Index

A

ActivateDropoutNode (class in quark.torch.quantization.graph.optimization.activate_dropout) AdaroundConstants (class in quark.onnx.finetuning.create_torch.quant_base_ops) AdaroundQuantizer (class in quark.onnx.finetuning.create_torch.quant_base_ops) AddQDQToQOPTransform (class in quark.onnx.optimizations.convert_transforms) adjust_quantize_info() (in module quark.onnx.refine) align_quantize_info() (in module quark.onnx.refine) allow_exported_model_train_eval() (in module quark.torch.quantization.graph.torch_utils) allow_multi_consumers (quark.onnx.graph_transformations.transforms.Transform property)

apply() (quark.onnx.graph_transformations.transforms_pipeline.TransformsPipeline method) (quark.onnx.optimizations.convert_transforms_pipeline.ConvertQDQToQOPTransformsPipeline method) (quark.onnx.optimizations.convert_transforms_pipeline.RemoveQDQTransformsPipeline method) augment_graph() (quark.onnx.calibrate.PowOfTwoCalibrater method) auto_mixprecision() (in module quark.onnx.finetuning.auto_mixprecision) average_L2() (in module quark.onnx.finetuning.onnx_evaluate) AWQConfig (class in quark.torch.quantization.config.config) AwqProcessor (class in quark.torch.algorithm.awq.awq)

B

backward() (quark.onnx.finetuning.create_torch.quant_base_ops.BaseDeQuantFunction static method) (quark.onnx.finetuning.create_torch.quant_base_ops.BaseQuantDequantFunction static method) (quark.onnx.finetuning.create_torch.quant_base_ops.BaseQuantFunction static method) (quark.onnx.finetuning.create_torch.quant_base_ops.INTDeQuantFunction static method) (quark.onnx.finetuning.create_torch.quant_base_ops.INTQuantDequantFunction static method) (quark.onnx.finetuning.create_torch.quant_base_ops.INTQuantFunction static method) (quark.onnx.finetuning.create_torch.quant_base_ops.RoundHalfToEven static method) (quark.torch.kernel.DequantE4M3Function static method) (quark.torch.kernel.FakeQuantizeFunction static method) (quark.torch.kernel.QuantE4M3Function static method) (quark.torch.kernel.RealQuantizeFunction static method)

BaseAlgoProcessor (class in quark.torch.algorithm.processor) BaseDeQuantFunction (class in quark.onnx.finetuning.create_torch.quant_base_ops) BaseQuantDequantFunction (class in quark.onnx.finetuning.create_torch.quant_base_ops) BaseQuantFunction (class in quark.onnx.finetuning.create_torch.quant_base_ops) BaseVocab (class in quark.torch.export.gguf_export.utils) BpeVocab (class in quark.torch.export.gguf_export.utils)

C

CachedDataReader (class in quark.onnx.quant_utils) CachedDataset (class in quark.onnx.finetuning.torch_utils) calc_recon_loss() (quark.onnx.finetuning.train_torch.train_model_loss.TrainLoss static method) calc_round_loss() (quark.onnx.finetuning.train_torch.train_model_loss.TrainLoss class method) calculate_qparams() (quark.torch.quantization.observer.observer.PerGroupMinMaxObserver method) (quark.torch.quantization.observer.observer.UniformScalingObserver method) calculate_quantization_params() (quark.onnx.onnx_quantizer.VitisONNXQuantizer method) check_hard_sigmoid_condition() (in module quark.onnx.quant_utils) check_min_max_valid() (in module quark.torch.quantization.nn.utils) check_model_quantizable() (in module quark.onnx.quant_utils) check_onnx_model() (in module quark.onnx.quant_utils) check_reduce_mean_condition() (in module quark.onnx.quant_utils) check_relu_like_node() (in module quark.onnx.quant_utils) CLE_PAIR_TYPE (class in quark.onnx.equalization) cle_transforms() (in module quark.onnx.equalization) Clip (class in quark.onnx.finetuning.create_torch.create_model_ops) collect() (quark.onnx.calibrate.PowOfTwoCollector method) collect_data() (quark.onnx.calibrate.PowOfTwoCalibrater method) compute_collection_result() (quark.onnx.calibrate.PowOfTwoCollector method) compute_range() (quark.onnx.calibrate.PowOfTwoCalibrater method) compute_scale_zp() (in module quark.onnx.quant_utils) compute_scale_zp_fp() (in module quark.onnx.quant_utils) Config (class in quark.onnx.quantization.config.config) (class in quark.torch.quantization.config.config) convert_act() (in module quark.onnx.finetuning.create_torch.create_model_ops) convert_bn_to_conv() (quark.onnx.optimize.Optimize method) convert_clip_to_relu() (quark.onnx.optimize.Optimize method) convert_conv() (in module quark.onnx.finetuning.create_torch.create_model_ops)

