kohya-ss/sd-scripts

18.7

Adjusted Score

18.7

Raw Score

100%

Time Factor

2026-08-02

Last Push

7.2K

Stars

Python

Language

118.7K

Lines of Code

256

Files

1.6K

Pattern Hits

2026-08-02

Scan Date

0.29

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 75MEDIUM 138LOW 1429

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 1642 distinct pattern matches across 16 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 Identifiers474 hits · 476 pts

Severity	File	Line	Snippet	Context
LOW⚡	train_textual_inversion.py	126	def get_latents_caching_strategy(self, args):	CODE
LOW⚡	train_textual_inversion.py	135	def get_text_encoding_strategy(self, args):	CODE
LOW⚡	train_textual_inversion.py	138	def get_models_for_text_encoding(self, args, accelerator, text_encoders) -> List[Any]:	CODE
LOW	sd3_train_network.py	156	def get_latents_caching_strategy(self, args):	CODE
LOW	sd3_train_network.py	162	def get_text_encoding_strategy(self, args):	CODE
LOW⚡	sd3_train_network.py	180	def get_models_for_text_encoding(self, args, accelerator, text_encoders):	CODE
LOW⚡	sd3_train_network.py	189	def get_text_encoders_train_flags(self, args, text_encoders):	CODE
LOW⚡	sd3_train_network.py	192	def get_text_encoder_outputs_caching_strategy(self, args):	CODE
LOW	sd3_train_network.py	206	def cache_text_encoder_outputs_if_needed(	CODE
LOW	sd3_train_network.py	317	def get_noise_pred_and_target(	CODE
LOW	sd3_train_network.py	412	def is_text_encoder_not_needed_for_training(self, args):	CODE
LOW	sd3_train_network.py	415	def prepare_text_encoder_grad_ckpt_workaround(self, index, text_encoder):	CODE
LOW	sd3_train_network.py	470	def prepare_unet_with_accelerator(	CODE
LOW	sdxl_train.py	59	def get_block_params_to_optimize(unet: SdxlUNet2DConditionModel, block_lrs: List[float]) -> List[dict]:	CODE
LOW	sdxl_train.py	254	def set_diffusers_xformers_flag(model, valid):	CODE
LOW	anima_train_control_net_lllite.py	90	def _build_inpaint_cond_image(	CODE
LOW	anima_train_control_net_lllite.py	108	def _generate_random_masks_for_batch(	CODE
LOW	anima_train_control_net_lllite.py	122	def _make_lllite_sample_hooks(args, lllite, dit_dtype):	CODE
LOW	anima_train_control_net_lllite.py	201	def add_anima_lllite_arguments(parser: argparse.ArgumentParser):	CODE
LOW⚡	sdxl_train_network.py	78	def get_latents_caching_strategy(self, args):	CODE
LOW⚡	sdxl_train_network.py	84	def get_text_encoding_strategy(self, args):	CODE
LOW⚡	sdxl_train_network.py	87	def get_models_for_text_encoding(self, args, accelerator, text_encoders):	CODE
LOW⚡	sdxl_train_network.py	90	def get_text_encoder_outputs_caching_strategy(self, args):	CODE
LOW⚡	sdxl_train_network.py	98	def cache_text_encoder_outputs_if_needed(	CODE
LOW⚡	lumina_train_network.py	96	def get_latents_caching_strategy(self, args):	CODE
LOW⚡	lumina_train_network.py	99	def get_text_encoding_strategy(self, args):	CODE
LOW⚡	lumina_train_network.py	102	def get_text_encoders_train_flags(self, args, text_encoders):	CODE
LOW⚡	lumina_train_network.py	105	def get_text_encoder_outputs_caching_strategy(self, args):	CODE
LOW	lumina_train_network.py	117	def cache_text_encoder_outputs_if_needed(	CODE
LOW	lumina_train_network.py	240	def get_noise_pred_and_target(	CODE
LOW⚡	lumina_train_network.py	342	def is_text_encoder_not_needed_for_training(self, args):	CODE
LOW⚡	lumina_train_network.py	345	def prepare_text_encoder_grad_ckpt_workaround(self, index, text_encoder):	CODE
LOW⚡	lumina_train_network.py	353	def prepare_unet_with_accelerator(	CODE
LOW	fine_tune.py	157	def set_diffusers_xformers_flag(model, valid):	CODE
LOW	gen_img_diffusers.py	339	def replace_vae_attn_to_memory_efficient():	CODE
LOW	gen_img_diffusers.py	395	def replace_vae_attn_to_xformers():	CODE
LOW	gen_img_diffusers.py	663	def enable_xformers_memory_efficient_attention(self):	STRING
LOW	gen_img_diffusers.py	673	def disable_xformers_memory_efficient_attention(self):	STRING
LOW	gen_img_diffusers.py	696	def disable_attention_slicing(self):	STRING
LOW	gen_img_diffusers.py	704	def enable_sequential_cpu_offload(self):	STRING
LOW	gen_img_diffusers.py	1982	def get_unweighted_text_embeddings(	STRING
LOW	gen_img_diffusers.py	2043	def get_weighted_text_embeddings(	STRING
LOW	gen_img_diffusers.py	2238	def handle_dynamic_prompt_variants(prompt, repeat_count):	STRING
LOW⚡	hunyuan_image_train_network.py	421	def get_latents_caching_strategy(self, args):	CODE
LOW⚡	hunyuan_image_train_network.py	424	def get_text_encoding_strategy(self, args):	CODE
LOW⚡	hunyuan_image_train_network.py	430	def get_models_for_text_encoding(self, args, accelerator, text_encoders):	CODE
LOW⚡	hunyuan_image_train_network.py	436	def get_text_encoders_train_flags(self, args, text_encoders):	CODE
LOW⚡	hunyuan_image_train_network.py	440	def get_text_encoder_outputs_caching_strategy(self, args):	CODE
LOW⚡	hunyuan_image_train_network.py	448	def cache_text_encoder_outputs_if_needed(	CODE
LOW	hunyuan_image_train_network.py	524	def get_noise_pred_and_target(	CODE
LOW	hunyuan_image_train_network.py	596	def is_text_encoder_not_needed_for_training(self, args):	CODE
LOW	hunyuan_image_train_network.py	599	def prepare_text_encoder_grad_ckpt_workaround(self, index, text_encoder):	CODE
LOW	hunyuan_image_train_network.py	621	def prepare_unet_with_accelerator(	CODE
LOW⚡	anima_train_network.py	163	def get_latents_caching_strategy(self, args):	CODE
LOW⚡	anima_train_network.py	166	def get_text_encoding_strategy(self, args):	CODE
LOW⚡	anima_train_network.py	172	def get_models_for_text_encoding(self, args, accelerator, text_encoders):	CODE
LOW⚡	anima_train_network.py	177	def get_text_encoder_outputs_caching_strategy(self, args):	CODE
LOW⚡	anima_train_network.py	184	def cache_text_encoder_outputs_if_needed(	CODE
LOW	anima_train_network.py	268	def get_noise_pred_and_target(	CODE
LOW⚡	anima_train_network.py	416	def is_text_encoder_not_needed_for_training(self, args):	CODE
414 more matches not shown…

