Index

_

__quant_init__() (aimet_torch.v2.nn.QuantizationMixin method), [1]

A

ActivationType (class in aimet_common.defs) AdapterMetaData (class in aimet_torch.peft) AdaroundParameters (class in aimet_onnx.adaround.adaround_weight) (class in aimet_tensorflow.keras.adaround_weight) (class in aimet_torch.v1.adaround.adaround_weight) aimet_torch.v2.quantization.affine module aimet_torch.v2.quantization.float module allow_overwrite() (aimet_torch.v2.quantization.affine.quantizer.QuantizerBase method) analyze() (aimet_onnx.quant_analyzer.QuantAnalyzer method) (aimet_tensorflow.keras.quant_analyzer.QuantAnalyzer method) (aimet_torch.v1.quant_analyzer.QuantAnalyzer method)

apply_adaround() (in module aimet_onnx.adaround.adaround_weight.Adaround) (in module aimet_tensorflow.keras.adaround_weight.Adaround) (in module aimet_torch.v1.adaround.adaround_weight.Adaround) apply_gptvq() (in module aimet_torch.gptvq.gptvq_weight.GPTVQ) auto (aimet_tensorflow.keras.defs.SpatialSvdParameters.Mode attribute) (aimet_torch.defs.ChannelPruningParameters.Mode attribute) (aimet_torch.defs.SpatialSvdParameters.Mode attribute) (aimet_torch.defs.WeightSvdParameters.Mode attribute) AutoQuant (class in aimet_onnx.auto_quant_v2) (class in aimet_torch.v1.auto_quant)

B

bias_fold() (in module aimet_tensorflow.keras.cross_layer_equalization.HighBiasFold), [1] (in module aimet_torch.cross_layer_equalization.HighBiasFold), [1]

C

CallbackFunc (class in aimet_common.defs), [1], [2] (class in aimet_common.utils) channel_pruning (aimet_common.defs.CompressionScheme attribute) ChannelPruningParameters (class in aimet_torch.defs) ChannelPruningParameters.AutoModeParams (class in aimet_torch.defs) ChannelPruningParameters.ManualModeParams (class in aimet_torch.defs) ChannelPruningParameters.Mode (class in aimet_torch.defs) check_model_arch() (in module aimet_torch.arch_checker.arch_checker.ArchChecker) check_model_sensitivity_to_quantization() (aimet_onnx.quant_analyzer.QuantAnalyzer method) (aimet_torch.v1.quant_analyzer.QuantAnalyzer method) choose_fast_mixed_precision() (in module aimet_tensorflow.keras.mixed_precision) choose_mixed_precision() (in module aimet_onnx.mixed_precision) (in module aimet_tensorflow.keras.mixed_precision) (in module aimet_torch.mixed_precision) ClsSetInfo (class in aimet_tensorflow.keras.cross_layer_equalization) (class in aimet_torch.cross_layer_equalization)

ClsSetInfo.ClsSetLayerPairInfo (class in aimet_tensorflow.keras.cross_layer_equalization) (class in aimet_torch.cross_layer_equalization) compress_model() (aimet_tensorflow.keras.compress.ModelCompressor static method) (aimet_torch.compress.ModelCompressor static method) CompressionScheme (class in aimet_common.defs) compute_encodings() (aimet_onnx.quantsim.QuantizationSimModel method) (aimet_tensorflow.keras.quantsim.QuantizationSimModel method) (aimet_torch.v1.quantsim.QuantizationSimModel method) (aimet_torch.v2.nn.QuantizationMixin method), [1] (aimet_torch.v2.quantization.affine.quantizer.QuantizerBase method) (aimet_torch.v2.quantization.encoding_analyzer.EncodingAnalyzer method) ConvBnInfoType (class in aimet_common.bias_correction) correct_bias() (in module aimet_torch.bias_correction) CostMetric (class in aimet_common.defs) create_quantsim_and_encodings() (aimet_onnx.quant_analyzer.QuantAnalyzer method)

D

dequantize() (aimet_torch.v2.quantization.tensor.DequantizedTensor method) (aimet_torch.v2.quantization.tensor.QuantizedTensor method) (in module aimet_torch.v2.quantization.affine)

DequantizedTensor (class in aimet_torch.v2.quantization.tensor) disable_lora_adapters() (aimet_torch.peft.PeftQuantUtils method) display_comp_ratio_plot() (aimet_torch.visualize_serialized_data.VisualizeCompression method) display_eval_scores() (aimet_torch.visualize_serialized_data.VisualizeCompression method)

