kyutai-labs/pocket-tts

16.9

Adjusted Score

16.9

Raw Score

100%

Time Factor

2026-08-28

Last Push

9.2K

Stars

Python

Language

13.6K

Lines of Code

110

Files

161

Pattern Hits

2026-08-29

Scan Date

0.04

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 0HIGH 4MEDIUM 15LOW 142

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 161 distinct pattern matches across 10 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 Identifiers94 hits · 113 pts

Severity	File	Line	Snippet	Context
LOW	pocket_tts/quantization.py	52	def _ensure_quantization_engine():	CODE
LOW	pocket_tts/default_parameters.py	59	def get_default_text_for_language(language: str \| None) -> str:	CODE
LOW	pocket_tts/default_parameters.py	66	def get_default_voice_for_language(	CODE
LOW	pocket_tts/utils/weights_loading.py	82	def get_training_checkpoint_state_dicts(	CODE
LOW	pocket_tts/models/text_chunking.py	44	def _is_decimal_period_boundary(	CODE
LOW	pocket_tts/models/text_chunking.py	92	def _segments_from_boundaries(	CODE
LOW	pocket_tts/models/text_chunking.py	105	def split_into_best_sentences(	CODE
LOW	pocket_tts/models/tts_model.py	183	def _from_pydantic_config_with_weights(	CODE
LOW	pocket_tts/models/tts_model.py	365	def _run_flow_lm_and_increment_step(	CODE
LOW	pocket_tts/models/tts_model.py	676	def _generate_audio_stream_short_text(	CODE
LOW	pocket_tts/models/tts_model.py	788	def _autoregressive_generation(	CODE
LOW	pocket_tts/models/tts_model.py	825	def _cached_get_state_for_audio_prompt(	CODE
LOW	pocket_tts/models/tts_model.py	831	def get_state_for_audio_prompt(	CODE
LOW	pocket_tts/data/audio.py	44	def _audio_read_with_soundfile(filepath: Path) -> tuple[torch.Tensor, int]:	CODE
LOW	training/checkpointing.py	125	def export_pocket_safetensors(	CODE
LOW⚡	training/tests/test_dataloader.py	52	def test_rank_sharding_partitions_entries_without_overlap(tmp_path):	CODE
LOW⚡	training/tests/test_dataloader.py	59	def test_prompt_respects_the_configured_cap(tmp_path):	CODE
LOW⚡	training/tests/test_dataloader.py	64	def test_batches_have_the_requested_size(tmp_path):	CODE
LOW⚡	training/tests/test_dataloader.py	70	def test_target_audio_never_exceeds_max_duration(tmp_path):	CODE
LOW⚡	training/tests/test_dataloader.py	75	def test_unaligned_manifest_still_yields_batches(tmp_path):	CODE
LOW⚡	training/tests/test_dataloader.py	83	def test_entry_start_offsets_into_a_shared_file(tmp_path):	CODE
LOW	training/tests/test_smoke.py	137	def test_head_supports_every_time_cond_count(num_time_conds):	CODE
LOW	training/tests/test_smoke.py	149	def test_head_keeps_released_state_dict_layout():	CODE
LOW	training/tests/test_smoke.py	175	def test_generate_ragged_matches_batch_of_one():	CODE
LOW	training/tests/test_smoke.py	201	def test_generate_per_row_eos():	CODE
LOW	training/tests/test_smoke.py	214	def test_padded_batch_matches_unpadded_under_context_window():	CODE
LOW	training/tests/test_smoke.py	246	def test_ema_load_drops_untracked_keys():	CODE
LOW	training/tests/test_smoke.py	328	def test_grad_accum_matches_big_batch():	CODE
LOW⚡	training/tests/test_config_invariants.py	30	def test_scratch_reaches_the_effective_batch_floor():	CODE
LOW⚡	training/tests/test_config_invariants.py	38	def test_scratch_runs_past_the_quality_transition():	CODE
LOW⚡	training/tests/test_config_invariants.py	42	def test_scratch_builds_the_reference_teacher_depth():	CODE
LOW⚡	training/tests/test_config_invariants.py	48	def test_distill_teacher_is_deeper_than_its_student():	CODE
LOW⚡	training/tests/test_config_invariants.py	57	def test_distill_teacher_weights_point_at_the_scratch_run():	CODE
LOW⚡	training/tests/test_config_invariants.py	69	def test_num_ckpt_keep_zero_is_rejected(self):	CODE
LOW⚡	training/tests/test_config_invariants.py	73	def test_zero_frequencies_are_rejected(self):	CODE
LOW⚡	training/tests/test_config_invariants.py	78	def test_distillation_without_a_teacher_is_rejected(self):	CODE
LOW⚡	training/tests/test_config_invariants.py	82	def test_teacher_config_without_weights_is_rejected(self):	CODE
LOW⚡	training/tests/test_config_invariants.py	86	def test_unknown_keys_are_rejected(self):	CODE
LOW⚡	training/tests/test_units.py	20	def test_warmup_ends_at_the_configured_lr(self):	CODE
LOW⚡	training/tests/test_units.py	26	def test_cosine_decays_to_the_floor_at_max_steps(self):	CODE
LOW⚡	training/tests/test_units.py	32	def test_cosine_respects_lr_min_ratio(self):	CODE
LOW⚡	training/tests/test_units.py	45	def test_keeps_the_ends_of_the_stack(self):	CODE
LOW⚡	training/tests/test_units.py	48	def test_identity_when_depths_match(self):	CODE
LOW⚡	training/tests/test_units.py	51	def test_rejects_growing_the_stack(self):	CODE
LOW⚡	training/tests/test_units.py	55	def test_shrink_renumbers_kept_layers_contiguously(self):	CODE
LOW⚡	training/tests/test_units.py	76	def test_tracks_only_trainable_parameters(self):	CODE
LOW⚡	training/tests/test_units.py	82	def test_update_moves_towards_the_live_weights(self):	CODE
LOW⚡	training/tests/test_units.py	90	def test_loading_a_partial_shadow_keeps_untracked_weights(self):	CODE
LOW	training/tests/test_units.py	104	def test_skip_rate_matches_probability(self):	CODE
LOW	training/tests/test_units.py	111	def test_ranks_agree_on_every_draw(self):	CODE
LOW	training/tests/test_units.py	126	def test_eval_mode_always_computes_the_full_loss(self):	CODE
LOW	training/tests/test_eval_naming.py	19	def test_default_name_is_bare():	CODE
LOW⚡	training/tests/test_eval_naming.py	33	def test_each_knob_marks_the_name(over, fragment):	CODE
LOW⚡	training/tests/test_eval_naming.py	37	def test_seeds_do_not_collide():	CODE
LOW⚡	training/tests/test_eval_naming.py	43	def test_every_pair_of_settings_is_distinguishable():	CODE
LOW⚡	training/tests/test_eval_naming.py	57	def test_step_prefixes_do_not_collide_under_globbing():	CODE
LOW⚡	training/tests/test_eval_naming.py	64	def test_hf_prompt_root_tags_by_repo_name_not_snapshot_path():	CODE
LOW⚡	training/tests/test_eval_naming.py	73	def test_local_prompt_root_keeps_directory_tag():	CODE
LOW	training/tests/test_prepare_data.py	19	def test_utterances_in_one_chapter_share_the_manifest_path(tmp_path, monkeypatch):	CODE
LOW	training/tests/test_prepare_data.py	94	def test_hf_alignments_join_on_utterance_id_not_path(tmp_path, monkeypatch):	CODE
34 more matches not shown…

