TextGenerateLTX2Prompt 报错（model files match requirement）

オープン

#520 2026年6月23日に作成コメント 0 件リアクション 0 件担当者 0 名 GitHub で見る

まだ誰も着手していません。

評価

難易度: 4/5
見積もり時間: 1〜2日
初心者へのやさしさ: 30/100

issue の種類: バグ
明瞭さ: おおむね明確
活発さ: 静か
技術スタック: python, pytorch
領域: machine-learning

調査の方向性

The traceback ends in comfy/text_encoders/llama.py at sample_token (around line 975), where torch.multinomial receives probs; the logits come from generate (around line 918) and the Gemma path runs through comfy/text_encoders/lt.py (around line 97) and comfy_extras/nodes_textgen.py. The traceback paths point to ComfyUI core, not this repo, so the fix may belong there. Start by reproducing with the same sampling settings and checking where inf or nan first appears in the logits or probabilities. Done means the node runs without the RuntimeError and the cause is recorded in the issue.

索引モデルが issue の本文から書いたものです。

説明

RuntimeError: probability tensor contains either inf, nan or element < 0

ComfyUI Error Report

Error Details

Node ID: 320:325
Node Type: TextGenerateLTX2Prompt
Exception Type: RuntimeError
Exception Message: RuntimeError: probability tensor contains either inf, nan or element < 0

Stack Trace

  File "/opt/ComfyUI-master/execution.py", line 542, 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 "/opt/ComfyUI-master/execution.py", line 341, 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 "/opt/ComfyUI-master/execution.py", line 315, in _async_map_node_over_list
    await process_inputs(input_dict, i)

  File "/opt/ComfyUI-master/execution.py", line 303, in process_inputs
    result = f(**inputs)
             ^^^^^^^^^^^

  File "/opt/ComfyUI-master/comfy_api/internal/__init__.py", line 149, in wrapped_func
    return method(locked_class, **inputs)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

  File "/opt/ComfyUI-master/comfy_api/latest/_io.py", line 1900, in EXECUTE_NORMALIZED
    to_return = cls.execute(*args, **kwargs)
                ^^^^^^^^^^^^^^^^^^^^^^^^^^^^

  File "/opt/ComfyUI-master/comfy_extras/nodes_textgen.py", line 174, in execute
    return super().execute(clip, formatted_prompt, max_length, sampling_mode, image=image, thinking=thinking, use_default_template=use_default_template, video=video, audio=audio)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

  File "/opt/ComfyUI-master/comfy_extras/nodes_textgen.py", line 63, in execute
    generated_ids = clip.generate(
                    ^^^^^^^^^^^^^^

  File "/opt/ComfyUI-master/comfy/sd.py", line 465, in generate
    return self.cond_stage_model.generate(tokens, do_sample=do_sample, max_length=max_length, temperature=temperature, top_k=top_k, top_p=top_p, min_p=min_p, repetition_penalty=repetition_penalty, seed=seed, presence_penalty=presence_penalty)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

  File "/opt/ComfyUI-master/comfy/text_encoders/lt.py", line 192, in generate
    return self.gemma3_12b.generate(tokens["gemma3_12b"], do_sample, max_length, temperature, top_k, top_p, min_p, repetition_penalty, seed, presence_penalty)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

  File "/opt/ComfyUI-master/comfy/text_encoders/lt.py", line 97, in generate
    return self.transformer.generate(embeds, do_sample, max_length, temperature, top_k, top_p, min_p, repetition_penalty, seed, stop_tokens=[106], presence_penalty=presence_penalty)  # 106 is <end_of_turn>
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

  File "/opt/ComfyUI-master/comfy/text_encoders/llama.py", line 918, in generate
    next_token = self.sample_token(logits, temperature, top_k, top_p, min_p, repetition_penalty, initial_tokens + generated_token_ids, generator, do_sample=do_sample, presence_penalty=presence_penalty)
                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

  File "/opt/ComfyUI-master/comfy/text_encoders/llama.py", line 975, in sample_token
    return torch.multinomial(probs, num_samples=1, generator=generator)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