convert_customqdq_to_qdq() (in module quark.onnx.tools.convert_customqdq_to_qdq) convert_exported_model_to_gguf() (in module quark.torch.export.gguf_export.api) convert_float16_to_float() (in module quark.onnx.tools.float16) convert_float_to_float16() (in module quark.onnx.tools.float16) convert_float_to_float16_model_path() (in module quark.onnx.tools.float16) convert_gemm() (in module quark.onnx.finetuning.create_torch.create_model_ops) convert_initializers_to_float() (in module quark.onnx.tools.convert_quant_to_float) convert_lstm_to_customlstm() (in module quark.onnx.tools.convert_lstm_to_customlstm) convert_norm() (in module quark.onnx.finetuning.create_torch.create_model_ops) convert_np_to_float() (in module quark.onnx.tools.float16) convert_np_to_float16() (in module quark.onnx.tools.float16) convert_onnx_to_torch() (in module quark.onnx.finetuning.torch_utils) convert_ops_to_modules() (in module quark.onnx.finetuning.create_torch.create_model_ops) convert_qdq_to_qop() (in module quark.onnx.tools.convert_qdq_to_qop) convert_reduce_mean_to_global_avg_pool() (quark.onnx.optimize.Optimize method) convert_split_to_slice() (quark.onnx.optimize.Optimize method) convert_tensor_float16_to_float() (in module quark.onnx.tools.float16) convert_tensor_float_to_float16() (in module quark.onnx.tools.float16) convert_torch_to_onnx() (in module quark.onnx.finetuning.torch_utils) ConvertQDQToQOPTransformsPipeline (class in quark.onnx.optimizations.convert_transforms_pipeline) ConvQDQToQOPTransform (class in quark.onnx.optimizations.convert_transforms) create_calibrator_float_scale() (in module quark.onnx.calibrate) create_calibrator_power_of_two() (in module quark.onnx.calibrate) create_session() (in module quark.onnx.finetuning.onnx_evaluate) custom_ops_infer_shapes() (in module quark.onnx.tools.convert_customqdq_to_qdq) (in module quark.onnx.tools.convert_lstm_to_customlstm) CustomFormatter (class in quark.torch.utils.log) customqdq_to_contribqdq() (in module quark.onnx.quant_utils)

D

DequantE4M3Function (class in quark.torch.kernel) dequantize_data() (in module quark.onnx.quant_utils) DeviceType (class in quark.torch.quantization.config.type)

dpu_leaky_relu_alpha() (in module quark.onnx.quant_utils) Dtype (class in quark.torch.quantization.config.type) dump_model() (in module quark.onnx.quant_utils) DuplicateFilter (class in quark.torch.utils.log)

E

EmbeddingType (class in quark.torch.export.json_export.builder.llm_info) EntropyCalibrater (class in quark.onnx.calibrate) Equalization (class in quark.onnx.equalization) export_gguf_model() (quark.torch.export.api.ModelExporter method) export_model_info() (quark.torch.export.api.ModelExporter method) export_onnx_model() (quark.torch.export.api.ModelExporter method) ExportConv2d (class in quark.torch.export.nn.modules.export_operator) ExporterConfig (class in quark.torch.export.config.config)

ExportLinear (class in quark.torch.export.nn.modules.export_operator) extra_repr() (quark.torch.quantization.observer.observer.PlaceholderObserver method) (quark.torch.quantization.observer.observer.UniformScalingObserver method) (quark.torch.quantization.tensor_quantize.FakeQuantize method) extract_attr_values() (in module quark.onnx.finetuning.create_torch.create_model_ops) extract_padding_params() (in module quark.onnx.finetuning.create_torch.create_model_ops) extract_padding_params_for_conv() (in module quark.onnx.finetuning.create_torch.create_model_ops) extract_weight_and_bias() (in module quark.onnx.finetuning.create_torch.create_model_ops)

F

FakeQuantize (class in quark.torch.quantization.tensor_quantize) FakeQuantizeBase (class in quark.torch.quantization.tensor_quantize) FakeQuantizeFunction (class in quark.torch.kernel) fast_finetune() (in module quark.onnx.finetuning.fast_finetune) filter() (quark.torch.utils.log.DuplicateFilter method) find_int16_scale() (in module quark.onnx.quant_utils) find_quant_scale_zp() (quark.onnx.onnx_quantizer.VitisONNXQuantizer method) find_quantized_value() (quark.onnx.onnx_quantizer.VitisONNXQuantizer method) fold_batch_norm() (quark.onnx.optimize.Optimize method) fold_batch_norm_after_concat() (quark.onnx.optimize.Optimize method) format() (quark.torch.utils.log.CustomFormatter method) forward() (quark.onnx.finetuning.create_torch.create_model.TorchModel method) (quark.onnx.finetuning.create_torch.quant_base_ops.BaseDeQuantFunction static method) (quark.onnx.finetuning.create_torch.quant_base_ops.BaseQuantDequantFunction static method) (quark.onnx.finetuning.create_torch.quant_base_ops.BaseQuantFunction static method) (quark.onnx.finetuning.create_torch.quant_base_ops.INTDeQuantFunction static method) (quark.onnx.finetuning.create_torch.quant_base_ops.INTQuantDequantFunction static method) (quark.onnx.finetuning.create_torch.quant_base_ops.INTQuantFunction static method) (quark.onnx.finetuning.create_torch.quant_base_ops.RoundHalfToEven static method) (quark.torch.algorithm.rotation.rotation_utils.RMSNorm method) (quark.torch.kernel.DequantE4M3Function static method) (quark.torch.kernel.FakeQuantizeFunction static method) (quark.torch.kernel.QuantE4M3Function static method) (quark.torch.kernel.RealQuantizeFunction static method) (quark.torch.quantization.observer.observer.PerTensorHistogramObserver method) (quark.torch.quantization.observer.observer.PerTensorMinMaxObserver method)

freeze() (quark.torch.quantization.api.ModelQuantizer static method) FreezedFakeQuantize (class in quark.torch.quantization.tensor_quantize) fuse_instance_norm() (quark.onnx.optimize.Optimize method) fuse_l2_norm() (quark.onnx.optimize.Optimize method) fuse_layer_norm() (quark.onnx.optimize.Optimize method)

