diff --git a/optimized/I2IPlusRefiner/README.md b/optimized/I2IPlusRefiner/README.md index f241851..d607f46 100644 --- a/optimized/I2IPlusRefiner/README.md +++ b/optimized/I2IPlusRefiner/README.md @@ -18,12 +18,13 @@ text prompt, so it cannot replace the subject from text conditioning. | QRF-I2I-01 | `refiner_reference.jpg` | Canonical shared I2I prompt | NoctQ v4, 8 steps, CFG 1, Euler ancestral / beta | FLUX.3 Klein, 8 steps, guidance 1, Euler ancestral / simple, denoise 0.11 | 1088×1920 | 115.720 uncached; 0.274 cached | Completed | Base and refined outputs are nearly identical at denoise 0.11; [base PNG](results/QRF-I2I-01-base.png) vs [refined PNG](results/QRF-I2I-01.png) | | QRF-I2I-02 | `refiner_reference.jpg` | Canonical shared I2I prompt | Qwen Image 2.1, 50 steps, CFG 3.5, Euler CFG++ / beta | FLUX.3 Klein, 8 steps, guidance 1, Euler ancestral / simple, denoise 0.11 | Base 1080×1920; refined 1072×1920 | 145.129 | Completed | More realistic base; [base PNG](results/QRF-I2I-02-base.png) vs [refined PNG](results/QRF-I2I-02.png) | | QRF-I2I-03 | `refiner_reference.jpg` | Canonical shared I2I prompt | Qwen Image 2.1, 8 steps, CFG 1, Euler ancestral / beta | FLUX.3 Klein, 8 steps, guidance 1, Euler ancestral / simple, denoise 0.11 | Base 1080×1920; refined 1072×1920 | 60.252 | Completed | Realistic low-step base; white top/bottom bands; [base PNG](results/QRF-I2I-03-base.png) vs [refined PNG](results/QRF-I2I-03.png) | +| ZRF-I2I-01 | `refiner_reference.jpg` | Canonical shared I2I prompt | Qwen Image 2.1, 8 steps, CFG 1, Euler ancestral / beta | ZImage Turbo, 5 steps, CFG 1, DPM++ 2M SDE / beta, denoise 0.33, shift 3 | Base/refined 1080×1920; refiner internal canvas 1088×1920 | 33.571 | Completed | Clean refiner output preserves the subject and clothing; [base PNG](results/ZRF-I2I-01-base.png) vs [refined PNG](results/ZRF-I2I-01.png) | ## Planned follow-ups -- ZImage as the refiner, using the same base output, prompt, input image, and - resolution. - Additional refiner models only after the first comparison is recorded. -Record the resolved workflow, prompt ID, model names, runtime, dimensions, -hashes, and visual review in `results/QRF-I2I-01.json`. +The FLUX workflow is `workflow.json`; the ZImage variant is +`zimage-refiner-workflow.json`. Each resolved run records its prompt ID, model +names, runtime, dimensions, hashes, and visual review in its matching metadata +file under `results/`. diff --git a/optimized/I2IPlusRefiner/results/ZRF-I2I-01-base.png b/optimized/I2IPlusRefiner/results/ZRF-I2I-01-base.png new file mode 100644 index 0000000..4020911 Binary files /dev/null and b/optimized/I2IPlusRefiner/results/ZRF-I2I-01-base.png differ diff --git a/optimized/I2IPlusRefiner/results/ZRF-I2I-01.json b/optimized/I2IPlusRefiner/results/ZRF-I2I-01.json new file mode 100644 index 0000000..98781ca --- /dev/null +++ b/optimized/I2IPlusRefiner/results/ZRF-I2I-01.json @@ -0,0 +1,52 @@ +{ + "test_id": "ZRF-I2I-01", + "status": "completed", + "prompt_id": "ac1d835c-659e-4328-909d-b24ceb15343c", + "server": "http://192.168.1.202:8188", + "input_image": "refiner_reference.jpg", + "input_image_sha256": "f478ab8064bd0642401cbf6bab9aa95380ed4c05d50c6d2959bc9a0eb632354a", + "prompt_source": "../I2I/qwen/workflow.json#7", + "prompt_used_by_base": true, + "prompt_used_by_refiner": true, + "base_stage": { + "model": "qwen_image_2.1_int8_convrot.safetensors", + "text_encoder": "qwen3vl_8b_w4a8.safetensors", + "vae": "qwen_image_2.1_vae_bf16.safetensors", + "generation_width": 