System Information

ComfyUI Version: 0.25.0
Arguments: /opt/ComfyUI-master/main.py --enable-manager --listen 0.0.0.0 --port 8188
OS: linux
Python Version: 3.12.13 | packaged by Anaconda, Inc. | (main, Mar 19 2026, 20:20:58) [GCC 14.3.0]
Embedded Python: false
PyTorch Version: 2.11.0+cu130

Devices

Name: cuda:0 NVIDIA L20 : cudaMallocAsync
- Type: cuda
- VRAM Total: 47665709056
- VRAM Free: 47357165568
- Torch VRAM Total: 0
- Torch VRAM Free: 0
Name: cuda:1 NVIDIA L20 : cudaMallocAsync
- Type: cuda
- VRAM Total: 47665709056
- VRAM Free: 47357165568
- Torch VRAM Total: 0
- Torch VRAM Free: 0
Name: cuda:2 NVIDIA L20 : cudaMallocAsync
- Type: cuda
- VRAM Total: 47665709056
- VRAM Free: 47357165568
- Torch VRAM Total: 0
- Torch VRAM Free: 0
Name: cuda:3 NVIDIA L20 : cudaMallocAsync
- Type: cuda
- VRAM Total: 47665709056
- VRAM Free: 47357165568
- Torch VRAM Total: 0
- Torch VRAM Free: 0
Name: cuda:4 NVIDIA L20 : cudaMallocAsync
- Type: cuda
- VRAM Total: 47665709056
- VRAM Free: 47357165568
- Torch VRAM Total: 0
- Torch VRAM Free: 0
Name: cuda:5 NVIDIA L20 : cudaMallocAsync
- Type: cuda
- VRAM Total: 47665709056
- VRAM Free: 47357165568
- Torch VRAM Total: 0
- Torch VRAM Free: 0
Name: cuda:6 NVIDIA L20 : cudaMallocAsync
- Type: cuda
- VRAM Total: 47665709056
- VRAM Free: 47357165568
- Torch VRAM Total: 0
- Torch VRAM Free: 0
Name: cuda:7 NVIDIA L20 : cudaMallocAsync
- Type: cuda
- VRAM Total: 47665709056
- VRAM Free: 47357165568
- Torch VRAM Total: 0
- Torch VRAM Free: 0

