-
c07e962cae
Image tools: migrate to OWUI 0.9.0 async model accessors
main
William Gill
2026-04-26 06:16:02 -05:00
-
d034700af9
Image tools: work around OWUI 0.9.x files-event regression
William Gill
2026-04-26 06:03:21 -05:00
-
28f370a80b
Update deployments/ai-stack/openwebui-tools/smart_image_gen.py
57_Wolve
2026-04-20 17:08:32 +00:00
-
02a4bece5d
Ollama: keep loaded models resident until evicted (KEEP_ALIVE=-1)
William Gill
2026-04-19 19:25:42 -05:00
-
a1af88a632
smart_image_gen v0.7.8: per-job filename_prefix + source-image diagnostic
William Gill
2026-04-19 18:22:45 -05:00
-
d8c8421361
Image Studio: lock in working config — Native + enable_thinking=false
William Gill
2026-04-19 18:12:24 -05:00
-
f5a222fe6f
Image Studio: default base model → huihui_ai/qwen3-vl-abliterated:8b
William Gill
2026-04-19 17:56:16 -05:00
-
ec6108888a
smart_image_gen v0.7.7: enforce style inheritance for edit_image
William Gill
2026-04-19 17:53:32 -05:00
-
20d4bd5b72
Image Studio: switch base model to qwen3.5:9b (non-abliterated)
William Gill
2026-04-19 17:09:42 -05:00
-
1ae451ad5f
Add smart_image_pipe.py — deterministic Pipe for image gen / edit / inpaint
William Gill
2026-04-19 16:58:01 -05:00
-
011fade024
Image Studio prompt: forbid post-tool echo of the function call
William Gill
2026-04-19 16:41:51 -05:00
-
63917709c1
smart_image_gen v0.7.6: feather inpaint mask edges via GrowMask
William Gill
2026-04-19 16:40:39 -05:00
-
1ed2e7293e
Image Studio: ship function_calling=default — Native leaks Qwen 3.5 XML
William Gill
2026-04-19 16:36:05 -05:00
-
18a205d69d
smart_image_gen v0.7.5: fix black-image bug — fetch from SaveImage explicitly
William Gill
2026-04-19 16:30:04 -05:00
-
0fa8040251
Eliminate first-inpaint timeout: preseed SAM/GroundingDINO + 600s default
William Gill
2026-04-19 15:39:23 -05:00
-
e77666ea0f
Image Studio docs: require setting a separate Task Model after install
William Gill
2026-04-19 15:37:21 -05:00
-
6700f6ce33
smart_image_gen v0.7.3: edit_image inherits style from prior tool call
William Gill
2026-04-19 15:31:57 -05:00
-
def27087c1
Make every image tag in compose pinnable via .env
William Gill
2026-04-19 15:27:08 -05:00
-
2cecf77981
Pin transformers <5 — comfyui_segment_anything's GroundingDINO needs it
v0.2.1
William Gill
2026-04-19 15:21:21 -05:00
-
f26dfbee02
smart_image_gen v0.7.2: chat-DB fallback + diagnostic 'no image' msg
William Gill
2026-04-19 15:07:55 -05:00
-
06433d3815
smart_image_gen v0.7.1: rename edit_image arg + parse file id from URL
William Gill
2026-04-19 14:58:40 -05:00
-
780ce42711
Image Studio: move tool_choice into params.custom_params (correct field)
William Gill
2026-04-19 14:50:34 -05:00
-
f6f5690fcd
smart_image_gen v0.7: edit_image finds previously-emitted images
William Gill
2026-04-19 14:46:10 -05:00
-
d935e24624
Add text-targeted inpainting via GroundingDINO+SAM (mask_text param)
v0.2.0
William Gill
2026-04-19 14:43:52 -05:00
-
7c7897818e
Image Studio: bake tool_choice=required into the preset
William Gill
2026-04-19 14:30:49 -05:00
-
a1fca4d5d9
Image Studio: default base model → huihui_ai/qwen3.5-abliterated:9b
William Gill
2026-04-19 14:22:13 -05:00
-
0b1c2ee5b5
Image Studio: tighten system prompt, add /no_think for Qwen 3.x
William Gill
2026-04-19 14:18:49 -05:00
-
5a34ced8f1
Add S3 mirror path for Ollama models + mirror-ollama-model.sh helper
William Gill
2026-04-19 13:43:26 -05:00
-
f77f5993fb
Image Studio: enable vision capability + document upgrade path
William Gill
2026-04-19 13:31:17 -05:00
-
b604e3f509
smart_image_gen v0.5: surface images via files event (canonical path)
William Gill
2026-04-19 13:21:48 -05:00
-
4d996e1205
smart_image_gen v0.4: emit image to chat, return only confirmation
William Gill
2026-04-19 13:14:59 -05:00
-
6adf133558
Ship Image Studio as importable JSON in addition to markdown walkthrough
William Gill
2026-04-19 13:04:49 -05:00
-
d4e2058859
smart_image_gen v0.3: add edit_image (img2img) method
William Gill
2026-04-19 12:59:13 -05:00
-
41d571d8d1
Add Image Studio model preset — forces smart_image_gen tool use
William Gill
2026-04-19 12:54:13 -05:00
-
9e22de0328
smart_image_gen: tighten docstring + Literal style enum
William Gill
2026-04-19 12:52:26 -05:00
-
b815cd6a5f
Tune static workflows to CyberRealisticXL recommended settings
William Gill
2026-04-19 12:47:46 -05:00
-
45d5541be0
smart_image_gen v0.2: per-style sampler/CFG/steps/CLIP-skip + prompt prefixes
William Gill
2026-04-19 12:45:34 -05:00
-
cd0034cd99
Flesh out per-style negatives in smart_image_gen Tool
William Gill
2026-04-19 12:39:24 -05:00
-
c585e53ed4
Bake quality-focused default negatives into the static workflows
William Gill
2026-04-19 12:36:44 -05:00
-
392b26167f
Add smart_image_gen Tool for per-prompt checkpoint routing
William Gill
2026-04-19 12:17:02 -05:00
-
704bcfdf13
Default workflows to SDXL CyberRealistic; ship empty model preseed
William Gill
2026-04-19 12:02:26 -05:00
-
0ad99b6199
Add comfyui-model-init sidecar for ComfyUI model preseeding
William Gill
2026-04-19 11:57:24 -05:00
-
21e976e275
Externalise WEBUI_URL / LLM_URL to .env
v0.1.0
William Gill
2026-04-19 10:49:29 -05:00
-
97547c783c
Make ai-stack the only deployment shape
William Gill
2026-04-19 10:45:23 -05:00
-
b1c9bff15f
Match init-models.sh to the live preseed list
William Gill
2026-04-19 10:41:29 -05:00
-
5b61caa197
Add deployments/ai-stack — combined production-shape example
William Gill
2026-04-19 10:40:41 -05:00
-
09fcec7c07
Add Gitea release workflow; pull image from registry by default
William Gill
2026-04-19 10:26:06 -05:00
-
a73aa4a38a
Initial commit — ComfyUI on NVIDIA fronted by Open WebUI
William Gill
2026-04-19 10:21:43 -05:00