From 03ee67bfd34b9e872b33eb05fef5db83410b16f3 Mon Sep 17 00:00:00 2001 From: WDevelopsWebApps <97454358+WDevelopsWebApps@users.noreply.github.com> Date: Wed, 28 Sep 2022 10:53:40 +0200 Subject: add advanced saving for save button --- modules/ui.py | 37 +++++++++++++++++++++++++++++-------- 1 file changed, 29 insertions(+), 8 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/ui.py b/modules/ui.py index 7db8edbd..87a86a45 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -28,6 +28,7 @@ import modules.gfpgan_model import modules.codeformer_model import modules.styles import modules.generation_parameters_copypaste +from modules.images import apply_filename_pattern, get_next_sequence_number # this is a fix for Windows users. Without it, javascript files will be served with text/html content-type and the bowser will not show any UI mimetypes.init() @@ -90,13 +91,26 @@ def send_gradio_gallery_to_image(x): def save_files(js_data, images, index): - import csv - - os.makedirs(opts.outdir_save, exist_ok=True) - + import csv filenames = [] + #quick dictionary to class object conversion. Its neccesary due apply_filename_pattern requiring it + class MyObject: + def __init__(self, d=None): + if d is not None: + for key, value in d.items(): + setattr(self, key, value) + data = json.loads(js_data) + p = MyObject(data) + path = opts.outdir_save + save_to_dirs = opts.save_to_dirs + + if save_to_dirs: + dirname = apply_filename_pattern(opts.directories_filename_pattern or "[prompt_words]", p, p.seed, p.prompt) + path = os.path.join(opts.outdir_save, dirname) + + os.makedirs(path, exist_ok=True) if index > -1 and opts.save_selected_only and (index > 0 or not opts.return_grid): # ensures we are looking at a specific non-grid picture, and we have save_selected_only images = [images[index]] @@ -107,11 +121,18 @@ def save_files(js_data, images, index): writer = csv.writer(file) if at_start: writer.writerow(["prompt", "seed", "width", "height", "sampler", "cfgs", "steps", "filename", "negative_prompt"]) - - filename_base = str(int(time.time() * 1000)) + file_decoration = opts.samples_filename_pattern or "[seed]-[prompt_spaces]" + if file_decoration != "": + file_decoration = "-" + file_decoration.lower() + file_decoration = apply_filename_pattern(file_decoration, p, p.seed, p.prompt) + truncated = (file_decoration[:240] + '..') if len(file_decoration) > 240 else file_decoration + filename_base = truncated + + basecount = get_next_sequence_number(path, "") for i, filedata in enumerate(images): - filename = filename_base + ("" if len(images) == 1 else "-" + str(i + 1)) + ".png" - filepath = os.path.join(opts.outdir_save, filename) + file_number = f"{basecount+i:05}" + filename = file_number + filename_base + ".png" + filepath = os.path.join(path, filename) if filedata.startswith("data:image/png;base64,"): filedata = filedata[len("data:image/png;base64,"):] -- cgit v1.2.3 From abdbf1de646f007b6d76cfb3f416fdfaadb57903 Mon Sep 17 00:00:00 2001 From: Liam Date: Thu, 29 Sep 2022 14:40:47 -0400 Subject: token counters now update when roll artist and style buttons are pressed https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/1194#issuecomment-1261203893 --- javascript/ui.js | 28 ++++++++++++++++++++++------ modules/ui.py | 6 +++++- 2 files changed, 27 insertions(+), 7 deletions(-) (limited to 'modules/ui.py') diff --git a/javascript/ui.js b/javascript/ui.js index bfe02410..88fd45ae 100644 --- a/javascript/ui.js +++ b/javascript/ui.js @@ -199,12 +199,21 @@ let txt2img_textarea, img2img_textarea = undefined; let wait_time = 800 let token_timeout; -function submit_prompt(event, generate_button_id) { - if (event.altKey && event.keyCode === 13) { - event.preventDefault(); - gradioApp().getElementById(generate_button_id).click(); - return; - } +function roll_artist_txt2img(prompt_text) { + update_token_counter("txt2img_token_button") + return prompt_text; +} +function roll_artist_img2img(prompt_text) { + update_token_counter("img2img_token_button") + return prompt_text; +} +function update_style_txt2img(prompt_text, negative_prompt, style1, style2) { + update_token_counter("txt2img_token_button") + return [prompt_text, negative_prompt, style1, style2] +} +function update_style_img2img(prompt_text, negative_prompt, style1, style2) { + update_token_counter("img2img_token_button") + return [prompt_text, negative_prompt, style1, style2] } function update_token_counter(button_id) { @@ -212,3 +221,10 @@ function update_token_counter(button_id) { clearTimeout(token_timeout); token_timeout = setTimeout(() => gradioApp().getElementById(button_id)?.click(), wait_time); } +function submit_prompt(event, generate_button_id) { + if (event.altKey && event.keyCode === 13) { + event.preventDefault(); + gradioApp().getElementById(generate_button_id).click(); + return; + } +} \ No newline at end of file diff --git a/modules/ui.py b/modules/ui.py index 15572bb0..5eea1860 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -539,6 +539,7 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): roll.click( fn=roll_artist, + _js="roll_artist_txt2img", inputs=[ txt2img_prompt, ], @@ -743,6 +744,7 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): roll.click( fn=roll_artist, + _js="roll_artist_img2img", inputs=[ img2img_prompt, ], @@ -753,6 +755,7 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): prompts = [(txt2img_prompt, txt2img_negative_prompt), (img2img_prompt, img2img_negative_prompt)] style_dropdowns = [(txt2img_prompt_style, txt2img_prompt_style2), (img2img_prompt_style, img2img_prompt_style2)] + style_js_funcs = ["update_style_txt2img", "update_style_img2img"] for button, (prompt, negative_prompt) in zip([txt2img_save_style, img2img_save_style], prompts): button.click( @@ -764,9 +767,10 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): outputs=[txt2img_prompt_style, img2img_prompt_style, txt2img_prompt_style2, img2img_prompt_style2], ) - for button, (prompt, negative_prompt), (style1, style2) in zip([txt2img_prompt_style_apply, img2img_prompt_style_apply], prompts, style_dropdowns): + for button, (prompt, negative_prompt), (style1, style2), js_func in zip([txt2img_prompt_style_apply, img2img_prompt_style_apply], prompts, style_dropdowns, style_js_funcs): button.click( fn=apply_styles, + _js=js_func, inputs=[prompt, negative_prompt, style1, style2], outputs=[prompt, negative_prompt, style1, style2], ) -- cgit v1.2.3 From ff8dc1908af088d0ed43fb85baad662733c5ca9c Mon Sep 17 00:00:00 2001 From: Liam Date: Thu, 29 Sep 2022 15:47:06 -0400 Subject: fixed token counter for prompt editing --- modules/ui.py | 20 +++++++++++++------- 1 file changed, 13 insertions(+), 7 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/ui.py b/modules/ui.py index 5eea1860..6bf28562 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -11,6 +11,7 @@ import time import traceback import platform import subprocess as sp +from functools import reduce import numpy as np import torch @@ -32,6 +33,7 @@ import modules.gfpgan_model import modules.codeformer_model import modules.styles import modules.generation_parameters_copypaste +from modules.prompt_parser import get_learned_conditioning_prompt_schedules # this is a fix for Windows users. Without it, javascript files will be served with text/html content-type and the bowser will not show any UI mimetypes.init() @@ -345,8 +347,11 @@ def connect_reuse_seed(seed: gr.Number, reuse_seed: gr.Button, generation_info: outputs=[seed, dummy_component] ) -def update_token_counter(text): - tokens, token_count, max_length = model_hijack.tokenize(text) +def update_token_counter(text, steps): + prompt_schedules = get_learned_conditioning_prompt_schedules([text], steps) + flat_prompts = reduce(lambda list1, list2: list1+list2, prompt_schedules) + prompts = [prompt_text for step,prompt_text in flat_prompts] + tokens, token_count, max_length = max([model_hijack.tokenize(prompt) for prompt in prompts], key=lambda args: args[1]) style_class = ' class="red"' if (token_count > max_length) else "" return f"{token_count}/{max_length}" @@ -364,8 +369,7 @@ def create_toprow(is_img2img): roll = gr.Button(value=art_symbol, elem_id="roll", visible=len(shared.artist_db.artists) > 0) paste = gr.Button(value=paste_symbol, elem_id="paste") token_counter = gr.HTML(value="", elem_id=f"{id_part}_token_counter") - hidden_button = gr.Button(visible=False, elem_id=f"{id_part}_token_button") - hidden_button.click(fn=update_token_counter, inputs=[prompt], outputs=[token_counter]) + token_button = gr.Button(visible=False, elem_id=f"{id_part}_token_button") with gr.Column(scale=10, elem_id="style_pos_col"): prompt_style = gr.Dropdown(label="Style 1", elem_id=f"{id_part}_style_index", choices=[k for k, v in shared.prompt_styles.styles.items()], value=next(iter(shared.prompt_styles.styles.keys())), visible=len(shared.prompt_styles.styles) > 1) @@ -396,7 +400,7 @@ def create_toprow(is_img2img): prompt_style_apply = gr.Button('Apply style', elem_id="style_apply") save_style = gr.Button('Create style', elem_id="style_create") - return prompt, roll, prompt_style, negative_prompt, prompt_style2, submit, interrogate, prompt_style_apply, save_style, paste + return prompt, roll, prompt_style, negative_prompt, prompt_style2, submit, interrogate, prompt_style_apply, save_style, paste, token_counter, token_button def setup_progressbar(progressbar, preview, id_part): @@ -419,7 +423,7 @@ def setup_progressbar(progressbar, preview, id_part): def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): with gr.Blocks(analytics_enabled=False) as txt2img_interface: - txt2img_prompt, roll, txt2img_prompt_style, txt2img_negative_prompt, txt2img_prompt_style2, submit, _, txt2img_prompt_style_apply, txt2img_save_style, paste = create_toprow(is_img2img=False) + txt2img_prompt, roll, txt2img_prompt_style, txt2img_negative_prompt, txt2img_prompt_style2, submit, _, txt2img_prompt_style_apply, txt2img_save_style, paste, token_counter, token_button = create_toprow(is_img2img=False) dummy_component = gr.Label(visible=False) with gr.Row(elem_id='txt2img_progress_row'): @@ -568,9 +572,10 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): (hr_options, lambda d: gr.Row.update(visible="Denoising strength" in d)), ] modules.generation_parameters_copypaste.connect_paste(paste, txt2img_paste_fields, txt2img_prompt) + token_button.click(fn=update_token_counter, inputs=[txt2img_prompt, steps], outputs=[token_counter]) with gr.Blocks(analytics_enabled=False) as img2img_interface: - img2img_prompt, roll, img2img_prompt_style, img2img_negative_prompt, img2img_prompt_style2, submit, img2img_interrogate, img2img_prompt_style_apply, img2img_save_style, paste = create_toprow(is_img2img=True) + img2img_prompt, roll, img2img_prompt_style, img2img_negative_prompt, img2img_prompt_style2, submit, img2img_interrogate, img2img_prompt_style_apply, img2img_save_style, paste, token_counter, token_button = create_toprow(is_img2img=True) with gr.Row(elem_id='img2img_progress_row'): with gr.Column(scale=1): @@ -793,6 +798,7 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): (denoising_strength, "Denoising strength"), ] modules.generation_parameters_copypaste.connect_paste(paste, img2img_paste_fields, img2img_prompt) + token_button.click(fn=update_token_counter, inputs=[img2img_prompt, steps], outputs=[token_counter]) with gr.Blocks(analytics_enabled=False) as extras_interface: with gr.Row().style(equal_height=False): -- cgit v1.2.3 From 3c6a049fc3c6b54ada3736710a7e86663ea7f3d9 Mon Sep 17 00:00:00 2001 From: Liam Date: Fri, 30 Sep 2022 12:12:44 -0400 Subject: consolidated token counter functions --- javascript/ui.js | 21 +++++++++------------ modules/ui.py | 6 +++--- 2 files changed, 12 insertions(+), 15 deletions(-) (limited to 'modules/ui.py') diff --git a/javascript/ui.js b/javascript/ui.js index 88fd45ae..f94ed081 100644 --- a/javascript/ui.js +++ b/javascript/ui.js @@ -199,21 +199,18 @@ let txt2img_textarea, img2img_textarea = undefined; let wait_time = 800 let token_timeout; -function roll_artist_txt2img(prompt_text) { +function update_txt2img_tokens(...args) { update_token_counter("txt2img_token_button") - return prompt_text; + if (args.length == 2) + return args[0] + return args; } -function roll_artist_img2img(prompt_text) { - update_token_counter("img2img_token_button") - return prompt_text; -} -function update_style_txt2img(prompt_text, negative_prompt, style1, style2) { - update_token_counter("txt2img_token_button") - return [prompt_text, negative_prompt, style1, style2] -} -function update_style_img2img(prompt_text, negative_prompt, style1, style2) { + +function update_img2img_tokens(...args) { update_token_counter("img2img_token_button") - return [prompt_text, negative_prompt, style1, style2] + if (args.length == 2) + return args[0] + return args; } function update_token_counter(button_id) { diff --git a/modules/ui.py b/modules/ui.py index 6bf28562..40c08984 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -543,7 +543,7 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): roll.click( fn=roll_artist, - _js="roll_artist_txt2img", + _js="update_txt2img_tokens", inputs=[ txt2img_prompt, ], @@ -749,7 +749,7 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): roll.click( fn=roll_artist, - _js="roll_artist_img2img", + _js="update_img2img_tokens", inputs=[ img2img_prompt, ], @@ -760,7 +760,7 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): prompts = [(txt2img_prompt, txt2img_negative_prompt), (img2img_prompt, img2img_negative_prompt)] style_dropdowns = [(txt2img_prompt_style, txt2img_prompt_style2), (img2img_prompt_style, img2img_prompt_style2)] - style_js_funcs = ["update_style_txt2img", "update_style_img2img"] + style_js_funcs = ["update_txt2img_tokens", "update_img2img_tokens"] for button, (prompt, negative_prompt) in zip([txt2img_save_style, img2img_save_style], prompts): button.click( -- cgit v1.2.3 From 4f8490cd5630823ac44de8b5c5e4325bdbbea7fa Mon Sep 17 00:00:00 2001 From: DepFA <35278260+dfaker@users.noreply.github.com> Date: Sat, 1 Oct 2022 18:33:31 +0100 Subject: add restart button --- modules/ui.py | 15 ++++++++++++++- 1 file changed, 14 insertions(+), 1 deletion(-) (limited to 'modules/ui.py') diff --git a/modules/ui.py b/modules/ui.py index 15572bb0..ec6aaa28 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1002,6 +1002,17 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): _js='function(){}' ) + def request_restart(): + settings_interface.gradio_ref.do_restart = True + + restart_gradio = gr.Button(value='Restart Gradio and Refresh Scripts') + restart_gradio.click( + fn=request_restart, + inputs=[], + outputs=[], + _js='function(){document.body.innerHTML=\'

Reloading

\';setTimeout(function(){location.reload()},2000)}' + ) + if column is not None: column.__exit__() @@ -1026,7 +1037,9 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): css += css_hide_progressbar with gr.Blocks(css=css, analytics_enabled=False, title="Stable Diffusion") as demo: - + + settings_interface.gradio_ref = demo + with gr.Tabs() as tabs: for interface, label, ifid in interfaces: with gr.TabItem(label, id=ifid): -- cgit v1.2.3 From afaa03c5fd05f48ed9c9f15558ea6f0bc4f61628 Mon Sep 17 00:00:00 2001 From: DepFA <35278260+dfaker@users.noreply.github.com> Date: Sat, 1 Oct 2022 22:43:45 +0100 Subject: add redefinition guard to gradio_routes_templates_response --- modules/ui.py | 15 ++++++++------- 1 file changed, 8 insertions(+), 7 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/ui.py b/modules/ui.py index ec6aaa28..fd057916 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1219,12 +1219,13 @@ for filename in sorted(os.listdir(jsdir)): javascript += f"\n" -def template_response(*args, **kwargs): - res = gradio_routes_templates_response(*args, **kwargs) - res.body = res.body.replace(b'', f'{javascript}'.encode("utf8")) - res.init_headers() - return res +if 'gradio_routes_templates_response' not in globals(): + def template_response(*args, **kwargs): + res = gradio_routes_templates_response(*args, **kwargs) + res.body = res.body.replace(b'', f'{javascript}'.encode("utf8")) + res.init_headers() + return res + gradio_routes_templates_response = gradio.routes.templates.TemplateResponse + gradio.routes.templates.TemplateResponse = template_response -gradio_routes_templates_response = gradio.routes.templates.TemplateResponse -gradio.routes.templates.TemplateResponse = template_response -- cgit v1.2.3 From 6048002dade91b82b1ce9fea3c6ff5b5c1f8c990 Mon Sep 17 00:00:00 2001 From: DepFA <35278260+dfaker@users.noreply.github.com> Date: Sat, 1 Oct 2022 23:10:07 +0100 Subject: Add scope warning to refresh button --- modules/ui.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'modules/ui.py') diff --git a/modules/ui.py b/modules/ui.py index fd057916..72846a12 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1005,7 +1005,7 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): def request_restart(): settings_interface.gradio_ref.do_restart = True - restart_gradio = gr.Button(value='Restart Gradio and Refresh Scripts') + restart_gradio = gr.Button(value='Restart Gradio and Refresh components (Custom Scripts, ui.py, js and css only)', variant='primary') restart_gradio.click( fn=request_restart, inputs=[], -- cgit v1.2.3 From 027c5aae5546ff3650347cb3c2b87df4415ab900 Mon Sep 17 00:00:00 2001 From: DepFA <35278260+dfaker@users.noreply.github.com> Date: Sat, 1 Oct 2022 23:29:26 +0100 Subject: update reloading message style --- modules/ui.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'modules/ui.py') diff --git a/modules/ui.py b/modules/ui.py index 72846a12..7b2359c2 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1010,7 +1010,7 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): fn=request_restart, inputs=[], outputs=[], - _js='function(){document.body.innerHTML=\'

Reloading

\';setTimeout(function(){location.reload()},2000)}' + _js='function(){document.body.innerHTML=\'

Reloading...

\';setTimeout(function(){location.reload()},2000)}' ) if column is not None: -- cgit v1.2.3 From 0aa354bd5e811e2b41b17a3052cf5d4c8190d533 Mon Sep 17 00:00:00 2001 From: DepFA <35278260+dfaker@users.noreply.github.com> Date: Sun, 2 Oct 2022 00:13:47 +0100 Subject: remove styling from python side --- modules/ui.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'modules/ui.py') diff --git a/modules/ui.py b/modules/ui.py index 7b2359c2..cb859ac4 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1010,7 +1010,7 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): fn=request_restart, inputs=[], outputs=[], - _js='function(){document.body.innerHTML=\'

Reloading...

