Pass dataset folder option for moss TTS v1.5? [look at moss-tts-v1.5-8b-voice-acting model]

[krisheetu]

is it possible to pass a folder path that has audio files already chunked with text instead of attaching a lot of clips?

[diodiogod]

This is now available in v5.6.4 on main.

MOSS Dataset Prep now accepts a folder path directly in dataset_source, so you do not need to attach every clip individually. Put each audio file beside a non-empty .txt transcript with the same basename, for example:

my_dataset/
├── clip001.wav
├── clip001.txt
├── clip002.flac
└── clip002.txt

Supported audio formats are WAV, FLAC, MP3, OGG, and M4A. Direct folder scanning is the default; enable recursive_folder_scan if the clips are inside subfolders. Existing JSONL manifest workflows are unchanged.

This applies to the currently supported MOSS LoRA training models: MOSS-TTS Delay 8B v1.0 and v1.5. Local 1.7B and TTSD training are not currently supported by the integrated trainer.

Please update to v5.6.4 and test the folder workflow. Let us know whether your dataset layout is recognized correctly or provide the exact layout/error if it is not.

[krisheetu]

Great! so I loaded the default workflow but did not understand, first check the image:

so the MOSS Dataset Prep has an input and MOSS-TTS Engine has no standard model V1.5 delay?

and when I hit run I am getting these errors:

[INFO] got prompt
[ERROR] Failed to validate prompt for output 147:
[ERROR] * MossDatasetPrepNode 143:
[ERROR]   - Value 0 smaller than min of 1: prep_batch_size
[ERROR]   - Value 42.0 bigger than max of 0.5: validation_split
[ERROR] Output will be ignored
[WARNING] invalid prompt: {'type': 'prompt_outputs_failed_validation', 'message': 'Prompt outputs failed validation', 'details': '', 'extra_info': {}}

I havenot changed any value, and what values should be there as I did not find that in docs.

[krisheetu]

Following this, I put the values:
prep_batch_size = 8

validation_split = 0.1 for 10%

and got this error:

Node threw an error during execution.

# ComfyUI Error Report
## Error Details
- **Node ID:** 143
- **Node Type:** MossDatasetPrepNode
- **Exception Type:** RuntimeError
- **Exception Message:** RuntimeError: MOSS training supports the Delay 8B v1.0 and v1.5 models only. Selected variant 'MOSS-TTS-Local-Transformer' is not supported yet.

## Stack Trace

File "/root/comfy/ComfyUI/execution.py", line 545, in execute
output_data, output_ui, has_subgraph, has_pending_tasks = await get_output_data(prompt_id, unique_id, obj, input_data_all, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb, v3_data=v3_data)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

File "/root/comfy/ComfyUI/execution.py", line 344, in get_output_data
return_values = await _async_map_node_over_list(prompt_id, unique_id, obj, input_data_all, obj.FUNCTION, allow_interrupt=True, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb, v3_data=v3_data)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

File "/root/comfy/ComfyUI/execution.py", line 318, in _async_map_node_over_list
await process_inputs(input_dict, i)

File "/root/comfy/ComfyUI/execution.py", line 306, in process_inputs
result = f(**inputs)
^^^^^^^^^^^

File "/root/comfy/ComfyUI/custom_nodes/TTS-Audio-Suite/nodes/training/moss_dataset_prep_node.py", line 127, in prepare_dataset
dataset = handler.prepare_dataset(
^^^^^^^^^^^^^^^^^^^^^^^^

File "/root/comfy/ComfyUI/custom_nodes/TTS-Audio-Suite/engines/moss_tts/training/handler.py", line 56, in prepare_dataset
return prepare_moss_training_dataset(self._shared_settings(tts_engine), **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

File "/root/comfy/ComfyUI/custom_nodes/TTS-Audio-Suite/engines/moss_tts/training/handler.py", line 20, in _shared_settings
resolve_delay_training_variant(config)

File "/root/comfy/ComfyUI/custom_nodes/TTS-Audio-Suite/engines/moss_tts/training/common.py", line 177, in resolve_delay_training_variant
raise RuntimeError(

## System Information
- **ComfyUI Version:** 0.30.0
- **Arguments:** /root/comfy/ComfyUI/main.py --listen 0.0.0.0 --port 8001
- **OS:** linux
- **Python Version:** 3.12.10 (main, May 21 2025, 23:34:56) [GCC 12.2.0]
- **Embedded Python:** false
- **PyTorch Version:** 2.11.0+cu128
## Devices

- **Name:** cuda:0 NVIDIA L40S : cudaMallocAsync
  - **Type:** cuda
  - **VRAM Total:** 47665709056
  - **VRAM Free:** 47210430464
  - **Torch VRAM Total:** 0
  - **Torch VRAM Free:** 0

