Skip to content

Commit b953f82

Browse files
lsteinlstein
authored
Merge branch 'development' into fix-doc-typos
2 parents a6e28d2 + ef20588 commit b953f82

38 files changed

+2011
-634
lines changed

‎.github/CODEOWNERS

+4
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,4 @@
1+
ldm/invoke/pngwriter.py @CapableWeb
2+
ldm/invoke/server_legacy.py @CapableWeb
3+
scripts/legacy_api.py @CapableWeb
4+
tests/legacy_tests.sh @CapableWeb

‎.github/workflows/mkdocs-flow.yml

+3-3
Original file line numberDiff line numberDiff line change
@@ -3,9 +3,9 @@ on:
33
push:
44
branches:
55
- main
6-
pull_request:
7-
branches:
8-
- main
6+
# pull_request:
7+
# branches:
8+
# - main
99
jobs:
1010
build:
1111
name: Deploy docs to GitHub Pages

‎LICENSE

+1-1
Original file line numberDiff line numberDiff line change
@@ -1,6 +1,6 @@
11
MIT License
22

3-
Copyright (c) 2022 Lincoln D. Stein (https://github.com/lstein)
3+
Copyright (c) 2022 Lincoln Stein and InvokeAI Organization
44

55
This software is derived from a fork of the source code available from
66
https://github.com/pesser/stable-diffusion and

‎README.md

+38-41
Original file line numberDiff line numberDiff line change
@@ -2,14 +2,7 @@
22

33
# InvokeAI: A Stable Diffusion Toolkit
44

5-
_Note: This fork is rapidly evolving. Please use the
6-
[Issues](https://github.com/invoke-ai/InvokeAI/issues) tab to
7-
report bugs and make feature requests. Be sure to use the provided
8-
templates. They will help aid diagnose issues faster._
9-
10-
_This repository was formally known as lstein/stable-diffusion_
11-
12-
# **Table of Contents**
5+
_Formally known as lstein/stable-diffusion_
136

147
![project logo](docs/assets/logo.png)
158

@@ -46,8 +39,13 @@ This is a fork of
4639
the open source text-to-image generator. It provides a streamlined
4740
process with various new features and options to aid the image
4841
generation process. It runs on Windows, Mac and Linux machines, with
49-
GPU cards with as little as 4 GB or RAM. It provides both a polished
50-
Web interface, and an easy-to-use command-line interface.
42+
GPU cards with as little as 4 GB of RAM. It provides both a polished
43+
Web interface (see below), and an easy-to-use command-line interface.
44+
45+
**Quick links**: [<a href="https://discord.gg/NwVCmKwY">Discord Server</a>] [<a href="https://invoke-ai.github.io/InvokeAI/">Documentation and Tutorials</a>] [<a href="https://github.com/invoke-ai/InvokeAI/">Code and Downloads</a>] [<a href="https://github.com/invoke-ai/InvokeAI/issues">Bug Reports</a>] [<a href="https://github.com/invoke-ai/InvokeAI/discussions">Discussion, Ideas & Q&A</a>]
46+
47+
<div align="center"><img src="docs/assets/invoke-web-server-1.png" width=640></div>
48+
5149

5250
_Note: This fork is rapidly evolving. Please use the
5351
[Issues](https://github.com/invoke-ai/InvokeAI/issues) tab to report bugs and make feature
@@ -91,7 +89,7 @@ You wil need one of the following:
9189

9290
#### Disk
9391

94-
- At least 6 GB of free disk space for the machine learning model, Python, and all its dependencies.
92+
- At least 12 GB of free disk space for the machine learning model, Python, and all its dependencies.
9593

9694
**Note**
9795

@@ -136,39 +134,38 @@ you can try starting `invoke.py` with the `--precision=float32` flag:
136134

