OptimalScale/LMFlow

12.4

Adjusted Score

12.4

Raw Score

100%

Time Factor

2026-08-07

Last Push

8.5K

Stars

Python

Language

45.2K

Lines of Code

281

Files

402

Pattern Hits

2026-08-09

Scan Date

0.11

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

Longitudinal tracking requires multiple scan runs. Once this repository is re-scanned after new commits land, this chart will visualise how the synthetic code signal evolves over time — enabling you to detect whether AI authorship is growing, stabilising, or being actively corrected by human engineers.

No multi-scan history yet — run the scanner again to build trend data.

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 0HIGH 31MEDIUM 14LOW 357

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 402 distinct pattern matches across 18 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.

Cross-File Repetition28 hits · 140 pts

Severity	File	Snippet	Context
HIGH	…SA-diffusion/diffusion_dpo/train_diffusion_dpo_lisa.py	update target parameters to be closer to those of source parameters using an exponential moving average. :param target_p	STRING
HIGH	…ent_consistency_model/train_lcm_distill_sd_wds_lisa.py	update target parameters to be closer to those of source parameters using an exponential moving average. :param target_p	STRING
HIGH	…ent_consistency_model/train_lcm_distill_sd_wds_lora.py	update target parameters to be closer to those of source parameters using an exponential moving average. :param target_p	STRING
HIGH	tests/datasets/test_dataset.py	a one-line summary of the module or program, terminated by a period. leave one blank line. the rest of this docstring sh	STRING
HIGH	tests/models/test_hf_decoder_model.py	a one-line summary of the module or program, terminated by a period. leave one blank line. the rest of this docstring sh	STRING
HIGH	examples/evaluation.py	a one-line summary of the module or program, terminated by a period. leave one blank line. the rest of this docstring sh	STRING
HIGH	examples/finetune_multi_modal.py	a one-line summary of the module or program, terminated by a period. leave one blank line. the rest of this docstring sh	STRING
HIGH	examples/finetune.py	a one-line summary of the module or program, terminated by a period. leave one blank line. the rest of this docstring sh	STRING
HIGH	examples/benchmarking.py	a one-line summary of the module or program, terminated by a period. leave one blank line. the rest of this docstring sh	STRING
HIGH	src/lmflow/models/encoder_decoder_model.py	a one-line summary of the module or program, terminated by a period. leave one blank line. the rest of this docstring sh	STRING
HIGH	src/lmflow/models/decoder_model.py	a one-line summary of the module or program, terminated by a period. leave one blank line. the rest of this docstring sh	STRING
HIGH	contrib/tool-finetune/function_call_finetune.py	tokenize the full dataset. parameters ------------ dataset : lmflow.datasets.dataset. args : optional. positional argume	STRING
HIGH	src/lmflow/models/hf_decoder_model.py	tokenize the full dataset. parameters ------------ dataset : lmflow.datasets.dataset. args : optional. positional argume	STRING
HIGH	src/lmflow/models/hf_text_regression_model.py	tokenize the full dataset. parameters ------------ dataset : lmflow.datasets.dataset. args : optional. positional argume	STRING
HIGH	examples/chatbot_gradio.py	a simple shell chatbot implemented with lmflow apis.	STRING
HIGH	examples/inference.py	a simple shell chatbot implemented with lmflow apis.	STRING
HIGH	examples/chatbot.py	a simple shell chatbot implemented with lmflow apis.	STRING
HIGH	src/lmflow/optim/lamb.py	performs a single optimization step. arguments: closure: a closure that reevaluates the model and returns the loss.	STRING
HIGH	src/lmflow/optim/lars.py	performs a single optimization step. arguments: closure: a closure that reevaluates the model and returns the loss.	STRING
HIGH	src/lmflow/optim/adamp.py	performs a single optimization step. arguments: closure: a closure that reevaluates the model and returns the loss.	STRING
HIGH	src/lmflow/optim/yogi.py	performs a single optimization step. arguments: closure: a closure that reevaluates the model and returns the loss.	STRING
HIGH	src/lmflow/optim/radam.py	performs a single optimization step. arguments: closure: a closure that reevaluates the model and returns the loss.	STRING
HIGH	src/lmflow/optim/adabound.py	performs a single optimization step. arguments: closure: a closure that reevaluates the model and returns the loss.	STRING
HIGH	src/lmflow/optim/sgdp.py	performs a single optimization step. arguments: closure: a closure that reevaluates the model and returns the loss.	STRING
HIGH	src/lmflow/optim/sgd_schedule_free.py	performs a single optimization step. args: closure (callable, optional): a closure that reevaluates the model and return	STRING
HIGH	src/lmflow/optim/adabelief.py	performs a single optimization step. args: closure (callable, optional): a closure that reevaluates the model and return	STRING
HIGH	src/lmflow/optim/adamw_schedule_free.py	performs a single optimization step. args: closure (callable, optional): a closure that reevaluates the model and return	STRING
HIGH	src/lmflow/optim/muon.py	performs a single optimization step. args: closure (callable, optional): a closure that reevaluates the model and return	STRING

