Upload folder using huggingface_hub

11955a5 verified almost 2 years ago

14.3 kB

---
tags:
- stable-diffusion-xl
- stable-diffusion-xl-diffusers
- text-to-image
- diffusers
- lora
- template:sd-lora
widget:
- text: in the style of <s0><s1>, perlin noise terrain generation; The image is a
    map of a city with a repeating pattern of blue, green, and yellow colors. The
    blue and green colors are scattered across the map, creating a mosaic-like effect.
    The yellow and blue colors are larger and more prominent, with some areas being
    darker and others being lighter. The map appears to be overlaid on top of a larger
    map, with a white background. The overall color scheme of the map is predominantly
    blue, with hints of green and yellow.
  output:
    url: image-0.png
- text: in the style of <s0><s1>, procedural generation; The image is a 3D map of
    a small island in the middle of a body of water. The island is covered in green
    and brown vegetation, with patches of blue and green. The water is a deep blue-green
    color, and there are several small islands scattered throughout the island. In
    the center of the island, there is a large white area with a red dot in the center,
    which appears to be a small town or village. The red dot is likely a location
    marker or a marker for a specific location. The map is set against a dark blue
    background, making the colors of the vegetation stand out.
  output:
    url: image-1.png
- text: in the style of <s0><s1>, procedural generation; The image is an abstract
    digital art piece that appears to be made up of different colors and shapes. The
    background is divided into different shades of blue, green, orange, and yellow,
    creating a chaotic and dynamic composition. In the center of the image, there
    is a large, dark blue blob that is the focal point. The blob is outlined in a
    lighter shade of blue and has a pink dot in the middle. Surrounding the blob are
    various shapes and lines that create a sense of depth and dimension. The overall
    effect is one of movement and energy.
  output:
    url: image-2.png
- text: in the style of <s0><s1>, perlin noise terrain generation; The image is a
    map of the world, with different colors representing different regions. The colors
    are predominantly green, yellow, and blue, with some areas being darker and others
    being lighter. The map appears to be a topographic representation of the continents
    and oceans, with the majority of the colors being green and the majority being
    yellow.\n\nThe map is divided into different sections, with each section having
    a different color - yellow, orange, blue, and green. The yellow section is in
    the center of the map, with a darker shade of green on the left side and a lighter
    shade of blue on the right side. The blue section is on the top right corner,
    and the green section is at the bottom left corner. There are also several smaller
    blue areas scattered throughout the map. The overall color scheme of the image
    is warm and earthy, with shades of yellow, green, and orange.
  output:
    url: image-3.png
- text: in the style of <s0><s1>, perlin noise terrain generation; The image is a
    map of a small island in the middle of a blue ocean. The island is made up of
    different shades of green and pink, with a darker shade of green on the top and
    lighter shades of pink on the bottom. In the center of the island, there is a
    large white circle with a small white dot in the center. The circle is surrounded
    by smaller pink and green areas. On the right side of the image, there are two
    smaller pink areas, one on the left side and one in the bottom right corner. The
    background is a solid blue color.
  output:
    url: image-4.png
- text: in the style of <s0><s1>, perlin noise terrain generation; The image is a
    map of the island of Santorini, Greece. The map is color-coded, with different
    shades of green, beige, and brown representing different areas of the country.
    The green areas are darker in color, while the beige areas are lighter in color.
    The brown areas are scattered throughout the map, with some areas being darker
    and others being lighter.\n\nThe map is set against a blue background, with the
    ocean visible in the top left corner. There are also several smaller islands scattered
    throughout, including a small island in the center of the map. The islands are
    outlined in white, and there are a few smaller islands on the top right corner
    of the image. Overall, the map appears to be a detailed representation of the
    Mediterranean Sea, with various geographical features such as mountains, rivers,
    and islands visible.
  output:
    url: image-5.png
- text: in the style of <s0><s1>, perlin noise terrain generation; The image is a
    map of the Mediterranean Sea, which is a deep blue color. The map is color-coded,
    with different shades of green, orange, and yellow representing different areas
    of the sea. The green areas are densely packed together, while the orange areas
    are scattered throughout the map. The yellow areas are larger and more densely
    packed, with some areas appearing darker and others appearing lighter.\n\nThe
    map also shows the extent of the ocean floor, with the majority of the land covered
    in green and orange areas. There are also some areas that appear to be smaller
    and more flat, with a few smaller areas scattered throughout. The ocean floor
    is also colored in shades of blue and green, with hints of orange and yellow.
    The overall color scheme of the map is predominantly blue, green, and orange.