\';setTimeout(function(){location.reload()},2000)}' + _js='function(){restart_reload()}' ) if column is not None: -- cgit v1.2.3 From 07e40ad7f23472fc1c781fe1cc6c1ee403413918 Mon Sep 17 00:00:00 2001 From: DepFA <35278260+dfaker@users.noreply.github.com> Date: Sun, 2 Oct 2022 01:19:55 +0100 Subject: add custom script body only refresh option --- modules/ui.py | 11 +++++++++++ 1 file changed, 11 insertions(+) (limited to 'modules/ui.py') diff --git a/modules/ui.py b/modules/ui.py index cb859ac4..eb7c0585 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1012,6 +1012,17 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): outputs=[], _js='function(){restart_reload()}' ) + + def reload_scripts(): + modules.scripts.reload_script_body_only() + + reload_script_bodies = gr.Button(value='Reload custom script bodies (No ui updates, No restart)', variant='primary') + reload_script_bodies.click( + fn=reload_scripts, + inputs=[], + outputs=[], + _js='function(){}' + ) if column is not None: column.__exit__() -- cgit v1.2.3 From 2deea867814272f1f089b60e9ba8d587c16b2fb1 Mon Sep 17 00:00:00 2001 From: DepFA <35278260+dfaker@users.noreply.github.com> Date: Sun, 2 Oct 2022 01:36:30 +0100 Subject: Put reload buttons in row and add secondary style --- modules/ui.py | 23 +++++++++++++---------- 1 file changed, 13 insertions(+), 10 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/ui.py b/modules/ui.py index eb7c0585..963a2c61 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1002,27 +1002,30 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): _js='function(){}' ) - def request_restart(): - settings_interface.gradio_ref.do_restart = True + with gr.Row(): + reload_script_bodies = gr.Button(value='Reload custom script bodies (No ui updates, No restart)', variant='secondary') + restart_gradio = gr.Button(value='Restart Gradio and Refresh components (Custom Scripts, ui.py, js and css only)', variant='primary') - restart_gradio = gr.Button(value='Restart Gradio and Refresh components (Custom Scripts, ui.py, js and css only)', variant='primary') - restart_gradio.click( - fn=request_restart, - inputs=[], - outputs=[], - _js='function(){restart_reload()}' - ) def reload_scripts(): modules.scripts.reload_script_body_only() - reload_script_bodies = gr.Button(value='Reload custom script bodies (No ui updates, No restart)', variant='primary') reload_script_bodies.click( fn=reload_scripts, inputs=[], outputs=[], _js='function(){}' ) + + def request_restart(): + settings_interface.gradio_ref.do_restart = True + + restart_gradio.click( + fn=request_restart, + inputs=[], + outputs=[], + _js='function(){restart_reload()}' + ) if column is not None: column.__exit__() -- cgit v1.2.3 From 820f1dc96b1979d7e92170c161db281ee8bd988b Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Sun, 2 Oct 2022 15:03:39 +0300 Subject: initial support for training textual inversion --- .gitignore | 1 + javascript/progressbar.js | 1 + javascript/textualInversion.js | 8 + modules/devices.py | 3 +- modules/processing.py | 13 +- modules/sd_hijack.py | 324 ++++------------------ modules/sd_hijack_optimizations.py | 164 +++++++++++ modules/sd_models.py | 4 +- modules/shared.py | 3 +- modules/textual_inversion/dataset.py | 76 +++++ modules/textual_inversion/textual_inversion.py | 258 +++++++++++++++++ modules/textual_inversion/ui.py | 32 +++ modules/ui.py | 139 ++++++++-- style.css | 10 +- textual_inversion_templates/style.txt | 19 ++ textual_inversion_templates/style_filewords.txt | 19 ++ textual_inversion_templates/subject.txt | 27 ++ textual_inversion_templates/subject_filewords.txt | 27 ++ webui.py | 15 +- 19 files changed, 828 insertions(+), 315 deletions(-) create mode 100644 javascript/textualInversion.js create mode 100644 modules/sd_hijack_optimizations.py create mode 100644 modules/textual_inversion/dataset.py create mode 100644 modules/textual_inversion/textual_inversion.py create mode 100644 modules/textual_inversion/ui.py create mode 100644 textual_inversion_templates/style.txt create mode 100644 textual_inversion_templates/style_filewords.txt create mode 100644 textual_inversion_templates/subject.txt create mode 100644 textual_inversion_templates/subject_filewords.txt (limited to 'modules/ui.py') diff --git a/.gitignore b/.gitignore index 3532dab3..7afc9395 100644 --- a/.gitignore +++ b/.gitignore @@ -25,3 +25,4 @@ __pycache__ /.idea notification.mp3 /SwinIR +/textual_inversion diff --git a/javascript/progressbar.js b/javascript/progressbar.js index 21f25b38..1e297abb 100644 --- a/javascript/progressbar.js +++ b/javascript/progressbar.js @@ -30,6 +30,7 @@ function check_progressbar(id_part, id_progressbar, id_progressbar_span, id_inte onUiUpdate(function(){ check_progressbar('txt2img', 'txt2img_progressbar', 'txt2img_progress_span', 'txt2img_interrupt', 'txt2img_preview', 'txt2img_gallery') check_progressbar('img2img', 'img2img_progressbar', 'img2img_progress_span', 'img2img_interrupt', 'img2img_preview', 'img2img_gallery') + check_progressbar('ti', 'ti_progressbar', 'ti_progress_span', 'ti_interrupt', 'ti_preview', 'ti_gallery') }) function requestMoreProgress(id_part, id_progressbar_span, id_interrupt){ diff --git a/javascript/textualInversion.js b/javascript/textualInversion.js new file mode 100644 index 00000000..8061be08 --- /dev/null +++ b/javascript/textualInversion.js @@ -0,0 +1,8 @@ + + +function start_training_textual_inversion(){ + requestProgress('ti') + gradioApp().querySelector('#ti_error').innerHTML='' + + return args_to_array(arguments) +} diff --git a/modules/devices.py b/modules/devices.py index 07bb2339..ff82f2f6 100644 --- a/modules/devices.py +++ b/modules/devices.py @@ -32,10 +32,9 @@ def enable_tf32(): errors.run(enable_tf32, "Enabling TF32") - device = get_optimal_device() device_codeformer = cpu if has_mps else device - +dtype = torch.float16 def randn(seed, shape): # Pytorch currently doesn't handle setting randomness correctly when the metal backend is used. diff --git a/modules/processing.py b/modules/processing.py index 7eeb5191..8223423a 100644 --- a/modules/processing.py +++ b/modules/processing.py @@ -56,7 +56,7 @@ class StableDiffusionProcessing: self.prompt: str = prompt self.prompt_for_display: str = None self.negative_prompt: str = (negative_prompt or "") - self.styles: str = styles + self.styles: list = styles or [] self.seed: int = seed self.subseed: int = subseed self.subseed_strength: float = subseed_strength @@ -271,7 +271,7 @@ def create_infotext(p, all_prompts, all_seeds, all_subseeds, comments, iteration "Variation seed strength": (None if p.subseed_strength == 0 else p.subseed_strength), "Seed resize from": (None if p.seed_resize_from_w == 0 or p.seed_resize_from_h == 0 else f"{p.seed_resize_from_w}x{p.seed_resize_from_h}"), "Denoising strength": getattr(p, 'denoising_strength', None), - "Eta": (None if p.sampler.eta == p.sampler.default_eta else p.sampler.eta), + "Eta": (None if p.sampler is None or p.sampler.eta == p.sampler.default_eta else p.sampler.eta), } generation_params.update(p.extra_generation_params) @@ -295,8 +295,11 @@ def process_images(p: StableDiffusionProcessing) -> Processed: fix_seed(p) - os.makedirs(p.outpath_samples, exist_ok=True) - os.makedirs(p.outpath_grids, exist_ok=True) + if p.outpath_samples is not None: + os.makedirs(p.outpath_samples, exist_ok=True) + + if p.outpath_grids is not None: + os.makedirs(p.outpath_grids, exist_ok=True) modules.sd_hijack.model_hijack.apply_circular(p.tiling) @@ -323,7 +326,7 @@ def process_images(p: StableDiffusionProcessing) -> Processed: return create_infotext(p, all_prompts, all_seeds, all_subseeds, comments, iteration, position_in_batch) if os.path.exists(cmd_opts.embeddings_dir): - model_hijack.load_textual_inversion_embeddings(cmd_opts.embeddings_dir, p.sd_model) + model_hijack.embedding_db.load_textual_inversion_embeddings() infotexts = [] output_images = [] diff --git a/modules/sd_hijack.py b/modules/sd_hijack.py index fa7eaeb8..fd57e5c5 100644 --- a/modules/sd_hijack.py +++ b/modules/sd_hijack.py @@ -6,244 +6,41 @@ import torch import numpy as np from torch import einsum -from modules import prompt_parser +import modules.textual_inversion.textual_inversion +from modules import prompt_parser, devices, sd_hijack_optimizations, shared from modules.shared import opts, device, cmd_opts -from ldm.util import default -from einops import rearrange import ldm.modules.attention import ldm.modules.diffusionmodules.model +attention_CrossAttention_forward = ldm.modules.attention.CrossAttention.forward +diffusionmodules_model_nonlinearity = ldm.modules.diffusionmodules.model.nonlinearity +diffusionmodules_model_AttnBlock_forward = ldm.modules.diffusionmodules.model.AttnBlock.forward -# see https://github.com/basujindal/stable-diffusion/pull/117 for discussion -def split_cross_attention_forward_v1(self, x, context=None, mask=None): - h = self.heads - q = self.to_q(x) - context = default(context, x) - k = self.to_k(context) - v = self.to_v(context) - del context, x +def apply_optimizations(): + if cmd_opts.opt_split_attention_v1: + ldm.modules.attention.CrossAttention.forward = sd_hijack_optimizations.split_cross_attention_forward_v1 + elif not cmd_opts.disable_opt_split_attention and (cmd_opts.opt_split_attention or torch.cuda.is_available()): + ldm.modules.attention.CrossAttention.forward = sd_hijack_optimizations.split_cross_attention_forward + ldm.modules.diffusionmodules.model.nonlinearity = sd_hijack_optimizations.nonlinearity_hijack + ldm.modules.diffusionmodules.model.AttnBlock.forward = sd_hijack_optimizations.cross_attention_attnblock_forward - q, k, v = map(lambda t: rearrange(t, 'b n (h d) -> (b h) n d', h=h), (q, k, v)) - r1 = torch.zeros(q.shape[0], q.shape[1], v.shape[2], device=q.device) - for i in range(0, q.shape[0], 2): - end = i + 2 - s1 = einsum('b i d, b j d -> b i j', q[i:end], k[i:end]) - s1 *= self.scale +def undo_optimizations(): + ldm.modules.attention.CrossAttention.forward = attention_CrossAttention_forward + ldm.modules.diffusionmodules.model.nonlinearity = diffusionmodules_model_nonlinearity + ldm.modules.diffusionmodules.model.AttnBlock.forward = diffusionmodules_model_AttnBlock_forward - s2 = s1.softmax(dim=-1) - del s1 - - r1[i:end] = einsum('b i j, b j d -> b i d', s2, v[i:end]) - del s2 - - r2 = rearrange(r1, '(b h) n d -> b n (h d)', h=h) - del r1 - - return self.to_out(r2) - - -# taken from https://github.com/Doggettx/stable-diffusion -def split_cross_attention_forward(self, x, context=None, mask=None): - h = self.heads - - q_in = self.to_q(x) - context = default(context, x) - k_in = self.to_k(context) * self.scale - v_in = self.to_v(context) - del context, x - - q, k, v = map(lambda t: rearrange(t, 'b n (h d) -> (b h) n d', h=h), (q_in, k_in, v_in)) - del q_in, k_in, v_in - - r1 = torch.zeros(q.shape[0], q.shape[1], v.shape[2], device=q.device, dtype=q.dtype) - - stats = torch.cuda.memory_stats(q.device) - mem_active = stats['active_bytes.all.current'] - mem_reserved = stats['reserved_bytes.all.current'] - mem_free_cuda, _ = torch.cuda.mem_get_info(torch.cuda.current_device()) - mem_free_torch = mem_reserved - mem_active - mem_free_total = mem_free_cuda + mem_free_torch - - gb = 1024 ** 3 - tensor_size = q.shape[0] * q.shape[1] * k.shape[1] * q.element_size() - modifier = 3 if q.element_size() == 2 else 2.5 - mem_required = tensor_size * modifier - steps = 1 - - if mem_required > mem_free_total: - steps = 2 ** (math.ceil(math.log(mem_required / mem_free_total, 2))) - # print(f"Expected tensor size:{tensor_size/gb:0.1f}GB, cuda free:{mem_free_cuda/gb:0.1f}GB " - # f"torch free:{mem_free_torch/gb:0.1f} total:{mem_free_total/gb:0.1f} steps:{steps}") - - if steps > 64: - max_res = math.floor(math.sqrt(math.sqrt(mem_free_total / 2.5)) / 8) * 64 - raise RuntimeError(f'Not enough memory, use lower resolution (max approx. {max_res}x{max_res}). ' - f'Need: {mem_required / 64 / gb:0.1f}GB free, Have:{mem_free_total / gb:0.1f}GB free') - - slice_size = q.shape[1] // steps if (q.shape[1] % steps) == 0 else q.shape[1] - for i in range(0, q.shape[1], slice_size): - end = i + slice_size - s1 = einsum('b i d, b j d -> b i j', q[:, i:end], k) - - s2 = s1.softmax(dim=-1, dtype=q.dtype) - del s1 - - r1[:, i:end] = einsum('b i j, b j d -> b i d', s2, v) - del s2 - - del q, k, v - - r2 = rearrange(r1, '(b h) n d -> b n (h d)', h=h) - del r1 - - return self.to_out(r2) - -def nonlinearity_hijack(x): - # swish - t = torch.sigmoid(x) - x *= t - del t - - return x - -def cross_attention_attnblock_forward(self, x): - h_ = x - h_ = self.norm(h_) - q1 = self.q(h_) - k1 = self.k(h_) - v = self.v(h_) - - # compute attention - b, c, h, w = q1.shape - - q2 = q1.reshape(b, c, h*w) - del q1 - - q = q2.permute(0, 2, 1) # b,hw,c - del q2 - - k = k1.reshape(b, c, h*w) # b,c,hw - del k1 - - h_ = torch.zeros_like(k, device=q.device) - - stats = torch.cuda.memory_stats(q.device) - mem_active = stats['active_bytes.all.current'] - mem_reserved = stats['reserved_bytes.all.current'] - mem_free_cuda, _ = torch.cuda.mem_get_info(torch.cuda.current_device()) - mem_free_torch = mem_reserved - mem_active - mem_free_total = mem_free_cuda + mem_free_torch - - tensor_size = q.shape[0] * q.shape[1] * k.shape[2] * q.element_size() - mem_required = tensor_size * 2.5 - steps = 1 - - if mem_required > mem_free_total: - steps = 2**(math.ceil(math.log(mem_required / mem_free_total, 2))) - - slice_size = q.shape[1] // steps if (q.shape[1] % steps) == 0 else q.shape[1] - for i in range(0, q.shape[1], slice_size): - end = i + slice_size - - w1 = torch.bmm(q[:, i:end], k) # b,hw,hw w[b,i,j]=sum_c q[b,i,c]k[b,c,j] - w2 = w1 * (int(c)**(-0.5)) - del w1 - w3 = torch.nn.functional.softmax(w2, dim=2, dtype=q.dtype) - del w2 - - # attend to values - v1 = v.reshape(b, c, h*w) - w4 = w3.permute(0, 2, 1) # b,hw,hw (first hw of k, second of q) - del w3 - - h_[:, :, i:end] = torch.bmm(v1, w4) # b, c,hw (hw of q) h_[b,c,j] = sum_i v[b,c,i] w_[b,i,j] - del v1, w4 - - h2 = h_.reshape(b, c, h, w) - del h_ - - h3 = self.proj_out(h2) - del h2 - - h3 += x - - return h3 class StableDiffusionModelHijack: - ids_lookup = {} - word_embeddings = {} - word_embeddings_checksums = {} fixes = None comments = [] - dir_mtime = None layers = None circular_enabled = False clip = None - def load_textual_inversion_embeddings(self, dirname, model): - mt = os.path.getmtime(dirname) - if self.dir_mtime is not None and mt <= self.dir_mtime: - return - - self.dir_mtime = mt - self.ids_lookup.clear() - self.word_embeddings.clear() - - tokenizer = model.cond_stage_model.tokenizer - - def const_hash(a): - r = 0 - for v in a: - r = (r * 281 ^ int(v) * 997) & 0xFFFFFFFF - return r - - def process_file(path, filename): - name = os.path.splitext(filename)[0] - - data = torch.load(path, map_location="cpu") - - # textual inversion embeddings - if 'string_to_param' in data: - param_dict = data['string_to_param'] - if hasattr(param_dict, '_parameters'): - param_dict = getattr(param_dict, '_parameters') # fix for torch 1.12.1 loading saved file from torch 1.11 - assert len(param_dict) == 1, 'embedding file has multiple terms in it' - emb = next(iter(param_dict.items()))[1] - # diffuser concepts - elif type(data) == dict and type(next(iter(data.values()))) == torch.Tensor: - assert len(data.keys()) == 1, 'embedding file has multiple terms in it' - - emb = next(iter(data.values())) - if len(emb.shape) == 1: - emb = emb.unsqueeze(0) - - self.word_embeddings[name] = emb.detach().to(device) - self.word_embeddings_checksums[name] = f'{const_hash(emb.reshape(-1)*100)&0xffff:04x}' - - ids = tokenizer([name], add_special_tokens=False)['input_ids'][0] - - first_id = ids[0] - if first_id not in self.ids_lookup: - self.ids_lookup[first_id] = [] - self.ids_lookup[first_id].append((ids, name)) - - for fn in os.listdir(dirname): - try: - fullfn = os.path.join(dirname, fn) - - if os.stat(fullfn).st_size == 0: - continue - - process_file(fullfn, fn) - except Exception: - print(f"Error loading emedding {fn}:", file=sys.stderr) - print(traceback.format_exc(), file=sys.stderr) - continue - - print(f"Loaded a total of {len(self.word_embeddings)} textual inversion embeddings.") + embedding_db = modules.textual_inversion.textual_inversion.EmbeddingDatabase(cmd_opts.embeddings_dir) def hijack(self, m): model_embeddings = m.cond_stage_model.transformer.text_model.embeddings @@ -253,12 +50,7 @@ class StableDiffusionModelHijack: self.clip = m.cond_stage_model - if cmd_opts.opt_split_attention_v1: - ldm.modules.attention.CrossAttention.forward = split_cross_attention_forward_v1 - elif not cmd_opts.disable_opt_split_attention and (cmd_opts.opt_split_attention or torch.cuda.is_available()): - ldm.modules.attention.CrossAttention.forward = split_cross_attention_forward - ldm.modules.diffusionmodules.model.nonlinearity = nonlinearity_hijack - ldm.modules.diffusionmodules.model.AttnBlock.forward = cross_attention_attnblock_forward + apply_optimizations() def flatten(el): flattened = [flatten(children) for children in el.children()] @@ -296,7 +88,7 @@ class FrozenCLIPEmbedderWithCustomWords(torch.nn.Module): def __init__(self, wrapped, hijack): super().