Multi-lingual large voice generation model, providing inference, training and deployment full-stack ability.
This report summarizes scanner evidence for FunAudioLLM/CosyVoice, a Python project with 23,913 GitHub stars. SynthScan examined 26,042 lines of code across 136 source files, recording 292 pattern matches distributed across 15 syntactic categories. The legacy adjusted pattern signal is 17.8 (Elevated pattern signal); it is not an authorship estimate.
Use the findings to ask whether maintainers can explain intent, verify dependencies, simplify unnecessary scaffolding, and focus review effort. Risk Payload v1 adds ownership, supply-chain, over-engineering, and review-effort vectors when available. These are prompts for human verification, not probabilities, security findings, or measured maintenance costs.
Use each vector to guide manual verification of the underlying findings.
Can maintainers explain intent, assumptions, and trade-offs? 178 finding(s).
Are imports declared and package/API names verified? 61 finding(s).
Are wrappers, guards, and repeated structures necessary? 109 finding(s).
Which changes need focused human review? 292 finding(s).
This chart tracks the legacy adjusted pattern score across scan runs. Changes can reflect source changes, scan coverage, or detector/scoring versions; they do not measure AI adoption or prove that a team refactored code. Compare runs only when their scoring versions are compatible. The dashed line tracks the raw pattern-hit count.
Groups findings by configured severity weight. Severity is not a confidence estimate, security rating, or proof that a package is missing. Verify imports against the project's manifests and inspect each finding in its local design context.
This horizontal bar chart shows where configured pattern scores accumulate across directories. Treat the bars as locations to inspect, not as evidence of authorship or code quality. Use the file-level findings to decide whether a directory needs a focused ownership, dependency, complexity, or review check.
The scanner identified 292 distinct pattern matches across 15 syntactic categories. Each entry below represents a discrete location in the source code where the engine recorded a detector match to inspect in context. Expand any category row to inspect the individual file paths, line numbers, code snippets, and the lexical context (CODE, COMMENT, or STRING) in which each match was detected.
Reading the findings table: The Severity column indicates the configured finding weight (CRITICAL / HIGH / MEDIUM / LOW). The Context column identifies whether the match occurred inside executable code, an inline comment, or a string literal — legacy comment-context matches receive a ×1.5 score weight. The ⚡ bolt icon marks clustered matches: three or more patterns within a 10-line window, each receiving an additional ×1.5 density multiplier under the legacy scoring model.
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| HIGH | runtime/triton_trtllm/token2wav_dit.py | 0 | example usage cuda_visible_devices=0 \ python3 token2wav.py --enable-trt || exit 1 | STRING |
| HIGH | runtime/triton_trtllm/token2wav.py | 0 | example usage cuda_visible_devices=0 \ python3 token2wav.py --enable-trt || exit 1 | STRING |
| HIGH | …ton_trtllm/model_repo/token2wav_dit/1/token2wav_dit.py | 0 | example usage cuda_visible_devices=0 \ python3 token2wav.py --enable-trt || exit 1 | STRING |
| HIGH | runtime/triton_trtllm/offline_inference.py | 0 | extract speech ids from token strings like <|s_23456|> | STRING |
| HIGH | runtime/triton_trtllm/infer_cosyvoice3.py | 0 | extract speech ids from token strings like <|s_23456|> | STRING |
| HIGH | examples/grpo/cosyvoice2/infer_dataset.py | 0 | extract speech ids from token strings like <|s_23456|> | STRING |
| HIGH | runtime/triton_trtllm/model_repo/cosyvoice2/1/model.py | 0 | initialize the model. args: args: dictionary containing model configuration | STRING |