E

enable_adapter_and_load_weights() (aimet_torch.peft.PeftQuantUtils method) enable_per_layer_mse_loss() (aimet_onnx.quant_analyzer.QuantAnalyzer method) (aimet_torch.v1.quant_analyzer.QuantAnalyzer method) EncodingAnalyzer (class in aimet_torch.v2.quantization.encoding_analyzer) equalize_model() (in module aimet_onnx.cross_layer_equalization) (in module aimet_tensorflow.keras.cross_layer_equalization) (in module aimet_torch.cross_layer_equalization) EvalCallbackFactory (class in aimet_onnx.amp.mixed_precision_algo) (class in aimet_torch.amp.mixed_precision_algo)

export() (aimet_onnx.quantsim.QuantizationSimModel method) (aimet_tensorflow.keras.quantsim.QuantizationSimModel method) (aimet_torch.v1.quantsim.QuantizationSimModel method) export_adapter_weights() (aimet_torch.peft.PeftQuantUtils method) export_per_layer_encoding_min_max_range() (aimet_onnx.quant_analyzer.QuantAnalyzer method) (aimet_torch.v1.quant_analyzer.QuantAnalyzer method) export_per_layer_mse_loss() (aimet_onnx.quant_analyzer.QuantAnalyzer method) (aimet_torch.v1.quant_analyzer.QuantAnalyzer method) export_per_layer_stats_histogram() (aimet_onnx.quant_analyzer.QuantAnalyzer method) (aimet_torch.v1.quant_analyzer.QuantAnalyzer method)

F

FloatQuantizeDequantize (class in aimet_torch.v2.quantization.float), [1] fold_all_batch_norms() (in module aimet_tensorflow.keras.batch_norm_fold) (in module aimet_torch.batch_norm_fold) fold_all_batch_norms_to_scale() (in module aimet_tensorflow.keras.batch_norm_fold) (in module aimet_torch.batch_norm_fold) fold_given_batch_norms() (in module aimet_tensorflow.keras.batch_norm_fold) (in module aimet_torch.batch_norm_fold)

forward() (aimet_torch.v2.nn.QuantizationMixin method), [1] (aimet_torch.v2.quantization.affine.quantizer.Quantize method) (aimet_torch.v2.quantization.affine.quantizer.QuantizeDequantize method) freeze_base_model() (aimet_torch.peft.PeftQuantUtils method) freeze_base_model_activation_quantizers() (aimet_torch.peft.PeftQuantUtils method) freeze_base_model_param_quantizers() (aimet_torch.peft.PeftQuantUtils method) from_module() (aimet_torch.v2.nn.QuantizationMixin class method)

G

generate_layer_outputs() (aimet_onnx.layer_output_utils.LayerOutputUtil method) (aimet_tensorflow.keras.layer_output_utils.LayerOutputUtil method) (aimet_torch.layer_output_utils.LayerOutputUtil method) get_activation_quantizers() (aimet_onnx.amp.quantizer_groups.QuantizerGroup method) get_active_quantizers() (aimet_onnx.amp.quantizer_groups.QuantizerGroup method) (aimet_torch.amp.quantizer_groups.QuantizerGroup method) get_candidate() (aimet_onnx.amp.quantizer_groups.QuantizerGroup method) (aimet_torch.amp.quantizer_groups.QuantizerGroup method) get_default_kernel() (aimet_torch.v2.nn.QuantizationMixin class method)

get_encoding() (aimet_torch.v2.quantization.affine.quantizer.QuantizerBase method) get_fp_lora_layer() (aimet_torch.peft.PeftQuantUtils method) get_input_quantizer_modules() (aimet_torch.amp.quantizer_groups.QuantizerGroup method) get_kernel() (aimet_torch.v2.nn.QuantizationMixin method) get_legacy_encodings() (aimet_torch.v2.quantization.affine.quantizer.QuantizerBase method) get_param_quantizers() (aimet_onnx.amp.quantizer_groups.QuantizerGroup method) get_quant_scheme_candidates() (aimet_onnx.auto_quant_v2.AutoQuant method) get_quantized_lora_layer() (aimet_torch.peft.PeftQuantUtils method) GPTVQParameters (class in aimet_torch.gptvq.defs) GreedySelectionParameters (class in aimet_common.defs)

