onnxsim/onnxsim

Pattern Findings

The scanner identified 14244 distinct pattern matches across 20 syntactic categories. Each entry below represents a discrete location in the source code where the engine recorded a detector match to inspect in context. Expand any category row to inspect the individual file paths, line numbers, code snippets, and the lexical context (CODE, COMMENT, or STRING) in which each match was detected.

Reading the findings table: The Severity column indicates the configured finding weight (CRITICAL / HIGH / MEDIUM / LOW). The Context column identifies whether the match occurred inside executable code, an inline comment, or a string literal — legacy comment-context matches receive a ×1.5 score weight. The ⚡ bolt icon marks clustered matches: three or more patterns within a 10-line window, each receiving an additional ×1.5 density multiplier under the legacy scoring model.

Cross-File Repetition1191 hits · 5955 pts

Severity	File	Line	Snippet	Context
HIGH	README.md	0	<domain: "com.onnxsim.test", opset_import: ["" : 18]> matmul_add_pattern (x, w, b) => (y) { t = matmul(x, w) y = add(t,	STRING
HIGH	tests/test_function_rewriter_vs_onnxscript.py	0	<domain: "com.onnxsim.test", opset_import: ["" : 18]> matmul_add_pattern (x, w, b) => (y) { t = matmul(x, w) y = add(t,	STRING
HIGH	tests/test_function_rewriter.py	0	<domain: "com.onnxsim.test", opset_import: ["" : 18]> matmul_add_pattern (x, w, b) => (y) { t = matmul(x, w) y = add(t,	STRING
HIGH	README.md	0	<domain: "com.onnxsim.test", opset_import: ["" : 18]> gemm_replacement (x, w, b) => (y) { y = gemm(x, w, b) }	STRING
HIGH	tests/test_function_rewriter_vs_onnxscript.py	0	<domain: "com.onnxsim.test", opset_import: ["" : 18]> gemm_replacement (x, w, b) => (y) { y = gemm(x, w, b) }	STRING
HIGH	tests/test_function_rewriter.py	0	<domain: "com.onnxsim.test", opset_import: ["" : 18]> gemm_replacement (x, w, b) => (y) { y = gemm(x, w, b) }	STRING
HIGH	tests/test_axera_div_quant_fields_two_live_inputs.py	0	`computeasymmetricuint8quantparams` (`onnxsim/passes/static_ quantize_matmul.h`), the plain float32-throughout formula t	STRING
HIGH	tests/test_axera_add_trivial_x2_zeropoint.py	0	`computeasymmetricuint8quantparams` (`onnxsim/passes/static_ quantize_matmul.h`), the plain float32-throughout formula t	STRING
HIGH	…s/test_axera_two_live_input_operand_order_synthesis.py	0	`computeasymmetricuint8quantparams` (`onnxsim/passes/static_ quantize_matmul.h`), the plain float32-throughout formula t	STRING
HIGH	tests/test_axera_two_live_input_cluster_synthesis.py	0	`computeasymmetricuint8quantparams` (`onnxsim/passes/static_ quantize_matmul.h`), the plain float32-throughout formula t	STRING
HIGH	tests/test_axera_div_shape64_x2_locator.py	0	`computeasymmetricuint8quantparams` (`onnxsim/passes/static_ quantize_matmul.h`), the plain float32-throughout formula t	STRING
HIGH	tests/test_axera_sub_operand_order.py	0	`computeasymmetricuint8quantparams` (`onnxsim/passes/static_ quantize_matmul.h`), the plain float32-throughout formula t	STRING
HIGH	tests/test_axera_mul_shape64_x2_locator_reexamined.py	0	`computeasymmetricuint8quantparams` (`onnxsim/passes/static_ quantize_matmul.h`), the plain float32-throughout formula t	STRING
HIGH	tests/test_axera_sub_div_reflow_extended_seeds.py	0	`computeasymmetricuint8quantparams` (`onnxsim/passes/static_ quantize_matmul.h`), the plain float32-throughout formula t	STRING
HIGH	tests/test_axera_add_shape64_x2_locator_reexamined.py	0	`computeasymmetricuint8quantparams` (`onnxsim/passes/static_ quantize_matmul.h`), the plain float32-throughout formula t	STRING
HIGH	tests/test_axera_sub_quant_fields_two_live_inputs.py	0	`computeasymmetricuint8quantparams` (`onnxsim/passes/static_ quantize_matmul.h`), the plain float32-throughout formula t	STRING
HIGH	tests/test_axera_mul_quant_fields_two_live_inputs.py	0	`computeasymmetricuint8quantparams` (`onnxsim/passes/static_ quantize_matmul.h`), the plain float32-throughout formula t	STRING
HIGH	…sts/test_axera_shape64_seed12_absence_investigation.py	0	`computeasymmetricuint8quantparams` (`onnxsim/passes/static_ quantize_matmul.h`), the plain float32-throughout formula t	STRING
HIGH	tests/test_axera_div_operand_order.py	0	`computeasymmetricuint8quantparams` (`onnxsim/passes/static_ quantize_matmul.h`), the plain float32-throughout formula t	STRING
HIGH	tests/test_axera_add_two_live_shape_generality.py	0	`computeasymmetricuint8quantparams` (`onnxsim/passes/static_ quantize_matmul.h`), the plain float32-throughout formula t	STRING
HIGH	tests/test_axera_add_operand_order.py	0	`computeasymmetricuint8quantparams` (`onnxsim/passes/static_ quantize_matmul.h`), the plain float32-throughout formula t	STRING
HIGH	tests/test_axera_sub_mul_trivial_x2_zeropoint.py	0	`computeasymmetricuint8quantparams` (`onnxsim/passes/static_ quantize_matmul.h`), the plain float32-throughout formula t	STRING
HIGH	tests/test_axera_conv_matmul_trivial_zeropoint.py	0	`computeasymmetricuint8quantparams` (`onnxsim/passes/static_ quantize_matmul.h`), the plain float32-throughout formula t	STRING
HIGH	tests/test_axera_two_live_input_four_op_synthesis.py	0	`computeasymmetricuint8quantparams` (`onnxsim/passes/static_ quantize_matmul.h`), the plain float32-throughout formula t	STRING
HIGH	tests/test_axera_sub_shape64_x2_locator.py	0	`computeasymmetricuint8quantparams` (`onnxsim/passes/static_ quantize_matmul.h`), the plain float32-throughout formula t	STRING
HIGH	tests/test_axera_mul_operand_order.py	0	`computeasymmetricuint8quantparams` (`onnxsim/passes/static_ quantize_matmul.h`), the plain float32-throughout formula t	STRING
HIGH	…/test_axera_shape64_x2_locator_correction_synthesis.py	0	`computeasymmetricuint8quantparams` (`onnxsim/passes/static_ quantize_matmul.h`), the plain float32-throughout formula t	STRING
HIGH	tests/test_axera_add_output_operand_order.py	0	`computeasymmetricuint8quantparams` (`onnxsim/passes/static_ quantize_matmul.h`), the plain float32-throughout formula t	STRING
HIGH	tests/test_axera_gemm_trivial_zeropoint.py	0	`computeasymmetricuint8quantparams` (`onnxsim/passes/static_ quantize_matmul.h`), the plain float32-throughout formula t	STRING
HIGH	tests/test_axera_add_quant_fields_two_live_inputs.py	0	`computeasymmetricuint8quantparams` (`onnxsim/passes/static_ quantize_matmul.h`), the plain float32-throughout formula t	STRING
HIGH	tests/test_axera_x1_trivial_zeropoint.py	0	`computeasymmetricuint8quantparams` (`onnxsim/passes/static_ quantize_matmul.h`), the plain float32-throughout formula t	STRING
HIGH	tests/test_weight_only_quantize_int4.py	0	< ir_version: {ir_version}, opset_import: ["": {opset}] > {body}	STRING
HIGH	tests/test_halide_integration.py	0	< ir_version: {ir_version}, opset_import: ["": {opset}] > {body}	STRING
HIGH	tests/test_ibert_softmax.py	0	< ir_version: {ir_version}, opset_import: ["": {opset}] > {body}	STRING
HIGH	tests/test_quant_error_eval.py	0	< ir_version: {ir_version}, opset_import: ["": {opset}] > {body}	STRING
HIGH	tests/test_deepseek_fp8_cpp.py	0	< ir_version: {ir_version}, opset_import: ["": {opset}] > {body}	STRING
HIGH	tests/test_modelopt_integration.py	0	< ir_version: {ir_version}, opset_import: ["": {opset}] > {body}	STRING
HIGH	tests/test_deepseek_fp8.py	0	< ir_version: {ir_version}, opset_import: ["": {opset}] > {body}	STRING
HIGH	tests/test_formal_verify_gguf_ternary_quant.py	0	< ir_version: {ir_version}, opset_import: ["": {opset}] > {body}	STRING
HIGH	tests/test_qoperator_quantize_matmul.py	0	< ir_version: {ir_version}, opset_import: ["": {opset}] > {body}	STRING
HIGH	tests/test_d2quant.py	0	< ir_version: {ir_version}, opset_import: ["": {opset}] > {body}	STRING
HIGH	tests/test_awq_cpp.py	0	< ir_version: {ir_version}, opset_import: ["": {opset}] > {body}	STRING
HIGH	tests/test_outlier_suppression.py	0	< ir_version: {ir_version}, opset_import: ["": {opset}] > {body}	STRING
HIGH	tests/test_partial_shape_eval_rank.py	0	< ir_version: {ir_version}, opset_import: ["": {opset}] > {body}	STRING
HIGH	tests/test_intactkv.py	0	< ir_version: {ir_version}, opset_import: ["": {opset}] > {body}	STRING
HIGH	tests/test_formal_verify_eliminate_shape_op.py	0	< ir_version: {ir_version}, opset_import: ["": {opset}] > {body}	STRING
HIGH	tests/test_kv_cache_quantization.py	0	< ir_version: {ir_version}, opset_import: ["": {opset}] > {body}	STRING
HIGH	tests/test_gptaq_cpp.py	0	< ir_version: {ir_version}, opset_import: ["": {opset}] > {body}	STRING
HIGH	tests/test_bias_correction.py	0	< ir_version: {ir_version}, opset_import: ["": {opset}] > {body}	STRING
HIGH	tests/test_quantize_bf16.py	0	< ir_version: {ir_version}, opset_import: ["": {opset}] > {body}	STRING
HIGH	tests/test_formal_verify_gguf_q4_1.py	0	< ir_version: {ir_version}, opset_import: ["": {opset}] > {body}	STRING
HIGH	tests/test_model_checking.py	0	< ir_version: {ir_version}, opset_import: ["": {opset}] > {body}	STRING
HIGH	tests/test_hqq_cpp.py	0	< ir_version: {ir_version}, opset_import: ["": {opset}] > {body}	STRING
HIGH	tests/test_spqr_cpp.py	0	< ir_version: {ir_version}, opset_import: ["": {opset}] > {body}	STRING
HIGH	tests/test_formal_verify_quarot.py	0	< ir_version: {ir_version}, opset_import: ["": {opset}] > {body}	STRING
HIGH	tests/test_mse_calibration.py	0	< ir_version: {ir_version}, opset_import: ["": {opset}] > {body}	STRING
HIGH	tests/test_weight_only_quantize_int16.py	0	< ir_version: {ir_version}, opset_import: ["": {opset}] > {body}	STRING
HIGH	tests/test_mixed_precision_cpp.py	0	< ir_version: {ir_version}, opset_import: ["": {opset}] > {body}	STRING
HIGH	tests/test_formal_verify_static_quantize_int16_conv.py	0	< ir_version: {ir_version}, opset_import: ["": {opset}] > {body}	STRING
HIGH	tests/test_formal_verify_static_quantize_matmul.py	0	< ir_version: {ir_version}, opset_import: ["": {opset}] > {body}	STRING
1131 more matches not shown…