| HIGH | …ime/triton_trtllm/model_repo/cosyvoice2_dit/1/model.py | 0 | initialize the model. args: args: dictionary containing model configuration | STRING |
| HIGH | runtime/triton_trtllm/model_repo/token2wav/1/model.py | 0 | initialize the model. args: args: dictionary containing model configuration | STRING |
| HIGH | …me/triton_trtllm/model_repo/audio_tokenizer/1/model.py | 0 | initialize the model. args: args: dictionary containing model configuration | STRING |
| HIGH | …time/triton_trtllm/model_repo/token2wav_dit/1/model.py | 0 | initialize the model. args: args: dictionary containing model configuration | STRING |
| HIGH | …/triton_trtllm/model_repo/speaker_embedding/1/model.py | 0 | initialize the model. args: args: dictionary containing model configuration | STRING |
| HIGH | …rtllm/model_repo_cosyvoice3/audio_tokenizer/1/model.py | 0 | initialize the model. args: args: dictionary containing model configuration | STRING |
| HIGH | …llm/model_repo_cosyvoice3/speaker_embedding/1/model.py | 0 | initialize the model. args: args: dictionary containing model configuration | STRING |
| HIGH | runtime/triton_trtllm/model_repo/cosyvoice2/1/model.py | 0 | execute inference on the batched requests. args: requests: list of inference requests returns: list of inference respons | STRING |
| HIGH | …ime/triton_trtllm/model_repo/cosyvoice2_dit/1/model.py | 0 | execute inference on the batched requests. args: requests: list of inference requests returns: list of inference respons | STRING |
| HIGH | runtime/triton_trtllm/model_repo/token2wav/1/model.py | 0 | execute inference on the batched requests. args: requests: list of inference requests returns: list of inference respons | STRING |
| HIGH | …time/triton_trtllm/model_repo/token2wav_dit/1/model.py | 0 | execute inference on the batched requests. args: requests: list of inference requests returns: list of inference respons | STRING |
| HIGH | …me/triton_trtllm/model_repo/audio_tokenizer/1/model.py | 0 | triton python model for audio tokenization. this model takes reference audio input and extracts semantic tokens using s3 | STRING |
| HIGH | …/triton_trtllm/model_repo/speaker_embedding/1/model.py | 0 | triton python model for audio tokenization. this model takes reference audio input and extracts semantic tokens using s3 | STRING |
| HIGH | …rtllm/model_repo_cosyvoice3/audio_tokenizer/1/model.py | 0 | triton python model for audio tokenization. this model takes reference audio input and extracts semantic tokens using s3 | STRING |
| HIGH | …llm/model_repo_cosyvoice3/speaker_embedding/1/model.py | 0 | triton python model for audio tokenization. this model takes reference audio input and extracts semantic tokens using s3 | STRING |
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| LOW | webui.py | 1 | # Copyright (c) 2024 Alibaba Inc (authors: Xiang Lyu, Liu Yue) | COMMENT |
| LOW | cosyvoice/llm/llm.py | 1 | # Copyright (c) 2024 Alibaba Inc (authors: Xiang Lyu, Zhihao Du) | COMMENT |
| LOW | cosyvoice/bin/export_jit.py | 1 | # Copyright (c) 2024 Alibaba Inc (authors: Xiang Lyu) | COMMENT |
| LOW | cosyvoice/bin/average_model.py | 1 | # Copyright (c) 2020 Mobvoi Inc (Di Wu) | COMMENT |
| LOW | cosyvoice/bin/train.py | 1 | # Copyright (c) 2024 Alibaba Inc (authors: Xiang Lyu) | COMMENT |
| LOW | cosyvoice/bin/export_onnx.py | 1 | # Copyright (c) 2024 Antgroup Inc (authors: Zhoubofan, hexisyztem@icloud.com) | COMMENT |
| LOW | cosyvoice/dataset/processor.py | 1 | # Copyright (c) 2024 Alibaba Inc (authors: Xiang Lyu) | COMMENT |
| LOW | cosyvoice/dataset/dataset.py | 1 | # Copyright (c) 2021 Mobvoi Inc. (authors: Binbin Zhang) | COMMENT |
| LOW | cosyvoice/hifigan/generator.py | 1 | # Copyright (c) 2024 Alibaba Inc (authors: Xiang Lyu, Kai Hu) | COMMENT |
| LOW | cosyvoice/hifigan/f0_predictor.py | 1 | # Copyright (c) 2024 Alibaba Inc (authors: Xiang Lyu, Kai Hu) | COMMENT |
| LOW | cosyvoice/transformer/decoder.py | 1 | # Copyright (c) 2021 Mobvoi Inc. (authors: Binbin Zhang, Di Wu) | COMMENT |
| LOW | cosyvoice/transformer/attention.py | 1 | # Copyright (c) 2019 Shigeki Karita | COMMENT |