Logs

2026-06-23T04:50:00.063744 - [32m[INFO][0m setup plugin alembic.autogenerate.schemas
2026-06-23T04:50:00.063815 - [32m[INFO][0m setup plugin alembic.autogenerate.tables
2026-06-23T04:50:00.063856 - [32m[INFO][0m setup plugin alembic.autogenerate.types
2026-06-23T04:50:00.063890 - [32m[INFO][0m setup plugin alembic.autogenerate.constraints
2026-06-23T04:50:00.063922 - [32m[INFO][0m setup plugin alembic.autogenerate.defaults
2026-06-23T04:50:00.063959 - [32m[INFO][0m setup plugin alembic.autogenerate.comments
2026-06-23T04:50:00.267166 - [START] Security scan2026-06-23T04:50:00.267178 - 
2026-06-23T04:50:00.293979 - [32m[INFO][0m [ComfyUI-Manager] Using uv as Python module for pip operations.
2026-06-23T04:50:00.372638 - [DONE] Security scan2026-06-23T04:50:00.372646 - 
2026-06-23T04:50:00.373289 - ** ComfyUI startup time:2026-06-23T04:50:00.373339 -  2026-06-23T04:50:00.373366 - 2026-06-23 04:50:00.3722026-06-23T04:50:00.373389 - 
2026-06-23T04:50:00.373416 - ** Platform:2026-06-23T04:50:00.373437 -  2026-06-23T04:50:00.373457 - Linux2026-06-23T04:50:00.373476 - 
2026-06-23T04:50:00.373497 - ** Python version:2026-06-23T04:50:00.373517 -  2026-06-23T04:50:00.373537 - 3.12.13 | packaged by Anaconda, Inc. | (main, Mar 19 2026, 20:20:58) [GCC 14.3.0]2026-06-23T04:50:00.373557 - 
2026-06-23T04:50:00.373577 - ** Python executable:2026-06-23T04:50:00.373596 -  2026-06-23T04:50:00.373615 - /home/test/anaconda3/envs/wan/bin/python2026-06-23T04:50:00.373633 - 
2026-06-23T04:50:00.373652 - ** ComfyUI Path:2026-06-23T04:50:00.373670 -  2026-06-23T04:50:00.373689 - /opt/ComfyUI-master2026-06-23T04:50:00.373707 - 
2026-06-23T04:50:00.373727 - ** ComfyUI Base Folder Path:2026-06-23T04:50:00.373745 -  2026-06-23T04:50:00.373763 - /opt/ComfyUI-master2026-06-23T04:50:00.373781 - 
2026-06-23T04:50:00.373800 - ** User directory:2026-06-23T04:50:00.373818 -  2026-06-23T04:50:00.373841 - /opt/ComfyUI-master/user2026-06-23T04:50:00.373860 - 
2026-06-23T04:50:00.373880 - ** ComfyUI-Manager config path:2026-06-23T04:50:00.373898 -  2026-06-23T04:50:00.373917 - /opt/ComfyUI-master/user/__manager/config.ini2026-06-23T04:50:00.373935 - 
2026-06-23T04:50:00.373959 - ** Log path:2026-06-23T04:50:00.373978 -  2026-06-23T04:50:00.373997 - /opt/ComfyUI-master/user/comfyui.log2026-06-23T04:50:00.374016 - 
2026-06-23T04:50:00.437805 - [32m[INFO][0m [PRE] ComfyUI-Manager
2026-06-23T04:50:02.946338 - [32m[INFO][0m Found comfy_kitchen backend cuda: {'available': True, 'disabled': False, 'unavailable_reason': None, 'capabilities': ['apply_rope', 'apply_rope1', 'apply_rope_split_half', 'apply_rope_split_half1', 'dequantize_nvfp4', 'dequantize_per_tensor_fp8', 'gemv_awq_w4a16', 'quantize_mxfp8', 'quantize_nvfp4', 'quantize_per_tensor_fp8', 'quantize_svdquant_w4a4', 'scaled_mm_nvfp4', 'scaled_mm_svdquant_w4a4', 'stochastic_rounding_fp8']}
2026-06-23T04:50:02.946470 - [32m[INFO][0m Found comfy_kitchen backend triton: {'available': True, 'disabled': True, 'unavailable_reason': None, 'capabilities': ['apply_rope', 'apply_rope1', 'apply_rope_split_half', 'apply_rope_split_half1', 'dequantize_nvfp4', 'dequantize_per_tensor_fp8', 'quantize_mxfp8', 'quantize_nvfp4', 'quantize_per_tensor_fp8']}