  output:
    url: image-6.png
- text: in the style of <s0><s1>, simplex noise terrain generation
  output:
    url: image-7.png
- text: 'in the style of <s0><s1>, simplex noise terrain generation; '
  output:
    url: image-8.png
- text: in the style of <s0><s1>, simplex noise terrain generation
  output:
    url: image-9.png
- text: in the style of <s0><s1>, simplex noise terrain generation
  output:
    url: image-10.png
- text: in the style of <s0><s1>, simplex noise terrain generation
  output:
    url: image-11.png
- text: in the style of <s0><s1>, voronoi noise terrain generation, landform
  output:
    url: image-12.png
- text: in the style of <s0><s1>, voronoi noise terrain generation, landform
  output:
    url: image-13.png
- text: in the style of <s0><s1>, voronoi noise terrain generation, abstract pattern
  output:
    url: image-14.png
- text: in the style of <s0><s1>, voronoi noise terrain generation, landform
  output:
    url: image-15.png
- text: in the style of <s0><s1>, voronoi noise terrain generation, abstract pattern
  output:
    url: image-16.png
- text: in the style of <s0><s1>, pixelated grid landform, cellular automata; The
    image is a pixelated map of the world. The map is made up of small squares of
    different colors, including blue, green, orange, and gray. The squares are arranged
    in a grid-like pattern, with some overlapping each other. In the center of the
    map, there is a large orange flower with a yellow center. The flower appears to
    be floating in the air, with its petals spread out. The background of the image
    is dark blue, making the colors of the flower stand out.
  output:
    url: image-17.png
- text: in the style of <s0><s1>, pixelated grid landform, cellular automata; The
    image is a pixelated map of a city or town. The map is made up of small squares
    of different colors, including green, blue, and gray. The squares are arranged
    in a grid-like pattern, with some overlapping each other. In the center of the
    map, there is a blue square with a white outline. The blue square is slightly
    larger than the green square, and it appears to be floating in the air. The gray
    squares are scattered throughout the map. The overall color scheme of the image
    is green, gray, and blue.
  output:
    url: image-18.png
- text: in the style of <s0><s1>, pixelated grid landform, cellular automata
  output:
    url: image-19.png
- text: in the style of <s0><s1>, simplex noise terrain generation; The image is a
    satellite view of a large body of water, possibly a lake or a river. The water
    is a deep blue color and appears to be calm and still. The surface of the water
    is covered in patches of green and yellow vegetation, with some areas appearing
    to be densely packed together. The vegetation is mostly green, with patches of
    yellow and brown scattered throughout. The image is taken from a top-down perspective,
    looking down on the water and the surrounding area. The overall color scheme of
    the image is predominantly green, blue, and yellow.
  output:
    url: image-20.png
- text: in the style of <s0><s1>, procedural generation
  output:
    url: image-21.png
- text: in the style of <s0><s1>, perlin noise terrain generation; The image is a
    map of the world in a camouflage pattern. The map is divided into different shades
    of blue, green, yellow, and brown, creating a camouflage-like effect. The colors
    are arranged in a way that creates a sense of depth and dimension, with the blue
    representing the ocean and the green representing the land. The brown represents
    the land, while the yellow represents the sea. The overall color scheme of the
    map is predominantly blue and green, with some areas being darker and others being
    lighter.
  output:
    url: image-22.png
- text: in the style of <s0><s1>, perlin noise terrain generation; The image is a
    seamless pattern with a camouflage-like design. The background is a combination
    of brown and yellow colors, with the majority of the colors being a darker shade
    of brown. The pattern is made up of small, irregularly shaped spots in various
    shades of yellow and purple. The spots are arranged in a random, overlapping manner,
    creating a sense of depth and dimension. The overall effect is a textured, abstract
    design with a mix of different shades and textures.
  output:
    url: image-23.png
- text: 'in the style of <s0><s1>, simplex noise terrain generation; '
  output:
    url: image-24.png
- text: in the style of <s0><s1>, simplex noise terrain generation
  output:
    url: image-25.png
- text: in the style of <s0><s1>, voronoi noise terrain generation, abstract pattern
  output:
    url: image-26.png
- text: in the style of <s0><s1>, simplex noise terrain generation; The image is a
    square with a green background and a white border. In the center of the square,
    there is a map of the world in a pixelated style. The map is divided into different
    shades of green, yellow, and blue, with some areas in gray and others in blue.
    The colors are arranged in a way that creates a mosaic-like effect, with the green
    being the dominant color and the yellow being the smallest. The blue areas are
    scattered throughout the map, creating a sense of depth and dimension. The overall
    color scheme of the image is predominantly green and yellow, with hints of blue
    and gray.'
  output:
    url: image-27.png
base_model: stabilityai/stable-diffusion-xl-base-1.0
instance_prompt: in the style of <s0><s1>
license: openrail++
---
# SDXL LoRA DreamBooth - chechiamah/proceduralgenerationforms