Hyper-Verbose Identifiers114 hits · 120 pts

Severity	File	Line	Snippet	Context
LOW	…al/LISA-diffusion/diffusion_dpo/train_diffusion_dpo.py	69	def import_model_class_from_model_name_or_path(pretrained_model_name_or_path: str, revision: str):	CODE
LOW	…SA-diffusion/diffusion_dpo/train_diffusion_dpo_lisa.py	90	def import_model_class_from_model_name_or_path(pretrained_model_name_or_path: str, revision: str):	CODE
LOW	…ent_consistency_model/train_lcm_distill_sd_wds_lisa.py	134	def tarfile_to_samples_nothrow(src, handler=wds.warn_and_continue):	CODE
LOW	…ent_consistency_model/train_lcm_distill_sd_wds_lisa.py	375	def scalings_for_boundary_conditions(timestep, sigma_data=0.5, timestep_scaling=10.0):	CODE
LOW	…ent_consistency_model/train_lcm_distill_sd_wds_lisa.py	383	def get_predicted_original_sample(model_output, timesteps, sample, prediction_type, alphas, sigmas):	CODE
LOW	…ent_consistency_model/train_lcm_distill_sd_wds_lisa.py	468	def import_model_class_from_model_name_or_path(	CODE
LOW	…ent_consistency_model/train_lcm_distill_sd_wds_lora.py	77	def get_module_kohya_state_dict(module, prefix: str, dtype: torch.dtype, adapter_name: str = "default"):	CODE
LOW	…ent_consistency_model/train_lcm_distill_sd_wds_lora.py	129	def tarfile_to_samples_nothrow(src, handler=wds.warn_and_continue):	CODE
LOW	…ent_consistency_model/train_lcm_distill_sd_wds_lora.py	356	def scalings_for_boundary_conditions(timestep, sigma_data=0.5, timestep_scaling=10.0):	CODE
LOW	…ent_consistency_model/train_lcm_distill_sd_wds_lora.py	364	def get_predicted_original_sample(model_output, timesteps, sample, prediction_type, alphas, sigmas):	CODE
LOW	…ent_consistency_model/train_lcm_distill_sd_wds_lora.py	448	def import_model_class_from_model_name_or_path(	CODE
LOW	tests/pipeline/test_evaluator_messages.py	9	def test_nll_unsupported_dataset_message_includes_type(self):	CODE
LOW⚡	tests/pipeline/test_auto_pipeline.py	14	def test_get_evaluator_pipeline(self):	CODE
LOW⚡	tests/pipeline/test_auto_pipeline.py	23	def test_get_finetuner_pipeline(self):	CODE
LOW⚡	tests/pipeline/test_auto_pipeline.py	32	def test_get_inferencer_pipeline(self):	CODE
LOW⚡	tests/pipeline/test_auto_pipeline.py	41	def test_get_unsupported_pipeline(self):	CODE
LOW	tests/pipeline/test_sglang_infernecer.py	22	def sglang_test_inferencer_args() -> InferencerArguments:	CODE
LOW	tests/pipeline/test_memory_safe_vllm_inferencer.py	66	def test_inference_detokenize(self):	CODE
LOW	tests/pipeline/test_vllm_inferencer.py	69	def test_override_with_inference_args(self, model_args, inferencer_args):	CODE
LOW	tests/pipeline/test_vllm_inferencer.py	131	def test_creates_dataproto_with_repeat(self):	CODE
LOW	tests/pipeline/test_vllm_inferencer.py	167	def vllm_test_inferencer_args() -> InferencerArguments:	CODE
LOW	tests/pipeline/test_vllm_inferencer.py	226	def test_vllm_inferencer_save_load(	CODE
LOW	tests/datasets/conftest.py	8	def dataset_inference_conversation() -> Dataset:	CODE
LOW	tests/datasets/conftest.py	17	def dataset_inference_conversation_batch() -> Dataset:	CODE
LOW	tests/datasets/test_dataset.py	56	def test_create_from_dict_bad_type(self):	CODE
LOW⚡	tests/utils/test_envs.py	17	def test_is_accelerate_env_false_without_prefix(self):	CODE
LOW⚡	tests/utils/test_envs.py	21	def test_is_accelerate_env_true_with_prefix(self):	CODE
LOW⚡	tests/utils/test_envs.py	25	def test_is_accelerate_env_false_when_accelerate_not_prefix(self):	CODE