Hyper-Verbose Identifiers7469 hits · 5930 pts

Severity	File	Line	Snippet	Context
LOW	bench/nncf_comparison.py	131	def _referenced_initializer_bytes(model: onnx.ModelProto) -> int:	CODE
LOW	…nnx-cardputer-flash/tests/test_onnx_to_tflite_micro.py	27	def test_round_trips_arbitrary_bytes():	CODE
LOW	…nnx-cardputer-flash/tests/test_onnx_to_tflite_micro.py	39	def test_declares_length_and_alignment():	CODE
LOW	…nnx-cardputer-flash/tests/test_onnx_to_tflite_micro.py	46	def test_rejects_non_identifier_var_name():	CODE
LOW	…nnx-cardputer-flash/tests/test_onnx_to_tflite_micro.py	57	def test_main_writes_plain_tflite_for_dot_tflite_output(tmp_path):	CODE
LOW	…nnx-cardputer-flash/tests/test_onnx_to_tflite_micro.py	68	def test_main_writes_c_header_for_dot_h_output(tmp_path):	CODE
LOW	…rdputer-flash/tests/test_onnx_to_tflite_flatbuffers.py	83	def test_softmax_sandwich_emits_a_single_softmax_op(tmp_path):	CODE
LOW	…rdputer-flash/tests/test_onnx_to_tflite_flatbuffers.py	151	def test_real_model_matches_onnx_reference(	CODE
LOW⚡	tools/onnx-k210-flash/web/ncc/ncc.js	1	var create_ncc=(()=>{var _scriptName=globalThis.document?.currentScript?.src;return async function(moduleArg={}){var mod	CODE
LOW	tools/onnx-k210-flash/tests/test_onnx_legalizer.py	34	def _assert_legalizes_to_equivalent_graph(model: onnx.ModelProto, op_type: str, feeds: dict[str, np.ndarray], *, atol=1e	CODE
LOW	tools/onnx-k210-flash/tests/test_onnx_legalizer.py	46	def test_supported_ops_matches_registered_legalizers():	CODE
LOW	tools/onnx-k210-flash/tests/test_onnx_legalizer.py	51	def test_leaves_already_supported_ops_alone():	CODE
LOW	tools/onnx-k210-flash/tests/test_onnx_legalizer.py	69	def test_gelu_tanh_approximate():	CODE
LOW	tools/onnx-k210-flash/tests/test_onnx_legalizer.py	113	def test_isinf_both_directions():	CODE
LOW	tools/onnx-k210-flash/tests/test_onnx_legalizer.py	125	def test_isinf_neither_direction():	CODE
LOW	tools/onnx-k210-flash/tests/test_onnx_legalizer.py	131	def test_mean_variance_normalization_default_axes():	CODE
LOW	tools/onnx-k210-flash/tests/test_onnx_legalizer.py	142	def test_legalize_is_a_noop_the_second_time():	CODE
LOW	tools/onnx-k210-flash/tests/test_onnx_legalizer.py	152	def test_legalizes_a_node_nested_among_supported_ops():	CODE
LOW	tools/onnx-k210-flash/tests/test_onnx_to_kmodel.py	28	def _make_model_with_batch_dim(batch_dim_param: str \| None, batch_dim_value: int) -> onnx.ModelProto:	CODE
LOW⚡	tools/onnx-k210-flash/tests/test_onnx_to_kmodel.py	50	def test_pin_batch_dim_replaces_symbolic_dim_param():	CODE
LOW⚡	tools/onnx-k210-flash/tests/test_onnx_to_kmodel.py	58	def test_pin_batch_dim_replaces_zero_dim_value():	CODE
LOW⚡	tools/onnx-k210-flash/tests/test_onnx_to_kmodel.py	68	def test_pin_batch_dim_leaves_a_fixed_positive_batch_alone():	CODE
LOW⚡	tools/onnx-k210-flash/tests/test_onnx_to_kmodel.py	74	def test_convert_to_kmodel_real_hf_model(tmp_path):	CODE
LOW	tools/onnx-k210-flash/scripts/onnx_legalizer.py	195	def _legalize_meanvariancenormalization(node: onnx.NodeProto) -> list[onnx.NodeProto]:	CODE
LOW	tools/onnx-deploy/scripts/make_toy_seq2seq.py	142	def make_decoder_with_past_model(enc_seq_len, vocab_size):	CODE
LOW	tools/onnx-deploy/scripts/make_toy_int8_kv_decoder.py	55	def _cache_sum_logits_subgraph(prefix, cache_name, vocab_size):	CODE
LOW	tools/onnx-deploy/scripts/make_toy_int8_kv_decoder.py	113	def make_decoder_with_past_model(vocab_size):	CODE
LOW	…ls/onnx-finetune/tests/test_distillation_graph_grad.py	228	def test_gradient_matches_finite_differences(toy_models, target):	CODE
LOW	…ls/onnx-finetune/tests/test_distillation_graph_grad.py	246	def test_gradient_matches_tinygrad_autodiff(toy_models, target):	CODE
LOW	…ls/onnx-finetune/tests/test_distillation_graph_grad.py	286	def test_gradient_matches_finite_differences_across_batch_sizes(toy_models):	CODE
LOW	…ls/onnx-finetune/tests/test_distillation_graph_grad.py	308	def test_step_graph_has_no_training_only_ops(toy_models):	CODE
LOW	…ls/onnx-finetune/tests/test_distillation_graph_grad.py	318	def test_step_graph_trains_on_plain_onnxruntime(toy_models):	CODE
LOW	…ls/onnx-finetune/tests/test_distillation_graph_grad.py	381	def test_frozen_prefixes_stay_constants_and_train_the_rest(toy_models):	CODE
LOW	tools/onnx-finetune/tests/test_lora.py	181	def test_matmul_injection_noop_at_init_and_round_trips(tmp_path):	CODE
LOW	tools/onnx-finetune/tests/test_lora.py	205	def test_gemm_transb_injection_noop_at_init(tmp_path):	CODE
LOW	tools/onnx-finetune/tests/test_lora.py	225	def test_gemm_transa_injection_noop_at_init(tmp_path):	CODE
LOW	tools/onnx-finetune/tests/test_lora.py	265	def test_target_contains_filters_by_name(tmp_path):	CODE
LOW	tools/onnx-finetune/tests/test_lora.py	286	def test_prepare_qlora_excludes_adapter_from_quantization(tmp_path):	CODE
LOW	tools/onnx-finetune/tests/test_lora.py	320	def test_export_onnx_adapter_matches_merged_model_via_native_lora_adapter(tmp_path):	CODE
LOW	tools/onnx-finetune/scripts/generate_nas_step_graphs.py	199	def write_manifest_and_initial_state(	CODE
LOW	…finetune/scripts/generate_darts_supernet_step_graph.py	106	def build_darts_supernet_step(	CODE
LOW	…finetune/scripts/generate_darts_supernet_step_graph.py	232	def write_manifest_and_initial_state(	CODE
LOW	…x-finetune/scripts/generate_distillation_step_graph.py	260	def _build_forward_loss_and_grads(	CODE
LOW	…x-finetune/scripts/generate_distillation_step_graph.py	449	def build_distillation_step_graph(	CODE
LOW	…x-finetune/scripts/generate_distillation_step_graph.py	528	def write_manifest_and_initial_state(	CODE
LOW	…finetune/scripts/generate_federated_lora_step_graph.py	105	def build_federated_lora_step_graph(	CODE
LOW	…finetune/scripts/generate_federated_lora_step_graph.py	180	def write_manifest_and_initial_state(	CODE
LOW	tools/onnx-remote/python/onnx_remote_client.py	303	def encode_request_payload_parts(	CODE
LOW⚡	tests/test_axera_mcode_segments_fix.py	56	def test_padded_and_unpadded_blobs_have_the_right_header(name, pad, header):	CODE
LOW⚡	tests/test_axera_mcode_segments_fix.py	66	def test_both_padding_groups_are_represented():	CODE
LOW⚡	tests/test_axera_mcode_segments_fix.py	72	def test_every_segment_starts_on_its_stream_header(name):	CODE
LOW⚡	tests/test_axera_mcode_segments_fix.py	81	def test_non_zero_padding_is_rejected():	CODE
LOW	tests/test_axera_div_quant_fields_two_live_inputs.py	305	def test_all_three_trivial_fixtures_decode_cleanly(self):	CODE
LOW	tests/test_axera_div_quant_fields_two_live_inputs.py	333	def test_no_mul_style_requant_multiplier_field_group_exists(self):	CODE
LOW	tests/test_axera_div_quant_fields_two_live_inputs.py	349	def test_plain_y_scale_is_bit_exact_at_a_raw_byte_offset_four_times(self):	CODE
LOW	tests/test_axera_div_quant_fields_two_live_inputs.py	364	def test_zp_y_matches_after_excluding_the_unrelated_constant_hit(self):	CODE
LOW	tests/test_axera_div_quant_fields_two_live_inputs.py	406	def test_no_decode_errors_on_any_seed_pair(self):	CODE
LOW	tests/test_weight_only_quantize_int4.py	107	def test_quantize_gemm_transb_with_bias():	STRING
LOW	tests/test_weight_only_quantize_int4.py	134	def test_quantize_scale_shape_matches_block_count():	STRING
LOW	tests/test_weight_only_quantize_int4.py	160	def test_quantize_skips_k_not_divisible_by_block_size():	STRING
7409 more matches not shown…

Over-Commented Block1981 hits · 1910 pts

Severity	File	Line	Snippet	Context
LOW	build_wasm_pyodide.sh	1	#!/usr/bin/env bash	COMMENT
LOW	build_wasm_pyodide.sh	21	# KNOWN GOOD for onnxsim's vendored protobuf (clang-18/Emscripten	COMMENT
LOW	build_wasm_pyodide.sh	41	# .../xbuildenv/pyodide-root/cpython/installs/python-3.13.2/include/python3.13	COMMENT
LOW	build_wasm_pyodide.sh	61	# can fail on a minor-version mismatch against the target headers.	COMMENT
LOW	build_wasm_pyodide.sh	121	# a DIFFERENT hint-variable namespace, so all three need to be seeded with	COMMENT
LOW	CMakeLists.txt	1	cmake_minimum_required(VERSION 3.22)	COMMENT
LOW	CMakeLists.txt	61	# XNNPACK instead of ONNX Runtime, for a small, explicitly-supported subset of	COMMENT
LOW	CMakeLists.txt	81	# code #ifdef'd out) and no ORT source is fetched or compiled.	COMMENT
LOW	CMakeLists.txt	201		COMMENT
LOW	CMakeLists.txt	281	# remote_executor.h exposes the transport profiling level to native clients.	COMMENT
LOW	CMakeLists.txt	461	# With the onnx stack built static, the linker prunes archive members	COMMENT
LOW	CMakeLists.txt	521	# linker response file) via get_filename_component(... CMAKE_CURRENT_LIST_FILE	COMMENT
LOW	CMakeLists.txt	541	get_filename_component(ONNXSIM_NANOBIND_PACKAGE_ROOT	COMMENT
LOW	CMakeLists.txt	561	# symbols leak out of this module even under the project-wide hidden preset.	COMMENT
LOW	CMakeLists.txt	701	add_test(NAME precision_estimator_test COMMAND precision_estimator_test)	COMMENT
LOW	CMakeLists.txt	721	# third_party/onnx/onnx/inliner/CMakeLists.txt) for the inlining	COMMENT
LOW	CMakeLists.txt	781	add_test(NAME qat_entry_test COMMAND qat_entry_test)	COMMENT
LOW	pyproject.toml	41	# by raising this floor, since every other onnxruntime use in the package	COMMENT
LOW	pyproject.toml	61	# with that same algebra on concrete models. Unrelated to onnxsim's own	COMMENT
LOW	pyproject.toml	81	# Needed by onnxsim.export_sam2_model(), which traces a Meta SAM 2	COMMENT
LOW	pyproject.toml	101	# onnxsim.compile_training_loop -- training itself still runs entirely on	COMMENT
LOW	pyproject.toml	121	# pywebnn is rustnn's (https://github.com/rustnn/rustnn) Python binding: a	COMMENT
LOW	setup.py	41	# reachable from Simplify() itself.	COMMENT
LOW	setup.py	161	return	COMMENT
LOW	setup.py	221	# Without an override, onnx's own CMakeLists.txt (see its	COMMENT
LOW	setup.py	441	packages = setuptools.find_packages()	COMMENT
LOW	bench/graph_shape_inference_teardown_bench.cpp	1	// Microbenchmark for the arena change in graph_shape_inference.cc (ProcessNode).	COMMENT
LOW	bench/graph_shape_inference_teardown_bench.cpp	21	// [full cycle] build + destroy one node-visit's messages per iteration, a	COMMENT
LOW	bench/graph_shape_inference_teardown_bench.cpp	41	#include <chrono>	COMMENT
LOW	bench/graph_shape_inference_gperftools_repro.py	81	# elementwise ops on every block (Add/Mul/Sub), so ConstantDataFor +	COMMENT
LOW	bench/fold_teardown_bench.sh	1	#!/usr/bin/env bash	COMMENT
LOW	bench/fold_teardown_bench.cpp	1	// Microbenchmark for the arena change in onnxsim/onnxsim.cpp (RunOp).	COMMENT
LOW	bench/fold_teardown_bench.cpp	21	// [teardown] build N sub-models first, then time only their destruction	COMMENT
LOW	bench/graph_shape_inference_teardown_bench.sh	1	#!/usr/bin/env bash	COMMENT
LOW	…ols/onnx-cardputer-flash/firmware/runtime/src/main.cpp	1	// Generic TFLite Micro runtime for M5Stack Cardputer.	COMMENT
LOW	…-cardputer-flash/scripts/onnx_to_tflite_flatbuffers.py	761	# inference is batch-1 anyway).	COMMENT
LOW	…ter-flash/scripts/tflite_schema/schema_py_generated.py	1	# Vendored from the `ai-edge-litert` PyPI package (2.2.0)	COMMENT
LOW	tools/onnx-k210-flash/firmware/runtime/build.sh	1	#!/usr/bin/env bash	COMMENT
LOW	tools/onnx-k210-flash/firmware/runtime/src/main.cpp	1	// Generic K210 kmodel runtime -- the K210 counterpart to	COMMENT
LOW	tools/onnx-k210-flash/firmware/runtime/src/main.cpp	21	// onnx-cardputer-flash's TFLite Micro runtime has via	COMMENT
LOW	tools/onnx-k210-flash/firmware/runtime/src/w25qxx.h	21	#define DATALENGTH 8	COMMENT
LOW	tools/onnx-k210-flash/firmware/runtime/src/w25qxx.h	41	#define FAST_READ_QUAL_IO 0xEB	COMMENT
LOW	…x-k210-flash/firmware/camera_demo/src/board_m5stickv.c	1	// M5StickV board support: AXP192 power rails, the 135x240 ST7789 LCD on	COMMENT
LOW	…x-k210-flash/firmware/camera_demo/src/board_m5stickv.c	81	// ------------------------------------------------------------------ LCD ----	COMMENT
LOW	…/onnx-k210-flash/firmware/camera_demo/src/board_cube.c	1	// Sipeed Maix Cube board support: the 240x240 IPS LCD on octal SPI0 (a	COMMENT
LOW	…/onnx-k210-flash/firmware/camera_demo/src/board_cube.c	21		COMMENT
LOW	…/onnx-k210-flash/firmware/camera_demo/src/board_cube.c	61	i2c_send_data(I2C_DEVICE_1, b, 2);	COMMENT
LOW	…/onnx-k210-flash/firmware/camera_demo/src/board_cube.c	81	// MADCTL and the panel offset inside the ST7789's 240x320 RAM. MaixPy's	COMMENT
LOW	…/onnx-k210-flash/firmware/camera_demo/src/board_cube.c	101	//	COMMENT
LOW	…/onnx-k210-flash/firmware/camera_demo/src/board_cube.c	121	// a uniform pink/purple instead of the intended solid colour, and a later	COMMENT
LOW	…nx-k210-flash/firmware/camera_demo/src/camera_ov7740.c	1	// OV7740 camera + DVP capture, shared by every board this demo supports (the	COMMENT
LOW	…nx-k210-flash/firmware/camera_demo/src/camera_ov7740.c	21		COMMENT
LOW	tools/onnx-k210-flash/firmware/camera_demo/src/board.h	1	// Board and camera interface for the camera demo. One firmware per board,	COMMENT
LOW	tools/onnx-k210-flash/firmware/camera_demo/src/board.h	21	// CAM_W x CAM_H, so camera pixels map 1:1 onto LCD pixels.	COMMENT
LOW	tools/onnx-k210-flash/firmware/camera_demo/src/board.h	41	int board_button_b(void);	COMMENT
LOW	tools/onnx-k210-flash/firmware/camera_demo/src/main.cpp	1	// M5StickV camera demo: live camera -> the .kmodel flashed at 0x00C00000 (KPU)	COMMENT
LOW	…-deploy/include/onnx_deploy/static_kv_cache_pipeline.h	1	// onnx_deploy/static_kv_cache_pipeline.h	COMMENT
LOW	…-deploy/include/onnx_deploy/static_kv_cache_pipeline.h	21	// same buffers, via `Ort::IoBinding` binding the identical `Ort::Value` as	COMMENT
LOW	…-deploy/include/onnx_deploy/static_kv_cache_pipeline.h	121	std::to_string(shape[2]) + ") doesn't match max_cache_len (" +	COMMENT
LOW	…ls/onnx-deploy/include/onnx_deploy/kv_cache_pipeline.h	1	// onnx_deploy/kv_cache_pipeline.h	COMMENT
1921 more matches not shown…