Decorative Section Separators104 hits · 349 pts

Severity	File	Line	Snippet	Context
MEDIUM⚡	inpainting_minimal_inference.py	55	# -------------------------------------------------------------------------	COMMENT
MEDIUM⚡	inpainting_minimal_inference.py	57	# -------------------------------------------------------------------------	COMMENT
MEDIUM⚡	inpainting_minimal_inference.py	67	# -------------------------------------------------------------------------	COMMENT
MEDIUM⚡	inpainting_minimal_inference.py	69	# -------------------------------------------------------------------------	COMMENT
MEDIUM⚡	inpainting_minimal_inference.py	271	# ------------------------------------------------------------------	COMMENT
MEDIUM⚡	inpainting_minimal_inference.py	273	# ------------------------------------------------------------------	COMMENT
MEDIUM⚡	inpainting_minimal_inference.py	281	# ------------------------------------------------------------------	COMMENT
MEDIUM⚡	inpainting_minimal_inference.py	283	# ------------------------------------------------------------------	COMMENT
MEDIUM	inpainting_minimal_inference.py	136	# -------------------------------------------------------------------------	COMMENT
MEDIUM	inpainting_minimal_inference.py	138	# -------------------------------------------------------------------------	COMMENT
MEDIUM	inpainting_minimal_inference.py	152	# -------------------------------------------------------------------------	COMMENT
MEDIUM	inpainting_minimal_inference.py	154	# -------------------------------------------------------------------------	COMMENT
MEDIUM	inpainting_minimal_inference.py	191	# -------------------------------------------------------------------------	COMMENT
MEDIUM	inpainting_minimal_inference.py	193	# -------------------------------------------------------------------------	COMMENT
MEDIUM	inpainting_minimal_inference.py	225	# ------------------------------------------------------------------	COMMENT
MEDIUM	inpainting_minimal_inference.py	227	# ------------------------------------------------------------------	COMMENT
MEDIUM	inpainting_minimal_inference.py	314	# ------------------------------------------------------------------	COMMENT
MEDIUM	inpainting_minimal_inference.py	316	# ------------------------------------------------------------------	COMMENT
MEDIUM	inpainting_minimal_inference.py	341	# ------------------------------------------------------------------	COMMENT
MEDIUM	inpainting_minimal_inference.py	343	# ------------------------------------------------------------------	COMMENT
MEDIUM⚡	anima_minimal_inference_control_net_lllite.py	55	# ---------------------------------------------------------------------------	COMMENT
MEDIUM⚡	anima_minimal_inference_control_net_lllite.py	57	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	anima_minimal_inference_control_net_lllite.py	107	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	anima_minimal_inference_control_net_lllite.py	109	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	anima_minimal_inference_control_net_lllite.py	238	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	anima_minimal_inference_control_net_lllite.py	240	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	anima_minimal_inference_control_net_lllite.py	278	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	anima_minimal_inference_control_net_lllite.py	280	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	anima_minimal_inference_control_net_lllite.py	370	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	anima_minimal_inference_control_net_lllite.py	372	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	anima_minimal_inference_control_net_lllite.py	426	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	anima_minimal_inference_control_net_lllite.py	428	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	tools/dev/visualize_masks.py	52	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	tools/dev/visualize_masks.py	54	# ---------------------------------------------------------------------------	COMMENT
MEDIUM⚡	tools/dev/visualize_masks.py	120	# ---------------------------------------------------------------------------	COMMENT
MEDIUM⚡	tools/dev/visualize_masks.py	122	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	tools/dev/visualize_masks.py	169	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	tools/dev/visualize_masks.py	171	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	tools/dev/visualize_masks.py	274	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	tools/dev/visualize_masks.py	276	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	tools/dev/manual_test_anima_lllite_dryrun.py	35	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	tools/dev/manual_test_anima_lllite_dryrun.py	39	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	tests/test_mask_generator.py	29	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	tests/test_mask_generator.py	31	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	tests/test_mask_generator.py	44	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	tests/test_mask_generator.py	46	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	tests/test_mask_generator.py	84	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	tests/test_mask_generator.py	86	# ---------------------------------------------------------------------------	COMMENT
MEDIUM⚡	tests/test_mask_generator.py	123	# ---------------------------------------------------------------------------	COMMENT
MEDIUM⚡	tests/test_mask_generator.py	125	# ---------------------------------------------------------------------------	COMMENT
MEDIUM⚡	tests/test_mask_generator.py	162	# ---------------------------------------------------------------------------	COMMENT
MEDIUM⚡	tests/test_mask_generator.py	164	# ---------------------------------------------------------------------------	COMMENT
MEDIUM⚡	tests/test_mask_generator.py	188	# ---------------------------------------------------------------------------	COMMENT
MEDIUM⚡	tests/test_mask_generator.py	190	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	tests/test_expand_unet_to_inpainting.py	28	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	tests/test_expand_unet_to_inpainting.py	30	# ---------------------------------------------------------------------------	COMMENT
MEDIUM⚡	tests/test_expand_unet_to_inpainting.py	60	# ---------------------------------------------------------------------------	COMMENT
MEDIUM⚡	tests/test_expand_unet_to_inpainting.py	62	# ---------------------------------------------------------------------------	COMMENT
MEDIUM⚡	tests/test_expand_unet_to_inpainting.py	92	# ---------------------------------------------------------------------------	COMMENT
MEDIUM⚡	tests/test_expand_unet_to_inpainting.py	94	# ---------------------------------------------------------------------------	COMMENT
44 more matches not shown…