LOW⚡	tests/utils/test_envs.py	31	def test_get_device_name_cpu_when_cuda_unavailable(self, _mock_cuda: object):	CODE
LOW⚡	tests/utils/test_envs.py	35	def test_get_device_name_cuda_when_cuda_available(self, _mock_cuda: object):	CODE
LOW⚡	tests/utils/test_envs.py	38	def test_get_torch_device_matches_device_name(self):	CODE
LOW⚡	tests/utils/test_envs.py	48	def test_get_torch_device_fallback_returns_cuda_on_attribute_error(self, _mock_name: object):	CODE
LOW⚡	tests/utils/test_envs.py	56	def test_require_cuda_for_gpu_mode_raises_when_cuda_unavailable(self, _mock_cuda: object):	CODE
LOW⚡	tests/utils/test_envs.py	63	def test_set_cuda_device_calls_torch_set_device(self, mock_set_device: object, _mock_cuda: object):	CODE
LOW⚡	tests/utils/test_envs.py	69	def test_set_cuda_device_raises_without_cuda(self, mock_set_device: object, _mock_cuda: object):	CODE
LOW	tests/utils/test_conversation_template.py	131	def test_encode_conversation_singleturn(self):	CODE
LOW	tests/utils/test_conversation_template.py	140	def test_encode_conversation_multiturn(self):	CODE
LOW	tests/utils/test_conversation_template.py	156	def test_encode_conversation_singleturn(self):	CODE
LOW	tests/utils/test_conversation_template.py	165	def test_encode_conversation_multiturn(self):	CODE
LOW	tests/utils/test_conversation_formatter.py	7	def test_format_string_component(self):	CODE
LOW	tests/models/test_hf_decoder_model.py	290	def make_gt_from_conversation_ids(conversation_ids):	CODE
LOW	tests/models/test_hf_decoder_model.py	303	def make_gt_from_conversation_ids_batch(batched_conversation_ids):	CODE
LOW	tests/models/test_hf_decoder_model.py	367	def test_tokenize_text_only_multiple(self):	CODE
LOW	tests/models/test_hf_decoder_model.py	411	def test_tokenize_conversation(self):	CODE
LOW	tests/models/test_hf_decoder_model.py	516	def test_tokenize_conversation_multiple(self):	CODE
LOW	tests/models/test_auto_model.py	24	def test_get_text_regression_model(self):	STRING
LOW	tests/models/test_auto_model.py	40	def test_get_unsupported_model(self):	CODE
LOW	contrib/tool-finetune/function_call_finetune.py	158	def conversation_tokenize_function(	CODE
LOW	examples/benchmarking.py	115	def is_lmflow_local_benchmarking(dataset_name):	CODE
LOW⚡	examples/benchmarking.py	142	def is_lm_evaluation_benchmarking(dataset_name):	CODE
LOW⚡	examples/benchmarking.py	149	def run_lmflow_local_benchmarking(	CODE
LOW	examples/benchmarking.py	191	def run_lm_evaluation_benchmarking(dataset_name, model_name):	CODE
LOW	scripts/archive/convert_llama_weights_to_hf.py	59	def compute_intermediate_size(n):	CODE
LOW	src/lmflow/pipeline/dpo_aligner.py	55	def return_prompt_and_responses(samples) -> dict[str, str]:	CODE
LOW	src/lmflow/pipeline/iterative_dpo_aligner.py	120	def _do_target_model_inference(	CODE
LOW	src/lmflow/pipeline/iterative_dpo_aligner.py	145	def _build_response_generator(	CODE
LOW	src/lmflow/pipeline/iterative_dpo_aligner.py	168	def _dataproto_to_text_to_textlist_instances(res: DataProto) -> list[dict]:	CODE
LOW	src/lmflow/pipeline/iterative_dpo_aligner.py	185	def _do_reward_model_inference(	CODE
LOW	src/lmflow/pipeline/iterative_dpo_aligner.py	230	def _parse_target_model_inference_args(	CODE
LOW	src/lmflow/pipeline/iterative_dpo_aligner.py	239	def _parse_reward_model_inference_args(	CODE
54 more matches not shown…