__init__() self.wrapped = wrapped - self.hijack = hijack + self.hijack: StableDiffusionModelHijack = hijack self.tokenizer = wrapped.tokenizer self.max_length = wrapped.max_length self.token_mults = {} @@ -317,7 +109,6 @@ class FrozenCLIPEmbedderWithCustomWords(torch.nn.Module): if mult != 1.0: self.token_mults[ident] = mult - def tokenize_line(self, line, used_custom_terms, hijack_comments): id_start = self.wrapped.tokenizer.bos_token_id id_end = self.wrapped.tokenizer.eos_token_id @@ -339,28 +130,19 @@ class FrozenCLIPEmbedderWithCustomWords(torch.nn.Module): while i < len(tokens): token = tokens[i] - possible_matches = self.hijack.ids_lookup.get(token, None) + embedding = self.hijack.embedding_db.find_embedding_at_position(tokens, i) - if possible_matches is None: + if embedding is None: remade_tokens.append(token) multipliers.append(weight) + i += 1 else: - found = False - for ids, word in possible_matches: - if tokens[i:i + len(ids)] == ids: - emb_len = int(self.hijack.word_embeddings[word].shape[0]) - fixes.append((len(remade_tokens), word)) - remade_tokens += [0] * emb_len - multipliers += [weight] * emb_len - i += len(ids) - 1 - found = True - used_custom_terms.append((word, self.hijack.word_embeddings_checksums[word])) - break - - if not found: - remade_tokens.append(token) - multipliers.append(weight) - i += 1 + emb_len = int(embedding.vec.shape[0]) + fixes.append((len(remade_tokens), embedding)) + remade_tokens += [0] * emb_len + multipliers += [weight] * emb_len + used_custom_terms.append((embedding.name, embedding.checksum())) + i += emb_len if len(remade_tokens) > maxlen - 2: vocab = {v: k for k, v in self.wrapped.tokenizer.get_vocab().items()} @@ -431,32 +213,23 @@ class FrozenCLIPEmbedderWithCustomWords(torch.nn.Module): while i < len(tokens): token = tokens[i] - possible_matches = self.hijack.ids_lookup.get(token, None) + embedding = self.hijack.embedding_db.find_embedding_at_position(tokens, i) mult_change = self.token_mults.get(token) if opts.enable_emphasis else None if mult_change is not None: mult *= mult_change - elif possible_matches is None: + i += 1 + elif embedding is None: remade_tokens.append(token) multipliers.append(mult) + i += 1 else: - found = False - for ids, word in possible_matches: - if tokens[i:i+len(ids)] == ids: - emb_len = int(self.hijack.word_embeddings[word].shape[0]) - fixes.append((len(remade_tokens), word)) - remade_tokens += [0] * emb_len - multipliers += [mult] * emb_len - i += len(ids) - 1 - found = True - used_custom_terms.append((word, self.hijack.word_embeddings_checksums[word])) - break - - if not found: - remade_tokens.append(token) - multipliers.append(mult) - - i += 1 + emb_len = int(embedding.vec.shape[0]) + fixes.append((len(remade_tokens), embedding)) + remade_tokens += [0] * emb_len + multipliers += [mult] * emb_len + used_custom_terms.append((embedding.name, embedding.checksum())) + i += emb_len if len(remade_tokens) > maxlen - 2: vocab = {v: k for k, v in self.wrapped.tokenizer.get_vocab().items()} @@ -464,6 +237,7 @@ class FrozenCLIPEmbedderWithCustomWords(torch.nn.Module): overflowing_words = [vocab.get(int(x), "") for x in ovf] overflowing_text = self.wrapped.tokenizer.convert_tokens_to_string(''.join(overflowing_words)) hijack_comments.append(f"Warning: too many input tokens; some ({len(overflowing_words)}) have been truncated:\n{overflowing_text}\n") + token_count = len(remade_tokens) remade_tokens = remade_tokens + [id_end] * (maxlen - 2 - len(remade_tokens)) remade_tokens = [id_start] + remade_tokens[0:maxlen-2] + [id_end] @@ -484,7 +258,6 @@ class FrozenCLIPEmbedderWithCustomWords(torch.nn.Module): else: batch_multipliers, remade_batch_tokens, used_custom_terms, hijack_comments, hijack_fixes, token_count = self.process_text(text) - self.hijack.fixes = hijack_fixes self.hijack.comments = hijack_comments @@ -517,14 +290,19 @@ class EmbeddingsWithFixes(torch.nn.Module): inputs_embeds = self.wrapped(input_ids) - if batch_fixes is not None: - for fixes, tensor in zip(batch_fixes, inputs_embeds): - for offset, word in fixes: - emb = self.embeddings.word_embeddings[word] - emb_len = min(tensor.shape[0]-offset-1, emb.shape[0]) - tensor[offset+1:offset+1+emb_len] = self.embeddings.word_embeddings[word][0:emb_len] + if batch_fixes is None or len(batch_fixes) == 0 or max([len(x) for x in batch_fixes]) == 0: + return inputs_embeds + + vecs = [] + for fixes, tensor in zip(batch_fixes, inputs_embeds): + for offset, embedding in fixes: + emb = embedding.vec + emb_len = min(tensor.shape[0]-offset-1, emb.shape[0]) + tensor = torch.cat([tensor[0:offset+1], emb[0:emb_len], tensor[offset+1+emb_len:]]) + + vecs.append(tensor) - return inputs_embeds + return torch.stack(vecs) def add_circular_option_to_conv_2d(): diff --git a/modules/sd_hijack_optimizations.py b/modules/sd_hijack_optimizations.py new file mode 100644 index 00000000..9c079e57 --- /dev/null +++ b/modules/sd_hijack_optimizations.py @@ -0,0 +1,164 @@ +import math +import torch +from torch import einsum + +from ldm.util import default +from einops import rearrange + + +# see https://github.com/basujindal/stable-diffusion/pull/117 for discussion +def split_cross_attention_forward_v1(self, x, context=None, mask=None): + h = self.heads + + q = self.to_q(x) + context = default(context, x) + k = self.to_k(context) + v = self.to_v(context) + del context, x + + q, k, v = map(lambda t: rearrange(t, 'b n (h d) -> (b h) n d', h=h), (q, k, v)) + + r1 = torch.zeros(q.shape[0], q.shape[1], v.shape[2], device=q.device) + for i in range(0, q.shape[0], 2): + end = i + 2 + s1 = einsum('b i d, b j d -> b i j', q[i:end], k[i:end]) + s1 *= self.scale + + s2 = s1.softmax(dim=-1) + del s1 + + r1[i:end] = einsum('b i j, b j d -> b i d', s2, v[i:end]) + del s2 + + r2 = rearrange(r1, '(b h) n d -> b n (h d)', h=h) + del r1 + + return self.to_out(r2) + + +# taken from https://github.com/Doggettx/stable-diffusion +def split_cross_attention_forward(self, x, context=None, mask=None): + h = self.heads + + q_in = self.to_q(x) + context = default(context, x) + k_in = self.to_k(context) * self.scale + v_in = self.to_v(context) + del context, x + + q, k, v = map(lambda t: rearrange(t, 'b n (h d) -> (b h) n d', h=h), (q_in, k_in, v_in)) + del q_in, k_in, v_in + + r1 = torch.zeros(q.shape[0], q.shape[1], v.shape[2], device=q.device, dtype=q.dtype) + + stats = torch.cuda.memory_stats(q.device) + mem_active = stats['active_bytes.all.current'] + mem_reserved = stats['reserved_bytes.all.current'] + mem_free_cuda, _ = torch.cuda.mem_get_info(torch.cuda.current_device()) + mem_free_torch = mem_reserved - mem_active + mem_free_total = mem_free_cuda + mem_free_torch + + gb = 1024 ** 3 + tensor_size = q.shape[0] * q.shape[1] * k.shape[1] * q.element_size() + modifier = 3 if q.element_size() == 2 else 2.5 + mem_required = tensor_size * modifier + steps = 1 + + if mem_required > mem_free_total: + steps = 2 ** (math.ceil(math.log(mem_required / mem_free_total, 2))) + # print(f"Expected tensor size:{tensor_size/gb:0.1f}GB, cuda free:{mem_free_cuda/gb:0.1f}GB " + # f"torch free:{mem_free_torch/gb:0.1f} total:{mem_free_total/gb:0.1f} steps:{steps}") + + if steps > 64: + max_res = math.floor(math.sqrt(math.sqrt(mem_free_total / 2.5)) / 8) * 64 + raise RuntimeError(f'Not enough memory, use lower resolution (max approx. {max_res}x{max_res}). ' + f'Need: {mem_required / 64 / gb:0.1f}GB free, Have:{mem_free_total / gb:0.1f}GB free') + + slice_size = q.shape[1] // steps if (q.shape[1] % steps) == 0 else q.shape[1] + for i in range(0, q.shape[1], slice_size): + end = i + slice_size + s1 = einsum('b i d, b j d -> b i j', q[:, i:end], k) + + s2 = s1.softmax(dim=-1, dtype=q.dtype) + del s1 + + r1[:, i:end] = einsum('b i j, b j d -> b i d', s2, v) + del s2 + + del q, k, v + + r2 = rearrange(r1, '(b h) n d -> b n (h d)', h=h) + del r1 + + return self.to_out(r2) + +def nonlinearity_hijack(x): + # swish + t = torch.sigmoid(x) + x *= t + del t + + return x + +def cross_attention_attnblock_forward(self, x): + h_ = x + h_ = self.norm(h_) + q1 = self.q(h_) + k1 = self.k(h_) + v = self.v(h_) + + # compute attention + b, c, h, w = q1.shape + + q2 = q1.reshape(b, c, h*w) + del q1 + + q = q2.permute(0, 2, 1) # b,hw,c + del q2 + + k = k1.reshape(b, c, h*w) # b,c,hw + del k1 + + h_ = torch.zeros_like(k, device=q.device) + + stats = torch.cuda.memory_stats(q.device) + mem_active = stats['active_bytes.all.current'] + mem_reserved = stats['reserved_bytes.all.current'] + mem_free_cuda, _ = torch.cuda.mem_get_info(torch.cuda.current_device()) + mem_free_torch = mem_reserved - mem_active + mem_free_total = mem_free_cuda + mem_free_torch + + tensor_size = q.shape[0] * q.shape[1] * k.shape[2] * q.element_size() + mem_required = tensor_size * 2.5 + steps = 1 + + if mem_required > mem_free_total: + steps = 2**(math.ceil(math.log(mem_required / mem_free_total, 2))) + + slice_size = q.shape[1] // steps if (q.shape[1] % steps) == 0 else q.shape[1] + for i in range(0, q.shape[1], slice_size): + end = i + slice_size + + w1 = torch.bmm(q[:, i:end], k) # b,hw,hw w[b,i,j]=sum_c q[b,i,c]k[b,c,j] + w2 = w1 * (int(c)**(-0.5)) + del w1 + w3 = torch.nn.functional.softmax(w2, dim=2, dtype=q.dtype) + del w2 + + # attend to values + v1 = v.reshape(b, c, h*w) + w4 = w3.permute(0, 2, 1) # b,hw,hw (first hw of k, second of q) + del w3 + + h_[:, :, i:end] = torch.bmm(v1, w4) # b, c,hw (hw of q) h_[b,c,j] = sum_i v[b,c,i] w_[b,i,j] + del v1, w4 + + h2 = h_.reshape(b, c, h, w) + del h_ + + h3 = self.proj_out(h2) + del h2 + + h3 += x + + return h3 diff --git a/modules/sd_models.py b/modules/sd_models.py index 2539f14c..5b3dbdc7 100644 --- a/modules/sd_models.py +++ b/modules/sd_models.py @@ -8,7 +8,7 @@ from omegaconf import OmegaConf from ldm.util import instantiate_from_config -from modules import shared, modelloader +from modules import shared, modelloader, devices from modules.paths import models_path model_dir = "Stable-diffusion" @@ -134,6 +134,8 @@ def load_model_weights(model, checkpoint_file, sd_model_hash): if not shared.cmd_opts.no_half: model.half() + devices.dtype = torch.float32 if shared.cmd_opts.no_half else torch.float16 + model.sd_model_hash = sd_model_hash model.sd_model_checkpint = checkpoint_file diff --git a/modules/shared.py b/modules/shared.py index ac968b2d..ac0bc480 100644 --- a/modules/shared.py +++ b/modules/shared.py @@ -78,6 +78,7 @@ class State: current_latent = None current_image = None current_image_sampling_step = 0 + textinfo = None def interrupt(self): self.interrupted = True @@ -88,7 +89,7 @@ class State: self.current_image_sampling_step = 0 def get_job_timestamp(self): - return datetime.datetime.now().strftime("%Y%m%d%H%M%S") + return datetime.datetime.now().strftime("%Y%m%d%H%M%S") # shouldn't this return job_timestamp? state = State() diff --git a/modules/textual_inversion/dataset.py b/modules/textual_inversion/dataset.py new file mode 100644 index 00000000..7e134a08 --- /dev/null +++ b/modules/textual_inversion/dataset.py @@ -0,0 +1,76 @@ +import os +import numpy as np +import PIL +import torch +from PIL import Image +from torch.utils.data import Dataset +from torchvision import transforms + +import random +import tqdm + + +class PersonalizedBase(Dataset): + def __init__(self, data_root, size=None, repeats=100, flip_p=0.5, placeholder_token="*", width=512, height=512, model=None, device=None, template_file=None): + + self.placeholder_token = placeholder_token + + self.size = size + self.width = width + self.height = height + self.flip = transforms.RandomHorizontalFlip(p=flip_p) + + self.dataset = [] + + with open(template_file, "r") as file: + lines = [x.strip() for x in file.readlines()] + + self.lines = lines + + assert data_root, 'dataset directory not specified' + + self.image_paths = [os.path.join(data_root, file_path) for file_path in os.listdir(data_root)] + print("Preparing dataset...") + for path in tqdm.tqdm(self.image_paths): + image = Image.open(path) + image = image.convert('RGB') + image = image.resize((self.width, self.height), PIL.Image.BICUBIC) + + filename = os.path.basename(path) + filename_tokens = os.path.splitext(filename)[0].replace('_', '-').replace(' ', '-').split('-') + filename_tokens = [token for token in filename_tokens if token.isalpha()] + + npimage = np.array(image).astype(np.uint8) + npimage = (npimage / 127.5 - 1.0).astype(np.float32) + + torchdata = torch.from_numpy(npimage).to(device=device, dtype=torch.float32) + torchdata = torch.moveaxis(torchdata, 2, 0) + + init_latent = model.get_first_stage_encoding(model.encode_first_stage(torchdata.unsqueeze(dim=0))).squeeze() + + self.dataset.append((init_latent, filename_tokens)) + + self.length = len(self.dataset) * repeats + + self.initial_indexes = np.arange(self.length) % len(self.dataset) + self.indexes = None + self.shuffle() + + def shuffle(self): + self.indexes = self.initial_indexes[torch.randperm(self.initial_indexes.shape[0])] + + def __len__(self): + return self.length + + def __getitem__(self, i): + if i % len(self.dataset) == 0: + self.shuffle() + + index = self.indexes[i % len(self.indexes)] + x, filename_tokens = self.dataset[index] + + text = random.choice(self.lines) + text = text.replace("[name]", self.placeholder_token) + text = text.replace("[filewords]", ' '.join(filename_tokens)) + + return x, text diff --git a/modules/textual_inversion/textual_inversion.py b/modules/textual_inversion/textual_inversion.py new file mode 100644 index 00000000..c0baaace --- /dev/null +++ b/modules/textual_inversion/textual_inversion.py @@ -0,0 +1,258 @@ +import os +import sys +import traceback + +import torch +import tqdm +import html +import datetime + +from modules import shared, devices, sd_hijack, processing +import modules.textual_inversion.dataset + + +class Embedding: + def __init__(self, vec, name, step=None): + self.vec = vec + self.name = name + self.step = step + self.cached_checksum = None + + def save(self, filename): + embedding_data = { + "string_to_token": {"*": 265}, + "string_to_param": {"*": self.vec}, + "name": self.name, + "step": self.step, + } + + torch.save(embedding_data, filename) + + def checksum(self): + if self.cached_checksum is not None: + return self.cached_checksum + + def const_hash(a): + r = 0 + for v in a: + r = (r * 281 ^ int(v) * 997) & 0xFFFFFFFF + return r + + self.cached_checksum = f'{const_hash(self.vec.reshape(-1) * 100) & 0xffff:04x}' + return self.cached_checksum + +class EmbeddingDatabase: + def __init__(self, embeddings_dir): + self.ids_lookup = {} + self.word_embeddings = {} + self.dir_mtime = None + self.embeddings_dir = embeddings_dir + + def register_embedding(self, embedding, model): + + self.word_embeddings[embedding.name] = embedding + + ids = model.cond_stage_model.tokenizer([embedding.name], add_special_tokens=False)['input_ids'][0] + + first_id = ids[0] + if first_id not in self.ids_lookup: + self.ids_lookup[first_id] = [] + self.ids_lookup[first_id].append((ids, embedding)) + + return embedding + + def load_textual_inversion_embeddings(self): + mt = os.path.getmtime(self.embeddings_dir) + if self.dir_mtime is not None and mt <= self.dir_mtime: + return + + self.dir_mtime = mt + self.ids_lookup.clear() + self.word_embeddings.clear() + + def process_file(path, filename): + name = os.path.splitext(filename)[0] + + data = torch.load(path, map_location="cpu") + + # textual inversion embeddings + if 'string_to_param' in data: + param_dict = data['string_to_param'] + if hasattr(param_dict, '_parameters'): + param_dict = getattr(param_dict, '_parameters') # fix for torch 1.12.1 loading saved file from torch 1.11 + assert len(param_dict) == 1, 'embedding file has multiple terms in it' + emb = next(iter(param_dict.items()))[1] + # diffuser concepts + elif type(data) == dict and type(next(iter(data.values()))) == torch.Tensor: + assert len(data.keys()) == 1, 'embedding file has multiple terms in it' + + emb = next(iter(data.values())) + if len(emb.shape) == 1: + emb = emb.unsqueeze(0) + else: + raise Exception(f"Couldn't identify {filename} as neither textual inversion embedding nor diffuser concept.") + + vec = emb.detach().to(devices.device, dtype=torch.float32) + embedding = Embedding(vec, name) + embedding.step = data.get('step', None) + self.register_embedding(embedding, shared.sd_model) + + for fn in os.listdir(self.embeddings_dir): + try: + fullfn = os.path.join(self.embeddings_dir, fn) + + if os.stat(fullfn).st_size == 0: + continue + + process_file(fullfn, fn) + except Exception: + print(f"Error loading emedding {fn}:", file=sys.stderr) + print(traceback.format_exc(), file=sys.stderr) + continue + + print(f"Loaded a total of {len(self.word_embeddings)} textual inversion embeddings.") + + def find_embedding_at_position(self, tokens, offset): + token = tokens[offset] + possible_matches = self.ids_lookup.get(token, None) + + if possible_matches is None: + return None + + for ids, embedding in possible_matches: + if tokens[offset:offset + len(ids)] == ids: + return embedding + + return None + + + +def create_embedding(name, num_vectors_per_token): + init_text = '*' + + cond_model = shared.sd_model.cond_stage_model + embedding_layer = cond_model.wrapped.transformer.text_model.embeddings + + ids = cond_model.tokenizer(init_text, max_length=num_vectors_per_token, return_tensors="pt", add_special_tokens=False)["input_ids"] + embedded = embedding_layer(ids.to(devices.device)).squeeze(0) + vec = torch.zeros((num_vectors_per_token, embedded.shape[1]), device=devices.device) + + for i in range(num_vectors_per_token): + vec[i] = embedded[i * int(embedded.shape[0]) // num_vectors_per_token] + + fn = os.path.join(shared.cmd_opts.embeddings_dir, f"{name}.pt") + assert not os.path.exists(fn), f"file {fn} already exists" + + embedding = Embedding(vec, name) + embedding.step = 0 + embedding.save(fn) + + return fn + + +def train_embedding(embedding_name, learn_rate, data_root, log_directory, steps, create_image_every, save_embedding_every, template_file): + assert embedding_name, 'embedding not selected' + + shared.state.textinfo = "Initializing textual inversion training..." + shared.state.job_count = steps + + filename = os.path.join(shared.cmd_opts.embeddings_dir, f'{embedding_name}.pt') + + log_directory = os.path.join(log_directory, datetime.datetime.now().strftime("%Y-%d-%m"), embedding_name) + + if save_embedding_every > 0: + embedding_dir = os.path.join(log_directory, "embeddings") + os.makedirs(embedding_dir, exist_ok=True) + else: + embedding_dir = None + + if create_image_every > 0: + images_dir = os.path.join(log_directory, "images") + os.makedirs(images_dir, exist_ok=True) + else: + images_dir = None + + cond_model = shared.sd_model.cond_stage_model + + shared.state.textinfo = f"Preparing dataset from {html.escape(data_root)}..." + with torch.autocast("cuda"): + ds = modules.textual_inversion.dataset.PersonalizedBase(data_root=data_root, size=512, placeholder_token=embedding_name, model=shared.sd_model, device=devices.device, template_file=template_file) + + hijack = sd_hijack.model_hijack + + embedding = hijack.embedding_db.word_embeddings[embedding_name] + embedding.vec.requires_grad = True + + optimizer = torch.optim.AdamW([embedding.vec], lr=learn_rate) + + losses = torch.zeros((32,)) + + last_saved_file = "" + last_saved_image = "" + + ititial_step = embedding.step or 0 + if ititial_step > steps: + return embedding, filename + + pbar = tqdm.tqdm(enumerate(ds), total=steps-ititial_step) + for i, (x, text) in pbar: + embedding.step = i + ititial_step + + if embedding.step > steps: + break + + if shared.state.interrupted: + break + + with torch.autocast("cuda"): + c = cond_model([text]) + loss = shared.sd_model(x.unsqueeze(0), c)[0] + + losses[embedding.step % losses.shape[0]] = loss.item() + + optimizer.zero_grad() + loss.backward() + optimizer.step() + + pbar.set_description(f"loss: {losses.mean():.7f}") + + if embedding.step > 0 and embedding_dir is not None and embedding.step % save_embedding_every == 0: + last_saved_file = os.path.join(embedding_dir, f'{embedding_name}-{embedding.step}.pt') + embedding.save(last_saved_file) + + if embedding.step > 0 and images_dir is not None and embedding.step % create_image_every == 0: + last_saved_image = os.path.join(images_dir, f'{embedding_name}-{embedding.step}.png') + + p = processing.StableDiffusionProcessingTxt2Img( + sd_model=shared.sd_model, + prompt=text, + steps=20, + do_not_save_grid=True, + do_not_save_samples=True, + ) + + processed = processing.process_images(p) + image = processed.images[0] + + shared.state.current_image = image + image.save(last_saved_image) + + last_saved_image += f", prompt: {text}" + + shared.state.job_no = embedding.step + + shared.state.textinfo = f""" +