<Gallery />

## Model description

### These are chechiamah/proceduralgenerationforms LoRA adaption weights for stabilityai/stable-diffusion-xl-base-1.0.

## Download model

### Use it with UIs such as AUTOMATIC1111, Comfy UI, SD.Next, Invoke

- **LoRA**: download **[`proceduralgenerationforms.safetensors` here 💾](/chechiamah/proceduralgenerationforms/blob/main/proceduralgenerationforms.safetensors)**.
    - Place it on your `models/Lora` folder.
    - On AUTOMATIC1111, load the LoRA by adding `<lora:proceduralgenerationforms:1>` to your prompt. On ComfyUI just [load it as a regular LoRA](https://comfyanonymous.github.io/ComfyUI_examples/lora/).
- *Embeddings*: download **[`proceduralgenerationforms_emb.safetensors` here 💾](/chechiamah/proceduralgenerationforms/blob/main/proceduralgenerationforms_emb.safetensors)**.
    - Place it on it on your `embeddings` folder
    - Use it by adding `proceduralgenerationforms_emb` to your prompt. For example, `in the style of proceduralgenerationforms_emb`
    (you need both the LoRA and the embeddings as they were trained together for this LoRA)
    
## Use it with the [🧨 diffusers library](https://github.com/huggingface/diffusers)

```py
from diffusers import AutoPipelineForText2Image
import torch
from huggingface_hub import hf_hub_download
from safetensors.torch import load_file
        
pipeline = AutoPipelineForText2Image.from_pretrained('stabilityai/stable-diffusion-xl-base-1.0', torch_dtype=torch.float16).to('cuda')
pipeline.load_lora_weights('chechiamah/proceduralgenerationforms', weight_name='pytorch_lora_weights.safetensors')
embedding_path = hf_hub_download(repo_id='chechiamah/proceduralgenerationforms', filename='proceduralgenerationforms_emb.safetensors' repo_type="model")
state_dict = load_file(embedding_path)
pipeline.load_textual_inversion(state_dict["clip_l"], token=["<s0>", "<s1>"], text_encoder=pipeline.text_encoder, tokenizer=pipeline.tokenizer)
pipeline.load_textual_inversion(state_dict["clip_g"], token=["<s0>", "<s1>"], text_encoder=pipeline.text_encoder_2, tokenizer=pipeline.tokenizer_2)
        
image = pipeline('in the style of <s0><s1>').images[0]
```

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)

## Trigger words

To trigger image generation of trained concept(or concepts) replace each concept identifier in you prompt with the new inserted tokens:

to trigger concept `TOK` → use `<s0><s1>` in your prompt 



## Details
All [Files & versions](/chechiamah/proceduralgenerationforms/tree/main).

The weights were trained using [🧨 diffusers Advanced Dreambooth Training Script](https://github.com/huggingface/diffusers/blob/main/examples/advanced_diffusion_training/train_dreambooth_lora_sdxl_advanced.py).

LoRA for the text encoder was enabled. False.

Pivotal tuning was enabled: True.

Special VAE used for training: madebyollin/sdxl-vae-fp16-fix.