576, + "generation_height": 1024, + "steps": 8, + "cfg": 1.0, + "sampler": "euler_ancestral", + "scheduler": "beta", + "denoise": 1.0 + }, + "refiner_stage": { + "model": "z_image_turbo_int8_convrot.safetensors", + "text_encoder": "uncensoredQwen34BLLMForZ_fp8.safetensors", + "vae": "ae.safetensors", + "internal_width": 1088, + "internal_height": 1920, + "steps": 5, + "cfg": 1.0, + "sampler": "dpmpp_2m_sde", + "scheduler": "beta", + "denoise": 0.33, + "sampling_shift": 3.0 + }, + "generation_time_seconds": 33.571, + "outputs": { + "base": { + "path": "results/ZRF-I2I-01-base.png", + "width": 1080, + "height": 1920, + "sha256": "f6161ea1edb31ed9a85254a2360f61f102248348243e7f958bf17cd13dd18021" + }, + "refined": { + "path": "results/ZRF-I2I-01.png", + "width": 1080, + "height": 1920, + "sha256": "a3751a19d49ced3df4c624bd7dcab430387674c7139286502b44abd2c29a84dc" + } + }, + "review": "Successful ZImage refiner run. The first attempt used TextEncodeZImageOmni with the base image as a second reference and produced a visible grid/ghosting artifact. The final run uses the standard ZImage image-to-image path: the same canonical prompt plus the Qwen base latent, with an internal 1088x1920 aligned canvas and a final 1080x1920 resize. The refined image preserves the subject, glasses, hair, suit, and portrait framing without the artifact." +} diff --git a/optimized/I2IPlusRefiner/results/ZRF-I2I-01.png b/optimized/I2IPlusRefiner/results/ZRF-I2I-01.png new file mode 100644 index 0000000..16fb130 Binary files /dev/null and b/optimized/I2IPlusRefiner/results/ZRF-I2I-01.png differ diff --git a/optimized/I2IPlusRefiner/variables.json b/optimized/I2IPlusRefiner/variables.json index 4e8427a..e246fd1 100644 --- a/optimized/I2IPlusRefiner/variables.json +++ b/optimized/I2IPlusRefiner/variables.json @@ -24,6 +24,19 @@ "scheduler": "simple", "denoise": 0.11 }, + "zimage_refiner_stage": { + "workflow": "zimage-refiner-workflow.json", + "model": "z_image_turbo_int8_convrot.safetensors", + "text_encoder": "uncensoredQwen34BLLMForZ_fp8.safetensors", + "vae": "ae.safetensors", + "internal_resolution": {"width": 1088, "height": 1920}, + "steps": 5, + "cfg": 1.0, + "sampler": "dpmpp_2m_sde", + "scheduler": "beta", + "denoise": 0.33, + "sampling_shift": 3.0 + }, "status": "completed", "results": [ { @@ -56,6 +69,17 @@ "refined": "results/QRF-I2I-03.png" }, "metadata": "results/QRF-I2I-03.json" + }, + { + "id": "ZRF-I2I-01", + "status": "completed", + "generation_time_seconds": 33.571, + "outputs": { + "base": "results/ZRF-I2I-01-base.png", + "refined": "results/ZRF-I2I-01.png" + }, + "metadata": "results/ZRF-I2I-01.json", + "refiner_workflow": "zimage-refiner-workflow.json" } ] } diff --git a/optimized/I2IPlusRefiner/zimage-refiner-workflow.json b/optimized/I2IPlusRefiner/zimage-refiner-workflow.json new file mode 100644 index 0000000..9147630 --- /dev/null +++ b/optimized/I2IPlusRefiner/zimage-refiner-workflow.json @@ -0,0 +1,29 @@ +{ + "1": {"class_type":"UNETLoader","inputs":{"unet_name":"qwen_image_2.1_int8_convrot.safetensors","weight_dtype":"default"}}, + "2": {"class_type":"CLIPLoader","inputs":{"clip_name":"qwen3vl_8b_w4a8.