Repository Analysis

vllm-project/vllm-omni

A framework for efficient model inference with omni-modality models

26.4 Moderate AI signal View on GitHub

Analysis Overview

This report presents the forensic synthetic code analysis of vllm-project/vllm-omni, a Python project with 6,436 GitHub stars. SynthScan v2.0 examined 944,682 lines of code across 3201 source files, recording 19064 pattern matches distributed across 24 syntactic categories. The overall adjusted score of 26.4 places this repository in the Moderate AI signal band.

The scanner applied 160+ deterministic lexical heuristics, multi-line block detectors, abstract syntax tree depth profilers, and a cross-file Jaccard similarity matrix to construct a statistically normalised synthetic code estimate. All matches are individually weighted by severity coefficient and contextual multiplier before summation, and the resulting headline score is temporally discounted to account for the repository's development history relative to the commercial emergence of large language model coding tooling (November 2022 onward).

26.4
Adjusted Score
26.4
Raw Score
100%
Time Factor
2026-08-29
Last Push
6.4K
Stars
Python
Language
944.7K
Lines of Code
3.2K
Files
19.1K
Pattern Hits
2026-08-29
Scan Date
0.10
HC Hit Rate

What These Metrics Mean

Adjusted Score
Primary synthetic code indicator. Raw score normalised per 1,000 lines of code and multiplied by the temporal discount factor. This is the definitive comparative metric — use it to rank repositories by AI authorship density.
Raw Score
The unmodified sum of all severity-weighted, context-multiplied pattern match scores before temporal discounting. Reflects the absolute signal strength independent of when the repository was last active.
Time Factor
The temporal discount multiplier (0–100%) applied to the raw score. Repositories last updated before ChatGPT's launch (Nov 2022) receive a 5% factor. Full signal is only assigned to repositories active in the post-adoption era (Jan 2024+).
Pattern Hits
Total count of individual pattern matches across all files and categories. A high hit count with a low score may indicate a very large codebase with isolated AI snippets; a low count with a high score indicates dense, concentrated AI signatures.
HC Hit Rate
High+Critical pattern hits per file, averaged across the repository. This orthogonal signal catches repositories where a few files are densely packed with high-severity AI tells — a strong indicator even when the normalised score appears moderate due to codebase size.
Lines of Code / Files
Total lines and files analysed. The scanner examines 94 file extensions. These denominators are used to normalise the score, enabling fair comparison between repositories of vastly different sizes.

Score History

This chart maps the temporal evolution of the adjusted synthetic code score across successive scan runs. An upward trajectory indicates ongoing incorporation of AI-generated code or expanding LLM-assisted scaffolding; a stable or declining trajectory may reflect active human refactoring, code removal, or the adoption of stricter authorship policies. The dashed secondary line (right axis) independently tracks total raw pattern hit count, which can diverge from the normalised score when codebase size changes significantly between scans.

Severity Breakdown

Classifies detected patterns by their diagnostic confidence and structural impact. CRITICAL patterns (coefficient 10) represent definitive synthetic signatures — hallucinated imports, explicit LLM attribution metadata — virtually never produced by human authors. HIGH (5) indicates strong structural tells such as cross-file repetition or cross-linguistic idioms. MEDIUM (2) covers recognisable conversational padding and AI-specific vocabulary. LOW (1) captures subtle indicators like tautological comments and generic boilerplate that require density to carry independent signal.

CRITICAL 6HIGH 304MEDIUM 2323LOW 16431

Directory Score Breakdown

This horizontal bar chart decomposes the repository's raw synthetic code score by top-level directory, allowing you to pinpoint precisely which modules or components carry the highest AI authorship density. Directories with disproportionately high scores relative to their size warrant targeted manual review: concentrated AI signatures often trace back to mass-generated configuration layers, auto-ported test suites, LLM-scaffolded boilerplate classes, or entire subsystems authored under heavy copilot assistance. Use this view to prioritise your human code-review effort.

Pattern Findings

The scanner identified 19064 distinct pattern matches across 24 syntactic categories. Each entry below represents a discrete location in the source code where the engine recorded a statistically significant AI authorship indicator. 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 diagnostic confidence level (CRITICAL / HIGH / MEDIUM / LOW). The Context column identifies whether the match occurred inside executable code, an inline comment, or a string literal — comment-context matches receive a ×1.5 weight because LLMs systematically over-annotate. The ⚡ bolt icon marks clustered matches: three or more patterns within a 10-line window, each receiving an additional ×1.5 density multiplier as dense clusters constitute far stronger evidence of synthetic authorship than isolated hits.