G

generate_initializer() (in module quark.onnx.utils.model_utils) generate_input_initializer() (in module quark.onnx.graph_transformations.model_transformer_test) get_annotate_tensors() (in module quark.onnx.quant_utils) get_bias() (quark.onnx.finetuning.create_torch.create_model.TorchModel method) get_clip_min_max() (in module quark.onnx.quant_utils) get_configs() (quark.onnx.graph_transformations.transforms_pipeline.TransformsPipeline method) get_datatype_shape() (in module quark.onnx.quant_utils) get_exclude_nodes() (in module quark.onnx.quant_utils) get_hadamard_matrices() (in module quark.torch.algorithm.rotation.hadamard) get_min_max_by_mse() (quark.torch.quantization.observer.observer.PerTensorMSEObserver method) get_min_max_by_percentile() (quark.torch.quantization.observer.observer.PerTensorPercentileObserver method) get_modules_optimized_bias() (in module quark.onnx.finetuning.create_torch.create_model_ops) get_modules_optimized_weight() (in module quark.onnx.finetuning.create_torch.create_model_ops)

get_nested_attr_from_module() (in module quark.torch.algorithm.utils.module) get_next() (quark.onnx.quant_utils.CachedDataReader method) (quark.onnx.quant_utils.PathDataReader method) (quark.onnx.quant_utils.RandomDataReader method) (quark.onnx.tools.save_tensor_hist.HistDataReader method) get_qdq_to_remove() (in module quark.onnx.quant_utils) get_qmin_qmax_for_qType() (in module quark.onnx.quant_utils) get_qrange_for_qType() (in module quark.onnx.quant_utils) get_rotation_matrix() (in module quark.torch.algorithm.rotation.rotation_utils) get_tensor_value() (in module quark.onnx.utils.model_utils) get_weight() (quark.onnx.finetuning.create_torch.create_model.TorchModel method) GPTQConfig (class in quark.torch.quantization.config.config) GptqProcessor (class in quark.torch.algorithm.gptq.gptq) GraphTransform (class in quark.torch.quantization.graph.fx.base)

H

hardmard_transform() (in module quark.torch.algorithm.rotation.hadamard)

HistDataReader (class in quark.onnx.tools.save_tensor_hist)

I

infer_shape() (in module quark.onnx.quant_utils) inference_model() (in module quark.onnx.finetuning.onnx_evaluate) initialize_alpha() (quark.onnx.finetuning.create_torch.quant_base_ops.AdaroundQuantizer method) insert_quantizer() (in module quark.torch.quantization.graph.processor.insert_quantizer) Int16Method (class in quark.onnx.quant_utils) INTDeQuantFunction (class in quark.onnx.finetuning.create_torch.quant_base_ops) INTQuantDequantFunction (class in quark.onnx.finetuning.create_torch.quant_base_ops) INTQuantFunction (class in quark.onnx.finetuning.create_torch.quant_base_ops)

is_approximately_equal() (in module quark.onnx.quant_utils) is_clip_with_min_max() (in module quark.onnx.quant_utils) is_conv1d_node() (in module quark.torch.quantization.graph.torch_utils) is_conv2d_node() (in module quark.torch.quantization.graph.torch_utils) is_conv3d_node() (in module quark.torch.quantization.graph.torch_utils) is_leaky_relu_with_alpha() (in module quark.onnx.quant_utils) is_node_needs_annotated() (in module quark.onnx.quant_utils) is_ort_version_below() (in module quark.onnx.quant_utils)

J

JsonExporterConfig (class in quark.torch.export.config.config)

L

LayerNormType (class in quark.torch.export.json_export.builder.llm_info) LlamaHfVocab (class in quark.torch.export.gguf_export.utils)

LlamaModelWriter (class in quark.torch.export.gguf_export.gguf_model_writer) load_weight_and_bias() (in module quark.onnx.finetuning.create_torch.create_model_ops)