| LOW | cosyvoice/transformer/attention.py | 161 | Returns: | COMMENT |
| LOW | cosyvoice/transformer/attention.py | 281 | # cache(1, head, 0, d_k * 2) (16/-1, -1/-1, 16/0 mode) | COMMENT |
| LOW | cosyvoice/transformer/subsampling.py | 1 | # Copyright (c) 2021 Mobvoi Inc (Binbin Zhang, Di Wu) | COMMENT |
| LOW | cosyvoice/transformer/convolution.py | 1 | # Copyright (c) 2020 Mobvoi Inc. (authors: Binbin Zhang, Di Wu) | COMMENT |
| LOW | cosyvoice/transformer/upsample_encoder.py | 1 | # Copyright (c) 2021 Mobvoi Inc (Binbin Zhang, Di Wu) | COMMENT |
| LOW | cosyvoice/transformer/embedding.py | 1 | # Copyright (c) 2020 Mobvoi Inc. (authors: Binbin Zhang, Di Wu) | COMMENT |
| LOW | cosyvoice/transformer/encoder_layer.py | 1 | # Copyright (c) 2021 Mobvoi Inc (Binbin Zhang, Di Wu) | COMMENT |
| LOW | cosyvoice/transformer/encoder.py | 1 | # Copyright (c) 2021 Mobvoi Inc (Binbin Zhang, Di Wu) | COMMENT |
| LOW | cosyvoice/transformer/activation.py | 1 | # Copyright (c) 2020 Johns Hopkins University (Shinji Watanabe) | COMMENT |
| LOW | cosyvoice/transformer/positionwise_feed_forward.py | 1 | # Copyright (c) 2019 Shigeki Karita | COMMENT |
| LOW | cosyvoice/transformer/decoder_layer.py | 1 | # Copyright (c) 2019 Shigeki Karita | COMMENT |
| LOW | cosyvoice/transformer/label_smoothing_loss.py | 1 | # Copyright (c) 2019 Shigeki Karita | COMMENT |
| LOW | cosyvoice/vllm/cosyvoice2.py | 1 | # SPDX-License-Identifier: Apache-2.0 | COMMENT |
| LOW | cosyvoice/utils/class_utils.py | 1 | # Copyright [2023-11-28] <sxc19@mails.tsinghua.edu.cn, Xingchen Song> | COMMENT |
| LOW | cosyvoice/utils/mask.py | 1 | # Copyright (c) 2019 Shigeki Karita | COMMENT |
| LOW | cosyvoice/utils/train_utils.py | 1 | # Copyright (c) 2021 Mobvoi Inc. (authors: Binbin Zhang) | COMMENT |
| LOW | cosyvoice/utils/common.py | 1 | # Copyright (c) 2020 Mobvoi Inc (Binbin Zhang) | COMMENT |
| LOW | cosyvoice/utils/file_utils.py | 1 | # Copyright (c) 2021 Mobvoi Inc. (authors: Binbin Zhang) | COMMENT |
| LOW | cosyvoice/utils/scheduler.py | 1 | # Copyright (c) 2020 Mobvoi Inc (Binbin Zhang) | COMMENT |
| LOW | cosyvoice/utils/frontend_utils.py | 1 | # Copyright (c) 2024 Alibaba Inc (authors: Xiang Lyu, Zhihao Du) | COMMENT |
| LOW | cosyvoice/utils/executor.py | 1 | # Copyright (c) 2020 Mobvoi Inc (Binbin Zhang) | COMMENT |
| LOW | cosyvoice/cli/frontend.py | 1 | # Copyright (c) 2024 Alibaba Inc (authors: Xiang Lyu) | COMMENT |
| LOW | cosyvoice/cli/cosyvoice.py | 1 | # Copyright (c) 2024 Alibaba Inc (authors: Xiang Lyu) | COMMENT |
| LOW | cosyvoice/cli/model.py | 1 | # Copyright (c) 2024 Alibaba Inc (authors: Xiang Lyu) | COMMENT |
| LOW | cosyvoice/flow/decoder.py | 1 | # Copyright (c) 2024 Alibaba Inc (authors: Xiang Lyu, Zhihao Du) | COMMENT |
| LOW | cosyvoice/flow/flow.py | 1 | # Copyright (c) 2024 Alibaba Inc (authors: Xiang Lyu, Zhihao Du) | COMMENT |
| LOW | cosyvoice/flow/length_regulator.py | 1 | # Copyright (c) 2024 Alibaba Inc (authors: Xiang Lyu, Zhihao Du) | COMMENT |
| LOW | cosyvoice/flow/flow_matching.py | 1 | # Copyright (c) 2024 Alibaba Inc (authors: Xiang Lyu, Zhihao Du) | COMMENT |
| LOW | tools/make_parquet_list.py | 1 | #!/usr/bin/env python3 | COMMENT |
| LOW | tools/extract_speech_token.py | 1 | #!/usr/bin/env python3 | COMMENT |
| LOW | tools/extract_embedding.py | 1 | #!/usr/bin/env python3 | COMMENT |
| LOW | runtime/python/grpc/server.py | 1 | # Copyright (c) 2024 Alibaba Inc (authors: Xiang Lyu) | COMMENT |
| LOW | runtime/python/grpc/client.py | 1 | # Copyright (c) 2024 Alibaba Inc (authors: Xiang Lyu) | COMMENT |
| LOW | runtime/python/fastapi/server.py | 1 | # Copyright (c) 2024 Alibaba Inc (authors: Xiang Lyu) | COMMENT |
| LOW | runtime/python/fastapi/client.py | 1 | # Copyright (c) 2024 Alibaba Inc (authors: Xiang Lyu) | COMMENT |
| LOW | runtime/triton_trtllm/token2wav_dit.py | 1 | # SPDX-FileCopyrightText: Copyright (c) 2025, NVIDIA CORPORATION. All rights reserved. | COMMENT |
