Frogger40 commited on
Commit
659dfea
·
verified ·
1 Parent(s): b723c7c

Upload README.md with huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +25 -48
README.md CHANGED
@@ -1,73 +1,50 @@
1
  ---
2
  base_model: krea/Krea-2-Raw
3
- library_name: diffusers
4
- license: apache-2.0
5
- instance_prompt: m4r4kr342
6
- widget: []
7
  tags:
8
  - text-to-image
9
- - diffusers-training
10
  - diffusers
11
  - lora
12
  - krea2
13
- - krea2-diffusers
14
  - template:sd-lora
 
 
 
 
 
 
 
 
 
15
  ---
16
 
17
- <!-- This model card has been generated automatically according to the information the training script had access to. You
18
- should probably proofread and complete it, then remove this comment. -->
19
-
20
-
21
- # Krea 2 DreamBooth LoRA - Frogger40/mara2
22
 
23
  <Gallery />
24
 
25
- ## Model description
26
 
27
- These are Frogger40/mara2 DreamBooth LoRA weights, trained on krea/Krea-2-Raw.
28
 
29
- The weights were trained using [DreamBooth](https://dreambooth.github.io/) with the [Krea 2 diffusers trainer](https://github.com/huggingface/diffusers/blob/main/examples/dreambooth/README_krea2.md).
30
 
31
- Krea 2 ships as two checkpoints: **RAW** (the non-distilled base you fine-tune on) and **Turbo** (an 8-step distilled checkpoint for fast, high-quality inference). Train your LoRA on RAW and run it on Turbo — LoRAs trained on RAW express strongly on Turbo.
32
 
33
- ## Trigger words
34
 
35
- You should use `m4r4kr342` to trigger the image generation.
36
 
37
- ## Download model
38
 
39
- [Download the *.safetensors LoRA](Frogger40/mara2/tree/main) in the Files & versions tab.
40
 
41
- ## Use it with the [🧨 diffusers library](https://github.com/huggingface/diffusers)
42
 
43
  ```py
44
- >>> import torch
45
- >>> from diffusers import Krea2Pipeline
46
 
47
- >>> # Load the LoRA onto Krea 2 Turbo (the distilled inference model)
48
- >>> pipe = Krea2Pipeline.from_pretrained("krea/Krea-2-Turbo", torch_dtype=torch.bfloat16).to("cuda")
49
- >>> pipe.load_lora_weights("Frogger40/mara2")
50
-
51
- >>> # Turbo recipe: 8 steps, no classifier-free guidance
52
- >>> image = pipe("m4r4kr342", num_inference_steps=8, guidance_scale=0.0).images[0]
53
- >>> image.save("output.png")
54
  ```
55
-
56
- For more details, including weighting, merging and fusing LoRAs, check the [documentation on loading LoRAs in diffusers](https://huggingface.co/docs/diffusers/main/en/using-diffusers/loading_adapters)
57
-
58
-
59
- ## Intended uses & limitations
60
-
61
- #### How to use
62
-
63
- ```python
64
- # TODO: add an example code snippet for running this diffusion pipeline
65
- ```
66
-
67
- #### Limitations and bias
68
-
69
- [TODO: provide examples of latent issues and potential remediations]
70
-
71
- ## Training details
72
-
73
- [TODO: describe the data used to train the model]
 
1
  ---
2
  base_model: krea/Krea-2-Raw
 
 
 
 
3
  tags:
4
  - text-to-image
 
5
  - diffusers
6
  - lora
7
  - krea2
 
8
  - template:sd-lora
9
+ license: apache-2.0
10
+ instance_prompt: "m4r4kr342"
11
+ widget:
12
+ - text: "m4r4kr34, A highly realistic candid iPhone photo captures a young woman with beautiful, shoulder-length dark brown hair, smiling warmly directly at the camera. She is seated comfortably on a dark, black slatted public bench, leaning slightly forward in a relaxed, inviting posture. She is dressed in a casual yet stylish outfit: a fitted dark brown leather jacket worn over a crisp white button-down shirt, paired with loose-fitting, light grey wide-leg jeans. On her feet, she wears simple green flip-flop sandals. She is holding a white takeaway coffee cup with both hands, resting it near her lap. The setting is urban and architectural, positioned outside a classic European-style building with a light-colored stone facade (the number '28' is visible on the left). Behind her, a large, ornate window reflects the surrounding street scene, showing a glimpse of the building across the street and a blurred figure (a man in a light beige jacket) sitting inside, perhaps at a café table. The lighting is bright, soft natural daylight, giving the scene a warm, golden, and very approachable feel. The photo is framed in a medium close-up, capturing her from the waist up, and feels like a genuine, everyday lifestyle snapshot with realistic skin texture, subtle pores, and natural color grading"
13
+ output:
14
+ url: sample_0.png
15
+ - text: "m4r4kr34 A confident 23-year-old woman with black hair, soft lighting, sitting in a cafe, sunny morning, photrealistic"
16
+ output:
17
+ url: sample_1.png
18
  ---
19
 
