[GH-ISSUE #329] TagComplete Does Not Show a Dropdown. #231

New issue

Closed

opened 2026-10-30 21:21:20 +00:00 by bsakaguchi · 2 comments

bsakaguchi commented

2026-10-30 21:21:20 +00:00

Owner

Originally created by @DeeJayMhyst on GitHub (Aug 25, 2025).
Original GitHub issue: https://github.com/DominikDoom/a1111-sd-webui-tagcomplete/issues/329

I'm not getting a dropdown when starting a LoRA or wildcard. I just get a blank line. And if I arrow down on my keyboard, it will populate a LoRA, but I have no idea which one it will choose. If I hit enter, it will populate whatever was supposed to be shown if it was working properly. It is as if it still can find LoRA and Wildcards, but refuses to show them in the dropdown window. I included my log if that helps.

Here's a couple of notes:

<edit: removed as irrelevant>

I don't have the openpose-editor extension installed.

I isolated the tagcomplete extension by disabling all other extensions (built-in and user) and still had this issue.

Here is my console log:

Python 3.10.11 (tags/v3.10.11:7d4cc5a, Apr 5 2023, 00:38:17) [MSC v.1929 64 bit (AMD64)]
Version: f2.0.1v1.10.1-previous-669-gdfdcbab6
Commit hash: dfdcbab685e57677014f05a3309b48cc87383167
D:\StabilityMatrix-win-x64\Data\Packages\Stable Diffusion WebUI Forge\extensions-builtin\forge_legacy_preprocessors\install.py:2: DeprecationWarning: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html
import pkg_resources
D:\StabilityMatrix-win-x64\Data\Packages\Stable Diffusion WebUI Forge\extensions-builtin\sd_forge_controlnet\install.py:2: DeprecationWarning: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html
import pkg_resources
Launching Web UI with arguments: --cuda-malloc --gradio-allowed-path 'D:\StabilityMatrix-win-x64\Data\Images'
Using cudaMallocAsync backend.
Total VRAM 16303 MB, total RAM 32144 MB
pytorch version: 2.8.0+cu128
Set vram state to: NORMAL_VRAM
Device: cuda:0 NVIDIA GeForce RTX 5080 Laptop GPU : cudaMallocAsync
VAE dtype preferences: [torch.bfloat16, torch.float32] -> torch.bfloat16
CUDA Using Stream: False
W0825 09:12:08.989000 27184 venv\Lib\site-packages\torch\distributed\elastic\multiprocessing\redirects.py:29] NOTE: Redirects are currently not supported in Windows or MacOs.
Using pytorch cross attention
Using pytorch attention for VAE
ControlNet preprocessor location: D:\StabilityMatrix-win-x64\Data\Packages\Stable Diffusion WebUI Forge\models\ControlNetPreprocessor
Tag Autocomplete: Could not locate model-keyword extension, Lora trigger word completion will be limited to those added through the extra networks menu.
[-] ADetailer initialized. version: 25.3.0, num models: 10
2025-08-25 09:12:14,204 - ControlNet - INFO - ControlNet UI callback registered.
Model selected: {'checkpoint_info': {'filename': 'D:\StabilityMatrix-win-x64\Data\Packages\Stable Diffusion WebUI Forge\models\Stable-diffusion\sd\cyberillustrious_v60.safetensors', 'hash': 'bc0255a5'}, 'additional_modules': [], 'unet_storage_dtype': None}
Using online LoRAs in FP16: False

