Model card auto-generated by SimpleTuner
Browse files
README.md
CHANGED
@@ -18,7 +18,7 @@ widget:
|
|
18 |
negative_prompt: 'blurry, cropped, ugly'
|
19 |
output:
|
20 |
url: ./assets/image_0_0.png
|
21 |
-
- text: 'A
|
22 |
parameters:
|
23 |
negative_prompt: 'blurry, cropped, ugly'
|
24 |
output:
|
@@ -32,7 +32,7 @@ This is a LyCORIS adapter derived from [black-forest-labs/FLUX.1-dev](https://hu
|
|
32 |
|
33 |
The main validation prompt used during training was:
|
34 |
```
|
35 |
-
A
|
36 |
```
|
37 |
|
38 |
|
@@ -58,10 +58,10 @@ You may reuse the base model text encoder for inference.
|
|
58 |
|
59 |
## Training settings
|
60 |
|
61 |
-
- Training epochs:
|
62 |
-
- Training steps:
|
63 |
- Learning rate: 0.0001
|
64 |
-
- Learning rate schedule:
|
65 |
- Warmup steps: 100
|
66 |
- Max grad norm: 2.0
|
67 |
- Effective batch size: 1
|
@@ -146,7 +146,7 @@ lora_scale = 1.0
|
|
146 |
wrapper, _ = create_lycoris_from_weights(lora_scale, adapter_file_path, pipeline.transformer)
|
147 |
wrapper.merge_to()
|
148 |
|
149 |
-
prompt = "A
|
150 |
|
151 |
|
152 |
## Optional: quantise the model to save on vram.
|
|
|
18 |
negative_prompt: 'blurry, cropped, ugly'
|
19 |
output:
|
20 |
url: ./assets/image_0_0.png
|
21 |
+
- text: 'A slender, clear glass tumbler brimming with a beautifully crafted latte, showcasing a delightful gradient. The top is adorned with velvety foam that appears almost cloud-like, while beneath it lies a warm, inviting caramel hue that hints at the rich espresso blended with steamed milk. The glass rests elegantly on a crisp white plate, accompanied by a vibrant yellow napkin elegantly folded beneath.'
|
22 |
parameters:
|
23 |
negative_prompt: 'blurry, cropped, ugly'
|
24 |
output:
|
|
|
32 |
|
33 |
The main validation prompt used during training was:
|
34 |
```
|
35 |
+
A slender, clear glass tumbler brimming with a beautifully crafted latte, showcasing a delightful gradient. The top is adorned with velvety foam that appears almost cloud-like, while beneath it lies a warm, inviting caramel hue that hints at the rich espresso blended with steamed milk. The glass rests elegantly on a crisp white plate, accompanied by a vibrant yellow napkin elegantly folded beneath.
|
36 |
```
|
37 |
|
38 |
|
|
|
58 |
|
59 |
## Training settings
|
60 |
|
61 |
+
- Training epochs: 249
|
62 |
+
- Training steps: 500
|
63 |
- Learning rate: 0.0001
|
64 |
+
- Learning rate schedule: constant
|
65 |
- Warmup steps: 100
|
66 |
- Max grad norm: 2.0
|
67 |
- Effective batch size: 1
|
|
|
146 |
wrapper, _ = create_lycoris_from_weights(lora_scale, adapter_file_path, pipeline.transformer)
|
147 |
wrapper.merge_to()
|
148 |
|
149 |
+
prompt = "A slender, clear glass tumbler brimming with a beautifully crafted latte, showcasing a delightful gradient. The top is adorned with velvety foam that appears almost cloud-like, while beneath it lies a warm, inviting caramel hue that hints at the rich espresso blended with steamed milk. The glass rests elegantly on a crisp white plate, accompanied by a vibrant yellow napkin elegantly folded beneath."
|
150 |
|
151 |
|
152 |
## Optional: quantise the model to save on vram.
|