	---
	tags:
	- stable-diffusion-xl
	- stable-diffusion-xl-diffusers
	- text-to-image
	- diffusers
	- lora
	- template:sd-lora
	widget:
	- text: in the style of <s0><s1>, perlin noise terrain generation; The image is a
	map of a city with a repeating pattern of blue, green, and yellow colors. The
	blue and green colors are scattered across the map, creating a mosaic-like effect.
	The yellow and blue colors are larger and more prominent, with some areas being
	darker and others being lighter. The map appears to be overlaid on top of a larger
	map, with a white background. The overall color scheme of the map is predominantly
	blue, with hints of green and yellow.
	output:
	url: image-0.png
	- text: in the style of <s0><s1>, procedural generation; The image is a 3D map of
	a small island in the middle of a body of water. The island is covered in green
	and brown vegetation, with patches of blue and green. The water is a deep blue-green
	color, and there are several small islands scattered throughout the island. In
	the center of the island, there is a large white area with a red dot in the center,
	which appears to be a small town or village. The red dot is likely a location
	marker or a marker for a specific location. The map is set against a dark blue
	background, making the colors of the vegetation stand out.
	output:
	url: image-1.png
	- text: in the style of <s0><s1>, procedural generation; The image is an abstract
	digital art piece that appears to be made up of different colors and shapes. The
	background is divided into different shades of blue, green, orange, and yellow,
	creating a chaotic and dynamic composition. In the center of the image, there
	is a large, dark blue blob that is the focal point. The blob is outlined in a
	lighter shade of blue and has a pink dot in the middle. Surrounding the blob are
	various shapes and lines that create a sense of depth and dimension. The overall
	effect is one of movement and energy.
	output:
	url: image-2.png
	- text: in the style of <s0><s1>, perlin noise terrain generation; The image is a
	map of the world, with different colors representing different regions. The colors
	are predominantly green, yellow, and blue, with some areas being darker and others
	being lighter. The map appears to be a topographic representation of the continents
	and oceans, with the majority of the colors being green and the majority being
	yellow.\n\nThe map is divided into different sections, with each section having
	a different color - yellow, orange, blue, and green. The yellow section is in
	the center of the map, with a darker shade of green on the left side and a lighter
	shade of blue on the right side. The blue section is on the top right corner,
	and the green section is at the bottom left corner. There are also several smaller
	blue areas scattered throughout the map. The overall color scheme of the image
	is warm and earthy, with shades of yellow, green, and orange.
	output:
	url: image-3.png
	- text: in the style of <s0><s1>, perlin noise terrain generation; The image is a
	map of a small island in the middle of a blue ocean. The island is made up of
	different shades of green and pink, with a darker shade of green on the top and
	lighter shades of pink on the bottom. In the center of the island, there is a
	large white circle with a small white dot in the center. The circle is surrounded
	by smaller pink and green areas. On the right side of the image, there are two
	smaller pink areas, one on the left side and one in the bottom right corner. The
	background is a solid blue color.
	output:
	url: image-4.png
	- text: in the style of <s0><s1>, perlin noise terrain generation; The image is a
	map of the island of Santorini, Greece. The map is color-coded, with different
	shades of green, beige, and brown representing different areas of the country.
	The green areas are darker in color, while the beige areas are lighter in color.
	The brown areas are scattered throughout the map, with some areas being darker
	and others being lighter.\n\nThe map is set against a blue background, with the
	ocean visible in the top left corner. There are also several smaller islands scattered
	throughout, including a small island in the center of the map. The islands are
	outlined in white, and there are a few smaller islands on the top right corner
	of the image. Overall, the map appears to be a detailed representation of the
	Mediterranean Sea, with various geographical features such as mountains, rivers,
	and islands visible.
	output:
	url: image-5.png
	- text: in the style of <s0><s1>, perlin noise terrain generation; The image is a
	map of the Mediterranean Sea, which is a deep blue color. The map is color-coded,
	with different shades of green, orange, and yellow representing different areas
	of the sea. The green areas are densely packed together, while the orange areas
	are scattered throughout the map. The yellow areas are larger and more densely
	packed, with some areas appearing darker and others appearing lighter.\n\nThe
	map also shows the extent of the ocean floor, with the majority of the land covered
	in green and orange areas. There are also some areas that appear to be smaller
	and more flat, with a few smaller areas scattered throughout. The ocean floor
	is also colored in shades of blue and green, with hints of orange and yellow.
	The overall color scheme of the map is predominantly blue, green, and orange.