I

implements() (aimet_torch.v2.nn.QuantizationMixin class method)

input_quantizers (aimet_torch.v2.nn.QuantizationMixin attribute), [1] is_initialized() (aimet_torch.v2.quantization.affine.quantizer.QuantizerBase method)

L

LayerOutputUtil (class in aimet_onnx.layer_output_utils) (class in aimet_tensorflow.keras.layer_output_utils) (class in aimet_torch.layer_output_utils)

load_checkpoint() (aimet_torch.v1.quantsim method)

M

mac (aimet_common.defs.CostMetric attribute) manual (aimet_tensorflow.keras.defs.SpatialSvdParameters.Mode attribute) (aimet_torch.defs.ChannelPruningParameters.Mode attribute) (aimet_torch.defs.SpatialSvdParameters.Mode attribute) (aimet_torch.defs.WeightSvdParameters.Mode attribute) memory (aimet_common.defs.CostMetric attribute) MinMaxEncodingAnalyzer (class in aimet_torch.v2.quantization.encoding_analyzer)

ModelCompressor (class in aimet_tensorflow.keras.compress) (class in aimet_torch.compress) module aimet_torch.v2.quantization.affine aimet_torch.v2.quantization.float ModuleCompRatioPair (class in aimet_tensorflow.keras.defs) (class in aimet_torch.defs)

N

NamingScheme (class in aimet_torch.layer_output_utils)

no_activation (aimet_common.defs.ActivationType attribute)

O

ONNX (aimet_torch.layer_output_utils.NamingScheme attribute)

optimize() (aimet_onnx.auto_quant_v2.AutoQuant method) output_quantizers (aimet_torch.v2.nn.QuantizationMixin attribute), [1]

P

param_quantizers (aimet_torch.v2.nn.QuantizationMixin attribute), [1] PeftQuantUtils (class in aimet_torch.peft) PercentileEncodingAnalyzer (class in aimet_torch.v2.quantization.encoding_analyzer) perform_per_layer_analysis_by_disabling_quant_wrappers() (aimet_torch.v1.quant_analyzer.QuantAnalyzer method) perform_per_layer_analysis_by_disabling_quantizers() (aimet_onnx.quant_analyzer.QuantAnalyzer method) perform_per_layer_analysis_by_enabling_quant_wrappers() (aimet_torch.v1.quant_analyzer.QuantAnalyzer method)

perform_per_layer_analysis_by_enabling_quantizers() (aimet_onnx.quant_analyzer.QuantAnalyzer method) post_training_percentile (aimet_common.defs.QuantScheme attribute), [1], [2] post_training_tf (aimet_common.defs.QuantScheme attribute), [1], [2] post_training_tf_enhanced (aimet_common.defs.QuantScheme attribute), [1], [2] prepare_model() (in module aimet_tensorflow.keras.model_preparer) (in module aimet_torch.model_preparer) PYTORCH (aimet_torch.layer_output_utils.NamingScheme attribute)

Q

QuantAnalyzer (class in aimet_onnx.quant_analyzer) (class in aimet_tensorflow.keras.quant_analyzer) (class in aimet_torch.v1.quant_analyzer) QuantizationMixin (class in aimet_torch.v2.nn), [1] QuantizationSimModel (class in aimet_onnx.quantsim) (class in aimet_tensorflow.keras.quantsim) (class in aimet_torch.v1.quantsim) Quantize (class in aimet_torch.v2.quantization.affine) (class in aimet_torch.v2.quantization.affine.quantizer) quantize() (aimet_torch.v2.quantization.tensor.DequantizedTensor method) (aimet_torch.v2.quantization.tensor.QuantizedTensor method) (in module aimet_torch.v2.quantization.affine)

quantize_dequantize() (in module aimet_torch.v2.quantization.affine) quantize_lora_scale_with_fixed_range() (aimet_torch.peft.PeftQuantUtils method) quantized_repr() (aimet_torch.v2.quantization.tensor.DequantizedTensor method) (aimet_torch.v2.quantization.tensor.QuantizedTensor method) QuantizeDequantize (class in aimet_torch.v2.quantization.affine) (class in aimet_torch.v2.quantization.affine.quantizer) (class in aimet_torch.v2.quantization.float), [1] QuantizedTensor (class in aimet_torch.v2.quantization.tensor) QuantizerBase (class in aimet_torch.v2.quantization.affine.quantizer) QuantizerGroup (class in aimet_onnx.amp.quantizer_groups) (class in aimet_torch.amp.quantizer_groups) QuantParams (class in aimet_torch.v1.quantsim) QuantScheme (class in aimet_common.defs), [1], [2]