Cross-File Repetition68 hits · 340 pts

Severity	File	Line	Snippet	Context
HIGH	gen_img_diffusers.py	0	pipeline for text-to-image generation using stable diffusion without tokens length limit, and support parsing weighting	STRING
HIGH	library/lpw_stable_diffusion.py	0	pipeline for text-to-image generation using stable diffusion without tokens length limit, and support parsing weighting	STRING
HIGH	library/sdxl_lpw_stable_diffusion.py	0	pipeline for text-to-image generation using stable diffusion without tokens length limit, and support parsing weighting	STRING
HIGH	gen_img_diffusers.py	0	function invoked when calling the pipeline for generation. args: prompt (`str` or `list[str]`): the prompt or prompts to	STRING
HIGH	library/lpw_stable_diffusion.py	0	function invoked when calling the pipeline for generation. args: prompt (`str` or `list[str]`): the prompt or prompts to	STRING
HIGH	library/sdxl_lpw_stable_diffusion.py	0	function invoked when calling the pipeline for generation. args: prompt (`str` or `list[str]`): the prompt or prompts to	STRING
HIGH	gen_img_diffusers.py	0	function for text-to-image generation. args: prompt (`str` or `list[str]`): the prompt or prompts to guide the image gen	STRING
HIGH	library/lpw_stable_diffusion.py	0	function for text-to-image generation. args: prompt (`str` or `list[str]`): the prompt or prompts to guide the image gen	STRING
HIGH	library/sdxl_lpw_stable_diffusion.py	0	function for text-to-image generation. args: prompt (`str` or `list[str]`): the prompt or prompts to guide the image gen	STRING
HIGH	gen_img_diffusers.py	0	function for image-to-image generation. args: image (`torch.floattensor` or `pil.image.image`): `image`, or tensor repre	STRING
HIGH	library/lpw_stable_diffusion.py	0	function for image-to-image generation. args: image (`torch.floattensor` or `pil.image.image`): `image`, or tensor repre	STRING
HIGH	library/sdxl_lpw_stable_diffusion.py	0	function for image-to-image generation. args: image (`torch.floattensor` or `pil.image.image`): `image`, or tensor repre	STRING
HIGH	gen_img_diffusers.py	0	function for inpaint. args: image (`torch.floattensor` or `pil.image.image`): `image`, or tensor representing an image b	STRING
HIGH	library/lpw_stable_diffusion.py	0	function for inpaint. args: image (`torch.floattensor` or `pil.image.image`): `image`, or tensor representing an image b	STRING
HIGH	library/sdxl_lpw_stable_diffusion.py	0	function for inpaint. args: image (`torch.floattensor` or `pil.image.image`): `image`, or tensor representing an image b	STRING
HIGH	gen_img_diffusers.py	0	\\\(\| \\\)\| \\\[\| \\]\| \\\\\| \\\| \(\| \[\| :([+-]?[.\d]+)\)\| \)\| ]\| [^\\()\[\]:]+\| :	STRING
HIGH	sdxl_gen_img.py	0	\\\(\| \\\)\| \\\[\| \\]\| \\\\\| \\\| \(\| \[\| :([+-]?[.\d]+)\)\| \)\| ]\| [^\\()\[\]:]+\| :	STRING
HIGH	gen_img.py	0	\\\(\| \\\)\| \\\[\| \\]\| \\\\\| \\\| \(\| \[\| :([+-]?[.\d]+)\)\| \)\| ]\| [^\\()\[\]:]+\| :	STRING
HIGH	library/strategy_base.py	0	\\\(\| \\\)\| \\\[\| \\]\| \\\\\| \\\| \(\| \[\| :([+-]?[.\d]+)\)\| \)\| ]\| [^\\()\[\]:]+\| :	STRING
HIGH	library/lpw_stable_diffusion.py	0	\\\(\| \\\)\| \\\[\| \\]\| \\\\\| \\\| \(\| \[\| :([+-]?[.\d]+)\)\| \)\| ]\| [^\\()\[\]:]+\| :	STRING
HIGH	library/sdxl_lpw_stable_diffusion.py	0	\\\(\| \\\)\| \\\[\| \\]\| \\\\\| \\\| \(\| \[\| :([+-]?[.\d]+)\)\| \)\| ]\| [^\\()\[\]:]+\| :	STRING
HIGH	library/custom_train_functions.py	0	\\\(\| \\\)\| \\\[\| \\]\| \\\\\| \\\| \(\| \[\| :([+-]?[.\d]+)\)\| \)\| ]\| [^\\()\[\]:]+\| :	STRING
HIGH	gen_img_diffusers.py	0	parses a string with attention tokens and returns a list of pairs: text and its associated weight. accepted tokens are:	STRING
HIGH	sdxl_gen_img.py	0	parses a string with attention tokens and returns a list of pairs: text and its associated weight. accepted tokens are:	STRING
HIGH	gen_img.py	0	parses a string with attention tokens and returns a list of pairs: text and its associated weight. accepted tokens are:	STRING
HIGH	library/strategy_base.py	0	parses a string with attention tokens and returns a list of pairs: text and its associated weight. accepted tokens are:	STRING
HIGH	library/lpw_stable_diffusion.py	0	parses a string with attention tokens and returns a list of pairs: text and its associated weight. accepted tokens are:	STRING
HIGH	library/sdxl_lpw_stable_diffusion.py	0	parses a string with attention tokens and returns a list of pairs: text and its associated weight. accepted tokens are:	STRING
HIGH	library/custom_train_functions.py	0	parses a string with attention tokens and returns a list of pairs: text and its associated weight. accepted tokens are:	STRING
HIGH	gen_img_diffusers.py	0	tokenize a list of prompts and return its tokens with weights of each token. no padding, starting or ending token is inc	STRING
HIGH	sdxl_gen_img.py	0	tokenize a list of prompts and return its tokens with weights of each token. no padding, starting or ending token is inc	STRING
HIGH	gen_img.py	0	tokenize a list of prompts and return its tokens with weights of each token. no padding, starting or ending token is inc	STRING
HIGH	library/lpw_stable_diffusion.py	0	tokenize a list of prompts and return its tokens with weights of each token. no padding, starting or ending token is inc	STRING
HIGH	library/sdxl_lpw_stable_diffusion.py	0	tokenize a list of prompts and return its tokens with weights of each token. no padding, starting or ending token is inc	STRING
HIGH	library/custom_train_functions.py	0	tokenize a list of prompts and return its tokens with weights of each token. no padding, starting or ending token is inc	STRING
HIGH	gen_img_diffusers.py	0	pad the tokens (with starting and ending tokens) and weights (with 1.0) to max_length.	STRING
HIGH	sdxl_gen_img.py	0	pad the tokens (with starting and ending tokens) and weights (with 1.0) to max_length.	STRING
HIGH	gen_img.py	0	pad the tokens (with starting and ending tokens) and weights (with 1.0) to max_length.	STRING
HIGH	library/strategy_base.py	0	pad the tokens (with starting and ending tokens) and weights (with 1.0) to max_length.	STRING
HIGH	library/lpw_stable_diffusion.py	0	pad the tokens (with starting and ending tokens) and weights (with 1.0) to max_length.	STRING
HIGH	library/sdxl_lpw_stable_diffusion.py	0	pad the tokens (with starting and ending tokens) and weights (with 1.0) to max_length.	STRING
HIGH	library/custom_train_functions.py	0	pad the tokens (with starting and ending tokens) and weights (with 1.0) to max_length.	STRING
HIGH	gen_img_diffusers.py	0	when the length of tokens is a multiple of the capacity of the text encoder, it should be split into chunks and sent to	STRING
HIGH	sdxl_gen_img.py	0	when the length of tokens is a multiple of the capacity of the text encoder, it should be split into chunks and sent to	STRING
HIGH	gen_img.py	0	when the length of tokens is a multiple of the capacity of the text encoder, it should be split into chunks and sent to	STRING
HIGH	library/lpw_stable_diffusion.py	0	when the length of tokens is a multiple of the capacity of the text encoder, it should be split into chunks and sent to	STRING
HIGH	library/sdxl_lpw_stable_diffusion.py	0	when the length of tokens is a multiple of the capacity of the text encoder, it should be split into chunks and sent to	STRING
HIGH	library/custom_train_functions.py	0	when the length of tokens is a multiple of the capacity of the text encoder, it should be split into chunks and sent to	STRING
HIGH	gen_img_diffusers.py	0	prompts can be assigned with local weights using brackets. for example, prompt 'a (very beautiful) masterpiece' highligh	STRING
HIGH	library/lpw_stable_diffusion.py	0	prompts can be assigned with local weights using brackets. for example, prompt 'a (very beautiful) masterpiece' highligh	STRING
HIGH	library/sdxl_lpw_stable_diffusion.py	0	prompts can be assigned with local weights using brackets. for example, prompt 'a (very beautiful) masterpiece' highligh	STRING
HIGH	library/custom_train_functions.py	0	prompts can be assigned with local weights using brackets. for example, prompt 'a (very beautiful) masterpiece' highligh	STRING
HIGH	finetune/make_captions_by_git.py	0	collate function that allows to remove corrupted examples in the dataloader. it expects that the dataloader returns 'non	STRING
HIGH	finetune/make_captions.py	0	collate function that allows to remove corrupted examples in the dataloader. it expects that the dataloader returns 'non	STRING
HIGH	finetune/prepare_buckets_latents.py	0	collate function that allows to remove corrupted examples in the dataloader. it expects that the dataloader returns 'non	STRING
HIGH	networks/dylora.py	0	replaces forward method of the original linear, instead of replacing the original linear module.	STRING
HIGH	networks/oft_flux.py	0	replaces forward method of the original linear, instead of replacing the original linear module.	STRING
HIGH	networks/lora_diffusers.py	0	replaces forward method of the original linear, instead of replacing the original linear module.	STRING
HIGH	networks/lora_flux.py	0	replaces forward method of the original linear, instead of replacing the original linear module.	STRING
HIGH	networks/oft.py	0	replaces forward method of the original linear, instead of replacing the original linear module.	STRING
8 more matches not shown…