+Loss: {losses.mean():.7f}
+Step: {embedding.step}
+Last prompt: {html.escape(text)}
+Last saved embedding: {html.escape(last_saved_file)}
+Last saved image: {html.escape(last_saved_image)}
+

+""" + + embedding.cached_checksum = None + embedding.save(filename) + + return embedding, filename + diff --git a/modules/textual_inversion/ui.py b/modules/textual_inversion/ui.py new file mode 100644 index 00000000..ce3677a9 --- /dev/null +++ b/modules/textual_inversion/ui.py @@ -0,0 +1,32 @@ +import html + +import gradio as gr + +import modules.textual_inversion.textual_inversion as ti +from modules import sd_hijack, shared + + +def create_embedding(name, nvpt): + filename = ti.create_embedding(name, nvpt) + + sd_hijack.model_hijack.embedding_db.load_textual_inversion_embeddings() + + return gr.Dropdown.update(choices=sorted(sd_hijack.model_hijack.embedding_db.word_embeddings.keys())), f"Created: {filename}", "" + + +def train_embedding(*args): + + try: + sd_hijack.undo_optimizations() + + embedding, filename = ti.train_embedding(*args) + + res = f""" +Training {'interrupted' if shared.state.interrupted else 'finished'} after {embedding.step} steps. +Embedding saved to {html.escape(filename)} +""" + return res, "" + except Exception: + raise + finally: + sd_hijack.apply_optimizations() diff --git a/modules/ui.py b/modules/ui.py index 15572bb0..57aef6ff 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -21,6 +21,7 @@ import gradio as gr import gradio.utils import gradio.routes +from modules import sd_hijack from modules.paths import script_path from modules.shared import opts, cmd_opts import modules.shared as shared @@ -32,6 +33,7 @@ import modules.gfpgan_model import modules.codeformer_model import modules.styles import modules.generation_parameters_copypaste +import modules.textual_inversion.ui # this is a fix for Windows users. Without it, javascript files will be served with text/html content-type and the bowser will not show any UI mimetypes.init() @@ -142,8 +144,8 @@ def save_files(js_data, images, index): return '', '', plaintext_to_html(f"Saved: {filenames[0]}") -def wrap_gradio_call(func): - def f(*args, **kwargs): +def wrap_gradio_call(func, extra_outputs=None): + def f(*args, extra_outputs_array=extra_outputs, **kwargs): run_memmon = opts.memmon_poll_rate > 0 and not shared.mem_mon.disabled if run_memmon: shared.mem_mon.monitor() @@ -159,7 +161,10 @@ def wrap_gradio_call(func): shared.state.job = "" shared.state.job_count = 0 - res = [None, '', f"
{plaintext_to_html(type(e).__name__+': '+str(e))}
"] + if extra_outputs_array is None: + extra_outputs_array = [None, ''] + + res = extra_outputs_array + [f"
{plaintext_to_html(type(e).__name__+': '+str(e))}
"] elapsed = time.perf_counter() - t @@ -179,6 +184,7 @@ def wrap_gradio_call(func): res[-1] += f"

Time taken: {elapsed:.2f}s

{vram_html}
" shared.state.interrupted = False + shared.state.job_count = 0 return tuple(res) @@ -187,7 +193,7 @@ def wrap_gradio_call(func): def check_progress_call(id_part): if shared.state.job_count == 0: - return "", gr_show(False), gr_show(False) + return "", gr_show(False), gr_show(False), gr_show(False) progress = 0 @@ -219,13 +225,19 @@ def check_progress_call(id_part): else: preview_visibility = gr_show(True) - return f"

{progressbar}

", preview_visibility, image + if shared.state.textinfo is not None: + textinfo_result = gr.HTML.update(value=shared.state.textinfo, visible=True) + else: + textinfo_result = gr_show(False) + + return f"

{progressbar}

", preview_visibility, image, textinfo_result def check_progress_call_initial(id_part): shared.state.job_count = -1 shared.state.current_latent = None shared.state.current_image = None + shared.state.textinfo = None return check_progress_call(id_part) @@ -399,13 +411,16 @@ def create_toprow(is_img2img): return prompt, roll, prompt_style, negative_prompt, prompt_style2, submit, interrogate, prompt_style_apply, save_style, paste -def setup_progressbar(progressbar, preview, id_part): +def setup_progressbar(progressbar, preview, id_part, textinfo=None): + if textinfo is None: + textinfo = gr.HTML(visible=False) + check_progress = gr.Button('Check progress', elem_id=f"{id_part}_check_progress", visible=False) check_progress.click( fn=lambda: check_progress_call(id_part), show_progress=False, inputs=[], - outputs=[progressbar, preview, preview], + outputs=[progressbar, preview, preview, textinfo], ) check_progress_initial = gr.Button('Check progress (first)', elem_id=f"{id_part}_check_progress_initial", visible=False) @@ -413,11 +428,14 @@ def setup_progressbar(progressbar, preview, id_part): fn=lambda: check_progress_call_initial(id_part), show_progress=False, inputs=[], - outputs=[progressbar, preview, preview], + outputs=[progressbar, preview, preview, textinfo], ) -def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): +def create_ui(wrap_gradio_gpu_call): + import modules.img2img + import modules.txt2img + with gr.Blocks(analytics_enabled=False) as txt2img_interface: txt2img_prompt, roll, txt2img_prompt_style, txt2img_negative_prompt, txt2img_prompt_style2, submit, _, txt2img_prompt_style_apply, txt2img_save_style, paste = create_toprow(is_img2img=False) dummy_component = gr.Label(visible=False) @@ -483,7 +501,7 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): connect_reuse_seed(subseed, reuse_subseed, generation_info, dummy_component, is_subseed=True) txt2img_args = dict( - fn=txt2img, + fn=wrap_gradio_gpu_call(modules.txt2img.txt2img), _js="submit", inputs=[ txt2img_prompt, @@ -675,7 +693,7 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): ) img2img_args = dict( - fn=img2img, + fn=wrap_gradio_gpu_call(modules.img2img.img2img), _js="submit_img2img", inputs=[ dummy_component, @@ -828,7 +846,7 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): open_extras_folder = gr.Button('Open output directory', elem_id=button_id) submit.click( - fn=run_extras, + fn=wrap_gradio_gpu_call(modules.extras.run_extras), _js="get_extras_tab_index", inputs=[ dummy_component, @@ -878,7 +896,7 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): pnginfo_send_to_img2img = gr.Button('Send to img2img') image.change( - fn=wrap_gradio_call(run_pnginfo), + fn=wrap_gradio_call(modules.extras.run_pnginfo), inputs=[image], outputs=[html, generation_info, html2], ) @@ -887,7 +905,7 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): with gr.Row().style(equal_height=False): with gr.Column(variant='panel'): gr.HTML(value="

A merger of the two checkpoints will be generated in your checkpoint directory.

") - + with gr.Row(): primary_model_name = gr.Dropdown(modules.sd_models.checkpoint_tiles(), elem_id="modelmerger_primary_model_name", label="Primary Model Name") secondary_model_name = gr.Dropdown(modules.sd_models.checkpoint_tiles(), elem_id="modelmerger_secondary_model_name", label="Secondary Model Name") @@ -896,10 +914,96 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): interp_method = gr.Radio(choices=["Weighted Sum", "Sigmoid", "Inverse Sigmoid"], value="Weighted Sum", label="Interpolation Method") save_as_half = gr.Checkbox(value=False, label="Safe as float16") modelmerger_merge = gr.Button(elem_id="modelmerger_merge", label="Merge", variant='primary') - + with gr.Column(variant='panel'): submit_result = gr.Textbox(elem_id="modelmerger_result", show_label=False) + sd_hijack.model_hijack.embedding_db.load_textual_inversion_embeddings() + + with gr.Blocks() as textual_inversion_interface: + with gr.Row().style(equal_height=False): + with gr.Column(): + with gr.Group(): + gr.HTML(value="

Create a new embedding

") + + new_embedding_name = gr.Textbox(label="Name") + nvpt = gr.Slider(label="Number of vectors per token", minimum=1, maximum=75, step=1, value=1) + + with gr.Row(): + with gr.Column(scale=3): + gr.HTML(value="") + + with gr.Column(): + create_embedding = gr.Button(value="Create", variant='primary') + + with gr.Group(): + gr.HTML(value="

Train an embedding; must specify a directory with a set of 512x512 images

") + train_embedding_name = gr.Dropdown(label='Embedding', choices=sorted(sd_hijack.model_hijack.embedding_db.word_embeddings.keys())) + learn_rate = gr.Number(label='Learning rate', value=5.0e-03) + dataset_directory = gr.Textbox(label='Dataset directory', placeholder="Path to directory with input images") + log_directory = gr.Textbox(label='Log directory', placeholder="Path to directory where to write outputs", value="textual_inversion") + template_file = gr.Textbox(label='Prompt template file', value=os.path.join(script_path, "textual_inversion_templates", "style_filewords.txt")) + steps = gr.Number(label='Max steps', value=100000, precision=0) + create_image_every = gr.Number(label='Save an image to log directory every N steps, 0 to disable', value=1000, precision=0) + save_embedding_every = gr.Number(label='Save a copy of embedding to log directory every N steps, 0 to disable', value=1000, precision=0) + + with gr.Row(): + with gr.Column(scale=2): + gr.HTML(value="") + + with gr.Column(): + with gr.Row(): + interrupt_training = gr.Button(value="Interrupt") + train_embedding = gr.Button(value="Train", variant='primary') + + with gr.Column(): + progressbar = gr.HTML(elem_id="ti_progressbar") + ti_output = gr.Text(elem_id="ti_output", value="", show_label=False) + + ti_gallery = gr.Gallery(label='Output', show_label=False, elem_id='ti_gallery').style(grid=4) + ti_preview = gr.Image(elem_id='ti_preview', visible=False) + ti_progress = gr.HTML(elem_id="ti_progress", value="") + ti_outcome = gr.HTML(elem_id="ti_error", value="") + setup_progressbar(progressbar, ti_preview, 'ti', textinfo=ti_progress) + + create_embedding.click( + fn=modules.textual_inversion.ui.create_embedding, + inputs=[ + new_embedding_name, + nvpt, + ], + outputs=[ + train_embedding_name, + ti_output, + ti_outcome, + ] + ) + + train_embedding.click( + fn=wrap_gradio_gpu_call(modules.textual_inversion.ui.train_embedding, extra_outputs=[gr.update()]), + _js="start_training_textual_inversion", + inputs=[ + train_embedding_name, + learn_rate, + dataset_directory, + log_directory, + steps, + create_image_every, + save_embedding_every, + template_file, + ], + outputs=[ + ti_output, + ti_outcome, + ] + ) + + interrupt_training.click( + fn=lambda: shared.state.interrupt(), + inputs=[], + outputs=[], + ) + def create_setting_component(key): def fun(): return opts.data[key] if key in opts.data else opts.data_labels[key].default @@ -1011,6 +1115,7 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): (extras_interface, "Extras", "extras"), (pnginfo_interface, "PNG Info", "pnginfo"), (modelmerger_interface, "Checkpoint Merger", "modelmerger"), + (textual_inversion_interface, "Textual inversion", "ti"), (settings_interface, "Settings", "settings"), ] @@ -1044,11 +1149,11 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): def modelmerger(*args): try: - results = run_modelmerger(*args) + results = modules.extras.run_modelmerger(*args) except Exception as e: print("Error loading/saving model file:", file=sys.stderr) print(traceback.format_exc(), file=sys.stderr) - modules.sd_models.list_models() #To remove the potentially missing models from the list + modules.sd_models.list_models() # to remove the potentially missing models from the list return ["Error loading/saving model file. It doesn't exist or the name contains illegal characters"] + [gr.Dropdown.update(choices=modules.sd_models.checkpoint_tiles()) for _ in range(3)] return results diff --git a/style.css b/style.css index 79d6bb0d..39586bf1 100644 --- a/style.css +++ b/style.css @@ -157,7 +157,7 @@ button{ max-width: 10em; } -#txt2img_preview, #img2img_preview{ +#txt2img_preview, #img2img_preview, #ti_preview{ position: absolute; width: 320px; left: 0; @@ -172,18 +172,18 @@ button{ } @media screen and (min-width: 768px) { - #txt2img_preview, #img2img_preview { + #txt2img_preview, #img2img_preview, #ti_preview { position: absolute; } } @media screen and (max-width: 767px) { - #txt2img_preview, #img2img_preview { + #txt2img_preview, #img2img_preview, #ti_preview { position: relative; } } -#txt2img_preview div.left-0.top-0, #img2img_preview div.left-0.top-0{ +#txt2img_preview div.left-0.top-0, #img2img_preview div.left-0.top-0, #ti_preview div.left-0.top-0{ display: none; } @@ -247,7 +247,7 @@ input[type="range"]{ #txt2img_negative_prompt, #img2img_negative_prompt{ } -#txt2img_progressbar, #img2img_progressbar{ +#txt2img_progressbar, #img2img_progressbar, #ti_progressbar{ position: absolute; z-index: 1000; right: 0; diff --git a/textual_inversion_templates/style.txt b/textual_inversion_templates/style.txt new file mode 100644 index 00000000..15af2d6b --- /dev/null +++ b/textual_inversion_templates/style.txt @@ -0,0 +1,19 @@ +a painting, art by [name] +a rendering, art by [name] +a cropped painting, art by [name] +the painting, art by [name] +a clean painting, art by [name] +a dirty painting, art by [name] +a dark painting, art by [name] +a picture, art by [name] +a cool painting, art by [name] +a close-up painting, art by [name] +a bright painting, art by [name] +a cropped painting, art by [name] +a good painting, art by [name] +a close-up painting, art by [name] +a rendition, art by [name] +a nice painting, art by [name] +a small painting, art by [name] +a weird painting, art by [name] +a large painting, art by [name] diff --git a/textual_inversion_templates/style_filewords.txt b/textual_inversion_templates/style_filewords.txt new file mode 100644 index 00000000..b3a8159a --- /dev/null +++ b/textual_inversion_templates/style_filewords.txt @@ -0,0 +1,19 @@ +a painting of [filewords], art by [name] +a rendering of [filewords], art by [name] +a cropped painting of [filewords], art by [name] +the painting of [filewords], art by [name] +a clean painting of [filewords], art by [name] +a dirty painting of [filewords], art by [name] +a dark painting of [filewords], art by [name] +a picture of [filewords], art by [name] +a cool painting of [filewords], art by [name] +a close-up painting of [filewords], art by [name] +a bright painting of [filewords], art by [name] +a cropped painting of [filewords], art by [name] +a good painting of [filewords], art by [name] +a close-up painting of [filewords], art by [name] +a rendition of [filewords], art by [name] +a nice painting of [filewords], art by [name] +a small painting of [filewords], art by [name] +a weird painting of [filewords], art by [name] +a large painting of [filewords], art by [name] diff --git a/textual_inversion_templates/subject.txt b/textual_inversion_templates/subject.txt new file mode 100644 index 00000000..79f36aa0 --- /dev/null +++ b/textual_inversion_templates/subject.txt @@ -0,0 +1,27 @@ +a photo of a [name] +a rendering of a [name] +a cropped photo of the [name] +the photo of a [name] +a photo of a clean [name] +a photo of a dirty [name] +a dark photo of the [name] +a photo of my [name] +a photo of the cool [name] +a close-up photo of a [name] +a bright photo of the [name] +a cropped photo of a [name] +a photo of the [name] +a good photo of the [name] +a photo of one [name] +a close-up photo of the [name] +a rendition of the [name] +a photo of the clean [name] +a rendition of a [name] +a photo of a nice [name] +a good photo of a [name] +a photo of the nice [name] +a photo of the small [name] +a photo of the weird [name] +a photo of the large [name] +a photo of a cool [name] +a photo of a small [name] diff --git a/textual_inversion_templates/subject_filewords.txt b/textual_inversion_templates/subject_filewords.txt new file mode 100644 index 00000000..008652a6 --- /dev/null +++ b/textual_inversion_templates/subject_filewords.txt @@ -0,0 +1,27 @@ +a photo of a [name], [filewords] +a rendering of a [name], [filewords] +a cropped photo of the [name], [filewords] +the photo of a [name], [filewords] +a photo of a clean [name], [filewords] +a photo of a dirty [name], [filewords] +a dark photo of the [name], [filewords] +a photo of my [name], [filewords] +a photo of the cool [name], [filewords] +a close-up photo of a [name], [filewords] +a bright photo of the [name], [filewords] +a cropped photo of a [name], [filewords] +a photo of the [name], [filewords] +a good photo of the [name], [filewords] +a photo of one [name], [filewords] +a close-up photo of the [name], [filewords] +a rendition of the [name], [filewords] +a photo of the clean [name], [filewords] +a rendition of a [name], [filewords] +a photo of a nice [name], [filewords] +a good photo of a [name], [filewords] +a photo of the nice [name], [filewords] +a photo of the small [name], [filewords] +a photo of the weird [name], [filewords] +a photo of the large [name], [filewords] +a photo of a cool [name], [filewords] +a photo of a small [name], [filewords] diff --git a/webui.py b/webui.py index b8cccd54..19fdcdd4 100644 --- a/webui.py +++ b/webui.py @@ -12,7 +12,6 @@ import modules.bsrgan_model as bsrgan import modules.extras import modules.face_restoration import modules.gfpgan_model as gfpgan -import modules.img2img import modules.ldsr_model as ldsr import modules.lowvram import modules.realesrgan_model as realesrgan @@ -21,7 +20,6 @@ import modules.sd_hijack import modules.sd_models import modules.shared as shared import modules.swinir_model as swinir -import modules.txt2img import modules.ui from modules import modelloader from modules.paths import script_path @@ -46,7 +44,7 @@ def wrap_queued_call(func): return f -def wrap_gradio_gpu_call(func): +def wrap_gradio_gpu_call(func, extra_outputs=None): def f(*args, **kwargs): devices.torch_gc() @@ -58,6 +56,7 @@ def wrap_gradio_gpu_call(func): shared.state.current_image = None shared.state.current_image_sampling_step = 0 shared.state.interrupted = False + shared.state.textinfo = None with queue_lock: res = func(*args, **kwargs) @@ -69,7 +68,7 @@ def wrap_gradio_gpu_call(func): return res - return modules.ui.wrap_gradio_call(f) + return modules.ui.wrap_gradio_call(f, extra_outputs=extra_outputs) modules.scripts.load_scripts(os.path.join(script_path, "scripts")) @@ -86,13 +85,7 @@ def webui(): signal.signal(signal.SIGINT, sigint_handler) - demo = modules.ui.create_ui( - txt2img=wrap_gradio_gpu_call(modules.txt2img.txt2img), - img2img=wrap_gradio_gpu_call(modules.img2img.img2img), - run_extras=wrap_gradio_gpu_call(modules.extras.run_extras), - run_pnginfo=modules.extras.run_pnginfo, - run_modelmerger=modules.extras.run_modelmerger - ) + demo = modules.ui.create_ui(wrap_gradio_gpu_call=wrap_gradio_gpu_call) demo.launch( share=cmd_opts.share, -- cgit v1.2.3 From 88ec0cf5571883d84abd09196652b3679e359f2e Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Sun, 2 Oct 2022 19:40:51 +0300 Subject: fix for incorrect embedding token length calculation (will break seeds that use embeddings, you're welcome!) add option to input initialization text for embeddings --- modules/sd_hijack.py | 8 ++++---- modules/textual_inversion/textual_inversion.py | 13 +++++-------- modules/textual_inversion/ui.py | 4 ++-- modules/ui.py | 2 ++ 4 files changed, 13 insertions(+), 14 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/sd_hijack.py b/modules/sd_hijack.py index fd57e5c5..3fa06242 100644 --- a/modules/sd_hijack.py +++ b/modules/sd_hijack.py @@ -130,7 +130,7 @@ class FrozenCLIPEmbedderWithCustomWords(torch.nn.Module): while i < len(tokens): token = tokens[i] - embedding = self.hijack.embedding_db.find_embedding_at_position(tokens, i) + embedding, embedding_length_in_tokens = self.hijack.embedding_db.find_embedding_at_position(tokens, i) if embedding is None: remade_tokens.append(token) @@ -142,7 +142,7 @@ class FrozenCLIPEmbedderWithCustomWords(torch.nn.Module): remade_tokens += [0] * emb_len multipliers += [weight] * emb_len used_custom_terms.append((embedding.name, embedding.checksum())) - i += emb_len + i += embedding_length_in_tokens if len(remade_tokens) > maxlen - 2: vocab = {v: k for k, v in self.wrapped.tokenizer.get_vocab().items()} @@ -213,7 +213,7 @@ class FrozenCLIPEmbedderWithCustomWords(torch.nn.Module): while i < len(tokens): token = tokens[i] - embedding = self.hijack.embedding_db.find_embedding_at_position(tokens, i) + embedding, embedding_length_in_tokens = self.hijack.embedding_db.find_embedding_at_position(tokens, i) mult_change = self.token_mults.get(token) if opts.enable_emphasis else None if mult_change is not None: @@ -229,7 +229,7 @@ class FrozenCLIPEmbedderWithCustomWords(torch.nn.Module): remade_tokens += [0] * emb_len multipliers += [mult] * emb_len used_custom_terms.append((embedding.name, embedding.checksum())) - i += emb_len + i += embedding_length_in_tokens if len(remade_tokens) > maxlen - 2: vocab = {v: k for k, v in self.wrapped.tokenizer.get_vocab().items()} diff --git a/modules/textual_inversion/textual_inversion.py b/modules/textual_inversion/textual_inversion.py index c0baaace..0c50161d 100644 --- a/modules/textual_inversion/textual_inversion.py +++ b/modules/textual_inversion/textual_inversion.py @@ -117,24 +117,21 @@ class EmbeddingDatabase: possible_matches = self.ids_lookup.get(token, None) if possible_matches is None: - return None + return None, None for ids, embedding in possible_matches: if tokens[offset:offset + len(ids)] == ids: - return embedding + return embedding, len(ids) - return None + return None, None - -def create_embedding(name, num_vectors_per_token): - init_text = '*' - +def create_embedding(name, num_vectors_per_token, init_text='*'): cond_model = shared.sd_model.cond_stage_model embedding_layer = cond_model.wrapped.transformer.text_model.embeddings ids = cond_model.tokenizer(init_text, max_length=num_vectors_per_token, return_tensors="pt", add_special_tokens=False)["input_ids"] - embedded = embedding_layer(ids.to(devices.device)).squeeze(0) + embedded = embedding_layer.token_embedding.wrapped(ids.to(devices.device)).squeeze(0) vec = torch.zeros((num_vectors_per_token, embedded.shape[1]), device=devices.device) for i in range(num_vectors_per_token): diff --git a/modules/textual_inversion/ui.py b/modules/textual_inversion/ui.py index ce3677a9..66c43ffb 100644 --- a/modules/textual_inversion/ui.py +++ b/modules/textual_inversion/ui.py @@ -6,8 +6,8 @@ import modules.textual_inversion.textual_inversion as ti from modules import sd_hijack, shared -def create_embedding(name, nvpt): - filename = ti.create_embedding(name, nvpt) +def create_embedding(name, initialization_text, nvpt): + filename = ti.create_embedding(name, nvpt, init_text=initialization_text) sd_hijack.model_hijack.embedding_db.load_textual_inversion_embeddings() diff --git a/modules/ui.py b/modules/ui.py index 3b81a4f7..eca50df0 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -954,6 +954,7 @@ def create_ui(wrap_gradio_gpu_call): gr.HTML(value="