| LOW | runtime/triton_trtllm/run_stepaudio2_dit_token2wav.sh | 101 | if [ $stage -le 3 ] && [ $stop_stage -ge 3 ]; then | COMMENT |
| LOW | runtime/triton_trtllm/client_http.py | 1 | # Copyright 2025, NVIDIA CORPORATION & AFFILIATES. All rights reserved. | COMMENT |
| LOW | runtime/triton_trtllm/offline_inference.py | 1 | # SPDX-FileCopyrightText: Copyright (c) 2025, NVIDIA CORPORATION. All rights reserved. | COMMENT |
| LOW | runtime/triton_trtllm/client_grpc.py | 1 | # Copyright 2022 Xiaomi Corp. (authors: Fangjun Kuang) | COMMENT |
| LOW | runtime/triton_trtllm/token2wav.py | 1 | # SPDX-FileCopyrightText: Copyright (c) 2025, NVIDIA CORPORATION. All rights reserved. | COMMENT |
| LOW | runtime/triton_trtllm/model_repo/cosyvoice2/1/model.py | 1 | # Copyright 2025, NVIDIA CORPORATION & AFFILIATES. All rights reserved. | COMMENT |
| LOW | …ime/triton_trtllm/model_repo/cosyvoice2_dit/1/model.py | 1 | # Copyright 2025, NVIDIA CORPORATION & AFFILIATES. All rights reserved. | COMMENT |
| LOW | runtime/triton_trtllm/model_repo/token2wav/1/model.py | 1 | # Copyright 2025, NVIDIA CORPORATION & AFFILIATES. All rights reserved. | COMMENT |
| LOW | …me/triton_trtllm/model_repo/audio_tokenizer/1/model.py | 1 | # Copyright 2025, NVIDIA CORPORATION & AFFILIATES. All rights reserved. | COMMENT |
| LOW | …ton_trtllm/model_repo/token2wav_dit/1/token2wav_dit.py | 1 | # SPDX-FileCopyrightText: Copyright (c) 2025, NVIDIA CORPORATION. All rights reserved. | COMMENT |
| LOW | …time/triton_trtllm/model_repo/token2wav_dit/1/model.py | 1 | # Copyright 2025, NVIDIA CORPORATION & AFFILIATES. All rights reserved. | COMMENT |
| LOW | …/triton_trtllm/model_repo/speaker_embedding/1/model.py | 1 | # Copyright 2025, NVIDIA CORPORATION & AFFILIATES. All rights reserved. | COMMENT |
| 10 more matches not shown… | ||||
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| LOW | webui.py | 19 | CODE | |
| LOW | webui.py | 22 | CODE | |
| LOW | cosyvoice/llm/llm.py | 18 | CODE | |
| LOW | cosyvoice/llm/llm.py | 20 | CODE | |
| LOW | cosyvoice/bin/export_jit.py | 15 | CODE | |
| LOW | cosyvoice/bin/train.py | 15 | CODE | |
| LOW | cosyvoice/bin/export_onnx.py | 16 | CODE | |
| LOW | cosyvoice/vllm/cosyvoice2.py | 35 | CODE | |
| LOW | cosyvoice/flow/DiT/modules.py | 11 | CODE | |
| LOW | cosyvoice/flow/DiT/dit.py | 11 | CODE | |
| LOW | runtime/triton_trtllm/offline_inference.py | 34 | CODE | |
| LOW | runtime/triton_trtllm/offline_inference.py | 39 | CODE | |
| LOW | runtime/triton_trtllm/streaming_inference.py | 7 | CODE | |
| LOW | runtime/triton_trtllm/token2wav_cosyvoice3.py | 12 | CODE | |
| LOW | runtime/triton_trtllm/token2wav_cosyvoice3.py | 14 | CODE | |
| LOW | runtime/triton_trtllm/token2wav_cosyvoice3.py | 19 | CODE | |
| LOW | runtime/triton_trtllm/token2wav_cosyvoice3.py | 20 | CODE | |
| LOW | …ime/triton_trtllm/model_repo/cosyvoice2_dit/1/model.py | 28 | CODE | |
| LOW | …ime/triton_trtllm/model_repo/cosyvoice2_dit/1/model.py | 29 | CODE | |
| LOW | …ime/triton_trtllm/model_repo/cosyvoice2_dit/1/model.py | 32 | CODE | |
| LOW | …ime/triton_trtllm/model_repo/cosyvoice2_dit/1/model.py | 32 | CODE | |
| LOW | …ime/triton_trtllm/model_repo/cosyvoice2_dit/1/model.py | 32 | CODE | |
| LOW | …ime/triton_trtllm/model_repo/cosyvoice2_dit/1/model.py | 42 | CODE | |
| LOW | runtime/triton_trtllm/model_repo/token2wav/1/model.py | 40 | CODE | |
| LOW | …me/triton_trtllm/model_repo/audio_tokenizer/1/model.py | 33 | CODE | |
| LOW | …time/triton_trtllm/model_repo/token2wav_dit/1/model.py | 28 | CODE | |
| LOW | …time/triton_trtllm/model_repo/token2wav_dit/1/model.py | 31 | CODE | |
| LOW | …time/triton_trtllm/model_repo/token2wav_dit/1/model.py | 31 | CODE | |
| LOW | …time/triton_trtllm/model_repo/token2wav_dit/1/model.py | 35 | CODE | |
| LOW | …time/triton_trtllm/model_repo/token2wav_dit/1/model.py | 39 | CODE | |
| LOW | …time/triton_trtllm/model_repo/token2wav_dit/1/model.py | 40 | CODE | |
| LOW | …time/triton_trtllm/model_repo/token2wav_dit/1/model.py | 41 | CODE | |
| LOW | …time/triton_trtllm/model_repo/token2wav_dit/1/model.py | 41 | CODE | |