M

MatMulQDQToQOPTransform (class in quark.onnx.optimizations.convert_transforms) MinMaxCalibrater (class in quark.onnx.calibrate) ModelExporter (class in quark.torch.export.api) ModelOptimizer (class in quark.onnx.finetuning.train_torch.train_model) ModelQuantizer (class in quark.onnx.quantization.api) (class in quark.torch.quantization.api) ModelTransformer (class in quark.onnx.graph_transformations.model_transformer) ModelTransformer.NodeType (class in quark.onnx.graph_transformations.model_transformer) ModelTransformerTest (class in quark.onnx.graph_transformations.model_transformer_test) ModelTransformerTest.RemoveRelu (class in quark.onnx.graph_transformations.model_transformer_test) ModelTransformerTest.ReplaceWholeModel (class in quark.onnx.graph_transformations.model_transformer_test) ModelWriter (class in quark.torch.export.gguf_export.gguf_model_writer) modified_annotate_input() (in module quark.onnx.quant_utils) module quark quark.onnx quark.onnx.bias_correction quark.onnx.calibrate quark.onnx.cpu_quantizer quark.onnx.equalization quark.onnx.finetuning quark.onnx.finetuning.auto_mixprecision quark.onnx.finetuning.create_torch quark.onnx.finetuning.create_torch.create_model quark.onnx.finetuning.create_torch.create_model_ops quark.onnx.finetuning.create_torch.create_model_test quark.onnx.finetuning.create_torch.create_model_utils quark.onnx.finetuning.create_torch.quant_base_ops quark.onnx.finetuning.create_torch.quant_conv_ops quark.onnx.finetuning.create_torch.quant_gemm_ops quark.onnx.finetuning.create_torch.quant_norm_ops quark.onnx.finetuning.fast_finetune quark.onnx.finetuning.onnx_evaluate quark.onnx.finetuning.onnx_subgraph quark.onnx.finetuning.torch_utils quark.onnx.finetuning.torch_utils_test quark.onnx.finetuning.train_torch quark.onnx.finetuning.train_torch.train_model quark.onnx.finetuning.train_torch.train_model_loss quark.onnx.finetuning.train_torch.train_model_param quark.onnx.graph_transformations quark.onnx.graph_transformations.model_transformer quark.onnx.graph_transformations.model_transformer_test quark.onnx.graph_transformations.transforms quark.onnx.graph_transformations.transforms_pipeline quark.onnx.mprecision quark.onnx.mprecision.mixed_bfp quark.onnx.onnx_quantizer quark.onnx.operators quark.onnx.operators.custom_ops quark.onnx.operators.custom_ops.test_custom_ops_bfp quark.onnx.operators.custom_ops.test_custom_ops_in quark.onnx.operators.custom_ops.test_custom_ops_qdq quark.onnx.operators.custom_ops.vai_custom_op quark.onnx.operators.vai_ops quark.onnx.operators.vai_ops.concat quark.onnx.operators.vai_ops.hardsigmoid quark.onnx.operators.vai_ops.layernorm quark.onnx.operators.vai_ops.prelu quark.onnx.operators.vai_ops.qdq_ops quark.onnx.operators.vai_ops.softmax quark.onnx.optimizations quark.onnx.optimizations.convert_transforms quark.onnx.optimizations.convert_transforms_pipeline quark.onnx.optimize quark.onnx.qdq_quantizer quark.onnx.quant_utils quark.onnx.quantization quark.onnx.quantization.api quark.onnx.quantization.config quark.onnx.quantization.config.config quark.onnx.quantization.config.custom_config quark.onnx.quantize quark.onnx.refine quark.onnx.registry quark.onnx.simulate_dpu quark.onnx.simulate_dpu_softmax quark.onnx.smooth_quant quark.onnx.tools quark.onnx.tools.convert_a8w8_npu_to_a8w8_cpu quark.onnx.tools.convert_customqdq_to_qdq quark.onnx.tools.convert_dynamic_to_fixed quark.onnx.tools.convert_fp16_to_fp32 quark.onnx.tools.convert_lstm_to_customlstm quark.onnx.tools.convert_nchw_to_nhwc quark.onnx.tools.convert_onnx_to_onnxtxt quark.onnx.tools.convert_onnxtxt_to_onnx quark.onnx.tools.convert_qdq_to_qop quark.onnx.tools.convert_quant_to_float quark.onnx.tools.convert_s8s8_to_u8s8 quark.onnx.tools.convert_u16s8_to_s16s8 quark.onnx.tools.float16 quark.onnx.tools.print_a16w8_a8w8_nodes quark.onnx.tools.random_quantize quark.onnx.tools.remove_initializer_from_input quark.onnx.tools.remove_qdq quark.onnx.tools.save_tensor_hist quark.onnx.tools.save_weights_hist quark.onnx.utils quark.onnx.utils.model_utils quark.torch quark.torch.algorithm quark.torch.algorithm.api quark.torch.algorithm.awq quark.torch.algorithm.awq.awq quark.torch.algorithm.awq.modules quark.torch.algorithm.awq.modules.act quark.torch.algorithm.awq.scale quark.torch.algorithm.awq.smooth quark.torch.algorithm.gptq quark.torch.algorithm.gptq.gptq quark.torch.algorithm.gptq.quantizer quark.torch.algorithm.processor quark.torch.algorithm.rotation quark.torch.algorithm.rotation.hadamard quark.torch.algorithm.rotation.rotation quark.torch.algorithm.rotation.rotation_utils quark.torch.algorithm.utils quark.torch.algorithm.utils.module quark.torch.algorithm.utils.prepare quark.torch.algorithm.utils.utils quark.torch.export quark.torch.export.api quark.torch.export.config quark.torch.export.config.config quark.torch.export.config.custom_config quark.torch.export.gguf_export quark.torch.export.gguf_export.api quark.torch.export.gguf_export.gguf_model_converter quark.torch.export.gguf_export.gguf_model_writer quark.torch.export.gguf_export.tensor_convert quark.torch.export.gguf_export.utils quark.torch.export.json_export quark.torch.export.json_export.builder quark.torch.export.json_export.builder.llm_info quark.torch.export.json_export.builder.llm_info_builder quark.torch.export.json_export.builder.native_model_info_builder quark.torch.export.json_export.converter quark.torch.export.json_export.converter.llm_info_converter quark.torch.export.json_export.utils quark.torch.export.json_export.utils.utils quark.torch.export.main_export quark.torch.export.main_export.model_post_process quark.torch.export.main_export.quant_config_parser quark.torch.export.nn quark.torch.export.nn.modules quark.torch.export.nn.modules.export_operator quark.torch.export.onnx quark.torch.export.utils quark.torch.kernel quark.torch.kernel.hw_emulation quark.torch.kernel.hw_emulation.extensions quark.torch.kernel.hw_emulation.hw_emulation_interface quark.torch.quantization quark.torch.quantization.api quark.torch.quantization.config quark.torch.quantization.config.config quark.torch.quantization.config.config_verification quark.torch.quantization.config.type quark.torch.quantization.graph quark.torch.quantization.graph.fx quark.torch.quantization.graph.fx.base quark.torch.quantization.graph.optimization quark.torch.quantization.graph.optimization.activate_dropout quark.torch.quantization.graph.optimization.remove_dropout_node quark.torch.quantization.graph.optimization.replace_conv2d_to_qtconv2d quark.torch.quantization.graph.optimization.replace_conv_bn_to_qt_model quark.torch.quantization.graph.optimization.replace_linear_to_qtlinear quark.torch.quantization.graph.optimization.utils quark.torch.quantization.graph.processor quark.torch.quantization.graph.processor.insert_quantizer quark.torch.quantization.graph.processor.processor quark.torch.quantization.graph.processor.processor_utils quark.torch.quantization.graph.torch_utils quark.torch.quantization.model_transformation quark.torch.quantization.nn quark.torch.quantization.nn.modules quark.torch.quantization.nn.modules.mixin quark.torch.quantization.nn.modules.quantize_conv quark.torch.quantization.nn.modules.quantize_conv_bn_fused quark.torch.quantization.nn.modules.quantize_linear quark.torch.quantization.nn.utils quark.torch.quantization.observer quark.torch.quantization.observer.observer quark.torch.quantization.tensor_quantize quark.torch.quantization.utils quark.torch.utils quark.torch.utils.log quark.torch.utils.pack

