neural_compressor.torch.utils.bit_packer) pack_array_with_numba_b4_c32() (in module neural_compressor.torch.utils.bit_packer) pack_array_with_numba_b4_c64() (in module neural_compressor.torch.utils.bit_packer) pack_array_with_numba_b4_c8() (in module neural_compressor.torch.utils.bit_packer) pack_array_with_numba_b8_c16() (in module neural_compressor.torch.utils.bit_packer) pack_array_with_numba_b8_c32() (in module neural_compressor.torch.utils.bit_packer) pack_array_with_numba_b8_c64() (in module neural_compressor.torch.utils.bit_packer) pack_array_with_numba_b8_c8() (in module neural_compressor.torch.utils.bit_packer) Packer (class in neural_compressor.torch.algorithms.weight_only.hqq.bitpack) packing (neural_compressor.torch.algorithms.weight_only.hqq.qtensor.QTensorMetaInfo attribute)

ParamLevel (class in neural_compressor.common.tuning_param) params_list (neural_compressor.common.base_config.BaseConfig attribute) parse_cfgs() (in module neural_compressor.torch.algorithms.static_quant.utility) patch_hqq_moduile() (in module neural_compressor.torch.algorithms.weight_only.hqq.quantizer) PatchedModuleBase (class in neural_compressor.torch.algorithms.fp8_quant.patched_module_base) PickleError PicklingError postprocess_model() (in module neural_compressor.torch.utils.utility) prepare() (in module neural_compressor.torch.quantization.quantize) prepare_deserialized_quantized_model() (in module neural_compressor.jax.quantization.saving) prepare_qat() (in module neural_compressor.torch.quantization.quantize) preprocess_quant_config() (in module neural_compressor.torch.quantization.quantize) print_model() (in module neural_compressor.jax.utils.utility) ProcessorType (class in neural_compressor.common.utils.utility)

Q

qdq_quantize() (in module neural_compressor.torch.algorithms.smooth_quant.smooth_quant) qdq_weight_actor() (in module neural_compressor.torch.algorithms.weight_only.utility) qdq_weight_asym() (in module neural_compressor.torch.algorithms.weight_only.utility) qdq_weight_sym() (in module neural_compressor.torch.algorithms.weight_only.utility) QDQLayer (class in neural_compressor.torch.algorithms.layer_wise.utils) (class in neural_compressor.torch.algorithms.weight_only.modules) QDynamicCachedGemma3Attention (class in neural_compressor.jax.quantization.layers_dynamic) QDynamicConv2D (class in neural_compressor.jax.quantization.layers_dynamic) QDynamicConv2DMixin (class in neural_compressor.jax.quantization.layers_dynamic) QDynamicDense (class in neural_compressor.jax.quantization.layers_dynamic) QDynamicDenseMixin (class in neural_compressor.jax.quantization.layers_dynamic) QDynamicEinsumDense (class in neural_compressor.jax.quantization.layers_dynamic) QDynamicGemma3VisionAttention (class in neural_compressor.jax.quantization.layers_dynamic) QDynamicMultiHeadAttention (class in neural_compressor.jax.quantization.layers_dynamic) QDynamicReversibleEmbedding (class in neural_compressor.jax.quantization.layers_dynamic) QStaticCachedGemma3Attention (class in neural_compressor.jax.quantization.layers_static) QStaticConv2d (class in neural_compressor.jax.quantization.layers_static) QStaticConv2DMixin (class in neural_compressor.jax.quantization.layers_static) QStaticDense (class in neural_compressor.jax.quantization.layers_static)