Deep Nesting94 hits · 86 pts

Severity	File	Line	Context
LOW	experimental/RAFT-diffusion/train_text_to_image_lora.py	361	CODE
LOW	…fusion/instruct_pix2pix/train_instruct_pix2pix_lisa.py	404	CODE
LOW	…al/LISA-diffusion/diffusion_dpo/train_diffusion_dpo.py	465	CODE
LOW	…SA-diffusion/diffusion_dpo/train_diffusion_dpo_lisa.py	494	CODE
LOW	…ent_consistency_model/train_lcm_distill_sd_wds_lisa.py	241	CODE
LOW	…ent_consistency_model/train_lcm_distill_sd_wds_lisa.py	849	CODE
LOW	…ent_consistency_model/train_lcm_distill_sd_wds_lisa.py	874	CODE
LOW	…ent_consistency_model/train_lcm_distill_sd_wds_lora.py	236	CODE
LOW	…ent_consistency_model/train_lcm_distill_sd_wds_lora.py	842	CODE
LOW	…ent_consistency_model/train_lcm_distill_sd_wds_lora.py	867	CODE
LOW	contrib/langchain/retrieval_chatbot.py	51	CODE
LOW	contrib/langchain/retrieval_chatbot.py	85	CODE
LOW	contrib/text2image/diffuser_finetuner.py	99	CODE
LOW	contrib/tool-finetune/function_call_finetune.py	158	CODE
LOW	contrib/tool-finetune/function_call_finetune.py	35	CODE
LOW	examples/vis_chatbot.py	63	CODE
LOW	examples/chatbot.py	42	CODE
LOW	examples/vis_chatbot_gradio.py	155	CODE
LOW	examples/vis_chatbot_gradio.py	214	CODE
LOW	examples/benchmarking.py	115	CODE
LOW	scripts/archive/export_llama_state_dict_checkpoint.py	19	CODE
LOW	src/lmflow/pipeline/inferencer.py	128	CODE
LOW	src/lmflow/pipeline/dpov2_aligner.py	228	CODE
LOW	src/lmflow/pipeline/dpov2_aligner.py	250	CODE
LOW	src/lmflow/pipeline/evaluator.py	179	CODE
LOW	src/lmflow/pipeline/evaluator.py	287	CODE
LOW	src/lmflow/pipeline/evaluator.py	440	CODE
LOW	src/lmflow/pipeline/utils/dpov2_dataprocessor.py	133	CODE
LOW	src/lmflow/pipeline/utils/dpov2_trainer.py	102	CODE
LOW	src/lmflow/pipeline/utils/raft_trainer.py	282	CODE
LOW	src/lmflow/pipeline/utils/raft_trainer.py	865	CODE
LOW	src/lmflow/pipeline/utils/raft_trainer.py	1003	CODE
LOW	src/lmflow/pipeline/utils/raft_trainer.py	1059	CODE
LOW	src/lmflow/pipeline/utils/raft_trainer.py	1180	CODE
LOW	src/lmflow/pipeline/utils/raft_trainer.py	1266	CODE
LOW	src/lmflow/pipeline/utils/raft_trainer.py	1327	CODE
LOW	src/lmflow/pipeline/utils/raft_trainer.py	1613	CODE
LOW	src/lmflow/pipeline/utils/raft_trainer.py	2053	CODE
LOW	src/lmflow/pipeline/utils/raft_trainer.py	2073	CODE
LOW	src/lmflow/pipeline/utils/raft_trainer.py	2131	CODE
LOW	src/lmflow/pipeline/utils/raft_trainer.py	2282	CODE
LOW	src/lmflow/pipeline/utils/raft_trainer.py	2391	CODE
LOW	src/lmflow/pipeline/utils/raft_trainer.py	2561	CODE
LOW	src/lmflow/pipeline/utils/raft_trainer.py	2714	CODE
LOW	src/lmflow/pipeline/utils/raft_trainer.py	3270	CODE
LOW	src/lmflow/datasets/multi_modal_dataset.py	94	CODE
LOW	src/lmflow/datasets/multi_modal_dataset.py	44	CODE
LOW	src/lmflow/datasets/dataset.py	72	CODE
LOW	src/lmflow/datasets/dataset.py	150	CODE
LOW	src/lmflow/optim/adan.py	75	CODE
LOW	src/lmflow/optim/lars.py	84	CODE
LOW	src/lmflow/optim/sgd_schedule_free.py	61	CODE
LOW	src/lmflow/optim/sgd_schedule_free.py	73	CODE
LOW	src/lmflow/optim/sgd_schedule_free.py	85	CODE
LOW	src/lmflow/optim/adabelief.py	88	CODE
LOW	src/lmflow/optim/novograd.py	31	CODE
LOW	src/lmflow/optim/utils.py	10	CODE
LOW	src/lmflow/optim/utils.py	13	CODE
LOW	src/lmflow/optim/radam.py	65	CODE
LOW	src/lmflow/optim/adamw_schedule_free.py	58	CODE
34 more matches not shown…