MulQDQToQOPTransform (class in quark.onnx.optimizations.convert_transforms)

N

NodeTree (class in quark.onnx.graph_transformations.transforms)

NoVocab (class in quark.torch.export.gguf_export.utils)

O

ObserverBase (class in quark.torch.quantization.observer.observer) OnnxExporterConfig (class in quark.torch.export.config.config) ONNXQuantizer (class in quark.onnx.onnx_quantizer)

Optimize (class in quark.onnx.optimize) optimize() (in module quark.onnx.optimize) optimize_module() (in module quark.onnx.finetuning.torch_utils) OpTypePattern (class in quark.onnx.graph_transformations.transforms)

P

param_is_symmetric() (in module quark.onnx.finetuning.create_torch.create_model_ops) parse_options_to_params() (in module quark.onnx.finetuning.torch_utils) PathDataReader (class in quark.onnx.quant_utils) pattern() (quark.onnx.graph_transformations.model_transformer_test.ModelTransformerTest.RemoveRelu method) (quark.onnx.graph_transformations.model_transformer_test.ModelTransformerTest.ReplaceWholeModel method) (quark.onnx.graph_transformations.transforms.Transform method) (quark.onnx.optimizations.convert_transforms.AddQDQToQOPTransform method) (quark.onnx.optimizations.convert_transforms.ConvQDQToQOPTransform method) (quark.onnx.optimizations.convert_transforms.MatMulQDQToQOPTransform method) (quark.onnx.optimizations.convert_transforms.MulQDQToQOPTransform method) (quark.onnx.optimizations.convert_transforms.RemoveQDQTransform method) (quark.onnx.optimizations.convert_transforms.SigmoidQDQToQOPTransform method) PerBlockMXObserver (class in quark.torch.quantization.observer.observer)

PercentileCalibrater (class in quark.onnx.calibrate) PerChannelMinMaxObserver (class in quark.torch.quantization.observer.observer) PerGroupMinMaxObserver (class in quark.torch.quantization.observer.observer) PerTensorHistogramObserver (class in quark.torch.quantization.observer.observer) PerTensorMinMaxObserver (class in quark.torch.quantization.observer.observer) PerTensorMSEObserver (class in quark.torch.quantization.observer.observer) PerTensorPercentileObserver (class in quark.torch.quantization.observer.observer) PlaceholderObserver (class in quark.torch.quantization.observer.observer) pos2scale() (in module quark.onnx.quant_utils) PowerOfTwoMethod (class in quark.onnx.quant_utils) PowOfTwoCalibrater (class in quark.onnx.calibrate) PowOfTwoCollector (class in quark.onnx.calibrate) print_quantize_info() (in module quark.onnx.quant_utils)