## Logs

2026-08-03T13:30:49.482515 - �[32m[INFO]�[0m comfyui-frontend-package version: 1.47.12
2026-08-03T13:30:49.482634 - �[32m[INFO]�[0m comfyui-workflow-templates version: 0.11.27
2026-08-03T13:30:49.482746 - �[32m[INFO]�[0m comfyui-embedded-docs version: 0.5.9
2026-08-03T13:30:49.482818 - �[32m[INFO]�[0m comfy-kitchen version: 0.2.26
2026-08-03T13:30:49.482885 - �[32m[INFO]�[0m comfy-aimdo version: 0.4.11
2026-08-03T13:30:49.484276 - �[32m[INFO]�[0m [Prompt Server] web root: /usr/local/lib/python3.12/site-packages/comfyui_frontend_package/static
2026-08-03T13:30:49.485343 - �[32m[INFO]�[0m Asset seeder disabled
2026-08-03T13:30:50.841998 - �[32m[INFO]�[0m No OpenGL_accelerate module loaded: No module named 'OpenGL_accelerate'
2026-08-03T13:30:58.366993 - �[34m[ComfyUI-Easy-Use] server: �[0mv1.3.7 �[92mLoaded�[0m2026-08-03T13:30:58.367066 -
2026-08-03T13:30:58.367108 - �[34m[ComfyUI-Easy-Use] web root: �[0m/root/comfy/ComfyUI/custom_nodes/ComfyUI-Easy-Use/web_version/v2 �[92mLoaded�[0m2026-08-03T13:30:58.367133 -
2026-08-03T13:30:58.418380 - �[32m[INFO]�[0m ### Loading: ComfyUI-Manager (V3.41)
2026-08-03T13:30:58.419047 - �[32m[INFO]�[0m [ComfyUI-Manager] network_mode: public
2026-08-03T13:30:58.419213 - �[32m[INFO]�[0m [ComfyUI-Manager] ComfyUI per-queue preview override detected (PR #11261). Manager's preview method feature is disabled. Use ComfyUI's --preview-method CLI option or 'Settings > Execution > Live preview method'.
2026-08-03T13:30:58.816472 - �[32m[INFO]�[0m ### ComfyUI Version: v0.30.0-1-g14b05228 | Released on '2026-08-02'
2026-08-03T13:30:58.888824 - �[32m[INFO]�[0m [ComfyUI-Manager] default cache updated: https://raw.githubusercontent.com/ltdrdata/ComfyUI-Manager/main/alter-list.json
2026-08-03T13:30:59.025256 - �[32m[INFO]�[0m [ComfyUI-Manager] default cache updated: https://raw.githubusercontent.com/ltdrdata/ComfyUI-Manager/main/model-list.json
2026-08-03T13:30:59.070720 - �[32m[INFO]�[0m [ComfyUI-Manager] default cache updated: https://raw.githubusercontent.com/ltdrdata/ComfyUI-Manager/main/github-stats.json
2026-08-03T13:30:59.178023 - �[32m[INFO]�[0m [ComfyUI-Manager] default cache updated: https://raw.githubusercontent.com/ltdrdata/ComfyUI-Manager/main/extension-node-map.json
2026-08-03T13:30:59.301644 - �[32m[INFO]�[0m [ComfyUI-Manager] default cache updated: https://raw.githubusercontent.com/ltdrdata/ComfyUI-Manager/main/custom-node-list.json
2026-08-03T13:30:59.996573 - [FishAudioS2] 'dac' not found — auto-installing from: descript-audio-codec --no-deps
ImportError: No module named 'dac'
sys.path: ['/root/comfy/ComfyUI/custom_nodes/ComfyUI-FishAudioS2/fish_speech_src', '/root/comfy/ComfyUI/comfy', '/root/comfy/ComfyUI', '/pkg', '/root', '/usr/local/lib/python312.zip', '/usr/local/lib/python3.12', '/usr/local/lib/python3.12/lib-dynload', '/usr/local/lib/python3.12/site-packages', '/_modal/deps', '/root/comfy/ComfyUI/custom_nodes/ComfyUI-Manager/glob', '/root/comfy/ComfyUI/custom_nodes/ComfyUI-Manager/glob', '/root/comfy/ComfyUI/custom_nodes/ComfyUI-Manager/glob', '../..']
sys.modules entry:
2026-08-03T13:31:00.278036 - [FishAudioS2] Running: /usr/local/bin/python -m pip install descript-audio-codec --no-deps
2026-08-03T13:31:01.176632 - [FishAudioS2] Successfully installed: descript-audio-codec --no-deps
2026-08-03T13:31:01.185836 - [FishAudioS2] torch 2.11.0+cu128 is a CUDA build — no restore needed.
2026-08-03T13:31:01.193282 - [FishAudioS2] Models folder registered: /root/comfy/ComfyUI/models/fishaudioS2
2026-08-03T13:31:01.207975 - [FishAudioS2] Registered 4 nodes (v0.5.3): Fish S2 TTS, Fish S2 Voice Clone TTS, Fish S2 Multi-Speaker TTS, Fish S2 Multi-Speaker Split TTS
2026-08-03T13:31:01.249769 - /root/comfy/ComfyUI/custom_nodes/ComfyUI_essentials/conditioning.py:234: SyntaxWarning: invalid escape sequence '.'