Create a new embedding

") new_embedding_name = gr.Textbox(label="Name") + initialization_text = gr.Textbox(label="Initialization text", value="*") nvpt = gr.Slider(label="Number of vectors per token", minimum=1, maximum=75, step=1, value=1) with gr.Row(): @@ -997,6 +998,7 @@ def create_ui(wrap_gradio_gpu_call): fn=modules.textual_inversion.ui.create_embedding, inputs=[ new_embedding_name, + initialization_text, nvpt, ], outputs=[ -- cgit v1.2.3 From e808096cf641d868f88465515d70d40fc46125d4 Mon Sep 17 00:00:00 2001 From: DepFA <35278260+dfaker@users.noreply.github.com> Date: Sun, 2 Oct 2022 19:26:06 +0100 Subject: correct indent --- modules/scripts.py | 48 +++++++++++++++++++++++++----------------------- modules/ui.py | 25 ++++++++++++------------- 2 files changed, 37 insertions(+), 36 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/scripts.py b/modules/scripts.py index 788397f5..45230f9a 100644 --- a/modules/scripts.py +++ b/modules/scripts.py @@ -163,37 +163,39 @@ class ScriptRunner: return processed def reload_sources(self): - for si,script in list(enumerate(self.scripts)): - with open(script.filename, "r", encoding="utf8") as file: - args_from = script.args_from - args_to = script.args_to - filename = script.filename - text = file.read() + for si, script in list(enumerate(self.scripts)): + with open(script.filename, "r", encoding="utf8") as file: + args_from = script.args_from + args_to = script.args_to + filename = script.filename + text = file.read() - from types import ModuleType - compiled = compile(text, filename, 'exec') - module = ModuleType(script.filename) - exec(compiled, module.__dict__) + from types import ModuleType - for key, script_class in module.__dict__.items(): - if type(script_class) == type and issubclass(script_class, Script): - self.scripts[si] = script_class() - self.scripts[si].filename = filename - self.scripts[si].args_from = args_from - self.scripts[si].args_to = args_to + compiled = compile(text, filename, 'exec') + module = ModuleType(script.filename) + exec(compiled, module.__dict__) + + for key, script_class in module.__dict__.items(): + if type(script_class) == type and issubclass(script_class, Script): + self.scripts[si] = script_class() + self.scripts[si].filename = filename + self.scripts[si].args_from = args_from + self.scripts[si].args_to = args_to scripts_txt2img = ScriptRunner() scripts_img2img = ScriptRunner() def reload_script_body_only(): - scripts_txt2img.reload_sources() - scripts_img2img.reload_sources() + scripts_txt2img.reload_sources() + scripts_img2img.reload_sources() + def reload_scripts(basedir): - global scripts_txt2img,scripts_img2img + global scripts_txt2img, scripts_img2img - scripts_data.clear() - load_scripts(basedir) + scripts_data.clear() + load_scripts(basedir) - scripts_txt2img = ScriptRunner() - scripts_img2img = ScriptRunner() + scripts_txt2img = ScriptRunner() + scripts_img2img = ScriptRunner() diff --git a/modules/ui.py b/modules/ui.py index 963a2c61..6b30f84b 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1003,12 +1003,12 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): ) with gr.Row(): - reload_script_bodies = gr.Button(value='Reload custom script bodies (No ui updates, No restart)', variant='secondary') - restart_gradio = gr.Button(value='Restart Gradio and Refresh components (Custom Scripts, ui.py, js and css only)', variant='primary') + reload_script_bodies = gr.Button(value='Reload custom script bodies (No ui updates, No restart)', variant='secondary') + restart_gradio = gr.Button(value='Restart Gradio and Refresh components (Custom Scripts, ui.py, js and css only)', variant='primary') def reload_scripts(): - modules.scripts.reload_script_body_only() + modules.scripts.reload_script_body_only() reload_script_bodies.click( fn=reload_scripts, @@ -1018,7 +1018,7 @@ def create_ui(txt2img, img2img, run_extras, run_pnginfo, run_modelmerger): ) def request_restart(): - settings_interface.gradio_ref.do_restart = True + settings_interface.gradio_ref.do_restart = True restart_gradio.click( fn=request_restart, @@ -1234,12 +1234,11 @@ for filename in sorted(os.listdir(jsdir)): if 'gradio_routes_templates_response' not in globals(): - def template_response(*args, **kwargs): - res = gradio_routes_templates_response(*args, **kwargs) - res.body = res.body.replace(b'', f'{javascript}'.encode("utf8")) - res.init_headers() - return res - - gradio_routes_templates_response = gradio.routes.templates.TemplateResponse - gradio.routes.templates.TemplateResponse = template_response - + def template_response(*args, **kwargs): + res = gradio_routes_templates_response(*args, **kwargs) + res.body = res.body.replace(b'', f'{javascript}'.encode("utf8")) + res.init_headers() + return res + + gradio_routes_templates_response = gradio.routes.templates.TemplateResponse + gradio.routes.templates.TemplateResponse = template_response -- cgit v1.2.3 From 91f327f22bb2feb780c424c74723cc0629dc34a1 Mon Sep 17 00:00:00 2001 From: Lopyter Date: Sun, 2 Oct 2022 18:15:31 +0200 Subject: make save to dirs optional for imgs saved from ui --- modules/shared.py | 1 + modules/ui.py | 2 +- 2 files changed, 2 insertions(+), 1 deletion(-) (limited to 'modules/ui.py') diff --git a/modules/shared.py b/modules/shared.py index 1bf7a6c1..785e7af6 100644 --- a/modules/shared.py +++ b/modules/shared.py @@ -173,6 +173,7 @@ options_templates.update(options_section(('saving-to-dirs', "Saving to a directo "grid_save_to_dirs": OptionInfo(False, "Save grids to subdirectory"), "directories_filename_pattern": OptionInfo("", "Directory name pattern"), "directories_max_prompt_words": OptionInfo(8, "Max prompt words", gr.Slider, {"minimum": 1, "maximum": 20, "step": 1}), + "use_save_to_dirs_for_ui": OptionInfo(False, "Use \"Save images to a subdirectory\" option for images saved from UI"), })) options_templates.update(options_section(('upscaling', "Upscaling"), { diff --git a/modules/ui.py b/modules/ui.py index 78a15d83..8912deff 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -113,7 +113,7 @@ def save_files(js_data, images, index): p = MyObject(data) path = opts.outdir_save - save_to_dirs = opts.save_to_dirs + save_to_dirs = opts.use_save_to_dirs_for_ui if save_to_dirs: dirname = apply_filename_pattern(opts.directories_filename_pattern or "[prompt_words]", p, p.seed, p.prompt) -- cgit v1.2.3 From c7543d4940da672d970124ae8f2fec9de7bdc1da Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Sun, 2 Oct 2022 22:41:21 +0300 Subject: preprocessing for textual inversion added --- modules/interrogate.py | 1 + modules/textual_inversion/preprocess.py | 75 ++++++++++++++++++++++++++ modules/textual_inversion/textual_inversion.py | 1 + modules/textual_inversion/ui.py | 14 +++-- modules/ui.py | 36 +++++++++++++ 5 files changed, 124 insertions(+), 3 deletions(-) create mode 100644 modules/textual_inversion/preprocess.py (limited to 'modules/ui.py') diff --git a/modules/interrogate.py b/modules/interrogate.py index f62a4745..eed87144 100644 --- a/modules/interrogate.py +++ b/modules/interrogate.py @@ -21,6 +21,7 @@ Category = namedtuple("Category", ["name", "topn", "items"]) re_topn = re.compile(r"\.top(\d+)\.") + class InterrogateModels: blip_model = None clip_model = None diff --git a/modules/textual_inversion/preprocess.py b/modules/textual_inversion/preprocess.py new file mode 100644 index 00000000..209e928f --- /dev/null +++ b/modules/textual_inversion/preprocess.py @@ -0,0 +1,75 @@ +import os +from PIL import Image, ImageOps +import tqdm + +from modules import shared, images + + +def preprocess(process_src, process_dst, process_flip, process_split, process_caption): + size = 512 + src = os.path.abspath(process_src) + dst = os.path.abspath(process_dst) + + assert src != dst, 'same directory specified as source and desitnation' + + os.makedirs(dst, exist_ok=True) + + files = os.listdir(src) + + shared.state.textinfo = "Preprocessing..." + shared.state.job_count = len(files) + + if process_caption: + shared.interrogator.load() + + def save_pic_with_caption(image, index): + if process_caption: + caption = "-" + shared.interrogator.generate_caption(image) + else: + caption = "" + + image.save(os.path.join(dst, f"{index:05}-{subindex[0]}{caption}.png")) + subindex[0] += 1 + + def save_pic(image, index): + save_pic_with_caption(image, index) + + if process_flip: + save_pic_with_caption(ImageOps.mirror(image), index) + + for index, imagefile in enumerate(tqdm.tqdm(files)): + subindex = [0] + filename = os.path.join(src, imagefile) + img = Image.open(filename).convert("RGB") + + if shared.state.interrupted: + break + + ratio = img.height / img.width + is_tall = ratio > 1.35 + is_wide = ratio < 1 / 1.35 + + if process_split and is_tall: + img = img.resize((size, size * img.height // img.width)) + + top = img.crop((0, 0, size, size)) + save_pic(top, index) + + bot = img.crop((0, img.height - size, size, img.height)) + save_pic(bot, index) + elif process_split and is_wide: + img = img.resize((size * img.width // img.height, size)) + + left = img.crop((0, 0, size, size)) + save_pic(left, index) + + right = img.crop((img.width - size, 0, img.width, size)) + save_pic(right, index) + else: + img = images.resize_image(1, img, size, size) + save_pic(img, index) + + shared.state.nextjob() + + if process_caption: + shared.interrogator.send_blip_to_ram() diff --git a/modules/textual_inversion/textual_inversion.py b/modules/textual_inversion/textual_inversion.py index 1183aab7..d4e250d8 100644 --- a/modules/textual_inversion/textual_inversion.py +++ b/modules/textual_inversion/textual_inversion.py @@ -7,6 +7,7 @@ import tqdm import html import datetime + from modules import shared, devices, sd_hijack, processing, sd_models import modules.textual_inversion.dataset diff --git a/modules/textual_inversion/ui.py b/modules/textual_inversion/ui.py index 633037d8..f19ac5e0 100644 --- a/modules/textual_inversion/ui.py +++ b/modules/textual_inversion/ui.py @@ -2,24 +2,31 @@ import html import gradio as gr -import modules.textual_inversion.textual_inversion as ti +import modules.textual_inversion.textual_inversion +import modules.textual_inversion.preprocess from modules import sd_hijack, shared def create_embedding(name, initialization_text, nvpt): - filename = ti.create_embedding(name, nvpt, init_text=initialization_text) + filename = modules.textual_inversion.textual_inversion.create_embedding(name, nvpt, init_text=initialization_text) sd_hijack.model_hijack.embedding_db.load_textual_inversion_embeddings() return gr.Dropdown.update(choices=sorted(sd_hijack.model_hijack.embedding_db.word_embeddings.keys())), f"Created: {filename}", "" +def preprocess(*args): + modules.textual_inversion.preprocess.preprocess(*args) + + return "Preprocessing finished.", "" + + def train_embedding(*args): try: sd_hijack.undo_optimizations() - embedding, filename = ti.train_embedding(*args) + embedding, filename = modules.textual_inversion.textual_inversion.train_embedding(*args) res = f""" Training {'interrupted' if shared.state.interrupted else 'finished'} at {embedding.step} steps. @@ -30,3 +37,4 @@ Embedding saved to {html.escape(filename)} raise finally: sd_hijack.apply_optimizations() + diff --git a/modules/ui.py b/modules/ui.py index 8912deff..e7bde53b 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -961,6 +961,8 @@ def create_ui(wrap_gradio_gpu_call): with gr.Row().style(equal_height=False): with gr.Column(): with gr.Group(): + gr.HTML(value="

See wiki for detailed explanation.