Decorative Section Separators568 hits · 1795 pts

Severity	File	Line	Snippet	Context
MEDIUM	build_wasm_pyodide.sh	17	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	build_wasm_pyodide.sh	71	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	tests/test_halide_integration.py	188	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	tests/test_halide_integration.py	197	# ---------------------------------------------------------------------------	COMMENT
MEDIUM⚡	tests/test_gpu_backend.py	138	# --------------------------------------------------------------------------	COMMENT
MEDIUM⚡	tests/test_gpu_backend.py	140	# --------------------------------------------------------------------------	COMMENT
MEDIUM⚡	tests/test_gpu_backend.py	194	# --------------------------------------------------------------------------	COMMENT
MEDIUM⚡	tests/test_gpu_backend.py	196	# --------------------------------------------------------------------------	COMMENT
MEDIUM⚡	tests/test_gpu_backend.py	239	# --------------------------------------------------------------------------	COMMENT
MEDIUM⚡	tests/test_gpu_backend.py	241	# --------------------------------------------------------------------------	COMMENT
MEDIUM⚡	tests/test_gpu_backend.py	385	# --------------------------------------------------------------------------	COMMENT
MEDIUM⚡	tests/test_gpu_backend.py	387	# --------------------------------------------------------------------------	COMMENT
MEDIUM⚡	tests/test_gpu_backend.py	556	# --------------------------------------------------------------------------	COMMENT
MEDIUM⚡	tests/test_gpu_backend.py	558	# --------------------------------------------------------------------------	COMMENT
MEDIUM⚡	tests/test_gpu_backend.py	696	# --------------------------------------------------------------------------	COMMENT
MEDIUM⚡	tests/test_gpu_backend.py	698	# --------------------------------------------------------------------------	COMMENT
MEDIUM	tests/test_gpu_backend.py	46	# --------------------------------------------------------------------------	COMMENT
MEDIUM	tests/test_gpu_backend.py	48	# --------------------------------------------------------------------------	COMMENT
MEDIUM	tests/test_gpu_backend.py	607	# --------------------------------------------------------------------------	COMMENT
MEDIUM	tests/test_gpu_backend.py	609	# --------------------------------------------------------------------------	COMMENT
MEDIUM⚡	tests/test_formal_verify_gguf_ternary_quant.py	208	# ============================================================================	COMMENT
MEDIUM⚡	tests/test_formal_verify_gguf_ternary_quant.py	211	# ============================================================================	COMMENT
MEDIUM⚡	tests/test_formal_verify_gguf_ternary_quant.py	275	# ============================================================================	COMMENT
MEDIUM⚡	tests/test_formal_verify_gguf_ternary_quant.py	278	# ============================================================================	COMMENT
MEDIUM⚡	tests/test_formal_verify_gguf_ternary_quant.py	326	# ============================================================================	COMMENT
MEDIUM⚡	tests/test_formal_verify_gguf_ternary_quant.py	329	# ============================================================================	COMMENT
MEDIUM⚡	tests/test_formal_verify_gguf_ternary_quant.py	441	# ============================================================================	COMMENT
MEDIUM⚡	tests/test_formal_verify_gguf_ternary_quant.py	444	# ============================================================================	COMMENT
MEDIUM	tests/test_formal_verify_gguf_ternary_quant.py	560	# ============================================================================	COMMENT
MEDIUM	tests/test_formal_verify_gguf_ternary_quant.py	562	# ============================================================================	COMMENT
MEDIUM⚡	tests/test_formal_verify_rewrite_bev_pool_to_scatter.py	139	# ---------------------------------------------------------------------------	COMMENT
MEDIUM⚡	tests/test_formal_verify_rewrite_bev_pool_to_scatter.py	141	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	tests/test_formal_verify_rewrite_bev_pool_to_scatter.py	277	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	tests/test_formal_verify_rewrite_bev_pool_to_scatter.py	279	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	tests/test_formal_verify_rewrite_bev_pool_to_scatter.py	426	# ---------------------------------------------------------------------------	STRING
MEDIUM	tests/test_formal_verify_rewrite_bev_pool_to_scatter.py	430	# ---------------------------------------------------------------------------	STRING
MEDIUM	tests/test_d2quant.py	88	# ---------------------------------------------------------------------------	STRING
MEDIUM	tests/test_d2quant.py	90	# ---------------------------------------------------------------------------	STRING
MEDIUM	tests/test_d2quant.py	242	# ---------------------------------------------------------------------------	STRING
MEDIUM	tests/test_d2quant.py	244	# ---------------------------------------------------------------------------	STRING
MEDIUM⚡	…ts/test_formal_verify_qoperator_quantize_activation.py	154	# ---------------------------------------------------------------------------	COMMENT
MEDIUM⚡	…ts/test_formal_verify_qoperator_quantize_activation.py	156	# ---------------------------------------------------------------------------	COMMENT
MEDIUM⚡	…ts/test_formal_verify_qoperator_quantize_activation.py	260	# ---------------------------------------------------------------------------	COMMENT
MEDIUM⚡	…ts/test_formal_verify_qoperator_quantize_activation.py	262	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	…ts/test_formal_verify_qoperator_quantize_activation.py	386	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	…ts/test_formal_verify_qoperator_quantize_activation.py	388	# ---------------------------------------------------------------------------	COMMENT
MEDIUM⚡	…ts/test_formal_verify_qoperator_quantize_activation.py	670	# ---------------------------------------------------------------------------	STRING
MEDIUM⚡	…ts/test_formal_verify_qoperator_quantize_activation.py	672	# ---------------------------------------------------------------------------	STRING
MEDIUM	tests/test_mlx_integration.py	160	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	tests/test_mlx_integration.py	169	# ---------------------------------------------------------------------------	COMMENT
MEDIUM⚡	tests/test_imatrix_quant.py	67	# ---------------------------------------------------------------------------	STRING
MEDIUM⚡	tests/test_imatrix_quant.py	69	# ---------------------------------------------------------------------------	STRING
MEDIUM⚡	tests/test_imatrix_quant.py	159	# ---------------------------------------------------------------------------	STRING
MEDIUM⚡	tests/test_imatrix_quant.py	161	# ---------------------------------------------------------------------------	STRING
MEDIUM⚡	tests/test_imatrix_quant.py	200	# ---------------------------------------------------------------------------	STRING
MEDIUM⚡	tests/test_imatrix_quant.py	202	# ---------------------------------------------------------------------------	STRING
MEDIUM⚡	…/test_formal_verify_rewrite_trt_batched_rotated_nms.py	135	# ===========================================================================	COMMENT
MEDIUM⚡	…/test_formal_verify_rewrite_trt_batched_rotated_nms.py	137	# ===========================================================================	COMMENT
MEDIUM⚡	…/test_formal_verify_rewrite_trt_batched_rotated_nms.py	764	# ===========================================================================	COMMENT
MEDIUM⚡	…/test_formal_verify_rewrite_trt_batched_rotated_nms.py	766	# ===========================================================================	COMMENT
508 more matches not shown…