R

reestimate_bn_stats() (in module aimet_tensorflow.keras.bn_reestimation) (in module aimet_torch.bn_reestimation) register_quantization_parameter() (aimet_torch.v2.quantization.affine.quantizer.QuantizerBase method) relu (aimet_common.defs.ActivationType attribute)

relu6 (aimet_common.defs.ActivationType attribute) replace_lora_layers_with_quantizable_layers() (aimet_torch.peft method) reset_stats() (aimet_torch.v2.quantization.encoding_analyzer.EncodingAnalyzer method) run_inference() (aimet_onnx.auto_quant_v2.AutoQuant method)

S

save_checkpoint() (aimet_torch.v1.quantsim method) scale_cls_sets() (in module aimet_tensorflow.keras.cross_layer_equalization.CrossLayerScaling) (in module aimet_torch.cross_layer_equalization.CrossLayerScaling) scale_model() (in module aimet_tensorflow.keras.cross_layer_equalization.CrossLayerScaling) (in module aimet_torch.cross_layer_equalization.CrossLayerScaling) ScatterDense (class in aimet_torch.nn.modules.custom) set_activation_quantizers_to_float() (in module aimet_torch.v2.quantsim.config_utils) set_adaround_params() (aimet_onnx.auto_quant_v2.AutoQuant method) set_bitwidth_for_lora_adapters() (aimet_torch.peft.PeftQuantUtils method) set_blockwise_quantization_for_weights() (in module aimet_torch.v2.quantsim.config_utils) set_default_kernel() (aimet_torch.v2.nn.QuantizationMixin class method) set_grouped_blockwise_quantization_for_weights() (in module aimet_torch.v2.quantsim.config_utils) set_kernel() (aimet_torch.v2.nn.QuantizationMixin method) set_legacy_encodings() (aimet_torch.v2.quantization.affine.quantizer.QuantizerBase method) set_quant_scheme_candidates() (aimet_onnx.auto_quant_v2.AutoQuant method)

set_quantizers_to_candidate() (aimet_onnx.amp.quantizer_groups.QuantizerGroup method) (aimet_torch.amp.quantizer_groups.QuantizerGroup method) SparseTensorWrapper (class in aimet_torch.nn.modules.custom) spatial_svd (aimet_common.defs.CompressionScheme attribute) SpatialSvdParameters (class in aimet_tensorflow.keras.defs) (class in aimet_torch.defs) SpatialSvdParameters.AutoModeParams (class in aimet_tensorflow.keras.defs) (class in aimet_torch.defs) SpatialSvdParameters.ManualModeParams (class in aimet_tensorflow.keras.defs) (class in aimet_torch.defs) SpatialSvdParameters.Mode (class in aimet_tensorflow.keras.defs) (class in aimet_torch.defs) sqnr() (aimet_onnx.amp.mixed_precision_algo.EvalCallbackFactory method) (aimet_torch.amp.mixed_precision_algo.EvalCallbackFactory method) SqnrEncodingAnalyzer (class in aimet_torch.v2.quantization.encoding_analyzer)

T

to_list() (aimet_onnx.amp.quantizer_groups.QuantizerGroup method) (aimet_torch.amp.quantizer_groups.QuantizerGroup method) TORCHSCRIPT (aimet_torch.layer_output_utils.NamingScheme attribute)

track_lora_meta_data() (aimet_torch.peft method) training_range_learning_with_tf_enhanced_init (aimet_common.defs.QuantScheme attribute), [1], [2] training_range_learning_with_tf_init (aimet_common.defs.QuantScheme attribute), [1], [2]

U

update_stats() (aimet_torch.v2.quantization.encoding_analyzer.EncodingAnalyzer method)

V

visualize_changes_after_optimization() (in module aimet_torch.visualize_model) visualize_relative_weight_ranges_to_identify_problematic_layers() (in module aimet_torch.visualize_model)

visualize_stats() (in module aimet_torch.v2.visualization_tools) visualize_weight_ranges() (in module aimet_torch.visualize_model) VisualizeCompression (class in aimet_torch.visualize_serialized_data)

W

weight_svd (aimet_common.defs.CompressionScheme attribute) WeightSvdParameters (class in aimet_torch.defs)

WeightSvdParameters.AutoModeParams (class in aimet_torch.defs) WeightSvdParameters.ManualModeParams (class in aimet_torch.defs) WeightSvdParameters.Mode (class in aimet_torch.defs)