Unused Imports370 hits · 320 pts

Severity	File	Line	Context
LOW	sd3_train_network.py	2	CODE
LOW	sd3_train_network.py	4	CODE
LOW	sd3_train_network.py	9	CODE
LOW	sd3_train_network.py	15	CODE
LOW	train_db.py	39	CODE
LOW	train_db.py	39	CODE
LOW	train_db.py	39	CODE
LOW	anima_train_control_net_lllite.py	10	CODE
LOW	anima_train_control_net_lllite.py	23	CODE
LOW	anima_train_control_net_lllite.py	53	CODE
LOW	flux_train_control_net.py	17	CODE
LOW	flux_train_control_net.py	18	CODE
LOW	flux_train_control_net.py	20	CODE
LOW	flux_train_control_net.py	20	CODE
LOW	flux_train_control_net.py	20	CODE
LOW	flux_train_control_net.py	20	CODE
LOW	flux_train_control_net.py	24	CODE
LOW	sdxl_train_network.py	10	CODE
LOW	sdxl_train_control_net_lllite.py	5	CODE
LOW	sdxl_train_control_net_lllite.py	9	CODE
LOW	sdxl_train_control_net_lllite.py	11	CODE
LOW	sdxl_train_control_net_lllite.py	24	CODE
LOW	sdxl_train_control_net_lllite.py	25	CODE
LOW	sdxl_train_control_net_lllite.py	26	CODE
LOW	sdxl_train_control_net_lllite.py	38	CODE
LOW	sdxl_train_control_net_lllite.py	54	CODE
LOW	sdxl_train_control_net_lllite.py	54	CODE
LOW	fine_tune.py	44	CODE
LOW	lumina_minimal_inference.py	6	CODE
LOW	lumina_minimal_inference.py	12	CODE
LOW	lumina_minimal_inference.py	13	CODE
LOW	lumina_minimal_inference.py	15	CODE
LOW	lumina_minimal_inference.py	18	CODE
LOW	gen_img_diffusers.py	50	CODE
LOW	gen_img_diffusers.py	56	CODE
LOW	gen_img_diffusers.py	60	CODE
LOW	gen_img_diffusers.py	73	CODE
LOW	gen_img_diffusers.py	92	CODE
LOW	anima_train.py	4	CODE
LOW	anima_train.py	10	CODE
LOW	anima_train.py	16	CODE
LOW	anima_train.py	23	CODE
LOW	train_control_net.py	2	CODE
LOW	train_control_net.py	6	CODE
LOW	train_control_net.py	37	CODE
LOW	flux_minimal_inference.py	34	CODE
LOW	sdxl_train_textual_inversion.py	5	CODE
LOW	flux_train.py	14	CODE
LOW	flux_train.py	19	CODE
LOW	flux_train.py	20	CODE
LOW	flux_train.py	20	CODE
LOW	flux_train.py	20	CODE
LOW	flux_train.py	20	CODE
LOW	flux_train.py	26	CODE
LOW	anima_minimal_inference_control_net_lllite.py	35	CODE
LOW	train_network.py	19	CODE
LOW	train_network.py	49	CODE
LOW	flux_train_network.py	3	CODE
LOW	flux_train_network.py	4	CODE
LOW	sd3_minimal_inference.py	7	CODE
310 more matches not shown…

Deep Nesting253 hits · 249 pts

Severity	File	Line	Context
LOW	train_textual_inversion.py	193	CODE
LOW	sd3_train_network.py	206	CODE
LOW	train_db.py	60	CODE
LOW	sdxl_train.py	59	CODE
LOW	sdxl_train.py	87	CODE
LOW	sdxl_train.py	106	CODE
LOW	anima_train_control_net_lllite.py	282	CODE
LOW	flux_train_control_net.py	68	CODE
LOW	lumina_train_network.py	117	CODE
LOW	sdxl_train_control_net_lllite.py	86	CODE
LOW	fine_tune.py	54	CODE
LOW	gen_img_diffusers.py	1817	CODE
LOW	gen_img_diffusers.py	1906	CODE
LOW	gen_img_diffusers.py	1957	CODE
LOW	gen_img_diffusers.py	1982	CODE
LOW	gen_img_diffusers.py	2238	CODE
LOW	gen_img_diffusers.py	2371	CODE
LOW	gen_img_diffusers.py	630	CODE
LOW	gen_img_diffusers.py	726	CODE
LOW	gen_img_diffusers.py	3043	CODE
LOW	anima_train.py	50	CODE
LOW	hunyuan_image_train_network.py	46	CODE
LOW	hunyuan_image_train_network.py	149	CODE
LOW	hunyuan_image_train_network.py	209	CODE
LOW	hunyuan_image_train_network.py	448	CODE
LOW	train_control_net.py	71	CODE
LOW	anima_train_network.py	184	CODE
LOW	flux_minimal_inference.py	201	CODE
LOW	flux_minimal_inference.py	282	CODE
LOW	flux_train.py	62	CODE
LOW	anima_minimal_inference_control_net_lllite.py	242	CODE
LOW	train_network.py	72	CODE
LOW	train_network.py	371	CODE
LOW	train_network.py	499	CODE
LOW	train_network.py	898	CODE
LOW	train_network.py	1561	CODE
LOW	lumina_train.py	64	CODE
LOW	sdxl_train_leco.py	103	CODE
LOW	flux_train_network.py	231	CODE
LOW	sdxl_train_control_net.py	76	CODE
LOW	train_leco.py	101	CODE
LOW	hunyuan_image_minimal_inference.py	169	CODE
LOW	hunyuan_image_minimal_inference.py	655	CODE
LOW	hunyuan_image_minimal_inference.py	942	CODE
LOW	hunyuan_image_minimal_inference.py	1086	CODE
LOW	hunyuan_image_minimal_inference.py	1170	CODE
LOW	sdxl_gen_img.py	962	CODE
LOW	sdxl_gen_img.py	1051	CODE
LOW	sdxl_gen_img.py	1102	CODE
LOW	sdxl_gen_img.py	1127	CODE
LOW	sdxl_gen_img.py	1344	CODE
LOW	sdxl_gen_img.py	1482	CODE
LOW	sdxl_gen_img.py	396	CODE
LOW	sdxl_gen_img.py	2100	CODE
LOW	anima_minimal_inference.py	150	CODE
LOW	anima_minimal_inference.py	518	CODE
LOW	anima_minimal_inference.py	766	CODE
LOW	anima_minimal_inference.py	894	CODE
LOW	anima_minimal_inference.py	976	CODE
LOW	gen_img.py	962	CODE
193 more matches not shown…