Hyper-Verbose Identifiers10775 hits · 10177 pts
SeverityFileLineSnippetContext
LOWcollect_env.py135def run_and_parse_first_match(run_lambda, command, regex):CODE
LOWcollect_env.py146def run_and_return_first_line(run_lambda, command):CODE
LOWcollect_env.py179def get_nvidia_driver_version(run_lambda):CODE
LOWcollect_env.py318def summarize_vllm_build_flags():CODE
LOWcollect_env.py473def get_cachingallocator_config():CODE
LOWcollect_env.py478def get_cuda_module_loading_config():CODE
LOWtools/run_jobs_common.sh526def timeout_kill_after_seconds() -> int:CODE
LOWtools/run_jobs_common.sh535def prepend_timeout_to_pytest(pytest_line: str, timeout_min: int | None) -> str:CODE
LOWtools/run_jobs_common.sh547def _write_job_timeouts_manifest(jobs_dir: Path, job_timeouts: dict[str, int]) -> None:CODE
LOWtools/nightly/buildkite_testcase_statistics.py86def extract_pytest_targets_from_line(line: str) -> list[str]:CODE
LOWtools/nightly/buildkite_testcase_statistics.py138def get_pytest_targets_from_step(step: dict) -> list[tuple[str, str]]:CODE
LOWtools/nightly/buildkite_testcase_statistics.py158def _strip_ignore_clauses_from_args_fragment(fragment: str) -> str:CODE
LOWtools/nightly/buildkite_testcase_statistics.py163def _strip_test_config_file_clauses_from_args_fragment(fragment: str) -> str:CODE
LOWtools/nightly/buildkite_testcase_statistics.py168def _pytest_collect_sidecar_args(raw_line: str) -> list[str]:CODE
LOWtools/nightly/buildkite_testcase_statistics.py208def _parse_extra_args_from_line(raw_line: str) -> list[str]:CODE
LOWtools/nightly/buildkite_testcase_statistics.py213def _parse_collect_only_stdout(stdout: str, *, raise_on_empty: bool = True, stderr: str = "") -> list[str]:CODE
LOWtools/nightly/buildkite_testcase_statistics.py268def _format_pytest_collect_failure(CODE
LOWtools/nightly/buildkite_testcase_statistics.py383def get_docstring_for_node_id(repo_root: Path, node_id: str) -> str:CODE
LOWtools/nightly/buildkite_testcase_statistics.py426def get_test_file_for_node_id(node_id: str) -> str:CODE
LOWtools/nightly/generate_nightly_perf_excel.py219def _ensure_omni_summary_columns(summary_columns: list[str]) -> list[str]:CODE
LOWtools/nightly/generate_nightly_perf_excel.py262def _default_diffusion_input_dir(input_dir: str) -> str:CODE
LOWtools/nightly/generate_nightly_perf_excel.py456def _parse_diffusion_result_from_filename(filename: str) -> dict[str, Any]:CODE
LOWtools/nightly/generate_nightly_perf_excel.py516def _collect_diffusion_records(diffusion_input_dir: str) -> list[dict[str, Any]]:CODE
LOWtools/nightly/generate_nightly_perf_excel.py545def _process_diffusion_record(record: dict[str, Any]) -> dict[str, Any]:CODE
LOWtools/nightly/generate_nightly_perf_excel.py561def _apply_build_metadata_to_latest_only(CODE
LOWtools/nightly/generate_nightly_perf_excel.py599def _sort_records_for_summary(records: list[dict[str, Any]]) -> list[dict[str, Any]]:CODE
LOWtools/nightly/generate_nightly_perf_excel.py608def _sort_diffusion_records_for_summary(records: list[dict[str, Any]]) -> list[dict[str, Any]]:CODE
LOWtools/nightly/generate_nightly_perf_html.py52def _default_diffusion_input_dir(input_dir: str) -> str:CODE
LOWtools/nightly/generate_nightly_perf_html.py173def _parse_diffusion_from_filename(filename: str) -> dict[str, Any]:CODE
LOWtools/nightly/generate_nightly_perf_html.py192def _iter_diffusion_json_records(input_dir: str) -> Iterable[dict[str, Any]]:CODE
LOWtools/nightly/generate_nightly_perf_html.py248def _collect_diffusion_records(input_dir: str) -> list[dict[str, Any]]:CODE
LOWtools/nightly/run_nightly_jobs.sh565def timeout_kill_after_seconds() -> int:CODE
LOWtools/nightly/run_nightly_jobs.sh574def prepend_timeout_to_pytest(pytest_line: str, timeout_min: int | None) -> str:CODE
LOWtools/nightly/run_nightly_jobs.sh586def _write_job_timeouts_manifest(jobs_dir: Path, job_timeouts: dict[str, int]) -> None:CODE
LOWtools/nightly/run_nightly_jobs.sh660def local_test_files_by_filename(substr: str) -> list[str]:CODE
LOWtools/nightly/run_nightly_jobs.sh694def perf_json_matches_model_type(json_basename: str, model_types: list[str]) -> bool:CODE
LOWtools/pre_commit/check_test_marks.py80def get_files_missing_markers(CODE
LOWtools/wan22/assemble_wan22_i2v_diffusers.py62def _load_shard_files_from_index(index_file: Path, role: str) -> tuple[Path, ...]:CODE
LOWtests/metrics/test_stats.py18def test_orchestrator_aggregator_builds_summary() -> None:CODE
LOWtests/metrics/test_stats.py94def test_build_and_log_summary_multiple_requests() -> None:CODE
LOWtests/metrics/test_prometheus.py53 def test_all_metric_families_present(self, scrape_output: str) -> None:CODE
LOWtests/metrics/test_prometheus.py102 def test_pipeline_metrics_carry_model_name(self, scrape_output: str) -> None:CODE
LOWtests/metrics/test_prometheus.py107 def test_no_legacy_engine_label(self, scrape_output: str) -> None:CODE
LOWtests/metrics/test_prometheus.py110 def test_no_legacy_seconds_or_ms_families(self, scrape_output: str) -> None:CODE
LOWtests/metrics/test_prometheus.py120 def test_omni_metrics_in_default_registry(self, scrape_output: str) -> None:CODE
LOWtests/metrics/test_prometheus.py124 def test_process_metrics_in_default_registry(self, scrape_output: str) -> None:CODE
LOWtests/metrics/test_prometheus.py147 def test_running_and_waiting_zero_after_request_completes(self, registry: CollectorRegistry) -> None:CODE
LOWtests/metrics/test_prometheus.py176 def test_gauges_reflect_remaining_requests_after_one_completes(self, registry: CollectorRegistry) -> None:CODE
LOWtests/metrics/test_metrics_utils.py25def test_count_audio_chunk_frames(audio_chunk: object, expected: int) -> None:CODE
LOWtests/metrics/test_metrics_utils.py29def test_count_audio_frames_sums_audio_chunks() -> None:CODE
LOWtests/metrics/test_metrics_utils.py40def test_count_audio_frames_supports_model_outputs() -> None:CODE
LOWtests/metrics/test_metrics_utils.py52def test_count_audio_frames_returns_zero_without_audio(mm_out: dict[str, object]) -> None:CODE
LOWtests/metrics/test_metrics_utils.py72def test_count_image_pixels_sums_nested_values() -> None:CODE
LOWtests/metrics/test_metrics_utils.py84def test_count_tokens_from_outputs() -> None:CODE
LOWtests/metrics/test_transfer.py42 def test_all_four_families_present(self, tx: OmniTransferMetrics) -> None:CODE
LOWtests/metrics/test_transfer.py52 def test_legacy_ms_family_names_dropped(self, tx: OmniTransferMetrics) -> None:CODE
LOWtests/metrics/test_transfer.py68 def test_size_observed_with_correct_labels(self, tx: OmniTransferMetrics) -> None:CODE
LOWtests/metrics/test_transfer.py124 def test_multiple_edges_produce_independent_series(self, tx: OmniTransferMetrics) -> None:CODE
LOWtests/metrics/test_transfer.py157 def test_size_uses_bytes_buckets(self, tx: OmniTransferMetrics) -> None:CODE
LOWtests/metrics/test_transfer.py165 def test_time_families_use_seconds_fast_buckets(self, tx: OmniTransferMetrics) -> None:CODE
10715 more matches not shown…
Decorative Section Separators2097 hits · 7110 pts
SeverityFileLineSnippetContext
MEDIUMtools/run_jobs_common.sh26# ---------------------------------------------------------------------------COMMENT
MEDIUMtools/run_jobs_common.sh28# ---------------------------------------------------------------------------COMMENT
MEDIUMtools/run_jobs_common.sh179# ---------------------------------------------------------------------------COMMENT
MEDIUMtools/run_jobs_common.sh186# ---------------------------------------------------------------------------COMMENT
MEDIUMtests/metrics/test_transfer.py36# ---------------------------------------------------------------------------COMMENT
MEDIUMtests/metrics/test_transfer.py38# ---------------------------------------------------------------------------COMMENT
MEDIUMtests/metrics/test_transfer.py62# ---------------------------------------------------------------------------COMMENT
MEDIUMtests/metrics/test_transfer.py64# ---------------------------------------------------------------------------COMMENT
MEDIUMtests/metrics/test_transfer.py118# ---------------------------------------------------------------------------COMMENT
MEDIUMtests/metrics/test_transfer.py120# ---------------------------------------------------------------------------COMMENT
MEDIUMtests/metrics/test_transfer.py151# ---------------------------------------------------------------------------COMMENT
MEDIUMtests/metrics/test_transfer.py153# ---------------------------------------------------------------------------COMMENT
MEDIUMtests/metrics/test_transfer.py173# ---------------------------------------------------------------------------COMMENT
MEDIUMtests/metrics/test_transfer.py175# ---------------------------------------------------------------------------COMMENT
MEDIUMtests/metrics/test_modality.py183# ---------------------------------------------------------------------------COMMENT
MEDIUMtests/metrics/test_modality.py185# ---------------------------------------------------------------------------COMMENT
MEDIUMtests/metrics/test_modality.py227# ---------------------------------------------------------------------------COMMENT
MEDIUMtests/metrics/test_modality.py229# ---------------------------------------------------------------------------COMMENT
MEDIUMtests/metrics/test_modality.py38# ---------------------------------------------------------------------------COMMENT
MEDIUMtests/metrics/test_modality.py40# ---------------------------------------------------------------------------COMMENT
MEDIUMtests/metrics/test_modality.py88# ---------------------------------------------------------------------------COMMENT
MEDIUMtests/metrics/test_modality.py90# ---------------------------------------------------------------------------COMMENT
MEDIUMtests/metrics/test_stat_logger.py70# ---------------------------------------------------------------------------COMMENT
MEDIUMtests/metrics/test_stat_logger.py72# ---------------------------------------------------------------------------COMMENT
MEDIUMtests/metrics/test_stat_logger.py92# ---------------------------------------------------------------------------COMMENT
MEDIUMtests/metrics/test_stat_logger.py94# ---------------------------------------------------------------------------COMMENT
MEDIUMtests/metrics/test_stat_logger.py219# ---------------------------------------------------------------------------COMMENT
MEDIUMtests/metrics/test_stat_logger.py221# ---------------------------------------------------------------------------COMMENT
MEDIUMtests/metrics/test_stat_logger.py255# ---------------------------------------------------------------------------COMMENT
MEDIUMtests/metrics/test_stat_logger.py257# ---------------------------------------------------------------------------COMMENT
MEDIUMtests/metrics/test_stat_logger.py278# ---------------------------------------------------------------------------COMMENT
MEDIUMtests/metrics/test_stat_logger.py280# ---------------------------------------------------------------------------COMMENT
MEDIUMtests/metrics/test_stat_logger.py305# ---------------------------------------------------------------------------COMMENT
MEDIUMtests/metrics/test_stat_logger.py309# ---------------------------------------------------------------------------COMMENT
MEDIUMtests/metrics/test_stat_logger.py371# ---------------------------------------------------------------------------COMMENT
MEDIUMtests/metrics/test_stat_logger.py378# ---------------------------------------------------------------------------COMMENT
MEDIUMtests/metrics/test_stat_logger.py404# ---------------------------------------------------------------------------COMMENT
MEDIUMtests/metrics/test_stat_logger.py411# ---------------------------------------------------------------------------COMMENT
MEDIUMtests/metrics/test_stat_logger.py35# ---------------------------------------------------------------------------COMMENT
MEDIUMtests/metrics/test_stat_logger.py37# ---------------------------------------------------------------------------COMMENT
MEDIUMtests/metrics/test_emit_calls.py28# ---------------------------------------------------------------------------COMMENT
MEDIUMtests/metrics/test_emit_calls.py30# ---------------------------------------------------------------------------COMMENT
MEDIUMtests/metrics/test_emit_calls.py365# ---------------------------------------------------------------------------COMMENT
MEDIUMtests/metrics/test_emit_calls.py368# ---------------------------------------------------------------------------COMMENT
MEDIUMtests/metrics/test_emit_calls.py417# ---------------------------------------------------------------------------COMMENT
MEDIUMtests/metrics/test_emit_calls.py419# ---------------------------------------------------------------------------COMMENT
MEDIUM…/omni_connectors/test_mooncake_transfer_engine_rdma.py443# ---------------------------------------------------------------------------COMMENT
MEDIUM…/omni_connectors/test_mooncake_transfer_engine_rdma.py445# ---------------------------------------------------------------------------COMMENT
MEDIUM…/omni_connectors/test_mooncake_transfer_engine_rdma.py35# ---------------------------------------------------------------------------COMMENT
MEDIUM…/omni_connectors/test_mooncake_transfer_engine_rdma.py37# ---------------------------------------------------------------------------COMMENT
MEDIUM…/omni_connectors/test_mooncake_transfer_engine_rdma.py190# ---------------------------------------------------------------------------COMMENT
MEDIUM…/omni_connectors/test_mooncake_transfer_engine_rdma.py192# ---------------------------------------------------------------------------COMMENT
MEDIUM…/omni_connectors/test_mooncake_transfer_engine_rdma.py250# ---------------------------------------------------------------------------COMMENT
MEDIUM…/omni_connectors/test_mooncake_transfer_engine_rdma.py252# ---------------------------------------------------------------------------COMMENT
MEDIUM…/omni_connectors/test_mooncake_transfer_engine_rdma.py471# ---------------------------------------------------------------------------COMMENT
MEDIUM…/omni_connectors/test_mooncake_transfer_engine_rdma.py473# ---------------------------------------------------------------------------COMMENT
MEDIUM…/omni_connectors/test_mooncake_transfer_engine_rdma.py519# ---------------------------------------------------------------------------COMMENT
MEDIUM…/omni_connectors/test_mooncake_transfer_engine_rdma.py521# ---------------------------------------------------------------------------COMMENT
MEDIUM…ributed/omni_connectors/test_chunk_transfer_adapter.py1785# ---------------------------------------------------------------COMMENT
MEDIUM…ributed/omni_connectors/test_chunk_transfer_adapter.py1787# ---------------------------------------------------------------COMMENT
2037 more matches not shown…
Unused Imports1962 hits · 1850 pts
SeverityFileLineSnippetContext
LOWtools/configure_stage_memory.py19CODE
LOWtools/nightly/send_nightly_email.py9CODE
LOWtools/nightly/buildkite_testcase_statistics.py18CODE
LOWtools/nightly/generate_nightly_perf_excel.py7CODE
LOWtools/nightly/generate_nightly_perf_html.py7CODE
LOWtools/pre_commit/mypy.py18CODE
LOWtools/pre_commit/check_tts_adapter.py20CODE
LOWtools/pre_commit/check_buildkite.py12CODE
LOWtools/wan22/assemble_wan22_i2v_diffusers.py22CODE
LOWtools/baselines/omni_hash_baseline.py44CODE
LOWtests/conftest.py13CODE
LOWtests/metrics/test_stats.py1CODE