") + gr.HTML(value="

Create a new embedding

") new_embedding_name = gr.Textbox(label="Name") @@ -974,6 +976,24 @@ def create_ui(wrap_gradio_gpu_call): with gr.Column(): create_embedding = gr.Button(value="Create", variant='primary') + with gr.Group(): + gr.HTML(value="

Preprocess images

") + + process_src = gr.Textbox(label='Source directory') + process_dst = gr.Textbox(label='Destination directory') + + with gr.Row(): + process_flip = gr.Checkbox(label='Flip') + process_split = gr.Checkbox(label='Split into two') + process_caption = gr.Checkbox(label='Add caption') + + with gr.Row(): + with gr.Column(scale=3): + gr.HTML(value="") + + with gr.Column(): + run_preprocess = gr.Button(value="Preprocess", variant='primary') + with gr.Group(): gr.HTML(value="

Train an embedding; must specify a directory with a set of 512x512 images

") train_embedding_name = gr.Dropdown(label='Embedding', choices=sorted(sd_hijack.model_hijack.embedding_db.word_embeddings.keys())) @@ -1018,6 +1038,22 @@ def create_ui(wrap_gradio_gpu_call): ] ) + run_preprocess.click( + fn=wrap_gradio_gpu_call(modules.textual_inversion.ui.preprocess, extra_outputs=[gr.update()]), + _js="start_training_textual_inversion", + inputs=[ + process_src, + process_dst, + process_flip, + process_split, + process_caption, + ], + outputs=[ + ti_output, + ti_outcome, + ], + ) + train_embedding.click( fn=wrap_gradio_gpu_call(modules.textual_inversion.ui.train_embedding, extra_outputs=[gr.update()]), _js="start_training_textual_inversion", -- cgit v1.2.3 From 6785331e22d6a488fbf5905fab56d7fec867e038 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Sun, 2 Oct 2022 22:59:01 +0300 Subject: keep textual inversion dataset latents in CPU memory to save a bit of VRAM --- modules/textual_inversion/dataset.py | 2 ++ modules/textual_inversion/textual_inversion.py | 3 +++ modules/ui.py | 4 ++-- 3 files changed, 7 insertions(+), 2 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/textual_inversion/dataset.py b/modules/textual_inversion/dataset.py index 7e134a08..e8394ff6 100644 --- a/modules/textual_inversion/dataset.py +++ b/modules/textual_inversion/dataset.py @@ -8,6 +8,7 @@ from torchvision import transforms import random import tqdm +from modules import devices class PersonalizedBase(Dataset): @@ -47,6 +48,7 @@ class PersonalizedBase(Dataset): torchdata = torch.moveaxis(torchdata, 2, 0) init_latent = model.get_first_stage_encoding(model.encode_first_stage(torchdata.unsqueeze(dim=0))).squeeze() + init_latent = init_latent.to(devices.cpu) self.dataset.append((init_latent, filename_tokens)) diff --git a/modules/textual_inversion/textual_inversion.py b/modules/textual_inversion/textual_inversion.py index d4e250d8..8686f534 100644 --- a/modules/textual_inversion/textual_inversion.py +++ b/modules/textual_inversion/textual_inversion.py @@ -212,7 +212,10 @@ def train_embedding(embedding_name, learn_rate, data_root, log_directory, steps, with torch.autocast("cuda"): c = cond_model([text]) + + x = x.to(devices.device) loss = shared.sd_model(x.unsqueeze(0), c)[0] + del x losses[embedding.step % losses.shape[0]] = loss.item() diff --git a/modules/ui.py b/modules/ui.py index e7bde53b..d9d02ece 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1002,8 +1002,8 @@ def create_ui(wrap_gradio_gpu_call): log_directory = gr.Textbox(label='Log directory', placeholder="Path to directory where to write outputs", value="textual_inversion") template_file = gr.Textbox(label='Prompt template file', value=os.path.join(script_path, "textual_inversion_templates", "style_filewords.txt")) steps = gr.Number(label='Max steps', value=100000, precision=0) - create_image_every = gr.Number(label='Save an image to log directory every N steps, 0 to disable', value=1000, precision=0) - save_embedding_every = gr.Number(label='Save a copy of embedding to log directory every N steps, 0 to disable', value=1000, precision=0) + create_image_every = gr.Number(label='Save an image to log directory every N steps, 0 to disable', value=500, precision=0) + save_embedding_every = gr.Number(label='Save a copy of embedding to log directory every N steps, 0 to disable', value=500, precision=0) with gr.Row(): with gr.Column(scale=2): -- cgit v1.2.3 From e615d4f9d101e2712c7c2d0e3e8feb19cb430c74 Mon Sep 17 00:00:00 2001 From: Hanusz Leszek Date: Sun, 2 Oct 2022 21:08:23 +0200 Subject: Convert folder icon surrogate pair to valid utf8 --- javascript/hints.js | 2 +- modules/ui.py | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) (limited to 'modules/ui.py') diff --git a/javascript/hints.js b/javascript/hints.js index 84694eeb..e72e9338 100644 --- a/javascript/hints.js +++ b/javascript/hints.js @@ -15,7 +15,7 @@ titles = { "\u267b\ufe0f": "Reuse seed from last generation, mostly useful if it was randomed", "\u{1f3a8}": "Add a random artist to the prompt.", "\u2199\ufe0f": "Read generation parameters from prompt into user interface.", - "\uD83D\uDCC2": "Open images output directory", + "\u{1f4c2}": "Open images output directory", "Inpaint a part of image": "Draw a mask over an image, and the script will regenerate the masked area with content according to prompt", "SD upscale": "Upscale image normally, split result into tiles, improve each tile using img2img, merge whole image back", diff --git a/modules/ui.py b/modules/ui.py index d9d02ece..16432151 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -69,7 +69,7 @@ random_symbol = '\U0001f3b2\ufe0f' # 🎲️ reuse_symbol = '\u267b\ufe0f' # ♻️ art_symbol = '\U0001f3a8' # 🎨 paste_symbol = '\u2199\ufe0f' # ↙ -folder_symbol = '\uD83D\uDCC2' +folder_symbol = '\U0001f4c2' # 📂 def plaintext_to_html(text): text = "

" + "
\n".join([f"{html.escape(x)}" for x in text.split('\n')]) + "

" -- cgit v1.2.3 From 43a74fa595003321200a40bd2431e56c245e75ed Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Mon, 3 Oct 2022 11:48:19 +0300 Subject: batch processing for img2img with an empty output directory, by request --- modules/img2img.py | 7 +++++-- modules/ui.py | 2 +- 2 files changed, 6 insertions(+), 3 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/img2img.py b/modules/img2img.py index f4455c90..2ff8e261 100644 --- a/modules/img2img.py +++ b/modules/img2img.py @@ -23,8 +23,10 @@ def process_batch(p, input_dir, output_dir, args): print(f"Will process {len(images)} images, creating {p.n_iter * p.batch_size} new images for each.") + save_normally = output_dir == '' + p.do_not_save_grid = True - p.do_not_save_samples = True + p.do_not_save_samples = not save_normally state.job_count = len(images) * p.n_iter @@ -48,7 +50,8 @@ def process_batch(p, input_dir, output_dir, args): left, right = os.path.splitext(filename) filename = f"{left}-{n}{right}" - processed_image.save(os.path.join(output_dir, filename)) + if not save_normally: + processed_image.save(os.path.join(output_dir, filename)) def img2img(mode: int, prompt: str, negative_prompt: str, prompt_style: str, prompt_style2: str, init_img, init_img_with_mask, init_img_inpaint, init_mask_inpaint, mask_mode, steps: int, sampler_index: int, mask_blur: int, inpainting_fill: int, restore_faces: bool, tiling: bool, n_iter: int, batch_size: int, cfg_scale: float, denoising_strength: float, seed: int, subseed: int, subseed_strength: float, seed_resize_from_h: int, seed_resize_from_w: int, seed_enable_extras: bool, height: int, width: int, resize_mode: int, inpaint_full_res: bool, inpaint_full_res_padding: int, inpainting_mask_invert: int, img2img_batch_input_dir: str, img2img_batch_output_dir: str, *args): diff --git a/modules/ui.py b/modules/ui.py index 16432151..55f7aa95 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -658,7 +658,7 @@ def create_ui(wrap_gradio_gpu_call): with gr.TabItem('Batch img2img', id='batch'): hidden = '
Disabled when launched with --hide-ui-dir-config.' if shared.cmd_opts.hide_ui_dir_config else '' - gr.HTML(f"

Process images in a directory on the same machine where the server is running.{hidden}

") + gr.HTML(f"

Process images in a directory on the same machine where the server is running.
Use an empty output directory to save pictures normally instead of writing to the output directory.{hidden}

") img2img_batch_input_dir = gr.Textbox(label="Input directory", **shared.hide_dirs) img2img_batch_output_dir = gr.Textbox(label="Output directory", **shared.hide_dirs) -- cgit v1.2.3 From d5bba20a58f43a9f984bb67b4e17f48661f6b818 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Tue, 4 Oct 2022 14:35:12 +0300 Subject: ignore errors in parse for purposes of token counting for #1564 --- modules/ui.py | 12 ++++++++++-- 1 file changed, 10 insertions(+), 2 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/ui.py b/modules/ui.py index 55f7aa95..20dc8c37 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -386,14 +386,22 @@ def connect_reuse_seed(seed: gr.Number, reuse_seed: gr.Button, generation_info: outputs=[seed, dummy_component] ) + def update_token_counter(text, steps): - prompt_schedules = get_learned_conditioning_prompt_schedules([text], steps) + try: + prompt_schedules = get_learned_conditioning_prompt_schedules([text], steps) + except Exception: + # a parsing error can happen here during typing, and we don't want to bother the user with + # messages related to it in console + prompt_schedules = [[[steps, text]]] + flat_prompts = reduce(lambda list1, list2: list1+list2, prompt_schedules) - prompts = [prompt_text for step,prompt_text in flat_prompts] + prompts = [prompt_text for step, prompt_text in flat_prompts] tokens, token_count, max_length = max([model_hijack.tokenize(prompt) for prompt in prompts], key=lambda args: args[1]) style_class = ' class="red"' if (token_count > max_length) else "" return f"{token_count}/{max_length}" + def create_toprow(is_img2img): id_part = "img2img" if is_img2img else "txt2img" -- cgit v1.2.3 From b32852ef037251eb3d846af76e2965594e1ac7a5 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Tue, 4 Oct 2022 20:49:54 +0300 Subject: add editor to img2img --- modules/shared.py | 1 + modules/ui.py | 2 +- style.css | 4 ++++ 3 files changed, 6 insertions(+), 1 deletion(-) (limited to 'modules/ui.py') diff --git a/modules/shared.py b/modules/shared.py index ff4e5fa3..e52c9b1d 100644 --- a/modules/shared.py +++ b/modules/shared.py @@ -55,6 +55,7 @@ parser.add_argument("--hide-ui-dir-config", action='store_true', help="hide dire parser.add_argument("--ui-settings-file", type=str, help="filename to use for ui settings", default=os.path.join(script_path, 'config.json')) parser.add_argument("--gradio-debug", action='store_true', help="launch gradio with --debug option") parser.add_argument("--gradio-auth", type=str, help='set gradio authentication like "username:password"; or comma-delimit multiple like "u1:p1,u2:p2,u3:p3"', default=None) +parser.add_argument("--gradio-img2img-tool", type=str, help='gradio image uploader tool: can be either editor for ctopping, or color-sketch for drawing', choices=["color-sketch", "editor"], default="color-sketch") parser.add_argument("--opt-channelslast", action='store_true', help="change memory type for stable diffusion to channels last") parser.add_argument("--styles-file", type=str, help="filename to use for styles", default=os.path.join(script_path, 'styles.csv')) parser.add_argument("--autolaunch", action='store_true', help="open the webui URL in the system's default browser upon launch", default=False) diff --git a/modules/ui.py b/modules/ui.py index 20dc8c37..6cd6761b 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -644,7 +644,7 @@ def create_ui(wrap_gradio_gpu_call): with gr.Tabs(elem_id="mode_img2img") as tabs_img2img_mode: with gr.TabItem('img2img', id='img2img'): - init_img = gr.Image(label="Image for img2img", show_label=False, source="upload", interactive=True, type="pil") + init_img = gr.Image(label="Image for img2img", elem_id="img2img_image", show_label=False, source="upload", interactive=True, type="pil", tool=cmd_opts.gradio_img2img_tool) with gr.TabItem('Inpaint', id='inpaint'): init_img_with_mask = gr.Image(label="Image for inpainting with mask", show_label=False, elem_id="img2maskimg", source="upload", interactive=True, type="pil", tool="sketch", image_mode="RGBA") diff --git a/style.css b/style.css index 39586bf1..e8f4cb75 100644 --- a/style.css +++ b/style.css @@ -403,3 +403,7 @@ input[type="range"]{ .red { color: red; } + +#img2img_image div.h-60{ + height: 480px; +} \ No newline at end of file -- cgit v1.2.3 From ef40e4cd4d383a3405e03f1da3f5b5a1820a8f53 Mon Sep 17 00:00:00 2001 From: xpscyho Date: Tue, 4 Oct 2022 15:12:38 -0400 Subject: Display time taken in mins, secs when relevant Fixes #1656 --- modules/ui.py | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) (limited to 'modules/ui.py') diff --git a/modules/ui.py b/modules/ui.py index 6cd6761b..de6342a4 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -196,6 +196,11 @@ def wrap_gradio_call(func, extra_outputs=None): res = extra_outputs_array + [f"
{plaintext_to_html(type(e).__name__+': '+str(e))}
"] elapsed = time.perf_counter() - t + elapsed_m = int(elapsed // 60) + elapsed_s = elapsed % 60 + elapsed_text = f"{elapsed_s:.2f}s" + if (elapsed_m > 0): + elapsed_text = f"{elapsed_m}m "+elapsed_text if run_memmon: mem_stats = {k: -(v//-(1024*1024)) for k, v in shared.mem_mon.stop().items()} @@ -210,7 +215,7 @@ def wrap_gradio_call(func, extra_outputs=None): vram_html = '' # last item is always HTML - res[-1] += f"

Time taken: {elapsed:.2f}s

{vram_html}
" + res[-1] += f"