Modern Structural Boilerplate50 hits · 50 pts

Severity	File	Line	Snippet	Context
LOW	tests/pipeline/test_memory_safe_vllm_inferencer.py	16	logger = logging.getLogger(__name__)	CODE
LOW	contrib/text2image/t2i_dataset.py	15	logger = logging.getLogger(__name__)	CODE
LOW	contrib/text2image/diffuser_finetuner.py	28	logger = logging.getLogger(__name__)	CODE
LOW	contrib/tool-finetune/function_call_finetune.py	31	logger = logging.getLogger(__name__)	CODE
LOW	examples/iterative_dpo_train.py	19	logger = logging.getLogger(__name__)	CODE
LOW	examples/dpov2_train.py	20	logger = logging.getLogger(__name__)	CODE
LOW	examples/multistage_finetune.py	33	logger = logging.getLogger(__name__)	CODE
LOW	examples/reward_modeling.py	20	logger = logging.getLogger(__name__)	CODE
LOW	examples/sglang_inference.py	18	logger = logging.getLogger(__name__)	CODE
LOW	examples/vllm_inference.py	18	logger = logging.getLogger(__name__)	CODE
LOW	examples/rm_inference.py	18	logger = logging.getLogger(__name__)	CODE
LOW	examples/benchmarking.py	32	logger = logging.getLogger(__name__)	CODE
LOW	src/lmflow/__init__.py	21	__all__ = ["args", "datasets", "models", "pipeline", "utils"]	CODE
LOW	src/lmflow/args.py	31	logger = logging.getLogger(__name__)	CODE
LOW	src/lmflow/pipeline/iterative_dpo_aligner.py	26	logger = logging.getLogger(__name__)	CODE
LOW	src/lmflow/pipeline/sglang_inferencer.py	20	logger = logging.getLogger(__name__)	CODE
LOW	src/lmflow/pipeline/vllm_inferencer.py	28	logger = logging.getLogger(__name__)	CODE
LOW	src/lmflow/pipeline/inferencer.py	42	logger = logging.getLogger(__name__)	CODE
LOW	src/lmflow/pipeline/raft_aligner.py	31	logger = logging.getLogger(__name__)	CODE
LOW	src/lmflow/pipeline/rm_inferencer.py	38	logger = logging.getLogger(__name__)	CODE
LOW	src/lmflow/pipeline/rm_tuner.py	13	logger = logging.getLogger(__name__)	CODE
LOW	src/lmflow/pipeline/dpov2_aligner.py	22	logger = logging.getLogger(__name__)	CODE
LOW	src/lmflow/pipeline/finetuner.py	37	logger = logging.getLogger(__name__)	CODE
LOW	src/lmflow/pipeline/utils/dpov2_dataprocessor.py	12	logger = logging.getLogger(__name__)	CODE
LOW	src/lmflow/pipeline/utils/rm_dataprocessor.py	8	logger = logging.getLogger(__name__)	CODE
LOW	src/lmflow/pipeline/utils/dpov2_trainer.py	26	logger = logging.getLogger(__name__)	CODE
LOW	src/lmflow/pipeline/utils/memory_safe_dpov2_align.py	18	logger = logging.getLogger(__name__)	CODE
LOW	src/lmflow/pipeline/utils/memory_safe_vllm_inference.py	24	logger = logging.getLogger(__name__)	CODE
LOW	src/lmflow/datasets/dataset.py	33	logger = logging.getLogger(__name__)	CODE
LOW	src/lmflow/optim/__init__.py	3	__all__ = ["create_customized_optimizer"]	CODE
LOW	src/lmflow/utils/envs.py	11	logger = logging.getLogger(__name__)	CODE
LOW	src/lmflow/utils/envs.py	13	__all__ = [	CODE
LOW	src/lmflow/utils/envs.py	37	def set_cuda_device(local_rank: int) -> None:	CODE
LOW	src/lmflow/utils/protocol.py	26	logger = logging.getLogger(__name__)	CODE
LOW	src/lmflow/utils/deprecated.py	10	__all__ = ["deprecated_args"]	CODE
LOW	src/lmflow/utils/model.py	9	logger = logging.getLogger(__name__)	CODE
LOW	src/lmflow/utils/common.py	10	logger = logging.getLogger(__name__)	CODE
LOW	src/lmflow/utils/versioning.py	10	logger = logging.getLogger(__name__)	CODE
LOW	src/lmflow/utils/conversation_template/__init__.py	28	__all__ = [	CODE
LOW	src/lmflow/utils/conversation_template/__init__.py	33	logger = logging.getLogger(__name__)	CODE
LOW	src/lmflow/utils/conversation_template/llama.py	13	logger = logging.getLogger(__name__)	CODE
LOW	src/lmflow/utils/conversation_template/zephyr.py	11	logger = logging.getLogger(__name__)	CODE
LOW	src/lmflow/utils/conversation_template/gemma.py	8	logger = logging.getLogger(__name__)	CODE
LOW	src/lmflow/utils/conversation_template/base.py	14	logger = logging.getLogger(__name__)	CODE
LOW	src/lmflow/models/hf_decoder_model.py	45	logger = logging.getLogger(__name__)	CODE
LOW	src/lmflow/models/hf_text_regression_model.py	42	logger = logging.getLogger(__name__)	CODE
LOW	src/lmflow/models/hf_model_mixin.py	30	logger = logging.getLogger(__name__)	CODE
LOW	src/lmflow/models/vision_encoder/__init__.py	3	__all__ = ["build_vision_tower"]	CODE
LOW	src/lmflow/tokenization/hf_decoder_model.py	15	logger = logging.getLogger(__name__)	CODE
LOW	src/lmflow/tokenization/hf_text_regression_model.py	15	logger = logging.getLogger(__name__)	CODE

