nodes/NeuroSenko-ComfyUI_LLM_SDXL_Adapter
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
37 Commits 2 Branches 0 Tags
develop
Commit Graph
37 Commits
This Branch
This Branch
All Branches
Author SHA1 Message Date
NeuroSenko b6c757d45c Try to fix T5GEMMALoaderS node 2025-11-11 01:27:52 +03:00
NeuroSenko c1a9902dd1 Remove hardcoded dimensions and crop parameters from T5GEMMATextEncoderV2 2025-11-10 22:55:04 +03:00
NeuroSenko a05d090994 Add T5GEMMALoaderS for loading T5Gemma models from safetensors 2025-11-08 04:00:26 +03:00
NeuroSenko 76cdfe0706 Remove initialization print statements from __init__.py for cleaner startup output 2025-11-08 02:54:58 +03:00
NeuroSenko 803b7e30c8 Add T5GEMMATextEncoderV2 node for simplified text encoding and adapter application 2025-11-08 02:54:46 +03:00
NeuroSenko 77c1e6cd6a Update version to 3.0.1 2025-10-24 21:40:19 +03:00
NeuroSenko 56753b3e05 Append end-of-sequence token "<eos>" to text input in T5GEMMATextEncoder for improved tokenization 2025-10-24 21:39:44 +03:00
NeuroSenko 21e6d7fafc Update version to 3.0.0 2025-10-22 23:21:56 +03:00
HaiderRe 24b286e8e0 Fixed t5gemma_text_encoder and added t5gemma_apply_llm_to_sdxl_adapter 
t5gemma_text_encoder.py:
- Added custom type LLM_ATTENTION_MASK as a return type
- Changed encode_text function to not use a chat template since T5Gemma doesn't use one.
- Applied attention_mask to inputs of encode_text
t5gemma_apply_llm_to_sdxl_adapter.py:
- Added separate apply_llm_to_sdxl_adapter node for t5gemma since it uses an attention mask.
- Also allows for comparison of old code vs new code.
__init__.py:
- Added mandatory setup code for new node.
- Added custom type  "LLM_ATTENTION_MASK".
 2025-10-22 23:18:21 +03:00
NeuroSenko ccc40b8ff7 Update version to 2.0.0 2025-08-16 23:39:30 +03:00
NeuroSenko 6b105907f8 Added presets for adapters types. Added LLMAdapterLoaderCustom for custom adapters configs 2025-08-16 22:59:51 +03:00
NeuroSenko 239f8569e3 Added t5gemma support 2025-08-16 22:29:06 +03:00
NeuroSenko 6873299793 Update README.md to remove installation requirements and add requirements.txt for dependency management 2025-08-16 21:18:58 +03:00
NeuroSenko 425ce4aed8 Added LLMGGUFModelLoader 2025-08-16 21:14:53 +03:00
NeuroSenko 85fe64fc45 Update version to 1.2.2 2025-07-19 05:42:27 +03:00
NeuroSenko e42173e2ec Modify README.md to reflect changes in logging configuration 2025-07-19 05:40:02 +03:00
NeuroSenko 89402c5371 Refactor logging setup to use a consistent logger name "LLM-SDXL-Adapter" across multiple modules and update logging level in __init__.py 2025-07-19 05:39:51 +03:00
NeuroSenko 11e19f339b remove redundant info from __init__.py docstring 2025-07-19 05:24:42 +03:00
NeuroSenko 33f513f4ac Update version to 1.2.1 2025-07-18 11:56:58 +03:00
NeuroSenko 644aa4d482 Update README.md 2025-07-18 11:52:00 +03:00
draconicdragon c5624b9853 move pooled_output tensors also to cpu 
to fix ConditioningAverage
 2025-07-18 11:30:52 +03:00
draconicdragon 116a897a7b fix: remove IS_CHANGED classmethods 2025-07-18 11:30:52 +03:00
NeuroSenko 4c751ebb80 Update version to 1.2.0 2025-07-17 16:19:32 +03:00
NeuroSenko 1c48091045 Support both 'llm'/'LLM' directories; Update README to recommend lowercase naming 2025-07-17 16:15:11 +03:00
LupusKT 424f6f2de7 Add compatability with ComfyUI extra_model_paths (#2) 
add ability to load models from paths specified in extra_model_paths.yaml

llm and llm_adapters keys for dictionary

original load functionality is preserved
 2025-07-17 16:00:04 +03:00
Drac 60cc6c5ea8 fix: move conditioning to cpu for comfyui conditioning methods (#3) 
such as "Conditioning (Concat)" node
 2025-07-17 14:55:21 +03:00
NeuroSenko d9f0ac957f Add license information to pyproject.toml 2025-07-17 14:20:45 +03:00
NeuroSenko 7a2a683ff0 Create LICENSE 2025-07-17 14:18:52 +03:00
NeuroSenko 9cd5d80a17 added pyproject.toml and GitHub Actions workflow for publishing to Comfy registry 2025-07-17 14:16:07 +03:00
NeuroSenko 6eafc58baa Update README.md 2025-07-17 04:52:41 +03:00
NeuroSenko 9197ec2d1e Update README.md 2025-07-17 04:51:27 +03:00
NeuroSenko c305a665a0 added info about rouwei-gemma adapter 2025-07-17 04:47:39 +03:00
NeuroSenko ec74d9214f fix llm/adapter loading path issues 2025-07-17 04:47:06 +03:00
NeuroSenko 2f7e8d6eb2 Update README.md 2025-07-09 15:34:34 +03:00
NeuroSenko a937d7a6d6 Update README.md 2025-07-09 05:27:37 +03:00
NeuroSenko 432e448110 fix repo path 2025-07-09 05:22:08 +03:00
NeuroSenko 3fdb5dc3b8 init 2025-07-09 05:20:56 +03:00