Unused Imports1356 hits · 1096 pts

Severity	File	Line	Context
LOW	bench/nncf_comparison.py	65	CODE
LOW	bench/onnxslim_comparison.py	17	CODE
LOW	bench/opset_conversion_profile.py	25	CODE
LOW	bench/opset_conversion_profile.py	31	CODE
LOW	bench/opset_conversion_profile.py	35	CODE
LOW	bench/sg2002/profile_model.py	8	CODE
LOW	…-cardputer-flash/scripts/onnx_to_tflite_flatbuffers.py	95	CODE
LOW	…s/onnx-cardputer-flash/scripts/onnx_to_tflite_micro.py	22	CODE
LOW	…ter-flash/scripts/tflite_schema/schema_py_generated.py	917	CODE
LOW	…ter-flash/scripts/tflite_schema/schema_py_generated.py	1177	CODE
LOW	…ter-flash/scripts/tflite_schema/schema_py_generated.py	1443	CODE
LOW	…ter-flash/scripts/tflite_schema/schema_py_generated.py	1637	CODE
LOW	…ter-flash/scripts/tflite_schema/schema_py_generated.py	1759	CODE
LOW	…ter-flash/scripts/tflite_schema/schema_py_generated.py	1881	CODE
LOW	…ter-flash/scripts/tflite_schema/schema_py_generated.py	2232	CODE
LOW	…ter-flash/scripts/tflite_schema/schema_py_generated.py	2413	CODE
LOW	…ter-flash/scripts/tflite_schema/schema_py_generated.py	2692	CODE
LOW	…ter-flash/scripts/tflite_schema/schema_py_generated.py	2692	CODE
LOW	…ter-flash/scripts/tflite_schema/schema_py_generated.py	3015	CODE
LOW	…ter-flash/scripts/tflite_schema/schema_py_generated.py	3202	CODE
LOW	…ter-flash/scripts/tflite_schema/schema_py_generated.py	3456	CODE
LOW	…ter-flash/scripts/tflite_schema/schema_py_generated.py	3796	CODE
LOW	…ter-flash/scripts/tflite_schema/schema_py_generated.py	4280	CODE
LOW	…ter-flash/scripts/tflite_schema/schema_py_generated.py	4496	CODE
LOW	…ter-flash/scripts/tflite_schema/schema_py_generated.py	4684	CODE
LOW	…ter-flash/scripts/tflite_schema/schema_py_generated.py	4997	CODE
LOW	…ter-flash/scripts/tflite_schema/schema_py_generated.py	5170	CODE
LOW	…ter-flash/scripts/tflite_schema/schema_py_generated.py	5362	CODE
LOW	…ter-flash/scripts/tflite_schema/schema_py_generated.py	5866	CODE
LOW	…ter-flash/scripts/tflite_schema/schema_py_generated.py	6278	CODE
LOW	…ter-flash/scripts/tflite_schema/schema_py_generated.py	6454	CODE
LOW	…ter-flash/scripts/tflite_schema/schema_py_generated.py	7355	CODE
LOW	…ter-flash/scripts/tflite_schema/schema_py_generated.py	9536	CODE
LOW	…ter-flash/scripts/tflite_schema/schema_py_generated.py	10756	CODE
LOW	…ter-flash/scripts/tflite_schema/schema_py_generated.py	16450	CODE
LOW	…ter-flash/scripts/tflite_schema/schema_py_generated.py	17634	CODE
LOW	…ter-flash/scripts/tflite_schema/schema_py_generated.py	18070	CODE
LOW	…ter-flash/scripts/tflite_schema/schema_py_generated.py	18423	CODE
LOW	…ter-flash/scripts/tflite_schema/schema_py_generated.py	18635	CODE
LOW	…ter-flash/scripts/tflite_schema/schema_py_generated.py	19228	CODE
LOW	…ter-flash/scripts/tflite_schema/schema_py_generated.py	19612	CODE
LOW	tools/onnx-k210-flash/tests/test_onnx_legalizer.py	17	CODE
LOW	tools/onnx-k210-flash/scripts/onnx_to_kmodel.py	52	CODE
LOW	tools/onnx-k210-flash/scripts/onnx_legalizer.py	33	CODE
LOW	tools/onnx-deploy/scripts/make_toy_seq2seq.py	51	CODE
LOW	tools/onnx-finetune/scripts/generate_nas_step_graphs.py	54	CODE
LOW	tools/onnx-finetune/scripts/federated_lora_aggregate.py	37	CODE
LOW	tools/onnx-finetune/scripts/apply_trained_weights.py	21	CODE
LOW	…finetune/scripts/generate_darts_supernet_step_graph.py	80	CODE
LOW	…x-finetune/scripts/generate_distillation_step_graph.py	94	CODE
LOW	…finetune/scripts/generate_federated_lora_step_graph.py	48	CODE
LOW	tools/onnx-remote/python/onnx_remote_client.py	55	CODE
LOW	tests/test_axera_llm_row_decomposition.py	10	CODE
LOW	tests/test_axera_step_runner.py	10	CODE
LOW	tests/test_graph_grad_templates.py	45	CODE
LOW	tests/test_axera_safe_masked_div_hardware.py	3	CODE
LOW	tests/test_crown_bab.py	33	CODE
LOW	…/test_formal_verify_rewrite_trt_batched_rotated_nms.py	130	CODE
LOW	…/test_formal_verify_rewrite_trt_batched_rotated_nms.py	132	CODE
LOW	tests/test_hmx_gemm.py	8	CODE
1296 more matches not shown…

Empty Error Handlers114 hits · 538 pts

Severity	File	Line	Snippet	Context
HIGH	bench/peak_memory.py	173	except SystemExit:	CODE
HIGH	bench/sg2002/profile_model.py	291	except Exception:	CODE
HIGH	bench/sg2002/profile_model.py	287	except OSError:	CODE
HIGH	…ter-flash/scripts/tflite_schema/schema_py_generated.py	918	except:	CODE
HIGH	…ter-flash/scripts/tflite_schema/schema_py_generated.py	1178	except:	CODE
HIGH	…ter-flash/scripts/tflite_schema/schema_py_generated.py	1444	except:	CODE
HIGH	…ter-flash/scripts/tflite_schema/schema_py_generated.py	1638	except:	CODE
HIGH	…ter-flash/scripts/tflite_schema/schema_py_generated.py	1760	except:	CODE
HIGH	…ter-flash/scripts/tflite_schema/schema_py_generated.py	1882	except:	CODE
HIGH	…ter-flash/scripts/tflite_schema/schema_py_generated.py	2037	except:	CODE
HIGH	…ter-flash/scripts/tflite_schema/schema_py_generated.py	2233	except:	CODE
HIGH	…ter-flash/scripts/tflite_schema/schema_py_generated.py	2414	except:	CODE
HIGH	…ter-flash/scripts/tflite_schema/schema_py_generated.py	2693	except:	CODE
HIGH	…ter-flash/scripts/tflite_schema/schema_py_generated.py	3016	except:	CODE
HIGH	…ter-flash/scripts/tflite_schema/schema_py_generated.py	3203	except:	CODE
HIGH	…ter-flash/scripts/tflite_schema/schema_py_generated.py	3457	except:	CODE
HIGH	…ter-flash/scripts/tflite_schema/schema_py_generated.py	3797	except:	CODE
HIGH	…ter-flash/scripts/tflite_schema/schema_py_generated.py	4281	except:	CODE
HIGH	…ter-flash/scripts/tflite_schema/schema_py_generated.py	4497	except:	CODE
HIGH	…ter-flash/scripts/tflite_schema/schema_py_generated.py	4685	except:	CODE
HIGH	…ter-flash/scripts/tflite_schema/schema_py_generated.py	4998	except:	CODE
HIGH	…ter-flash/scripts/tflite_schema/schema_py_generated.py	5171	except:	CODE
HIGH	…ter-flash/scripts/tflite_schema/schema_py_generated.py	5363	except:	CODE
HIGH	…ter-flash/scripts/tflite_schema/schema_py_generated.py	5867	except:	CODE
HIGH	…ter-flash/scripts/tflite_schema/schema_py_generated.py	6279	except:	CODE
HIGH	…ter-flash/scripts/tflite_schema/schema_py_generated.py	6455	except:	CODE
HIGH	…ter-flash/scripts/tflite_schema/schema_py_generated.py	7356	except:	CODE
HIGH	…ter-flash/scripts/tflite_schema/schema_py_generated.py	9537	except:	CODE
HIGH	…ter-flash/scripts/tflite_schema/schema_py_generated.py	10757	except:	CODE
HIGH	…ter-flash/scripts/tflite_schema/schema_py_generated.py	16451	except:	CODE
HIGH	…ter-flash/scripts/tflite_schema/schema_py_generated.py	17635	except:	CODE
HIGH	…ter-flash/scripts/tflite_schema/schema_py_generated.py	18071	except:	CODE
HIGH	…ter-flash/scripts/tflite_schema/schema_py_generated.py	18424	except:	CODE
HIGH	…ter-flash/scripts/tflite_schema/schema_py_generated.py	18636	except:	CODE
HIGH	…ter-flash/scripts/tflite_schema/schema_py_generated.py	19229	except:	CODE
HIGH	…ter-flash/scripts/tflite_schema/schema_py_generated.py	19613	except:	CODE
HIGH	…s/onnx-k210-flash/firmware/camera_demo/scripts/grab.py	128	except ImportError:	CODE
HIGH	tests/test_axera_patch_scales.py	123	except ValueError:	CODE
HIGH	tests/test_build_edsr_train_step.py	103	except ValueError:	CODE
HIGH	tests/test_axera_mcode_structure.py	3180	except Exception:	CODE
HIGH	scripts/vnncomp_bench.py	342	except Exception: # noqa: BLE001 - torch missing / op unsupported: sampling only	CODE
HIGH	scripts/vnncomp_bench.py	147	except ValueError:	CODE
HIGH	scripts/gpu_bench.py	434	except ImportError:	CODE
HIGH	scripts/axelera/legalize.py	143	except Exception:	CODE
HIGH	scripts/rknn/nanopc_npu_utilization.py	131	except Exception:	CODE
HIGH	scripts/android/deploy/stages/partition.py	21	except Exception: # noqa: BLE001 -- best effort	CODE
HIGH	scripts/android/tinygrad_aot/export_cl.py	136	except CompileError: pass	CODE
HIGH	…pts/android/vision_models_probe/summarize_partition.py	24	except Exception: # noqa: BLE001 -- best effort; large models may not infer cleanly	CODE
HIGH	scripts/android/maskrcnn_e2e/onnx_compat.py	15	except Exception:	CODE
HIGH	scripts/axera/matmul_record_emit.py	630	except CalibrationError:	CODE
HIGH	scripts/axera/matmul_record_emit.py	681	except CalibrationError:	CODE
HIGH	scripts/axera/matmul_record_emit.py	645	except CalibrationError:	CODE
HIGH	scripts/axera/axcl_session.py	464	except OSError:	CODE
HIGH	scripts/axera/axcl_session.py	421	except OSError:	CODE
HIGH	scripts/axera/run_llm_rpc.py	77	except OSError:	CODE
HIGH	scripts/axera/run_llm_rpc.py	87	except OSError:	CODE
HIGH	scripts/axera/pulsar2_docker.py	132	except PermissionError:	CODE
HIGH	scripts/axera/pulsar2_docker.py	788	except subprocess.TimeoutExpired:	CODE
HIGH	scripts/axera/legalize.py	1222	except Exception:	CODE
HIGH	scripts/axera/tinygrad_ax_backend.py	1353	except ValueError:	CODE
54 more matches not shown…

Deep Nesting542 hits · 498 pts

Severity	File	Line	Context
LOW	setup.py	392	CODE
LOW	bench/sg2002/profile_model.py	142	CODE
LOW	…-cardputer-flash/scripts/onnx_to_tflite_flatbuffers.py	746	CODE
LOW	…finetune/scripts/generate_federated_lora_step_graph.py	61	CODE
LOW	tools/onnx-remote/python/onnx_remote_client.py	207	CODE
LOW	tests/test_axera_div_quant_fields_two_live_inputs.py	269	CODE
LOW	tests/test_halide_integration.py	339	CODE
LOW	tests/test_formal_verify_rewrite_bev_pool_to_scatter.py	144	CODE
LOW	tests/test_weight_only_quantize_matmul_nbits.py	237	CODE
LOW	tests/test_symexpr_kv_cache_consistency.py	233	CODE
LOW	tests/test_axera_step_runner.py	716	CODE
LOW	tests/test_axera_step_runner.py	2527	CODE
LOW	tests/test_cross_layer_equalization.py	76	CODE
LOW	tests/test_sparsegpt_pruning_cpp.py	692	CODE
LOW	tests/test_voyager_sdk_patterns.py	67	CODE
LOW	tests/test_voyager_sdk_patterns.py	80	CODE
LOW	tests/_bev_torch_custom_ops.py	180	CODE
LOW	tests/test_affinequant_cpp.py	107	CODE
LOW	tests/test_xdna_layer_engine.py	15	CODE
LOW	tests/test_xdna_blocked_stage.py	51	CODE
LOW	tests/test_axera_resource_model_census.py	203	CODE
LOW	tests/test_deform_conv_to_gather.py	101	CODE
LOW	…st_axera_conv_binary_cluster_quantization_synthesis.py	279	CODE
LOW	tests/test_mlx_integration.py	229	CODE
LOW	tests/test_axera_conv_dilation_receptive_field.py	128	CODE
LOW	tests/test_quark_int_presets_parity.py	112	CODE
LOW	tests/test_quark_int_presets_parity.py	373	CODE
LOW	tests/test_quark_int_presets_parity.py	504	CODE
LOW	tests/test_axera_sub_quant_fields_two_live_inputs.py	215	CODE
LOW	tests/test_axera_mul_quant_fields_two_live_inputs.py	268	CODE
LOW	…est_formal_verify_weight_only_quantize_matmul_nbits.py	513	CODE
LOW	…est_formal_verify_weight_only_quantize_matmul_nbits.py	629	CODE
LOW	…est_formal_verify_weight_only_quantize_matmul_nbits.py	746	CODE
LOW	tests/test_quant_int_verify.py	946	CODE
LOW	tests/test_quark_xint8_parity.py	352	CODE
LOW	tests/test_quark_xint8_parity.py	396	CODE
LOW	tests/test_quark_xint8_parity.py	1015	CODE
LOW	tests/test_quark_xint8_parity.py	1379	CODE
LOW	tests/test_quark_xint8_parity.py	409	CODE
LOW	tests/test_quark_finetune_parity.py	139	CODE
LOW	…st_formal_verify_rewrite_msdeformattn_to_gridsample.py	307	CODE
LOW	tests/test_simple.py	226	CODE
LOW	tests/test_build_resnet50_layer4_step.py	66	CODE
LOW	tests/test_shape_cost.py	232	CODE
LOW	tests/test_shape_cost.py	927	CODE
LOW	tests/test_qwen3_5_reconstruct.py	472	CODE
LOW	tests/test_axera_llm_layer_loop.py	410	CODE
LOW	tests/test_quark_int_presets.py	31	CODE
LOW	tests/test_quark_npu.py	756	CODE
LOW	tests/test_quarot_cpp.py	185	CODE
LOW	tests/test_quark_amp_mixing.py	314	CODE
LOW	tests/test_axera_conv_reg232_zeropoint_verification.py	128	CODE
LOW	tests/test_axera_mcode_structure.py	445	CODE
LOW	tests/test_axera_mcode_structure.py	4053	CODE
LOW	tests/test_axera_mcode_structure.py	4312	CODE
LOW	tests/test_axera_mcode_structure.py	5608	CODE
LOW	tests/test_axera_mcode_structure.py	5670	CODE
LOW	tests/test_axera_mcode_structure.py	6123	CODE
LOW	tests/test_axera_mcode_structure.py	5629	CODE
LOW	tests/test_axera_mcode_structure.py	6135	CODE
482 more matches not shown…