Decorative Section Separators14 hits · 45 pts

Severity	File	Line	Snippet	Context
MEDIUM⚡	pocket_tts/main.py	39	# ------------------------------------------------------	COMMENT
MEDIUM⚡	pocket_tts/main.py	41	# ------------------------------------------------------	COMMENT
MEDIUM	pocket_tts/main.py	220	# ------------------------------------------------------	COMMENT
MEDIUM	pocket_tts/main.py	222	# ------------------------------------------------------	COMMENT
MEDIUM	pocket_tts/main.py	350	# ----------------------------------------------	COMMENT
MEDIUM	pocket_tts/main.py	352	# ----------------------------------------------	COMMENT
MEDIUM	scripts/evaluate_quantization.py	210	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	scripts/evaluate_quantization.py	212	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	scripts/evaluate_quantization.py	352	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	scripts/evaluate_quantization.py	354	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	scripts/evaluate_quantization.py	389	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	scripts/evaluate_quantization.py	391	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	scripts/evaluate_quantization.py	511	# ---------------------------------------------------------------------------	COMMENT
MEDIUM	scripts/evaluate_quantization.py	513	# ---------------------------------------------------------------------------	COMMENT

Docstring Block Structure4 hits · 20 pts

Severity	File	Line	Snippet	Context
HIGH	pocket_tts/models/tts_model.py	268	Load a pre-trained TTS model with specified configuration. This class method loads a complete TTS model includi	STRING
HIGH	pocket_tts/models/tts_model.py	527	Generate complete audio tensor from text input. This method generates the full audio output for the given text	STRING
HIGH	pocket_tts/models/tts_model.py	595	Generate audio streaming chunks from text input. This method generates audio from text and yields chunks as the	STRING
HIGH	pocket_tts/models/tts_model.py	834	Create model state conditioned on audio prompt for continuation. This method processes an audio prompt and crea	STRING