Unused Imports26 hits · 24 pts

Severity	File	Line	Context
LOW	tests/models/test_hf_decoder_model.py	15	CODE
LOW	tests/models/test_hf_decoder_model.py	20	CODE
LOW	tests/models/test_hf_decoder_model.py	27	CODE
LOW	tests/models/test_hf_decoder_model.py	27	CODE
LOW	tests/models/test_hf_decoder_model.py	31	CODE
LOW	src/lmflow/__init__.py	8	CODE
LOW	src/lmflow/__init__.py	8	CODE
LOW	src/lmflow/__init__.py	8	CODE
LOW	src/lmflow/__init__.py	8	CODE
LOW	src/lmflow/__init__.py	8	CODE
LOW	src/lmflow/pipeline/utils/raft_trainer.py	23	CODE
LOW	src/lmflow/pipeline/utils/raft_trainer.py	23	CODE
LOW	src/lmflow/pipeline/utils/raft_trainer.py	23	CODE
LOW	src/lmflow/pipeline/utils/raft_trainer.py	23	CODE
LOW	src/lmflow/pipeline/utils/raft_trainer.py	280	CODE
LOW	src/lmflow/pipeline/utils/raft_trainer.py	280	CODE
LOW	src/lmflow/pipeline/utils/raft_trainer.py	280	CODE
LOW	src/lmflow/pipeline/utils/raft_trainer.py	280	CODE
LOW	src/lmflow/datasets/__init__.py	8	CODE
LOW	src/lmflow/datasets/__init__.py	12	CODE
LOW	src/lmflow/optim/__init__.py	1	CODE
LOW	src/lmflow/utils/test_utils.py	4	CODE
LOW	src/lmflow/utils/conversation_template/__init__.py	7	CODE
LOW	src/lmflow/utils/conversation_template/__init__.py	7	CODE
LOW	src/lmflow/models/hf_text_regression_model.py	39	CODE
LOW	src/lmflow/models/vision_encoder/__init__.py	1	CODE

Excessive Try-Catch Wrapping18 hits · 22 pts

Severity	File	Line	Snippet	Context
LOW	…ent_consistency_model/train_lcm_distill_sd_wds_lisa.py	158	except Exception:	CODE
MEDIUM	…ent_consistency_model/train_lcm_distill_sd_wds_lisa.py	147	def __call__(self, x):	CODE
LOW	…ent_consistency_model/train_lcm_distill_sd_wds_lora.py	153	except Exception:	CODE
MEDIUM	…ent_consistency_model/train_lcm_distill_sd_wds_lora.py	142	def __call__(self, x):	CODE
LOW	src/lmflow/pipeline/inferencer.py	91	except Exception:	CODE
MEDIUM	src/lmflow/pipeline/inferencer.py	92	print("Error in setting hidden size, use the default size 1024")	CODE
LOW	src/lmflow/pipeline/inferencer.py	346	except Exception:	STRING
MEDIUM	src/lmflow/pipeline/inferencer.py	347	print("Error in setting hidden size for draft model, use the default size 1024")	STRING
LOW	src/lmflow/pipeline/evaluator.py	97	except Exception:	CODE
MEDIUM	src/lmflow/pipeline/evaluator.py	98	print("Error in setting hidden size, use the default size 1024")	CODE
LOW	src/lmflow/pipeline/evaluator.py	403	except Exception:	CODE
LOW	src/lmflow/pipeline/evaluator.py	483	except Exception:	CODE
LOW	src/lmflow/pipeline/utils/raft_trainer.py	2274	except Exception as e:	CODE
LOW	src/lmflow/utils/data_utils.py	264	except Exception:	CODE
LOW	src/lmflow/utils/versioning.py	22	except Exception as e:	CODE
LOW	src/lmflow/models/hf_model_mixin.py	418	except Exception:	CODE
LOW	src/lmflow/models/hf_model_mixin.py	580	except Exception:	CODE
LOW	src/lmflow/tokenization/hf_text_regression_model.py	186	except Exception:	CODE