| LOW | …time/triton_trtllm/model_repo/token2wav_dit/1/model.py | 42 | CODE | |
| LOW | …time/triton_trtllm/model_repo/token2wav_dit/1/model.py | 43 | CODE | |
| LOW | …time/triton_trtllm/model_repo/token2wav_dit/1/model.py | 44 | CODE | |
| LOW | …/triton_trtllm/model_repo/speaker_embedding/1/model.py | 33 | CODE | |
| LOW | …iton_trtllm/model_repo_cosyvoice3/token2wav/1/model.py | 7 | CODE | |
| LOW | …rtllm/model_repo_cosyvoice3/audio_tokenizer/1/model.py | 33 | CODE | |
| LOW | …llm/model_repo_cosyvoice3/speaker_embedding/1/model.py | 33 | CODE | |
| LOW | runtime/triton_trtllm/scripts/test_llm.py | 19 | CODE | |
| LOW | …time/triton_trtllm/scripts/convert_cosyvoice3_to_hf.py | 39 | CODE | |
| LOW | …time/triton_trtllm/scripts/convert_cosyvoice3_to_hf.py | 39 | CODE | |
| LOW | …time/triton_trtllm/scripts/convert_cosyvoice3_to_hf.py | 75 | CODE | |
| LOW | examples/grpo/cosyvoice2/prepare_data.py | 20 | CODE | |
| LOW | examples/grpo/cosyvoice2/infer_dataset.py | 30 | CODE | |
| LOW | examples/grpo/cosyvoice2/infer_dataset.py | 33 | CODE | |
| LOW | examples/grpo/cosyvoice2/infer_dataset.py | 37 | CODE | |
| LOW | examples/grpo/cosyvoice2/infer_dataset.py | 43 | CODE | |
| LOW | examples/grpo/cosyvoice2/pretrained_to_huggingface.py | 27 | CODE | |
| LOW | examples/grpo/cosyvoice2/token2wav_asr_server.py | 20 | CODE | |
| LOW | examples/grpo/cosyvoice2/token2wav_asr_server.py | 25 | CODE | |
| LOW | examples/grpo/cosyvoice2/token2wav_asr_server.py | 27 | CODE | |
| LOW | examples/grpo/cosyvoice2/token2wav_asr_server.py | 33 | CODE | |
| LOW | examples/grpo/cosyvoice2/reward_tts.py | 19 | CODE | |
| LOW | examples/grpo/cosyvoice2/reward_tts.py | 119 | CODE | |
| LOW | …amples/grpo/cosyvoice2/scripts/offline-decode-files.py | 87 | CODE |
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| LOW | cosyvoice/dataset/processor.py | 51 | except Exception as ex: | CODE |
| MEDIUM | runtime/triton_trtllm/client_grpc.py | 100 | print(f"An error occurred in the stream callback: {error}") | CODE |
| MEDIUM⚡ | runtime/triton_trtllm/client_grpc.py | 576 | print(f"Error: Audio file not found for item {i}: {item['audio_filepath']}") | CODE |
| LOW⚡ | runtime/triton_trtllm/client_grpc.py | 577 | except Exception as e: | CODE |
| MEDIUM⚡ | runtime/triton_trtllm/client_grpc.py | 578 | print(f"Error processing item {i} ({item['target_audio_path']}): {e}") | CODE |
| LOW⚡ | runtime/triton_trtllm/client_grpc.py | 588 | except Exception as e: | CODE |
| LOW | runtime/triton_trtllm/client_grpc.py | 742 | except Exception as e: | CODE |
| LOW⚡ | runtime/triton_trtllm/client_grpc.py | 903 | except Exception as e: | CODE |
| LOW⚡ | runtime/triton_trtllm/client_grpc.py | 910 | except Exception as e: | CODE |
| MEDIUM⚡ | runtime/triton_trtllm/client_grpc.py | 911 | print(f"Error closing async stats client: {e}") | CODE |
| LOW⚡ | runtime/triton_trtllm/client_grpc.py | 918 | except Exception as e: | CODE |
| MEDIUM⚡ | runtime/triton_trtllm/client_grpc.py | 919 | print(f"An error occurred in main: {e}") | CODE |
| MEDIUM | runtime/triton_trtllm/client_grpc.py | 915 | def run_main(): | CODE |
| LOW | …ton_trtllm/model_repo_cosyvoice3/cosyvoice3/1/model.py | 427 | except Exception as e: | CODE |
| LOW | …ton_trtllm/model_repo_cosyvoice3/cosyvoice3/1/model.py | 483 | except Exception as e: | CODE |
| LOW | runtime/triton_trtllm/scripts/convert_checkpoint.py | 293 | except Exception as e: | CODE |
| LOW | examples/grpo/cosyvoice2/reward_tts.py | 106 | except Exception as e: | CODE |
| LOW | examples/grpo/cosyvoice2/reward_tts.py | 212 | except Exception as e: | CODE |
| MEDIUM | examples/grpo/cosyvoice2/reward_tts.py | 213 | print(f"Error computing score: {e}") | CODE |
| MEDIUM⚡ | examples/grpo/cosyvoice2/reward_tts.py | 229 | print(f"Error: File not found - {args.input}") | CODE |
| LOW⚡ | examples/grpo/cosyvoice2/reward_tts.py | 232 | except Exception as e: | CODE |
| MEDIUM⚡ | examples/grpo/cosyvoice2/reward_tts.py | 233 | print(f"Error: {e}") | CODE |