	output:
	url: image-6.png
	- text: in the style of <s0><s1>, simplex noise terrain generation
	output:
	url: image-7.png
	- text: 'in the style of <s0><s1>, simplex noise terrain generation; '
	output:
	url: image-8.png
	- text: in the style of <s0><s1>, simplex noise terrain generation
	output:
	url: image-9.png
	- text: in the style of <s0><s1>, simplex noise terrain generation
	output:
	url: image-10.png
	- text: in the style of <s0><s1>, simplex noise terrain generation
	output:
	url: image-11.png
	- text: in the style of <s0><s1>, voronoi noise terrain generation, landform
	output:
	url: image-12.png
	- text: in the style of <s0><s1>, voronoi noise terrain generation, landform
	output:
	url: image-13.png
	- text: in the style of <s0><s1>, voronoi noise terrain generation, abstract pattern
	output:
	url: image-14.png
	- text: in the style of <s0><s1>, voronoi noise terrain generation, landform
	output:
	url: image-15.png
	- text: in the style of <s0><s1>, voronoi noise terrain generation, abstract pattern
	output:
	url: image-16.png
	- text: in the style of <s0><s1>, pixelated grid landform, cellular automata; The
	image is a pixelated map of the world. The map is made up of small squares of
	different colors, including blue, green, orange, and gray. The squares are arranged
	in a grid-like pattern, with some overlapping each other. In the center of the
	map, there is a large orange flower with a yellow center. The flower appears to
	be floating in the air, with its petals spread out. The background of the image
	is dark blue, making the colors of the flower stand out.
	output:
	url: image-17.png
	- text: in the style of <s0><s1>, pixelated grid landform, cellular automata; The
	image is a pixelated map of a city or town. The map is made up of small squares
	of different colors, including green, blue, and gray. The squares are arranged
	in a grid-like pattern, with some overlapping each other. In the center of the
	map, there is a blue square with a white outline. The blue square is slightly
	larger than the green square, and it appears to be floating in the air. The gray
	squares are scattered throughout the map. The overall color scheme of the image
	is green, gray, and blue.
	output:
	url: image-18.png
	- text: in the style of <s0><s1>, pixelated grid landform, cellular automata
	output:
	url: image-19.png
	- text: in the style of <s0><s1>, simplex noise terrain generation; The image is a
	satellite view of a large body of water, possibly a lake or a river. The water
	is a deep blue color and appears to be calm and still. The surface of the water
	is covered in patches of green and yellow vegetation, with some areas appearing
	to be densely packed together. The vegetation is mostly green, with patches of
	yellow and brown scattered throughout. The image is taken from a top-down perspective,
	looking down on the water and the surrounding area. The overall color scheme of
	the image is predominantly green, blue, and yellow.
	output:
	url: image-20.png
	- text: in the style of <s0><s1>, procedural generation
	output:
	url: image-21.png
	- text: in the style of <s0><s1>, perlin noise terrain generation; The image is a
	map of the world in a camouflage pattern. The map is divided into different shades
	of blue, green, yellow, and brown, creating a camouflage-like effect. The colors
	are arranged in a way that creates a sense of depth and dimension, with the blue
	representing the ocean and the green representing the land. The brown represents
	the land, while the yellow represents the sea. The overall color scheme of the
	map is predominantly blue and green, with some areas being darker and others being
	lighter.
	output:
	url: image-22.png
	- text: in the style of <s0><s1>, perlin noise terrain generation; The image is a
	seamless pattern with a camouflage-like design. The background is a combination
	of brown and yellow colors, with the majority of the colors being a darker shade
	of brown. The pattern is made up of small, irregularly shaped spots in various
	shades of yellow and purple. The spots are arranged in a random, overlapping manner,
	creating a sense of depth and dimension. The overall effect is a textured, abstract
	design with a mix of different shades and textures.
	output:
	url: image-23.png
	- text: 'in the style of <s0><s1>, simplex noise terrain generation; '
	output:
	url: image-24.png
	- text: in the style of <s0><s1>, simplex noise terrain generation
	output:
	url: image-25.png
	- text: in the style of <s0><s1>, voronoi noise terrain generation, abstract pattern
	output:
	url: image-26.png
	- text: in the style of <s0><s1>, simplex noise terrain generation; The image is a
	square with a green background and a white border. In the center of the square,
	there is a map of the world in a pixelated style. The map is divided into different
	shades of green, yellow, and blue, with some areas in gray and others in blue.
	The colors are arranged in a way that creates a mosaic-like effect, with the green
	being the dominant color and the yellow being the smallest. The blue areas are
	scattered throughout the map, creating a sense of depth and dimension. The overall
	color scheme of the image is predominantly green and yellow, with hints of blue
	and gray.'
	output:
	url: image-27.png
	base_model: stabilityai/stable-diffusion-xl-base-1.0
	instance_prompt: in the style of <s0><s1>
	license: openrail++
	---