Q

QConv1d (class in quark.onnx.finetuning.create_torch.quant_conv_ops) QConv2d (class in quark.onnx.finetuning.create_torch.quant_conv_ops) QConv3d (class in quark.onnx.finetuning.create_torch.quant_conv_ops) QConvTranspose1d (class in quark.onnx.finetuning.create_torch.quant_conv_ops) QConvTranspose2d (class in quark.onnx.finetuning.create_torch.quant_conv_ops) QConvTranspose3d (class in quark.onnx.finetuning.create_torch.quant_conv_ops) QDQModule (class in quark.onnx.finetuning.create_torch.quant_base_ops) QDQNPUTransformerQuantizer (class in quark.onnx.qdq_quantizer) QDQQuantizer (class in quark.onnx.qdq_quantizer) QGemm (class in quark.onnx.finetuning.create_torch.quant_gemm_ops) QInstanceNorm1d (class in quark.onnx.finetuning.create_torch.quant_norm_ops) QInstanceNorm2d (class in quark.onnx.finetuning.create_torch.quant_norm_ops) QInstanceNorm3d (class in quark.onnx.finetuning.create_torch.quant_norm_ops) QSchemeType (class in quark.torch.quantization.config.type) QuantConv2d (class in quark.torch.quantization.nn.modules.quantize_conv) QuantE4M3Function (class in quark.torch.kernel) QuantizationConfig (class in quark.onnx.quantization.config.config) (class in quark.torch.quantization.config.config) QuantizationMode (class in quark.torch.quantization.config.type) QuantizationSpec (class in quark.torch.quantization.config.config) quantize_bias_static() (quark.onnx.onnx_quantizer.VitisONNXQuantizer method) quantize_data_pof2s() (in module quark.onnx.quant_utils) quantize_initializer() (quark.onnx.onnx_quantizer.VitisONNXQuantizer method) quantize_model() (quark.onnx.quantization.api.ModelQuantizer method) (quark.torch.quantization.api.ModelQuantizer method) quantize_weight() (quark.onnx.onnx_quantizer.VitisONNXQuantizer method) quantize_weight_per_channel() (quark.onnx.onnx_quantizer.VitisONNXQuantizer method) QuantizedConvBatchNorm2d (class in quark.torch.quantization.nn.modules.quantize_conv_bn_fused) Quantizer (class in quark.onnx.finetuning.create_torch.quant_base_ops) (class in quark.torch.algorithm.gptq.quantizer) QuantizeWrapper (class in quark.onnx.finetuning.create_torch.quant_base_ops) QuantLinear (class in quark.torch.quantization.nn.modules.quantize_linear) QuantMixin (class in quark.torch.quantization.nn.modules.mixin) quark module quark.onnx module quark.onnx.bias_correction module quark.onnx.calibrate module quark.onnx.cpu_quantizer module quark.onnx.equalization module quark.onnx.finetuning module quark.onnx.finetuning.auto_mixprecision module quark.onnx.finetuning.create_torch module quark.onnx.finetuning.create_torch.create_model module quark.onnx.finetuning.create_torch.create_model_ops module quark.onnx.finetuning.create_torch.create_model_test module quark.onnx.finetuning.create_torch.create_model_utils module quark.onnx.finetuning.create_torch.quant_base_ops module quark.onnx.finetuning.create_torch.quant_conv_ops module quark.onnx.finetuning.create_torch.quant_gemm_ops module quark.onnx.finetuning.create_torch.quant_norm_ops module quark.onnx.finetuning.fast_finetune module quark.onnx.finetuning.onnx_evaluate module quark.onnx.finetuning.onnx_subgraph module quark.onnx.finetuning.torch_utils module quark.onnx.finetuning.torch_utils_test module quark.onnx.finetuning.train_torch module quark.onnx.finetuning.train_torch.train_model module quark.onnx.finetuning.train_torch.train_model_loss module quark.onnx.finetuning.train_torch.train_model_param module quark.onnx.graph_transformations module quark.onnx.graph_transformations.model_transformer module quark.onnx.graph_transformations.model_transformer_test module quark.onnx.graph_transformations.transforms module quark.onnx.graph_transformations.transforms_pipeline module quark.onnx.mprecision module quark.onnx.mprecision.mixed_bfp module quark.onnx.onnx_quantizer module quark.onnx.operators module quark.onnx.operators.custom_ops module quark.onnx.operators.custom_ops.test_custom_ops_bfp module quark.onnx.operators.custom_ops.test_custom_ops_in module quark.onnx.operators.custom_ops.test_custom_ops_qdq module quark.onnx.operators.custom_ops.vai_custom_op module quark.onnx.operators.vai_ops module quark.onnx.operators.vai_ops.concat module quark.onnx.operators.vai_ops.hardsigmoid module quark.onnx.operators.vai_ops.layernorm module quark.onnx.operators.vai_ops.prelu module quark.onnx.operators.vai_ops.qdq_ops module quark.onnx.operators.vai_ops.softmax module quark.onnx.optimizations module quark.onnx.optimizations.convert_transforms module quark.onnx.optimizations.convert_transforms_pipeline module quark.onnx.optimize module quark.onnx.qdq_quantizer module quark.onnx.quant_utils module quark.onnx.quantization module quark.onnx.quantization.api module quark.onnx.quantization.config module quark.onnx.quantization.config.config module quark.onnx.quantization.config.custom_config module quark.onnx.quantize module quark.onnx.refine module quark.onnx.registry module quark.onnx.simulate_dpu module quark.onnx.simulate_dpu_softmax module quark.onnx.smooth_quant module quark.onnx.tools module quark.onnx.tools.convert_a8w8_npu_to_a8w8_cpu module quark.onnx.tools.convert_customqdq_to_qdq module quark.onnx.tools.convert_dynamic_to_fixed module quark.onnx.tools.convert_fp16_to_fp32 module quark.onnx.tools.convert_lstm_to_customlstm module quark.onnx.tools.convert_nchw_to_nhwc module quark.onnx.tools.convert_onnx_to_onnxtxt module quark.onnx.tools.convert_onnxtxt_to_onnx module quark.onnx.tools.convert_qdq_to_qop module quark.onnx.tools.convert_quant_to_float module quark.onnx.tools.convert_s8s8_to_u8s8 module quark.onnx.tools.convert_u16s8_to_s16s8 module quark.onnx.tools.float16 module quark.onnx.tools.print_a16w8_a8w8_nodes module