LOWtests/metrics/test_prometheus.py1CODE
LOWtests/metrics/test_metrics_utils.py1CODE
LOWtests/metrics/test_transfer.py1CODE
LOWtests/metrics/test_modality.py4CODE
LOWtests/metrics/test_stat_logger.py1CODE
LOWtests/metrics/test_emit_calls.py9CODE
LOWtests/platforms/npu/test_310p_patches.py9CODE
LOWtests/platforms/npu/test_step_audio2_token2wav.py5CODE
LOW…/platforms/npu/test_diffusion_attn_backend_selector.py15CODE
LOWtests/platforms/npu/quant/test_kv_quant_npu.py10CODE
LOWtests/platforms/npu/quant/test_kv_quant_npu.py55CODE
LOWtests/tools/test_check_tts_adapter.py9CODE
LOW…e/sched/test_omni_ar_scheduler_free_request_cleanup.py8CODE
LOW…e/sched/test_omni_ar_scheduler_free_request_cleanup.py17CODE
LOWtests/core/sched/test_omni_ar_scheduler_stale_drain.py27CODE
LOWtests/core/sched/test_omni_ar_scheduler_stale_drain.py37CODE
LOW…sched/test_omni_scheduler_input_coordinator_cleanup.py1CODE
LOW…s/core/sched/test_omni_sched_prefill_stats_finalize.py13CODE
LOW…s/core/sched/test_omni_sched_prefill_stats_finalize.py22CODE
LOW…sched/test_omni_generation_scheduler_update_session.py9CODE
LOW…sched/test_omni_generation_scheduler_update_session.py20CODE
LOWtests/core/sched/test_omni_scheduling_coordinator.py12CODE
LOWtests/core/sched/test_omni_scheduler_mixin_timeouts.py17CODE
LOWtests/core/sched/test_omni_sched_deferred_free_fence.py24CODE
LOWtests/core/sched/test_omni_sched_deferred_free_fence.py33CODE
LOWtests/core/sched/test_sched_package_imports.py4CODE
LOW…ched/test_omni_ar_scheduler_unified_decode_deferral.py1CODE
LOWtests/core/sched/test_omni_ar_scheduler_streaming.py6CODE
LOWtests/core/sched/test_omni_ar_scheduler_streaming.py16CODE
LOWtests/core/sched/test_omni_sched_ec_request_finish.py15CODE
LOWtests/core/sched/test_omni_sched_ec_request_finish.py24CODE
LOWtests/core/sched/test_omni_ar_scheduler_logprobs.py3CODE
LOW…ore/sched/test_omni_scheduler_finish_requests_purge.py17CODE
LOWtests/host_weight_runtime/test_contracts.py5CODE
LOWtests/host_weight_runtime/test_runtime_resolution.py5CODE
LOWtests/host_weight_runtime/test_filesystem_store.py5CODE
LOWtests/config/test_omni_config.py5CODE
LOWtests/config/test_config_import_cycle.py5CODE
LOWtests/config/test_forced_aligner_injection.py3CODE
LOWtests/config/composable_parallel/test_translator.py5CODE
LOW…config/composable_parallel/test_factory_integration.py5CODE
LOW…sts/config/composable_parallel/test_strategy_loader.py5CODE
LOWtests/config/composable_parallel/test_spec_types.py5CODE
LOWtests/config/composable_parallel/test_apply.py5CODE
LOWtests/model_executor/test_cfg_pairing.py11CODE
LOWtests/model_executor/helpers.py27CODE
LOWtests/model_executor/models/registry.py10CODE
LOW…sts/model_executor/models/test_encoder_quant_config.py18CODE
1902 more matches not shown…
Excessive Try-Catch Wrapping1322 hits · 1374 pts
SeverityFileLineSnippetContext
LOWcollect_env.py499 except Exception:CODE
MEDIUMcollect_env.py293def get_vllm_omni_version(run_lambda):CODE
LOWsetup.py39 except Exception as e:CODE
LOWsetup.py94 except Exception:CODE
LOWsetup.py103 except Exception:CODE
LOWsetup.py112 except Exception:CODE
LOWsetup.py150 except Exception as e:CODE
LOWtools/configure_stage_memory.py39 except Exception:CODE
LOWtools/configure_stage_memory.py65 except Exception as e:CODE
MEDIUMtools/configure_stage_memory.py317 print(f"Error: {config_path} not found", file=sys.stderr)STRING
LOWtools/nightly/send_nightly_email.py277 except Exception as e:CODE
LOWtools/nightly/buildkite_testcase_statistics.py402 except Exception:CODE
LOWtools/pre_commit/check_buildkite.py109 except Exception as exc: # noqa: BLE001 - surface expander errors as hook failuresCODE
LOWtools/wan22/assemble_wan22_i2v_diffusers.py66 except Exception as exc:CODE
LOWtools/baselines/omni_hash_baseline.py90 except Exception:CODE
LOWtools/baselines/omni_hash_baseline.py99 except Exception:CODE
MEDIUMtools/baselines/omni_hash_baseline.py87def _git_commit() -> str:CODE
LOWtests/core/sched/test_generation_scheduler_restore.py60 except Exception:CODE
LOWtests/core/sched/test_generation_scheduler_restore.py83 except Exception:CODE
MEDIUM…/omni_connectors/test_mooncake_transfer_engine_rdma.py402def worker(rid, data):CODE
MEDIUM…/omni_connectors/test_mooncake_transfer_engine_rdma.py613def producer_worker(rid, tensor):CODE
LOW…/omni_connectors/test_mooncake_transfer_engine_rdma.py75 except Exception:CODE
LOW…/omni_connectors/test_mooncake_transfer_engine_rdma.py124 except Exception as exc:CODE
LOW…/omni_connectors/test_mooncake_transfer_engine_rdma.py157 except Exception as exc:CODE
LOW…/omni_connectors/test_mooncake_transfer_engine_rdma.py408 except Exception as e:CODE
LOW…/omni_connectors/test_mooncake_transfer_engine_rdma.py562 except Exception as e:CODE
LOW…/omni_connectors/test_mooncake_transfer_engine_rdma.py624 except Exception as e:CODE
LOW…/omni_connectors/test_mooncake_transfer_engine_rdma.py657 except Exception as e:CODE
LOW…ributed/omni_connectors/test_chunk_transfer_adapter.py1173 except Exception as e:CODE
LOW…ributed/omni_connectors/test_chunk_transfer_adapter.py1182 except Exception as e:CODE
LOW…mni_connectors/test_mooncake_transfer_engine_buffer.py78 except Exception as e:CODE
MEDIUM…mni_connectors/test_mooncake_transfer_engine_buffer.py72def worker(worker_id):CODE
LOW…ributed/omni_connectors/test_omni_connector_configs.py89 except Exception as e:CODE
LOWtests/model_executor/helpers.py54 except Exception:CODE
LOW…s/model_executor/models/glm_image/test_glm_image_ar.py150 except Exception:CODE
LOW…l_executor/models/qwen3_tts/test_cuda_graph_decoder.py33except Exception:CODE
LOWtests/utils/test_speaker_cache.py127 except Exception as e:CODE
MEDIUMtests/utils/test_speaker_cache.py122def worker(tid):CODE
LOWtests/diffusion/test_diffusion_config_fields.py18except Exception:CODE
LOWtests/diffusion/test_diffusion_streaming_output.py86 except Exception as exc:CODE
LOWtests/diffusion/test_diffusion_streaming_output.py204 except Exception:CODE
MEDIUMtests/diffusion/test_diffusion_streaming_output.py83def step_streaming(self, request):CODE
LOW…iffusion/offloader/test_diffusion_layerwise_offload.py149 except Exception:CODE
LOWtests/diffusion/offloader/test_diffusion_cpu_offload.py123 except Exception:CODE
LOWtests/diffusion/attention/test_trtllm_attn.py28 except Exception:CODE
LOWtests/diffusion/attention/test_flash_attn.py735 except Exception as e:CODE
LOWtests/diffusion/attention/test_flash_attn.py744 except Exception as e:CODE
LOWtests/diffusion/ar_diffusion/test_paged_attention.py81 except Exception:CODE
LOWtests/diffusion/ar_diffusion/test_paged_attention.py92 except Exception:CODE
LOWtests/diffusion/models/pi0/test_pi0_parity.py354 except Exception as e: # noqa: BLE001CODE
LOW…s/hunyuan_image3/test_hunyuan_image3_it2i_ar_format.py114 except Exception:CODE
LOWtests/diffusion/batching/test_diffusion_batching.py514 except Exception as e:CODE
LOWtests/diffusion/batching/test_diffusion_batching.py338 except Exception as e:CODE
LOWtests/diffusion/batching/test_diffusion_batching.py378 except Exception as e:CODE
LOWtests/diffusion/batching/test_diffusion_batching.py480 except Exception as e:CODE
LOWtests/dfx/conftest.py172 except Exception as e:CODE
LOWtests/dfx/reliability/test_reliability_qwen3_omni.py300 except Exception: # noqa: BLE001CODE
LOWtests/dfx/reliability/test_reliability_qwen3_omni.py314 except Exception as exc: # noqa: BLE001CODE
LOWtests/dfx/reliability/test_reliability_qwen3_omni.py335 except Exception as exc: # noqa: BLE001CODE
LOWtests/dfx/reliability/test_reliability_qwen3_omni.py451 except Exception as exc:CODE
1262 more matches not shown…
Deep Nesting901 hits · 750 pts
SeverityFileLineSnippetContext
LOWcollect_env.py187CODE
LOWcollect_env.py356CODE
LOWsetup.py43CODE
LOWsetup.py123CODE
LOWtools/configure_stage_memory.py192CODE
LOWtools/nightly/send_nightly_email.py211CODE
LOWtools/nightly/buildkite_testcase_statistics.py324CODE
LOWtools/nightly/buildkite_testcase_statistics.py383CODE
LOWtools/nightly/buildkite_testcase_statistics.py437CODE
LOWtools/nightly/buildkite_testcase_statistics.py988CODE
LOWtools/nightly/generate_nightly_perf_excel.py727CODE
LOWtools/nightly/generate_nightly_perf_html.py192CODE
LOWtools/pre_commit/check_tts_adapter.py82CODE
LOWtools/pre_commit/check_spdx_header.py115CODE
LOWtools/pre_commit/check_spdx_header.py164CODE
LOWtools/baselines/omni_hash_baseline.py144CODE
LOWtests/host_weight_runtime/test_contracts.py208CODE
LOWtests/host_weight_runtime/test_filesystem_store.py1728CODE
LOWtests/host_weight_runtime/test_filesystem_store.py291CODE
LOW…/omni_connectors/test_mooncake_transfer_engine_rdma.py40CODE
LOW…/omni_connectors/test_mooncake_transfer_engine_rdma.py80CODE
LOW…/omni_connectors/test_mooncake_transfer_engine_rdma.py360CODE
LOW…/omni_connectors/test_mooncake_transfer_engine_rdma.py597CODE
LOW…/omni_connectors/test_mooncake_transfer_engine_rdma.py630CODE
LOWtests/config/test_environment_variables.py26CODE
LOWtests/config/test_environment_variables.py44CODE
LOWtests/config/test_environment_variables.py71CODE
LOWtests/config/test_environment_variables.py114CODE
LOWtests/config/test_environment_variables.py136CODE
LOWtests/model_executor/models/test_gepard_wiring.py168CODE
LOWtests/model_executor/models/test_omni_processing.py162CODE
LOW…executor/models/qwen3_tts/test_code_predictor_dtype.py821CODE
LOW…executor/models/qwen3_tts/test_code_predictor_dtype.py934CODE
LOW…executor/models/qwen3_tts/test_code_predictor_dtype.py1061CODE
LOW…el_executor/models/qwen2_5_omni/test_weight_loading.py7CODE
LOWtests/utils/test_mm_outputs.py32CODE
LOWtests/model_tests/conftest.py11CODE
LOWtests/model_tests/diffusion/test_common_online.py37CODE
LOWtests/model_tests/diffusion/test_common_offline.py30CODE
LOWtests/model_tests/diffusion/task_runners.py240CODE
LOWtests/model_tests/diffusion/task_runners.py266CODE
LOWtests/diffusion/test_multiproc_engine_concurrency.py100CODE
LOWtests/diffusion/test_multiproc_engine_concurrency.py105CODE
LOWtests/diffusion/test_diffusion_engine_metrics.py87CODE
LOWtests/diffusion/test_diffusion_step_pipeline.py437CODE
LOWtests/diffusion/attention/test_ulysses_uaa.py57CODE
LOWtests/diffusion/attention/test_flash_attn.py536CODE
LOWtests/diffusion/cache/test_cache_dit.py96CODE
LOWtests/diffusion/ar_diffusion/test_paged_attention.py63CODE
LOW…sion/models/lingbot_world/test_lingbot_world_typing.py21CODE
LOW…ion/models/hunyuan_image3/test_diffusion_kv_request.py240CODE
LOW…ts/diffusion/models/minimax_h3/test_minimax_h3_lora.py46CODE
LOWtests/diffusion/quantization/test_fp8_config.py336CODE
LOWtests/diffusion/quantization/test_quantization_fp8.py191CODE
LOWtests/diffusion/batching/test_diffusion_batching.py284CODE
LOWtests/diffusion/lora/test_loader.py133CODE
LOW…/diffusion/diffusion_backend/test_diffusers_backend.py147CODE
LOWtests/dfx/conftest.py38CODE
LOWtests/dfx/stability/helpers.py32CODE
LOWtests/dfx/stability/helpers.py65CODE
841 more matches not shown…
Cross-Language Confusion119 hits · 649 pts
SeverityFileLineSnippetContext
HIGHtools/nightly/buildkite_testcase_statistics.py877 const fileName = (fileDetail.dataset.testFile || fileDetail.textContent || "").toLowerCase();CODE
HIGHtools/nightly/buildkite_testcase_statistics.py878 const pathOk = !query || fileName.includes(query);CODE
HIGHtools/nightly/buildkite_testcase_statistics.py883 const matched = pathOk && anyRowVisible;CODE
HIGHtools/nightly/generate_nightly_perf_excel.py120# and max_concurrency (null); leave those as-is. If they become float in the future, they areCOMMENT
HIGHtools/nightly/generate_nightly_perf_html.py648 if (value === null || value === undefined) return "";CODE
HIGHtools/nightly/generate_nightly_perf_html.py648 if (value === null || value === undefined) return "";CODE
HIGHtools/nightly/generate_nightly_perf_html.py658 const items = new Set(arr.filter(v => v !== null && v !== undefined && String(v).trim() !== ""));CODE
HIGHtools/nightly/generate_nightly_perf_html.py663 if (v === null || v === undefined) return null;CODE
HIGHtools/nightly/generate_nightly_perf_html.py663 if (v === null || v === undefined) return null;CODE
HIGHtools/nightly/generate_nightly_perf_html.py664 if (typeof v === "number") return Number.isFinite(v) ? v : null;CODE
HIGHtools/nightly/generate_nightly_perf_html.py666 if (!s || s.toLowerCase() === "inf") return null;CODE
HIGHtools/nightly/generate_nightly_perf_html.py666 if (!s || s.toLowerCase() === "inf") return null;CODE
HIGHtools/nightly/generate_nightly_perf_html.py668 return Number.isFinite(n) ? n : null;CODE
HIGHtools/nightly/generate_nightly_perf_html.py1005 if (x < padLeft || x > padLeft + innerW || y < padTop || y > padTop + innerH) {{CODE
HIGHtools/nightly/generate_nightly_perf_html.py1012 let best = null;CODE
HIGHtools/nightly/generate_nightly_perf_html.py1018 if (!best || dist < best.dist) {{CODE
HIGHtools/nightly/generate_nightly_perf_html.py1027 .filter(([, value]) => value !== null && value !== undefined && String(value).trim() !== "")CODE
HIGHtools/nightly/generate_nightly_perf_html.py1179 if (filters.model && String(row[filters.modelKey] || "") !== filters.model) return false;CODE
HIGHtools/nightly/generate_nightly_perf_html.py1180 if (filters.testName && String(row.test_name || "") !== filters.testName) return false;CODE
HIGHtools/nightly/generate_nightly_perf_html.py1181 if (filters.datasetName && String(row[filters.datasetKey] || "") !== filters.datasetName) return false;CODE
HIGHtools/nightly/generate_nightly_perf_html.py704 row.max_concurrency !== undefined && row.max_concurrency !== null ? `c${{row.max_concurrency}}` : "",CODE
HIGHtools/nightly/generate_nightly_perf_html.py705 row.num_prompts !== undefined && row.num_prompts !== null ? `p${{row.num_prompts}}` : "",CODE
HIGHtools/nightly/generate_nightly_perf_html.py720 row.max_concurrency !== undefined && row.max_concurrency !== null ? `c${{row.max_concurrency}}` : "",CODE
HIGHtools/nightly/generate_nightly_perf_html.py721 row.num_prompts !== undefined && row.num_prompts !== null ? `p${{row.num_prompts}}` : "",CODE
HIGHtools/nightly/generate_nightly_perf_html.py733 const raw = value === null || value === undefined ? "" : String(value);CODE
HIGHtools/nightly/generate_nightly_perf_html.py786 if (previous && values.map((v) => String(v)).includes(String(previous))) {{CODE
HIGHtools/nightly/generate_nightly_perf_html.py871 if (display.title && display.title !== display.text) {{CODE
HIGHtools/nightly/generate_nightly_perf_html.py890 const dpr = window.devicePixelRatio || 1;CODE
HIGHtools/nightly/generate_nightly_perf_html.py912 xLabels.push(p.x);CODE
HIGHtools/nightly/generate_nightly_perf_html.py1057 .filter((value) => value !== null && value !== undefined && String(value).trim() !== "")CODE
HIGHtools/nightly/generate_nightly_perf_html.py1062 if (y === null) return;CODE
HIGHtools/nightly/generate_nightly_perf_html.py1074 seriesByKey.get(key).points.push({{ x: dateStr, y, meta }});CODE
HIGHtools/nightly/generate_nightly_perf_html.py1100 if (latest && latest.date) {{CODE
HIGHtools/nightly/generate_nightly_perf_html.py1101 lines.push(`date: <code>${{escapeHtml(formatDate(latest.date))}}</code>`);CODE
HIGHtools/nightly/generate_nightly_perf_html.py1112 lines.push(`${{escapeHtml(keyName)}}: <code>${{escapeHtml(String(latest[keyName]))}}</code>`);CODE
HIGHtools/nightly/generate_nightly_perf_html.py1116 if (value) lines.push(`${{escapeHtml(metric)}}: <code>${{value}}</code>`);CODE