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| HIGH | cosyvoice/utils/mask.py | 18 | def subsequent_mask( size: int, device: torch.device = torch.device("cpu"), ) -> torch.Tensor: """C | STRING |
| HIGH | cosyvoice/utils/mask.py | 57 | Create mask for subsequent steps (size, size). This mask is used only in decoder which works in an auto-regressive | STRING |
| HIGH | cosyvoice/utils/mask.py | 95 | Create mask for subsequent steps (size, size) with chunk size, this is for streaming encoder Args: s | STRING |
| HIGH | cosyvoice/utils/mask.py | 133 | Create mask for subsequent steps (size, size) with chunk size, this is for streaming encoder Args: s | STRING |
| HIGH | cosyvoice/utils/mask.py | 240 | Make mask tensor containing indices of padded part. See description of make_non_pad_mask. Args: length | STRING |
| HIGH | cosyvoice/utils/common.py | 57 | Perform padding for the list of tensors. Args: xs (List): List of Tensors [(T_1, `*`), (T_2, `*`), ..., (T_ | STRING |
| HIGH | cosyvoice/flow/decoder.py | 211 | Forward pass of the UNet1DConditional model. Args: x (torch.Tensor): shape (batch_size, in_channels | STRING |
| HIGH | cosyvoice/flow/decoder.py | 406 | Forward pass of the UNet1DConditional model. Args: x (torch.Tensor): shape (batch_size, in_channels | STRING |
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| LOW | webui.py | 50 | CODE | |
| LOW | cosyvoice/llm/llm.py | 302 | CODE | |
| LOW | cosyvoice/llm/llm.py | 505 | CODE | |
| LOW | cosyvoice/llm/llm.py | 552 | CODE | |
| LOW | cosyvoice/bin/average_model.py | 43 | CODE | |
| LOW | cosyvoice/dataset/processor.py | 31 | CODE | |
| LOW | cosyvoice/utils/mask.py | 161 | CODE | |
| LOW | cosyvoice/utils/train_utils.py | 111 | CODE | |
| LOW | cosyvoice/utils/train_utils.py | 291 | CODE | |
| LOW | cosyvoice/cli/model.py | 101 | CODE | |
| LOW | cosyvoice/cli/model.py | 175 | CODE | |
| LOW | cosyvoice/flow/decoder.py | 196 | CODE | |
| LOW | runtime/python/grpc/client.py | 30 | CODE | |
| LOW | runtime/triton_trtllm/token2wav_dit.py | 175 | CODE | |
| LOW | runtime/triton_trtllm/offline_inference.py | 325 | CODE | |
| LOW | runtime/triton_trtllm/client_grpc.py | 112 | CODE | |
| LOW | runtime/triton_trtllm/client_grpc.py | 159 | CODE | |
| LOW | runtime/triton_trtllm/client_grpc.py | 412 | CODE | |
| LOW | runtime/triton_trtllm/client_grpc.py | 501 | CODE | |
| LOW | runtime/triton_trtllm/client_grpc.py | 687 | CODE | |
| LOW | runtime/triton_trtllm/token2wav_cosyvoice3.py | 193 | CODE | |
| LOW | runtime/triton_trtllm/token2wav.py | 127 | CODE | |
| LOW | runtime/triton_trtllm/infer_cosyvoice3.py | 219 | CODE | |
| LOW | runtime/triton_trtllm/model_repo/cosyvoice2/1/model.py | 315 | CODE | |
| LOW | …ime/triton_trtllm/model_repo/cosyvoice2_dit/1/model.py | 121 | CODE | |
| LOW | …ime/triton_trtllm/model_repo/cosyvoice2_dit/1/model.py | 295 | CODE | |
| LOW | …ton_trtllm/model_repo/token2wav_dit/1/token2wav_dit.py | 175 | CODE | |
| LOW | …/triton_trtllm/model_repo/speaker_embedding/1/model.py | 91 | CODE | |
| LOW | …ton_trtllm/model_repo_cosyvoice3/cosyvoice3/1/model.py | 84 | CODE | |
| LOW | …ton_trtllm/model_repo_cosyvoice3/cosyvoice3/1/model.py | 305 | CODE | |
| LOW | …llm/model_repo_cosyvoice3/speaker_embedding/1/model.py | 91 | CODE | |
| LOW | runtime/triton_trtllm/scripts/fill_template.py | 6 | CODE | |
| LOW | runtime/triton_trtllm/scripts/convert_checkpoint.py | 152 | CODE | |
| LOW | runtime/triton_trtllm/scripts/convert_checkpoint.py | 186 | CODE | |
| LOW | runtime/triton_trtllm/scripts/convert_checkpoint.py | 282 | CODE | |
| LOW | examples/grpo/cosyvoice2/infer_dataset.py | 282 | CODE | |
| LOW | …amples/grpo/cosyvoice2/scripts/offline-decode-files.py | 137 | CODE |
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| LOW | cosyvoice/transformer/decoder.py | 178 | def forward_layers_checkpointed(self, x: torch.Tensor, | CODE |
| LOW | cosyvoice/transformer/encoder.py | 173 | def forward_layers_checkpointed(self, xs: torch.Tensor, | CODE |
| LOW | cosyvoice/utils/mask.py | 89 | def subsequent_chunk_mask_deprecated( | CODE |