Modern Structural Boilerplate145 hits · 144 pts

Severity	File	Line	Snippet	Context
LOW	train_textual_inversion.py	50	logger = logging.getLogger(__name__)	CODE
LOW	sd3_train_network.py	25	logger = logging.getLogger(__name__)	CODE
LOW	train_db.py	55	logger = logging.getLogger(__name__)	CODE
LOW	sdxl_train.py	36	logger = logging.getLogger(__name__)	CODE
LOW	anima_train_control_net_lllite.py	68	logger = logging.getLogger(__name__)	CODE
LOW	flux_train_control_net.py	56	logger = logging.getLogger(__name__)	CODE
LOW	sdxl_train_network.py	20	logger = logging.getLogger(__name__)	CODE
LOW	lumina_train_network.py	32	logger = logging.getLogger(__name__)	CODE
LOW	sdxl_train_control_net_lllite.py	69	logger = logging.getLogger(__name__)	CODE
LOW	fine_tune.py	26	logger = logging.getLogger(__name__)	CODE
LOW	lumina_minimal_inference.py	36	logger = logging.getLogger(__name__)	CODE
LOW	gen_img_diffusers.py	110	logger = logging.getLogger(__name__)	CODE
LOW	anima_train.py	39	logger = logging.getLogger(__name__)	CODE
LOW	hunyuan_image_train_network.py	39	logger = logging.getLogger(__name__)	CODE
LOW	train_control_net.py	54	logger = logging.getLogger(__name__)	CODE
LOW	anima_train_network.py	34	logger = logging.getLogger(__name__)	CODE
LOW⚡	inpainting_minimal_inference.py	53	logger = logging.getLogger(__name__)	CODE
LOW	flux_minimal_inference.py	31	logger = logging.getLogger(__name__)	CODE
LOW	flux_train.py	50	logger = logging.getLogger(__name__)	CODE
LOW⚡	anima_minimal_inference_control_net_lllite.py	52	logger = logging.getLogger(__name__)	CODE
LOW	train_network.py	63	logger = logging.getLogger(__name__)	CODE
LOW	lumina_train.py	52	logger = logging.getLogger(__name__)	CODE
LOW	sdxl_train_leco.py	40	logger = logging.getLogger(__name__)	CODE
LOW	flux_train_network.py	32	logger = logging.getLogger(__name__)	CODE
LOW	sdxl_train_control_net.py	59	logger = logging.getLogger(__name__)	CODE
LOW	sd3_minimal_inference.py	28	logger = logging.getLogger(__name__)	CODE
LOW	train_leco.py	39	logger = logging.getLogger(__name__)	CODE
LOW	hunyuan_image_minimal_inference.py	38	logger = logging.getLogger(__name__)	CODE
LOW	sdxl_gen_img.py	69	logger = logging.getLogger(__name__)	CODE
LOW	anima_minimal_inference.py	39	logger = logging.getLogger(__name__)	CODE
LOW	sdxl_minimal_inference.py	32	logger = logging.getLogger(__name__)	CODE
LOW	gen_img.py	73	logger = logging.getLogger(__name__)	CODE
LOW	sd3_train.py	44	logger = logging.getLogger(__name__)	CODE
LOW	tools/latent_upscaler.py	23	logger = logging.getLogger(__name__)	CODE
LOW	tools/original_control_net.py	13	logger = logging.getLogger(__name__)	CODE
LOW	tools/convert_diffusers20_original_sd.py	12	logger = logging.getLogger(__name__)	CODE
LOW	tools/merge_sd3_safetensors.py	15	logger = logging.getLogger(__name__)	CODE
LOW	tools/detect_face_rotate.py	21	logger = logging.getLogger(__name__)	CODE
LOW	tools/cache_text_encoder_outputs.py	38	logger = logging.getLogger(__name__)	CODE
LOW	tools/canny.py	7	logger = logging.getLogger(__name__)	CODE
LOW	tools/show_metadata.py	7	logger = logging.getLogger(__name__)	CODE
LOW	tools/cache_latents.py	28	logger = logging.getLogger(__name__)	CODE
LOW	tools/cache_latents.py	31	def set_tokenize_strategy(is_sd: bool, is_sdxl: bool, is_flux: bool, args: argparse.Namespace) -> None:	CODE
LOW	tools/convert_diffusers_to_flux.py	39	logger = logging.getLogger(__name__)	CODE
LOW	tools/merge_models.py	11	logger = logging.getLogger(__name__)	CODE
LOW	tools/resize_images_to_resolution.py	12	logger = logging.getLogger(__name__)	CODE
LOW	library/anima_models.py	132	logger = logging.getLogger(__name__)	CODE
LOW	library/hunyuan_image_models.py	19	logger = logging.getLogger(__name__)	CODE
LOW	library/sdxl_original_unet.py	38	logger = logging.getLogger(__name__)	CODE
LOW⚡	library/sdxl_original_unet.py	1050	def set_use_memory_efficient_attention(self, xformers: bool, mem_eff: bool) -> None:	STRING
LOW⚡	library/sdxl_original_unet.py	1058	def set_use_sdpa(self, sdpa: bool) -> None:	STRING
LOW	library/flux_train_utils.py	30	logger = logging.getLogger(__name__)	CODE
LOW	library/strategy_flux.py	18	logger = logging.getLogger(__name__)	CODE
LOW	library/compile_utils.py	25	logger = logging.getLogger(__name__)	CODE
LOW	library/strategy_base.py	18	logger = logging.getLogger(__name__)	CODE
LOW	library/strategy_sdxl.py	15	logger = logging.getLogger(__name__)	CODE
LOW	library/fp8_optimization_utils.py	18	logger = logging.getLogger(__name__)	CODE
LOW	library/hunyuan_image_text_encoder.py	23	logger = logging.getLogger(__name__)	CODE
LOW	library/qwen_image_autoencoder_kl.py	38	logger = logging.getLogger(__name__)	CODE
LOW	library/model_util.py	22	logger = logging.getLogger(__name__)	CODE
85 more matches not shown…