2026-06-23T04:50:02.946519 - [32m[INFO][0m Found comfy_kitchen backend eager: {'available': True, 'disabled': False, 'unavailable_reason': None, 'capabilities': ['apply_rope', 'apply_rope1', 'apply_rope_split_half', 'apply_rope_split_half1', 'dequantize_mxfp8', 'dequantize_nvfp4', 'dequantize_per_tensor_fp8', 'gemv_awq_w4a16', 'quantize_mxfp8', 'quantize_nvfp4', 'quantize_per_tensor_fp8', 'quantize_svdquant_w4a4', 'scaled_mm_mxfp8', 'scaled_mm_nvfp4', 'scaled_mm_svdquant_w4a4', 'stochastic_rounding_fp8']}
2026-06-23T04:50:02.949737 - [32m[INFO][0m Checkpoint files will always be loaded safely.
2026-06-23T04:50:03.128973 - [32m[INFO][0m Total VRAM 45458 MB, total RAM 773323 MB
2026-06-23T04:50:03.129078 - [32m[INFO][0m pytorch version: 2.11.0+cu130
2026-06-23T04:50:03.129304 - [32m[INFO][0m Set vram state to: NORMAL_VRAM
2026-06-23T04:50:03.129595 - [32m[INFO][0m Device: cuda:0 NVIDIA L20 : cudaMallocAsync
2026-06-23T04:50:03.129703 - [32m[INFO][0m Device: cuda:1 NVIDIA L20 : cudaMallocAsync
2026-06-23T04:50:03.129753 - [32m[INFO][0m Device: cuda:2 NVIDIA L20 : cudaMallocAsync
2026-06-23T04:50:03.129795 - [32m[INFO][0m Device: cuda:3 NVIDIA L20 : cudaMallocAsync
2026-06-23T04:50:03.129835 - [32m[INFO][0m Device: cuda:4 NVIDIA L20 : cudaMallocAsync
2026-06-23T04:50:03.129875 - [32m[INFO][0m Device: cuda:5 NVIDIA L20 : cudaMallocAsync
2026-06-23T04:50:03.129914 - [32m[INFO][0m Device: cuda:6 NVIDIA L20 : cudaMallocAsync
2026-06-23T04:50:03.129952 - [32m[INFO][0m Device: cuda:7 NVIDIA L20 : cudaMallocAsync
2026-06-23T04:50:03.130051 - [32m[INFO][0m Using async weight offloading with 2 streams
2026-06-23T04:50:03.130178 - [32m[INFO][0m Enabled pinned memory 695990.0
2026-06-23T04:50:03.232491 - [32m[INFO][0m Using pytorch attention
2026-06-23T04:50:04.563863 - [32m[INFO][0m DynamicVRAM support detected and enabled
2026-06-23T04:50:04.563961 - [32m[INFO][0m Python version: 3.12.13 | packaged by Anaconda, Inc. | (main, Mar 19 2026, 20:20:58) [GCC 14.3.0]
2026-06-23T04:50:04.564003 - [32m[INFO][0m ComfyUI version: 0.25.0
2026-06-23T04:50:04.564350 - [32m[INFO][0m comfy-aimdo version: 0.4.10
2026-06-23T04:50:04.564738 - [32m[INFO][0m comfy-kitchen version: 0.2.10
2026-06-23T04:50:04.566703 - [32m[INFO][0m comfyui-frontend-package version: 1.45.19
2026-06-23T04:50:04.566786 - [32m[INFO][0m comfyui-workflow-templates version: 0.10.0
2026-06-23T04:50:04.566836 - [32m[INFO][0m comfyui-embedded-docs version: 0.5.4
2026-06-23T04:50:04.566882 - [32m[INFO][0m comfy-kitchen version: 0.2.10
2026-06-23T04:50:04.566923 - [32m[INFO][0m comfy-aimdo version: 0.4.10
2026-06-23T04:50:04.567403 - [32m[INFO][0m [Prompt Server] web root: /home/test/anaconda3/envs/wan/lib/python3.12/site-packages/comfyui_frontend_package/static
2026-06-23T04:50:04.568234 - [32m[INFO][0m Asset seeder disabled
2026-06-23T04:50:04.568337 - [32m[INFO][0m [START] ComfyUI-Manager
2026-06-23T04:50:04.599777 - [ComfyUI-Manager] Using GitPython backend2026-06-23T04:50:04.599817 - 
2026-06-23T04:50:04.637258 - [32m[INFO][0m [ComfyUI-Manager] network_mode: personal_cloud
2026-06-23T04:50:04.639184 - [1m[33m[WARNING][0m [ComfyUI-Manager] The matrix sharing feature has been disabled because the `matrix-nio` dependency is not installed.
	To use this feature, please run the following command:
	/home/test/anaconda3/envs/wan/bin/python -m pip install matrix-nio