QStaticDenseMixin (class in neural_compressor.jax.quantization.layers_static) QStaticEinsumDense (class in neural_compressor.jax.quantization.layers_static) QStaticGemma3VisionAttention (class in neural_compressor.jax.quantization.layers_static) QStaticMultiHeadAttention (class in neural_compressor.jax.quantization.layers_static) QStaticReversibleEmbedding (class in neural_compressor.jax.quantization.layers_static) QTensor (class in neural_compressor.torch.algorithms.weight_only.hqq.qtensor) QTensorConfig (class in neural_compressor.torch.algorithms.weight_only.hqq.config) QTensorMetaInfo (class in neural_compressor.torch.algorithms.weight_only.hqq.qtensor) quant_dequant_w_v1() (in module neural_compressor.torch.algorithms.smooth_quant.utility) quant_dequant_x_v1() (in module neural_compressor.torch.algorithms.smooth_quant.utility) quant_tensor() (in module neural_compressor.torch.algorithms.weight_only.utility) quant_weight_w_scale() (in module neural_compressor.torch.algorithms.weight_only.utility) quantize() (in module neural_compressor.torch.quantization.quantize) quantize_4bit() (in module neural_compressor.torch.algorithms.weight_only.utility) quantize_elemwise_op() (in module neural_compressor.torch.algorithms.mx_quant.utils) quantize_model() (in module neural_compressor.jax.quantization.quantize) quantize_mx_op() (in module neural_compressor.torch.algorithms.mx_quant.utils) Quantizer (class in neural_compressor.torch.algorithms.base_algorithm) (class in neural_compressor.torch.algorithms.weight_only.gptq) QuantLinear (class in neural_compressor.torch.algorithms.qat.quant_linear)

R

RAWGPTQuantizer (class in neural_compressor.torch.algorithms.weight_only.gptq) read_json_file() (in module neural_compressor.torch.utils.utility) recover_forward() (in module neural_compressor.torch.algorithms.weight_only.utility) (in module neural_compressor.torch.utils.block_wise) recover_model_from_json() (in module neural_compressor.torch.algorithms.smooth_quant.save_load) register_accelerator() (in module neural_compressor.torch.utils.auto_accelerator) register_algo() (in module neural_compressor.jax.utils.utility) (in module neural_compressor.torch.utils.utility) register_autotune() (in module neural_compressor.torch.algorithms.smooth_quant.utility) register_config() (in module neural_compressor.common.base_config) register_dynamic_quantized_layer() (in module neural_compressor.jax.quantization.layers_dynamic) register_observer() (in module neural_compressor.torch.algorithms.fp8_quant.observer) register_pack_func() (in module neural_compressor.torch.utils.bit_packer) register_patched_module() (in module neural_compressor.torch.algorithms.fp8_quant.patched_module_base)

register_scaling_methods() (in module neural_compressor.torch.algorithms.fp8_quant.scaling_method_base) register_static_quantized_layer() (in module neural_compressor.jax.quantization.layers_static) register_supported_configs_for_fwk() (in module neural_compressor.common.base_config) register_weight_hooks() (in module neural_compressor.torch.algorithms.layer_wise.utils) replace_forward() (in module neural_compressor.torch.algorithms.weight_only.utility) (in module neural_compressor.torch.utils.block_wise) replace_with_quant_linear() (in module neural_compressor.torch.algorithms.qat.quant_utils) replacement_fn() (in module neural_compressor.torch.algorithms.weight_only.hqq.quantizer) reshape_in_channel_to_last() (in module neural_compressor.torch.algorithms.smooth_quant.utility) reshape_scale_as_input() (in module neural_compressor.torch.algorithms.smooth_quant.utility) reshape_scale_as_weight() (in module neural_compressor.torch.algorithms.smooth_quant.utility) RoundingMode (class in neural_compressor.torch.algorithms.mx_quant.utils) rtn_entry() (in module neural_compressor.torch.quantization.algorithm_entry) RTNConfig (class in neural_compressor.torch.quantization.config) RTNQuantizer (class in neural_compressor.torch.algorithms.weight_only.rtn)