Excessive Try-Catch Wrapping407 hits · 411 pts

Severity	File	Line	Snippet	Context
LOW	bench/nncf_comparison.py	368	except Exception as exc: # pragma: no cover - reported, not raised	CODE
LOW	bench/nncf_comparison.py	375	except Exception as exc: # pragma: no cover - reported, not raised	CODE
LOW	bench/amd_npu_roofline.py	301	except Exception:	CODE
LOW	bench/onnxslim_comparison.py	69	except Exception as e: # noqa: BLE001	CODE
LOW	bench/onnxslim_comparison.py	82	except Exception as e: # noqa: BLE001	CODE
LOW	bench/sg2002/profile_model.py	271	except Exception as error: # noqa: BLE001 - present transport and device errors to the user	STRING
LOW	bench/sg2002/profile_model.py	291	except Exception:	STRING
LOW	…rdputer-flash/tests/test_onnx_to_tflite_flatbuffers.py	175	except Exception as e:	CODE
LOW	tools/onnx-k210-flash/tests/test_onnx_to_kmodel.py	88	except Exception as e:	CODE
LOW	tests/test_quark_low_opset_parity.py	48	except Exception as e: # pragma: no cover - environment dependent	CODE
LOW	tests/test_quark_low_opset_parity.py	143	except Exception:	CODE
LOW	tests/test_export_transformers.py	44	except Exception as e: # network/hub errors surface as a variety of types	CODE
LOW	tests/test_export_transformers.py	74	except Exception as e:	CODE
LOW	tests/test_export_transformers.py	94	except Exception as e:	CODE
LOW	tests/test_optimum_export_deploy.py	90	except Exception as e: # network/hub errors surface as a variety of types	CODE
LOW⚡	tests/test_docs_ranges.py	48	except Exception as e: # say which block of the page broke	CODE
LOW	tests/test_tvm_integration.py	180	except Exception:	CODE
LOW	tests/test_torch_integration.py	203	except Exception:	CODE
LOW⚡	tests/test_to_torch.py	439	except Exception as e: # onnxruntime build without these contrib kernels	STRING
LOW⚡	tests/test_to_torch.py	456	except Exception as e:	STRING
LOW⚡	tests/test_to_torch.py	667	except Exception as e:	STRING
LOW⚡	tests/test_to_torch.py	684	except Exception as e:	STRING
LOW	tests/test_to_torch.py	328	except Exception as e: # older onnxruntime: no opset-23 Attention kernel	STRING
LOW	tests/test_to_torch.py	513	except Exception as e:	STRING
LOW	tests/test_to_torch.py	568	except Exception as e:	STRING
LOW	tests/test_to_torch.py	636	except Exception as e:	STRING
LOW	tests/test_mlx_integration.py	291	except Exception:	CODE
LOW	tests/test_quark_constants_misc_parity.py	48	except Exception as e: # pragma: no cover - environment dependent	CODE
LOW	tests/test_export_transformers_whisper.py	51	except Exception as e: # network/hub errors surface as a variety of types	CODE
LOW	tests/test_quant_int_verify.py	728	except Exception as e: # pragma: no cover - API drift	CODE
LOW	tests/test_quark_xint8_parity.py	29	except Exception as e: # pragma: no cover - environment dependent	CODE
LOW	tests/test_function_rewriter_compile_rule.py	100	except Exception as exc: # pragma: no cover - environment/version guard	CODE
LOW	tests/test_quark_finetune_parity.py	70	except Exception as e: # pragma: no cover - environment dependent	CODE
LOW	tests/test_rknn3588_compat.py	46	except Exception:	CODE
LOW⚡	tests/test_docs_fp_error.py	31	except Exception as e: # say which block of the page broke	CODE
LOW	tests/test_static_kv_cache_export.py	55	except Exception as e: # network/hub errors surface as a variety of types	CODE
LOW	tests/test_static_kv_cache_export.py	186	except Exception as e:	CODE
MEDIUM	tests/test_static_kv_cache_export.py	52	def hf_model_config():	CODE
LOW⚡	tests/test_quark_hadamard_parity.py	361	except Exception as e: # pragma: no cover - environment dependent	CODE
LOW	tests/test_quark_hadamard_parity.py	36	except Exception: # pragma: no cover - environment dependent	CODE
LOW	tests/test_axera_mcode_structure.py	3180	except Exception:	CODE
LOW	tests/test_axera_mcode_structure.py	5886	except Exception:	CODE
LOW	tests/test_quark_finetune_coverage_parity.py	179	except Exception: # Quark cannot convert it (and draws nothing)	CODE
LOW	tests/test_quark_algo_parity.py	40	except Exception as e: # pragma: no cover - environment dependent	CODE
LOW	tests/test_quark_algo_parity.py	594	except Exception as e: # pragma: no cover - needs quark's torch extras	CODE
LOW⚡	tests/test_quark_parity.py	2062	except Exception: # Quark itself rejects this graph	CODE
LOW	tests/test_quark_parity.py	34	except Exception as e: # pragma: no cover - environment dependent	CODE
LOW	tests/test_quark_parity.py	153	except Exception: # pragma: no cover - environment dependent	CODE
LOW	tests/test_quark_parity.py	400	except Exception: # Quark itself rejects this graph	CODE
LOW	tests/test_quark_parity.py	596	except Exception as e: # pragma: no cover - op library missing etc.	CODE
LOW	tests/test_export_diffusion.py	46	except Exception as e: # network/hub errors surface as a variety of types	CODE
LOW⚡	tests/test_export_diffusion.py	69	except Exception as e:	CODE
LOW⚡	tests/test_export_diffusion.py	84	except Exception as e:	CODE
LOW	tests/test_torch_training.py	55	except Exception: # noqa: BLE001 -- any build failure means unavailable	CODE
LOW	tests/test_fuse_lp_normalization.py	94	except Exception:	STRING
LOW	tests/test_compile_training.py	61	except Exception: # noqa: BLE001 -- any build failure means unavailable	CODE
LOW	…ts/test_speculative_decoding_drafter_hub_checkpoint.py	213	except Exception as e: # network/hub errors surface as a variety of types	CODE
LOW	tests/test_quark_block_preproc_parity.py	52	except Exception as e: # pragma: no cover - environment dependent	CODE
LOW	tests/test_quark_block_preproc_parity.py	228	except Exception: # pragma: no cover - environment dependent	CODE
LOW	tests/test_quark_block_preproc_parity.py	256	except Exception: # pragma: no cover - a build of the op library that cannot load	CODE
347 more matches not shown…

Cross-Language Confusion81 hits · 380 pts

Severity	File	Line	Snippet	Context
HIGH	bench/sg2002/profile_model.py	83	state=$(cat "$proc" 2>/dev/null \|\| true)	CODE
HIGH	bench/sg2002/profile_model.py	92	while kill -0 "$runpid" 2>/dev/null; do	CODE
HIGH	tests/test_formal_verify_eliminate_shape_op.py	8	non-negative (``dim.is_int && dim.dim >= 0``).	STRING
HIGH	tests/test_formal_verify_eliminate_shape_gather.py	25	indices_val].is_int \|\| dims[indices_val].dim == -1``). Otherwise it builds a	STRING
HIGH⚡	tests/test_axera_mcode_validator.py	516	assert null == 0, (name, null)	CODE
HIGH	tests/test_axera_mcode_validator.py	259	assert null <= max(2, len(quintets) // 4), (name, len(quintets), null)	CODE
HIGH	tests/test_axera_mcode_validator.py	383	assert len(null) <= max(2, (len(prefixes) + len(doublets)) // 4), (	CODE
HIGH	tests/test_axera_mcode_validator.py	386	len(null),	CODE
HIGH	tests/test_axera_mcode_validator.py	638	assert null <= 2 * len(takes) + 2, (name, len(takes), null)	CODE
HIGH	tests/test_axera_mcode_validator.py	800	assert null == 0, (name, null)	CODE
HIGH	tests/test_axera_mcode_validator.py	966	assert null == 0, (name, null)	CODE
HIGH	tests/test_axera_mcode_validator.py	1043	assert null == 0, (name, null)	CODE
HIGH⚡	tests/test_amax_calibration.py	30	hist.push(np.array([v]))	CODE
HIGH⚡	tests/test_amax_calibration.py	36	hist.push(np.array([0.001]))	CODE
HIGH⚡	tests/test_amax_calibration.py	37	hist.push(np.array([0.001]))	CODE
HIGH⚡	tests/test_amax_calibration.py	38	hist.push(np.array([0.001]))	CODE
HIGH	tests/test_axera_llm_layer_loop.py	13	(`cmake -S tools/onnx-remote -B build/onnx-remote && cmake --build build/onnx-remote`,	STRING
HIGH	tests/test_axera_mcode_structure.py	2861	assert observed >= 40 and null > 0 and observed / null >= 2.0, (observed, null)	CODE
HIGH	tests/test_axera_mcode_structure.py	3524	assert null_a1 == 0 or null[(0xA1, 0)] <= 0.8 * null_a1 + 2, dict(null)	CODE
HIGH	tests/test_axera_mcode_structure.py	6208	assert real[1] >= 2 * null[1] and real[2] >= 2 * null[2], (name, real, null)	CODE
HIGH	tests/test_axera_mcode_structure.py	2835	shuffled null); the undecoded residue left after removing every	STRING
HIGH	tests/test_axera_mcode_structure.py	2897	that rested on the wrong null); and blocks A and B share a 158-byte	STRING
HIGH	tests/test_formal_verify_eliminate_nop_with_unit.py	13	all-zeros -- ``index == 1 && isAllOf(tensor, 0)``. Index 0 (the minuend)	STRING
HIGH	…ts/test_formal_verify_eliminate_nop_monotone_argmax.py	179	# and check_n=0 (Log(negative) is undefined, so a numeric equivalence	COMMENT
HIGH	tests/test_formal_verify_magnitude_pruning_matmul.py	373	# never be matched: FetchConstantTensor returns null, so	STRING
HIGH	tests/test_formal_verify_eliminate_nop_split.py	21	(!sizes[axis].is_int \|\| sizes[axis].dim != split[0])) {	STRING
HIGH	tests/test_formal_verify_magnitude_pruning_conv.py	359	# never be matched: FetchConstantTensor returns null, so	STRING
HIGH	tests/test_formal_verify_eliminate_nop_flatten.py	11	(``input_shape[0].is_int && input_shape[0].dim == 1``).	STRING
HIGH	tests/test_formal_verify_magnitude_pruning_attention.py	419	# never be matched: FetchConstantTensor returns null, so	STRING
HIGH	scripts/stable_relu_prune_eval.py	116	"""top-1 agreement, mean KL(a \|\| b), max \|logit diff\|, and accuracy of both if labels given."""	STRING
HIGH	scripts/rknn/overhead_probe_nanopc.py	272	"timer_floor_ms": round(statistics.median(null), 6),	CODE
HIGH	scripts/edgetpu/benchmark_device.py	19	sudo udevadm control --reload-rules && sudo udevadm trigger	STRING
HIGH	scripts/allwinner/compile_nbg.py	203	'trap "chmod -R a+rwX . 2>/dev/null \|\| true" EXIT', # a Docker run leaves root-owned files behind	CODE
HIGH	scripts/allwinner/trace_npu.py	84	"for c in 0 1 2 3 4 5 6 7; do echo $c $(cat /sys/devices/system/cpu/cpu$c/cpu_capacity 2>/dev/null); done",	STRING
HIGH	scripts/nvidia/imagenette_data.py	9	curl -LO https://s3.amazonaws.com/fast-ai-imageclas/imagenette2-320.tgz && tar xzf it.tgz	STRING
HIGH	scripts/openpilot_dsp/adaround_conv.py	8	min_V \|\| conv(X, W_q) - conv(X, W) \|\|^2 + lam * sum(1 - \|2 h(V) - 1\|^beta) (beta annealed 20 -> 2)	STRING
HIGH	scripts/android/hexagon_sim_harness.py	79	if (d > worst \|\| d != d) worst = d;	CODE
HIGH	scripts/android/deploy/deploy.py	358	device.push(ctx, ctx.d("push"))	CODE
HIGH	scripts/android/deploy/stages/device.py	171	adb(ctx, "shell", f"rm -rf {dd}/res && mkdir -p {dd}/res")	CODE
HIGH	scripts/android/webgpu_ops/texdirect_tune/tune_1x1.py	24	out = subprocess.run(["bash", "-c", f"~/.cache/android-phone/phone-run bash -c 'adb push {local} {D}/job.sh >/dev/nu	CODE
HIGH	scripts/android/webgpu_ops/texdirect_tune/validate.py	36	out = subprocess.run(["bash", "-c", f"~/.cache/android-phone/phone-run bash -c 'adb push {local} {D}/job.sh >/dev/nu	CODE
HIGH	scripts/android/webgpu_ops/texdirect_tune/tune.py	42	["bash", "-c", f"~/.cache/android-phone/phone-run bash -c 'adb push {local} {D}/job.sh >/dev/null; adb shell \"s	CODE
HIGH	scripts/android/webgpu_ops/texdirect_tune/accuracy.py	10	cmd = (f"cd {D} && export LD_LIBRARY_PATH=. && rm -rf dcpu dtd dbase && mkdir dcpu dtd dbase && "	CODE
HIGH	scripts/android/webgpu_ops/texdirect_tune/accuracy.py	16	subprocess.run(["bash", "-c", f"~/.cache/android-phone/phone-run bash -c 'adb push {tmp}/job.sh {D}/acc.sh >/dev/nul	CODE
HIGH	scripts/android/webgpu_ops/texdirect_tune/tune_cc.py	21	out = subprocess.run(["bash", "-c", f"~/.cache/android-phone/phone-run bash -c 'adb push {local} {D}/job.sh >/dev/nu	CODE
HIGH	scripts/android/webgpu_ops/attr/run_ablation2.py	15	subprocess.run(["bash", "-c", f"{phone_run} bash -c 'adb push {here}/ab2.sh /data/local/tmp/wg-attr2/ >/dev/null; adb pu	CODE
HIGH	scripts/android/webgpu_ops/attr/run_ablation2.py	31	cmd = f"{phone_run} bash -c 'adb push {work}/{var} /data/local/tmp/wg-attr2/m/ >/dev/null; adb shell \"sh /data/loca	CODE
HIGH	scripts/android/webgpu_ops/attr/run_ablation.py	14	subprocess.run(["bash", "-c", f"{phone_run} bash -c 'adb push {here}/ab.sh /data/local/tmp/wg-attr/ >/dev/null; adb push	CODE
HIGH	scripts/android/webgpu_ops/attr/run_ablation.py	30	cmd = f"{phone_run} bash -c 'adb push {work}/{var} /data/local/tmp/wg-attr/m/ >/dev/null; adb shell \"sh /data/local	CODE
HIGH	scripts/android/tinygrad_kernels/tune.py	42	f"adb shell 'mkdir -p {PHONE_DIR}/{name}' >/dev/null; adb push {runner_dir}/{runner} {PHONE_DIR}/ >/dev/null; "	CODE
HIGH	scripts/android/vision_models/fastbev/e2e_phone.py	88	adb -s {SERIAL} shell "chmod 755 {R}/fastbev_run && cd {R} && {env} ./fastbev_run {a.fam} {R}/{a.fam} {len(frames)} {a.i	CODE
HIGH	scripts/android/vision_models/nfru/nfru_cl_check.py	164	vp, sw = self.null, srch	CODE
HIGH	scripts/android/vision_models/nfru/nfru_cl_check.py	183	won if hint_mv is not None else self.null,	CODE
HIGH	scripts/android/vision_models/nfru/nfru_cl_check.py	270	g.null,	CODE
HIGH	scripts/android/vision_models/nfru/nfru_cl_check.py	271	g.null,	CODE
HIGH	scripts/android/vision_models/nfru/nfru_cl_check.py	546	g.null,	CODE
HIGH	scripts/android/vision_models/sparse4d/phone_chain.py	45	got = adb("shell", f"stat -c %s {dst} 2>/dev/null \|\| true", capture=True).strip()	CODE
HIGH	scripts/android/vision_models/rfdetr/phone_eval.py	106	f"cd {rd} && rm -rf ctx.onnx res && mkdir res\n"	CODE
HIGH	scripts/android/vision_models/rfdetr/phone_eval.py	136	f"rm -rf {stage}/pulled && adb -s {SERIAL} pull -q {rd}/res {stage}/pulled"	CODE
HIGH	scripts/android/vision_models/rtdetr/phone_eval.py	131	f"cd {rd} && rm -rf ctx.onnx res && mkdir res\n"	CODE
21 more matches not shown…

