Tip
Starting from v1.3, to support auto-detection on Civitai, calculate_model_hash
will be renamed as
calculate_hash
and it will be enabled by default. Due to the addition of temporary storage of the model hash values,
the first image generated after switching to the new model will take more time to calculate the hash value,
but it will not affect the generation speed afterwards until the server is restarted.
Note
comfyui-prompt-reader-node is now published on the Comfy Registry.
- Add
Any Converter
node #76 #78 - Fix the input validation of the
Parameter Generator
node #89 - Fix the input validation of the
Parameter Extractor
node #90 - Fix the
Parameter Generator
unable to choose standard resolutions #82 - Fix the
Prompt Saver
unable to correctly handle filenames containing.
#80 - Update the
Parameter Extractor
node to support parsing quoted parameters #87 @spagatticode - Update PyProject Toml - License #85 @haohaocreates
Tip
Starting from v1.3, to support auto-detection on Civitai, calculate_model_hash
will be renamed as
calculate_hash
and it will be enabled by default. Due to the addition of temporary storage of the model hash values,
the first image generated after switching to the new model will take more time to calculate the hash value,
but it will not affect the generation speed afterwards until the server is restarted.
Note
comfyui-prompt-reader-node is now published on the Comfy Registry.
- Add pyproject.toml for Custom Node Registry #75 @haohaocreates
Starting from v1.3, to support auto-detection on Civitai,
calculate_model_hash
will be renamed ascalculate_hash
and it will be enabled by default. Due to the addition of temporary storage of the model hash values, the first image generated after switching to the new model will take more time to calculate the hash value, but it will not affect the generation speed afterwards until the server is restarted.
- Add support for pasted images to the
Prompt Reader
node #64 - Update core to 1.3.5
- Update nodes.py save_images to have %width and %height #66 @COvert96
- Update the
Prompt Reader
node to the new exception handling method
Starting from v1.3, to support auto-detection on Civitai,
calculate_model_hash
will be renamed ascalculate_hash
and it will be enabled by default. Due to the addition of temporary storage of the model hash values, the first image generated after switching to the new model will take more time to calculate the hash value, but it will not affect the generation speed afterwards until the server is restarted.
- Add user-friendly warnings and remove the exception raised during the reading of metadata #58 #63
- Add JPEG to the image extension list
- Fix the issue where the
Prompt Saver
node cannot consume filename from anotherPrompt Saver
node #62 - Fix the
Prompt Saver
node freezes workflow when name includes%model
#61 - Remove ckpt validation of the
Prompt Reader
node #59
Starting from this version, to support auto-detection on Civitai,
calculate_model_hash
will be renamed ascalculate_hash
and it will be enabled by default. Due to the addition of temporary storage of the model hash values, the first image generated after switching to the new model will take more time to calculate the hash value, but it will not affect the generation speed afterwards until the server is restarted.
- Add
Lora Loader
node andLora Selector
node - Add
VAE_NAME
output to theParameter Generator
node and addvae_name
input to thePrompt Saver
node #39 - Add
lora_name
input to thePrompt Saver
node - Add
resource_hash
switch to thePrompt Saver
node - Add resources hashes to metadata for auto-detection on Civitai #35
- Add temporary storage for model hashes
- Add
WEB_DIRECTORY
and remove old js directory - Fix adding the
Parameter Generator
node to a workflow disables the queue button #45 - Fix the
Parameter Generator
node doesn't load seed value from generated images #44 - Fix the input validation of the
Batch Loader
node #37 #38 - Fix the parsing error caused by the loss of model data #43
- Fix the error message caused by aspect_ratio
- Rename
calculate_model_hash
ascalculate_hash
- Update seedGen.js to rgthree's latest code
- Update core to cc3c8b2
- Fix the input validation of the
Prompt Reader
node #37 - Fix the
Batch Loader
node not working when not connected to any node
- Add
Parameter Extractor
node #24 - Add a model matching feature to the
Prompt Reader
node #24 - Add
save_metadata_file
option to thePrompt Saver
node #30 - Add
FILE_PATH
output to thePrompt Saver
node #26 - Fix the issue where the
Prompt Merger
node threw an error when merging empty strings #29 - Enhance the
Batch Loader
node to support processing either a single file or a list of files #26 - Update core to 1.3.4.post1
- Add
Batch Loader
node #13 - Add
MODEL_NAME
output to thePrompt Reader
node #23 - Add VAE selector to the
Parameter Generator
node #15 - Add pixel dimensions display to the
aspect_ratio
in theParameter Generator
node #6 - Add Positive and Negative Aesthetic Score to the
PARAMETERS
in theParameter Generator
node #8 - Add
FILENAME
andMETADATA
output to thePrompt Saver
node #16
- Add a new file-naming mechanism to ensure naming uniqueness
- Fix
%counter
overwriting existing images #11 #14 - Fix the bug causing the
Prompt Saver
node fails to save images in jpg and webp formats #10 - Use relative paths for js imports #12
- Initial release