2026-06-23T04:50:05.746220 - [36;20m[/opt/ComfyUI-master/custom_nodes/comfyui_controlnet_aux] | INFO -> Using ckpts path: /opt/ComfyUI-master/custom_nodes/comfyui_controlnet_aux/ckpts[0m
2026-06-23T04:50:05.746445 - [36;20m[/opt/ComfyUI-master/custom_nodes/comfyui_controlnet_aux] | INFO -> Using symlinks: False[0m
2026-06-23T04:50:05.746597 - [36;20m[/opt/ComfyUI-master/custom_nodes/comfyui_controlnet_aux] | INFO -> Using ort providers: ['CUDAExecutionProvider', 'DirectMLExecutionProvider', 'OpenVINOExecutionProvider', 'ROCMExecutionProvider', 'CPUExecutionProvider', 'CoreMLExecutionProvider'][0m
2026-06-23T04:50:06.155280 - DWPose: Onnxruntime with acceleration providers detected2026-06-23T04:50:06.155318 - 
2026-06-23T04:50:06.160969 - [32m[INFO][0m 
Import times for custom nodes:
2026-06-23T04:50:06.161054 - [32m[INFO][0m    0.0 seconds: /opt/ComfyUI-master/custom_nodes/websocket_image_save.py
2026-06-23T04:50:06.161096 - [32m[INFO][0m    0.4 seconds: /opt/ComfyUI-master/custom_nodes/comfyui_controlnet_aux
2026-06-23T04:50:06.161130 - [32m[INFO][0m 
2026-06-23T04:50:06.163255 - [32m[INFO][0m Context impl SQLiteImpl.
2026-06-23T04:50:06.163343 - [32m[INFO][0m Will assume non-transactional DDL.
2026-06-23T04:50:06.203161 - [32m[INFO][0m Using RAM pressure cache.
2026-06-23T04:50:06.203484 - [32m[INFO][0m Starting server

