Index A | B | C | D | E | F | G | H | I | K | L | M | N | O | P | Q | R | S | T | W | Z A ActAddPattern (class in embedl_deploy.tensorrt.patterns) activation (embedl_deploy.quantize.QuantConfig attribute) AdaptiveAvgPoolPattern (class in embedl_deploy.tensorrt.patterns) alpha (embedl_deploy.quantize.SmoothQuantConfig attribute) apply_transformation_plan() (in module embedl_deploy) B Backend (class in embedl_deploy.backend) C calibrate_qdq() (in module embedl_deploy.quantize) calibrate_smooth_quant() (in module embedl_deploy.quantize) calibration_method (embedl_deploy.quantize.TensorQuantConfig attribute) CalibrationMethod (class in embedl_deploy.quantize) compute_parameters() (embedl_deploy.quantize.QuantStub method) configure() (in module embedl_deploy.quantize) consumers (embedl_deploy.quantize.WeightFakeQuantize attribute) ConvBNActPattern (class in embedl_deploy.tensorrt.patterns) ConvBNAddActPattern (class in embedl_deploy.tensorrt.patterns) ConvBNPattern (class in embedl_deploy.tensorrt.patterns) conversion_patterns (embedl_deploy.backend.Backend attribute) D DecomposeMultiheadAttentionPattern (class in embedl_deploy.tensorrt.patterns) disable_fake_quant() (in module embedl_deploy.quantize) discover_backends() (in module embedl_deploy.backend) E embedl_deploy module embedl_deploy.backend module embedl_deploy.lattice module embedl_deploy.lattice.modules module embedl_deploy.lattice.patterns module embedl_deploy.quantize module embedl_deploy.tensorrt module embedl_deploy.tensorrt.modules module embedl_deploy.tensorrt.patterns module embedl_deploy.version module embedl_deploy.version.public module enable_fake_quant() (in module embedl_deploy.quantize) F FlattenLinearToConv1x1Pattern (class in embedl_deploy.tensorrt.patterns) FlattenToReshapePattern (class in embedl_deploy.lattice.patterns) forward() (embedl_deploy.quantize.QuantStub method) (embedl_deploy.quantize.WeightFakeQuantize method) (embedl_deploy.tensorrt.modules.FusedAdaptiveAvgPool2d method) (embedl_deploy.tensorrt.modules.FusedConvBN method) (embedl_deploy.tensorrt.modules.FusedConvBNAct method) (embedl_deploy.tensorrt.modules.FusedConvBNActMaxPool method) (embedl_deploy.tensorrt.modules.FusedConvBNAddAct method) (embedl_deploy.tensorrt.modules.FusedLayerNorm method) (embedl_deploy.tensorrt.modules.FusedLinear method) (embedl_deploy.tensorrt.modules.FusedLinearAct method) (embedl_deploy.tensorrt.modules.FusedMHAInProjection method) (embedl_deploy.tensorrt.modules.FusedScaledDotProductAttention method) freeze() (embedl_deploy.quantize.WeightFakeQuantize method) freeze_bn_stats() (in module embedl_deploy.quantize) freeze_weight_quantization() (in module embedl_deploy.quantize) FusedAdaptiveAvgPool2d (class in embedl_deploy.tensorrt.modules) FusedConvBN (class in embedl_deploy.tensorrt.modules) FusedConvBNAct (class in embedl_deploy.tensorrt.modules) FusedConvBNActMaxPool (class in embedl_deploy.tensorrt.modules) FusedConvBNAddAct (class in embedl_deploy.tensorrt.modules) FusedLayerNorm (class in embedl_deploy.tensorrt.modules) FusedLinear (class in embedl_deploy.tensorrt.modules) FusedLinearAct (class in embedl_deploy.tensorrt.modules) FusedMHAInProjection (class in