HIGHtools/nightly/generate_nightly_perf_html.py1134 if (!r || r.width < 10 || r.height < 10) return;CODE
HIGHtools/nightly/generate_nightly_perf_html.py1194 if (filters.backend && String(row[filters.backendKey] || "") !== filters.backend) return false;CODE
HIGHtools/nightly/generate_nightly_perf_html.py1211 const values = rows.map((row) => toNumber(row[key])).filter((v) => v !== null);CODE
HIGHtools/nightly/generate_nightly_perf_html.py1400 if (extraKey && extraInput && !extraInput.value.trim()) {{CODE
HIGHtools/nightly/generate_nightly_perf_html.py1412 const latest = filteredAsc[filteredAsc.length - 1] || null;CODE
HIGHtests/diffusion/ar_diffusion/test_slot_mapping.py61 Every slot for the in-flight chunk must land inside a real (non-null) blockSTRING
HIGHtests/examples/test_minicpmo_realtime_web_static.py41 assert "event.event || event" in sourceCODE
HIGHtests/examples/test_minicpmo_realtime_web_static.py147 this.port = { onmessage: null, postMessage: () => {} };STRING
HIGHtests/examples/test_minicpmo_realtime_web_static.py150 let Processor = null;STRING
HIGHtests/examples/test_minicpmo_realtime_web_static.py227 onmessage: null,STRING
HIGHtests/examples/test_minicpmo_realtime_web_static.py228 postMessage: (message) => messages.push(message),STRING
HIGHtests/examples/test_minicpmo_realtime_web_static.py232 let Processor = null;STRING
HIGHtests/examples/test_minicpmo_realtime_web_static.py260 tail.every((sample, index) => index === 0 || sample <= tail[index - 1]),STRING
HIGHtests/dfx/stability/scripts/generate_report.py305 charts.push(new Chart(canvas, {{CODE
HIGHtests/entrypoints/test_stream_finish_reason.py123 """text+audio: text finish sends finish_reason=null, audio sends stop."""STRING
HIGHtests/entrypoints/openai_api/test_serving_speech.py3641 '{"error":{"message":"engine dead","type":"InternalServerError","param":null,'CODE
HIGH…s/entrypoints/openai_api/test_serving_speech_stream.py454 # Real-time audio chunk (timestamps null), then the timestamp frame.COMMENT
HIGH…ne_serving/text_to_speech/moss_tts_nano/gradio_demo.py55 this.buf = null;CODE
HIGH…ne_serving/text_to_speech/moss_tts_nano/gradio_demo.py60 if (e.data && e.data.type === 'clear') {CODE
HIGH…ne_serving/text_to_speech/moss_tts_nano/gradio_demo.py61 this.queue = []; this.buf = null; this.pos = 0; this.played = 0;CODE
HIGH…ne_serving/text_to_speech/moss_tts_nano/gradio_demo.py182 if (mTtfp && st.ttfp != null) mTtfp.textContent = st.ttfp.toFixed(0) + 'ms';CODE
HIGH…ne_serving/text_to_speech/moss_tts_nano/gradio_demo.py182 if (mTtfp && st.ttfp != null) mTtfp.textContent = st.ttfp.toFixed(0) + 'ms';CODE
HIGH…ne_serving/text_to_speech/moss_tts_nano/gradio_demo.py185 if (dur > 0 && elapsed > 0) {{CODE
HIGH…ne_serving/text_to_speech/moss_tts_nano/gradio_demo.py73 this.queue.push(mono);CODE
59 more matches not shown…
Cross-File Repetition121 hits · 605 pts
SeverityFileLineSnippetContext
HIGH…buted/omni_connectors/test_bagel_mooncake_connector.py0validate that image pixels match expected reference values. args: image: the pil image to validate. reference_pixels: liSTRING
HIGH…/omni_connectors/test_bagel_shared_memory_connector.py0validate that image pixels match expected reference values. args: image: the pil image to validate. reference_pixels: liSTRING
HIGH…line_inference/test_sensenova_u1_text2img_expansion.py0validate that image pixels match expected reference values. args: image: the pil image to validate. reference_pixels: liSTRING
HIGH…fline_inference/test_sensenova_u1_img2img_expansion.py0validate that image pixels match expected reference values. args: image: the pil image to validate. reference_pixels: liSTRING
HIGH…l_executor/models/voxcpm2/test_talker_output_marker.py0defer talker import (pulls vllm model_executor) until first use.STRING
HIGH…_executor/models/voxcpm2/test_talker_state_eviction.py0defer talker import (pulls vllm model_executor) until first use.STRING
HIGHtests/model_executor/models/dots_tts/test_talker.py0defer talker import (pulls vllm model_executor) until first use.STRING
HIGH…del_executor/models/moss_tts/test_per_request_codes.py0defer talker import (pulls vllm model_executor) until first use.STRING
HIGH…s/e2e/offline_inference/test_qwen2_5_omni_expansion.py0test audio and text input processing and text/audio output generation via openai api. deploy setting: default yaml inputSTRING
HIGHtests/e2e/online_serving/test_qwen3_tts_customvoice.py0test audio and text input processing and text/audio output generation via openai api. deploy setting: default yaml inputSTRING
HIGHtests/e2e/online_serving/test_qwen3_tts_base.py0test audio and text input processing and text/audio output generation via openai api. deploy setting: default yaml inputSTRING
HIGHtests/e2e/online_serving/test_qwen2_5_omni_expansion.py0test audio and text input processing and text/audio output generation via openai api. deploy setting: default yaml inputSTRING
HIGHtests/e2e/online_serving/test_qwen3_omni.py0test audio and text input processing and text/audio output generation via openai api. deploy setting: default yaml inputSTRING
HIGHdocs/contributing/ci/test_writing_guide.md0test audio and text input processing and text/audio output generation via openai api. deploy setting: default yaml inputSTRING
HIGHtests/e2e/online_serving/test_mimo_audio_expansion.py0test audio and text input processing and text/audio output generation via openai api. deploy setting: default yaml inputSTRING
HIGHtests/e2e/online_serving/test_wan22_t2v.py0return a single default ``omniserverparams`` row (no extra ``server_args``).STRING
HIGHtests/e2e/online_serving/test_qwen_image.py0return a single default ``omniserverparams`` row (no extra ``server_args``).STRING
HIGHtests/e2e/online_serving/test_qwen_image_layered.py0return a single default ``omniserverparams`` row (no extra ``server_args``).STRING
HIGHtests/e2e/online_serving/test_boogu_image.py0return a single default ``omniserverparams`` row (no extra ``server_args``).STRING
HIGH…ts/e2e/online_serving/test_qwen3_tts_base_expansion.py0e2e online tests for qwen3-tts model with text input and audio output. these tests verify the /v1/audio/speech endpoint STRING
HIGH…online_serving/test_qwen3_tts_customvoice_expansion.py0e2e online tests for qwen3-tts model with text input and audio output. these tests verify the /v1/audio/speech endpoint STRING
HIGHtests/e2e/online_serving/test_qwen3_tts_customvoice.py0e2e online tests for qwen3-tts model with text input and audio output. these tests verify the /v1/audio/speech endpoint STRING
HIGHtests/e2e/online_serving/test_qwen3_tts_base.py0e2e online tests for qwen3-tts model with text input and audio output. these tests verify the /v1/audio/speech endpoint STRING
HIGH…ts/e2e/online_serving/test_qwen3_tts_base_expansion.py0text prompt for text-to-audio tests (same as test_qwen3_omni - beijing test case).STRING
HIGH…online_serving/test_qwen3_tts_customvoice_expansion.py0text prompt for text-to-audio tests (same as test_qwen3_omni - beijing test case).STRING
HIGHtests/e2e/online_serving/test_qwen3_tts_customvoice.py0text prompt for text-to-audio tests (same as test_qwen3_omni - beijing test case).STRING
HIGHtests/e2e/online_serving/test_qwen3_tts_base.py0text prompt for text-to-audio tests (same as test_qwen3_omni - beijing test case).STRING
HIGH…ts/e2e/online_serving/test_qwen3_tts_base_expansion.py0batch size for concurrent requests (same as test_qwen3_omni).STRING
HIGH…online_serving/test_qwen3_tts_customvoice_expansion.py0batch size for concurrent requests (same as test_qwen3_omni).STRING
HIGHtests/e2e/online_serving/test_qwen3_tts_customvoice.py0batch size for concurrent requests (same as test_qwen3_omni).STRING
HIGHtests/e2e/online_serving/test_qwen3_tts_base.py0batch size for concurrent requests (same as test_qwen3_omni).STRING
HIGH…ts/e2e/online_serving/test_qwen3_tts_base_expansion.py0test text input processing and audio output via openai api. deploy setting: default yaml input modal: text output modal:STRING
HIGH…online_serving/test_qwen3_tts_customvoice_expansion.py0test text input processing and audio output via openai api. deploy setting: default yaml input modal: text output modal:STRING
HIGHtests/e2e/online_serving/test_qwen3_tts_customvoice.py0test text input processing and audio output via openai api. deploy setting: default yaml input modal: text output modal:STRING
HIGHtests/e2e/online_serving/test_qwen3_tts_base.py0test text input processing and audio output via openai api. deploy setting: default yaml input modal: text output modal:STRING
HIGHtests/e2e/online_serving/test_qwen2_5_omni_expansion.py0test text input processing and text-only output generation via openai api. deploy setting: default yaml input modal: texSTRING
HIGHtests/e2e/online_serving/test_qwen3_omni.py0test text input processing and text-only output generation via openai api. deploy setting: default yaml input modal: texSTRING
HIGHtests/e2e/online_serving/test_mimo_audio_expansion.py0test text input processing and text-only output generation via openai api. deploy setting: default yaml input modal: texSTRING
HIGH…ne_serving/text_to_speech/moss_tts_nano/gradio_demo.py0<div id="tts-player"> <div style="display:flex; align-items:center; gap:10px;"> <div id="tts-status-dot" style="width:10STRING
HIGH…online_serving/text_to_speech/qwen3_tts/gradio_demo.py0<div id="tts-player"> <div style="display:flex; align-items:center; gap:10px;"> <div id="tts-status-dot" style="width:10STRING
HIGH…s/online_serving/text_to_speech/voxcpm2/gradio_demo.py0<div id="tts-player"> <div style="display:flex; align-items:center; gap:10px;"> <div id="tts-status-dot" style="width:10STRING
HIGH…ne_serving/text_to_speech/moss_tts_nano/gradio_demo.py0<script> const sr = {sample_rate}; const wc = {json.dumps(worklet_js)}; let ctx = null, node = null, abort = null, gen =STRING
HIGH…s/online_serving/text_to_speech/voxcpm2/gradio_demo.py0<script> const sr = {sample_rate}; const wc = {json.dumps(worklet_js)}; let ctx = null, node = null, abort = null, gen =STRING
HIGH…online_serving/text_to_speech/qwen3_tts/gradio_demo.py0<script> const sr = {sample_rate}; const wc = {json.dumps(worklet_js)}; let ctx = null, node = null, abort = null, gen =STRING
HIGH…ne_serving/text_to_speech/moss_tts_nano/gradio_demo.py0encode gradio audio input (sample_rate, numpy_array) to base64 data url.STRING
HIGH…s/online_serving/text_to_speech/glm_tts/gradio_demo.py0encode gradio audio input (sample_rate, numpy_array) to base64 data url.STRING
HIGH…line_serving/text_to_speech/fish_speech/gradio_demo.py0encode gradio audio input (sample_rate, numpy_array) to base64 data url.STRING
HIGH…/online_serving/text_to_speech/qwen3_tts/tts_common.py0encode gradio audio input (sample_rate, numpy_array) to base64 data url.STRING
HIGH…ne_serving/text_to_speech/moss_tts_nano/gradio_demo.py0#generate-btn button { width: 100%; } #streaming-player { border: 1px solid var(--border-color-primary) !important; bordSTRING
HIGH…s/online_serving/text_to_speech/voxcpm2/gradio_demo.py0#generate-btn button { width: 100%; } #streaming-player { border: 1px solid var(--border-color-primary) !important; bordSTRING
HIGH…online_serving/text_to_speech/qwen3_tts/gradio_demo.py0#generate-btn button { width: 100%; } #streaming-player { border: 1px solid var(--border-color-primary) !important; bordSTRING
HIGH…ne_serving/text_to_speech/moss_tts_nano/gradio_demo.py0<div style="text-align:center; padding:8px 0; margin-top:4px;"> <a href="https://github.com/vllm-project/vllm-omni" targSTRING
HIGH…s/online_serving/text_to_speech/voxcpm2/gradio_demo.py0<div style="text-align:center; padding:8px 0; margin-top:4px;"> <a href="https://github.com/vllm-project/vllm-omni" targSTRING
HIGH…online_serving/text_to_speech/qwen3_tts/gradio_demo.py0<div style="text-align:center; padding:8px 0; margin-top:4px;"> <a href="https://github.com/vllm-project/vllm-omni" targSTRING
HIGHbenchmarks/glm_image/benchmark_glm_image.py0return path to prompt.json, downloading from huggingface if needed.STRING
HIGHbenchmarks/glm_image/vllm-omni/inference.py0return path to prompt.json, downloading from huggingface if needed.STRING
HIGHbenchmarks/glm_image/huggingface/inference.py0return path to prompt.json, downloading from huggingface if needed.STRING
HIGH…model_executor/models/hunyuan_image3/hunyuan_image3.py0embeds scalar timesteps into vector representations.STRING
HIGH…odels/indextts2/s2mel/modules/diffusion_transformer.py0embeds scalar timesteps into vector representations.STRING
HIGH…on/models/hunyuan_image3/hunyuan_image3_transformer.py0embeds scalar timesteps into vector representations.STRING
61 more matches not shown…
Modern Structural Boilerplate552 hits · 551 pts
SeverityFileLineSnippetContext
LOWtools/nightly/generate_nightly_perf_excel.py753def _set_column_width(ws, columns: Sequence[str], col_name: str, width: int) -> None:CODE
LOWtests/config/test_config_factory.py3036 def set_files(self, files: list[tuple[str, str]]) -> None:CODE
LOW…utor/stage_input_processors/test_nemotron_voicechat.py341 def set_init_inputs(self, speaker_name: str) -> None:CODE
LOWtests/model_tests/diffusion/utils.py10logger = logging.getLogger(__name__)CODE
LOWtests/diffusion/test_diffusion_step_pipeline.py67def _update_environment_variables(envs_dict: dict[str, str]) -> None:CODE
LOWtests/diffusion/test_diffusion_step_pipeline.py897 def set_active_adapter(self, adapter, scale: float = 1.0) -> None:CODE
LOWtests/diffusion/distributed/test_comm.py33def _update_environment_variables(envs_dict: dict[str, str]) -> None:CODE
LOWtests/diffusion/distributed/test_pipeline_parallel.py320 def _set_rank(monkeypatch, world_size: int, first_stage: bool) -> None:CODE
LOWtests/diffusion/distributed/test_cfg_parallel.py34def _set_random_seeds(seed: int) -> None:CODE
LOWtests/diffusion/distributed/test_cfg_parallel.py65def _update_environment_variables(envs_dict: dict[str, str]) -> None:CODE
LOW…n/models/lingbot_world/test_lingbot_world_attention.py61 def set_weight_attrs(weight: torch.Tensor, attrs: dict) -> None:CODE
LOW…n/models/lingbot_world/test_lingbot_world_attention.py267def _set_identity_attention(attention: nn.Module) -> None:CODE
LOWtests/diffusion/models/wan2_2/conftest.py36 def set_timesteps(self, num_steps: int, device: torch.device) -> None:CODE
LOW…diffusion/models/wan2_2/test_wan22_pipeline_diffuse.py39 def set_timesteps(self, num_steps: int, device: torch.device) -> None:CODE
LOWtests/diffusion/diffusion_kv/test_initialization.py35 def set_kv_cache_configs(self, configs, resolved_max_model_len) -> None:CODE
LOWtests/dfx/reliability/helpers.py34logger = logging.getLogger(__name__)CODE
LOW…/entrypoints/openai_api/test_serving_audio_generate.py26logger = logging.getLogger(__name__)CODE
LOWtests/entrypoints/openai_api/test_serving_speech.py68logger = logging.getLogger(__name__)CODE
LOWtests/e2e/features/helpers/custom_pipeline.py37logger = logging.getLogger(__name__)CODE
LOWtests/e2e/features/helpers/custom_pipeline.py262def set_death_signal() -> None:CODE
LOWtests/e2e/features/helpers/verl_omni_server.py34logger = logging.getLogger(__name__)CODE
LOWtests/e2e/accuracy/test_qwen_image.py99def _set_reference_attention_backend(pipe: DiffusionPipeline) -> None:CODE
LOW…sts/e2e/accuracy/test_hunyuan_image3_pixel_accuracy.py23logger = logging.getLogger(__name__)CODE
LOWtests/e2e/accuracy/test_diffusers_backend_similarity.py52def _set_matched_attention_backend(pipe: DiffusionPipeline) -> None:CODE
