Public/automatic
1
0
Fork 0
mirror of https://github.com/vladmandic/automatic synced 2026-09-19 17:24:32 +02:00
Code Issues Packages Projects Releases Wiki Activity

fix: correct comments and cleanup model descriptions

Browse Source 
- Fix Klein text encoder comment to specify correct sizes per variant
- Lock TAESD decode logging behind SD_PREVIEW_DEBUG env var
- Fix misleading comment about FLUX.2 128-channel reshape (is fallback)
- Remove VRAM requirements from model descriptions in reference files
This commit is contained in:
CalamitousFelicitousness
2026-01-16 03:24:39 +00:00
parent 5e2bc01367
commit eaa8dbcd42
4 changed files with 12 additions and 8 deletions
Download Patch File Download Diff File Expand all files Collapse all files
html/reference-distilled.json
+2 -2
View File
@@ -165,7 +165,7 @@
"Black Forest Labs FLUX.2 Klein 4B": {
"path": "black-forest-labs/FLUX.2-klein-4B",
"preview": "black-forest-labs--FLUX.2-klein-4B.jpg",
"desc": "FLUX.2-klein-4B is a 4 billion parameter size-distilled version of FLUX.2-dev optimized for consumer GPUs. Achieves sub-second inference with 4 steps while fitting in ~13GB VRAM. Supports both text-to-image generation and multi-reference image editing. Apache 2.0 licensed.",
"desc": "FLUX.2-klein-4B is a 4 billion parameter size-distilled version of FLUX.2-dev optimized for consumer GPUs. Achieves sub-second inference with 4 steps. Supports both text-to-image generation and multi-reference image editing. Apache 2.0 licensed.",
"skip": true,
"tags": "distilled",
"extras": "sampler: Default, cfg_scale: 4.0, steps: 4",
@@ -175,7 +175,7 @@
"Black Forest Labs FLUX.2 Klein 9B": {
"path": "black-forest-labs/FLUX.2-klein-9B",
"preview": "black-forest-labs--FLUX.2-klein-9B.jpg",
"desc": "FLUX.2-klein-9B is a 9 billion parameter size-distilled version of FLUX.2-dev. Higher quality than 4B variant with sub-second inference using 4 steps. Requires ~29GB VRAM. Supports text-to-image and multi-reference editing. Non-commercial license.",
"desc": "FLUX.2-klein-9B is a 9 billion parameter size-distilled version of FLUX.2-dev. Higher quality than 4B variant with sub-second inference using 4 steps. Supports text-to-image and multi-reference editing. Non-commercial license.",
"skip": true,
"tags": "distilled",
"extras": "sampler: Default, cfg_scale: 4.0, steps: 4",
html/reference.json
+3 -3
View File
@@ -127,7 +127,7 @@
"Black Forest Labs FLUX.2 Klein Base 4B": {
"path": "black-forest-labs/FLUX.2-klein-base-4B",
"preview": "black-forest-labs--FLUX.2-klein-base-4B.jpg",
"desc": "FLUX.2-klein-base-4B is the undistilled 4 billion parameter base model of FLUX.2-klein. Requires 50 inference steps for full quality but offers flexibility for fine-tuning. Fits in ~13GB VRAM. Supports text-to-image and multi-reference editing. Apache 2.0 licensed.",
"desc": "FLUX.2-klein-base-4B is the undistilled 4 billion parameter base model of FLUX.2-klein. Requires 50 inference steps for full quality but offers flexibility for fine-tuning. Supports text-to-image and multi-reference editing. Apache 2.0 licensed.",
"skip": true,
"extras": "sampler: Default, cfg_scale: 4.0, steps: 50",
"size": 8.5,
@@ -136,7 +136,7 @@
"Black Forest Labs FLUX.2 Klein Base 9B": {
"path": "black-forest-labs/FLUX.2-klein-base-9B",
"preview": "black-forest-labs--FLUX.2-klein-base-9B.jpg",
"desc": "FLUX.2-klein-base-9B is the undistilled 9 billion parameter base model of FLUX.2-klein. Requires 50 inference steps for full quality but offers flexibility for fine-tuning. Requires ~29GB VRAM. Supports text-to-image and multi-reference editing. Non-commercial license.",
"desc": "FLUX.2-klein-base-9B is the undistilled 9 billion parameter base model of FLUX.2-klein. Requires 50 inference steps for full quality but offers flexibility for fine-tuning. Supports text-to-image and multi-reference editing. Non-commercial license.",
"skip": true,
"extras": "sampler: Default, cfg_scale: 4.0, steps: 50",
"size": 18.5,
@@ -146,7 +146,7 @@
"Z-Image-Turbo": {
"path": "Tongyi-MAI/Z-Image-Turbo",
"preview": "Tongyi-MAI--Z-Image-Turbo.jpg",
"desc": "Z-Image-Turbo, a distilled version of Z-Image that matches or exceeds leading competitors with only 8 NFEs (Number of Function Evaluations). It offers sub-second inference latency on enterprise-grade H800 GPUs and fits comfortably within 16G VRAM consumer devices. It excels in photorealistic image generation, bilingual text rendering (English & Chinese), and robust instruction adherence.",
"desc": "Z-Image-Turbo, a distilled version of Z-Image that matches or exceeds leading competitors with only 8 NFEs (Number of Function Evaluations). It excels in photorealistic image generation, bilingual text rendering (English & Chinese), and robust instruction adherence.",
"skip": true,
"extras": "sampler: Default, cfg_scale: 1.0, steps: 9",
"size": 20.3,