Hallucination Indicators29 hits · 330 pts

Severity	File	Line	Snippet	Context
CRITICAL	tests/test_axera_mcode_structure.py	5929	inp.type.tensor_type.shape.dim.add().dim_value = value	CODE
CRITICAL	scripts/apple/train_block_step_coreml.py	189	vi.type.tensor_type.shape.dim.add().dim_value = 1	CODE
CRITICAL	scripts/apple/train_mlp_step_coreml.py	170	vi.type.tensor_type.shape.dim.add().dim_value = 1	CODE
CRITICAL⚡	scripts/tinyengine/models.py	97	x_type.type.tensor_type.shape.dim.add().dim_param = "N"	CODE
CRITICAL⚡	scripts/tinyengine/models.py	99	x_type.type.tensor_type.shape.dim.add().dim_value = d	CODE
CRITICAL⚡	scripts/tinyengine/models.py	101	y_type.type.tensor_type.shape.dim.add().dim_param = "N"	CODE
CRITICAL⚡	scripts/tinyengine/models.py	103	y_type.type.tensor_type.shape.dim.add().dim_value = d	CODE
CRITICAL	scripts/convertmodel/test/make_edgeai_tidl_fixture.py	87	x_type.type.tensor_type.shape.dim.add().dim_param = "N"	CODE
CRITICAL	scripts/convertmodel/test/make_edgeai_tidl_fixture.py	89	x_type.type.tensor_type.shape.dim.add().dim_value = d	CODE
CRITICAL	scripts/allwinner/compile_nbg.py	120	v.type.tensor_type.shape.dim.add().dim_value = d	CODE
CRITICAL	scripts/android/webgpu_ops/model_flops.py	24	i.type.tensor_type.shape.dim.add().dim_value = int(d)	CODE
CRITICAL	scripts/android/vision_models/nfru/nfru.py	402	o.type.tensor_type.shape.dim.add().dim_value = v	CODE
CRITICAL	scripts/android/vision_models/nss/nss.py	413	o.type.tensor_type.shape.dim.add().dim_value = v	CODE
CRITICAL⚡	scripts/edgeai/models.py	82	x_type.type.tensor_type.shape.dim.add().dim_param = "N"	CODE
CRITICAL⚡	scripts/edgeai/models.py	84	x_type.type.tensor_type.shape.dim.add().dim_value = d	CODE
CRITICAL⚡	scripts/edgeai/models.py	86	y_type.type.tensor_type.shape.dim.add().dim_param = "N"	CODE
CRITICAL⚡	scripts/edgeai/models.py	88	y_type.type.tensor_type.shape.dim.add().dim_value = d	CODE
CRITICAL	scripts/axera/build_calib_signal_probe.py	165	inp.type.tensor_type.shape.dim.add().dim_value = 1	CODE
CRITICAL	scripts/axera/reshape_dma_emit.py	76	value_info.type.tensor_type.shape.dim.add().dim_value = size	CODE
CRITICAL	scripts/axera/reshape_emit.py	146	value_info.type.tensor_type.shape.dim.add().dim_value = size	CODE
CRITICAL	scripts/axera/teng2_cluster_patch_npoint.py	150	value_info.type.tensor_type.shape.dim.add().dim_value = dim	CODE
CRITICAL	scripts/axera/build_resident_train_step.py	814	inp.type.tensor_type.shape.dim.add().dim_value = 1	CODE
CRITICAL	scripts/axera/memory_emit.py	123	value_info.type.tensor_type.shape.dim.add().dim_value = size	CODE
CRITICAL	scripts/axera/legalize.py	569	value.type.tensor_type.shape.dim.add().dim_value = 1	CODE
CRITICAL	scripts/axera/transpose_emit.py	70	value_info.type.tensor_type.shape.dim.add().dim_value = size	CODE
CRITICAL	scripts/axera/teng2_cluster_patch.py	185	value_info.type.tensor_type.shape.dim.add().dim_value = dim	CODE
CRITICAL	scripts/axera/reshape_record_emit.py	409	value_info.type.tensor_type.shape.dim.add().dim_value = size	CODE
CRITICAL	onnxsim/interval.py	1329	d = vi.type.tensor_type.shape.dim.add()	CODE
CRITICAL	onnxsim/shape_cost.py	1244	vi.type.tensor_type.shape.dim.add().dim_value = v	CODE

AI Slop Vocabulary82 hits · 215 pts

Severity	File	Line	Snippet	Context
MEDIUM	tests/test_halide_integration.py	211	"""Kahn's algorithm over node I/O names; robust to any valid node order."""	STRING
MEDIUM	tests/test_formal_verify_gguf_ternary_quant.py	727	# Structural sanity, robust to however many fixed-point iterations ran:	STRING
MEDIUM	tests/test_axera_sub_div_reflow_extended_seeds.py	49	#1681's own sample) -- the "never reflows" observation is more robust	STRING
MEDIUM	tests/test_gguf_q5_k_cpp.py	130	# A coarser naive baseline keeps this test meaningful and robust.	STRING
MEDIUM	tests/test_qoperator_quantize_concat.py	114	# own static values, not force-fed through the calibration harness.	STRING
MEDIUM	tests/test_deform_conv_to_gather.py	155	# Simplify + run-and-compare harness.	COMMENT
MEDIUM	tests/test_mlx_integration.py	185	"""Kahn's algorithm over node I/O names; robust to any valid node order."""	STRING
MEDIUM	tests/test_edgeai_tidl_compat.py	20	# The TIDL harness lives under scripts/edgeai; reuse it rather than duplicate.	COMMENT
MEDIUM	tests/test_qnn_compat.py	22	# The QNN harness lives under scripts/qualcomm; reuse it rather than duplicate.	COMMENT
MEDIUM⚡	…/test_formal_verify_rewrite_trt_batched_rotated_nms.py	765	# Threshold-sweep harness: reads the real pass's internal IoU indirectly.	COMMENT
MEDIUM	tests/test_rknn_compat.py	26	# The RKNN harness lives under scripts/rknn; reuse it rather than duplicate.	COMMENT
MEDIUM	tests/test_formal_verify_qoperator_quantize_concat.py	435	# static values instead of force-fed through the calibration harness as	STRING
MEDIUM	tests/test_formal_verify_qoperator_quantize_where.py	441	# harness as if it varied at inference time.	STRING
MEDIUM	tests/test_openvino_compat.py	20	# The OpenVINO harness lives under scripts/intel; reuse it rather than duplicate.	COMMENT
MEDIUM	tests/test_shape_cost.py	169	# --------------------------------------------------------------------------- harness	COMMENT
MEDIUM	tests/test_hqq.py	118	# robust (Lp<2) fit should recover a tighter, more accurate zero-point	STRING
MEDIUM	tests/test_hqq.py	144	# HQQ's robust fit should show its benefit most clearly.	STRING
MEDIUM	tests/test_qoperator_quantize_elementwise.py	177	# through the calibration harness.	STRING
MEDIUM	tests/test_yolo.py	19	# dominates YOLO's node-count reduction. The standalone compatibility harness	COMMENT
MEDIUM	tests/test_qoperator_quantize_where.py	113	# own static values, not force-fed through the calibration harness.	STRING
MEDIUM	tests/test_nanochat.py	16	# harness in scripts/nanochat/simplify_nanochat.py exercises nanochat's own	COMMENT
MEDIUM	tests/test_formal_verify_eliminate_nop_reshape.py	279	# input, so onnxsim's random-input test harness -- which special-cases a	COMMENT
MEDIUM	tests/test_tinygrad_hexagon_compat.py	21	# The harness lives under scripts/hexagon; reuse it rather than duplicate.	COMMENT
MEDIUM	tests/test_coreml_compat.py	19	# The Core ML harness lives under scripts/apple; reuse it rather than duplicate.	COMMENT
LOW	tests/test_quantize_fp16.py	269	# quantized model, not just pass the (more lenient) checker.	STRING
MEDIUM	tests/test_axera_sparse_resource_clustering.py	356	# robust population-level gap. Redesigned here (see module	COMMENT
MEDIUM	tests/test_quarot.py	253	# pair. Empirically (see this session's verification harness) the	STRING
MEDIUM	tests/test_gptaq.py	214	# standalone harness simulating cross-platform numerical noise: every	STRING
MEDIUM	tests/test_zeroquant.py	201	# robust to platform-specific float32 MatMul reduction-order	STRING
MEDIUM	tests/test_hexagon_tinygrad.py	577	# EXACT on the phone; tinygrad_hexagon_bridge/README.md, TopK section), so the harness exits 1 by	COMMENT
MEDIUM	tests/test_adpq_cpp.py	108	# wide margin (small noise -> small robust sigma -> small threshold),	STRING
MEDIUM	tests/test_hexagon_qfloat.py	161	("f32", False, 1e-5), # control: stock TVM through the same harness	CODE
MEDIUM	tests/test_pulsar2_compat.py	19	# The Pulsar2 harness lives under scripts/axera; reuse it rather than duplicate.	COMMENT
MEDIUM⚡	tests/test_trt_batched_rotated_nms.py	242	# Threshold-sweep harness against the actual pass (see module docstring).	COMMENT
MEDIUM	tests/test_migraphx_compat.py	22	# The MIGraphX harness lives under scripts/amd; reuse it rather than duplicate.	COMMENT
MEDIUM	scripts/make_qat_parity_fixtures.py	561	# silent divergence this harness exists to prevent, and it happened.	COMMENT
MEDIUM	scripts/shape_cost_validation.py	31	import test_shape_cost as T # noqa: E402 (the models and the run harness live with the tests)	CODE
MEDIUM⚡	scripts/gpu_bench.py	187	# cases: each returns a dict of results; the harness adds time and memory	COMMENT
MEDIUM	scripts/rknn/rknn_backend.py	104	# The target platform used for every build in this harness. Purely a	COMMENT
MEDIUM	scripts/rknn/rknn_backend.py	180	# do_quantization=False: this harness checks graph-compile and	COMMENT
MEDIUM	scripts/rknn3/rknn3_backend.py	209	# in the onnxsim repo checkout during development of this harness) --	COMMENT
MEDIUM	scripts/android/webgpu_ops/attr/opbench.py	29	# calibration entry: a tiny Relu, its per-copy cost is the harness floor (keeper Slice + graph output copy)	COMMENT
MEDIUM	…pts/android/tinygrad_hexagon_bridge/hex_gemm_kernel.py	66	# first. Depending on every range (M can never degenerate) makes this robust.	COMMENT
LOW	scripts/regression/worker.py	67	# isn't (many repos just call it "model.onnx").	COMMENT
MEDIUM	…s/regression/paddle2onnx/run_paddle2onnx_regression.py	194	# Verification harness failure is not an onnxsim failure; note it but	COMMENT
MEDIUM	scripts/yolo/simplify_yolo.py	54	# Nano scale keeps the harness fast; the graph structure (what onnxsim	COMMENT
MEDIUM	.github/workflows/model-regression.yml	58	# Validate the harness itself whenever it changes.	COMMENT
MEDIUM	.github/workflows/model-regression.yml	378	# pull_request runs, which only validate the harness.	COMMENT
MEDIUM	.github/workflows/amd-integration.yml	48	# installed wheel on `import onnxsim`. Reference the harness by absolute	COMMENT
MEDIUM	.github/workflows/big-endian.yml	14	# requests that touch either the harness or the code that actually reads and	COMMENT
MEDIUM	.github/workflows/big-endian.yml	87	# The harness, so changes to it are validated before they are relied on.	COMMENT
MEDIUM	.github/workflows/ep-integration.yml	21	# Scheduled + on demand, plus on PRs that touch any harness so they stay	COMMENT
MEDIUM	.github/workflows/ep-integration.yml	99	# installed wheel on `import onnxsim`. Reference the harness by absolute	COMMENT
MEDIUM	.github/workflows/ep-integration.yml	141	# installed wheel on `import onnxsim`. Reference the harness by absolute	COMMENT
MEDIUM	.github/workflows/ep-integration.yml	189	# installed wheel on `import onnxsim`. Reference the harness by absolute	COMMENT
MEDIUM	.github/workflows/apple-integration.yml	14	# either harness so they stay valid.	COMMENT
MEDIUM	.github/workflows/apple-integration.yml	60	# installed wheel on `import onnxsim`. Reference the harness by absolute	COMMENT
MEDIUM	.github/workflows/apple-integration.yml	97	# installed wheel on `import onnxsim`. Reference the harness by absolute	COMMENT
MEDIUM	.github/workflows/rknn-integration.yml	9	# NCHW-vs-NHWC compatibility issues this harness works around.	COMMENT
MEDIUM	.github/workflows/rknn-integration.yml	11	# Scheduled + on demand, plus on PRs that touch the harness so it stays	COMMENT
22 more matches not shown…