137135
### Latest Changes
138136

139-
- vNEXT (TODO 2022)
140-
137+
- v2.0.1 (13 October 2022)
138+
- fix noisy images at high step count when using k* samplers
139+
- dream.py script now calls invoke.py module directly rather than
140+
via a new python process (which could break the environment)
141+
142+
- v2.0.0 (9 October 2022)
143+
144+
- `dream.py` script renamed `invoke.py`. A `dream.py` script wrapper remains
145+
for backward compatibility.
146+
- Completely new WebGUI - launch with `python3 scripts/invoke.py --web`
147+
- Support for <a href="https://github.com/invoke-ai/InvokeAI/blob/main/docs/features/INPAINTING.md">inpainting</a> and <a href="https://github.com/invoke-ai/InvokeAI/blob/main/docs/features/OUTPAINTING.md">outpainting</a>
148+
- img2img runs on all k* samplers
149+
- Support for <a href="https://github.com/invoke-ai/InvokeAI/blob/main/docs/features/PROMPTS.md#negative-and-unconditioned-prompts">negative prompts</a>
150+
- Support for CodeFormer face reconstruction
151+
- Support for Textual Inversion on Macintoshes
152+
- Support in both WebGUI and CLI for <a href="https://github.com/invoke-ai/InvokeAI/blob/main/docs/features/POSTPROCESS.md">post-processing of previously-generated images</a>
153+
using facial reconstruction, ESRGAN upscaling, outcropping (similar to DALL-E infinite canvas),
154+
and "embiggen" upscaling. See the `!fix` command.
155+
- New `--hires` option on `invoke>` line allows <a href="https://github.com/invoke-ai/InvokeAI/blob/main/docs/features/CLI.md#this-is-an-example-of-txt2img">larger images to be created without duplicating elements</a>, at the cost of some performance.
156+
- New `--perlin` and `--threshold` options allow you to add and control variation
157+
during image generation (see <a href="https://github.com/invoke-ai/InvokeAI/blob/main/docs/features/OTHER.md#thresholding-and-perlin-noise-initialization-options">Thresholding and Perlin Noise Initialization</a>
158+
- Extensive metadata now written into PNG files, allowing reliable regeneration of images
159+
and tweaking of previous settings.
160+
- Command-line completion in `invoke.py` now works on Windows, Linux and Mac platforms.
161+
- Improved <a href="https://github.com/invoke-ai/InvokeAI/blob/main/docs/features/CLI.md">command-line completion behavior</a>.
162+
New commands added:
163+
* List command-line history with `!history`
164+
* Search command-line history with `!search`
165+
* Clear history with `!clear`
141166
- Deprecated `--full_precision` / `-F`. Simply omit it and `invoke.py` will auto
142167
configure. To switch away from auto use the new flag like `--precision=float32`.
143168

144-
- v1.14 (11 September 2022)
145-
146-
- Memory optimizations for small-RAM cards. 512x512 now possible on 4 GB GPUs.
147-
- Full support for Apple hardware with M1 or M2 chips.
148-
- Add "seamless mode" for circular tiling of image. Generates beautiful effects.
149-
([prixt](https://github.com/prixt)).
150-
- Inpainting support.
151-
- Improved web server GUI.
152-
- Lots of code and documentation cleanups.
153-
154-
- v1.13 (3 September 2022
155-
156-
- Support image variations (see [VARIATIONS](docs/features/VARIATIONS.md)
157-
([Kevin Gibbons](https://github.com/bakkot) and many contributors and reviewers)
158-
- Supports a Google Colab notebook for a standalone server running on Google hardware
159-
[Arturo Mendivil](https://github.com/artmen1516)
160-
- WebUI supports GFPGAN/ESRGAN facial reconstruction and upscaling
161-
[Kevin Gibbons](https://github.com/bakkot)
162-
- WebUI supports incremental display of in-progress images during generation
163-
[Kevin Gibbons](https://github.com/bakkot)
164-
- A new configuration file scheme that allows new models (including upcoming
165-
stable-diffusion-v1.5) to be added without altering the code.
166-
([David Wager](https://github.com/maddavid12))
167-
- Can specify --grid on invoke.py command line as the default.
168-
- Miscellaneous internal bug and stability fixes.
169-
- Works on M1 Apple hardware.
170-
- Multiple bug fixes.
171-
172169
For older changelogs, please visit the **[CHANGELOG](docs/features/CHANGELOG.md)**.
173170

174171
### Troubleshooting

‎docs/CHANGELOG.md

+78-23
Original file line numberDiff line numberDiff line change
@@ -1,18 +1,73 @@
1-
# **Changelog**
2-
3-
## v1.13 (in process)
1+
---
2+
title: Changelog
3+
---
44