LOW…ts/e2e/online_serving/test_cosyvoice3_tts_expansion.py53def _set_rocm_request_timeout(request_config: dict) -> None:CODE
LOWtests/engine/test_orchestrator_stage_input_bridge.py73 def set_engine_outputs(self, _outputs) -> None:CODE
LOWtests/engine/test_orchestrator.py166 def set_engine_outputs(self, outputs) -> None:CODE
LOWtests/engine/test_orchestrator.py262 def update_scheduler_stats(self, _scheduler_stats) -> None:CODE
LOWtests/helpers/media.py32logger = logging.getLogger(__name__)CODE
LOWtests/helpers/media.py866__all__ = [CODE
LOWtests/helpers/clean.py20logger = logging.getLogger(__name__)CODE
LOWtests/helpers/clean.py255__all__ = [CODE
LOWtests/helpers/monitor.py55__all__ = ["DeviceMemoryMonitor"]CODE
LOWtests/helpers/client.py2129__all__ = [CODE
LOWtests/helpers/assertions.py1032__all__ = [CODE
LOWtests/helpers/runtime.py1208__all__ = [CODE
LOWtests/helpers/stage_config.py746__all__ = [CODE
LOW…ence/custom_pipeline/image_to_image/custom_pipeline.py12logger = logging.getLogger(__name__)CODE
LOW…es/offline_inference/qwen3_omni/end2end_async_chunk.py46logger = logging.getLogger(__name__)CODE
LOW…es/offline_inference/text_to_speech/glm_tts/end2end.py48logger = logging.getLogger(__name__)CODE
LOW…ffline_inference/text_to_speech/fish_speech/end2end.py38logger = logging.getLogger(__name__)CODE
LOW…/offline_inference/text_to_speech/qwen3_tts/end2end.py21logger = logging.getLogger(__name__)CODE
LOW…ffline_inference/text_to_speech/voxtral_tts/end2end.py35logger = logging.getLogger(__name__)CODE
LOWexamples/online_serving/minicpmo/realtime_web/server.py21logger = logging.getLogger(__name__)CODE
LOW…mples/online_serving/minicpmo/realtime_web/__init__.py5__all__ = ["build_app"]CODE
LOW…e_serving/step_audio2/openai_chat_completion_client.py34logger = logging.getLogger(__name__)CODE
LOW…ne_serving/text_to_speech/moss_tts_nano/gradio_demo.py38logger = logging.getLogger(__name__)CODE
LOW…s/online_serving/text_to_speech/voxcpm2/gradio_demo.py32logger = logging.getLogger(__name__)CODE
LOW…rving/text_to_speech/qwen3_tts/word_timestamps_demo.py47logger = logging.getLogger(__name__)CODE
LOW…online_serving/text_to_speech/qwen3_tts/gradio_demo.py45logger = logging.getLogger(__name__)CODE
LOWbenchmarks/kernels/mot_linear_benchmarks.py56logger = logging.getLogger(__name__)CODE
LOWbenchmarks/diffusion/diffusion_benchmark_serving.py112logger = logging.getLogger(__name__)CODE
LOWbenchmarks/accuracy/image_to_image/gedit_bench.py41logger = logging.getLogger(__name__)CODE
LOWvllm_omni/version.py51__all__ = ["__version__", "__version_tuple__"]CODE
LOWvllm_omni/__init__.py59__all__ = [CODE
LOWvllm_omni/metrics/definitions.py26logger = logging.getLogger(__name__)CODE
LOWvllm_omni/metrics/__init__.py10__all__ = [CODE
LOWvllm_omni/metrics/prometheus.py122 def set_running(self, n: int) -> None:CODE
LOWvllm_omni/metrics/prometheus.py127 def set_waiting(self, n: int) -> None:CODE
LOWvllm_omni/metrics/prometheus.py183 def set_stage_waiting_requests(self, stage: int, n_waiting: int) -> None:CODE
492 more matches not shown…
Over-Commented Block383 hits · 356 pts
SeverityFileLineSnippetContext
LOWtools/run_merge_jobs.sh1#!/usr/bin/env bashCOMMENT
LOWtools/run_merge_jobs.sh21# Usage:COMMENT
LOWtools/run_ready_jobs.sh1#!/usr/bin/env bashCOMMENT
LOWtools/run_ready_jobs.sh21# Usage:COMMENT
LOWtools/run_jobs_common.sh1#!/usr/bin/env bashCOMMENT
LOWtools/nightly/run_nightly_jobs.sh1#!/usr/bin/env bashCOMMENT
LOWtools/nightly/run_nightly_jobs.sh21# function — label has neither "Perf Test" nor "Accuracy Test" (incl. Doc, Multi-Replica, etc.)COMMENT
LOWtools/nightly/run_nightly_jobs.sh41# From repo root: pytest -sv -m "<markers> and local_model" (markers from MODEL_TYPE: omni, tts,COMMENT
LOWtools/nightly/run_nightly_jobs.sh61# • Set REPO_ROOT (or pass --repo-root) — default YAML is $REPO_ROOT/.buildkite/cuda/test-nightly.ymlCOMMENT
LOWtests/attention/test_fish_kvcache_attn.py261 assert captured["key"] == expectedCOMMENT
LOWtests/model_executor/models/voxcpm2/__init__.py1# SPDX-License-Identifier: Apache-2.0COMMENT
LOWtests/model_executor/models/dots_tts/__init__.py1# SPDX-License-Identifier: Apache-2.0COMMENT
LOW…l_executor/models/common/test_fused_group_norm_silu.py41 H, W = spatial_sizeCOMMENT
LOWtests/model_executor/models/minicpmo_4_5/__init__.py1# SPDX-License-Identifier: Apache-2.0COMMENT
LOWtests/model_tests/diffusion/config_types.py61COMMENT
LOWtests/model_tests/diffusion/model_settings.py1from tests.model_tests.diffusion import diff_model_buildersCOMMENT
LOWtests/diffusion/attention/test_flash_attn.py461 query = torch.randn(1, 8, 2, 4)COMMENT
LOWtests/diffusion/models/pi0/__init__.py1# SPDX-License-Identifier: Apache-2.0COMMENT
LOWtests/diffusion/models/pi0/test_pi0_units.py481 masks = [bool(m.item()) for m in image_masks]COMMENT
LOW…s/hunyuan_image3/test_hunyuan_image3_it2i_ar_format.py61# it kept skipping because instantiatingCOMMENT
LOWtests/diffusion/models/lingbot_video/__init__.py1# SPDX-License-Identifier: Apache-2.0COMMENT
LOWtests/diffusion/models/hunyuan_video/__init__.py1# SPDX-License-Identifier: Apache-2.0COMMENT
LOWtests/diffusion/models/minimax_h3/__init__.py1# SPDX-License-Identifier: Apache-2.0COMMENT
LOWtests/diffusion/models/cosmos3/__init__.py1# SPDX-License-Identifier: Apache-2.0COMMENT
LOWtests/diffusion/quantization/__init__.py1# SPDX-License-Identifier: Apache-2.0COMMENT
LOWtests/diffusion/quantization/test_mxfp4_config.py361# Row-parallel (to_out, ffn.net_2): input is sharded (K/TP), output is full (N).COMMENT
LOWtests/dfx/stability/scripts/resource_monitor.sh1#!/bin/bashCOMMENT
LOWtests/entrypoints/test_async_omni.py361#COMMENT
LOWtests/entrypoints/test_pd_disaggregation.py1161# Tests: Failure mode & memory leak preventionCOMMENT
LOWtests/e2e/offline_inference/run_quantization_e2e.sh1#!/bin/bashCOMMENT
LOW…eatures/fullduplex/engine/test_duplex_deploy_config.py81 # Guards the admission probe. A capacity edit, a config that declares none,COMMENT
LOWtests/e2e/accuracy/sana_video/test_sana_video_golden.py321 # The native prompt cross-attention uses vLLM-Omni Attention ratherCOMMENT
LOWtests/helpers/runtime.py1041 print("OmniRunner started successfully")COMMENT
LOWrequirements/common.txt1# Common dependencies for all platformsCOMMENT
LOWexamples/offline_inference/lance/gradio_demo.py41# canonical (W, H) via:COMMENT
LOWexamples/offline_inference/lance/gradio_demo.py1021 )COMMENT
LOWexamples/offline_inference/audex/speech_to_speech.py41from vllm_omni.model_executor.models.audex.prompt import build_cond_promptCOMMENT
LOW…_inference/qwen3_omni/run_single_prompt_async_chunk.sh1#!/bin/bashCOMMENT
LOW…ine_inference/qwen2_5_omni/strategy_stage_replica.yaml1# Composable-parallel strategy for Qwen2.5-Omni.COMMENT
LOW…mples/offline_inference/qwen2_5_omni/strategy_tp2.yaml1# Composable-parallel strategy for Qwen2.5-Omni: tensor-parallel the thinker.COMMENT
LOW…s/offline_inference/text_to_speech/moss_tts/end2end.py161 gc.collect()COMMENT
LOW…line_serving/replica_data_parallel/wan2_2_ti2v_dp.yaml1# Replica data-parallel recipe for Wan2.2 TI2V-5B (video DiT).COMMENT
LOW…ples/online_serving/dreamzero/droid_sim_eval_client.py61 except ImportError as exc: # pragma: no cover - runtime dependency guardCOMMENT
LOW…/online_serving/joyvl_interaction/scripts/start_all.sh1#!/bin/bashCOMMENT
LOWexamples/online_serving/audex/client.py261 # The chat-pass recipe is size-conditional (see the TEXT_INSTRUCTIONCOMMENT
LOWexamples/online_serving/audex/run_server.sh1#!/bin/bashCOMMENT
LOW…amples/online_serving/text_to_video/run_server_ltx2.sh1#!/bin/bashCOMMENT
LOW…ne_serving/text_to_speech/higgs_audio_v2/run_server.sh1#!/bin/bashCOMMENT
LOW…ne_serving/text_to_speech/qwen3_tts/run_gradio_demo.sh1#!/bin/bashCOMMENT
LOWbenchmarks/kernels/mot_linear_benchmarks.py321 # Mask predicate calculation, TMA state machine, etc.COMMENT
LOW.buildkite/npu/pipeline-npu-a3.yml1# steps:COMMENT
LOW.buildkite/npu/pipeline-npu-a3.yml21# vault.hashicorp.com/agent-inject-template-ca.pem: "{{- with secret \"internal/data/ascend/buildkitd\" -}COMMENT
LOW.buildkite/npu/pipeline-npu-a3.yml41# path: tokenCOMMENT
LOW.buildkite/npu/pipeline-npu-a3.yml61# tar -xzf /tmp/buildkit.tar.gz -C /tmp/buildkitCOMMENT
LOW.buildkite/npu/pipeline-npu-a3.yml81# echo "$${IMAGE_REGISTRY}/$${IMAGE_NAME}:$${VLLM_IMAGE_TAG}"COMMENT
LOW.buildkite/npu/pipeline-npu-a3.yml101 # build.pull_request.labels includes "omni-test" ||COMMENT
LOW.buildkite/release/scripts/publish-release-images.sh1#!/bin/bashCOMMENT
LOW.buildkite/amd/test-amd-ready.yml201# mirror_hardwares: [amdproduction]COMMENT
LOW.buildkite/amd/test-amd-ready.yml221COMMENT
LOW.buildkite/amd/test-amd-merge.yml241# grade: BlockingCOMMENT
323 more matches not shown…
Docstring Block Structure63 hits · 315 pts
SeverityFileLineSnippetContext
HIGH…buted/omni_connectors/test_bagel_mooncake_connector.py122Generate an image using Bagel model with configured parameters. Args: omni: The Omni instance to use for geSTRING
HIGH…/omni_connectors/test_bagel_shared_memory_connector.py163Generate an image using Bagel model with img2img pipeline. Args: omni: The Omni instance to use for generatSTRING
HIGH…line_inference/test_sensenova_u1_text2img_expansion.py120Generate an image using SenseNova-U1 model with configured parameters. Args: omni: The Omni instance to useSTRING
HIGH…fline_inference/test_sensenova_u1_img2img_expansion.py127Generate an edited image using SenseNova-U1 model with img2img pipeline. Args: omni: The Omni instance to uSTRING
HIGHvllm_omni/distributed/omni_coordinator/load_balancer.py47Route a task to one of the available replicas. Args: task: The task to route. Not used by the randoSTRING
HIGHvllm_omni/config/composable_parallel/apply.py220Overlay per-role strategy specs onto a merged stage list. Args: stages: the ``list[StageConfig]`` returned STRING
HIGHvllm_omni/model_executor/models/cosyvoice3/utils.py253Make mask tensor containing indices of padded part. See description of make_non_pad_mask. Args: lengthSTRING
HIGH…/nemotron_voicechat/nemo_vendored/ear_tts_vae_codec.py175 Converts a spectrogram back into a waveform using the overlap-add method. This function is an approximate inverSTRING
HIGH…els/nemotron_voicechat/nemo_vendored/duplex_ear_tts.py444 Retrieve a cached audio prompt latent and adapt it to the requested batch size. This fetches a latent STRING
HIGH…els/nemotron_voicechat/nemo_vendored/duplex_ear_tts.py651 Returns a dictionary of initial inputs for inference, using registered buffers. Args: B (iSTRING
HIGH…r/models/nemotron_voicechat/nemo_vendored/tokenizer.py303 Adds a dictionary of special tokens (eos, pad, cls...). If special tokens are NOT in the vocabulary, they are STRING
HIGH…utor/models/nemotron_voicechat/nemo_vendored/fusion.py346 Factory function to create the appropriate fusion module. Args: fuse_method: One of None, "add", "concSTRING
HIGH…ls/nemotron_voicechat/nemo_vendored/ear_tts_commons.py50 Loads a configuration from a JSON or Python file. - For JSON files (`*.json`), it parses the file directly. STRING
HIGH…ls/nemotron_voicechat/nemo_vendored/ear_tts_commons.py230 Finds the path of the latest checkpoint file or directory in a directory. The latest checkpoint is determined STRING
HIGH…m_omni/model_executor/models/minimax_music3/weights.py89Load the audio embedding table and the RVQ depth decoder in place. Args: vllm_config: The stage's config, uSTRING
HIGH…m_omni/model_executor/models/minimax_music3/staging.py46Stage ``count`` entries of each row and return the device view. Args: values: One sequence of PythoSTRING
HIGHvllm_omni/model_executor/models/minimax_music3/dit.py412Solve an aligned condition into a vocoder latent ``[1, 128, T_mel]``. The previous window's latent is re-imposeSTRING
HIGHvllm_omni/model_executor/models/minimax_music3/dav.py173Decode a latent to interleaved stereo. Args: latent: ``[B, 128, T]`` vocoder latent. ReturSTRING
HIGH…m_omni/model_executor/models/indextts2/utils/common.py66Make mask tensor containing indices of padded part. See description of make_non_pad_mask. Args: lengthSTRING
HIGH…model_executor/models/qwen3_tts/qwen3_tts_tokenizer.py68 Initialize tokenizer with HuggingFace `from_pretrained` style. Args: pretrained_model_nameSTRING
HIGH…odel_executor/stage_input_processors/minimax_music3.py166Emit the unconditioned companion for a guided MiniMax Music 3 request. Classifier-free guidance is not optional forSTRING
HIGHvllm_omni/diffusion/registry.py393Initialize a diffusion model from the registry. This function: 1. Loads the model class from the registry 2STRING
HIGHvllm_omni/diffusion/data.py1002 Find an available port with retry logic. Args: port: Initial port to check porSTRING
HIGHvllm_omni/diffusion/attention/backends/registry.py124Register or override a diffusion backend implementation. Args: backend: The DiffusionAttentionBackendEnum mSTRING
HIGHvllm_omni/diffusion/cache/selector.py12Get cache backend instance based on cache_backend string. This is a selector function that routes to the appropriatSTRING
HIGHvllm_omni/diffusion/cache/teacache/extractors.py1458 Get extractor function for given transformer class. This function looks up the extractor based on the exact trSTRING
HIGH…omni/diffusion/cache/teacache/coefficient_estimator.py221Estimate polynomial coefficients from collected data. Args: poly_order: Order of polynomial fit (deSTRING
HIGHvllm_omni/diffusion/cache/magcache/strategy.py473Get strategy function for given transformer class. This function looks up the strategy based on the exact transformSTRING
HIGHvllm_omni/diffusion/distributed/sp_sharding.py32Shard a tensor along the specified dimension for sequence parallelism. The tensor is split into world_size chunks aSTRING
HIGHvllm_omni/diffusion/distributed/parallel_state.py505 Initialize Ulysses, Ring, and AllGather-KV process groups. AllGather-KV is mutually exclusive with Ulysses andSTRING
HIGH…on/models/hunyuan_image3/hunyuan_image3_transformer.py2956 The call function to the pipeline for generation. Args: prompt (`str` or `List[str]`): STRING
HIGH…ni/diffusion/models/glm_image/glm_image_transformer.py64Validate GLM-Image TP constraints without requiring a distributed context. Args: dim: Model hidden dimensioSTRING
HIGH…ni/diffusion/models/glm_image/glm_image_transformer.py428Get cache for a specific layer. Args: layer_idx: Index of the layer (0-indexed). Returns: STRING
HIGH…mni/diffusion/models/sana_video/pipeline_sana_video.py813 Function invoked when calling the pipeline for generation. Args: prompt (`str` or `list[stSTRING
HIGH…lm_omni/diffusion/models/omnigen2/pipeline_omnigen2.py817 Get prompt embeddings from the Qwen2 text encoder. Args: prompt: The prompt or list of proSTRING
HIGH…on/models/gr00t/dataio/state_action/action_chunking.py131 Convert a relative action chunking to an absolute action chunking by applying the relative poses on topSTRING
HIGH…on/models/gr00t/dataio/state_action/action_chunking.py168 Convert action chunking to the specified action format. Must be implemented by subclasses. ArgSTRING
HIGH…on/models/gr00t/dataio/state_action/action_chunking.py259 Interpolate the joint action chunking to generate intermediate configurations. Uses linear interpolatiSTRING
HIGH…on/models/gr00t/dataio/state_action/action_chunking.py338 Convert a relative joint action chunking to an absolute action chunking by adding the relative joint poSTRING
