 arcum42
|
b49f765196
|
Add in nodes for switching text on and off, text substitution, and selecting one line of a text box. Make text weight have a spot for what separator is used.
|
2025-07-20 16:45:17 -07:00 |
|
|
arcum42
|
3adcfb28f8
|
Add a custom node to choose a random line, and an alternate image crop node.
|
2025-07-19 21:13:31 -07:00 |
|
|
arcum42
|
1fb8c589a0
|
Add some initial training nodes. Work in progress, probably shouldn't currently be used. Rework input dir parsing for Load Image node.
|
2025-07-14 19:06:54 -07:00 |
|
|
arcum42
|
3886dd22ce
|
Add in a variant of the lora stack wo/ clip weight (set to 1.0). Add x3/6/9 variants, and x6 variant of the main version.
|
2025-07-07 16:43:56 -07:00 |
|
|
arcum42
|
1f2e989c44
|
Remove add checkpoint from model info, as there's a new equivalent node, and it hasn't existed long enough to have been used.
|
2025-07-06 00:58:30 -07:00 |
|
|
arcum42
|
6e23875d02
|
Redo lora load nodes to use expansion.
|
2025-07-06 00:53:38 -07:00 |
|
|
arcum42
|
68472c2d64
|
Rework the triple lora loader internally to use graphbuilder, expanding out to multiple nodes. Remove unused imports.
|
2025-07-05 16:55:30 -07:00 |
|
|
arcum42
|
2db191d8f0
|
Add Reference Image node, which is a combination of a Reference Latent and a VAE Encode node.
|
2025-07-02 16:34:18 -07:00 |
|
|
arcum42
|
dc3eeb14a4
|
Added in individual nodes for loading each info selector, partially for easier troubleshooting. Some reorg.
|
2025-06-29 19:10:16 -07:00 |
|
|
arcum42
|
f75201f79b
|
Lots of work on radical changes to model_info, and getting unet, clip, and vae working in my nodes. Not ready for testing. Currently errors out in tests at the KSampler, and probably doesn't work right with metadata nodes or nodes for pulling model info.
|
2025-06-29 17:55:45 -07:00 |
|
|
arcum42
|
3bb88839a1
|
Loader that just gives a model, clip, and vae from model_info.
|
2025-06-28 14:34:35 -07:00 |
|
|
arcum42
|
8a5ff68c5d
|
Add halt and free memory nodes.
|
2025-06-28 11:55:25 -07:00 |
|
|
arcum42
|
bb3e0758fa
|
Added routes to view the cache via api. Started adding notes code, currently disabled.
|
2025-06-22 17:30:34 -07:00 |
|
|
arcum42
|
f58b769264
|
Import rework.
|
2025-06-22 01:47:52 -07:00 |
|
|
arcum42
|
eaf5e15ec2
|
Add node to save a string as a text file.
|
2025-06-20 17:21:58 -07:00 |
|
|
arcum42
|
339658cfc4
|
Add a version of the Set Text node that supports dynamic prompts, reading any wildcards in a new wildcards directory, with a built in clean option.
|
2025-06-20 16:04:42 -07:00 |
|
|
arcum42
|
8a6b3b86e5
|
Add in a modified version of the Image Resize node from ComfyUI Essentials.
|
2025-06-20 15:02:52 -07:00 |
|
|
arcum42
|
dc59a4686c
|
Add a node with common resolutions, and one to pick a common width and height based on the aspect ratio of the width and height give to it (largely for scaling images or making a picture with a similar ratio to an existing one).
|
2025-06-20 13:59:15 -07:00 |
|
|
arcum42
|
fd973f34c9
|
Cache if ollama models are vision models or not.
|
2025-06-19 20:26:52 -07:00 |
|
|
arcum42
|
d8ea03380d
|
Add a quick node to pick resolutions by aspect ratio. Everyone's got one of these.
|
2025-06-19 14:03:15 -07:00 |
|
|
arcum42
|
7b9410fa5a
|
Add 'refine' LLM nodes, which will act like a vision node, then feed its response with a separate prompt into another model, and return both responses. Intended to have one model enhance another models description.
|
2025-06-17 22:30:00 -07:00 |
|
|
arcum42
|
a75a443a24
|
Working on the LLM code and the ability to call them with custom host urls.
|
2025-06-12 18:02:13 -07:00 |
|
|
arcum42
|
1651057e57
|
Remove most of the depreciated nodes.
|
2025-06-11 22:30:40 -07:00 |
|
|
arcum42
|
5a505137ee
|
Enable LM Studio support. Revise workflow.
|
2025-06-05 18:27:11 -07:00 |
|
|
arcum42
|
10d62afae2
|
Add extra prompt node.