double_blocks.0.(img|txt)_(mod|attn|mlp).(lin|qkv|proj|0|2).(weight|bias)=1.1
2026-08-03T13:31:02.002820 - �[32m[INFO]�[0m generated new fontManager
2026-08-03T13:31:02.286383 - �[93m[RNDNANTHU] : �[37mTHANKS FOR USING�[0m2026-08-03T13:31:02.286491 -
2026-08-03T13:31:02.286545 - �[93m[RNDNANTHU] : �[37mKINDLY GIVE YOUR SUPPORT TO �[93mRNDNANTHU�[0m2026-08-03T13:31:02.286572 -
2026-08-03T13:31:02.317140 - �[35m
⣏⡉ ⡇ ⡎⠑ ⢀⡀ ⢀⣀ ⡀⢀ ⡇⢸ ⢀⡀ ⠄ ⢀⣀ ⢀⡀ ⢉⡹
⠇ ⠧⠤ ⠣⠔ ⠣⠜ ⠭⠕ ⣑⡺ ⠸⠃ ⠣⠜ ⠇ ⠣⠤ ⠣⠭ ⠤⠜
�[0m2026-08-03T13:31:02.317209 -
2026-08-03T13:31:02.317235 - FL CosyVoice3 Custom Nodes Loaded - Version 1.2.12026-08-03T13:31:02.317256 -
2026-08-03T13:31:02.364041 - [VibeVoice] Using embedded VibeVoice (MIT licensed)
2026-08-03T13:31:04.608277 - [VibeVoice] VibeVoice nodes registered successfully
2026-08-03T13:31:04.792744 - [PromptManager] Nodes registered: Prompt Manager, Prompt Composer, Prompt Generator, Prompt Generator Kill Switch, Prompt Extractor, Recipe Extractor, Recipe Builder, Recipe Renderer, Recipe Relay, Recipe Manager2026-08-03T13:31:04.792867 -
2026-08-03T13:31:04.818523 - �[32m[INFO]�[0m [ComfyUI-DramaBox] Nodes registered: ['DramaBoxTTS', 'DramaBoxOptions', 'DramaBoxTextEncoderLoader', 'DramaBoxUnload']
2026-08-03T13:31:04.857994 - 🔧 torchaudio.save/load globally patched (scipy for WAV files, no TorchCodec)2026-08-03T13:31:04.858081 -
2026-08-03T13:31:04.858410 - 🔇 Suppressed torchaudio 2.9 migration warnings2026-08-03T13:31:04.858466 -
2026-08-03T13:31:20.858040 - 🔬 Numba compatibility setup: 16.00s2026-08-03T13:31:20.858197 -
2026-08-03T13:31:20.858255 - ✅ Numba JIT working properly (tested in 16.00s)2026-08-03T13:31:20.858300 -
2026-08-03T13:31:20.866675 - ℹ️ Critical package versions: NumPy 2.4.4, Librosa 0.11.0, Numba 0.66.0, PyTorch 2.11.0+cu128, TorchAudio 2.11.0+cu128, Transformers 4.57.6, Accelerate 1.14.0, SoundFile 0.14.02026-08-03T13:31:20.866827 -
2026-08-03T13:31:21.624396 - ❌ Vocal/Noise Removal failed: No module named 'audio_separator'2026-08-03T13:31:21.624512 -
2026-08-03T13:31:21.708411 - ======================================================================2026-08-03T13:31:21.708521 -
2026-08-03T13:31:21.708576 - 🚀 TTS Audio Suite v5.6.42026-08-03T13:31:21.708621 -
2026-08-03T13:31:21.708663 - Universal multi-engine TTS extension for ComfyUI2026-08-03T13:31:21.708732 -
2026-08-03T13:31:21.711043 - ✅ TTS Audio Suite v5.6.4 loaded with 55 nodes:2026-08-03T13:31:21.711117 -
2026-08-03T13:31:21.711203 - • ⚙️ ChatterBox Official 23-Lang Engine2026-08-03T13:31:21.711252 -
2026-08-03T13:31:21.711298 - • ⚙️ ChatterBox TTS Engine2026-08-03T13:31:21.711342 -
2026-08-03T13:31:21.711393 - • ⚙️ CosyVoice3 Engine2026-08-03T13:31:21.711443 -
2026-08-03T13:31:21.711484 - • ⚙️ Dots TTS Engine2026-08-03T13:31:21.711531 -
2026-08-03T13:31:21.711572 - • ⚙️ DramaBox Engine2026-08-03T13:31:21.711609 -
2026-08-03T13:31:21.711654 - • ⚙️ Echo-TTS Engine2026-08-03T13:31:21.711721 -
2026-08-03T13:31:21.711780 - • ⚙️ F5 TTS Engine2026-08-03T13:31:21.711819 -
2026-08-03T13:31:21.711863 - • ⚙️ Fish Audio S2 Pro Engine2026-08-03T13:31:21.711902 -
2026-08-03T13:31:21.711944 - • ⚙️ Granite ASR Engine2026-08-03T13:31:21.711980 -
2026-08-03T13:31:21.712016 - • ⚙️ Higgs Audio 2 Engine2026-08-03T13:31:21.712065 -
2026-08-03T13:31:21.712106 - • ⚙️ Higgs Audio v3 Engine2026-08-03T13:31:21.712147 -
2026-08-03T13:31:21.712187 - • ⚙️ IndexTTS-2 Engine2026-08-03T13:31:21.712226 -
2026-08-03T13:31:21.712266 - • ⚙️ MOSS SoundEffect v2 Engine2026-08-03T13:31:21.712305 -
2026-08-03T13:31:21.712344 - • ⚙️ MOSS-TTS Engine2026-08-03T13:31:21.712379 -
2026-08-03T13:31:21.712416 - • ⚙️ OmniVoice Engine2026-08-03T13:31:21.712453 -