embedl_deploy.tensorrt.modules) FusedScaledDotProductAttention (class in embedl_deploy.tensorrt.modules) fusion_patterns (embedl_deploy.backend.Backend attribute) G get_backend() (in module embedl_deploy.backend) get_transformation_plan() (in module embedl_deploy) graft (embedl_deploy.lattice.patterns.FlattenToReshapePattern attribute) (embedl_deploy.lattice.patterns.LatticeConv2dPattern attribute) (embedl_deploy.lattice.patterns.LatticeGlobalAvgPoolPattern attribute) (embedl_deploy.lattice.patterns.LatticeLeakyReLUPattern attribute) (embedl_deploy.lattice.patterns.LatticeMaxPool2dPattern attribute) (embedl_deploy.tensorrt.patterns.ActAddPattern attribute) (embedl_deploy.tensorrt.patterns.AdaptiveAvgPoolPattern attribute) (embedl_deploy.tensorrt.patterns.ConvBNActPattern attribute) (embedl_deploy.tensorrt.patterns.ConvBNAddActPattern attribute) (embedl_deploy.tensorrt.patterns.ConvBNPattern attribute) (embedl_deploy.tensorrt.patterns.DecomposeMultiheadAttentionPattern attribute) (embedl_deploy.tensorrt.patterns.FlattenLinearToConv1x1Pattern attribute) (embedl_deploy.tensorrt.patterns.LayerNormPattern attribute) (embedl_deploy.tensorrt.patterns.LinearActPattern attribute) (embedl_deploy.tensorrt.patterns.LinearPattern attribute) (embedl_deploy.tensorrt.patterns.MHAInProjectionPattern attribute) (embedl_deploy.tensorrt.patterns.RemoveAssertPattern attribute) (embedl_deploy.tensorrt.patterns.RemoveIdentityAdaptiveAvgPoolPattern attribute) (embedl_deploy.tensorrt.patterns.RemoveIdentityPattern attribute) (embedl_deploy.tensorrt.patterns.ScaledDotProductAttentionPattern attribute) (embedl_deploy.tensorrt.patterns.StemConvBNActMaxPoolPattern attribute) H HISTOGRAM (embedl_deploy.quantize.CalibrationMethod attribute) I inputs_to_quantize (embedl_deploy.tensorrt.modules.FusedAdaptiveAvgPool2d attribute) (embedl_deploy.tensorrt.modules.FusedConvBN attribute) (embedl_deploy.tensorrt.modules.FusedConvBNAct attribute) (embedl_deploy.tensorrt.modules.FusedConvBNActMaxPool attribute) (embedl_deploy.tensorrt.modules.FusedConvBNAddAct attribute) (embedl_deploy.tensorrt.modules.FusedLayerNorm attribute) (embedl_deploy.tensorrt.modules.FusedLinear attribute) (embedl_deploy.tensorrt.modules.FusedLinearAct attribute) (embedl_deploy.tensorrt.modules.FusedMHAInProjection attribute) (embedl_deploy.tensorrt.modules.FusedScaledDotProductAttention attribute) is_compatible() (embedl_deploy.lattice.modules.LatticeAdaptiveAvgPool2d class method) (embedl_deploy.lattice.modules.LatticeConv2d class method) (embedl_deploy.lattice.modules.LatticeLeakyReLU class method) (embedl_deploy.lattice.modules.LatticeMaxPool2d class method) K KERNEL_SIZE (embedl_deploy.lattice.modules.LatticeMaxPool2d attribute) KERNEL_SIZES (embedl_deploy.lattice.modules.LatticeConv2d attribute) L LatticeAdaptiveAvgPool2d (class in embedl_deploy.lattice.modules) LatticeConv2d (class in embedl_deploy.lattice.modules) LatticeConv2dPattern (class in embedl_deploy.lattice.patterns) LatticeGlobalAvgPoolPattern (class in embedl_deploy.lattice.patterns) LatticeLeakyReLU (class in embedl_deploy.lattice.modules) LatticeLeakyReLUPattern (class in embedl_deploy.lattice.patterns) LatticeMaxPool2d (class in