|
2025-06-04 16:59:14 -07:00 |
|
|
arcum42
|
89cd079834
|
Ollama text and vision nodes are working reasonably. Major changes to the llm_prompt.json file.
|
2025-06-04 15:38:32 -07:00 |
|
|
arcum42
|
6049212d84
|
Work on LLMs. Disabling lm studio support because it doesn't work.
|
2025-06-04 12:33:30 -07:00 |
|
|
arcum42
|
e31183f043
|
Ignore errors and set encoding when loading json. Experimental prompt node that is not ready to use and will change substantially.
|
2025-06-03 14:52:04 -07:00 |
|
|
arcum42
|
bb719962c4
|
Generalize the code for configuration and asset files.
|
2025-06-02 19:44:55 -07:00 |
|
|
arcum42
|
7249f16dee
|
Initial work on ollama support. Currently disabled.
|
2025-06-02 18:38:06 -07:00 |
|
|
arcum42
|
9d1f368000
|
Reorg.
|
2025-05-31 11:21:58 -07:00 |
|
|
arcum42
|
8f81b8534c
|
More cleanup.
|
2025-05-27 00:53:30 -07:00 |
|
|
arcum42
|
67ba20d8d8
|
Delete unused test node.
|
2025-05-27 00:48:21 -07:00 |
|
|
arcum42
|
01a339a3cc
|
Rework mappings and do some shuffling around.
|
2025-05-27 00:44:00 -07:00 |
|
|
arcum42
|
dd36668b59
|
Add a node to check lora stack for updates. Fix logic. Recheck file if it's been modified since it was last checked.
|
2025-05-18 18:29:29 -07:00 |
|
|
arcum42
|
62574dd111
|
Add Audio VAE to KSampler. Add Model Shifts options to Model + Lora.
|
2025-05-11 19:08:19 -07:00 |
|
|
arcum42
|
ada34c266f
|
Add HiDream E1 node.
|
2025-04-30 14:55:47 -07:00 |
|
|
arcum42
|
3be919c5d9
|
Rename a few nodes. Add a node that lets you connect as many models as you want, then shoose one by number. Will currently let you choose numbers higher then the number of models.
|
2025-04-03 01:45:35 -07:00 |
|
|
arcum42
|
41f234ac79
|
Add a Text Weight node and a Pony Style node. Add an option to Join Text to put the separator at the end as well. Remove unused option on Pony Prefixes..
|
2025-04-01 11:58:58 -07:00 |
|
|
arcum42
|
e5ccb278b8
|
CheckpointInfo -> Checkpoint Selector, to be a little less confusing. Fix TripleLoraLoader.
|
2025-03-29 18:12:45 -07:00 |
|
|
arcum42
|
4ff5a9b07e
|
Simplifying names for things and refactoring.
|
2025-03-22 20:57:47 -07:00 |
|
|
arcum42
|
c410d5f3b5
|
Rework structure a bit and clean things up.
|
2025-03-22 19:48:38 -07:00 |
|
|
arcum42
|
7e21ff9770
|
Move a few things around. Add a file for depreciated nodes, and get rid of a no longer needed one.
|
2025-03-21 03:26:51 -07:00 |
|
|
arcum42
|
9399d5a8bb
|
Changing the new version of View Text to be View Anything As Text, and including an old version of View Text, because otherwise it badly breaks any workflows with the old version.
|
2025-03-20 17:20:58 -07:00 |
|
|
arcum42
|
9e4e303749
|
Add a Model + Lora Stack Loader, which takes in a model_info input and a lora_stack, loads the checkpoint, and applies the loras. Don't pull the model_info from a load checkpoint node, or you'll load it twice.
|
2025-03-19 02:29:29 -07:00 |
|
|
arcum42
|
c0adf62132
|
View Text is now somewhat of a misnomer, as it'll print *anything* hooked to it in the output as text.
|
2025-03-19 00:40:44 -07:00 |
|
|
arcum42
|
55e3a8811f
|
Add a text compare node, and some reformatting.
|
2025-03-16 12:35:02 -07:00 |
|
|
arcum42
|
b5853c9baf
|
Rework the triple stack lora code to be more flexible. Add a Checkpoint Info node that *only* chooses a checkpoint and gives model info without loading it.
|
2025-03-16 01:38:51 -07:00 |
|
|
arcum42
|
44c546cbaf
|
Add in a KSampler + Decoder node, KSampler + Tiled Decoder, and Tiling Info nodes. Combines the VAE Decode node into the KSampler, as well as the tiled version.
|
2025-03-12 00:28:55 -07:00 |
|
|
arcum42
|
9d463816d5
|
Add a triple lora loader. Add a test node, because I want to remember the code for future use.
|
2025-03-08 12:50:47 -08:00 |
|