HIGH…on/models/gr00t/dataio/state_action/action_chunking.py373 Convert trajectory to the desired format. Args: action_format: The desired output format STRING
HIGH…on/models/gr00t/dataio/state_action/action_chunking.py479 Interpolate the action chunking to generate intermediate poses. Uses linear interpolation for translatSTRING
HIGH…on/models/gr00t/dataio/state_action/action_chunking.py635 Convert trajectory to the desired format. Args: action_format: The desired output format STRING
HIGH…mni/diffusion/models/gr00t/dataio/state_action/pose.py101 Compute relative transformation between two poses. For EndEffectorPose: Computes the relative transforSTRING
HIGH…mni/diffusion/models/gr00t/dataio/state_action/pose.py243 Compute relative joint displacement. Args: other: Reference joint pose Returns: STRING
HIGH…ls/gr00t/dataio/state_action/state_action_processor.py227 Reverse state processing (denormalization). Args: state: Dict mapping joint_group -> proceSTRING
HIGH…ls/gr00t/dataio/state_action/state_action_processor.py286 Apply action processing (absolute->relative conversion, normalization). Processing order: 1. CSTRING
HIGH…ls/gr00t/dataio/state_action/state_action_processor.py373 Reverse action processing (denormalization, relative->absolute conversion). Processing order: STRING
HIGH…iffusion/models/ernie_image/ernie_image_transformer.py60Validate ErnieImage TP constraints without requiring a distributed context. Args: heads: Number of attentioSTRING
HIGHvllm_omni/diffusion/hooks/sequence_parallel.py80Get a parameter value from args or kwargs by name. Args: identifier: The parameter name to look up.STRING
HIGHvllm_omni/diffusion/hooks/sequence_parallel.py538Get a submodule by dotted name, supporting wildcards. Args: model: The root module. name: Dotted paSTRING
HIGHvllm_omni/diffusion/worker/diffusion_worker.py1616 Execute a method on the worker. Args: method: Method name (str) or serialized callable (bySTRING
HIGHvllm_omni/entrypoints/async_omni.py480Generate outputs for the given prompt(s) asynchronously. Coordinates multi-stage pipeline execution. Processes STRING
HIGHvllm_omni/entrypoints/utils.py93Filter dict-like object by removing callables and recursively converting values. Converts dict-like objects to reguSTRING
HIGHvllm_omni/entrypoints/utils.py261Resolve the stage/deploy config file path from the model name. Resolves configuration path based on the model type STRING
HIGHvllm_omni/entrypoints/openai/image_api_utils.py21Parse size string to width and height tuple. Args: size_str: Size in format "WIDTHxHEIGHT" (e.g., "1024x102STRING
HIGHvllm_omni/entrypoints/openai/api_server.py1336Generate speech audio from text using the loaded TTS model. Args: request: Speech synthesis request in OpenSTRING
HIGHvllm_omni/entrypoints/openai/api_server.py1864Generate images from text prompts using diffusion models. OpenAI DALL-E compatible endpoint for text-to-image generSTRING
HIGHvllm_omni/entrypoints/openai/api_server.py3629Retrieve metadata for a previously created video job. Args: video_id: Identifier returned by ``POST /v1/vidSTRING
HIGHvllm_omni/entrypoints/openai/api_server.py3657Delete a stored video job and any generated output. If the job is still queued or running, this endpoint first atteSTRING
HIGHvllm_omni/entrypoints/openai/api_server.py3709Download the generated file for a completed video job. Args: video_id: Identifier of the video job whose ouSTRING
3 more matches not shown…
AI Structural Patterns325 hits · 298 pts
SeverityFileLineSnippetContext
LOWtests/distributed/omni_connectors/test_tp_rank_aware.py41CODE
LOW…stage_input_processors/test_fish_speech_async_chunk.py14CODE
LOW…r/stage_input_processors/test_qwen3_tts_async_chunk.py47CODE
LOWtests/diffusion/attention/test_attention_sp.py122CODE
LOW…on/models/lingbot_world/test_pipeline_lingbot_world.py220CODE
LOW…ffusion/models/hunyuan_image3/test_multi_resolution.py564CODE
LOWtests/diffusion/quantization/test_quantization_fp8.py112CODE
LOW…/diffusion/diffusion_backend/test_diffusers_backend.py379CODE
LOWtests/profile/test_omni_torch_profiler.py42CODE
LOWtests/entrypoints/openai_api/test_duplex_handler.py81CODE
LOWtests/e2e/features/helpers/custom_pipeline.py74CODE
LOWtests/e2e/features/helpers/custom_pipeline.py484CODE
LOWtests/engine/test_orchestrator.py118CODE
LOW…les/offline_inference/mimo_audio/process_speechdata.py131CODE
LOW…line_inference/text_to_speech/moss_tts_nano/end2end.py52CODE
LOW…/offline_inference/text_to_speech/indextts2/end2end.py43CODE
LOWexamples/online_serving/lance/openai_chat_client.py17CODE
LOW…es/online_serving/image_to_image/openai_chat_client.py31CODE
LOW…les/online_serving/text_to_image/openai_chat_client.py17CODE
LOWbenchmarks/accuracy/common.py141CODE
LOWbenchmarks/accuracy/image_to_image/gedit_bench.py419CODE
LOWvllm_omni/tokenizers/mammoth_moda2_tokenizer.py103CODE
LOWvllm_omni/experimental/fullduplex/client.py502CODE
LOWvllm_omni/experimental/fullduplex/joyvl/memory/brain.py21CODE
LOW…_omni/experimental/fullduplex/joyvl/decision/policy.py26CODE
LOWvllm_omni/platforms/npu/worker/npu_model_runner.py112CODE
LOW…ni/platforms/npu/worker/npu_generation_model_runner.py623CODE
LOWvllm_omni/core/sched/omni_scheduler_mixin.py462CODE
LOW…m_omni/model_executor/models/voxcpm2/minicpm4_paged.py123CODE
LOW…model_executor/models/hunyuan_image3/hunyuan_image3.py2189CODE
LOWvllm_omni/model_executor/models/cosyvoice3/utils.py23CODE
LOW…/model_executor/models/cosyvoice3/code2wav_core/cfm.py233CODE
LOW…el_executor/models/cosyvoice3/code2wav_core/hifigan.py383CODE
LOW…el_executor/models/cosyvoice3/code2wav_core/hifigan.py612CODE
LOW…el_executor/models/cosyvoice3/code2wav_core/hifigan.py859CODE
LOWvllm_omni/model_executor/models/glm_tts/glm_tts_dit.py436CODE
LOWvllm_omni/model_executor/models/glm_tts/glm_tts_dit.py659CODE
LOW…i/model_executor/models/glm_tts/glm_tts_dit_wrapper.py565CODE
LOW…m_omni/model_executor/models/glm_image/glm_image_ar.py1800CODE
LOW…i/model_executor/models/gepard/configuration_gepard.py40CODE
LOW…ecutor/models/fish_speech/configuration_fish_speech.py23CODE
LOW…ecutor/models/fish_speech/configuration_fish_speech.py77CODE
LOW…odel_executor/models/mimo_audio/mimo_audio_code2wav.py997CODE
LOW…executor/models/mimo_audio/modeling_audio_tokenizer.py1135CODE
LOW…executor/models/mimo_audio/modeling_audio_tokenizer.py1176CODE
LOW…lm_omni/model_executor/models/mimo_audio/mimo_audio.py766CODE
LOW…/model_executor/models/mimo_audio/config_mimo_audio.py17CODE
LOW…/model_executor/models/mimo_audio/config_mimo_audio.py122CODE
LOW…lm_omni/model_executor/models/dots_tts/dots_tts_dit.py124CODE
LOW…mni/model_executor/models/dots_tts/dots_tts_vocoder.py90CODE
LOW…mni/model_executor/models/dots_tts/dots_tts_vocoder.py133CODE
LOW…mni/model_executor/models/dots_tts/dots_tts_vocoder.py235CODE
LOW…mni/model_executor/models/dots_tts/dots_tts_vocoder.py727CODE
LOW…omni/model_executor/models/ming_tts/config_ming_tts.py113CODE
LOW…omni/model_executor/models/ming_tts/prompt_assembly.py257CODE
LOW…/nemotron_voicechat/nemo_vendored/ear_tts_vae_codec.py519CODE
LOW…/nemotron_voicechat/nemo_vendored/ear_tts_vae_codec.py598CODE
LOW…r/models/nemotron_voicechat/nemo_vendored/tokenizer.py131CODE
LOW…dels/nemotron_voicechat/nemo_vendored/ear_tts_model.py786CODE
LOW…dels/nemotron_voicechat/nemo_vendored/ear_tts_model.py1228CODE
265 more matches not shown…
Structural Annotation Overuse132 hits · 216 pts
SeverityFileLineSnippetContext
LOWtests/core/sched/test_generation_scheduler_restore.py52 # Step 1: process_pending_chunks moves req-B outCOMMENT
LOWtests/core/sched/test_generation_scheduler_restore.py57 # Step 2: simulate the try/except/finally patternCOMMENT
LOWtests/core/sched/test_generation_scheduler_restore.py66 # Step 3: verify request is restoredCOMMENT
LOWtests/diffusion/attention/test_attention_sp.py473 # Step 1: Run without SP (baseline with ulysses_degree=1, ring_degree=1)COMMENT
LOWtests/diffusion/attention/test_attention_sp.py502 # Step 2: Run with SP enabledCOMMENT
LOWtests/diffusion/attention/test_attention_sp.py534 # Step 3: Verify input consistency and compare outputsCOMMENT
LOWtests/e2e/offline_inference/test_moss_tts_realtime.py166 # Step 1: locate the realtime processor module in the snapshot.COMMENT
LOWtests/e2e/offline_inference/test_moss_tts_realtime.py191 # Step 2: encode the reference audio clip via MOSS-Audio-Tokenizer.COMMENT
LOWtests/e2e/offline_inference/test_moss_tts_realtime.py209 # Step 3: build the (L, 17) prefill grid.COMMENT
LOW…ures/custom_pipeline/test_async_omni_collective_rpc.py126 # Step 1: list_loras (the call that was crashing)COMMENT
LOW…ures/custom_pipeline/test_async_omni_collective_rpc.py130 # Step 2: generate (should still work after list_loras)COMMENT
LOW.claude/skills/vllm-omni-test/SKILL.md31### Step 1: Classify Test GoalCOMMENT
LOW.claude/skills/vllm-omni-test/SKILL.md37### Step 2: Select Test LevelCOMMENT
LOW.claude/skills/vllm-omni-test/SKILL.md45### Step 3: Pick MarkersCOMMENT
LOW.claude/skills/vllm-omni-test/SKILL.md91### Step 4: Generate Test Case SkeletonCOMMENT
LOW.claude/skills/vllm-omni-test/SKILL.md627### Step 5: Wire Buildkite (when CI must run the new test)COMMENT
LOW.claude/skills/vllm-omni-test/SKILL.md829### Step 6: Run TestsCOMMENT
LOW.claude/skills/vllm-omni-test/SKILL.md856### Step 7: Validate Result QualityCOMMENT
LOW.claude/skills/precheck-pr/SKILL.md21### Step 1: Detect Base BranchCOMMENT
LOW.claude/skills/precheck-pr/SKILL.md31### Step 2: Validate PR TitleCOMMENT
LOW.claude/skills/precheck-pr/SKILL.md50### Step 3: Categorize the PRCOMMENT
LOW.claude/skills/precheck-pr/SKILL.md62### Step 4: Run ChecklistCOMMENT
LOW.claude/skills/precheck-pr/SKILL.md74### Step 5: Print ReportCOMMENT
LOW.claude/skills/add-diffusion-model/SKILL.md21## Step 0: Classify the Migration PathCOMMENT
LOW.claude/skills/add-diffusion-model/SKILL.md263### Step 4: Register Model in registry.pyCOMMENT
LOW.claude/skills/add-diffusion-model/SKILL.md283### Step 5: Run, Test, DebugCOMMENT
LOW.claude/skills/add-diffusion-model/SKILL.md301### Step 6: Add Example ScriptsCOMMENT
LOW.claude/skills/add-diffusion-model/SKILL.md308### Step 7: Update DocumentationCOMMENT
LOW.claude/skills/add-diffusion-model/SKILL.md318### Step 8: Add E2E TestsCOMMENT
LOW.claude/skills/add-diffusion-model/SKILL.md349### Step 9: Add Cache-DiT AccelerationCOMMENT
LOW.claude/skills/add-diffusion-model/SKILL.md453### Step 10: Add Parallelism SupportCOMMENT
LOW.claude/skills/add-diffusion-model/SKILL.md571### Step 11: Add CPU Offload SupportCOMMENT
LOW.claude/skills/add-diffusion-model/SKILL.md602### Step 12: Performance ProfilingCOMMENT
LOW…add-diffusion-model/references/parallelism-patterns.md186#### Step 1: Identify module boundariesCOMMENT
LOW…add-diffusion-model/references/parallelism-patterns.md208#### Step 2: Handle inline operationsCOMMENT
LOW…add-diffusion-model/references/parallelism-patterns.md227#### Step 3: Write `_sp_plan`COMMENT
LOW.claude/skills/vllm-omni-npu-upgrade/SKILL.md120### Step 1: PreparationCOMMENT
LOW.claude/skills/vllm-omni-npu-upgrade/SKILL.md137### Step 2: Analyze Omni-Specific LogicCOMMENT
LOW.claude/skills/vllm-omni-npu-upgrade/SKILL.md151### Step 3: Update Base Class (OmniNPUModelRunner)COMMENT
LOW.claude/skills/vllm-omni-npu-upgrade/SKILL.md173### Step 4: Update AR Model RunnerCOMMENT
LOW.claude/skills/vllm-omni-npu-upgrade/SKILL.md188### Step 5: Update Generation Model RunnerCOMMENT
LOW.claude/skills/vllm-omni-npu-upgrade/SKILL.md205### Step 6: Update ImportsCOMMENT
LOW.claude/skills/vllm-omni-npu-upgrade/SKILL.md226### Step 7: Sync GPU-Side Omni ChangesCOMMENT
LOW.claude/skills/vllm-omni-npu-upgrade/SKILL.md238### Step 8: ValidationCOMMENT
LOWdocs/design/feature/cfg_parallel.md84### Step 1: Inherit `CFGParallelMixin`COMMENT
LOWdocs/design/feature/cfg_parallel.md123### Step 2: Call `diffuse`COMMENT
LOWdocs/design/feature/tensor_parallel.md46### Step 1: Identify Linear LayersCOMMENT
LOWdocs/design/feature/tensor_parallel.md54### Step 2: Replace Linear Layers with Parallel EquivalentsCOMMENT
LOWdocs/design/feature/tensor_parallel.md148### Step 3: Validate TP ConstraintsCOMMENT
LOWdocs/design/feature/pipeline_parallel.md97### Step 1: Inherit `PipelineParallelMixin` and `CFGParallelMixin`COMMENT
LOWdocs/design/feature/pipeline_parallel.md123### Step 2: Make model forward and `predict_noise()` PP-awareCOMMENT
LOWdocs/design/feature/pipeline_parallel.md159### Step 3: Partition the transformer layersCOMMENT
LOWdocs/design/feature/pipeline_parallel.md235### Step 4: Use the standard denoising contractCOMMENT
LOWdocs/design/feature/hsdp.md44### Step 1: Identify Modules to ShardCOMMENT
LOWdocs/design/feature/hsdp.md55### Step 2: Define Shard ConditionsCOMMENT
LOWdocs/design/feature/sequence_parallel.md122### Step 1: Understand Module BoundariesCOMMENT
LOWdocs/design/feature/sequence_parallel.md145### Step 2: Handle Inline OperationsCOMMENT
LOWdocs/design/feature/sequence_parallel.md191### Step 3: Write `_sp_plan` for Your ModelCOMMENT
LOWdocs/design/feature/expert_parallel.md59### Step 1: Configure Expert Parallelism SettingsCOMMENT
LOWdocs/design/feature/expert_parallel.md72### Step 2: Use Sparse MoE Block to enable EP routing.COMMENT
72 more matches not shown…
Self-Referential Comments59 hits · 171 pts
SeverityFileLineSnippetContext
MEDIUM…ributed/omni_connectors/test_omni_connector_configs.py23 # This file is 4 levels deep from root if we count from tests?COMMENT
MEDIUMtests/config/test_config_factory.py908 # Create the modelCOMMENT
MEDIUMtests/config/test_config_factory.py949 # Create the modelCOMMENT
MEDIUMtests/config/test_config_factory.py1084 # Create a config with the autodetected type, and write theCOMMENT
MEDIUM…s/model_executor/models/glm_image/test_glm_image_ar.py137 # Create a mock for the parent parse_mm_dataCOMMENT
MEDIUMtests/diffusion/test_worker_wrapper_base.py465 # Create a mock worker class with re_init_pipelineCOMMENT
MEDIUMtests/diffusion/cache/test_cache_backends.py428 # Create a cache config with an scm policy & enable itCOMMENT
MEDIUM…iffusion/distributed/test_autoencoder_kl_wan_encode.py220 # Create a simple task with 2 temporal chunksCOMMENT
MEDIUMtests/diffusion/distributed/test_sp_plan_hooks.py517 # Create a test tensorCOMMENT
MEDIUMtests/diffusion/distributed/test_sp_plan_hooks.py534 # Create a test tensor with [text, image] concatenatedCOMMENT
MEDIUM…n/models/wan2_2/test_wan22_quant_config_propagation.py202 # Create a fake module that has WanRMS_normCOMMENT
MEDIUMtests/diffusion/models/ltx2/test_ltx2_pipeline.py2128 # Create a minimal od_config mockCOMMENT
MEDIUM…fusion/models/glm_image/test_glm_image_quantization.py544 # Create a minimal mock od_configCOMMENT
MEDIUM…els/sensenova_u1/test_sensenova_u1_logits_processor.py52 # Create a vLLM config with a wrapped diffusion configCOMMENT
MEDIUMtests/diffusion/quantization/test_hsdp_fp8.py53 # Create the quant-method instance without calling __init__ (avoidsCOMMENT
MEDIUMtests/entrypoints/openai_api/test_serving_speech.py548 # Create a file larger than 10MBCOMMENT
MEDIUMtests/entrypoints/openai_api/test_serving_speech.py878 # Create a TTS stageCOMMENT
MEDIUMtests/entrypoints/openai_api/test_image_server.py97 # Create a simple test imageCOMMENT
MEDIUMtests/worker/test_omni_gpu_model_runner.py216 # Create an instance without calling OmniGPUModelRunner.__init__COMMENT
MEDIUMtests/engine/test_arg_utils.py216 # Create a vLLM ModelConfigCOMMENT
MEDIUMtests/engine/test_arg_utils.py219 # Create a mock talking config with a dummy value for position_id_per_secondsCOMMENT