Self-Referential Comments31 hits · 104 pts

Severity	File	Line	Snippet	Context
MEDIUM	flux_train_network.py	168	# This method is called before `assert_extra_args`, so we cannot use `self.is_schnell` here.	COMMENT
MEDIUM	hunyuan_image_minimal_inference.py	512	# Define a function to move models to device if needed	COMMENT
MEDIUM	anima_minimal_inference.py	387	# Define a function to move models to device if needed	COMMENT
MEDIUM	tests/test_custom_offloading_utils.py	73	# Create a simple model with weights	COMMENT
MEDIUM	tests/test_custom_offloading_utils.py	194	# Create a fake future and test waiting	COMMENT
MEDIUM	tests/test_custom_offloading_utils.py	343	# Create a mini model with 4 blocks	COMMENT
MEDIUM	tests/library/test_sai_model_spec.py	187	# Create a tiny test PNG (1x1 pixel)	COMMENT
MEDIUM⚡	tests/library/test_sai_model_spec.py	225	# Create a test image file	COMMENT
MEDIUM	tests/library/test_strategy_lumina.py	25	# Create a mock output object with hidden states	COMMENT
MEDIUM	tests/library/test_strategy_lumina.py	77	# Create a mock model	COMMENT
MEDIUM⚡	tests/library/test_strategy_lumina.py	125	# Create a temporary directory for caching	COMMENT
MEDIUM⚡	tests/library/test_strategy_lumina.py	127	# Create a cache file path	COMMENT
MEDIUM⚡	tests/library/test_strategy_lumina.py	130	# Create the caching strategy	COMMENT
MEDIUM⚡	tests/library/test_strategy_lumina.py	137	# Create a mock class for ImageInfo	COMMENT
MEDIUM⚡	tests/library/test_strategy_lumina.py	143	# Create a sample input info	COMMENT
MEDIUM⚡	tests/library/test_strategy_lumina.py	187	# Create a temporary directory for caching	COMMENT
MEDIUM⚡	tests/library/test_strategy_lumina.py	195	# Create a smaller dummy image for testing	COMMENT
MEDIUM⚡	tests/library/test_strategy_lumina.py	198	# Create the caching strategy	COMMENT
MEDIUM⚡	tests/library/test_strategy_lumina.py	201	# Create a simple mock VAE	COMMENT
MEDIUM	tests/library/test_lumina_util.py	8	# Create a test tensor	COMMENT
MEDIUM	tests/library/test_lumina_util.py	23	# Create a test tensor	COMMENT
MEDIUM	tests/library/test_lumina_train_util.py	113	# Create a mock noise scheduler	COMMENT
MEDIUM	tests/library/test_lumina_train_util.py	182	# Create a mock scheduler	COMMENT
MEDIUM	tests/library/test_lumina_train_util.py	196	# Create a mock args and setup	COMMENT
MEDIUM	tests/library/test_lumina_models.py	179	# Create a small NextDiT model for testing	COMMENT
MEDIUM	library/strategy_sdxl.py	87	# Create a mask where the EOS tokens are	STRING
MEDIUM	library/fp8_optimization_utils.py	472	# Create a new forward method with the patched version.	COMMENT
MEDIUM	library/qwen_image_autoencoder_kl.py	533	# Create the components	STRING
MEDIUM	library/hidden_states.py	97	# Create a mask where the EOS tokens are	STRING
MEDIUM	library/lumina_train_util.py	70	# Create a key based on the parameters	COMMENT
MEDIUM⚡	library/sai_model_spec.py	480	# Create the dataclass	COMMENT

AI Structural Patterns67 hits · 65 pts

Severity	File	Line	Context
LOW	gen_img_diffusers.py	726	CODE
LOW	gen_img_diffusers.py	1302	CODE
LOW	gen_img_diffusers.py	1394	CODE
LOW	gen_img_diffusers.py	1485	CODE
LOW	sdxl_gen_img.py	1189	CODE
LOW	sdxl_gen_img.py	396	CODE
LOW	gen_img.py	1194	CODE
LOW	gen_img.py	407	CODE
LOW	tools/latent_upscaler.py	71	CODE
LOW	tools/dev/manual_test_anima_lllite_dryrun.py	98	CODE
LOW	library/anima_models.py	1041	CODE
LOW	library/sdxl_original_unet.py	512	CODE
LOW	library/sdxl_original_unet.py	535	CODE
LOW	library/sdxl_original_unet.py	553	CODE
LOW	library/sdxl_original_unet.py	762	CODE
LOW	library/fp8_optimization_utils.py	239	CODE
LOW	library/lpw_stable_diffusion.py	724	CODE
LOW	library/lpw_stable_diffusion.py	984	CODE
LOW	library/lpw_stable_diffusion.py	1079	CODE
LOW	library/lpw_stable_diffusion.py	1173	CODE
LOW	library/qwen_image_autoencoder_kl.py	574	CODE
LOW	library/qwen_image_autoencoder_kl.py	753	CODE
LOW	library/qwen_image_autoencoder_kl.py	868	CODE
LOW	library/sdxl_lpw_stable_diffusion.py	717	CODE
LOW	library/sdxl_lpw_stable_diffusion.py	1024	CODE
LOW	library/sdxl_lpw_stable_diffusion.py	1119	CODE
LOW	library/sdxl_lpw_stable_diffusion.py	1213	CODE
LOW	library/attention_processors.py	227	CODE
LOW	library/qwen_image_autoencoder_kl_2d.py	230	CODE
LOW	library/sd3_models.py	250	CODE
LOW	library/sd3_models.py	574	CODE
LOW	library/sd3_models.py	748	CODE
LOW	library/sd3_models.py	1331	CODE
LOW	library/original_unet.py	706	CODE
LOW	library/original_unet.py	729	CODE
LOW	library/original_unet.py	747	CODE
LOW	library/slicing_vae.py	178	CODE
LOW	library/slicing_vae.py	361	CODE
LOW	library/slicing_vae.py	563	CODE
LOW	library/mask_generator.py	304	CODE
LOW	library/lumina_models.py	827	CODE
LOW	library/lumina_models.py	1035	CODE
LOW	library/lumina_models.py	1219	CODE
LOW	library/hunyuan_image_modules.py	494	CODE
LOW	library/sai_model_spec.py	378	CODE
LOW	library/sai_model_spec.py	501	CODE
LOW	finetune/blip/med.py	393	CODE
LOW	finetune/blip/med.py	670	CODE
LOW	finetune/blip/med.py	830	CODE
LOW	finetune/blip/vit.py	91	CODE
LOW	finetune/blip/vit.py	118	CODE
LOW	networks/extract_lora_from_models.py	35	CODE
LOW	networks/control_net_lllite_anima.py	315	CODE
LOW	networks/dylora.py	121	CODE
LOW	networks/dylora.py	275	CODE
LOW	networks/oft_flux.py	130	CODE
LOW	networks/lokr.py	82	CODE
LOW	networks/lora_flux.py	38	CODE
LOW	networks/lora_flux.py	724	CODE
LOW	networks/flux_extract_lora.py	36	CODE
7 more matches not shown…