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":"refiner_reference.jpg"}}, + "5": {"class_type":"ResizeAndPadImage","inputs":{"image":["4",0],"target_width":576,"target_height":1024,"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. Infer identity, facial features, skin tone, hairstyle, clothing, accessories, colors, pose, body proportions, composition, and framing directly from the reference. Preserve those visual attributes faithfully while replacing the original rendering style with natural human anatomy, realistic skin and hair texture, physically accurate materials, believable fabric and surface detail, natural hands and fingers, cinematic soft light, realistic shadows, coherent perspective, premium editorial photography, high dynamic range, photorealistic, highly detailed. Do not redesign, restyle, 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]}}, + "28": {"class_type":"ResizeAndPadImage","inputs":{"image":["13",0],"target_width":1080,"target_height":1920,"padding_color":"white","interpolation":"lanczos"}}, + "14": {"class_type":"UNETLoader","inputs":{"unet_name":"z_image_turbo_int8_convrot.safetensors","weight_dtype":"default"}}, + "15": {"class_type":"CLIPLoader","inputs":{"clip_name":"uncensoredQwen34BLLMForZ_fp8.safetensors","type":"lumina2","device":"default"}}, + "16": {"class_type":"VAELoader","inputs":{"vae_name":"ae.safetensors"}}, + "29": {"class_type":"ModelSamplingAuraFlow","inputs":{"model":["14",0],"shift":3.0}}, + "30": {"class_type":"CLIPTextEncode","inputs":{"clip":["15",0],"text":"Transform the supplied reference image into a photorealistic image of the same subject. Infer identity, facial features, skin tone, hairstyle, clothing, accessories, colors, pose, body proportions, composition, and framing directly from the reference. Preserve those visual attributes faithfully while replacing the original rendering style with natural human anatomy, realistic skin and hair texture, physically accurate materials, believable fabric and surface detail, natural hands and fingers, cinematic soft light, realistic shadows, coherent perspective, premium editorial photography, high dynamic range, photorealistic, highly detailed. Do not redesign, restyle, substitute, add, remove, mirror, or crop the subject. No text or logos."}}, + "31": {"class_type":"CLIPTextEncode","inputs":{"text":"","clip":["15",0]}}, + "23": {"class_type":"VAEEncode","inputs":{"pixels":["35",0],"vae":["16",0]}}, + "32": {"class_type":"KSampler","inputs":{"seed":310101,"control_after_generate":"fixed","steps":5,"cfg":1.0,"sampler_name":"dpmpp_2m_sde","scheduler":"beta","positive":["30",0],"negative":["31",0],"latent_image":["23",0],"denoise":0.33,"model":["29",0]}}, + "33": {"class_type":"VAEDecode","inputs":{"samples":["32",0],"vae":["16",0]}}, + "35": {"class_type":"ResizeAndPadImage","inputs":{"image":["28",0],"target_width":1088,"target_height":1920,"padding_color":"white","interpolation":"lanczos"}}, + "36": {"class_type":"ResizeAndPadImage","inputs":{"image":["33",0],"target_width":1080,"target_height":1920,"padding_color":"white","interpolation":"lanczos"}}, + "27": {"class_type":"SaveImage","inputs":{"images":["28",0],"filename_prefix":"fluxExperiments/i2i_refiner/ZRF-I2I-01-base"}}, + "34": {"class_type":"SaveImage","inputs":{"images":["36",0],"filename_prefix":"fluxExperiments/i2i_refiner/ZRF-I2I-01"}} +}