2026-08-03T13:31:21.712491 - • ⚙️ Qwen3-TTS Engine2026-08-03T13:31:21.712533 -
2026-08-03T13:31:21.712575 - • ⚙️ RVC Engine2026-08-03T13:31:21.712617 -
2026-08-03T13:31:21.712656 - • ⚙️ Step Audio EditX Engine2026-08-03T13:31:21.712717 -
2026-08-03T13:31:21.712766 - • ⚙️ VibeVoice Engine2026-08-03T13:31:21.712809 -
2026-08-03T13:31:21.712915 - • ♻️ Refresh Voice Cache2026-08-03T13:31:21.713010 -
2026-08-03T13:31:21.713064 - • ✏️ ASR Transcribe2026-08-03T13:31:21.713116 -
2026-08-03T13:31:21.713161 - • 🌈 IndexTTS-2 Emotion Vectors2026-08-03T13:31:21.713201 -
2026-08-03T13:31:21.713241 - • 🌈 IndexTTS-2 Text Emotion2026-08-03T13:31:21.713278 -
2026-08-03T13:31:21.713316 - • 🌊 Audio Wave Analyzer2026-08-03T13:31:21.713351 -
2026-08-03T13:31:21.713386 - • 🌩️ Sound Effects2026-08-03T13:31:21.713429 -
2026-08-03T13:31:21.713468 - • 🎓 Model Training2026-08-03T13:31:21.713504 -
2026-08-03T13:31:21.713548 - • 🎙️ Voice Capture2026-08-03T13:31:21.713586 -
2026-08-03T13:31:21.713621 - • 🎛️ MOSS Training Config2026-08-03T13:31:21.713657 -
2026-08-03T13:31:21.713714 - • 🎛️ RVC Training Config2026-08-03T13:31:21.713761 -
2026-08-03T13:31:21.713810 - • 🎞️ MOSS Clip Staging2026-08-03T13:31:21.713855 -
2026-08-03T13:31:21.713901 - • 🎤 TTS Text2026-08-03T13:31:21.713939 -
2026-08-03T13:31:21.713977 - • 🎨 Step Audio EditX - Audio Editor2026-08-03T13:31:21.714011 -
2026-08-03T13:31:21.714054 - • 🎨 Voice Designer2026-08-03T13:31:21.714098 -
2026-08-03T13:31:21.714144 - • 🎭 Character Voices2026-08-03T13:31:21.714183 -
2026-08-03T13:31:21.714221 - • 🎭 Load RVC Character Model2026-08-03T13:31:21.714257 -
2026-08-03T13:31:21.714313 - • 🏷️ Multiline TTS Tag Editor2026-08-03T13:31:21.714355 -
2026-08-03T13:31:21.714391 - • 👄 F5-TTS Speech Editor2026-08-03T13:31:21.714427 -
2026-08-03T13:31:21.714474 - • 💾 Save Character Voice2026-08-03T13:31:21.714512 -
2026-08-03T13:31:21.714548 - • 📐 Visual Tag Builder2026-08-03T13:31:21.714589 -
2026-08-03T13:31:21.714625 - • 📝 ASR Punctuation / Truecase2026-08-03T13:31:21.714666 -
2026-08-03T13:31:21.714888 - • 📝 Phoneme Text Normalizer2026-08-03T13:31:21.714971 -
2026-08-03T13:31:21.715023 - • 📦 MOSS Dataset Prep2026-08-03T13:31:21.715066 -
2026-08-03T13:31:21.715106 - • 📦 RVC Dataset Prep2026-08-03T13:31:21.715145 -
2026-08-03T13:31:21.715184 - • 📺 TTS SRT2026-08-03T13:31:21.715220 -
2026-08-03T13:31:21.715258 - • 📺 Text to SRT Builder2026-08-03T13:31:21.715295 -
2026-08-03T13:31:21.715330 - • 🔄 Voice Changer2026-08-03T13:31:21.715367 -
2026-08-03T13:31:21.715405 - • 🔧 Audio Analyzer Options2026-08-03T13:31:21.715443 -
2026-08-03T13:31:21.715482 - • 🔧 F5-TTS Edit Options2026-08-03T13:31:21.715518 -
2026-08-03T13:31:21.715555 - • 🔧 RVC Pitch Extraction Options2026-08-03T13:31:21.715725 -
2026-08-03T13:31:21.715910 - • 🔧 SRT Advanced Options2026-08-03T13:31:21.715969 -
2026-08-03T13:31:21.716018 - • 🔧 Viseme Mouth Shape Options2026-08-03T13:31:21.716077 -
2026-08-03T13:31:21.716124 - • 🗣️ Silent Speech Analyzer2026-08-03T13:31:21.716167 -
2026-08-03T13:31:21.716209 - • 🤐 Voice Fixer2026-08-03T13:31:21.716248 -
2026-08-03T13:31:21.716287 - • 🥪 Merge Audio2026-08-03T13:31:21.716324 -
2026-08-03T13:31:21.716365 - • 🧾 MOSS Dataset Rows2026-08-03T13:31:21.716403 -
2026-08-03T13:31:21.716446 - ======================================================================2026-08-03T13:31:21.716484 -
2026-08-03T13:31:21.732098 - /root/comfy/ComfyUI/custom_nodes/comfyui-adaptiveprompts/py/generator.py:105: SyntaxWarning: invalid escape sequence '_'
Replace occurrences like _foo with unique placeholders.
2026-08-03T13:31:21.746146 - /root/comfy/ComfyUI/custom_nodes/comfyui-adaptiveprompts/py/prompt_generator.py:104: SyntaxWarning: invalid escape sequence '\L'
