Working Example with diffusers
#4
by
aggr8
- opened
Hi,
I've been trying to generate images using the checkpoint in diffusers but no matter what scheduler I try, I always get very blurry results. Can you please provide details on the exact scheduler and it's hyper-params that can generate images of the same quality as shown in examples.
Hello @aggr8 ! All images were generated using ComfyUI with HiResFix , DPM++ 3M SDE Karras, 30 Steps first Stage and 30 Steps in HiRes Stage. The HiRes denoise is set to 0.6. I can send you the ComfyUI json file. The prompts and the seeds as well. Please let me know if that would be helpful.
hammh0a
changed discussion status to
closed
hammh0a
changed discussion status to
open
Let me look into that, I have not tested this through diffusers yet - I don't seem to find DPM++ 3M SDE implemented there yet also I'm not aware of the HiRes-Fix there, give me sometime and I'll check it out.
I'd love the .json workflow :)
@huwhite you can either import from the attached image or the JSON.
{
"last_node_id": 19,
"last_link_id": 14,
"nodes": [
{
"id": 16,
"type": "LatentUpscaleBy",
"pos": [
1025,
130
],
"size": {
"0": 315,
"1": 82
},
"flags": {},
"order": 2,
"mode": 0,
"inputs": [
{
"name": "samples",
"type": "LATENT",
"link": 11
}
],
"outputs": [
{
"name": "LATENT",
"type": "LATENT",
"links": [
9
],
"shape": 3
}
],
"properties": {
"Node name for S&R": "LatentUpscaleBy"
},
"widgets_values": [
"nearest-exact",
1.5
]
},
{
"id": 19,
"type": "PreviewImage",
"pos": [
1193,
422
],
"size": {
"0": 210,
"1": 246
},
"flags": {},
"order": 5,
"mode": 0,
"inputs": [
{
"name": "images",
"type": "IMAGE",
"link": 14
}
],
"properties": {
"Node name for S&R": "PreviewImage"
}
},
{
"id": 11,
"type": "KSampler (Efficient)",
"pos": [
600,
130
],
"size": {
"0": 325,
"1": 562
},
"flags": {},
"order": 1,
"mode": 0,
"inputs": [
{
"name": "model",
"type": "MODEL",
"link": 1
},
{
"name": "positive",
"type": "CONDITIONING",
"link": 2
},
{
"name": "negative",
"type": "CONDITIONING",
"link": 3
},
{
"name": "latent_image",
"type": "LATENT",
"link": 4
},
{
"name": "optional_vae",
"type": "VAE",
"link": 5
},
{
"name": "script",
"type": "SCRIPT",
"link": null
}
],
"outputs": [
{
"name": "MODEL",
"type": "MODEL",
"links": [
6
],
"shape": 3
},
{
"name": "CONDITIONING+",
"type": "CONDITIONING",
"links": [
7
],
"shape": 3
},
{
"name": "CONDITIONING-",
"type": "CONDITIONING",
"links": [
8
],
"shape": 3
},
{
"name": "LATENT",
"type": "LATENT",
"links": [
11
],
"shape": 3
},
{
"name": "VAE",
"type": "VAE",
"links": [
10
],
"shape": 3
},
{
"name": "IMAGE",
"type": "IMAGE",
"links": [
13
],
"shape": 3,
"slot_index": 5
}
],
"properties": {
"Node name for S&R": "KSampler (Efficient)"
},
"widgets_values": [
392212875588156,
null,
30,
6,
"dpmpp_3m_sde",
"karras",
1,
"auto",
"true"
],
"color": "#443322",
"bgcolor": "#665533",
"shape": 1
},
{
"id": 14,
"type": "KSampler (Efficient)",
"pos": [
1440,
134
],
"size": {
"0": 325,
"1": 562
},
"flags": {},
"order": 4,
"mode": 0,
"inputs": [
{
"name": "model",
"type": "MODEL",
"link": 6
},
{
"name": "positive",
"type": "CONDITIONING",
"link": 7
},
{
"name": "negative",
"type": "CONDITIONING",
"link": 8
},
{
"name": "latent_image",
"type": "LATENT",
"link": 9
},
{
"name": "optional_vae",
"type": "VAE",
"link": 10
},
{
"name": "script",
"type": "SCRIPT",