Over-Commented Block19 hits · 19 pts

Severity	File	Line	Snippet	Context
LOW	experimental/RAFT-diffusion/train_text_to_image_lora.py	1	# Copyright 2023 The HuggingFace Inc. team. All rights reserved.	COMMENT
LOW	experimental/RAFT-diffusion/train_text_to_image_lora.py	441	# It's important to realize here how many attention weights will be added and of which sizes	COMMENT
LOW	experimental/RAFT-diffusion/train_text_to_image_lora.py	821	upload_folder(	COMMENT
LOW	experimental/RAFT-diffusion/train_text_to_image_lora.py	841	# for _ in range(args.num_validation_images):	COMMENT
LOW	…SA-diffusion/instruct_pix2pix/test_instruct_pix2pix.py	1	# Copyright 2024 HuggingFace Inc.	COMMENT
LOW	…fusion/instruct_pix2pix/train_instruct_pix2pix_lisa.py	1	#!/usr/bin/env python	COMMENT
LOW	…al/LISA-diffusion/diffusion_dpo/train_diffusion_dpo.py	1	#!/usr/bin/env python	COMMENT
LOW	…SA-diffusion/diffusion_dpo/train_diffusion_dpo_lisa.py	1	#!/usr/bin/env python	COMMENT
LOW	…ent_consistency_model/train_lcm_distill_sd_wds_lisa.py	1	#!/usr/bin/env python	COMMENT
LOW	…ent_consistency_model/train_lcm_distill_sd_wds_lora.py	1	#!/usr/bin/env python	COMMENT
LOW	scripts/archive/convert_llama_weights_to_hf.py	1	# Copyright 2022 EleutherAI and The HuggingFace Inc. team. All rights reserved.	COMMENT
LOW	scripts/archive/bash.sh	1	#!/bin/bash	COMMENT
LOW	scripts/multimodal/run_vis_chatbot_gradio_minigpt4.sh	1	#!/bin/bash	COMMENT
LOW	src/lmflow/pipeline/finetuner.py	181	# We drop the small remainder, we could add padding if the model	COMMENT
LOW	src/lmflow/utils/conversation_template/base.py	301		COMMENT
LOW	src/lmflow/utils/conversation_template/hymba.py	1	#!/usr/bin/env python	COMMENT
LOW	src/lmflow/utils/conversation_template/hymba.py	21	# {% endif %}	COMMENT
LOW	src/lmflow/models/hf_text_regression_model.py	341		COMMENT
LOW	src/lmflow/tokenization/__init__.py	1	#!/usr/bin/env python	COMMENT

Self-Referential Comments6 hits · 18 pts

Severity	File	Line	Snippet	Context
MEDIUM	…fusion/instruct_pix2pix/train_instruct_pix2pix_lisa.py	1054	# Create the pipeline using the trained modules and save it.	COMMENT
MEDIUM	…ent_consistency_model/train_lcm_distill_sd_wds_lisa.py	1489	# Create the pipeline using using the trained modules and save it.	COMMENT
MEDIUM	…ent_consistency_model/train_lcm_distill_sd_wds_lora.py	1440	# Create the pipeline using using the trained modules and save it.	COMMENT
MEDIUM	contrib/rlhflow/reward_modeling.py	108	## Define the trainer	COMMENT
MEDIUM	src/lmflow/datasets/dataset.py	9	# Importing necessary libraries and modules	COMMENT
MEDIUM	src/lmflow/utils/protocol.py	610	# Create a slice object	COMMENT

Structural Annotation Overuse8 hits · 17 pts

Severity	File	Line	Snippet	Context
LOW⚡	docs/readme/multi_node.md	9	#### Step 1: Set up single-node configurations	COMMENT
LOW⚡	docs/readme/multi_node.md	12	#### Step 2: Prepare multi-node communication tools	COMMENT
LOW⚡	docs/readme/multi_node.md	18	#### Step 3: Set up `~/.bashrc`	COMMENT
LOW⚡	docs/readme/multi_node.md	28	#### Step 4: Establish SSH trust between servers	COMMENT
LOW	docs/readme/multi_node.md	45	### Step 5: Update `hostfile`	COMMENT
LOW⚡	docs/readme/multi_node.md	65	#### Step 6: Check firewalls	COMMENT
LOW⚡	docs/readme/multi_node.md	71	#### Step 7: Running single-node training in each server	COMMENT
LOW⚡	docs/readme/multi_node.md	75	#### Step 8: Run the training script	COMMENT

AI Structural Patterns14 hits · 13 pts

Severity	File	Line	Context
LOW	src/lmflow/pipeline/inferencer.py	420	CODE
LOW	src/lmflow/pipeline/inferencer.py	626	CODE
LOW	src/lmflow/pipeline/raft_aligner.py	257	CODE
LOW	src/lmflow/pipeline/raft_aligner.py	390	CODE
LOW	src/lmflow/pipeline/finetuner.py	190	CODE
LOW	src/lmflow/pipeline/utils/dpov2_trainer.py	30	CODE
LOW	src/lmflow/pipeline/utils/raft_trainer.py	282	CODE
LOW	src/lmflow/pipeline/utils/raft_trainer.py	3433	CODE
LOW	src/lmflow/optim/adabelief.py	14	CODE
LOW	src/lmflow/optim/adamw_schedule_free.py	27	CODE
LOW	src/lmflow/optim/sgdp.py	20	CODE
LOW	src/lmflow/utils/protocol.py	1022	CODE
LOW	src/lmflow/models/hf_decoder_model.py	310	CODE
LOW	src/lmflow/models/vision2seq_model.py	189	CODE