5-
- Supports a Google Colab notebook for a standalone server running on Google hardware [Arturo Mendivil](https://github.com/artmen1516)
6-
- WebUI supports GFPGAN/ESRGAN facial reconstruction and upscaling [Kevin Gibbons](https://github.com/bakkot)
7-
- WebUI supports incremental display of in-progress images during generation [Kevin Gibbons](https://github.com/bakkot)
8-
- Output directory can be specified on the invoke> command line.
9-
- The grid was displaying duplicated images when not enough images to fill the final row [Muhammad Usama](https://github.com/SMUsamaShah)
10-
- Can specify --grid on invoke.py command line as the default.
11-
- Miscellaneous internal bug and stability fixes.
5+
# :octicons-log-16: **Changelog**
6+
7+
## v2.0.1 (13 October 2022)
8+
9+
- fix noisy images at high step count when using k* samplers
10+
- dream.py script now calls invoke.py module directly rather than
11+
via a new python process (which could break the environment)
12+
13+
## v2.0.0 <small>(9 October 2022)</small>
14+
15+
- `dream.py` script renamed `invoke.py`. A `dream.py` script wrapper remains
16+
for backward compatibility.
17+
- Completely new WebGUI - launch with `python3 scripts/invoke.py --web`
18+
- Support for <a href="https://github.com/invoke-ai/InvokeAI/blob/main/docs/features/INPAINTING.md">inpainting</a> and <a href="https://github.com/invoke-ai/InvokeAI/blob/main/docs/features/OUTPAINTING.md">outpainting</a>
19+
- img2img runs on all k* samplers
20+
- Support for <a href="https://github.com/invoke-ai/InvokeAI/blob/main/docs/features/PROMPTS.md#negative-and-unconditioned-prompts">negative prompts</a>
21+
- Support for CodeFormer face reconstruction
22+
- Support for Textual Inversion on Macintoshes
23+
- Support in both WebGUI and CLI for <a href="https://github.com/invoke-ai/InvokeAI/blob/main/docs/features/POSTPROCESS.md">post-processing of previously-generated images</a>
24+
using facial reconstruction, ESRGAN upscaling, outcropping (similar to DALL-E infinite canvas),
25+
and "embiggen" upscaling. See the `!fix` command.
26+
- New `--hires` option on `invoke>` line allows <a href="https://github.com/invoke-ai/InvokeAI/blob/main/docs/features/CLI.m#this-is-an-example-of-txt2img">larger images to be created without duplicating elements</a>, at the cost of some performance.
27+
- New `--perlin` and `--threshold` options allow you to add and control variation
28+
during image generation (see <a href="https://github.com/invoke-ai/InvokeAI/blob/main/docs/features/OTHER.md#thresholding-and-perlin-noise-initialization-options">Thresholding and Perlin Noise Initialization</a>
29+
- Extensive metadata now written into PNG files, allowing reliable regeneration of images
30+
and tweaking of previous settings.
31+
- Command-line completion in `invoke.py` now works on Windows, Linux and Mac platforms.
32+
- Improved <a href="https://github.com/invoke-ai/InvokeAI/blob/main/docs/features/CLI.m">command-line completion behavior</a>.
33+
New commands added:
34+
* List command-line history with `!history`
35+
* Search command-line history with `!search`
36+
* Clear history with `!clear`
37+
- Deprecated `--full_precision` / `-F`. Simply omit it and `invoke.py` will auto
38+
configure. To switch away from auto use the new flag like `--precision=float32`.
39+
40+
## v1.14 <small>(11 September 2022)</small>
41+
42+
- Memory optimizations for small-RAM cards. 512x512 now possible on 4 GB GPUs.
43+
- Full support for Apple hardware with M1 or M2 chips.
44+
- Add "seamless mode" for circular tiling of image. Generates beautiful effects.
45+
([prixt](https://github.com/prixt)).
46+
- Inpainting support.
47+
- Improved web server GUI.
48+
- Lots of code and documentation cleanups.
49+
50+
## v1.13 <small>(3 September 2022)</small>
51+
52+
- Support image variations (see [VARIATIONS](features/VARIATIONS.md)
53+
([Kevin Gibbons](https://github.com/bakkot) and many contributors and reviewers)
54+
- Supports a Google Colab notebook for a standalone server running on Google hardware
55+
[Arturo Mendivil](https://github.com/artmen1516)
56+
- WebUI supports GFPGAN/ESRGAN facial reconstruction and upscaling
57+
[Kevin Gibbons](https://github.com/bakkot)
58+
- WebUI supports incremental display of in-progress images during generation
59+
[Kevin Gibbons](https://github.com/bakkot)
60+
- A new configuration file scheme that allows new models (including upcoming
61+
stable-diffusion-v1.5) to be added without altering the code.
62+
([David Wager](https://github.com/maddavid12))
63+
- Can specify --grid on invoke.py command line as the default.
64+
- Miscellaneous internal bug and stability fixes.
65+
- Works on M1 Apple hardware.
66+
- Multiple bug fixes.
1267

