PseudoTerminal X
commited on
Model card auto-generated by SimpleTuner
Browse files
README.md
CHANGED
@@ -10,7 +10,7 @@ tags:
|
|
10 |
- template:sd-lora
|
11 |
inference: true
|
12 |
widget:
|
13 |
-
- text: '
|
14 |
parameters:
|
15 |
negative_prompt: 'blurry, cropped, ugly'
|
16 |
output:
|
@@ -24,7 +24,7 @@ This is a LoRA derived from [stabilityai/stable-diffusion-3-medium-diffusers](ht
|
|
24 |
The main validation prompt used during training was:
|
25 |
|
26 |
```
|
27 |
-
a
|
28 |
```
|
29 |
|
30 |
## Validation settings
|
@@ -49,7 +49,7 @@ You may reuse the base model text encoder for inference.
|
|
49 |
## Training settings
|
50 |
|
51 |
- Training epochs: 0
|
52 |
-
- Training steps:
|
53 |
- Learning rate: 1e-06
|
54 |
- Effective batch size: 1
|
55 |
- Micro-batch size: 1
|
@@ -82,10 +82,11 @@ You may reuse the base model text encoder for inference.
|
|
82 |
|
83 |
|
84 |
```python
|
85 |
-
|
|
|
86 |
|
87 |
model_id = "sd3-lora-test"
|
88 |
-
prompt = "a
|
89 |
negative_prompt = "malformed, disgusting, overexposed, washed-out"
|
90 |
|
91 |
pipeline = DiffusionPipeline.from_pretrained(model_id)
|
|
|
10 |
- template:sd-lora
|
11 |
inference: true
|
12 |
widget:
|
13 |
+
- text: 'a psychedelic man is surfing on top of a horse'
|
14 |
parameters:
|
15 |
negative_prompt: 'blurry, cropped, ugly'
|
16 |
output:
|
|
|
24 |
The main validation prompt used during training was:
|
25 |
|
26 |
```
|
27 |
+
a psychedelic man is surfing on top of a horse
|
28 |
```
|
29 |
|
30 |
## Validation settings
|
|
|
49 |
## Training settings
|
50 |
|
51 |
- Training epochs: 0
|
52 |
+
- Training steps: 94
|
53 |
- Learning rate: 1e-06
|
54 |
- Effective batch size: 1
|
55 |
- Micro-batch size: 1
|
|
|
82 |
|
83 |
|
84 |
```python
|
85 |
+
import torchfrom diffusers import StableDiffusion3Pipeline
|
86 |
+
|
87 |
|
88 |
model_id = "sd3-lora-test"
|
89 |
+
prompt = "a psychedelic man is surfing on top of a horse"
|
90 |
negative_prompt = "malformed, disgusting, overexposed, washed-out"
|
91 |
|
92 |
pipeline = DiffusionPipeline.from_pretrained(model_id)
|