2026-06-23T04:50:06.203764 - [32m[INFO][0m To see the GUI go to: http://0.0.0.0:8188
2026-06-23T04:50:11.519834 - [32m[INFO][0m [ComfyUI-Manager] The ComfyRegistry cache update is still in progress, so an outdated cache is being used.
2026-06-23T04:50:11.641838 - FETCH DATA from: /home/test/anaconda3/envs/wan/lib/python3.12/site-packages/comfyui_manager/custom-node-list.json2026-06-23T04:50:11.641872 - 2026-06-23T04:50:11.662156 -  [DONE]2026-06-23T04:50:11.662188 - 
2026-06-23T04:50:16.583948 - [32m[INFO][0m got prompt
2026-06-23T04:50:17.193804 - [32m[INFO][0m VAE load device: cuda:0, offload device: cpu, dtype: torch.float32
2026-06-23T04:50:17.506736 - [32m[INFO][0m Found quantization metadata version 1
2026-06-23T04:50:17.506874 - [32m[INFO][0m Detected mixed precision quantization
2026-06-23T04:50:17.507175 - [32m[INFO][0m Using mixed precision operations
2026-06-23T04:50:17.507242 - [32m[INFO][0m Native ops: float8_e4m3fn, float8_e5m2 , emulated ops: mxfp8, nvfp4
2026-06-23T04:50:17.909590 - [32m[INFO][0m model weight dtype torch.bfloat16, manual cast: torch.bfloat16
2026-06-23T04:50:17.927940 - [32m[INFO][0m model_type FLUX
2026-06-23T04:50:18.556054 - [32m[INFO][0m VAE load device: cuda:0, offload device: cpu, dtype: torch.bfloat16
2026-06-23T04:50:18.558129 - [1m[33m[WARNING][0m no CLIP/text encoder weights in checkpoint, the text encoder model will not be loaded.
2026-06-23T04:50:18.565508 - [32m[INFO][0m Requested to load VideoVAE
2026-06-23T04:50:18.593510 - [32m[INFO][0m Model VideoVAE prepared for dynamic VRAM loading. 1384MB Staged. 0 patches attached.
2026-06-23T04:50:19.137644 - [DEBUG] tokenizer_path = /opt/ComfyUI-master/comfy/sd1_tokenizer2026-06-23T04:50:19.137707 - 
2026-06-23T04:50:19.678410 - [32m[INFO][0m Requested to load LTXAVTEModel_
2026-06-23T04:50:19.745204 - [32m[INFO][0m loaded completely;  48679.09 MB loaded, full load: True
2026-06-23T04:50:19.754022 - [32m[INFO][0m CLIP/text encoder model load device: cpu, offload device: cpu, current: cpu, dtype: torch.float16
2026-06-23T04:50:27.342929 - [32m[INFO][0m got prompt
2026-06-23T04:50:38.113729 - [32m[INFO][0m Requested to load LTXAVTEModel_
2026-06-23T04:50:49.145460 - [32m[INFO][0m loaded completely;  48679.09 MB loaded, full load: True
2026-06-23T04:50:49.157552 - 
Generating tokens:   0%|          | 0/2048 [00:00<?, ?it/s]2026-06-23T04:51:02.128322 - 
Generating tokens:   0%|          | 0/2048 [00:12<?, ?it/s]2026-06-23T04:51:02.128369 - 
2026-06-23T04:51:02.128696 - [1m[31m[ERROR][0m !!! Exception during processing !!! probability tensor contains either `inf`, `nan` or element < 0
2026-06-23T04:51:02.131030 - [1m[31m[ERROR][0m Traceback (most recent call last):
  File "/opt/ComfyUI-master/execution.py", line 542, 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 "/opt/ComfyUI-master/execution.py", line 341, 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 "/opt/ComfyUI-master/execution.py", line 315, in _async_map_node_over_list
    await process_inputs(input_dict, i)
  File "/opt/ComfyUI-master/execution.py", line 303, in process_inputs
    result = f(**inputs)
             ^^^^^^^^^^^
  File "/opt/ComfyUI-master/comfy_api/internal/__init__.py", line 149, in wrapped_func
    return method(locked_class, **inputs)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/opt/ComfyUI-master/comfy_api/latest/_io.py", line 1900, in EXECUTE_NORMALIZED
    to_return = cls.execute(*args, **kwargs)
                ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/opt/ComfyUI-master/comfy_extras/nodes_textgen.py", line 174, in execute
    return super().execute(clip, formatted_prompt, max_length, sampling_mode, image=image, thinking=thinking, use_default_template=use_default_template, video=video, audio=audio)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/opt/ComfyUI-master/comfy_extras/nodes_textgen.py", line 63, in execute
    generated_ids = clip.generate(
                    ^^^^^^^^^^^^^^
  File "/opt/ComfyUI-master/comfy/sd.py", line 465, in generate
    return self.cond_stage_model.generate(tokens, do_sample=do_sample, max_length=max_length, temperature=temperature, top_k=top_k, top_p=top_p, min_p=min_p, repetition_penalty=repetition_penalty, seed=seed, presence_penalty=presence_penalty)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/opt/ComfyUI-master/comfy/text_encoders/lt.py", line 192, in generate
    return self.gemma3_12b.generate(tokens["gemma3_12b"], do_sample, max_length, temperature, top_k, top_p, min_p, repetition_penalty, seed, presence_penalty)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/opt/ComfyUI-master/comfy/text_encoders/lt.py", line 97, in generate
    return self.transformer.generate(embeds, do_sample, max_length, temperature, top_k, top_p, min_p, repetition_penalty, seed, stop_tokens=[106], presence_penalty=presence_penalty)  # 106 is <end_of_turn>
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/opt/ComfyUI-master/comfy/text_encoders/llama.py", line 918, in generate
    next_token = self.sample_token(logits, temperature, top_k, top_p, min_p, repetition_penalty, initial_tokens + generated_token_ids, generator, do_sample=do_sample, presence_penalty=presence_penalty)
                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/opt/ComfyUI-master/comfy/text_encoders/llama.py", line 975, in sample_token
    return torch.multinomial(probs, num_samples=1, generator=generator)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
RuntimeError: probability tensor contains either `inf`, `nan` or element < 0