| LOW | cosyvoice/utils/train_utils.py | 53 | def init_dataset_and_dataloader(args, configs, gan, dpo): | CODE |
| LOW | cosyvoice/utils/train_utils.py | 72 | def check_modify_and_save_config(args, configs): | CODE |
| LOW | cosyvoice/utils/train_utils.py | 111 | def init_optimizer_and_scheduler(args, configs, model, gan): | CODE |
| LOW | cosyvoice/utils/scheduler.py | 397 | def _linear_warmup_with_cosine_annealing(max_lr, warmup_steps, step, | CODE |
| LOW | cosyvoice/utils/scheduler.py | 543 | def _get_linear_warmup_with_cosine_annealing_lr(self, step): | CODE |
| LOW | cosyvoice/cli/frontend.py | 89 | def _extract_text_token_generator(self, text_generator): | CODE |
| LOW | runtime/triton_trtllm/token2wav_dit.py | 237 | def get_trt_kwargs_dynamic_batch(self, opt_batch_size=2, max_batch_size=64, streaming=False): | CODE |
| LOW | runtime/triton_trtllm/token2wav_dit.py | 264 | def prompt_audio_tokenization(self, prompt_audios_list: list[torch.Tensor]) -> list[list[int]]: | CODE |
| LOW | runtime/triton_trtllm/token2wav_dit.py | 371 | def get_prompt_audio_cache_for_streaming_tts( | CODE |
| LOW | runtime/triton_trtllm/offline_inference.py | 63 | async def send_batch_requests_async(api_base, model_name, chats, temperature, top_p, top_k): | CODE |
| LOW | runtime/triton_trtllm/client_grpc.py | 359 | def prepare_request_input_output( | STRING |
| LOW | runtime/triton_trtllm/client_grpc.py | 412 | def run_sync_streaming_inference( | CODE |
| LOW | runtime/triton_trtllm/token2wav_cosyvoice3.py | 219 | def prompt_audio_tokenization(self, prompt_audios_list): | CODE |
| LOW | runtime/triton_trtllm/token2wav.py | 185 | def get_trt_kwargs_dynamic_batch(self, opt_bs=2, max_batch_size=64): | CODE |
| LOW | runtime/triton_trtllm/token2wav.py | 193 | def prompt_audio_tokenization(self, prompt_audios_list: list[torch.Tensor]) -> list[list[int]]: | CODE |
| LOW | runtime/triton_trtllm/infer_cosyvoice3.py | 45 | async def send_batch_requests_async(api_base, model_name, chats, temperature, top_p, top_k): | CODE |
| LOW | runtime/triton_trtllm/model_repo/cosyvoice2/1/model.py | 201 | def forward_speaker_embedding(self, wav): | CODE |
| LOW | …ime/triton_trtllm/model_repo/cosyvoice2_dit/1/model.py | 52 | def parse_speech_token_string(response_text: str) -> List[int]: | CODE |
| LOW | …ime/triton_trtllm/model_repo/cosyvoice2_dit/1/model.py | 108 | def _convert_speech_tokens_to_str(self, speech_tokens: Union[torch.Tensor, List]) -> str: | CODE |
| LOW | …ime/triton_trtllm/model_repo/cosyvoice2_dit/1/model.py | 207 | def forward_speaker_embedding(self, wav): | CODE |
| LOW | …ton_trtllm/model_repo/token2wav_dit/1/token2wav_dit.py | 237 | def get_trt_kwargs_dynamic_batch(self, opt_batch_size=2, max_batch_size=64, streaming=False): | CODE |
| LOW | …ton_trtllm/model_repo/token2wav_dit/1/token2wav_dit.py | 264 | def prompt_audio_tokenization(self, prompt_audios_list: list[torch.Tensor]) -> list[list[int]]: | CODE |
| LOW | …ton_trtllm/model_repo/token2wav_dit/1/token2wav_dit.py | 371 | def get_prompt_audio_cache_for_streaming_tts( | CODE |
| LOW⚡ | …time/triton_trtllm/model_repo/token2wav_dit/1/model.py | 56 | def get_spk_id_from_prompt_audio(tensor: torch.Tensor) -> str: | CODE |
| LOW | …ton_trtllm/model_repo_cosyvoice3/cosyvoice3/1/model.py | 25 | def parse_speech_token_string(response_text): | CODE |
| LOW | …ton_trtllm/model_repo_cosyvoice3/cosyvoice3/1/model.py | 72 | def _convert_speech_tokens_to_str(self, speech_tokens): | CODE |
| LOW | …ton_trtllm/model_repo_cosyvoice3/cosyvoice3/1/model.py | 181 | def forward_speaker_embedding(self, wav): | CODE |
| LOW | …ton_trtllm/model_repo_cosyvoice3/cosyvoice3/1/model.py | 305 | async def _process_request_streaming(self, request): | CODE |
| LOW | runtime/triton_trtllm/scripts/convert_checkpoint.py | 186 | def update_quant_config_from_hf(quant_config, hf_config, | CODE |
| LOW | examples/grpo/cosyvoice2/token2wav_asr_server.py | 116 | def get_random_prompt_from_dataset(dataset): | CODE |