"tooltip": "Adaptive: Isolated RNG for strict prompt stability, allows re-arrangement.\Legacy: Sequential RNG which cascades changes (classic Dynamic Prompts)"
2026-08-03T13:31:21.756806 - /root/comfy/ComfyUI/custom_nodes/comfyui-adaptiveprompts/py/prompt_alias.py:17: SyntaxWarning: invalid escape sequence '('
either as literal: _(...), or escaped: _(...)
2026-08-03T13:31:21.787884 - �[32m[INFO]�[0m
Import times for custom nodes:
2026-08-03T13:31:21.788041 - �[32m[INFO]�[0m 0.0 seconds: /root/comfy/ComfyUI/custom_nodes/websocket_image_save.py
2026-08-03T13:31:21.788150 - �[32m[INFO]�[0m 0.0 seconds: /root/comfy/ComfyUI/custom_nodes/ComfyUI_ADV_CLIP_emb
2026-08-03T13:31:21.788214 - �[32m[INFO]�[0m 0.0 seconds: /root/comfy/ComfyUI/custom_nodes/ComfyUI-IndexTTS2
2026-08-03T13:31:21.788265 - �[32m[INFO]�[0m 0.0 seconds: /root/comfy/ComfyUI/custom_nodes/comfyui-moss-tts
2026-08-03T13:31:21.788314 - �[32m[INFO]�[0m 0.0 seconds: /root/comfy/ComfyUI/custom_nodes/ComfyUI-MelBandRoFormer
2026-08-03T13:31:21.788361 - �[32m[INFO]�[0m 0.0 seconds: /root/comfy/ComfyUI/custom_nodes/ComfyUI-DramaBox
2026-08-03T13:31:21.788431 - �[32m[INFO]�[0m 0.0 seconds: /root/comfy/ComfyUI/custom_nodes/ComfyUI_FL-CosyVoice3
2026-08-03T13:31:21.788477 - �[32m[INFO]�[0m 0.0 seconds: /root/comfy/ComfyUI/custom_nodes/ComfyUI_essentials
2026-08-03T13:31:21.788522 - �[32m[INFO]�[0m 0.1 seconds: /root/comfy/ComfyUI/custom_nodes/comfyui-adaptiveprompts
2026-08-03T13:31:21.788567 - �[32m[INFO]�[0m 0.2 seconds: /root/comfy/ComfyUI/custom_nodes/ComfyUI-Prompt-Manager
2026-08-03T13:31:21.788614 - �[32m[INFO]�[0m 0.5 seconds: /root/comfy/ComfyUI/custom_nodes/ComfyUI-Manager
2026-08-03T13:31:21.788659 - �[32m[INFO]�[0m 1.0 seconds: /root/comfy/ComfyUI/custom_nodes/ComfyUI-RndNanthu
2026-08-03T13:31:21.788744 - �[32m[INFO]�[0m 2.3 seconds: /root/comfy/ComfyUI/custom_nodes/VibeVoice-ComfyUI
2026-08-03T13:31:21.788805 - �[32m[INFO]�[0m 2.4 seconds: /root/comfy/ComfyUI/custom_nodes/ComfyUI-FishAudioS2
2026-08-03T13:31:21.788852 - �[32m[INFO]�[0m 5.9 seconds: /root/comfy/ComfyUI/custom_nodes/ComfyUI-Easy-Use
2026-08-03T13:31:21.788898 - �[32m[INFO]�[0m 16.9 seconds: /root/comfy/ComfyUI/custom_nodes/TTS-Audio-Suite
2026-08-03T13:31:21.788942 - �[32m[INFO]�[0m
2026-08-03T13:31:21.793323 - �[32m[INFO]�[0m Context impl SQLiteImpl.
2026-08-03T13:31:21.793640 - �[32m[INFO]�[0m Will assume non-transactional DDL.
2026-08-03T13:31:21.808525 - �[32m[INFO]�[0m Context impl SQLiteImpl.
2026-08-03T13:31:21.808645 - �[32m[INFO]�[0m Will assume non-transactional DDL.
2026-08-03T13:31:21.813086 - �[32m[INFO]�[0m Running upgrade -> 0001_assets, Initial assets schema
Revision ID: 0001_assets
Revises: None
Create Date: 2025-12-10 00:00:00
2026-08-03T13:31:21.834648 - �[32m[INFO]�[0m Running upgrade 0001_assets -> 0002_merge_to_asset_references, Merge AssetInfo and AssetCacheState into unified asset_references table.
2026-08-03T13:31:21.856426 - �[32m[INFO]�[0m Running upgrade 0002_merge_to_asset_references -> 0003_add_metadata_job_id, Add system_metadata and job_id columns to asset_references.
Change preview_id FK from assets.id to asset_references.id.
2026-08-03T13:31:21.887061 - �[32m[INFO]�[0m Running upgrade 0003_add_metadata_job_id -> 0004_drop_tag_type, Drop the vestigial tags.tag_type column.
2026-08-03T13:31:21.892428 - �[32m[INFO]�[0m Running upgrade 0004_drop_tag_type -> 0005_allow_case_sensitive_tags, Allow case-sensitive tag names.
2026-08-03T13:31:21.896568 - �[32m[INFO]�[0m Running upgrade 0005_allow_case_sensitive_tags -> 0006_add_loader_path, Add loader_path column to asset_references.
2026-08-03T13:31:21.898968 - �[32m[INFO]�[0m Database upgraded from None to 0006_add_loader_path
2026-08-03T13:31:22.121034 - �[32m[INFO]�[0m Using RAM pressure cache.
2026-08-03T13:31:22.121620 - �[32m[INFO]�[0m Starting server