AI Structural Patterns128 hits · 111 pts

Severity	File	Line	Context
LOW	…ter-flash/scripts/tflite_schema/schema_py_generated.py	2699	CODE
LOW	…ter-flash/scripts/tflite_schema/schema_py_generated.py	5873	CODE
LOW	…ter-flash/scripts/tflite_schema/schema_py_generated.py	6870	CODE
LOW	…ter-flash/scripts/tflite_schema/schema_py_generated.py	18077	CODE
LOW	…ter-flash/scripts/tflite_schema/schema_py_generated.py	19619	CODE
LOW	tests/test_cross_layer_equalization.py	76	CODE
LOW	tests/test_fuse_gqa.py	47	CODE
LOW	tests/test_deform_conv_to_gather.py	206	CODE
LOW	tests/test_fuse_attention.py	94	CODE
LOW	tests/test_fuse_attention.py	303	CODE
LOW	tests/test_fuse_attention.py	538	CODE
LOW	tests/test_structured_pruning_cpp.py	8334	CODE
LOW	tests/test_qmoe_pruning_cpp.py	236	CODE
LOW	tests/test_qmoe_pruning_cpp.py	447	CODE
LOW	tests/test_trt_batched_nms.py	70	CODE
LOW	tests/test_attention_head_wanda_pruning_cpp.py	2105	CODE
LOW	tests/test_attention_head_wanda_pruning_cpp.py	2412	CODE
LOW	tests/test_attention_head_wanda_pruning_cpp.py	2623	CODE
LOW	tests/test_attention_head_wanda_pruning_cpp.py	2766	CODE
LOW	tests/test_attention_head_wanda_pruning_cpp.py	2943	CODE
LOW	tests/test_attention_head_wanda_pruning_cpp.py	3035	CODE
LOW	tests/test_attention_head_wanda_pruning_cpp.py	3145	CODE
LOW	tests/test_attention_head_wanda_pruning_cpp.py	3497	CODE
LOW	tests/test_attention_head_wanda_pruning_cpp.py	3659	CODE
LOW	tests/test_attention_head_wanda_pruning_cpp.py	3799	CODE
LOW	tests/test_attention_head_wanda_pruning_cpp.py	3893	CODE
LOW	tests/test_attention_head_wanda_pruning_cpp.py	3973	CODE
LOW	tests/test_attention_head_wanda_pruning_cpp.py	4098	CODE
LOW	tests/test_attention_head_wanda_pruning_cpp.py	4165	CODE
LOW	tests/test_attention_head_wanda_pruning_cpp.py	4355	CODE
LOW	tests/test_dynamic_quantize_attention.py	73	CODE
LOW	tests/test_trt_batched_rotated_nms.py	78	CODE
LOW	tests/test_qwen_drive_perception_reconstruct.py	722	CODE
LOW	scripts/apple/train_block_step_coreml.py	51	CODE
LOW	scripts/apple/train_mlp_step_coreml.py	71	CODE
LOW	scripts/apple/export_llm_to_coreml.py	179	CODE
LOW	scripts/android/hexagon_resize2x.py	60	CODE
LOW	scripts/android/vision_models/rtdetr/model.py	94	CODE
LOW	scripts/axera/pulsar2_docker.py	240	CODE
LOW	scripts/axera/pulsar2_docker.py	581	CODE
LOW	scripts/axera/llm_layer_loop.py	506	CODE
LOW	scripts/axera/llm_layer_loop.py	680	CODE
LOW	scripts/axera/make_training_calib.py	63	CODE
LOW	scripts/axera/mcode.py	94	CODE
LOW	scripts/axera/finetune.py	92	CODE
LOW	scripts/axera/step_runner.py	2623	CODE
LOW	scripts/xdna/run_resnet_xdna.py	149	CODE
LOW	scripts/xdna/maxpool_design.py	24	CODE
LOW	scripts/xdna/quantized_add_relu_design.py	23	CODE
LOW	scripts/xdna/fused_bottleneck_design.py	28	CODE
LOW	scripts/xdna/layer_engine_host.py	47	CODE
LOW	scripts/xdna/parallel_projection_design.py	26	CODE
LOW	onnxsim/crown.py	1930	CODE
LOW	onnxsim/crown.py	2029	CODE
LOW	onnxsim/moequant.py	199	CODE
LOW	onnxsim/squeezellm.py	65	CODE
LOW	onnxsim/quark_weight_rounding.py	415	CODE
LOW	onnxsim/quark_weight_rounding.py	834	CODE
LOW	onnxsim/quark_weight_rounding.py	910	CODE
LOW	onnxsim/quark_fakequant_graph.py	490	CODE
68 more matches not shown…

Modern AI Meta-Vocabulary43 hits · 102 pts

Severity	File	Line	Snippet	Context
MEDIUM	tests/test_symexpr_kv_cache_consistency.py	3	# evaluator actually fold the shape scaffolding a real KV-cache transformer	COMMENT
MEDIUM	tests/test_symexpr_kv_cache_consistency.py	22	# * ``test_onnxsim_removes_kv_cache_shape_scaffolding``: the scaffolding	COMMENT
MEDIUM	tests/test_symexpr_kv_cache_consistency.py	112	# fix for both the correctness bug and the vacuous scaffolding	COMMENT
MEDIUM	tests/test_symexpr_kv_cache_consistency.py	223	# actually load-bearing: onnxsim must not make the scaffolding worse	COMMENT
MEDIUM	tests/test_graph_grad_templates.py	223	# and Identity (this test's own copy-out scaffolding, not something a	COMMENT
MEDIUM	tests/test_axera_conv_learn_128_widegap.py	79	# the scaffold region (everything else) is a separate, documented, partial result.	COMMENT
MEDIUM	tests/test_axera_conv_learn_128_widegap.py	153	# ~0.13% differed in the real campaign (the unresolved ~20% of scaffold bits,	COMMENT
MEDIUM	tests/test_fireredaudio.py	7	# generation (zero-shot/instruct TTS, semantic/acoustic speech editing, voice	COMMENT
MEDIUM⚡	tests/test_simple.py	371	# scaffolding standing. The SymExpr evaluator keeps the dim as `384*batch`,	STRING
MEDIUM⚡	tests/test_simple.py	374	# batch -- so the scaffolding collapses.	STRING
MEDIUM	tests/test_simple.py	301	# -- so the whole Shape -> Gather -> Concat scaffolding collapses away.	STRING
MEDIUM	tests/test_renesas_legalize.py	298	# --- legalize() / as_custom_rewriter() orchestration ---	STRING
MEDIUM	tests/test_axera_conv_learn_256to512.py	70	# scaffold region -- the weight bits themselves are byte-exact everywhere else.	COMMENT
MEDIUM	tests/test_axera_conv_learn_256to512.py	122	# that byte mapped) -- scaffold/const bytes may differ due to the second pass	COMMENT
MEDIUM	tests/test_moved_optimizer_passes.py	235	# scaffolding), so no batched MatMul remains.	STRING
MEDIUM	tests/test_moved_optimizer_passes.py	296	# scaffolding between two linear layers separated by an element-wise op)	STRING
MEDIUM	tests/test_moved_optimizer_passes.py	323	# The reshape scaffolding around the Relu is cancelled: at most the two	STRING
MEDIUM	tests/test_onnxnet_integration.py	132	# Node removal: the scaffolding Identity nodes carry no architectural	STRING
MEDIUM	tests/test_tensorrt_sparsity.py	99	# rank-agnostic reshape/slice scaffold must reproduce this exactly.	STRING
MEDIUM	tests/test_tensorrt_sparsity.py	126	# (rank-agnostic) scaffold rather than guessing 2-D. onnx.parser has no	STRING
MEDIUM⚡	tests/test_formal_verify_rewrite_trt_batched_nms.py	100	# Part 1: Z3 lemmas for the pass's own orchestration logic	COMMENT
MEDIUM⚡	tests/test_axera_conv_learn_wide.py	73	assert len(mismatched) > 0 # the scaffold region is real, not a fluke of this build	CODE
MEDIUM	tests/test_graph_grad.py	141	# so the allowlist check below can tell this scaffolding's own Identity	COMMENT
MEDIUM	tests/test_graph_grad.py	1462	# (GradAddTemplated's inlined body) is not scaffolding and must count.	COMMENT
MEDIUM	tests/test_axera_conv_weight_learn.py	86	# scaffolding this call does not touch.	COMMENT
MEDIUM	docs/axera-conv-compose-tiled-fix.md	1	# Does the tiled scaffold layout fix `docs/axera-conv-compose-real.md`'s composition bug? No -- but a different, real bu	COMMENT
MEDIUM	docs/axera-conv-weight-learn-256to512.md	1	# The widest Conv shapes in the ResNet18 step: weight codes hold, the scaffold does not	COMMENT
MEDIUM	docs/axera-conv-weight-learn-wide.md	55	## The requantisation/scaffold region: real, load-bearing, and only ~51% recovered	COMMENT
MEDIUM	docs/axera-conv-compose-real.md	101	## Finding 3: the scaffold does not patch correctly, even for the terminal Conv	COMMENT
MEDIUM	docs/axera-conv-weight-learn-128-and-widegap.md	1	# The 128/128 Conv shape, and does more data close the 256/256 scaffold gap?	COMMENT
MEDIUM	docs/axera-conv-weight-learn-128-and-widegap.md	46	### Device verification: the entire residual is the (separately known, unpatched) zero-point -- not the unresolved scaff	COMMENT
MEDIUM	docs/axera-conv-weight-learn-128-and-widegap.md	93	## Part B: does more data close 256/256's scaffold gap, or is it a plateau?	COMMENT
MEDIUM	docs/axera-conv-256to512-tiled-fix.md	1	# Fixing the two Conv shapes bit-permutation learning got wrong, with the tiled scaffold formula	COMMENT
MEDIUM	docs/axera-conv-256to512-tiled-fix.md	53	## Fact 2: Cin > 128 duplicates the scaffold	COMMENT
MEDIUM	docs/axera-conv-scaffold-arithmetic.md	1	# The wide-Cout Conv scaffold is a deterministic per-tile layout, not a bit-permutation-learned region	COMMENT
MEDIUM	docs/axera-matmul-scaffold-arithmetic.md	1	# Does MatMul/Gemm have a Conv-Wbt-style arithmetic scaffold? No -- and the existing evidence already shows why, precise	COMMENT
MEDIUM	scripts/axera/conv_256to512_tiled_fix.py	3	(PR #1784) arithmetic scaffold formula -- extended here with two real layout facts that	STRING
MEDIUM	scripts/axera/emitter.py	43	#: the weights; it is part of the shape's own scaffolding.	COMMENT
MEDIUM	scripts/axera/conv_compose_tiled_fix.py	25	#1783's Finding 2) and the scaffold's `M` now confirmed to match	STRING
MEDIUM	scripts/axera/conv_compose_tiled_fix.py	28	statistically indistinguishable from PR #1783's own buggy-scaffold	STRING
MEDIUM	scripts/axera/conv_compose_tiled_fix.py	124	"""Reproduces PR #1783's actual (buggy) scaffold M values: the formula	STRING
MEDIUM	scripts/regression/worker.py	65	# The parent directory name is orchestrate()'s `local` (model_id with	COMMENT
MEDIUM	onnxsim/tensorrt_sparsity.py	119	pass # rank unknown for every tensor -- every node falls back to the general scaffold	CODE