Modern Structural Boilerplate14 hits · 14 pts

Severity	File	Line	Snippet	Context
LOW	pocket_tts/quantization.py	18	logger = logging.getLogger(__name__)	CODE
LOW	pocket_tts/__init__.py	29	__all__ = ["TTSModel", "export_model_state"]	CODE
LOW⚡	pocket_tts/main.py	32	logger = logging.getLogger(__name__)	CODE
LOW	pocket_tts/models/text_chunking.py	9	logger = logging.getLogger(__name__)	CODE
LOW	pocket_tts/models/tts_model.py	52	logger = logging.getLogger(__name__)	CODE
LOW	pocket_tts/models/flow_lm.py	14	logger = logging.getLogger(__name__)	CODE
LOW	pocket_tts/conditioners/text.py	10	logger = logging.getLogger(__name__)	CODE
LOW	pocket_tts/data/audio.py	18	logger = logging.getLogger(__name__)	CODE
LOW	training/dataloader.py	26	logger = logging.getLogger(__name__)	CODE
LOW	training/checkpointing.py	16	logger = logging.getLogger(__name__)	CODE
LOW	training/modules/builders.py	27	logger = logging.getLogger(__name__)	CODE
LOW	training/modules/model.py	41	def _update_latent_stats(self, latents: torch.Tensor, mask: torch.Tensor) -> None:	CODE
LOW	training/modules/utils.py	52	def set_state_padding(model_state: dict, pad: torch.Tensor) -> None:	CODE
LOW	scripts/evaluate_quantization.py	58	logger = logging.getLogger(__name__)	CODE

Deep Nesting12 hits · 12 pts

Severity	File	Line	Context
LOW	pocket_tts/models/text_chunking.py	59	CODE
LOW	pocket_tts/models/tts_model.py	676	CODE
LOW	training/train.py	161	CODE
LOW	training/args.py	159	CODE
LOW	training/scripts/prepare_data.py	56	CODE
LOW	training/scripts/prepare_data.py	132	CODE
LOW	training/scripts/prepare_data.py	175	CODE
LOW	training/scripts/align_data.py	194	CODE
LOW	training/scripts/train_tokenizer.py	30	CODE
LOW	training/eval/librispeech.py	92	CODE
LOW	training/eval/librispeech.py	226	CODE
LOW	scripts/evaluate_quantization.py	516	CODE

Excessive Try-Catch Wrapping10 hits · 11 pts

Severity	File	Line	Snippet	Context
LOW	pocket_tts/models/tts_model.py	171	except Exception:	CODE
LOW	pocket_tts/models/tts_model.py	234	except Exception:	CODE
LOW	pocket_tts/models/tts_model.py	514	except Exception as e:	CODE
LOW	pocket_tts/models/tts_model.py	774	except Exception as e:	CODE
MEDIUM	pocket_tts/models/tts_model.py	769	def run_generation():	CODE
LOW	training/dataloader.py	200	except Exception as exc: # noqa: BLE001 — skip unreadable samples, whatever the cause	CODE
LOW	training/scripts/prepare_data.py	278	except Exception as exc: # noqa: BLE001 -- dead LibriVox links happen	CODE
LOW	training/scripts/align_data.py	258	except Exception as exc: # noqa: BLE001	CODE
LOW	scripts/evaluate_quantization.py	195	except Exception as e:	CODE
LOW	scripts/evaluate_quantization.py	264	except Exception as e:	CODE

AI Structural Patterns8 hits · 8 pts

Severity	File	Line	Context
LOW	pocket_tts/main.py	226	CODE
LOW	pocket_tts/models/tts_model.py	256	CODE
LOW	pocket_tts/models/mimi.py	97	CODE
LOW	pocket_tts/models/flow_lm.py	206	CODE
LOW	pocket_tts/modules/seanet.py	47	CODE
LOW	pocket_tts/modules/seanet.py	119	CODE
LOW	training/scripts/prepare_data.py	344	CODE
LOW	training/modules/samplers.py	70	CODE

Over-Commented Block2 hits · 2 pts

Severity	File	Line	Snippet	Context
LOW	pocket_tts/__init__.py	21	# Public methods:	COMMENT
LOW	scripts/generate_default_voices.py	21	model_state_copy = model.get_state_for_audio_prompt(dest)	COMMENT

Unused Imports2 hits · 2 pts

Severity	File	Line	Snippet	Context
LOW	pocket_tts/__init__.py	16		CODE
LOW	pocket_tts/__init__.py	17		CODE

Redundant / Tautological Comments1 hit · 2 pts

Severity	File	Line	Snippet	Context
LOW	pocket_tts/main.py	168	# Close the file before reading it back (required on Windows)	COMMENT

Analysis Overview

What These Metrics Mean

Score History

Severity Breakdown

Directory Score Breakdown

Pattern Findings