2026-06-23T04:51:02.132354 - [32m[INFO][0m [32mPrompt executed in 45.53 seconds[0m
2026-06-23T04:51:02.406268 - 
Generating tokens:   0%|          | 0/2048 [00:00<?, ?it/s]2026-06-23T04:51:15.619437 - 
Generating tokens:   0%|          | 0/2048 [00:13<?, ?it/s]2026-06-23T04:51:15.619493 - 
2026-06-23T04:51:15.619861 - [1m[31m[ERROR][0m !!! Exception during processing !!! probability tensor contains either `inf`, `nan` or element < 0
2026-06-23T04:51:15.620879 - [1m[31m[ERROR][0m Traceback (most recent call last):
  File "/opt/ComfyUI-master/execution.py", line 542, 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 "/opt/ComfyUI-master/execution.py", line 341, 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 "/opt/ComfyUI-master/execution.py", line 315, in _async_map_node_over_list
    await process_inputs(input_dict, i)
  File "/opt/ComfyUI-master/execution.py", line 303, in process_inputs
    result = f(**inputs)
             ^^^^^^^^^^^
  File "/opt/ComfyUI-master/comfy_api/internal/__init__.py", line 149, in wrapped_func
    return method(locked_class, **inputs)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/opt/ComfyUI-master/comfy_api/latest/_io.py", line 1900, in EXECUTE_NORMALIZED
    to_return = cls.execute(*args, **kwargs)
                ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/opt/ComfyUI-master/comfy_extras/nodes_textgen.py", line 174, in execute
    return super().execute(clip, formatted_prompt, max_length, sampling_mode, image=image, thinking=thinking, use_default_template=use_default_template, video=video, audio=audio)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/opt/ComfyUI-master/comfy_extras/nodes_textgen.py", line 63, in execute
    generated_ids = clip.generate(
                    ^^^^^^^^^^^^^^
  File "/opt/ComfyUI-master/comfy/sd.py", line 465, in generate
    return self.cond_stage_model.generate(tokens, do_sample=do_sample, max_length=max_length, temperature=temperature, top_k=top_k, top_p=top_p, min_p=min_p, repetition_penalty=repetition_penalty, seed=seed, presence_penalty=presence_penalty)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/opt/ComfyUI-master/comfy/text_encoders/lt.py", line 192, in generate
    return self.gemma3_12b.generate(tokens["gemma3_12b"], do_sample, max_length, temperature, top_k, top_p, min_p, repetition_penalty, seed, presence_penalty)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/opt/ComfyUI-master/comfy/text_encoders/lt.py", line 97, in generate
    return self.transformer.generate(embeds, do_sample, max_length, temperature, top_k, top_p, min_p, repetition_penalty, seed, stop_tokens=[106], presence_penalty=presence_penalty)  # 106 is <end_of_turn>
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/opt/ComfyUI-master/comfy/text_encoders/llama.py", line 918, in generate
    next_token = self.sample_token(logits, temperature, top_k, top_p, min_p, repetition_penalty, initial_tokens + generated_token_ids, generator, do_sample=do_sample, presence_penalty=presence_penalty)
                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/opt/ComfyUI-master/comfy/text_encoders/llama.py", line 975, in sample_token
    return torch.multinomial(probs, num_samples=1, generator=generator)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
RuntimeError: probability tensor contains either `inf`, `nan` or element < 0