| LOW | …amples/grpo/cosyvoice2/scripts/offline-decode-files.py | 588 | def normalize_text_alimeeting(text: str) -> str: | CODE |
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| LOW | cosyvoice/dataset/dataset.py | 104 | CODE | |
| LOW | cosyvoice/hifigan/generator.py | 383 | CODE | |
| LOW | cosyvoice/hifigan/generator.py | 577 | CODE | |
| LOW | cosyvoice/transformer/decoder.py | 58 | CODE | |
| LOW | cosyvoice/transformer/decoder.py | 276 | CODE | |
| LOW | cosyvoice/transformer/convolution.py | 151 | CODE | |
| LOW | cosyvoice/transformer/upsample_encoder.py | 108 | CODE | |
| LOW | cosyvoice/transformer/encoder.py | 39 | CODE | |
| LOW | cosyvoice/transformer/encoder.py | 341 | CODE | |
| LOW | cosyvoice/transformer/encoder.py | 390 | CODE | |
| LOW | cosyvoice/utils/mask.py | 158 | CODE | |
| LOW | cosyvoice/cli/model.py | 175 | CODE | |
| LOW | cosyvoice/cli/model.py | 328 | CODE | |
| LOW | cosyvoice/flow/decoder.py | 295 | CODE | |
| LOW | cosyvoice/flow/flow.py | 26 | CODE | |
| LOW | cosyvoice/flow/flow.py | 150 | CODE | |
| LOW | cosyvoice/flow/flow.py | 285 | CODE | |
| LOW | cosyvoice/flow/DiT/modules.py | 25 | CODE | |
| LOW | cosyvoice/flow/DiT/dit.py | 105 | CODE | |
| LOW | cosyvoice/flow/DiT/dit.py | 176 | CODE | |
| LOW | runtime/triton_trtllm/client_http.py | 143 | CODE | |
| LOW | examples/grpo/cosyvoice2/prepare_data.py | 68 | CODE |
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| HIGH | runtime/triton_trtllm/offline_inference.py | 52 | except RuntimeError: | CODE |
| HIGH | runtime/triton_trtllm/infer_cosyvoice3.py | 34 | except RuntimeError: | CODE |
| HIGH | examples/grpo/cosyvoice2/infer_dataset.py | 52 | except RuntimeError: | CODE |
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| HIGH | runtime/triton_trtllm/token2wav_dit.py | 17 | python3 token2wav.py --enable-trt || exit 1 | STRING |
| HIGH | runtime/triton_trtllm/token2wav_cosyvoice3.py | 3 | python3 token2wav_cosyvoice3.py --enable-trt || exit 1 | STRING |
| HIGH | runtime/triton_trtllm/token2wav.py | 17 | python3 token2wav.py --enable-trt || exit 1 | STRING |
| HIGH | …ton_trtllm/model_repo/token2wav_dit/1/token2wav_dit.py | 17 | python3 token2wav.py --enable-trt || exit 1 | STRING |
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| LOW | cosyvoice/transformer/convolution.py | 125 | # It's better we just return None if no cache is required, | COMMENT |
| MEDIUM | cosyvoice/utils/executor.py | 45 | # A context manager to be used in conjunction with an instance of | COMMENT |
| MEDIUM | cosyvoice/utils/executor.py | 97 | # A context manager to be used in conjunction with an instance of | COMMENT |
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| MEDIUM⚡ | …time/triton_trtllm/model_repo/token2wav_dit/1/model.py | 64 | # Create a SHA-256 hash of the byte string | COMMENT |
| MEDIUM | examples/grpo/cosyvoice2/run.sh | 9 | # This function is from espnet | COMMENT |
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| LOW | runtime/triton_trtllm/token2wav_cosyvoice3.py | 23 | logger = logging.getLogger(__name__) | CODE |
| LOW | runtime/triton_trtllm/model_repo/token2wav/1/model.py | 46 | logger = logging.getLogger(__name__) | CODE |
| LOW⚡ | …time/triton_trtllm/model_repo/token2wav_dit/1/model.py | 49 | logger = logging.getLogger(__name__) | CODE |
| LOW | …iton_trtllm/model_repo_cosyvoice3/token2wav/1/model.py | 13 | logger = logging.getLogger(__name__) | CODE |
| LOW | …triton_trtllm/model_repo_cosyvoice3/vocoder/1/model.py | 11 | logger = logging.getLogger(__name__) | CODE |
| LOW | …time/triton_trtllm/scripts/convert_cosyvoice3_to_hf.py | 45 | logger = logging.getLogger(__name__) | CODE |
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| MEDIUM | example.py | 60 | # bistream usage, you can use generator as input, this is useful when using text llm model as input | COMMENT |
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| MEDIUM | runtime/triton_trtllm/token2wav_cosyvoice3.py | 120 | # Speaker embedding model (campplus) | COMMENT |