diff --git a/optimized/I2I/README.md b/optimized/I2I/README.md index 6e06f13..5f712f4 100644 --- a/optimized/I2I/README.md +++ b/optimized/I2I/README.md @@ -1,6 +1,6 @@ # I2I workflows -Use [`flux/`](flux/), [`qwen/`](qwen/), or [`flux3klein/`](flux3klein/) for one-reference-image transformations. The workflows infer the subject from the input image instead of naming a specific character or scene. +Use [`flux/`](flux/), [`qwen/`](qwen/), or [`flux3klein/`](flux3klein/) for one-reference-image transformations. The workflows infer the subject from the input image instead of naming a specific character or scene. The [`qwen-noctq-flux3klein-refiner/`](qwen-noctq-flux3klein-refiner/) workflow chains a local Qwen NoctQ v4 generation into a low-denoise FLUX.3 Klein refinement pass. The shared reference asset and static variables are recorded in [`variables.json`](variables.json) and `character_reference.png`. The test plans live in each model README. diff --git a/optimized/I2I/qwen-noctq-flux3klein-refiner/README.md b/optimized/I2I/qwen-noctq-flux3klein-refiner/README.md new file mode 100644 index 0000000..6379ebe --- /dev/null +++ b/optimized/I2I/qwen-noctq-flux3klein-refiner/README.md @@ -0,0 +1,31 @@ +# Qwen NoctQ v4 → FLUX.3 Klein refiner I2I + +Two-stage local image-to-image workflow. Qwen NoctQ v4 generates the first +image, then the decoded pixels are passed to FLUX.3 Klein as both the image +reference and the VAE-encoded latent for a low-denoise refinement pass. + +## Fixed starter settings + +| Stage | Model | Steps | CFG / guidance | Sampler | Scheduler | Denoise | +|---|---|---:|---:|---|---|---:| +| Qwen base | `NoctQ_V4_int8_convrot.safetensors` | 8 | CFG 1.0 | `euler_ancestral` | `beta` | 1.0 | +| FLUX.3 Klein refiner | `flux3-image-klein-int8-convrot.safetensors` | 8 | Guidance 1.0 | `euler_ancestral` | `simple` | 0.11 | + +The text encoders are the installed local Heretic variants: +`qwen3-8b-heretic_int8.safetensors` for NoctQ and +`qwen3-vl-4b-heretic_int8.safetensors` for FLUX.3 Klein. No remote model API +is used. + +The input is `../character_reference.png` at 1080×1920 with seed `310101`. +The Qwen output is decoded with the Qwen VAE, then connected to the FLUX +reference conditioner and re-encoded with `flux2-vae.safetensors` before the +0.11-denoise refiner pass. + +## Test matrix + +| ID | Qwen settings | FLUX.3 Klein settings | Generation time (s) | Status | Review | +|---|---|---|---:|---|---| +| QNF3K-I2I-01 | 8 steps, CFG 1, euler_ancestral, beta | 8 steps, guidance 1, euler_ancestral, simple, denoise 0.11 | — | Planned | Starter refiner test | + +Record both stage runtimes, prompt ID, output dimensions, output path, model +names, and visual review notes in `results/QNF3K-I2I-01.json` after execution. diff --git a/optimized/I2I/qwen-noctq-flux3klein-refiner/variables.json b/optimized/I2I/qwen-noctq-flux3klein-refiner/variables.json new file mode 100644 index 0000000..754785d --- /dev/null +++ b/optimized/I2I/qwen-noctq-flux3klein-refiner/variables.json @@ -0,0 +1,28 @@ +{ + "workflow": "Qwen NoctQ v4 -> FLUX.3 Klein refiner", + "input_image": "../character_reference.png", + "resolution": {"width":1080,"height":1920}, + "seed": 310101, + "qwen_stage": { + "model": "NoctQ_V4_int8_convrot.safetensors", + "text_encoder": "qwen3-8b-heretic_int8.safetensors", + "vae": "qwen_image_2.1_vae_bf16.safetensors", + "steps": 8, + "cfg": 1.0, + "sampler": "euler_ancestral", + "scheduler": "beta", + "denoise": 1.0 + }, + "refiner_stage": { + "model": "flux3-image-klein-int8-convrot.safetensors", + "text_encoder": "qwen3-vl-4b-heretic_int8.safetensors", + "vae": "flux2-vae.safetensors", + "steps": 8, + "guidance": 1.0, + "sampler": "euler_ancestral", + "scheduler": "simple", + "denoise": 0.11 + }, + "status": "planned", + "results": [] +} diff --git a/optimized/I2I/qwen-noctq-flux3klein-refiner/workflow.json