2026-06-23T04:51:15.622246 - [32m[INFO][0m [32mPrompt executed in 13.24 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.

主要言語: Python
スター: 4.2k
フォーク: 508
平均マージ: 7時間 14分
マージ済み PR（30日）: 3

環境構築

このプロジェクトには開発コンテナ、Dockerfile、コントリビューションガイドがありません。まず README を読み、一般的な手順ははじめてのコントリビューションガイドを参照してください。

はじめの一歩

issue を最後まで読み、次にプロジェクトのコントリビューションガイドを読みます。
着手することを issue にコメントします — 二人が同じ作業をするのを防げます。
リポジトリをフォークし、ブランチを切って変更します。
issue 番号を参照したプルリクエストを送ります。

Lightricks/ComfyUI-LTXVideo のほかの issue

LTXVImgToVideoConditionOnly modifies its input latent in place, so cached runs can condition on a previous prompt's image
オープン
- 何か問題がありますか？
難易度 2/5 1〜3時間初心者へのやさしさ 74/100

Lightricks/ComfyUI-LTXVideo#569 · コメント 1 件 · 2026年9月27日に作成
Fix ImportError with newer Kornia versions (missing pad import)
オープン
- 何か問題がありますか？
難易度 1/5 1時間未満初心者へのやさしさ 72/100

Lightricks/ComfyUI-LTXVideo#541 · コメント 3 件 · リアクション 3 件 · 2026年8月1日に作成
Fix ImportError: cannot import name 'pad' from kornia ≥ 0.8.3
オープン
- 何か問題がありますか？
難易度 1/5 1時間未満初心者へのやさしさ 75/100

Lightricks/ComfyUI-LTXVideo#525 · コメント 6 件 · リアクション 3 件 · 2026年7月8日に作成
001
オープン
- 何か問題がありますか？
難易度 5/5 1週間以上初心者へのやさしさ 15/100

Lightricks/ComfyUI-LTXVideo#589 · 2026年11月11日に作成
001
オープン
- 何か問題がありますか？
難易度 5/5 1週間以上初心者へのやさしさ 15/100

Lightricks/ComfyUI-LTXVideo#588 · 2026年11月11日に作成

Lightricks/ComfyUI-LTXVideo の issue をすべて見る

似ている issue

[BUG] Elevated Group Discovery with PowerView: the test sample windows-xml.log is declared as Sysmon but holds Security events
オープン
- 何か問題がありますか？
bug

難易度 1/5 1時間未満初心者へのやさしさ 78/100

splunk/security_content#4418 · 2026年11月14日に作成

メンテナーはふだん 1 日以内に返信
[Bug / Enhancement]: Improve teleop_robot CLI with configurable timeout, robust input validation, and clean ROS shutdown
対応中かも @muskan-creates352 が今日担当しました。オープン
- 何か問題がありますか？
難易度 2/5 1〜3時間初心者へのやさしさ 72/100

open-rmf/rmf_demos#382 · 2026年11月14日に作成

メンテナーはふだん 3 日以内に返信
bug(pipelines): shortcode stripping deletes text between a {{< >}} shortcode and the next {{% %}}
対応中かも @Ashborne-lab が今日担当しました。オープン
- 何か問題がありますか？
難易度 2/5 1〜3時間初心者へのやさしさ 70/100

kubeflow/docs-agent#293 · 2026年11月14日に作成

メンテナーはふだん 3 日以内に返信
citation_compliance_agent: some checks still assume APA 7th for every citation style
オープン
- 何か問題がありますか？
難易度 2/5 1〜3時間初心者へのやさしさ 62/100

Imbad0202/academic-research-skills#1093 · 2026年11月14日に作成

メンテナーはふだん 1 日以内に返信
Input: placeholder drawn in the text colour when the empty Input is focused but not yet edited
オープン
- 何か問題がありますか？
難易度 2/5 1〜3時間初心者へのやさしさ 78/100

pierreaubert/gpui-toolkit#114 · 2026年11月14日に作成

メンテナーはふだん 1 日以内に返信

Python の issue をもっと見る

TextGenerateLTX2Prompt 报错 （model files match requirement）

何か問題がありますか？

評価

調査の方向性

説明

ComfyUI Error Report

Error Details

Stack Trace

System Information

Devices

Logs

Attached Workflow

環境構築

はじめの一歩

Lightricks/ComfyUI-LTXVideo のほかの issue

何か問題がありますか？

何か問題がありますか？

何か問題がありますか？

何か問題がありますか？

何か問題がありますか？

似ている issue

何か問題がありますか？

何か問題がありますか？

何か問題がありますか？

何か問題がありますか？

何か問題がありますか？

新しい issue をメールで受け取る

TextGenerateLTX2Prompt 报错（model files match requirement）