<edit: added information for clarity>

Originally created by @DeeJayMhyst on GitHub (Aug 25, 2025). Original GitHub issue: https://github.com/DominikDoom/a1111-sd-webui-tagcomplete/issues/329 I'm not getting a dropdown when starting a LoRA or wildcard. I just get a blank line. And if I arrow down on my keyboard, it will populate a LoRA, but I have no idea which one it will choose. If I hit enter, it will populate whatever was supposed to be shown if it was working properly. It is as if it still can find LoRA and Wildcards, but refuses to show them in the dropdown window. I included my log if that helps. <img width="1396" height="278" alt="Image" src="https://github.com/user-attachments/assets/ae4db734-428a-4073-a7bf-eca047ba6f41" /> Here's a couple of notes: <edit: removed as irrelevant> I don't have the openpose-editor extension installed. I isolated the tagcomplete extension by disabling all other extensions (built-in and user) and still had this issue. Here is my console log: Python 3.10.11 (tags/v3.10.11:7d4cc5a, Apr 5 2023, 00:38:17) [MSC v.1929 64 bit (AMD64)] Version: f2.0.1v1.10.1-previous-669-gdfdcbab6 Commit hash: dfdcbab685e57677014f05a3309b48cc87383167 D:\StabilityMatrix-win-x64\Data\Packages\Stable Diffusion WebUI Forge\extensions-builtin\forge_legacy_preprocessors\install.py:2: DeprecationWarning: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html import pkg_resources D:\StabilityMatrix-win-x64\Data\Packages\Stable Diffusion WebUI Forge\extensions-builtin\sd_forge_controlnet\install.py:2: DeprecationWarning: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html import pkg_resources Launching Web UI with arguments: --cuda-malloc --gradio-allowed-path 'D:\StabilityMatrix-win-x64\Data\Images' Using cudaMallocAsync backend. Total VRAM 16303 MB, total RAM 32144 MB pytorch version: 2.8.0+cu128 Set vram state to: NORMAL_VRAM Device: cuda:0 NVIDIA GeForce RTX 5080 Laptop GPU : cudaMallocAsync VAE dtype preferences: [torch.bfloat16, torch.float32] -> torch.bfloat16 CUDA Using Stream: False W0825 09:12:08.989000 27184 venv\Lib\site-packages\torch\distributed\elastic\multiprocessing\redirects.py:29] NOTE: Redirects are currently not supported in Windows or MacOs. Using pytorch cross attention Using pytorch attention for VAE ControlNet preprocessor location: D:\StabilityMatrix-win-x64\Data\Packages\Stable Diffusion WebUI Forge\models\ControlNetPreprocessor Tag Autocomplete: Could not locate model-keyword extension, Lora trigger word completion will be limited to those added through the extra networks menu. [-] ADetailer initialized. version: 25.3.0, num models: 10 2025-08-25 09:12:14,204 - ControlNet - INFO - ControlNet UI callback registered. Model selected: {'checkpoint_info': {'filename': 'D:\\StabilityMatrix-win-x64\\Data\\Packages\\Stable Diffusion WebUI Forge\\models\\Stable-diffusion\\sd\\cyberillustrious_v60.safetensors', 'hash': 'bc0255a5'}, 'additional_modules': [], 'unet_storage_dtype': None} Using online LoRAs in FP16: False <edit: added information for clarity>

bsakaguchi closed this issue

2026-10-30 21:21:20 +00:00

bsakaguchi commented

2026-10-30 21:21:21 +00:00

Author

Owner

@DominikDoom commented on GitHub (Aug 26, 2025):

With results being inserted but not visible it seems like a broken file load to me, or a wrong setting.

Just to make sure to exclude the latter, have you maybe accidentally set "Maximum results" to 0? Even if the "Show all results" option below that is used, maximum results will still be used to calculate the height of the popup box (i.e. how many are visible without scrolling). Although that would affect normal results too, not just wildcards / lora.

My other guess would be that the entries have no height because the browser is unable to retrieve some info and either directly or indirectly fails to set the display text due to it. Could you open the browser console (CTRL+SHIFT+I) and tell me if any errors are logged there, either at the start or after triggering the popup? Any failed file loads should appear in the logs.

If that is the case, it might be caused by this:

Launching Web UI with arguments: --cuda-malloc --gradio-allowed-path 'D:\StabilityMatrix-win-x64\Data\Images'

--gradio-allowed-path might be too restrictive and prevent TAC from loading some temporary files it relies on or something similar.

If that guess is wrong, the browser console should still show any potential other issues TAC had.

@DominikDoom commented on GitHub (Aug 26, 2025): With results being inserted but not visible it seems like a broken file load to me, or a wrong setting. Just to make sure to exclude the latter, have you maybe accidentally set "Maximum results" to 0? Even if the "Show all results" option below that is used, maximum results will still be used to calculate the height of the popup box (i.e. how many are visible without scrolling). Although that would affect normal results too, not just wildcards / lora. My other guess would be that the entries have no height because the browser is unable to retrieve some info and either directly or indirectly fails to set the display text due to it. Could you open the browser console (CTRL+SHIFT+I) and tell me if any errors are logged there, either at the start or after triggering the popup? Any failed file loads should appear in the logs. If that is the case, it might be caused by this: > Launching Web UI with arguments: --cuda-malloc --gradio-allowed-path 'D:\StabilityMatrix-win-x64\Data\Images' --gradio-allowed-path might be too restrictive and prevent TAC from loading some temporary files it relies on or something similar. If that guess is wrong, the browser console should still show any potential other issues TAC had.

bsakaguchi commented

2026-10-30 21:21:21 +00:00

Author

Owner

@DeeJayMhyst commented on GitHub (Aug 26, 2025):

The Maximum results was blank! changed to 10 and that fixed the problem. Thank you!!

@DeeJayMhyst commented on GitHub (Aug 26, 2025): The Maximum results was blank! changed to 10 and that fixed the problem. Thank you!!