quark.onnx.tools.random_quantize module quark.onnx.tools.remove_initializer_from_input module quark.onnx.tools.remove_qdq module quark.onnx.tools.save_tensor_hist module quark.onnx.tools.save_weights_hist module quark.onnx.utils module quark.onnx.utils.model_utils module quark.torch module quark.torch.algorithm module quark.torch.algorithm.api module quark.torch.algorithm.awq module quark.torch.algorithm.awq.awq module quark.torch.algorithm.awq.modules module quark.torch.algorithm.awq.modules.act module quark.torch.algorithm.awq.scale module quark.torch.algorithm.awq.smooth module quark.torch.algorithm.gptq module quark.torch.algorithm.gptq.gptq module quark.torch.algorithm.gptq.quantizer module quark.torch.algorithm.processor module quark.torch.algorithm.rotation module quark.torch.algorithm.rotation.hadamard module quark.torch.algorithm.rotation.rotation module quark.torch.algorithm.rotation.rotation_utils module quark.torch.algorithm.utils module quark.torch.algorithm.utils.module module quark.torch.algorithm.utils.prepare module quark.torch.algorithm.utils.utils module quark.torch.export module quark.torch.export.api module quark.torch.export.config module quark.torch.export.config.config module quark.torch.export.config.custom_config module quark.torch.export.gguf_export module quark.torch.export.gguf_export.api module quark.torch.export.gguf_export.gguf_model_converter module quark.torch.export.gguf_export.gguf_model_writer module quark.torch.export.gguf_export.tensor_convert module quark.torch.export.gguf_export.utils module quark.torch.export.json_export module quark.torch.export.json_export.builder module quark.torch.export.json_export.builder.llm_info module quark.torch.export.json_export.builder.llm_info_builder module quark.torch.export.json_export.builder.native_model_info_builder module quark.torch.export.json_export.converter module quark.torch.export.json_export.converter.llm_info_converter module quark.torch.export.json_export.utils module quark.torch.export.json_export.utils.utils module quark.torch.export.main_export module quark.torch.export.main_export.model_post_process module quark.torch.export.main_export.quant_config_parser module quark.torch.export.nn module quark.torch.export.nn.modules module quark.torch.export.nn.modules.export_operator module quark.torch.export.onnx module quark.torch.export.utils module quark.torch.kernel module quark.torch.kernel.hw_emulation module quark.torch.kernel.hw_emulation.extensions module quark.torch.kernel.hw_emulation.hw_emulation_interface module quark.torch.quantization module quark.torch.quantization.api module quark.torch.quantization.config module quark.torch.quantization.config.config module quark.torch.quantization.config.config_verification module quark.torch.quantization.config.type module quark.torch.quantization.graph module quark.torch.quantization.graph.fx module quark.torch.quantization.graph.fx.base module quark.torch.quantization.graph.optimization module quark.torch.quantization.graph.optimization.activate_dropout module quark.torch.quantization.graph.optimization.remove_dropout_node module quark.torch.quantization.graph.optimization.replace_conv2d_to_qtconv2d module quark.torch.quantization.graph.optimization.replace_conv_bn_to_qt_model module quark.torch.quantization.graph.optimization.replace_linear_to_qtlinear module quark.torch.quantization.graph.optimization.utils module quark.torch.quantization.graph.processor module quark.torch.quantization.graph.processor.insert_quantizer module quark.torch.quantization.graph.processor.processor module quark.torch.quantization.graph.processor.processor_utils module quark.torch.quantization.graph.torch_utils module quark.torch.quantization.model_transformation module quark.torch.quantization.nn module quark.torch.quantization.nn.modules module quark.torch.quantization.nn.modules.mixin module quark.torch.quantization.nn.modules.quantize_conv module quark.torch.quantization.nn.modules.quantize_conv_bn_fused module quark.torch.quantization.nn.modules.quantize_linear module quark.torch.quantization.nn.utils module quark.torch.quantization.observer module quark.torch.quantization.observer.observer module quark.torch.quantization.tensor_quantize module quark.torch.quantization.utils module quark.torch.utils module quark.torch.utils.log module quark.torch.utils.pack module