2026-08-03T13:31:22.121998 - �[32m[INFO]�[0m To see the GUI go to: http://0.0.0.0:8001
2026-08-03T13:31:22.229266 - [TTS Audio Suite] 🎭 Character voices: Found 26 characters, 10 aliases2026-08-03T13:31:22.229352 -
2026-08-03T13:31:23.719639 - 📋 System Dependencies (background check):2026-08-03T13:31:23.719779 -
2026-08-03T13:31:23.719844 - ⚠️ Some TTS Audio Suite engines are not installed correctly:2026-08-03T13:31:23.719886 -
2026-08-03T13:31:23.719940 - ChatterBox TTS:2026-08-03T13:31:23.719993 -
2026-08-03T13:31:23.720038 - • s3tokenizer2026-08-03T13:31:23.720110 -
2026-08-03T13:31:23.720157 - F5-TTS:2026-08-03T13:31:23.720195 -
2026-08-03T13:31:23.720229 - • cached-path2026-08-03T13:31:23.720268 -
2026-08-03T13:31:23.720305 - Higgs Audio 2:2026-08-03T13:31:23.720346 -
2026-08-03T13:31:23.720382 - • vector-quantize-pytorch2026-08-03T13:31:23.720420 -
2026-08-03T13:31:23.720480 - RVC Voice Conversion:2026-08-03T13:31:23.720526 -
2026-08-03T13:31:23.720565 - • torchcrepe2026-08-03T13:31:23.720606 -
2026-08-03T13:31:23.720644 - 🔧 To repair the installation:2026-08-03T13:31:23.720678 -
2026-08-03T13:31:23.720750 - 1. Close ComfyUI.2026-08-03T13:31:23.720789 -
2026-08-03T13:31:23.720826 - 2. Open a terminal.2026-08-03T13:31:23.720868 -
2026-08-03T13:31:23.720908 - 3. Run this command:2026-08-03T13:31:23.720947 -
2026-08-03T13:31:23.720984 - "/usr/local/bin/python" "/root/comfy/ComfyUI/custom_nodes/TTS-Audio-Suite/install.py"2026-08-03T13:31:23.721021 -
2026-08-03T13:31:23.721067 - 4. Start ComfyUI again.2026-08-03T13:31:23.721107 -
2026-08-03T13:32:14.266039 - �[1m�[33m[WARNING]�[0m [DEPRECATION WARNING] Detected import of deprecated legacy API: /scripts/ui.js. This is likely caused by a custom node extension using outdated APIs. Please update your extensions or contact the extension author for an updated version.
2026-08-03T13:32:14.313574 - �[1m�[33m[WARNING]�[0m [DEPRECATION WARNING] Detected import of deprecated legacy API: /extensions/core/groupNode.js. This is likely caused by a custom node extension using outdated APIs. Please update your extensions or contact the extension author for an updated version.
2026-08-03T13:32:14.404349 - [Adaptive Prompts] Config Updated: {'default_rng_mode': 'Adaptive'}2026-08-03T13:32:14.404475 -
2026-08-03T13:32:14.405733 - [Adaptive Prompts] Config Updated: {'resolution_strategy': 'Scoped'}2026-08-03T13:32:14.405810 -
2026-08-03T13:32:14.406868 - [Adaptive Prompts] Config Updated: {'search_depth_limit': 80}2026-08-03T13:32:14.406961 -
2026-08-03T13:32:14.428755 - [Adaptive Prompts] Config Updated: {'hide_comments': True}2026-08-03T13:32:14.429145 -
2026-08-03T13:32:14.463723 - [Adaptive Prompts] Config Updated: {'missing_wildcard_behavior': 'Inject Warning'}2026-08-03T13:32:14.463799 -
2026-08-03T13:32:15.490878 - [VibeVoice] Found 8 VibeVoice model(s) available
2026-08-03T13:32:15.548636 - �[33mCreated /root/comfy/ComfyUI/user/default/prompt_generator_data.json from default system prompts.�[0m2026-08-03T13:32:15.548776 -
2026-08-03T13:32:20.508937 - [Adaptive Prompts] Config Updated: {'search_depth_limit': 80}2026-08-03T13:32:20.509041 -
2026-08-03T13:32:20.526095 - [Adaptive Prompts] Config Updated: {'default_rng_mode': 'Adaptive'}2026-08-03T13:32:20.526163 -
2026-08-03T13:32:20.527157 - [Adaptive Prompts] Config Updated: {'hide_comments': True}2026-08-03T13:32:20.527218 -
2026-08-03T13:32:20.528095 - [Adaptive Prompts] Config Updated: {'resolution_strategy': 'Scoped'}2026-08-03T13:32:20.528192 -
2026-08-03T13:32:20.663134 - [Adaptive Prompts] Config Updated: {'missing_wildcard_behavior': 'Inject Warning'}2026-08-03T13:32:20.663212 -
2026-08-03T13:32:20.678902 - �[1m�[33m[WARNING]�[0m [DEPRECATION WARNING] Detected import of deprecated legacy API: /scripts/ui/components/buttonGroup.js. This is likely caused by a custom node extension using outdated APIs. Please update your extensions or contact the extension author for an updated version.