20
+ # Krea 2 LoRA Frogger40/mara2
 
 
 
 
21
 
22
  <Gallery />
23
 
24
+ A DreamBooth-LoRA for **Krea 2**, trained on **Krea 2 RAW** and shown on **Krea 2 Turbo**. The samples below were generated with this LoRA on Turbo (8 steps).
25
 
26
+ ## Trigger
27
 
28
+ Use the token `m4r4kr342` to invoke the concept.
29
 
30
+ ## Samples
31
 
32
+ ![sample](./sample_0.png)
33
 
34
+ > *"m4r4kr34, A highly realistic candid iPhone photo captures a young woman with beautiful, shoulder-length dark brown hair, smiling warmly directly at the camera. She is seated comfortably on a dark, black slatted public bench, leaning slightly forward in a relaxed, inviting posture. She is dressed in a casual yet stylish outfit: a fitted dark brown leather jacket worn over a crisp white button-down shirt, paired with loose-fitting, light grey wide-leg jeans. On her feet, she wears simple green flip-flop sandals. She is holding a white takeaway coffee cup with both hands, resting it near her lap. The setting is urban and architectural, positioned outside a classic European-style building with a light-colored stone facade (the number '28' is visible on the left). Behind her, a large, ornate window reflects the surrounding street scene, showing a glimpse of the building across the street and a blurred figure (a man in a light beige jacket) sitting inside, perhaps at a café table. The lighting is bright, soft natural daylight, giving the scene a warm, golden, and very approachable feel. The photo is framed in a medium close-up, capturing her from the waist up, and feels like a genuine, everyday lifestyle snapshot with realistic skin texture, subtle pores, and natural color grading"*
35
 
36
+ ![sample](./sample_1.png)
37
 
38
+ > *"m4r4kr34 A confident 23-year-old woman with black hair, soft lighting, sitting in a cafe, sunny morning, photrealistic"*
39
 
40
+ ## Use it with diffusers
41
 
42
  ```py
43
+ import torch
44
+ from diffusers import Krea2Pipeline
45
 
46
+ pipe = Krea2Pipeline.from_pretrained("krea/Krea-2-Turbo", torch_dtype=torch.bfloat16).to("cuda")
47
+ pipe.load_lora_weights("Frogger40/mara2")
48
+ image = pipe("m4r4kr34, A highly realistic candid iPhone photo captures a young woman with beautiful, shoulder-length dark brown hair, smiling warmly directly at the camera. She is seated comfortably on a dark, black slatted public bench, leaning slightly forward in a relaxed, inviting posture. She is dressed in a casual yet stylish outfit: a fitted dark brown leather jacket worn over a crisp white button-down shirt, paired with loose-fitting, light grey wide-leg jeans. On her feet, she wears simple green flip-flop sandals. She is holding a white takeaway coffee cup with both hands, resting it near her lap. The setting is urban and architectural, positioned outside a classic European-style building with a light-colored stone facade (the number '28' is visible on the left). Behind her, a large, ornate window reflects the surrounding street scene, showing a glimpse of the building across the street and a blurred figure (a man in a light beige jacket) sitting inside, perhaps at a café table. The lighting is bright, soft natural daylight, giving the scene a warm, golden, and very approachable feel. The photo is framed in a medium close-up, capturing her from the waist up, and feels like a genuine, everyday lifestyle snapshot with realistic skin texture, subtle pores, and natural color grading", num_inference_steps=8, guidance_scale=0.0).images[0]
49
+ image.save("output.png")
 
 
 
50
  ```