Time taken: {elapsed_text}

{vram_html}
" shared.state.interrupted = False shared.state.job_count = 0 -- cgit v1.2.3 From bbdbbd36eda870cf0bd49fdf28476c78919a123e Mon Sep 17 00:00:00 2001 From: DepFA <35278260+dfaker@users.noreply.github.com> Date: Wed, 5 Oct 2022 04:43:05 +0100 Subject: shared.state.interrupt when restart is requested --- modules/ui.py | 1 + 1 file changed, 1 insertion(+) (limited to 'modules/ui.py') diff --git a/modules/ui.py b/modules/ui.py index de6342a4..523ab25b 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1210,6 +1210,7 @@ def create_ui(wrap_gradio_gpu_call): ) def request_restart(): + shared.state.interrupt() settings_interface.gradio_ref.do_restart = True restart_gradio.click( -- cgit v1.2.3 From 59a2b9e5afc27d2fda72069ca0635070535d18fe Mon Sep 17 00:00:00 2001 From: Greendayle Date: Wed, 5 Oct 2022 20:50:10 +0200 Subject: deepdanbooru interrogator --- ... your deepbooru release project folder here.txt | 0 modules/deepbooru.py | 60 ++++++++++++++++++++++ modules/ui.py | 24 +++++++-- requirements.txt | 3 ++ requirements_versions.txt | 3 ++ style.css | 7 ++- 6 files changed, 91 insertions(+), 6 deletions(-) create mode 100644 models/deepbooru/Put your deepbooru release project folder here.txt create mode 100644 modules/deepbooru.py (limited to 'modules/ui.py') diff --git a/models/deepbooru/Put your deepbooru release project folder here.txt b/models/deepbooru/Put your deepbooru release project folder here.txt new file mode 100644 index 00000000..e69de29b diff --git a/modules/deepbooru.py b/modules/deepbooru.py new file mode 100644 index 00000000..958b1c3d --- /dev/null +++ b/modules/deepbooru.py @@ -0,0 +1,60 @@ +import os.path +from concurrent.futures import ProcessPoolExecutor + +import numpy as np +import deepdanbooru as dd +import tensorflow as tf + + +def _load_tf_and_return_tags(pil_image, threshold): + this_folder = os.path.dirname(__file__) + model_path = os.path.join(this_folder, '..', 'models', 'deepbooru', 'deepdanbooru-v3-20211112-sgd-e28') + if not os.path.exists(model_path): + return "Download https://github.com/KichangKim/DeepDanbooru/releases/download/v3-20211112-sgd-e28/deepdanbooru-v3-20211112-sgd-e28.zip unpack and put into models/deepbooru" + + tags = dd.project.load_tags_from_project(model_path) + model = dd.project.load_model_from_project( + model_path, compile_model=True + ) + + width = model.input_shape[2] + height = model.input_shape[1] + image = np.array(pil_image) + image = tf.image.resize( + image, + size=(height, width), + method=tf.image.ResizeMethod.AREA, + preserve_aspect_ratio=True, + ) + image = image.numpy() # EagerTensor to np.array + image = dd.image.transform_and_pad_image(image, width, height) + image = image / 255.0 + image_shape = image.shape + image = image.reshape((1, image_shape[0], image_shape[1], image_shape[2])) + + y = model.predict(image)[0] + + result_dict = {} + + for i, tag in enumerate(tags): + result_dict[tag] = y[i] + + + + result_tags_out = [] + result_tags_print = [] + for tag in tags: + if result_dict[tag] >= threshold: + result_tags_out.append(tag) + result_tags_print.append(f'{result_dict[tag]} {tag}') + + print('\n'.join(sorted(result_tags_print, reverse=True))) + + return ', '.join(result_tags_out) + + +def get_deepbooru_tags(pil_image, threshold=0.5): + with ProcessPoolExecutor() as executor: + f = executor.submit(_load_tf_and_return_tags, pil_image, threshold) + ret = f.result() # will rethrow any exceptions + return ret \ No newline at end of file diff --git a/modules/ui.py b/modules/ui.py index 20dc8c37..ae98219a 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -23,6 +23,7 @@ import gradio.utils import gradio.routes from modules import sd_hijack +from modules.deepbooru import get_deepbooru_tags from modules.paths import script_path from modules.shared import opts, cmd_opts import modules.shared as shared @@ -312,6 +313,11 @@ def interrogate(image): return gr_show(True) if prompt is None else prompt +def interrogate_deepbooru(image): + prompt = get_deepbooru_tags(image) + return gr_show(True) if prompt is None else prompt + + def create_seed_inputs(): with gr.Row(): with gr.Box(): @@ -439,15 +445,17 @@ def create_toprow(is_img2img): outputs=[], ) - with gr.Row(): + with gr.Row(scale=1): if is_img2img: - interrogate = gr.Button('Interrogate', elem_id="interrogate") + interrogate = gr.Button('Interrogate\nCLIP', elem_id="interrogate") + deepbooru = gr.Button('Interrogate\nDeepBooru', elem_id="deepbooru") else: interrogate = None + deepbooru = None prompt_style_apply = gr.Button('Apply style', elem_id="style_apply") save_style = gr.Button('Create style', elem_id="style_create") - return prompt, roll, prompt_style, negative_prompt, prompt_style2, submit, interrogate, prompt_style_apply, save_style, paste, token_counter, token_button + return prompt, roll, prompt_style, negative_prompt, prompt_style2, submit, interrogate, deepbooru, prompt_style_apply, save_style, paste, token_counter, token_button def setup_progressbar(progressbar, preview, id_part, textinfo=None): @@ -476,7 +484,7 @@ def create_ui(wrap_gradio_gpu_call): import modules.txt2img with gr.Blocks(analytics_enabled=False) as txt2img_interface: - txt2img_prompt, roll, txt2img_prompt_style, txt2img_negative_prompt, txt2img_prompt_style2, submit, _, txt2img_prompt_style_apply, txt2img_save_style, paste, token_counter, token_button = create_toprow(is_img2img=False) + txt2img_prompt, roll, txt2img_prompt_style, txt2img_negative_prompt, txt2img_prompt_style2, submit, _, _, txt2img_prompt_style_apply, txt2img_save_style, paste, token_counter, token_button = create_toprow(is_img2img=False) dummy_component = gr.Label(visible=False) with gr.Row(elem_id='txt2img_progress_row'): @@ -628,7 +636,7 @@ def create_ui(wrap_gradio_gpu_call): token_button.click(fn=update_token_counter, inputs=[txt2img_prompt, steps], outputs=[token_counter]) with gr.Blocks(analytics_enabled=False) as img2img_interface: - img2img_prompt, roll, img2img_prompt_style, img2img_negative_prompt, img2img_prompt_style2, submit, img2img_interrogate, img2img_prompt_style_apply, img2img_save_style, paste, token_counter, token_button = create_toprow(is_img2img=True) + img2img_prompt, roll, img2img_prompt_style, img2img_negative_prompt, img2img_prompt_style2, submit, img2img_interrogate, img2img_deepbooru, img2img_prompt_style_apply, img2img_save_style, paste, token_counter, token_button = create_toprow(is_img2img=True) with gr.Row(elem_id='img2img_progress_row'): with gr.Column(scale=1): @@ -785,6 +793,12 @@ def create_ui(wrap_gradio_gpu_call): outputs=[img2img_prompt], ) + img2img_deepbooru.click( + fn=interrogate_deepbooru, + inputs=[init_img], + outputs=[img2img_prompt], + ) + save.click( fn=wrap_gradio_call(save_files), _js="(x, y, z) => [x, y, selected_gallery_index()]", diff --git a/requirements.txt b/requirements.txt index 631fe616..cab101f8 100644 --- a/requirements.txt +++ b/requirements.txt @@ -23,3 +23,6 @@ resize-right torchdiffeq kornia lark +deepdanbooru +tensorflow +tensorflow-io diff --git a/requirements_versions.txt b/requirements_versions.txt index fdff2687..811953c6 100644 --- a/requirements_versions.txt +++ b/requirements_versions.txt @@ -22,3 +22,6 @@ resize-right==0.0.2 torchdiffeq==0.2.3 kornia==0.6.7 lark==1.1.2 +git+https://github.com/KichangKim/DeepDanbooru.git@edf73df4cdaeea2cf00e9ac08bd8a9026b7a7b26#egg=deepdanbooru[tensorflow] +tensorflow==2.10.0 +tensorflow-io==0.27.0 diff --git a/style.css b/style.css index 39586bf1..2fd351f9 100644 --- a/style.css +++ b/style.css @@ -103,7 +103,12 @@ #style_apply, #style_create, #interrogate{ margin: 0.75em 0.25em 0.25em 0.25em; - min-width: 3em; + min-width: 5em; +} + +#style_apply, #style_create, #deepbooru{ + margin: 0.75em 0.25em 0.25em 0.25em; + min-width: 5em; } #style_pos_col, #style_neg_col{ -- cgit v1.2.3 From c26732fbee2a57e621ac22bf70decf7496daa4cd Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Wed, 5 Oct 2022 23:16:27 +0300 Subject: added support for AND from https://energy-based-model.github.io/Compositional-Visual-Generation-with-Composable-Diffusion-Models/ --- modules/processing.py | 2 +- modules/prompt_parser.py | 114 ++++++++++++++++++++++++++++++++++++++++++++--- modules/sd_samplers.py | 35 ++++++++++----- modules/ui.py | 6 ++- 4 files changed, 138 insertions(+), 19 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/processing.py b/modules/processing.py index bb94033b..d8c6b8d5 100644 --- a/modules/processing.py +++ b/modules/processing.py @@ -360,7 +360,7 @@ def process_images(p: StableDiffusionProcessing) -> Processed: #c = p.sd_model.get_learned_conditioning(prompts) with devices.autocast(): uc = prompt_parser.get_learned_conditioning(shared.sd_model, len(prompts) * [p.negative_prompt], p.steps) - c = prompt_parser.get_learned_conditioning(shared.sd_model, prompts, p.steps) + c = prompt_parser.get_multicond_learned_conditioning(shared.sd_model, prompts, p.steps) if len(model_hijack.comments) > 0: for comment in model_hijack.comments: diff --git a/modules/prompt_parser.py b/modules/prompt_parser.py index a3b12421..f7420daf 100644 --- a/modules/prompt_parser.py +++ b/modules/prompt_parser.py @@ -97,10 +97,26 @@ def get_learned_conditioning_prompt_schedules(prompts, steps): ScheduledPromptConditioning = namedtuple("ScheduledPromptConditioning", ["end_at_step", "cond"]) -ScheduledPromptBatch = namedtuple("ScheduledPromptBatch", ["shape", "schedules"]) def get_learned_conditioning(model, prompts, steps): + """converts a list of prompts into a list of prompt schedules - each schedule is a list of ScheduledPromptConditioning, specifying the comdition (cond), + and the sampling step at which this condition is to be replaced by the next one. + + Input: + (model, ['a red crown', 'a [blue:green:5] jeweled crown'], 20) + + Output: + [ + [ + ScheduledPromptConditioning(end_at_step=20, cond=tensor([[-0.3886, 0.0229, -0.0523, ..., -0.4901, -0.3066, 0.0674], ..., [ 0.3317, -0.5102, -0.4066, ..., 0.4119, -0.7647, -1.0160]], device='cuda:0')) + ], + [ + ScheduledPromptConditioning(end_at_step=5, cond=tensor([[-0.3886, 0.0229, -0.0522, ..., -0.4901, -0.3067, 0.0673], ..., [-0.0192, 0.3867, -0.4644, ..., 0.1135, -0.3696, -0.4625]], device='cuda:0')), + ScheduledPromptConditioning(end_at_step=20, cond=tensor([[-0.3886, 0.0229, -0.0522, ..., -0.4901, -0.3067, 0.0673], ..., [-0.7352, -0.4356, -0.7888, ..., 0.6994, -0.4312, -1.2593]], device='cuda:0')) + ] + ] + """ res = [] prompt_schedules = get_learned_conditioning_prompt_schedules(prompts, steps) @@ -123,13 +139,75 @@ def get_learned_conditioning(model, prompts, steps): cache[prompt] = cond_schedule res.append(cond_schedule) - return ScheduledPromptBatch((len(prompts),) + res[0][0].cond.shape, res) + return res + + +re_AND = re.compile(r"\bAND\b") +re_weight = re.compile(r"^(.*?)(?:\s*:\s*([-+]?\s*(?:\d+|\d*\.\d+)?))?\s*$") + + +def get_multicond_prompt_list(prompts): + res_indexes = [] + + prompt_flat_list = [] + prompt_indexes = {} + + for prompt in prompts: + subprompts = re_AND.split(prompt) + + indexes = [] + for subprompt in subprompts: + text, weight = re_weight.search(subprompt).groups() + + weight = float(weight) if weight is not None else 1.0 + + index = prompt_indexes.get(text, None) + if index is None: + index = len(prompt_flat_list) + prompt_flat_list.append(text) + prompt_indexes[text] = index + + indexes.append((index, weight)) + + res_indexes.append(indexes) + + return res_indexes, prompt_flat_list, prompt_indexes + + +class ComposableScheduledPromptConditioning: + def __init__(self, schedules, weight=1.0): + self.schedules: list[ScheduledPromptConditioning] = schedules + self.weight: float = weight + + +class MulticondLearnedConditioning: + def __init__(self, shape, batch): + self.shape: tuple = shape # the shape field is needed to send this object to DDIM/PLMS + self.batch: list[list[ComposableScheduledPromptConditioning]] = batch -def reconstruct_cond_batch(c: ScheduledPromptBatch, current_step): - param = c.schedules[0][0].cond - res = torch.zeros(c.shape, device=param.device, dtype=param.dtype) - for i, cond_schedule in enumerate(c.schedules): +def get_multicond_learned_conditioning(model, prompts, steps) -> MulticondLearnedConditioning: + """same as get_learned_conditioning, but returns a list of ScheduledPromptConditioning along with the weight objects for each prompt. + For each prompt, the list is obtained by splitting the prompt using the AND separator. + + https://energy-based-model.github.io/Compositional-Visual-Generation-with-Composable-Diffusion-Models/ + """ + + res_indexes, prompt_flat_list, prompt_indexes = get_multicond_prompt_list(prompts) + + learned_conditioning = get_learned_conditioning(model, prompt_flat_list, steps) + + res = [] + for indexes in res_indexes: + res.append([ComposableScheduledPromptConditioning(learned_conditioning[i], weight) for i, weight in indexes]) + + return MulticondLearnedConditioning(shape=(len(prompts),), batch=res) + + +def reconstruct_cond_batch(c: list[list[ScheduledPromptConditioning]], current_step): + param = c[0][0].cond + res = torch.zeros((len(c),) + param.shape, device=param.device, dtype=param.dtype) + for i, cond_schedule in enumerate(c): target_index = 0 for current, (end_at, cond) in enumerate(cond_schedule): if current_step <= end_at: @@ -140,6 +218,30 @@ def reconstruct_cond_batch(c: ScheduledPromptBatch, current_step): return res +def reconstruct_multicond_batch(c: MulticondLearnedConditioning, current_step): + param = c.batch[0][0].schedules[0].cond + + tensors = [] + conds_list = [] + + for batch_no, composable_prompts in enumerate(c.batch): + conds_for_batch = [] + + for cond_index, composable_prompt in enumerate(composable_prompts): + target_index = 0 + for current, (end_at, cond) in enumerate(composable_prompt.schedules): + if current_step <= end_at: + target_index = current + break + + conds_for_batch.append((len(tensors), composable_prompt.weight)) + tensors.append(composable_prompt.schedules[target_index].cond) + + conds_list.append(conds_for_batch) + + return conds_list, torch.stack(tensors).to(device=param.device, dtype=param.dtype) + + re_attention = re.compile(r""" \\\(| \\\)| diff --git a/modules/sd_samplers.py b/modules/sd_samplers.py index dbf570d2..d27c547b 100644 --- a/modules/sd_samplers.py +++ b/modules/sd_samplers.py @@ -109,9 +109,12 @@ class VanillaStableDiffusionSampler: return 0 def p_sample_ddim_hook(self, x_dec, cond, ts, unconditional_conditioning, *args, **kwargs): - cond = prompt_parser.reconstruct_cond_batch(cond, self.step) + conds_list, tensor = prompt_parser.reconstruct_multicond_batch(cond, self.step) unconditional_conditioning = prompt_parser.reconstruct_cond_batch(unconditional_conditioning, self.step) + assert all([len(conds) == 1 for conds in conds_list]), 'composition via AND is not supported for DDIM/PLMS samplers' + cond = tensor + if self.mask is not None: img_orig = self.sampler.model.q_sample(self.init_latent, ts) x_dec = img_orig * self.mask + self.nmask * x_dec @@ -183,19 +186,31 @@ class CFGDenoiser(torch.nn.Module): self.step = 0 def forward(self, x, sigma, uncond, cond, cond_scale): - cond = prompt_parser.reconstruct_cond_batch(cond, self.step) + conds_list, tensor = prompt_parser.reconstruct_multicond_batch(cond, self.step) uncond = prompt_parser.reconstruct_cond_batch(uncond, self.step) + batch_size = len(conds_list) + repeats = [len(conds_list[i]) for i in range(batch_size)] + + x_in = torch.cat([torch.stack([x[i] for _ in range(n)]) for i, n in enumerate(repeats)] + [x]) + sigma_in = torch.cat([torch.stack([sigma[i] for _ in range(n)]) for i, n in enumerate(repeats)] + [sigma]) + cond_in = torch.cat([tensor, uncond]) + if shared.batch_cond_uncond: - x_in = torch.cat([x] * 2) - sigma_in = torch.cat([sigma] * 2) - cond_in = torch.cat([uncond, cond]) - uncond, cond = self.inner_model(x_in, sigma_in, cond=cond_in).chunk(2) - denoised = uncond + (cond - uncond) * cond_scale + x_out = self.inner_model(x_in, sigma_in, cond=cond_in) else: - uncond = self.inner_model(x, sigma, cond=uncond) - cond = self.inner_model(x, sigma, cond=cond) - denoised = uncond + (cond - uncond) * cond_scale + x_out = torch.zeros_like(x_in) + for batch_offset in range(0, x_out.shape[0], batch_size): + a = batch_offset + b = a + batch_size + x_out[a:b] = self.inner_model(x_in[a:b], sigma_in[a:b], cond=cond_in[a:b]) + + denoised_uncond = x_out[-batch_size:] + denoised = torch.clone(denoised_uncond) + + for i, conds in enumerate(conds_list): + for cond_index, weight in conds: + denoised[i] += (x_out[cond_index] - denoised_uncond[i]) * (weight * cond_scale) if self.mask is not None: denoised = self.init_latent * self.mask + self.nmask * denoised diff --git a/modules/ui.py b/modules/ui.py index 523ab25b..9620350f 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -34,7 +34,7 @@ import modules.gfpgan_model import modules.codeformer_model import modules.styles import modules.generation_parameters_copypaste -from modules.prompt_parser import get_learned_conditioning_prompt_schedules +from modules import prompt_parser from modules.images import apply_filename_pattern, get_next_sequence_number import modules.textual_inversion.ui @@ -394,7 +394,9 @@ def connect_reuse_seed(seed: gr.Number, reuse_seed: gr.Button, generation_info: def update_token_counter(text, steps): try: - prompt_schedules = get_learned_conditioning_prompt_schedules([text], steps) + _, prompt_flat_list, _ = prompt_parser.get_multicond_prompt_list([text]) + prompt_schedules = prompt_parser.get_learned_conditioning_prompt_schedules(prompt_flat_list, steps) + except Exception: # a parsing error can happen here during typing, and we don't want to bother the user with # messages related to it in console -- cgit v1.2.3 From 0bb458f0ca06a7be27cf1a1003c536d1f06a5bd3 Mon Sep 17 00:00:00 2001 From: Milly Date: Wed, 5 Oct 2022 01:19:50 +0900 Subject: Removed duplicate image saving codes Use `modules.images.save_image()` instead. --- modules/images.py | 7 ++++--- modules/ui.py | 46 ++++++++++------------------------------------ 2 files changed, 14 insertions(+), 39 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/images.py b/modules/images.py index c2fadab9..810f1446 100644 --- a/modules/images.py +++ b/modules/images.py @@ -353,7 +353,7 @@ def get_next_sequence_number(path, basename): return result + 1 -def save_image(image, path, basename, seed=None, prompt=None, extension='png', info=None, short_filename=False, no_prompt=False, grid=False, pnginfo_section_name='parameters', p=None, existing_info=None, forced_filename=None, suffix=""): +def save_image(image, path, basename, seed=None, prompt=None, extension='png', info=None, short_filename=False, no_prompt=False, grid=False, pnginfo_section_name='parameters', p=None, existing_info=None, forced_filename=None, suffix="", save_to_dirs=None): if short_filename or prompt is None or seed is None: file_decoration = "" elif opts.save_to_dirs: @@ -377,7 +377,8 @@ def save_image(image, path, basename, seed=None, prompt=None, extension='png', i else: pnginfo = None - save_to_dirs = (grid and opts.grid_save_to_dirs) or (not grid and opts.save_to_dirs and not no_prompt) + if save_to_dirs is None: + save_to_dirs = (grid and opts.grid_save_to_dirs) or (not grid and opts.save_to_dirs and not no_prompt) if save_to_dirs: dirname = apply_filename_pattern(opts.directories_filename_pattern or "[prompt_words]", p, seed, prompt).strip('\\ /') @@ -431,4 +432,4 @@ def save_image(image, path, basename, seed=None, prompt=None, extension='png', i with open(f"{fullfn_without_extension}.txt", "w", encoding="utf8") as file: file.write(info + "\n") - + return fullfn diff --git a/modules/ui.py b/modules/ui.py index 9620350f..4f18126f 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -35,7 +35,7 @@ import modules.codeformer_model import modules.styles import modules.generation_parameters_copypaste from modules import prompt_parser -from modules.images import apply_filename_pattern, get_next_sequence_number +from modules.images import save_image import modules.textual_inversion.ui # this is a fix for Windows users. Without it, javascript files will be served with text/html content-type and the bowser will not show any UI @@ -114,20 +114,13 @@ def save_files(js_data, images, index): p = MyObject(data) path = opts.outdir_save save_to_dirs = opts.use_save_to_dirs_for_ui - - if save_to_dirs: - dirname = apply_filename_pattern(opts.directories_filename_pattern or "[prompt_words]", p, p.seed, p.prompt) - path = os.path.join(opts.outdir_save, dirname) - - os.makedirs(path, exist_ok=True) - + extension: str = opts.samples_format + start_index = 0 if index > -1 and opts.save_selected_only and (index >= data["index_of_first_image"]): # ensures we are looking at a specific non-grid picture, and we have save_selected_only images = [images[index]] - infotexts = [data["infotexts"][index]] - else: - infotexts = data["infotexts"] + start_index = index with open(os.path.join(opts.outdir_save, "log.csv"), "a", encoding="utf8", newline='') as file: at_start = file.tell() == 0 @@ -135,37 +128,18 @@ def save_files(js_data, images, index): if at_start: writer.writerow(["prompt", "seed", "width", "height", "sampler", "cfgs", "steps", "filename", "negative_prompt"]) - file_decoration = opts.samples_filename_pattern or "[seed]-[prompt_spaces]" - if file_decoration != "": - file_decoration = "-" + file_decoration.lower() - file_decoration = apply_filename_pattern(file_decoration, p, p.seed, p.prompt) - truncated = (file_decoration[:240] + '..') if len(file_decoration) > 240 else file_decoration - filename_base = truncated - extension = opts.samples_format.lower() - - basecount = get_next_sequence_number(path, "") - for i, filedata in enumerate(images): - file_number = f"{basecount+i:05}" - filename = file_number + filename_base + f".{extension}" - filepath = os.path.join(path, filename) - - + for image_index, filedata in enumerate(images, start_index): if filedata.startswith("data:image/png;base64,"): filedata = filedata[len("data:image/png;base64,"):] image = Image.open(io.BytesIO(base64.decodebytes(filedata.encode('utf-8')))) - if opts.enable_pnginfo and extension == 'png': - pnginfo = PngImagePlugin.PngInfo() - pnginfo.add_text('parameters', infotexts[i]) - image.save(filepath, pnginfo=pnginfo) - else: - image.save(filepath, quality=opts.jpeg_quality) - if opts.enable_pnginfo and extension in ("jpg", "jpeg", "webp"): - piexif.insert(piexif.dump({"Exif": { - piexif.ExifIFD.UserComment: piexif.helper.UserComment.dump(infotexts[i], encoding="unicode") - }}), filepath) + is_grid = image_index < p.index_of_first_image + i = 0 if is_grid else (image_index - p.index_of_first_image) + + fullfn = save_image(image, path, "", seed=p.all_seeds[i], prompt=p.all_prompts[i], extension=extension, info=p.infotexts[image_index], grid=is_grid, p=p, save_to_dirs=save_to_dirs) + filename = os.path.relpath(fullfn, path) filenames.append(filename) writer.writerow([data["prompt"], data["seed"], data["width"], data["height"], data["sampler"], data["cfg_scale"], data["steps"], filenames[0], data["negative_prompt"]]) -- cgit v1.2.3 From 12c4d5c6b5bf9dd50d0601c36af4f99b65316d58 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Fri, 7 Oct 2022 23:22:22 +0300 Subject: hypernetwork training mk1 --- modules/hypernetwork.py | 88 --------- modules/hypernetwork/hypernetwork.py | 267 +++++++++++++++++++++++++++ modules/hypernetwork/ui.py | 43 +++++ modules/sd_hijack.py | 4 +- modules/sd_hijack_optimizations.py | 3 +- modules/shared.py | 13 +- modules/textual_inversion/ui.py | 1 - modules/ui.py | 58 +++++- scripts/xy_grid.py | 7 +- textual_inversion_templates/hypernetwork.txt | 27 +++ textual_inversion_templates/none.txt | 1 + webui.py | 9 + 12 files changed, 414 insertions(+), 107 deletions(-) delete mode 100644 modules/hypernetwork.py create mode 100644 modules/hypernetwork/hypernetwork.py create mode 100644 modules/hypernetwork/ui.py create mode 100644 textual_inversion_templates/hypernetwork.txt create mode 100644 textual_inversion_templates/none.txt (limited to 'modules/ui.py') diff --git a/modules/hypernetwork.py b/modules/hypernetwork.py deleted file mode 100644 index c7b86682..00000000 --- a/modules/hypernetwork.py +++ /dev/null @@ -1,88 +0,0 @@ -import glob -import os -import sys -import traceback - -import torch - -from ldm.util import default -from modules import devices, shared -import torch -from torch import einsum -from einops import rearrange, repeat - - -class HypernetworkModule(torch.nn.Module): - def __init__(self, dim, state_dict): - super().__init__() - - self.linear1 = torch.nn.Linear(dim, dim * 2) - self.linear2 = torch.nn.Linear(dim * 2, dim) - - self.load_state_dict(state_dict, strict=True) - self.to(devices.device) - - def forward(self, x): - return x + (self.linear2(self.linear1(x))) - - -class Hypernetwork: - filename = None - name = None - - def __init__(self, filename): - self.filename = filename - self.name = os.path.splitext(os.path.basename(filename))[0] - self.layers = {} - - state_dict = torch.load(filename, map_location='cpu') - for size, sd in state_dict.items(): - self.layers[size] = (HypernetworkModule(size, sd[0]), HypernetworkModule(size, sd[1])) - - -def load_hypernetworks(path): - res = {} - - for filename in glob.iglob(path + '**/*.pt', recursive=True): - try: - hn = Hypernetwork(filename) - res[hn.name] = hn - except Exception: - print(f"Error loading hypernetwork {filename}", file=sys.stderr) - print(traceback.format_exc(), file=sys.stderr) - - return res - - -def attention_CrossAttention_forward(self, x, context=None, mask=None): - h = self.heads - - q = self.to_q(x) - context = default(context, x) - - hypernetwork = shared.selected_hypernetwork() - hypernetwork_layers = (hypernetwork.layers if hypernetwork is not None else {}).get(context.shape[2], None) - - if hypernetwork_layers is not None: - k = self.to_k(hypernetwork_layers[0](context)) - v = self.to_v(hypernetwork_layers[1](context)) - else: - k = self.to_k(context) - v = self.to_v(context) - - q, k, v = map(lambda t: rearrange(t, 'b n (h d) -> (b h) n d', h=h), (q, k, v)) - - sim = einsum('b i d, b j d -> b i j', q, k) * self.scale - - if mask is not None: - mask = rearrange(mask, 'b ... -> b (...)') - max_neg_value = -torch.finfo(sim.dtype).max - mask = repeat(mask, 'b j -> (b h) () j', h=h) - sim.masked_fill_(~mask, max_neg_value) - - # attention, what we cannot get enough of - attn = sim.softmax(dim=-1) - - out = einsum('b i j, b j d -> b i d', attn, v) - out = rearrange(out, '(b h) n d -> b n (h d)', h=h) - return self.to_out(out) diff --git a/modules/hypernetwork/hypernetwork.py b/modules/hypernetwork/hypernetwork.py new file mode 100644 index 00000000..a3d6a47e --- /dev/null +++ b/modules/hypernetwork/hypernetwork.py @@ -0,0 +1,267 @@ +import datetime +import glob +import html +import os +import sys +import traceback +import tqdm + +import torch + +from ldm.util import default +from modules import devices, shared, processing, sd_models +import torch +from torch import einsum +from einops import rearrange, repeat +import modules.textual_inversion.dataset + + +class HypernetworkModule(torch.nn.Module): + def __init__(self, dim, state_dict=None): + super().__init__() + + self.linear1 = torch.nn.Linear(dim, dim * 2) + self.linear2 = torch.nn.Linear(dim * 2, dim) + + if state_dict is not None: + self.load_state_dict(state_dict, strict=True) + else: + self.linear1.weight.data.fill_(0.0001) + self.linear1.bias.data.fill_(0.0001) + self.linear2.weight.data.fill_(0.0001) + self.linear2.bias.data.fill_(0.0001) + + self.to(devices.device) + + def forward(self, x): + return x + (self.linear2(self.linear1(x))) + + +class Hypernetwork: + filename = None + name = None + + def __init__(self, name=None): + self.filename = None + self.name = name + self.layers = {} + self.step = 0 + self.sd_checkpoint = None + self.sd_checkpoint_name = None + + for size in [320, 640, 768, 1280]: + self.layers[size] = (HypernetworkModule(size), HypernetworkModule(size)) + + def weights(self): + res = [] + + for k, layers in self.layers.items(): + for layer in layers: + layer.train() + res += [layer.linear1.weight, layer.linear1.bias, layer.linear2.weight, layer.linear2.bias] + + return res + + def save(self, filename): + state_dict = {} + + for k, v in self.layers.items(): + state_dict[k] = (v[0].state_dict(), v[1].state_dict()) + + state_dict['step'] = self.step + state_dict['name'] = self.name + state_dict['sd_checkpoint'] = self.sd_checkpoint + state_dict['sd_checkpoint_name'] = self.sd_checkpoint_name + + torch.save(state_dict, filename) + + def load(self, filename): + self.filename = filename + if self.name is None: + self.name = os.path.splitext(os.path.basename(filename))[0] + + state_dict = torch.load(filename, map_location='cpu') + + for size, sd in state_dict.items(): + if type(size) == int: + self.layers[size] = (HypernetworkModule(size, sd[0]), HypernetworkModule(size, sd[1])) + + self.name = state_dict.get('name', self.name) + self.step = state_dict.get('step', 0) + self.sd_checkpoint = state_dict.get('sd_checkpoint', None) + self.sd_checkpoint_name = state_dict.get('sd_checkpoint_name', None) + + +def load_hypernetworks(path): + res = {} + + for filename in glob.iglob(path + '**/*.pt', recursive=True): + try: + hn = Hypernetwork() + hn.load(filename) + res[hn.name] = hn + except Exception: + print(f"Error loading hypernetwork {filename}", file=sys.stderr) + print(traceback.format_exc(), file=sys.stderr) + + return res + + +def attention_CrossAttention_forward(self, x, context=None, mask=None): + h = self.heads + + q = self.to_q(x) + context = default(context, x) + + hypernetwork_layers = (shared.hypernetwork.layers if shared.hypernetwork is not None else {}).get(context.shape[2], None) + + if hypernetwork_layers is not None: + hypernetwork_k, hypernetwork_v = hypernetwork_layers + + self.hypernetwork_k = hypernetwork_k + self.hypernetwork_v = hypernetwork_v + + context_k = hypernetwork_k(context) + context_v = hypernetwork_v(context) + else: + context_k = context + context_v = context + + k = self.to_k(context_k) + v = self.to_v(context_v) + + q, k, v = map(lambda t: rearrange(t, 'b n (h d) -> (b h) n d', h=h), (q, k, v)) + + sim = einsum('b i d, b j d -> b i j', q, k) * self.scale + + if mask is not None: + mask = rearrange(mask, 'b ... -> b (...)') + max_neg_value = -torch.finfo(sim.dtype).max + mask = repeat(mask, 'b j -> (b h) () j', h=h) + sim.masked_fill_(~mask, max_neg_value) + + # attention, what we cannot get enough of + attn = sim.softmax(dim=-1) + + out = einsum('b i j, b j d -> b i d', attn, v) + out = rearrange(out, '(b h) n d -> b n (h d)', h=h) + return self.to_out(out) + + +def train_hypernetwork(hypernetwork_name, learn_rate, data_root, log_directory, steps, create_image_every, save_hypernetwork_every, template_file, preview_image_prompt): + assert hypernetwork_name, 'embedding not selected' + + shared.hypernetwork = shared.hypernetworks[hypernetwork_name] + + shared.state.textinfo = "Initializing hypernetwork training..." + shared.state.job_count = steps + + filename = os.path.join(shared.cmd_opts.hypernetwork_dir, f'{hypernetwork_name}.pt') + + log_directory = os.path.join(log_directory, datetime.datetime.now().strftime("%Y-%m-%d"), hypernetwork_name) + + if save_hypernetwork_every > 0: + hypernetwork_dir = os.path.join(log_directory, "hypernetworks") + os.makedirs(hypernetwork_dir, exist_ok=True) + else: + hypernetwork_dir = None + + if create_image_every > 0: + images_dir = os.path.join(log_directory, "images") + os.makedirs(images_dir, exist_ok=True) + else: + images_dir = None + + cond_model = shared.sd_model.cond_stage_model + + shared.state.textinfo = f"Preparing dataset from {html.escape(data_root)}..." + with torch.autocast("cuda"): + ds = modules.textual_inversion.dataset.PersonalizedBase(data_root=data_root, size=512, placeholder_token=hypernetwork_name, model=shared.sd_model, device=devices.device, template_file=template_file) + + hypernetwork = shared.hypernetworks[hypernetwork_name] + weights = hypernetwork.weights() + for weight in weights: + weight.requires_grad = True + + optimizer = torch.optim.AdamW(weights, lr=learn_rate) + + losses = torch.zeros((32,)) + + last_saved_file = "" + last_saved_image = "" + + ititial_step = hypernetwork.step or 0 + if ititial_step > steps: + return hypernetwork, filename + + pbar = tqdm.tqdm(enumerate(ds), total=steps-ititial_step) + for i, (x, text) in pbar: + hypernetwork.step = i + ititial_step + + if hypernetwork.step > steps: + break + + if shared.state.interrupted: + break + + with torch.autocast("cuda"): + c = cond_model([text]) + + x = x.to(devices.device) + loss = shared.sd_model(x.unsqueeze(0), c)[0] + del x + + losses[hypernetwork.step % losses.shape[0]] = loss.item() + + optimizer.zero_grad() + loss.backward() + optimizer.step() + + pbar.set_description(f"loss: {losses.mean():.7f}") + + if hypernetwork.step > 0 and hypernetwork_dir is not None and hypernetwork.step % save_hypernetwork_every == 0: + last_saved_file = os.path.join(hypernetwork_dir, f'{hypernetwork_name}-{hypernetwork.step}.pt') + hypernetwork.save(last_saved_file) + + if hypernetwork.step > 0 and images_dir is not None and hypernetwork.step % create_image_every == 0: + last_saved_image = os.path.join(images_dir, f'{hypernetwork_name}-{hypernetwork.step}.png') + + preview_text = text if preview_image_prompt == "" else preview_image_prompt + + p = processing.StableDiffusionProcessingTxt2Img( + sd_model=shared.sd_model, + prompt=preview_text, + steps=20, + do_not_save_grid=True, + do_not_save_samples=True, + ) + + processed = processing.process_images(p) + image = processed.images[0] + + shared.state.current_image = image + image.save(last_saved_image) + + last_saved_image += f", prompt: {preview_text}" + + shared.state.job_no = hypernetwork.step + + shared.state.textinfo = f""" +