2026-08-03T13:32:20.898423 - �[1m�[33m[WARNING]�[0m [DEPRECATION WARNING] Detected import of deprecated legacy API: /scripts/ui/components/button.js. This is likely caused by a custom node extension using outdated APIs. Please update your extensions or contact the extension author for an updated version.
2026-08-03T13:32:21.543417 - 🔧 Settings endpoint called2026-08-03T13:32:21.543532 -
2026-08-03T13:32:21.543612 - 🔧 Received settings: precision=auto, device=auto, vc_engine=chatterbox_23lang, cosyvoice_variant=RL2026-08-03T13:32:21.543660 -
2026-08-03T13:32:21.543746 - 🎨 Step Audio EditX inline tags: precision=auto, device=auto2026-08-03T13:32:21.543798 -
2026-08-03T13:32:21.543840 - 🔄 Voice restoration engine: chatterbox_23lang2026-08-03T13:32:21.543877 -
2026-08-03T13:33:23.818245 - FETCH ComfyRegistry Data [DONE]2026-08-03T13:33:23.818381 -
2026-08-03T13:33:24.102812 - �[32m[INFO]�[0m [ComfyUI-Manager] default cache updated: https://api.comfy.org/nodes
2026-08-03T13:33:24.127772 - FETCH DATA from: https://raw.githubusercontent.com/ltdrdata/ComfyUI-Manager/main/custom-node-list.json2026-08-03T13:33:24.127847 - 2026-08-03T13:33:24.191554 - [DONE]2026-08-03T13:33:24.191680 -
2026-08-03T13:33:24.299137 - �[32m[INFO]�[0m [ComfyUI-Manager] All startup tasks have been completed.
2026-08-03T13:34:58.584227 - �[32m[INFO]�[0m got prompt
2026-08-03T13:35:00.631275 - 🔄 MOSS-TTS: Resolved model selection '1.7B' -> 'MOSS-TTS-Local-Transformer'2026-08-03T13:35:00.631444 -
2026-08-03T13:35:00.631742 - ⚙️ MOSS-TTS: Configured MOSS-TTS-Local-Transformer (tts) on auto2026-08-03T13:35:00.631863 -
2026-08-03T13:35:00.631978 - Mode: Custom Character Switching2026-08-03T13:35:00.632030 -
2026-08-03T13:35:00.632098 - Language=Auto, temp=1.0, top_p=0.95, top_k=50, rep_penalty=1.1, max_new_tokens=40962026-08-03T13:35:00.632169 -
2026-08-03T13:35:00.632220 - Chunking override: disabled (ignores Unified chunk settings)2026-08-03T13:35:00.632261 -
2026-08-03T13:35:00.707007 - �[1m�[31m[ERROR]�[0m !!! Exception during processing !!! MOSS training supports the Delay 8B v1.0 and v1.5 models only. Selected variant 'MOSS-TTS-Local-Transformer' is not supported yet.
2026-08-03T13:35:00.708823 - �[1m�[31m[ERROR]�[0m Traceback (most recent call last):
File "/root/comfy/ComfyUI/execution.py", line 545, in execute
output_data, output_ui, has_subgraph, has_pending_tasks = await get_output_data(prompt_id, unique_id, obj, input_data_all, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb, v3_data=v3_data)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/root/comfy/ComfyUI/execution.py", line 344, in get_output_data
return_values = await _async_map_node_over_list(prompt_id, unique_id, obj, input_data_all, obj.FUNCTION, allow_interrupt=True, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb, v3_data=v3_data)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/root/comfy/ComfyUI/execution.py", line 318, in _async_map_node_over_list
await process_inputs(input_dict, i)
File "/root/comfy/ComfyUI/execution.py", line 306, in process_inputs
result = f(**inputs)
^^^^^^^^^^^
File "/root/comfy/ComfyUI/custom_nodes/TTS-Audio-Suite/nodes/training/moss_dataset_prep_node.py", line 127, in prepare_dataset
dataset = handler.prepare_dataset(
^^^^^^^^^^^^^^^^^^^^^^^^
File "/root/comfy/ComfyUI/custom_nodes/TTS-Audio-Suite/engines/moss_tts/training/handler.py", line 56, in prepare_dataset
return prepare_moss_training_dataset(self._shared_settings(tts_engine), **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/root/comfy/ComfyUI/custom_nodes/TTS-Audio-Suite/engines/moss_tts/training/handler.py", line 20, in _shared_settings
resolve_delay_training_variant(config)
File "/root/comfy/ComfyUI/custom_nodes/TTS-Audio-Suite/engines/moss_tts/training/common.py", line 177, in resolve_delay_training_variant
raise RuntimeError(
RuntimeError: MOSS training supports the Delay 8B v1.0 and v1.5 models only. Selected variant 'MOSS-TTS-Local-Transformer' is not supported yet.