embedl_deploy.lattice.modules) LatticeMaxPool2dPattern (class in embedl_deploy.lattice.patterns) LayerNormPattern (class in embedl_deploy.tensorrt.patterns) LinearActPattern (class in embedl_deploy.tensorrt.patterns) LinearPattern (class in embedl_deploy.tensorrt.patterns) M matches (embedl_deploy.TransformationPlan attribute) (embedl_deploy.TransformationResult attribute) MHAInProjectionPattern (class in embedl_deploy.tensorrt.patterns) MINMAX (embedl_deploy.quantize.CalibrationMethod attribute) model (embedl_deploy.TransformationPlan attribute) (embedl_deploy.TransformationResult attribute) module embedl_deploy embedl_deploy.backend embedl_deploy.lattice embedl_deploy.lattice.modules embedl_deploy.lattice.patterns embedl_deploy.quantize embedl_deploy.tensorrt embedl_deploy.tensorrt.modules embedl_deploy.tensorrt.patterns embedl_deploy.version embedl_deploy.version.public ModulesToSkip (class in embedl_deploy.quantize) MOVING_AVERAGE_MINMAX (embedl_deploy.quantize.CalibrationMethod attribute) N n_bits (embedl_deploy.quantize.TensorQuantConfig attribute) name (embedl_deploy.backend.Backend attribute) NEGATIVE_SLOPE (embedl_deploy.lattice.modules.LatticeLeakyReLU attribute) O OUTPUT_SIZE (embedl_deploy.lattice.modules.LatticeAdaptiveAvgPool2d attribute) P PADDING (embedl_deploy.lattice.modules.LatticeMaxPool2d attribute) per_channel (embedl_deploy.quantize.TensorQuantConfig attribute) phase (embedl_deploy.lattice.patterns.FlattenToReshapePattern attribute) (embedl_deploy.lattice.patterns.LatticeConv2dPattern attribute) (embedl_deploy.lattice.patterns.LatticeGlobalAvgPoolPattern attribute) (embedl_deploy.lattice.patterns.LatticeLeakyReLUPattern attribute) (embedl_deploy.lattice.patterns.LatticeMaxPool2dPattern attribute) (embedl_deploy.tensorrt.patterns.ActAddPattern attribute) (embedl_deploy.tensorrt.patterns.AdaptiveAvgPoolPattern attribute) (embedl_deploy.tensorrt.patterns.ConvBNActPattern attribute) (embedl_deploy.tensorrt.patterns.ConvBNAddActPattern attribute) (embedl_deploy.tensorrt.patterns.ConvBNPattern attribute) (embedl_deploy.tensorrt.patterns.DecomposeMultiheadAttentionPattern attribute) (embedl_deploy.tensorrt.patterns.FlattenLinearToConv1x1Pattern attribute) (embedl_deploy.tensorrt.patterns.LayerNormPattern attribute) (embedl_deploy.tensorrt.patterns.LinearActPattern attribute) (embedl_deploy.tensorrt.patterns.LinearPattern attribute) (embedl_deploy.tensorrt.patterns.MHAInProjectionPattern attribute) (embedl_deploy.tensorrt.patterns.RemoveAssertPattern attribute) (embedl_deploy.tensorrt.patterns.RemoveIdentityAdaptiveAvgPoolPattern attribute) (embedl_deploy.tensorrt.patterns.RemoveIdentityPattern attribute) (embedl_deploy.tensorrt.patterns.ScaledDotProductAttentionPattern attribute) (embedl_deploy.tensorrt.patterns.StemConvBNActMaxPoolPattern attribute) prefers_fp_input (embedl_deploy.tensorrt.modules.FusedLayerNorm attribute) prepare_qat() (in module embedl_deploy.quantize) Q quant_max (embedl_deploy.quantize.TensorQuantConfig property) quant_min (embedl_deploy.quantize.TensorQuantConfig property) QuantConfig (class in embedl_deploy.quantize) quantize() (in module embedl_deploy.quantize) quantized_patterns (embedl_deploy.backend.Backend attribute) QuantStub (class in embedl_deploy.quantize) R RemoveAssertPattern (class in embedl_deploy.tensorrt.patterns) RemoveIdentityAdaptiveAvgPoolPattern (class in embedl_deploy.tensorrt.patterns) RemoveIdentityPattern (class in embedl_deploy.tensorrt.patterns) replace() (embedl_deploy.tensorrt.patterns.DecomposeMultiheadAttentionPattern method) report (embedl_deploy.TransformationResult attribute) reset_backends() (in module embedl_deploy.backend) S scale (embedl_deploy.quantize.QuantStub attribute) (embedl_deploy.quantize.WeightFakeQuantize attribute) ScaledDotProductAttentionPattern (class in embedl_deploy.tensorrt.patterns) set_backend() (in module embedl_deploy.backend) skip (embedl_deploy.quantize.QuantConfig attribute) smooth (embedl_deploy.quantize.ModulesToSkip attribute) smooth_patterns (embedl_deploy.backend.Backend attribute) smooth_quant (embedl_deploy.quantize.QuantConfig attribute) SmoothQuantConfig (class in embedl_deploy.quantize) snapped_params() (embedl_deploy.lattice.modules.LatticeConv2d class method) StemConvBNActMaxPoolPattern (class in embedl_deploy.tensorrt.patterns) STRIDE (embedl_deploy.lattice.modules.LatticeMaxPool2d attribute) STRIDES (embedl_deploy.lattice.modules.LatticeConv2d attribute) stub (embedl_deploy.quantize.ModulesToSkip attribute) symmetric (embedl_deploy.quantize.TensorQuantConfig attribute) T TensorQuantConfig (class in embedl_deploy.quantize) transform() (in module embedl_deploy) TransformationPlan (class in embedl_deploy) TransformationResult (class in embedl_deploy) tree (embedl_deploy.lattice.patterns.FlattenToReshapePattern attribute) (embedl_deploy.lattice.patterns.LatticeConv2dPattern attribute) (embedl_deploy.lattice.patterns.LatticeGlobalAvgPoolPattern attribute) (embedl_deploy.lattice.patterns.LatticeLeakyReLUPattern attribute) (embedl_deploy.lattice.patterns.LatticeMaxPool2dPattern attribute) (embedl_deploy.tensorrt.patterns.ActAddPattern attribute) (embedl_deploy.tensorrt.patterns.AdaptiveAvgPoolPattern attribute) (embedl_deploy.tensorrt.patterns.ConvBNActPattern attribute) (embedl_deploy.tensorrt.patterns.ConvBNAddActPattern attribute) (embedl_deploy.tensorrt.patterns.ConvBNPattern attribute) (embedl_deploy.tensorrt.patterns.DecomposeMultiheadAttentionPattern attribute) (embedl_deploy.tensorrt.patterns.FlattenLinearToConv1x1Pattern attribute) (embedl_deploy.tensorrt.patterns.LayerNormPattern attribute) (embedl_deploy.tensorrt.patterns.LinearActPattern attribute) (embedl_deploy.tensorrt.patterns.LinearPattern attribute) (embedl_deploy.tensorrt.patterns.MHAInProjectionPattern attribute) (embedl_deploy.tensorrt.patterns.RemoveAssertPattern attribute) (embedl_deploy.tensorrt.patterns.RemoveIdentityAdaptiveAvgPoolPattern attribute) (embedl_deploy.tensorrt.patterns.RemoveIdentityPattern attribute) (embedl_deploy.tensorrt.patterns.ScaledDotProductAttentionPattern attribute) (embedl_deploy.tensorrt.patterns.StemConvBNActMaxPoolPattern attribute) W weight (embedl_deploy.quantize.ModulesToSkip attribute) (embedl_deploy.quantize.QuantConfig attribute) WeightFakeQuantize (class in embedl_deploy.quantize) Z zero_point (embedl_deploy.quantize.QuantStub attribute) (embedl_deploy.quantize.WeightFakeQuantize attribute)