Excessive Try-Catch Wrapping55 hits · 60 pts

Severity	File	Line	Snippet	Context
LOW	anima_train_control_net_lllite.py	846	except Exception as e:	CODE
LOW	hunyuan_image_train_network.py	97	except Exception:	CODE
LOW	hunyuan_image_minimal_inference.py	1041	except Exception as e:	CODE
LOW	anima_minimal_inference.py	854	except Exception as e:	CODE
LOW	tools/merge_sd3_safetensors.py	56	except Exception as e:	CODE
LOW	tools/dev/manual_test_anima_real_training.py	102	except Exception as e:	CODE
LOW	tools/dev/manual_test_anima_real_training.py	231	except Exception:	CODE
LOW	tools/dev/manual_test_anima_cache.py	342	except Exception as e:	CODE
LOW	tools/dev/manual_test_anima_cache.py	527	except Exception as e:	CODE
LOW⚡	tools/dev/manual_test_anima_cache.py	573	except Exception as e:	CODE
LOW⚡	tools/dev/manual_test_anima_cache.py	582	except Exception as e:	CODE
LOW⚡	tools/dev/manual_test_anima_cache.py	591	except Exception as e:	CODE
LOW⚡	tools/dev/visualize_masks.py	115	except Exception as e:	CODE
LOW	tests/download_training_data.py	80	except Exception:	CODE
MEDIUM	tests/download_training_data.py	108	print("Error: 'datasets' package not installed. Run: pip install datasets", file=sys.stderr)	CODE
LOW	library/flux_train_utils.py	88	except Exception:	CODE
LOW	library/strategy_flux.py	132	except Exception as e:	CODE
LOW	library/strategy_base.py	471	except Exception as e:	STRING
LOW	library/strategy_sdxl.py	250	except Exception as e:	STRING
LOW	library/hunyuan_image_text_encoder.py	659	except Exception as e:	STRING
LOW	library/huggingface_util.py	38	except Exception as e: # とりあえずRepositoryNotFoundErrorは確認したが他にあると困るので	CODE
LOW	library/huggingface_util.py	61	except Exception as e: # RuntimeErrorを確認済みだが他にあると困るので	CODE
MEDIUM	library/huggingface_util.py	45	def uploader():	CODE
LOW	library/lumina_train_util.py	192	except Exception:	CODE
LOW	library/sd3_train_utils.py	423	except Exception:	CODE
LOW	library/strategy_sd3.py	307	except Exception as e:	CODE
LOW	library/anima_train_utils.py	568	except Exception:	CODE
LOW	library/dataset.py	919	except Exception as e:	STRING
LOW	library/caching.py	62	except Exception as e:	CODE
LOW	library/sampling.py	345	except Exception:	CODE
LOW⚡	library/device_utils.py	12	except Exception:	CODE
LOW⚡	library/device_utils.py	18	except Exception:	CODE
LOW⚡	library/device_utils.py	23	except Exception:	CODE
LOW⚡	library/device_utils.py	28	except Exception:	CODE
LOW	library/device_utils.py	107	except Exception as e:	STRING
LOW	library/strategy_hunyuan_image.py	118	except Exception as e:	CODE
LOW	library/config_util.py	691	except Exception:	CODE
LOW	library/config_util.py	699	except Exception:	CODE
LOW	library/strategy_lumina.py	195	except Exception as e:	CODE
LOW⚡	library/sai_model_spec.py	472	except Exception as e:	CODE
LOW	library/strategy_anima.py	193	except Exception as e:	CODE
LOW	library/ipex/hijacks.py	425	except Exception:	CODE
LOW	library/ipex/__init__.py	7	except Exception:	CODE
LOW	library/ipex/__init__.py	54	except Exception:	CODE
LOW	library/ipex/__init__.py	229	except Exception: # pylint: disable=broad-exception-caught	CODE
LOW	library/ipex/__init__.py	232	except Exception as e:	CODE
LOW	finetune/make_captions_by_git.py	141	except Exception as e:	STRING
LOW	finetune/tag_images_by_wd14_tagger.py	93	except Exception as e:	CODE
LOW	finetune/tag_images_by_wd14_tagger.py	187	except Exception:	CODE
LOW	finetune/make_captions.py	57	except Exception as e:	CODE
LOW	finetune/make_captions.py	148	except Exception as e:	CODE
LOW	finetune/image_loading_dataset.py	28	except Exception as e:	CODE
LOW	finetune/prepare_buckets_latents.py	152	except Exception as e:	CODE
LOW	networks/svd_merge_lora.py	232	except Exception:	CODE
MEDIUM	networks/svd_merge_lora.py	228	def format_lbws(lbws):	CODE