Self-Referential Comments174 hits · 96 pts

Severity	File	Line	Snippet	Context
MEDIUM	setup.py	126	# This file is generated by setup.py. DO NOT EDIT!	COMMENT
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	866	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	988	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	1096	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	1260	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	1586	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	1708	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	1830	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	1952	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	2118	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	2342	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	2492	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	2845	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	3151	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	3273	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	3603	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	3947	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	4041	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	4149	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	4229	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	4351	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	4445	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	4567	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	4793	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	4873	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	5109	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	5245	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	5471	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	6121	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	6403	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	6525	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	6605	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	6769	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	6947	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	7097	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	7261	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	7449	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	7529	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	7637	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	7731	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	7839	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	7961	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	8097	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	8177	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	8271	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	8365	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	8445	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	8525	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	8647	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	8783	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	8933	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	9083	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	9177	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	9271	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	9351	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	9418	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	9485	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	9607	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	9674	"""This method is deprecated. Please switch to GetRootAs."""	STRING
MEDIUM	…ter-flash/scripts/tflite_schema/schema_py_generated.py	9741	"""This method is deprecated. Please switch to GetRootAs."""	STRING
114 more matches not shown…

Modern Structural Boilerplate65 hits · 65 pts

Severity	File	Line	Snippet	Context
LOW	scripts/rknn3/llm_export.py	41	def _update_config(config, attr_names, value) -> None:	CODE
LOW	scripts/axera/reshape_dma_emit.py	73	def _set_dims(value_info, shape) -> None:	CODE
LOW	scripts/axera/reshape_emit.py	143	def _set_dims(value_info, shape: Shape) -> None:	CODE
LOW	scripts/axera/memory_emit.py	120	def _set_dims(value_info, shape: list[int]) -> None:	CODE
LOW	scripts/axera/transpose_emit.py	67	def _set_dims(value_info, shape: list[int]) -> None:	CODE
LOW	scripts/axera/elementwise_two_input_tile_predict.py	103	__all__ = ["predict_params"]	CODE
LOW	scripts/axera/reshape_record_emit.py	406	def _set_dims(value_info, shape: Sequence[int]) -> None:	CODE
LOW	scripts/xdna/__init__.py	60	__all__ = [	CODE
LOW	onnxsim/quark_boundary_qdq.py	395	__all__ = ["insert_boundary_quant_nodes"]	CODE
LOW	onnxsim/qat_interop.py	168	__all__ = [	CODE
LOW	onnxsim/qat_interop.py	1436	def _set_metadata(model: onnx.ModelProto, key: str, value: str) -> None:	CODE
LOW	onnxsim/_onnx_compat.py	19	__all__ = ["INT4", "UINT4"]	CODE
LOW	onnxsim/to_torch.py	52	__all__ = ["onnx_to_torch", "onnx_state_dict"]	CODE
LOW	onnxsim/quark_weight_rounding.py	545	def _set_attr(node: onnx.NodeProto, name: str, value: Optional[int]) -> None:	CODE
LOW	onnxsim/quark_weight_rounding.py	1040	__all__ = [	CODE
LOW	onnxsim/quark_fakequant_graph.py	694	__all__ = [	CODE
LOW	onnxsim/webgpu_kernel_metadata.py	310	def _set_node_metadata(node: onnx.NodeProto, key: str, value: str) -> None:	CODE
LOW	onnxsim/quark_fakequant_eval.py	268	__all__ = ["run_fake_quantized"]	CODE
LOW	onnxsim/activation_sensitivity.py	75	__all__ = [	CODE
LOW	onnxsim/webgpu_kernel_tuning.py	80	__all__ = ["KernelCandidates", "generate_kernel_candidates"]	CODE
LOW	onnxsim/quark_quarot.py	302	__all__ = ["make_rotation", "rotate_model"]	CODE
LOW	onnxsim/quark_compat.py	3265	__all__ = [	CODE
LOW	onnxsim/quark_finetune.py	1937	__all__ = [	CODE
LOW	onnxsim/einsum_decompose.py	36	__all__ = ["decompose_einsum"]	CODE
LOW	onnxsim/quark_fusions.py	63	__all__ = [	CODE
LOW	onnxsim/quark_fusions.py	97	def _set_nodes(graph: onnx.GraphProto, nodes: Sequence[onnx.NodeProto]) -> None:	CODE
LOW	onnxsim/rustnn_runtime.py	50	__all__ = [	CODE
LOW	onnxsim/bitwise_decompose.py	53	__all__ = ["decompose_bitwise_xor"]	CODE
LOW	onnxsim/quark_dynamic.py	282	__all__ = ["quantize_dynamic_integer"]	CODE
LOW	onnxsim/webnn_tinygrad_tuning.py	52	__all__ = [	CODE
LOW	onnxsim/webnn_tinygrad_tuning.py	372	def _set_node_metadata(node: onnx.NodeProto, key: str, value: str) -> None:	CODE
LOW	onnxsim/memory_planning.py	15	__all__ = [	CODE
LOW	onnxsim/profile_plot.py	37	__all__ = ["load_node_counts", "plot_node_reduction"]	CODE
LOW	onnxsim/__init__.py	497	__all__ = [	CODE
LOW	onnxsim/quark_preset_graphs.py	908	__all__ = [	CODE
LOW⚡	onnxsim/stable_relu_prune.py	64	__all__ = ["LayerReport", "PruneReport", "analyze", "apply", "prune_units"]	CODE
LOW⚡	onnxsim/stable_relu_prune.py	746	def _set_init(g: onnx.GraphProto, name: str, arr: np.ndarray) -> None:	CODE
LOW	onnxsim/quark_calibration.py	28	__all__ = ["QuarkHistogram", "lwp_select"]	CODE
LOW	onnxsim/quark_block_formats.py	279	__all__ = ["bf16_round", "bfp16", "bfp_prime", "fp16_round", "mx"]	CODE
LOW	onnxsim/quark_tools_extra.py	142	def _set_opset(model: onnx.ModelProto, domain: str, version: int) -> None:	CODE
LOW	onnxsim/quark_tools_extra.py	213	def set_nhwc(vi: onnx.ValueInfoProto, chw: Tuple[int, int, int]) -> None:	CODE
LOW	onnxsim/quark_tools_extra.py	1398	__all__ = [	CODE
LOW	onnxsim/ranges.py	56	def set_range(model: onnx.ModelProto, name: str, lo: Bound, hi: Bound) -> None:	CODE
LOW	onnxsim/quant_verify.py	1259	__all__ = [	CODE
LOW	onnxsim/quark_cle.py	215	__all__ = ["equalize_linear_layers"]	CODE
LOW	onnxsim/quark_smoothquant.py	142	__all__ = ["apply_smooth_quant_config", "smooth_quant"]	CODE
LOW	onnxsim/full_qdq.py	53	__all__ = ["quantize_full_qdq", "quantized_io", "sampling_coordinate_tensors"]	CODE
LOW	onnxsim/full_qdq.py	917	def set_qp(x: str) -> None:	CODE
LOW	onnxsim/shape_cost.py	1239	def set_type(vi: onnx.ValueInfoProto, n: str, et: int) -> None:	CODE
LOW	onnxsim/quark_equalization.py	650	__all__ = [	CODE
LOW	onnxsim/xnnpack_codegen.py	3	__all__ = [	CODE
LOW	onnxsim/model_info.py	44	__all__ = [	CODE
LOW	onnxsim/model_info.py	1116	def _set_metadata(proto, key: str, value: str) -> None:	CODE
LOW	onnxsim/quark_npu.py	206	def set_pos(self, node: onnx.NodeProto, new_pos: int) -> None:	CODE
LOW	onnxsim/gridsample_plan.py	32	__all__ = ["grid_sample_plan", "GridSampleTap"]	CODE
LOW	onnxsim/tinygrad_uop_export.py	112	__all__ = [	CODE
LOW	onnxsim/quark_hadamard.py	276	__all__ = [	CODE
LOW	onnxsim/webgpu_tinygrad_codegen.py	128	__all__ = [	CODE
LOW	onnxsim/onnx_simplifier.py	8755	def _set_metadata(model: onnx.ModelProto, key: str, value: str) -> None:	CODE
LOW	onnxsim/quark_mixing.py	309	def _set_init(self, name: str, value: np.ndarray) -> None:	CODE
5 more matches not shown…

Verbosity Indicators4 hits · 8 pts

Severity	File	Line	Snippet	Context
LOW⚡	…est_formal_verify_fuse_consecutive_reduce_unsqueeze.py	152	# Step 1: ONNX's own definition of keepdims -- keepdims=1 computes the	COMMENT
LOW⚡	…est_formal_verify_fuse_consecutive_reduce_unsqueeze.py	161	# Step 2: Unsqueeze(keepdims=0 result, axes) read at a full-rank index.	COMMENT
LOW	onnxsim/passes/rewrite_gathernd_to_gather.h	428	// Step 1: flatten data's leading b+k axes into one.	COMMENT
LOW	onnxsim/passes/rewrite_gathernd_to_gather.h	440	// Step 2: per-column extraction, negative-index normalization, and	COMMENT

Structural Annotation Overuse4 hits · 8 pts

Severity	File	Line	Snippet	Context
LOW⚡	…est_formal_verify_fuse_consecutive_reduce_unsqueeze.py	152	# Step 1: ONNX's own definition of keepdims -- keepdims=1 computes the	COMMENT
LOW⚡	…est_formal_verify_fuse_consecutive_reduce_unsqueeze.py	161	# Step 2: Unsqueeze(keepdims=0 result, axes) read at a full-rank index.	COMMENT
LOW	onnxsim/passes/rewrite_gathernd_to_gather.h	428	// Step 1: flatten data's leading b+k axes into one.	COMMENT
LOW	onnxsim/passes/rewrite_gathernd_to_gather.h	440	// Step 2: per-column extraction, negative-index normalization, and	COMMENT

Slop Phrases2 hits · 4 pts

Severity	File	Line	Snippet	Context
MEDIUM⚡	tools/onnx-k210-flash/web/ncc/ncc.js	1	var create_ncc=(()=>{var _scriptName=globalThis.document?.currentScript?.src;return async function(moduleArg={}){var mod	CODE
LOW⚡	tools/onnx-k210-flash/web/ncc/ncc.js	1	var create_ncc=(()=>{var _scriptName=globalThis.document?.currentScript?.src;return async function(moduleArg={}){var mod	CODE

Example Usage Blocks2 hits · 3 pts

Severity	File	Line	Snippet	Context
LOW	build_wasm_pyodide.sh	66	# Usage:	COMMENT
LOW	scripts/test_rust_package_standalone.sh	14	# Usage:	COMMENT

Redundant / Tautological Comments2 hits · 3 pts

Severity	File	Line	Snippet	Context
LOW	build_wasm.sh	4	# Check if emcmake is available	COMMENT
LOW	onnxsim/full_qdq.py	55	# Output values are a subset / convex combination of the data input's values	COMMENT

Maintenance and Governance Review

Risk Review · v3

What These Metrics Mean

Pattern-Signal History

Severity Breakdown

Directory Score Breakdown

Pattern Findings