R

random_hadamard_matrix() (in module quark.torch.algorithm.rotation.hadamard) RandomDataReader (class in quark.onnx.quant_utils) RealQuantizeFunction (class in quark.torch.kernel) remove_initializers() (in module quark.onnx.quant_utils) remove_nodes() (in module quark.onnx.quant_utils) remove_qdq() (in module quark.onnx.tools.remove_qdq) RemoveDropoutNode (class in quark.torch.quantization.graph.optimization.remove_dropout_node) RemoveQDQTransform (class in quark.onnx.optimizations.convert_transforms) RemoveQDQTransformsPipeline (class in quark.onnx.optimizations.convert_transforms_pipeline) replace_conv2d_qtconv2d() (in module quark.torch.quantization.graph.optimization.replace_conv2d_to_qtconv2d) replace_conv2dbn_quantizedconv_module() (in module quark.torch.quantization.graph.optimization.replace_conv_bn_to_qt_model) replace_linear_qtlinear() (in module quark.torch.quantization.graph.optimization.replace_linear_to_qtlinear) replacement() (quark.onnx.graph_transformations.model_transformer_test.ModelTransformerTest.RemoveRelu method) (quark.onnx.graph_transformations.model_transformer_test.ModelTransformerTest.ReplaceWholeModel method) (quark.onnx.graph_transformations.transforms.Transform method) (quark.onnx.optimizations.convert_transforms.AddQDQToQOPTransform method) (quark.onnx.optimizations.convert_transforms.ConvQDQToQOPTransform method) (quark.onnx.optimizations.convert_transforms.MatMulQDQToQOPTransform method) (quark.onnx.optimizations.convert_transforms.MulQDQToQOPTransform method) (quark.onnx.optimizations.convert_transforms.RemoveQDQTransform method) (quark.onnx.optimizations.convert_transforms.SigmoidQDQToQOPTransform method)

reset_iter() (quark.onnx.quant_utils.CachedDataReader method) reset_min_max_vals() (quark.torch.quantization.observer.observer.UniformScalingObserver method) RMSNorm (class in quark.torch.algorithm.rotation.rotation_utils) rotate_in_channels() (in module quark.torch.algorithm.rotation.rotation_utils) rotate_out_channels() (in module quark.torch.algorithm.rotation.rotation_utils) RotationConfig (class in quark.torch.quantization.config.config) RotationProcessor (class in quark.torch.algorithm.rotation.rotation) round_impl() (quark.onnx.finetuning.create_torch.quant_base_ops.AdaroundQuantizer method) (quark.onnx.finetuning.create_torch.quant_base_ops.Quantizer method) RoundHalfToEven (class in quark.onnx.finetuning.create_torch.quant_base_ops) RoundType (class in quark.torch.quantization.config.type) run() (quark.onnx.finetuning.train_torch.train_model.ModelOptimizer class method) run_onnx_model() (in module quark.onnx.quant_utils)

S

save_model() (in module quark.onnx.utils.model_utils) save_torch_model() (in module quark.onnx.finetuning.torch_utils) scale2pos() (in module quark.onnx.quant_utils) ScaledActivation (class in quark.torch.algorithm.awq.modules.act) ScaleType (class in quark.torch.quantization.config.type) SentencePieceTokenTypes (class in quark.torch.export.gguf_export.gguf_model_writer) set_bias() (quark.onnx.finetuning.create_torch.create_model.TorchModel method) set_modules_original_bias() (in module quark.onnx.finetuning.create_torch.create_model_ops) set_modules_original_weight() (in module quark.onnx.finetuning.create_torch.create_model_ops) set_op_by_name() (in module quark.torch.quantization.utils)

set_weight() (quark.onnx.finetuning.create_torch.create_model.TorchModel method) setup_seed() (in module quark.onnx.finetuning.torch_utils) SigmoidQDQToQOPTransform (class in quark.onnx.optimizations.convert_transforms) simulate_transforms() (in module quark.onnx.simulate_dpu) SmoothQuant (class in quark.onnx.smooth_quant) SmoothQuantConfig (class in quark.torch.quantization.config.config) SmoothQuantProcessor (class in quark.torch.algorithm.awq.smooth) split_large_kernel_pool() (quark.onnx.optimize.Optimize method) split_model_info() (in module quark.torch.export.json_export.utils.utils) Subgraph (class in quark.onnx.finetuning.onnx_subgraph)

T

tensor_sync() (quark.onnx.finetuning.create_torch.quant_base_ops.Quantizer method) test_custom_op() (in module quark.onnx.operators.custom_ops.test_custom_ops_qdq) TorchModel (class in quark.onnx.finetuning.create_torch.create_model) train_torch_module_api() (in module quark.onnx.finetuning.torch_utils) TrainLoss (class in quark.onnx.finetuning.train_torch.train_model_loss)

TrainParameters (class in quark.onnx.finetuning.train_torch.train_model_param) Transform (class in quark.onnx.graph_transformations.transforms) (class in quark.torch.quantization.graph.fx.base) transform() (quark.onnx.graph_transformations.model_transformer.ModelTransformer method) transform_for_annotation() (in module quark.torch.quantization.graph.processor.processor) TransformsPipeline (class in quark.onnx.graph_transformations.transforms_pipeline)

U

UniformScalingObserver (class in quark.torch.quantization.observer.observer)

update_buffer() (quark.torch.quantization.tensor_quantize.FakeQuantizeBase method)

V

VitisBFPQuantizer (class in quark.onnx.qdq_quantizer) VitisExtendedQuantizer (class in quark.onnx.qdq_quantizer) VitisONNXQuantizer (class in quark.onnx.onnx_quantizer) VitisQDQCPUQuantizer (class in quark.onnx.cpu_quantizer)

VitisQDQNPUCNNQuantizer (class in quark.onnx.qdq_quantizer) VitisQDQQuantizer (class in quark.onnx.qdq_quantizer) VitisQuantFormat (class in quark.onnx.quant_utils) VitisQuantType (class in quark.onnx.quant_utils) Vocab (class in quark.torch.export.gguf_export.utils)