MEDIUMdocs/mkdocs/hooks/generate_examples.py398 # Create the EXAMPLE_DOC_DIR if it doesn't existCOMMENT
MEDIUM…ctors/connectors/mooncake_transfer_engine_connector.py313 # Create a socket to determine the local IP used for external communicationCOMMENT
MEDIUM…del_executor/models/fish_speech/fish_speech_fast_ar.py532 # Create a seeded generator for deterministic residual codebook sampling.COMMENT
MEDIUM…odel_executor/models/mimo_audio/mimo_audio_code2wav.py1019 # Create an all-zero logits tensor (sampler will choose the first token, usually pad token)COMMENT
MEDIUM…/nemotron_voicechat/nemo_vendored/ear_tts_vae_codec.py96 # Create a range tensor from 0 to max_length - 1COMMENT
MEDIUM…/nemotron_voicechat/nemo_vendored/ear_tts_vae_codec.py148 # Create the window tensor on the same device as the waveform.COMMENT
MEDIUM…els/nemotron_voicechat/nemo_vendored/duplex_ear_tts.py375 # Create a random dropout decision for each BOS instanceCOMMENT
MEDIUM…els/nemotron_voicechat/nemo_vendored/duplex_ear_tts.py386 # Create a mask of the same shape as target_text_tokensCOMMENT
MEDIUM…dels/nemotron_voicechat/nemo_vendored/ear_tts_model.py258 # Create a range tensor from 0 to max_length - 1COMMENT
MEDIUM…dels/nemotron_voicechat/nemo_vendored/ear_tts_model.py431 # Create a new, dense character vocabulary sorted by the original token IDCOMMENT
MEDIUM…dels/nemotron_voicechat/nemo_vendored/ear_tts_model.py860 # Create a padded tensor for the character IDsCOMMENT
MEDIUM…odels/nemotron_voicechat/nemo_vendored/asr/features.py33# This file contains code artifacts adapted from https://github.com/ryanleary/patterCOMMENT
MEDIUM…otron_voicechat/nemo_vendored/asr/conformer_encoder.py490 # Create the self-attention and padding masksCOMMENT
MEDIUM…on_voicechat/nemo_vendored/asr/multi_head_attention.py596 # Create a helper tensor to find the local indices of global attentionCOMMENT
MEDIUM…_tts/tokenizer_25hz/modeling_qwen3_tts_tokenizer_v1.py210 """This class implements an attentive statistic pooling layer for each channel.STRING
MEDIUM…lm_omni/model_executor/models/qwen3_omni/qwen3_omni.py185 # Create a new vllm_config with thinker_config as the hf_configCOMMENT
MEDIUM…lm_omni/model_executor/models/qwen3_omni/qwen3_omni.py221 # Create a new vllm_config with talker_config as the hf_configCOMMENT
MEDIUM…lm_omni/model_executor/models/qwen3_omni/qwen3_omni.py256 # Create a new vllm_config with code2wav_config as the hf_configCOMMENT
MEDIUM…el_executor/models/qwen3_omni/qwen3_omni_moe_talker.py365 # Create a vllm_config for the talker's text modelCOMMENT
MEDIUM…executor/models/qwen2_5_omni/qwen2_5_omni_token2wav.py168 """This class implements an attentive statistic pooling layer for each channel.STRING
MEDIUM…_executor/models/minicpmo_4_5/minicpmo_4_5_omni_llm.py4381 # Create a sequence tensor of shape (batch_size, max_seq_len)COMMENT
MEDIUMvllm_omni/diffusion/layers/mot/ops/mot_gemm.py228 # Initialize VAE variables first and overwrite for TextCOMMENT
MEDIUMvllm_omni/diffusion/distributed/comm.py251 # Create a contiguous buffer for receivingCOMMENT
MEDIUMvllm_omni/diffusion/utils/kv_utils.py45 # Create the boolean mask for the input sequencesCOMMENT
MEDIUM…on/models/hunyuan_image3/hunyuan_image3_transformer.py686 # This class is just for type hinting purposes. Attribute `i` should be definedCOMMENT
MEDIUM…iffusion/models/qwen_image/autoencoder_kl_qwenimage.py347 # Create the componentsSTRING
MEDIUM…/scheduling_flow_match_euler_discrete_time_shifting.py12# This file is adapted by Boogu Team from prior open-source scheduler work.COMMENT
MEDIUM…mni/diffusion/models/nextstep_1_1/modeling_flux_vae.py295 # Create a config-like object for compatibilityCOMMENT
MEDIUM…on/models/gr00t/dataio/state_action/action_chunking.py203 # Create a joint action chunkingSTRING
MEDIUM…on/models/gr00t/dataio/state_action/action_chunking.py399 # Create an end-effector action chunkingSTRING
MEDIUM…mni/diffusion/models/gr00t/dataio/state_action/pose.py172 # Create a 6-DOF joint configurationSTRING
MEDIUMvllm_omni/diffusion/hooks/sequence_parallel.py5# This module is adapted from HuggingFace diffusers library:COMMENT
MEDIUMvllm_omni/diffusion/worker/diffusion_worker.py1480 # Create the actual worker instanceCOMMENT
MEDIUMvllm_omni/benchmarks/data_modules/daily_omni_dataset.py672 # Create an iterator-like wrapper for compatibilityCOMMENT
MEDIUMvllm_omni/entrypoints/utils.py716# The following code detects if the process is running in a container and ifCOMMENT
MEDIUMvllm_omni/engine/arg_utils.py291 # Create a temp dir with a patched config.jsonCOMMENT
MEDIUMvllm_omni/inputs/data.py352 # This class is changed to only represent a single prompt requestCOMMENT
MEDIUM…ps/ComfyUI-vLLM-Omni/comfyui_vllm_omni/utils/logger.py57 # Create a copy with redacted valuesCOMMENT
Modern AI Meta-Vocabulary51 hits · 158 pts
SeverityFileLineSnippetContext
MEDIUMtests/e2e/offline_inference/test_audex_s2s.py43# the official temp-1.0 seed-0 draw hallucinates off-language text (yet closesCOMMENT
MEDIUM…uracy/sana_video/test_sana_video_pipeline_alignment.py396 # Isolate scheduler orchestration from native-attention error byCOMMENT
MEDIUM…ts/e2e/online_serving/test_cosyvoice3_tts_expansion.py33# Official CosyVoice zero-shot prompt audio and its transcript. Vendored underCOMMENT
MEDIUMtests/engine/test_orchestrator_event_driven.py71# Parity: the legacy scenario matrix, re-run through the event-driven loopCOMMENT
MEDIUMtests/engine/test_orchestrator_event_driven.py89 # KV/queue gauges under the event-driven loop.COMMENT
MEDIUMtests/helpers/assertions.py755 # may hallucinate extra context around the short utterance. UseCOMMENT
MEDIUMrequirements/common.txt39# Required for running the Cosmos3 model with guardrails enabled (default behavior).COMMENT
MEDIUMdocs/design/module/entrypoints.md91### ENTRY-INV-001: Entrypoints adapt but do not orchestrateCOMMENT
MEDIUMdocs/design/module/engine_orchestration.md57# Engine orchestrationCOMMENT
MEDIUMdocs/contributing/profiling.md329[Speech API](../serving/speech_api.md#orchestration-loop-experimental)) an idleCODE
MEDIUMdocs/contributing/model/adding_tts_model.md600 # Build context window: left_context + chunkCOMMENT
MEDIUMexamples/offline_inference/audex/speech_to_speech.py54# the official temp-1.0 seed-0 draw hallucinates off-language text (yetCOMMENT
MEDIUM…mples/offline_inference/text_to_image/text_to_image.py477 # Cosmos3 loads its (gated) guardrail models at build time, so the guardrailsCOMMENT
MEDIUM…mples/offline_inference/text_to_image/text_to_image.py478 # gate is an engine-level config (offline analog of the server's --no-guardrails).COMMENT
MEDIUM…mples/offline_inference/text_to_video/text_to_video.py525 # Cosmos3 loads its (gated) guardrail models at build time, so the guardrailsCOMMENT
MEDIUM…mples/offline_inference/text_to_video/text_to_video.py526 # gate is an engine-level config (offline analog of the server's --no-guardrails).COMMENT
MEDIUMexamples/offline_inference/text_to_speech/README.md432### Quick start (zero-shot, default voice)COMMENT
MEDIUM…offline_inference/text_to_speech/cosyvoice3/end2end.py18# Upstream zero-shot reference clipCOMMENT
MEDIUM…les/offline_inference/image_to_video/image_to_video.py566 # Cosmos3 loads its (gated) guardrail models at build time, so the guardrailsCOMMENT
MEDIUM…les/offline_inference/image_to_video/image_to_video.py567 # gate is an engine-level config (offline analog of the server's --no-guardrails).COMMENT
MEDIUMexamples/online_serving/audex/client.py46# official temp-1.0 seed-0 draw measurably hallucinates off-language text onCOMMENT
MEDIUMexamples/online_serving/audex/client.py299 # hallucinate off-language answers (see the module-level comment).COMMENT
MEDIUM…ine_serving/text_to_speech/cosyvoice3/speech_client.py30# Official CosyVoice zero-shot prompt and its transcript.COMMENT
MEDIUM…ctors/connectors/mooncake_transfer_engine_connector.py172 # The orchestration layer (get_connectors_config_for_stage /COMMENT
MEDIUMvllm_omni/deploy/lance.yaml6# stage orchestration.COMMENT
MEDIUM…model_executor/models/hunyuan_image3/hunyuan_image3.py1972 # The <timestep> slot at the head of each per-image scaffold is NOTCOMMENT
MEDIUM…i/model_executor/models/gepard/configuration_gepard.py102 # Carried for the cloning follow-up; zero-shot uses null_prefix only.COMMENT
MEDIUM…omni/model_executor/models/dots_tts/dots_tts_prompt.py68 # V1 zero-shot leaves this empty; preprocess() ignores it.COMMENT
MEDIUM…omni/model_executor/models/dots_tts/dots_tts_talker.py704 # upstream zero-shot emits every DiT patch including the firstCOMMENT
MEDIUM…omni/model_executor/models/dots_tts/dots_tts_talker.py1158 # runner.py:1114). Otherwise prefill scaffold (24 hidden × 1536COMMENT
MEDIUM…el_executor/models/personaplex/personaplex_temporal.py200 # Capacity = context window (the mask truncates at `context` anyway).COMMENT
MEDIUM…_omni/diffusion/models/hunyuan_image3/system_prompt.py195 # Think task: use chain-of-thought recaption promptSTRING
MEDIUMvllm_omni/diffusion/models/lance/pipeline_lance.py1854 # the chat scaffolding.COMMENT
MEDIUM…iffusion/models/qwen_image/pipeline_qwen_image_edit.py443 # The edit template contains fixed multimodal scaffolding around theCOMMENT
MEDIUM…/models/longcat_video/pipeline_longcat_video_avatar.py53# Whisper uses a 30-second context window: 30s * 16kHz raw samples,COMMENT
MEDIUMvllm_omni/diffusion/models/cosmos3/guardrails.py65 # ``model_config["guardrails"]`` is False.COMMENT
MEDIUMvllm_omni/diffusion/models/cosmos3/guardrails.py133 # caller has opted in to guardrails.COMMENT
MEDIUM…quantization/tools/merge_mxfp4_dualscale_checkpoint.py331 # BF16 base: provides the scaffold for non-MXFP4 tensors (norms, embeddings,COMMENT
MEDIUM…quantization/tools/merge_mxfp4_dualscale_checkpoint.py378 # Merge: base_state provides BF16 scaffold; MXFP4 tensors override their BF16 counterpartsCOMMENT
MEDIUMvllm_omni/entrypoints/async_omni.py56# Blocking-wait interval for the event-driven final-output drainCOMMENT
MEDIUMvllm_omni/entrypoints/async_omni.py891 # event-driven loop (vllm_omni/engine/orchestrator.py).COMMENT
MEDIUMvllm_omni/entrypoints/cli/serve.py874 # Disable safety guardrails for this server (currently only applicable for Cosmos3)COMMENT
MEDIUMvllm_omni/entrypoints/openai/serving_chat.py668 # scaffold.COMMENT
MEDIUMvllm_omni/entrypoints/openai/tts_adapters/voxcpm2.py91 # VoxCPM2 has no predefined speaker presets — "default" means zero-shotCOMMENT
MEDIUMvllm_omni/engine/orchestrator.py76# VLLM_OMNI_EVENT_DRIVEN_ORCH=1 switches the orchestration loop (and theCOMMENT
MEDIUMvllm_omni/engine/orchestrator.py78# 1 ms poll cadence to event-driven wakeups: one reader task per live LLM stageCOMMENT
MEDIUMvllm_omni/engine/orchestrator.py84# How often the event-driven loop reconciles its reader-task set againstCOMMENT
MEDIUMvllm_omni/engine/orchestrator.py166 # min_tokens wider than the leftover context window never reachesCOMMENT
MEDIUMvllm_omni/engine/orchestrator.py593 # The orchestration loop isolates per-replica EngineDeadErrorCOMMENT
MEDIUMvllm_omni/engine/orchestrator.py675 # orchestration loop observes _shutdown_event and exits.COMMENT
MEDIUMvllm_omni/inputs/preprocess.py70 # defaults and scaffold).COMMENT
Redundant / Tautological Comments85 hits · 128 pts
SeverityFileLineSnippetContext
LOWtools/pre_commit/check_test_marks.py51# Check if a file is located under tests/ and matches test_<something>.pyCOMMENT
LOWtools/pre_commit/check_forbidden_imports.py249 # Check if it's allowedCOMMENT
LOWtests/distributed/omni_connectors/test_kv_flow.py126 # Check if data was put into connectorCOMMENT
LOW…_executor/models/higgs_audio_v3/test_higgs_audio_v3.py118 # Set weights to identity-like pattern so we can verify offsetsCOMMENT
LOW…ts/diffusion/models/flux2/test_flux2_transformer_tp.py243 # Check if to_out related weights are loadedCOMMENT
LOWtests/benchmarks/patch/test_patch.py470 # Check if attribute exists and has a value (should be 0.0 or similar default)COMMENT
LOWtests/e2e/offline_inference/compute_lpips.py95 # Print results tableCOMMENT
LOWtests/e2e/features/helpers/custom_pipeline.py312 # Check if model is a local pathCOMMENT
LOWtests/helpers/stage_config.py90 # Check if next key is a digit (list index) or string (dict key)COMMENT
LOWtests/helpers/stage_config.py224 # Check if this is a simple key (not dot-separated)COMMENT
LOWexamples/offline_inference/step_audio2/end2end.py203 # Check if it's a local pathCOMMENT
LOWexamples/offline_inference/image_to_image/image_edit.py639 # Check if this is a layered output (list of images)COMMENT
LOWexamples/offline_inference/image_to_image/image_edit.py650 # Check if this is a layered output (list of images)COMMENT
LOW…ai_chat_completion_client_for_multimodal_generation.py57 # Check if it's a URL (starts with http:// or https://)COMMENT
LOW…ai_chat_completion_client_for_multimodal_generation.py96 # Check if it's a URL (starts with http:// or https://)COMMENT
LOW…ai_chat_completion_client_for_multimodal_generation.py133 # Check if it's a URL (starts with http:// or https://)COMMENT
LOW…ai_chat_completion_client_for_multimodal_generation.py64 # Check if it's a URL (starts with http:// or https://)COMMENT
LOW…ai_chat_completion_client_for_multimodal_generation.py103 # Check if it's a URL (starts with http:// or https://)COMMENT
LOW…ai_chat_completion_client_for_multimodal_generation.py141 # Check if it's already a base64 data URLCOMMENT
LOW…ai_chat_completion_client_for_multimodal_generation.py145 # Check if it's a URL (starts with http:// or https://)COMMENT
LOWexamples/online_serving/lance/openai_chat_client.py49 # Check if local fileCOMMENT
LOWexamples/online_serving/qwen3_omni/run_gradio_demo.sh154 # Check if startup flag file exists (startup complete)COMMENT
LOWexamples/online_serving/qwen3_omni/run_gradio_demo.sh163 # Check if server process is still runningCOMMENT
LOWexamples/online_serving/qwen2_5_omni/run_gradio_demo.sh154 # Check if startup flag file exists (startup complete)COMMENT
LOWexamples/online_serving/qwen2_5_omni/run_gradio_demo.sh163 # Check if server process is still runningCOMMENT
LOW…online_serving/text_to_speech/qwen3_tts/gradio_demo.py365 # Check if this is a request ID referencing a stored payloadCOMMENT
LOWbenchmarks/build_dataset/extract_tts_prompts.py45 # Print result statsCOMMENT
LOWbenchmarks/build_dataset/extract_tts_prompts.py64 # Check if input file existsCOMMENT
LOWvllm_omni/platforms/cuda/platform.py91 # Check if FA packages are availableCOMMENT
LOWvllm_omni/platforms/musa/platform.py72 # Check if FA packages are availableCOMMENT
LOWvllm_omni/platforms/rocm/platform.py95 # Check if aiter is available for Flash Attention supportCOMMENT
LOW…ctors/connectors/mooncake_transfer_engine_connector.py1104 # Check if this is a query requestCOMMENT
LOW…mni/distributed/omni_connectors/utils/serialization.py210 # Check if this looks like an OmniRequestOutput (check before RequestOutputCOMMENT
LOW…mni/distributed/omni_connectors/utils/serialization.py215 # Check if this looks like a RequestOutputCOMMENT
LOW…mni/distributed/omni_connectors/utils/serialization.py219 # Check if this looks like a CompletionOutputCOMMENT
LOWvllm_omni/plugins/__init__.py34 # Check if the only discovered plugin is the default one.COMMENT
LOW…model_executor/models/hunyuan_image3/hunyuan_image3.py1813 # Check if T dimension is 1, then squeeze itCOMMENT