Redundant / Tautological Comments8 hits · 12 pts

Severity	File	Line	Snippet	Context
LOW	examples/detail_memory.py	11	# Check if the model name is provided as a command-line argument	COMMENT
LOW	src/lmflow/pipeline/inferencer.py	632	# Print the result	COMMENT
LOW	src/lmflow/pipeline/rm_tuner.py	125	# Check if it's time to switch active layers, including at step 0	COMMENT
LOW	src/lmflow/pipeline/utils/lisa_trainer.py	50	# Check if it's time to switch active layers, including at step 0	COMMENT
LOW	src/lmflow/pipeline/utils/raft_trainer.py	2046	# Check if saved optimizer or scheduler states exist	COMMENT
LOW	src/lmflow/pipeline/utils/raft_trainer.py	2869	# Check if we should delete older checkpoint(s)	COMMENT
LOW	src/lmflow/pipeline/utils/raft_trainer.py	3145	# Set back to None to begin a new accumulation	COMMENT
LOW	src/lmflow/pipeline/utils/raft_trainer.py	3691	# Set back to None to begin a new accumulation	COMMENT

Slop Phrases7 hits · 10 pts

Severity	File	Line	Snippet	Context
LOW	…fusion/instruct_pix2pix/train_instruct_pix2pix_lisa.py	543	# make sure to pop weight so that corresponding model is not saved again	COMMENT
LOW	…al/LISA-diffusion/diffusion_dpo/train_diffusion_dpo.py	598	# make sure to pop weight so that corresponding model is not saved again	COMMENT
LOW	…SA-diffusion/diffusion_dpo/train_diffusion_dpo_lisa.py	617	# make sure to pop weight so that corresponding model is not saved again	COMMENT
LOW	…ent_consistency_model/train_lcm_distill_sd_wds_lisa.py	1042	# make sure to pop weight so that corresponding model is not saved again	COMMENT
LOW	…ent_consistency_model/train_lcm_distill_sd_wds_lora.py	1048	# make sure to pop weight so that corresponding model is not saved again	COMMENT
LOW	tests/pipeline/test_finetuner_distributed_loss.py	23	# python subprocess takes list of strings as command, don't forget to split	COMMENT
LOW	tests/pipeline/test_finetuner_distributed_loss.py	23	# python subprocess takes list of strings as command, don't forget to split	COMMENT

Docstring Block Structure2 hits · 10 pts

Severity	File	Line	Snippet	Context
HIGH	src/lmflow/pipeline/utils/raft_trainer.py	3531	Upload self.model and self.tokenizer to the 🤗 model hub on the repo self.args.hub_model_id. Param	STRING
HIGH	src/lmflow/utils/protocol.py	583	Slice the DataProto and return a new DataProto object. This is an improved version of direct slicing wh	STRING

Cross-Language Confusion1 hit · 8 pts

Severity	File	Line	Snippet	Context
HIGH⚡	examples/benchmarking.py	132	os.system("cd data && ./download.sh common_sense_eval && cd -")	CODE

Decorative Section Separators2 hits · 6 pts

Severity	File	Line	Snippet	Context
MEDIUM	pyproject.toml	5	# -------------------------------	COMMENT
MEDIUM	pyproject.toml	7	# -------------------------------	COMMENT

AI Slop Vocabulary3 hits · 4 pts

Severity	File	Line	Snippet	Context
LOW	src/lmflow/pipeline/rm_inferencer.py	219	# For tensor_parallel_size == 1, we simply set num_gpus=1.	COMMENT
LOW	src/lmflow/pipeline/utils/raft_trainer.py	1807	# if loss is nan or inf simply add the average of previous logged losses	COMMENT
LOW	src/lmflow/pipeline/utils/raft_trainer.py	2020	# print("I just create a optimizer here!") # called	COMMENT

Modern AI Meta-Vocabulary1 hit · 3 pts

Severity	File	Line	Snippet	Context
MEDIUM	src/lmflow/pipeline/evaluator.py	480	# Gets context window length	COMMENT

Overly Generic Function Names1 hit · 0 pts

Severity	File	Line	Snippet	Context
LOW	src/lmflow/utils/deprecated.py	32	def my_function(inference_engine='huggingface', **kwargs):	STRING

Analysis Overview

What These Metrics Mean

Score History

Severity Breakdown

Directory Score Breakdown

Pattern Findings