1368
---
1469

15-
## v1.12 (28 August 2022)
70+
## v1.12 <small>(28 August 2022)</small>
1671

1772
- Improved file handling, including ability to read prompts from standard input.
1873
(kudos to [Yunsaki](https://github.com/yunsaki)
@@ -26,7 +81,7 @@
2681

2782
---
2883

29-
## v1.11 (26 August 2022)
84+
## v1.11 <small>(26 August 2022)</small>
3085

3186
- NEW FEATURE: Support upscaling and face enhancement using the GFPGAN module. (kudos to [Oceanswave](https://github.com/Oceanswave)
3287
- You now can specify a seed of -1 to use the previous image's seed, -2 to use the seed for the image generated before that, etc.
@@ -39,13 +94,13 @@
3994

4095
---
4196

42-
## v1.10 (25 August 2022)
97+
## v1.10 <small>(25 August 2022)</small>
4398

4499
- A barebones but fully functional interactive web server for online generation of txt2img and img2img.
45100

46101
---
47102

48-
## v1.09 (24 August 2022)
103+
## v1.09 <small>(24 August 2022)</small>
49104

50105
- A new -v option allows you to generate multiple variants of an initial image
51106
in img2img mode. (kudos to [Oceanswave](https://github.com/Oceanswave). [
@@ -55,7 +110,7 @@
55110

56111
---
57112

58-
## v1.08 (24 August 2022)
113+
## v1.08 <small>(24 August 2022)</small>
59114

60115
- Escape single quotes on the invoke> command before trying to parse. This avoids
61116
parse errors.
@@ -66,22 +121,22 @@
66121

67122
---
68123

69-
## v1.07 (23 August 2022)
124+
## v1.07 <small>(23 August 2022)</small>
70125

71126
- Image filenames will now never fill gaps in the sequence, but will be assigned the
72127
next higher name in the chosen directory. This ensures that the alphabetic and chronological
73128
sort orders are the same.
74129

75130
---
76131

77-
## v1.06 (23 August 2022)
132+
## v1.06 <small>(23 August 2022)</small>
78133

79134
- Added weighted prompt support contributed by [xraxra](https://github.com/xraxra)
80135
- Example of using weighted prompts to tweak a demonic figure contributed by [bmaltais](https://github.com/bmaltais)
81136

82137
---
83138

84-
## v1.05 (22 August 2022 - after the drop)
139+
## v1.05 <small>(22 August 2022 - after the drop)</small>
85140

86141
- Filenames now use the following formats:
87142
000010.95183149.png -- Two files produced by the same command (e.g. -n2),
@@ -99,22 +154,22 @@
99154

100155
---
101156

102-
## v1.04 (22 August 2022 - after the drop)
157+
## v1.04 <small>(22 August 2022 - after the drop)</small>
103158

104159
- Updated README to reflect installation of the released weights.
105160
- Suppressed very noisy and inconsequential warning when loading the frozen CLIP
106161
tokenizer.
107162

108163
---
109164

110-
## v1.03 (22 August 2022)
165+
## v1.03 <small>(22 August 2022)</small>
111166

112167
- The original txt2img and img2img scripts from the CompViz repository have been moved into
113168
a subfolder named "orig_scripts", to reduce confusion.
114169

115170
---
116171

117-
## v1.02 (21 August 2022)
172+
## v1.02 <small>(21 August 2022)</small>
118173

119174
- A copy of the prompt and all of its switches and options is now stored in the corresponding
120175
image in a tEXt metadata field named "Dream". You can read the prompt using scripts/images2prompt.py,
@@ -123,7 +178,7 @@
123178

124179
---
125180

126-
## v1.01 (21 August 2022)
181+
## v1.01 <small>(21 August 2022)</small>
127182

128183
- added k_lms sampling.
129184
**Please run "conda env update" to load the k_lms dependencies!!**
@@ -134,4 +189,4 @@
134189

135190
## Links
136191

137-
- **[Read Me](../readme.md)**
192+
- **[Read Me](index.md)**

‎docs/assets/invoke-web-server-9.png

1.08 MB
Loading

0 commit comments

Comments
 (0)