LOW…m_omni/model_executor/models/glm_image/glm_image_ar.py178 # Check if we're in a subdirectory (vision_language_encoder)COMMENT
LOW…m_omni/model_executor/models/glm_image/glm_image_ar.py891 # Check if we have images (i2i mode)COMMENT
LOW…m_omni/model_executor/models/glm_image/glm_image_ar.py945 # Check if there are any images to processCOMMENT
LOW…m_omni/model_executor/models/glm_image/glm_image_ar.py2808 # Check if this is a text-to-image request:COMMENT
LOW…m_omni/model_executor/models/glm_image/glm_image_ar.py2840 # Check if there's a token after this start and it's an image tokenCOMMENT
LOW…odel_executor/models/mimo_audio/mimo_audio_code2wav.py955 # Check if in CUDA graph capture phaseCOMMENT
LOW…lm_omni/model_executor/models/mimo_audio/mimo_audio.py879 # Check if in CUDA graph capture phaseCOMMENT
LOW…_executor/models/minicpmo_4_5/minicpmo_4_5_omni_llm.py461 # Check if the args are a device or a dtypeSTRING
LOW…mni/model_executor/stage_input_processors/glm_image.py338 # Check if elements are tensors or Python listsCOMMENT
LOW…_omni/model_executor/stage_input_processors/glm_tts.py340 # Check if enough tokens accumulated for next chunkCOMMENT
LOW…i/diffusion/offloader/distributed_layerwise_backend.py1196 # Check if this buffer is non-persistent on its OWNING moduleCOMMENT
LOWvllm_omni/diffusion/attention/parallel/ulysses.py366 # Check if Ring Attention is also active (Hybrid mode)COMMENT
LOWvllm_omni/diffusion/cache/cachedit/model_specific.py57 # Set timesteps to calculate the splitCOMMENT
LOWvllm_omni/diffusion/distributed/sp_plan.py403 # Check if it's an output specification (SequenceParallelOutput or list/tuple thereof)COMMENT
LOWvllm_omni/diffusion/distributed/cfg_parallel.py305 # Assign branches to ranks via round-robinCOMMENT
LOW…lm_omni/diffusion/models/wan2_2/pipeline_wan2_2_i2v.py228 # Check if this is a two-stage model (MoE with transformer_2)COMMENT
LOWvllm_omni/diffusion/models/wan2_2/pipeline_wan2_2.py352 # Check if this is a two-stage model (MoE with transformer_2)COMMENT
LOWvllm_omni/diffusion/models/flux/pipeline_flux.py106 # Check if model is a local pathCOMMENT
LOW…_omni/diffusion/models/glm_image/pipeline_glm_image.py753 # Check if prior_token_ids are provided externally (from AR stage in multistage mode)COMMENT
LOW…ni/diffusion/models/glm_image/glm_image_transformer.py582 # Check if SP is enabledCOMMENT
LOWvllm_omni/diffusion/models/flux2/pipeline_flux2.py235 # Check if it's a list of lists or a list of imagesCOMMENT
LOWvllm_omni/diffusion/models/flux2/pipeline_flux2.py381 # Check if model is a local pathCOMMENT
LOW…ion/models/qwen_image/pipeline_qwen_image_edit_plus.py214 # Check if model is a local pathCOMMENT
25 more matches not shown…
AI Slop Vocabulary31 hits · 78 pts
SeverityFileLineSnippetContext
MEDIUM…executor/models/qwen3_tts/test_code_predictor_dtype.py769 # The test harness mocks ``default_weight_loader`` as a no-op; install aCOMMENT
MEDIUM…executor/models/qwen3_tts/test_code_predictor_dtype.py830 # The test harness mocks ``default_weight_loader`` as a no-op; install aCOMMENT
MEDIUM…executor/models/qwen3_tts/test_code_predictor_dtype.py1136 # Find any actual self_attn layer prefix, to make the test robustCOMMENT
MEDIUMtests/model_tests/diffusion/test_alignment.py64 # The common harness does not yet define an image-to-video task or supplyCOMMENT
MEDIUMtests/diffusion/kernels/mot/test_mot_linear.py320 # Cosine similarity is the primary correctness gate: robust to scaleCOMMENT
LOW…iffusion/distributed/test_autoencoder_kl_wan_encode.py47 # Simple mock: just return the inputCOMMENT
MEDIUMtests/diffusion/models/pi0/test_pi0_units.py494# vision_tower / embed_scale version-robust paths against regressions.COMMENT
MEDIUM…s/e2e/offline_inference/test_sensenova_u1_expansion.py48# regardless of GQA, so this is the robust LoRA target. Verify against config.json.COMMENT
LOW…e/features/fullduplex/test_personaplex_server_lease.py63 # The drain must genuinely wait for the worker, not just return.COMMENT
MEDIUM…line_serving/text_to_speech/voxtral_tts/gradio_demo.py43# Default fallback voices - comprehensive listCOMMENT
LOWvllm_omni/metrics/stat_logger.py114 # rewritten label set already, so just pass through.COMMENT
LOWvllm_omni/platforms/cuda/platform.py319 # users can just use IR op priority directlyCOMMENT
MEDIUM…ni/platforms/npu/worker/npu_generation_model_runner.py253 # We simply utilize the implementation in vLLM.COMMENT
MEDIUMvllm_omni/platforms/npu/worker/npu_ar_model_runner.py610 # We simply utilize the implementation in vLLM.COMMENT
LOW…ctors/connectors/mooncake_transfer_engine_connector.py314 # We don't actually connect, just use the socket to get routing infoCOMMENT
LOWvllm_omni/config/model.py66 # would just use itself as prefix — safe but worthCOMMENT
MEDIUMvllm_omni/model_executor/models/utils.py10 """Make ``trust_remote_code`` weight loading robust to the transformers 5.9STRING
LOW…m_omni/model_executor/models/glm_image/glm_image_ar.py3140 # For intermediate tensors (PP), just return hidden statesCOMMENT
LOW…lm_omni/model_executor/models/mimo_audio/mimo_audio.py475 # If no bos/eos tokens, just use audio tokensCOMMENT
MEDIUM…els/nemotron_voicechat/nemo_vendored/duplex_ear_tts.py325 # EOS dropout to make the model more robustCOMMENT
MEDIUM…els/nemotron_voicechat/nemo_vendored/duplex_ear_tts.py342 # BOS dropout to make the model more robustCOMMENT
MEDIUM…els/nemotron_voicechat/nemo_vendored/duplex_ear_tts.py361 # BOS dropout to make the model more robustCOMMENT
LOW…del_executor/models/indextts2/gpt/conformer_encoder.py141 # It's better we just return None if no cache is required,COMMENT
MEDIUM…m_omni/diffusion/attention/backends/ring/ring_utils.py107 # Initialize LSE with robust logic (same as _update)COMMENT
MEDIUMvllm_omni/diffusion/models/pi0/modeling_pi0.py635 # embedding (version-robust: self-scaled at transformers ≥ 5.4, manuallyCOMMENT
LOWvllm_omni/diffusion/hooks/base.py313 # Otherwise just call the original forward.COMMENT
MEDIUMvllm_omni/diffusion/lora/layers/base_linear.py51 # `__dict__` for robust lookups in `__getattr__`.COMMENT
MEDIUMvllm_omni/benchmarks/data_modules/daily_omni_eval.py133 # ``\b`` after the letter avoids "Because"/"Definitely" false positivesCOMMENT
MEDIUMvllm_omni/entrypoints/openai/serving_chat.py1438 # Always track previous_texts for comprehensive output loggingCOMMENT
MEDIUMvllm_omni/entrypoints/openai/serving_chat.py1905 # Update for comprehensive logging even in simple caseCOMMENT
LOWvllm_omni/worker/gpu_ar_model_runner.py2100 # just use zeros for the draft tokens.COMMENT
Verbosity Indicators33 hits · 64 pts
SeverityFileLineSnippetContext
LOWtests/core/sched/test_generation_scheduler_restore.py52 # Step 1: process_pending_chunks moves req-B outCOMMENT
LOWtests/core/sched/test_generation_scheduler_restore.py57 # Step 2: simulate the try/except/finally patternCOMMENT
LOWtests/core/sched/test_generation_scheduler_restore.py66 # Step 3: verify request is restoredCOMMENT
LOWtests/diffusion/attention/test_attention_sp.py473 # Step 1: Run without SP (baseline with ulysses_degree=1, ring_degree=1)COMMENT
LOWtests/diffusion/attention/test_attention_sp.py502 # Step 2: Run with SP enabledCOMMENT
LOWtests/diffusion/attention/test_attention_sp.py534 # Step 3: Verify input consistency and compare outputsCOMMENT
LOWtests/e2e/offline_inference/test_moss_tts_realtime.py166 # Step 1: locate the realtime processor module in the snapshot.COMMENT
LOWtests/e2e/offline_inference/test_moss_tts_realtime.py191 # Step 2: encode the reference audio clip via MOSS-Audio-Tokenizer.COMMENT
LOWtests/e2e/offline_inference/test_moss_tts_realtime.py209 # Step 3: build the (L, 17) prefill grid.COMMENT
LOW…ures/custom_pipeline/test_async_omni_collective_rpc.py126 # Step 1: list_loras (the call that was crashing)COMMENT
LOW…ures/custom_pipeline/test_async_omni_collective_rpc.py130 # Step 2: generate (should still work after list_loras)COMMENT
LOWvllm_omni/core/prefix_cache.py211 # Step 1: consume the previous step's write FIRST. The previousCOMMENT
LOWvllm_omni/core/prefix_cache.py220 # Step 2: schedule the D2H on the copy stream. PyTorch implicitlyCOMMENT
LOWvllm_omni/core/prefix_cache.py243 # Step 3: stash for next step's consume.COMMENT
LOW…utor/models/ming_flash_omni/ming_flash_omni_thinker.py934 # Step 1: if vision/audio features were extracted, merge them intoCOMMENT
LOW…utor/models/ming_flash_omni/ming_flash_omni_thinker.py945 # Step 2: image-gen query tokens are appended at the tail of the promptCOMMENT
LOW…odel_executor/stage_input_processors/higgs_audio_v3.py139 # Step 1: Revert delay patternCOMMENT
LOW…odel_executor/stage_input_processors/higgs_audio_v3.py147 # Step 2: Replace out-of-range codes (BOC=1024, EOC=1025, -1) with 0.COMMENT
LOW…odel_executor/stage_input_processors/higgs_audio_v3.py157 # Step 3: Trim the last frame. After de-delay, the final frameCOMMENT
LOWvllm_omni/diffusion/layers/mot/ops/mot_rms_norm.py25 # Step 0: MoT RoutingCOMMENT
LOWvllm_omni/diffusion/layers/mot/ops/mot_rms_norm.py45 # Step 1: Compute sum of squares in float32 to avoid overflowCOMMENT
LOWvllm_omni/diffusion/layers/mot/ops/mot_rms_norm.py57 # Step 2: Compute RMS (root mean square) in float32COMMENT
LOWvllm_omni/diffusion/layers/mot/ops/mot_rms_norm.py62 # Step 3: Normalize and apply weightCOMMENT
LOWvllm_omni/diffusion/models/gr00t/policy.py336 # Step 1: Split batched observation into individual observationsCOMMENT
LOWvllm_omni/diffusion/models/gr00t/policy.py340 # Step 2: Process each observation through the VLA processorCOMMENT
LOWvllm_omni/diffusion/models/gr00t/policy.py348 # Step 3: Collate processed inputs into a single batch for modelCOMMENT
LOWvllm_omni/diffusion/models/gr00t/policy.py352 # Step 4: Run model inference to predict actionsCOMMENT
LOWvllm_omni/diffusion/models/gr00t/policy.py357 # Step 5: Decode actions from normalized space back to physical unitsCOMMENT
LOW…ls/gr00t/dataio/state_action/state_action_processor.py310 # Step 1: Convert absolute actions to relative (if needed)COMMENT
LOW…ls/gr00t/dataio/state_action/state_action_processor.py343 # Step 2: Normalize actionsCOMMENT
LOW…ls/gr00t/dataio/state_action/state_action_processor.py395 # Step 1: Unnormalize actionsCOMMENT
LOW…ls/gr00t/dataio/state_action/state_action_processor.py418 # Step 2: Convert relative actions to absolute (if needed)COMMENT
LOWvllm_omni/quantization/tools/merge_mxfp8_checkpoint.py231 # Step 1: Copy original model, skipping transformer dirs (will be replaced).COMMENT
Hallucination Indicators6 hits · 62 pts
SeverityFileLineSnippetContext
CRITICAL…executor/models/minicpmo_4_5/test_code2wav_batching.py364 token2wav.flow.decoder.estimator.cfg_batches.clear()CODE
CRITICALtests/helpers/client.py1479 with self.client.audio.speech.with_streaming_response.create(CODE
CRITICALtests/helpers/client.py1519 with self.client.audio.speech.with_streaming_response.create(CODE
CRITICALexamples/offline_inference/hunyuan_image3/README.md153The shared `vllm_omni.diffusion.models.hunyuan_image3.prompt_utils.build_prompt_tokens()`CODE
CRITICAL…executor/models/qwen2_5_omni/qwen2_5_omni_token2wav.py1231 self.codec_embed_size = self.code2wav_dit_model.text_embed.codec_embed.weight.size(0)STRING
CRITICALvllm_omni/diffusion/models/gr00t/modeling/gr00t_n1d7.py326 self.model.model.language_model.layers.pop(-1)CODE
Example Usage Blocks30 hits · 45 pts
SeverityFileLineSnippetContext
LOWtools/run_merge_jobs.sh21# Usage:COMMENT
LOWtools/run_ready_jobs.sh21# Usage:COMMENT
LOWtools/nightly/run_nightly_jobs.sh55# Usage:COMMENT
LOWtests/e2e/offline_inference/run_quantization_e2e.sh10# Usage:COMMENT
LOW…_inference/qwen3_omni/run_single_prompt_async_chunk.sh12# Usage:COMMENT
LOW…ference/qwen3_omni/run_multiple_prompts_async_chunk.sh8# Usage:COMMENT
LOW…ine_inference/qwen2_5_omni/strategy_stage_replica.yaml14# Usage:COMMENT
LOW…mples/offline_inference/qwen2_5_omni/strategy_tp2.yaml12# Usage:COMMENT
LOW…les/online_serving/replica_data_parallel/run_server.sh4# Usage:COMMENT
LOWexamples/online_serving/audex/run_server.sh13# Usage:COMMENT
LOW…amples/online_serving/text_to_video/run_server_ltx2.sh7# Usage:COMMENT
LOWexamples/online_serving/qwen3_omni/run_gradio_demo.sh4# Usage:COMMENT
LOWexamples/online_serving/qwen2_5_omni/run_gradio_demo.sh4# Usage:COMMENT
LOW…erving/text_to_speech/moss_tts_nano/run_gradio_demo.sh4# Usage:COMMENT
LOW…ine_serving/text_to_speech/moss_tts_nano/run_server.sh4# Usage:COMMENT
LOW…rving/text_to_speech/ming_flash_omni_tts/run_server.sh4# Usage:COMMENT
LOW…online_serving/text_to_speech/cosyvoice3/run_server.sh4# Usage:COMMENT
LOW…line_serving/text_to_speech/glm_tts/run_gradio_demo.sh4# Usage:COMMENT
LOW…es/online_serving/text_to_speech/glm_tts/run_server.sh4# Usage:COMMENT
LOW…/online_serving/text_to_speech/omnivoice/run_server.sh4# Usage:COMMENT
LOW…_serving/text_to_speech/fish_speech/run_gradio_demo.sh4# Usage:COMMENT
LOW…nline_serving/text_to_speech/fish_speech/run_server.sh4# Usage:COMMENT
LOW…s/online_serving/text_to_speech/ming_tts/run_server.sh4# Usage:COMMENT
LOW…ne_serving/text_to_speech/higgs_audio_v2/run_server.sh8# Usage:COMMENT
LOW…ne_serving/text_to_speech/higgs_audio_v3/run_server.sh6# Usage:COMMENT
LOW…ne_serving/text_to_speech/qwen3_tts/run_gradio_demo.sh4# Usage:COMMENT
LOW…/online_serving/text_to_speech/qwen3_tts/run_server.sh4# Usage:COMMENT
LOW.buildkite/release/scripts/publish-release-images.sh6# Usage:COMMENT
LOWvllm_omni/deploy/qwen3_omni_moe_mori_intranode.yaml26# Usage:COMMENT
LOWvllm_omni/deploy/indextts2_low_latency.yaml8# Usage:COMMENT
Dead Code8 hits · 16 pts
SeverityFileLineSnippetContext
MEDIUMtests/diffusion/test_inline_stage_diffusion_client.py142CODE
MEDIUM…/entrypoints/openai_api/test_serving_audio_generate.py418CODE
MEDIUM…/entrypoints/openai_api/test_serving_audio_generate.py477CODE
MEDIUM…/entrypoints/openai_api/test_serving_audio_generate.py492CODE
MEDIUMtests/entrypoints/openai_api/test_serving_speech.py3005CODE
MEDIUMtests/entrypoints/openai_api/test_serving_speech.py4815CODE
MEDIUMtests/entrypoints/openai_api/test_image_server.py615CODE
MEDIUMtests/entrypoints/openai_api/test_image_server.py937CODE
Magic Placeholder Names1 hit · 5 pts
SeverityFileLineSnippetContext
HIGH.github/ISSUE_TEMPLATE/400-bug-report.yml20 Consider redacting or replacing sensitive values with placeholders like `<YOUR_TOKEN_HERE>` when sharing configuraCODE
Overly Generic Function Names3 hits · 2 pts
SeverityFileLineSnippetContext
LOWtests/diffusion/test_stage_diffusion_proc.py100 async def run_task(req_data):CODE
LOWtests/diffusion/test_diffusion_engine.py811 async def run_task(rid):CODE
LOWexamples/offline_inference/lance/gradio_demo.py794def run_task(CODE
Fake / Example Data2 hits · 2 pts
SeverityFileLineSnippetContext
LOW…lm_omni/experimental/fullduplex/personaplex/runtime.py301 dummy_user = self._initial[9:].view(1, 8).expand(B, 8)CODE
LOW…lm_omni/experimental/fullduplex/personaplex/runtime.py309 ran, _ = self._tick(dummy_user, dummy_moshi, zero_text, all_rows, embed_override=(ov, all_rows))CODE
Slop Phrases3 hits · 2 pts
SeverityFileLineSnippetContext
LOW…_executor/models/minicpmo_4_5/minicpmo_4_5_omni_llm.py1904 # When lifting this restriction, don't forget to eitherSTRING
LOW…_executor/models/minicpmo_4_5/minicpmo_4_5_omni_llm.py1904 # When lifting this restriction, don't forget to eitherSTRING
MEDIUMvllm_omni/diffusion/models/bagel/bagel_transformer.py369 although it's worth noting that this is currently only usedSTRING