"link": null
}
],
"outputs": [
{
"name": "MODEL",
"type": "MODEL",
"links": null,
"shape": 3
},
{
"name": "CONDITIONING+",
"type": "CONDITIONING",
"links": null,
"shape": 3
},
{
"name": "CONDITIONING-",
"type": "CONDITIONING",
"links": null,
"shape": 3
},
{
"name": "LATENT",
"type": "LATENT",
"links": null,
"shape": 3
},
{
"name": "VAE",
"type": "VAE",
"links": null,
"shape": 3
},
{
"name": "IMAGE",
"type": "IMAGE",
"links": [
14
],
"shape": 3,
"slot_index": 5
}
],
"properties": {
"Node name for S&R": "KSampler (Efficient)"
},
"widgets_values": [
392212875588156,
null,
30,
6,
"dpmpp_3m_sde",
"karras",
0.5,
"auto",
"true"
],
"color": "#223322",
"bgcolor": "#335533",
"shape": 1
},
{
"id": 12,
"type": "Efficient Loader",
"pos": [
100,
130
],
"size": {
"0": 400,
"1": 462
},
"flags": {},
"order": 0,
"mode": 0,
"inputs": [
{
"name": "lora_stack",
"type": "LORA_STACK",
"link": null
},
{
"name": "cnet_stack",
"type": "CONTROL_NET_STACK",
"link": null
}
],
"outputs": [
{
"name": "MODEL",
"type": "MODEL",
"links": [
1
],
"shape": 3
},
{
"name": "CONDITIONING+",
"type": "CONDITIONING",
"links": [
2
],
"shape": 3
},
{
"name": "CONDITIONING-",
"type": "CONDITIONING",
"links": [
3
],
"shape": 3
},
{
"name": "LATENT",
"type": "LATENT",
"links": [
4
],
"shape": 3
},
{
"name": "VAE",
"type": "VAE",
"links": [
5
],
"shape": 3
},
{
"name": "CLIP",
"type": "CLIP",
"links": null,
"shape": 3
},
{
"name": "DEPENDENCIES",
"type": "DEPENDENCIES",
"links": null,
"shape": 3
}
],
"properties": {
"Node name for S&R": "Efficient Loader"
},
"widgets_values": [
"salle_sdv15_135K_steps_v1.safetensors",
"Baked VAE",
-1,
"None",
1,
1,
"A breathtaking 8K resolution photograph capturing a majestic dragon perched atop a rugged mountain peak, its scales shimmering under the setting sun. The scene exudes a powerful, yet serene atmosphere, with (Photorealistic:1.4), (Vivid colors:1.2), (Sharp focus:1.3), inspired by the fantasy landscapes of Roger Dean and the detailed creature design by Guillermo del Toro.\n",
"watermark, signature, worst quality, low quality, normal quality, lowres, simple background, inaccurate limb, extra fingers, fewer fingers, missing fingers, extra arms, (extra legs:1.3), inaccurate eyes, bad composition, bad anatomy, error, extra digit, fewer digits, cropped, low res, worst quality, low quality, normal quality, jpeg artifacts, extra digit, fewer digits, trademark, watermark, artist's name, username, signature, text, words, dark color, black and white, monochrome",
"length+mean",
"A1111",
512,
512,
1
],
"color": "#332233",
"bgcolor": "#553355",
"shape": 1
},
{
"id": 18,
"type": "PreviewImage",
"pos": [
956,
421
],
"size": {
"0": 210,
"1": 246
},
"flags": {},
"order": 3,
"mode": 0,
"inputs": [
{
"name": "images",
"type": "IMAGE",
"link": 13
}
],
"properties": {
"Node name for S&R": "PreviewImage"
}
}
],
"links": [
[
1,
12,
0,
11,
0,
"MODEL"
],
[
2,
12,
1,
11,
1,
"CONDITIONING"
],
[
3,
12,
2,
11,
2,
"CONDITIONING"
],
[
4,
12,
3,
11,
3,
"LATENT"
],
[
5,
12,
4,
11,
4,
"VAE"
],
[
6,
11,
0,
14,
0,
"MODEL"
],
[
7,
11,
1,
14,
1,
"CONDITIONING"
],
[
8,
11,
2,
14,
2,
"CONDITIONING"
],
[
9,
16,
0,
14,
3,
"LATENT"
],
[
10,
11,
4,
14,
4,
"VAE"
],
[
11,
11,
3,
16,
0,
"LATENT"
],
[
13,
11,
5,
18,
0,
"IMAGE"
],
[
14,
14,
5,
19,
0,
"IMAGE"
]
],
"groups": [],
"config": {},
"extra": {},
"version": 0.4
}