b/optimized/I2I/qwen-noctq-flux3klein-refiner/workflow.json new file mode 100644 index 0000000..82f0683 --- /dev/null +++ b/optimized/I2I/qwen-noctq-flux3klein-refiner/workflow.json @@ -0,0 +1,24 @@ +{ + "1": {"class_type":"UNETLoader","inputs":{"unet_name":"NoctQ_V4_int8_convrot.safetensors","weight_dtype":"default"}}, + "2": {"class_type":"CLIPLoader","inputs":{"clip_name":"qwen3-8b-heretic_int8.safetensors","type":"qwen_image","device":"default"}}, + "3": {"class_type":"VAELoader","inputs":{"vae_name":"qwen_image_2.1_vae_bf16.safetensors"}}, + "4": {"class_type":"LoadImage","inputs":{"image":"character_reference.png"}}, + "5": {"class_type":"ResizeAndPadImage","inputs":{"image":["4",0],"target_width":1080,"target_height":1920,"padding_color":"white","interpolation":"lanczos"}}, + "6": {"class_type":"QwenImage21Cache","inputs":{"model":["1",0],"device":"auto","dtype":"int8"}}, + "7": {"class_type":"TextEncodeQwenImage21","inputs":{"clip":["2",0],"prompt":"Transform the supplied reference image into a photorealistic image of the same subject. Preserve identity, facial features, hair, clothing, accessories, pose, composition, and framing faithfully. Use realistic anatomy, skin and hair texture, physically accurate materials, natural lighting, coherent perspective, premium editorial photography, high dynamic range, and fine detail. Do not redesign, substitute, add, remove, mirror, or crop the subject. No text or logos.","negative_prompt":"different person, changed identity, changed face, changed hairstyle, changed clothing, changed colors, altered accessories, changed pose, different composition, extra person, duplicate person, extra limbs, malformed hands, fused fingers, distorted anatomy, cropped head, cropped feet, mirrored image, cartoon, illustration, anime, CGI, plastic skin, waxy skin, blurry, low detail, text, logo, watermark","resolution":0,"images":{"image_1":["5",0]},"vae":["3",0]}}, + "8": {"class_type":"KSamplerSelect","inputs":{"sampler_name":"euler_ancestral"}}, + "9": {"class_type":"BasicScheduler","inputs":{"model":["6",0],"scheduler":"beta","steps":8,"denoise":1.0}}, + "10": {"class_type":"RandomNoise","inputs":{"noise_seed":310101}}, + "11": {"class_type":"CFGGuider","inputs":{"model":["6",0],"positive":["7",0],"negative":["7",1],"cfg":1.0}}, + "12": {"class_type":"SamplerCustomAdvanced","inputs":{"noise":["10",0],"guider":["11",0],"sampler":["8",0],"sigmas":["9",0],"latent_image":["7",2]}}, + "13": {"class_type":"VAEDecode","inputs":{"samples":["12",0],"vae":["3",0]}}, + "14": {"class_type":"UNETLoader","inputs":{"unet_name":"flux3-image-klein-int8-convrot.safetensors","weight_dtype":"default"}}, + "15": {"class_type":"CLIPLoader","inputs":{"clip_name":"qwen3-vl-4b-heretic_int8.safetensors","type":"flux3","device":"default"}}, + "16": {"class_type":"VAELoader","inputs":{"vae_name":"flux2-vae.safetensors"}}, + "17": {"class_type":"StarFlux2Conditioner","inputs":{"clip":["15",0],"vae":["16",0],"text":"Refine the supplied image while preserving the subject identity, composition, pose, clothing, colors, and overall visual content. Improve realism, detail, anatomy, materials, and lighting without redesigning or adding objects.","join_references":true,"image_1":["13",0]}}, + "18": {"class_type":"FluxGuidance","inputs":{"conditioning":["17",0],"guidance":1.0}}, + "19": {"class_type":"VAEEncode","inputs":{"pixels":["13",0],"vae":["16",0]}}, + "20": {"class_type":"FluxSampler","inputs":{"model":["14",0],"conditioning":["18",0],"latent_image":["19",0],"sampler_name":"euler_ancestral","scheduler":"simple","steps":8,"denoise":0.11,"noise_seed":310101}}, + "21": {"class_type":"VAEDecode","inputs":{"samples":["20",0],"vae":["16",0]}}, + "22": {"class_type":"SaveImage","inputs":{"images":["21",0],"filename_prefix":"fluxExperiments/qwenNoctQ_flux3klein_refiner/QNF3K-I2I-01"}} +}