S

Sampler (class in neural_compressor.common.base_tuning) save() (in module neural_compressor.torch.algorithms.static_quant.save_load) (in module neural_compressor.torch.algorithms.weight_only.save_load) (in module neural_compressor.torch.quantization.save_load_entry) save_config_mapping() (in module neural_compressor.common.utils.save_load) save_layers_in_shards_iteratively() (in module neural_compressor.torch.algorithms.layer_wise.utils) SaveableLayerMixin (class in neural_compressor.jax.quantization.saving) SaveLoadFormat (class in neural_compressor.torch.utils.constants) ScalingMethodBase (class in neural_compressor.torch.algorithms.fp8_quant.scaling_method_base) search_clip() (in module neural_compressor.torch.algorithms.weight_only.utility) select_kv_cache_dtype() (in module neural_compressor.jax.utils.utility) SequentialSampler (class in neural_compressor.common.base_tuning) set_module() (in module neural_compressor.torch.algorithms.smooth_quant.utility) (in module neural_compressor.torch.algorithms.weight_only.utility) (in module neural_compressor.torch.utils.utility) set_random_seed() (in module neural_compressor.common.utils.utility)

set_resume_from() (in module neural_compressor.common.utils.utility) set_tensorboard() (in module neural_compressor.common.utils.utility) set_workspace() (in module neural_compressor.common.utils.utility) shape (neural_compressor.torch.algorithms.weight_only.hqq.qtensor.QTensorMetaInfo attribute) simple_inference() (in module neural_compressor.torch.algorithms.static_quant.utility) singleton() (in module neural_compressor.common.utils.utility) smooth_quant_entry() (in module neural_compressor.torch.quantization.algorithm_entry) SmoothQuantConfig (class in neural_compressor.torch.quantization.config) SmoothQuantQuantizer (class in neural_compressor.torch.algorithms.smooth_quant.smooth_quant) SQLinearWrapper (class in neural_compressor.torch.algorithms.smooth_quant.utility) static_quant_entry() (in module neural_compressor.torch.quantization.algorithm_entry) static_quantize() (in module neural_compressor.jax.algorithms.static) StaticQDQLayer (class in neural_compressor.jax.quantization.layers_static) StaticQuantConfig (class in neural_compressor.jax.quantization.config) (class in neural_compressor.torch.quantization.config) StaticQuantQuantizer (class in neural_compressor.torch.algorithms.static_quant.static_quant) Statistics (class in neural_compressor.common.utils.utility)

T

TensorQuantizer (class in neural_compressor.torch.algorithms.qat.tensor_quantizer) teq_quantize_entry() (in module neural_compressor.torch.quantization.algorithm_entry) TEQConfig (class in neural_compressor.torch.quantization.config) TEQLinearFakeQuant (class in neural_compressor.torch.algorithms.weight_only.modules) TEQuantizer (class in neural_compressor.torch.algorithms.weight_only.teq) to_device() (in module neural_compressor.torch.utils.utility) to_dtype() (in module neural_compressor.torch.utils.utility) TorchBaseConfig (class in neural_compressor.torch.quantization.config)

TorchSmoothQuant (class in neural_compressor.torch.algorithms.smooth_quant.utility) trace_gptq_target_blocks() (in module neural_compressor.torch.algorithms.weight_only.gptq) TrainableEquivalentTransformation (class in neural_compressor.torch.algorithms.weight_only.teq) TransformerBasedModelBlockPatternDetector (class in neural_compressor.torch.algorithms.static_quant.utility) TuningConfig (class in neural_compressor.common.base_tuning) TuningLogger (class in neural_compressor.common.utils.logger) TuningMonitor (class in neural_compressor.common.base_tuning) TuningParam (class in neural_compressor.common.tuning_param)

U

UnpackedWeightOnlyLinearParams (class in neural_compressor.torch.algorithms.weight_only.modules) UnpicklingError

update_module() (in module neural_compressor.torch.algorithms.layer_wise.utils) update_sq_scale() (in module neural_compressor.torch.algorithms.smooth_quant.utility)

V

validate_modules() (in module neural_compressor.torch.utils.utility) verify_api() (in module neural_compressor.jax.utils.utility)

verify_replacement_api() (in module neural_compressor.jax.utils.utility) VersionManager (class in neural_compressor.jax.quantization.saving)

W

WeightOnlyLinear (class in neural_compressor.torch.algorithms.weight_only.modules) WOQModelLoader (class in neural_compressor.torch.algorithms.weight_only.save_load)

WrapperLayer (class in neural_compressor.torch.algorithms.smooth_quant.utility) write_json_file() (in module neural_compressor.torch.utils.utility)

X

XPU_Accelerator (class in neural_compressor.torch.utils.auto_accelerator)