2026-08-03T13:35:00.710519 - �[32m[INFO]�[0m �[32mPrompt executed in 0.08 seconds�[0m

## Attached Workflow
Please make sure that workflow does not contain any sensitive information such as API keys or passwords.

Workflow too large. Please manually upload the workflow from local file system.

## Additional Context
(Please add any additional context or steps to reproduce the error here)

[krisheetu]

so I selected the correct model and now it is working, the training is running, I will update and close the issue after that

[krisheetu]

hey, I just checked the training was not running and it was the dataset prepration only the error I am stuck on is:

MOSS dataset prep: target audio 48/52
MOSS dataset prep: target audio 52/52
[WARNING] [RANK 0] Detected kernel version 4.19.0, which is below the recommended minimum of 5.5.0; this can cause the process to hang. It is recommended to upgrade the kernel to the minimum version or higher.
[ERROR] !!! Exception during processing !!! Sizes of tensors must match except in dimension 0. Expected size 33 but got size 2 for tensor number 1 in the list.
[ERROR] Traceback (most recent call last):
  File "/root/comfy/ComfyUI/execution.py", line 545, in execute
    output_data, output_ui, has_subgraph, has_pending_tasks = await get_output_data(prompt_id, unique_id, obj, input_data_all, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb, v3_data=v3_data)
                                                              ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/root/comfy/ComfyUI/execution.py", line 344, in get_output_data
    return_values = await _async_map_node_over_list(prompt_id, unique_id, obj, input_data_all, obj.FUNCTION, allow_interrupt=True, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb, v3_data=v3_data)
                    ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/root/comfy/ComfyUI/execution.py", line 318, in _async_map_node_over_list
    await process_inputs(input_dict, i)
  File "/root/comfy/ComfyUI/execution.py", line 306, in process_inputs
    result = f(**inputs)
             ^^^^^^^^^^^
  File "/root/comfy/ComfyUI/custom_nodes/TTS-Audio-Suite/nodes/unified/model_training_node.py", line 127, in train_model
    artifacts = handler.train(
                ^^^^^^^^^^^^^^
  File "/root/comfy/ComfyUI/custom_nodes/TTS-Audio-Suite/engines/moss_tts/training/handler.py", line 71, in train
    return run_moss_training_job(
           ^^^^^^^^^^^^^^^^^^^^^^
  File "/root/comfy/ComfyUI/custom_nodes/TTS-Audio-Suite/engines/moss_tts/training/trainer.py", line 474, in run_moss_training_job
    for step, batch in enumerate(train_loader, start=1):
                       ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/local/lib/python3.12/site-packages/accelerate/data_loader.py", line 585, in __iter__
    current_batch = next(dataloader_iter)
                    ^^^^^^^^^^^^^^^^^^^^^
  File "/usr/local/lib/python3.12/site-packages/torch/utils/data/dataloader.py", line 741, in __next__
    data = self._next_data()
           ^^^^^^^^^^^^^^^^^
  File "/usr/local/lib/python3.12/site-packages/torch/utils/data/dataloader.py", line 801, in _next_data
    data = self._dataset_fetcher.fetch(index)  # may raise StopIteration
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/local/lib/python3.12/site-packages/torch/utils/data/_utils/fetch.py", line 54, in fetch
    data = [self.dataset[idx] for idx in possibly_batched_index]
            ~~~~~~~~~~~~^^^^^
  File "/root/comfy/ComfyUI/custom_nodes/TTS-Audio-Suite/engines/moss_tts/training/dataset.py", line 332, in __getitem__
    return self._pack_record(self.records[index])
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/root/comfy/ComfyUI/custom_nodes/TTS-Audio-Suite/engines/moss_tts/training/dataset.py", line 381, in _pack_record
    conversation = self.processor([[user_message, assistant_message]], mode="computing_loss", n_vq=target_n_vq)
                   ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/root/comfy/ComfyUI/custom_nodes/TTS-Audio-Suite/engines/moss_tts/impl/delay/processing_moss_tts.py", line 373, in __call__
    unified_codes = torch.cat(unified_codes)
                    ^^^^^^^^^^^^^^^^^^^^^^^^
RuntimeError: Sizes of tensors must match except in dimension 0. Expected size 33 but got size 2 for tensor number 1 in the list.

[diodiogod]

The compatibility regression is fixed in v5.6.5 on main.

The new recursive_folder_scan widget had been inserted in the middle of the existing MOSS Dataset Prep widgets. ComfyUI serializes widget values by position, so older workflows loaded every following value into the wrong field. This explains both sets of errors:

• validation_split = 42 and prep_batch_size = 0
• the later tensor mismatch (Expected size 33 but got size 2), because a shifted boolean value could make n_vq = 1 instead of using the model default

v5.6.5 restores the exact original widget order and appends recursive_folder_scan at the end, which preserves compatibility with workflows created before v5.6.4.

Please update/restart with v5.6.5 and reload the original workflow. Before running, verify these visible values once:

validation_split: 0.05 or 0.1
prep_batch_size: 8
n_vq: 0

Then retry MOSS-TTS Delay 8B v1.5 training. The folder import itself already completed successfully for all 52 clips; this fix targets the shifted training settings.

[krisheetu]

now it is working perfectly fine and the lora is trained well, thank you so much.

[diodiogod]

To be honest I never finished a lora training to test how good it is (the quality of using it with a lora). What is your opinion about it?

[krisheetu]

The LoRA performs well but requires more training, and both layers are necessary. I attempted to train a VibeVoice model, but MoSTTS v1.5 is superior, with fast and stable training.

The output is 95% accurate for my language, Hindi, which isn’t officially supported—yet this is the only model capable of handling it.

However, my training data lacks emotional expression. If possible, you could also implement the laion/moss-tts-v1.5-8b-voice-acting model for inference. I’m not familiar with the technical details, but it’s the same 8B model.

This is why I opened an issue for Dramabox training as well—hoping emotion support might become possible.