	# SDXL LoRA DreamBooth - chechiamah/proceduralgenerationforms

	<Gallery />

	## Model description

	### These are chechiamah/proceduralgenerationforms LoRA adaption weights for stabilityai/stable-diffusion-xl-base-1.0.

	## Download model

	### Use it with UIs such as AUTOMATIC1111, Comfy UI, SD.Next, Invoke

	- LoRA: download [`proceduralgenerationforms.safetensors` here 💾](/chechiamah/proceduralgenerationforms/blob/main/proceduralgenerationforms.safetensors).
	- Place it on your `models/Lora` folder.
	- On AUTOMATIC1111, load the LoRA by adding `<lora:proceduralgenerationforms:1>` to your prompt. On ComfyUI just [load it as a regular LoRA](https://comfyanonymous.github.io/ComfyUI_examples/lora/).
	- Embeddings: download [`proceduralgenerationforms_emb.safetensors` here 💾](/chechiamah/proceduralgenerationforms/blob/main/proceduralgenerationforms_emb.safetensors).
	- Place it on it on your `embeddings` folder
	- Use it by adding `proceduralgenerationforms_emb` to your prompt. For example, `in the style of proceduralgenerationforms_emb`
	(you need both the LoRA and the embeddings as they were trained together for this LoRA)


	## Use it with the [🧨 diffusers library](https://github.com/huggingface/diffusers)

	```py
	from diffusers import AutoPipelineForText2Image
	import torch
	from huggingface_hub import hf_hub_download
	from safetensors.torch import load_file

	pipeline = AutoPipelineForText2Image.from_pretrained('stabilityai/stable-diffusion-xl-base-1.0', torch_dtype=torch.float16).to('cuda')
	pipeline.load_lora_weights('chechiamah/proceduralgenerationforms', weight_name='pytorch_lora_weights.safetensors')
	embedding_path = hf_hub_download(repo_id='chechiamah/proceduralgenerationforms', filename='proceduralgenerationforms_emb.safetensors' repo_type="model")
	state_dict = load_file(embedding_path)
	pipeline.load_textual_inversion(state_dict["clip_l"], token=["<s0>", "<s1>"], text_encoder=pipeline.text_encoder, tokenizer=pipeline.tokenizer)
	pipeline.load_textual_inversion(state_dict["clip_g"], token=["<s0>", "<s1>"], text_encoder=pipeline.text_encoder_2, tokenizer=pipeline.tokenizer_2)

	image = pipeline('in the style of <s0><s1>').images[0]
	```

	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)

	## Trigger words

	To trigger image generation of trained concept(or concepts) replace each concept identifier in you prompt with the new inserted tokens:

	to trigger concept `TOK` → use `<s0><s1>` in your prompt



	## Details
	All [Files & versions](/chechiamah/proceduralgenerationforms/tree/main).

	The weights were trained using [🧨 diffusers Advanced Dreambooth Training Script](https://github.com/huggingface/diffusers/blob/main/examples/advanced_diffusion_training/train_dreambooth_lora_sdxl_advanced.py).

	LoRA for the text encoder was enabled. False.

	Pivotal tuning was enabled: True.

	Special VAE used for training: madebyollin/sdxl-vae-fp16-fix.