Over-Commented Block45 hits · 44 pts

Severity	File	Line	Snippet	Context
LOW	sd3_train_network.py	281	# def call_unet(self, args, accelerator, unet, noisy_latents, timesteps, text_conds, batch, weight_dtype):	COMMENT
LOW	sdxl_train_control_net_lllite.py	301		COMMENT
LOW	requirements.txt	21	voluptuous==0.15.2	COMMENT
LOW	requirements.txt	41	# protobuf==3.20.3	COMMENT
LOW	inpainting_minimal_inference.py	1	# Minimal inpainting inference script for SD1.5 and SDXL inpainting models.	COMMENT
LOW	train_network.py	1161	trainable_params = network.prepare_optimizer_params(text_encoder_lr, args.unet_lr)	COMMENT
LOW	lumina_train.py	321	# if args.blocks_to_swap is None:	COMMENT
LOW	hunyuan_image_minimal_inference.py	341		COMMENT
LOW	sdxl_gen_img.py	721	with torch.no_grad():	COMMENT
LOW	sdxl_gen_img.py	741		COMMENT
LOW	sdxl_gen_img.py	761		COMMENT
LOW	sdxl_gen_img.py	781	# latents.dtype	COMMENT
LOW	sdxl_gen_img.py	861	# # not working yet	COMMENT
LOW	sdxl_minimal_inference.py	121	# checkpointを読み込む。モデル変換についてはそちらの関数を参照	COMMENT
LOW	sd3_train.py	841	# Follow: Section 5 of https://arxiv.org/abs/2206.00364.	COMMENT
LOW	sd3_train.py	1041	# parser.add_argument(	COMMENT
LOW	tools/detect_face_rotate.py	61	# # 回転する分、すこし画像サイズを大きくする→とりあえず無効化	COMMENT
LOW	tools/convert_diffusers_to_flux.py	1	# This script converts the diffusers of a Flux model to a safetensors file of a Flux.1 model.	COMMENT
LOW	tests/run_sdxl_inpainting_test.sh	1	#!/usr/bin/env bash	COMMENT
LOW	tests/sd15_inpainting_test_lora.toml	1	# SD1.5 inpainting smoke-test config — lora mode.	COMMENT
LOW	tests/sd15_inpainting_test_ft.toml	1	# SD1.5 inpainting smoke-test config — ft (DreamBooth / full fine-tune) mode.	COMMENT
LOW	tests/sd15_inpainting_test_ft.toml	61	#sample_sampler = "euler_a"	COMMENT
LOW	tests/sdxl_inpainting_test_lora.toml	1	# SDXL inpainting smoke-test config — lora mode.	COMMENT
LOW	tests/sdxl_inpainting_test_ft.toml	1	# SDXL inpainting smoke-test config — ft (full fine-tune) mode.	COMMENT
LOW	tests/sdxl_inpainting_test_ft.toml	61	# --- sampling (disabled for smoke test) ---	COMMENT
LOW	tests/run_sd15_inpainting_test.sh	1	#!/usr/bin/env bash	COMMENT
LOW	library/anima_models.py	621		COMMENT
LOW	library/anima_models.py	1621	# def get_dit_config(state_dict, key_prefix=""):	COMMENT
LOW	library/anima_models.py	1641	# dit_config["max_fps"] = 30	COMMENT
LOW	library/anima_models.py	1661	# dit_config["extra_per_block_abs_pos_emb"] = False	COMMENT
LOW	library/qwen_image_autoencoder_kl.py	1	# Copied and modified from Diffusers (via Musubi-Tuner). Original copyright notice follows.	COMMENT
LOW	library/sdxl_train_util.py	361		COMMENT
LOW	library/dataset.py	621	min_size and max_size are ignored when enable_bucket is False	COMMENT
LOW	library/slicing_vae.py	1	# Modified from Diffusers to reduce VRAM usage	COMMENT
LOW	library/slicing_vae.py	81	if org_dtype == torch.float16:	COMMENT
LOW	library/args.py	101	)	COMMENT
LOW	library/args.py	841		COMMENT
LOW	library/lumina_models.py	1	# Copyright Alpha VLLM/Lumina Image 2.0 and contributors	COMMENT
LOW	library/hunyuan_image_modules.py	501		COMMENT
LOW	networks/control_net_lllite.py	401		COMMENT
LOW	networks/lora_flux.py	1061	del state_dict[key]	COMMENT
LOW	networks/lora_lumina.py	441	# rank = None	COMMENT
LOW	networks/lora_lumina.py	721	# # split qkv	COMMENT
LOW	networks/lora_lumina.py	741	# alpha = state_dict.pop(f"{lora_name}.alpha")	COMMENT
LOW	networks/control_net_lllite_for_train.py	441	# logger.info(info)	COMMENT

Redundant / Tautological Comments14 hits · 22 pts

Severity	File	Line	Snippet	Context
LOW	hunyuan_image_minimal_inference.py	1174	# Check if latents are provided	COMMENT
LOW	anima_minimal_inference.py	980	# Check if latents are provided	COMMENT
LOW	tools/resize_images_to_resolution.py	28	# Check if the image is png, jpg or webp etc...	COMMENT
LOW	tools/resize_images_to_resolution.py	49	# Check if the image needs resizing	COMMENT
LOW	tools/dev/manual_test_anima_real_training.py	85	# Check if it's a known error	COMMENT
LOW	tools/dev/manual_test_anima_cache.py	335	# Check if any precision was lost	COMMENT
LOW	library/fp8_optimization_utils.py	132	# Check if it's a weight key and matches target patterns	COMMENT
LOW	library/fp8_optimization_utils.py	286	# Check if weight key matches target patterns and does not match exclude patterns	COMMENT
LOW	library/fp8_optimization_utils.py	462	# Check if this module has a corresponding scale_weight	COMMENT
LOW	library/sai_model_spec.py	307	# Read file and encode as base64	COMMENT
LOW⚡	library/sai_model_spec.py	464	# Check if it's already a data URL or if it's a file path	COMMENT
LOW	finetune/tag_images_by_wd14_tagger.py	504	# Check if file exists	COMMENT
LOW	finetune/tag_images_by_wd14_tagger.py	507	# Read file and remove new lines	COMMENT
LOW	networks/lora_lumina.py	582	# Set dims to type_dims	COMMENT

Cross-Language Confusion4 hits · 18 pts

Severity	File	Line	Snippet	Context
HIGH	library/hunyuan_image_text_encoder.py	284	"image_token_id": null,	CODE
HIGH	library/hunyuan_image_text_encoder.py	334	"sliding_window": null,	CODE
HIGH	library/hunyuan_image_text_encoder.py	338	"video_token_id": null,	CODE
HIGH	library/original_unet.py	94	"num_class_embeds": null,	STRING

Docstring Block Structure3 hits · 15 pts

Severity	File	Line	Snippet	Context
HIGH	library/safetensors_utils.py	142	Load a tensor from the file with memory-efficient strategies. Note: If device is 'cuda' , the trans	STRING
HIGH	library/safetensors_utils.py	269	Convert byte tensor to float8 format if supported. Args: byte_tensor (torch.Tensor): Raw byte tenso	STRING
HIGH	library/utils.py	99	Convert a string to a torch.dtype Args: s: string representation of the dtype default_dtype: d	STRING

AI Slop Vocabulary6 hits · 9 pts

Severity	File	Line	Snippet	Context
LOW	hunyuan_image_train_network.py	101	# If only one device is available, just use the original prompt list. We don't need to care about the distributi	COMMENT
LOW	library/flux_train_utils.py	92	# If only one device is available, just use the original prompt list. We don't need to care about the distributi	COMMENT
LOW	library/lumina_train_util.py	198	# If only one device is available, just use the original prompt list. We don't need to care about the distributi	COMMENT
LOW	library/sd3_models.py	831	# just use a buffer already	COMMENT
LOW	library/sd3_train_utils.py	427	# If only one device is available, just use the original prompt list. We don't need to care about the distributi	COMMENT
LOW	library/sampling.py	349	# If only one device is available, just use the original prompt list. We don't need to care about the distributi	COMMENT

Example Usage Blocks2 hits · 3 pts

Severity	File	Line	Snippet	Context
LOW	tests/run_sdxl_inpainting_test.sh	14	# Usage:	COMMENT
LOW	tests/run_sd15_inpainting_test.sh	14	# Usage:	COMMENT

Slop Phrases1 hit · 2 pts

Severity	File	Line	Snippet	Context
LOW	sd3_minimal_inference.py	283	# make sure to move to the device: some tensors are created in the constructor on the CPU	COMMENT

Analysis Overview

What These Metrics Mean

Score History

Severity Breakdown

Directory Score Breakdown

Pattern Findings