SGDClassifier - m-cl - log - {‘zipmap’: False}#

Fitted on a problem type m-cl (see find_suitable_problem), method predict_proba matches output . Model was converted with additional parameter: <class 'sklearn.linear_model._stochastic_gradient.SGDClassifier'>={'zipmap': False}.

SGDClassifier(loss='log', n_jobs=8, random_state=0)

index

0

skl_nop

1

skl_ncoef

3

skl_nlin

1

onx_size

1421

onx_nnodes

18

onx_ninits

8

onx_doc_string

onx_ir_version

8

onx_domain

ai.onnx

onx_model_version

0

onx_producer_name

skl2onnx

onx_producer_version

1.11.1

onx_

13

onx_ai.onnx.ml

1

onx_op_Cast

3

onx_op_Reshape

1

onx_size_optim

1421

onx_nnodes_optim

18

onx_ninits_optim

8

fit_coef_.shape

(3, 4)

fit_intercept_.shape

3

fit_classes_.shape

3
%0 X X float((0, 4)) MatMul MatMul (MatMul) X->MatMul label label int64((0,)) probabilities probabilities float((0, 3)) ArgMax ArgMax (ArgMax) axis=1 probabilities->ArgMax classes classes int32((3,)) [0 1 2] ArrayFeatureExtractor ArrayFeatureExtractor (ArrayFeatureExtractor) classes->ArrayFeatureExtractor coef coef float32((4, 3)) [[  3.41621    -3.2010055 -91.58158  ] [ 20.75054... coef->MatMul intercept intercept float32((3,)) [   5.0744696  125.14089   -190.15077  ] Add Add (Add) intercept->Add negate negate float32(()) -1.0 Mul Mul (Mul) negate->Mul unity unity float32(()) 1.0 Add1 Add (Add1) unity->Add1 axis axis int64((1,)) [1] ReduceSum ReduceSum (ReduceSum) axis->ReduceSum num_classes num_classes float32(()) 3.0 Mul1 Mul (Mul1) num_classes->Mul1 shape_tensor shape_tensor int64((1,)) [-1] Reshape Reshape (Reshape) shape_tensor->Reshape matmul_result matmul_result matmul_result->Add MatMul->matmul_result score score score->Mul Add->score negated_scores negated_scores Exp Exp (Exp) negated_scores->Exp Mul->negated_scores exp_result exp_result exp_result->Add1 Exp->exp_result add_result add_result Reciprocal Reciprocal (Reciprocal) add_result->Reciprocal Add1->add_result proba proba proba->ReduceSum Add2 Add (Add2) proba->Add2 Reciprocal->proba reduced_proba reduced_proba Cast Cast (Cast) to=9 reduced_proba->Cast Add3 Add (Add3) reduced_proba->Add3 ReduceSum->reduced_proba bool_reduced_proba bool_reduced_proba Not Not (Not) bool_reduced_proba->Not Cast->bool_reduced_proba bool_not_reduced_proba bool_not_reduced_proba Cast1 Cast (Cast1) to=1 bool_not_reduced_proba->Cast1 Not->bool_not_reduced_proba not_reduced_proba not_reduced_proba not_reduced_proba->Mul1 not_reduced_proba->Add2 Cast1->not_reduced_proba mask mask mask->Add3 Mul1->mask proba_updated proba_updated Div Div (Div) proba_updated->Div Add2->proba_updated reduced_proba_updated reduced_proba_updated reduced_proba_updated->Div Add3->reduced_proba_updated Div->probabilities predicted_label predicted_label predicted_label->ArrayFeatureExtractor ArgMax->predicted_label final_label final_label final_label->Reshape ArrayFeatureExtractor->final_label reshaped_final_label reshaped_final_label Cast2 Cast (Cast2) to=7 reshaped_final_label->Cast2 Reshape->reshaped_final_label Cast2->label

previous

SGDClassifier - b-cl - log - {‘zipmap’: False}

next

SGDClassifier - ~b-cl-64 - log - {‘zipmap’: False}