+Loss: {losses.mean():.7f}
+Step: {hypernetwork.step}
+Last prompt: {html.escape(text)}
+Last saved embedding: {html.escape(last_saved_file)}
+Last saved image: {html.escape(last_saved_image)}
+

+""" + + checkpoint = sd_models.select_checkpoint() + + hypernetwork.sd_checkpoint = checkpoint.hash + hypernetwork.sd_checkpoint_name = checkpoint.model_name + hypernetwork.save(filename) + + return hypernetwork, filename + + diff --git a/modules/hypernetwork/ui.py b/modules/hypernetwork/ui.py new file mode 100644 index 00000000..525f978c --- /dev/null +++ b/modules/hypernetwork/ui.py @@ -0,0 +1,43 @@ +import html +import os + +import gradio as gr + +import modules.textual_inversion.textual_inversion +import modules.textual_inversion.preprocess +from modules import sd_hijack, shared + + +def create_hypernetwork(name): + fn = os.path.join(shared.cmd_opts.hypernetwork_dir, f"{name}.pt") + assert not os.path.exists(fn), f"file {fn} already exists" + + hypernetwork = modules.hypernetwork.hypernetwork.Hypernetwork(name=name) + hypernetwork.save(fn) + + shared.reload_hypernetworks() + shared.hypernetwork = shared.hypernetworks.get(shared.opts.sd_hypernetwork, None) + + return gr.Dropdown.update(choices=sorted([x for x in shared.hypernetworks.keys()])), f"Created: {fn}", "" + + +def train_hypernetwork(*args): + + initial_hypernetwork = shared.hypernetwork + + try: + sd_hijack.undo_optimizations() + + hypernetwork, filename = modules.hypernetwork.hypernetwork.train_hypernetwork(*args) + + res = f""" +Training {'interrupted' if shared.state.interrupted else 'finished'} at {hypernetwork.step} steps. +Hypernetwork saved to {html.escape(filename)} +""" + return res, "" + except Exception: + raise + finally: + shared.hypernetwork = initial_hypernetwork + sd_hijack.apply_optimizations() + diff --git a/modules/sd_hijack.py b/modules/sd_hijack.py index d68f89cc..ec8c9d4b 100644 --- a/modules/sd_hijack.py +++ b/modules/sd_hijack.py @@ -8,7 +8,7 @@ from torch import einsum from torch.nn.functional import silu import modules.textual_inversion.textual_inversion -from modules import prompt_parser, devices, sd_hijack_optimizations, shared, hypernetwork +from modules import prompt_parser, devices, sd_hijack_optimizations, shared from modules.shared import opts, device, cmd_opts import ldm.modules.attention @@ -32,6 +32,8 @@ def apply_optimizations(): def undo_optimizations(): + from modules.hypernetwork import hypernetwork + ldm.modules.attention.CrossAttention.forward = hypernetwork.attention_CrossAttention_forward ldm.modules.diffusionmodules.model.nonlinearity = diffusionmodules_model_nonlinearity ldm.modules.diffusionmodules.model.AttnBlock.forward = diffusionmodules_model_AttnBlock_forward diff --git a/modules/sd_hijack_optimizations.py b/modules/sd_hijack_optimizations.py index d9cca485..3f32e020 100644 --- a/modules/sd_hijack_optimizations.py +++ b/modules/sd_hijack_optimizations.py @@ -45,8 +45,7 @@ def split_cross_attention_forward(self, x, context=None, mask=None): q_in = self.to_q(x) context = default(context, x) - hypernetwork = shared.selected_hypernetwork() - hypernetwork_layers = (hypernetwork.layers if hypernetwork is not None else {}).get(context.shape[2], None) + hypernetwork_layers = (shared.hypernetwork.layers if shared.hypernetwork is not None else {}).get(context.shape[2], None) if hypernetwork_layers is not None: k_in = self.to_k(hypernetwork_layers[0](context)) diff --git a/modules/shared.py b/modules/shared.py index 879d8424..c5a893e8 100644 --- a/modules/shared.py +++ b/modules/shared.py @@ -13,7 +13,7 @@ import modules.memmon import modules.sd_models import modules.styles import modules.devices as devices -from modules import sd_samplers, hypernetwork +from modules import sd_samplers from modules.paths import models_path, script_path, sd_path sd_model_file = os.path.join(script_path, 'model.ckpt') @@ -28,6 +28,7 @@ parser.add_argument("--no-half", action='store_true', help="do not switch the mo parser.add_argument("--no-progressbar-hiding", action='store_true', help="do not hide progressbar in gradio UI (we hide it because it slows down ML if you have hardware acceleration in browser)") parser.add_argument("--max-batch-count", type=int, default=16, help="maximum batch count value for the UI") parser.add_argument("--embeddings-dir", type=str, default=os.path.join(script_path, 'embeddings'), help="embeddings directory for textual inversion (default: embeddings)") +parser.add_argument("--hypernetwork-dir", type=str, default=os.path.join(models_path, 'hypernetworks'), help="hypernetwork directory") parser.add_argument("--allow-code", action='store_true', help="allow custom script execution from webui") parser.add_argument("--medvram", action='store_true', help="enable stable diffusion model optimizations for sacrificing a little speed for low VRM usage") parser.add_argument("--lowvram", action='store_true', help="enable stable diffusion model optimizations for sacrificing a lot of speed for very low VRM usage") @@ -76,11 +77,15 @@ parallel_processing_allowed = not cmd_opts.lowvram and not cmd_opts.medvram config_filename = cmd_opts.ui_settings_file -hypernetworks = hypernetwork.load_hypernetworks(os.path.join(models_path, 'hypernetworks')) +def reload_hypernetworks(): + from modules.hypernetwork import hypernetwork + hypernetworks.clear() + hypernetworks.update(hypernetwork.load_hypernetworks(cmd_opts.hypernetwork_dir)) -def selected_hypernetwork(): - return hypernetworks.get(opts.sd_hypernetwork, None) + +hypernetworks = {} +hypernetwork = None class State: diff --git a/modules/textual_inversion/ui.py b/modules/textual_inversion/ui.py index f19ac5e0..c57de1f9 100644 --- a/modules/textual_inversion/ui.py +++ b/modules/textual_inversion/ui.py @@ -22,7 +22,6 @@ def preprocess(*args): def train_embedding(*args): - try: sd_hijack.undo_optimizations() diff --git a/modules/ui.py b/modules/ui.py index 4f18126f..051908c1 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -37,6 +37,7 @@ import modules.generation_parameters_copypaste from modules import prompt_parser from modules.images import save_image import modules.textual_inversion.ui +import modules.hypernetwork.ui # this is a fix for Windows users. Without it, javascript files will be served with text/html content-type and the bowser will not show any UI mimetypes.init() @@ -965,6 +966,18 @@ def create_ui(wrap_gradio_gpu_call): with gr.Column(): create_embedding = gr.Button(value="Create", variant='primary') + with gr.Group(): + gr.HTML(value="

Create a new hypernetwork

") + + new_hypernetwork_name = gr.Textbox(label="Name") + + with gr.Row(): + with gr.Column(scale=3): + gr.HTML(value="") + + with gr.Column(): + create_hypernetwork = gr.Button(value="Create", variant='primary') + with gr.Group(): gr.HTML(value="

Preprocess images

") @@ -986,6 +999,7 @@ def create_ui(wrap_gradio_gpu_call): with gr.Group(): gr.HTML(value="

Train an embedding; must specify a directory with a set of 512x512 images

") train_embedding_name = gr.Dropdown(label='Embedding', choices=sorted(sd_hijack.model_hijack.embedding_db.word_embeddings.keys())) + train_hypernetwork_name = gr.Dropdown(label='Hypernetwork', choices=[x for x in shared.hypernetworks.keys()]) learn_rate = gr.Number(label='Learning rate', value=5.0e-03) dataset_directory = gr.Textbox(label='Dataset directory', placeholder="Path to directory with input images") log_directory = gr.Textbox(label='Log directory', placeholder="Path to directory where to write outputs", value="textual_inversion") @@ -993,15 +1007,12 @@ def create_ui(wrap_gradio_gpu_call): steps = gr.Number(label='Max steps', value=100000, precision=0) create_image_every = gr.Number(label='Save an image to log directory every N steps, 0 to disable', value=500, precision=0) save_embedding_every = gr.Number(label='Save a copy of embedding to log directory every N steps, 0 to disable', value=500, precision=0) + preview_image_prompt = gr.Textbox(label='Preview prompt', value="") with gr.Row(): - with gr.Column(scale=2): - gr.HTML(value="") - - with gr.Column(): - with gr.Row(): - interrupt_training = gr.Button(value="Interrupt") - train_embedding = gr.Button(value="Train", variant='primary') + interrupt_training = gr.Button(value="Interrupt") + train_hypernetwork = gr.Button(value="Train Hypernetwork", variant='primary') + train_embedding = gr.Button(value="Train Embedding", variant='primary') with gr.Column(): progressbar = gr.HTML(elem_id="ti_progressbar") @@ -1027,6 +1038,18 @@ def create_ui(wrap_gradio_gpu_call): ] ) + create_hypernetwork.click( + fn=modules.hypernetwork.ui.create_hypernetwork, + inputs=[ + new_hypernetwork_name, + ], + outputs=[ + train_hypernetwork_name, + ti_output, + ti_outcome, + ] + ) + run_preprocess.click( fn=wrap_gradio_gpu_call(modules.textual_inversion.ui.preprocess, extra_outputs=[gr.update()]), _js="start_training_textual_inversion", @@ -1062,12 +1085,33 @@ def create_ui(wrap_gradio_gpu_call): ] ) + train_hypernetwork.click( + fn=wrap_gradio_gpu_call(modules.hypernetwork.ui.train_hypernetwork, extra_outputs=[gr.update()]), + _js="start_training_textual_inversion", + inputs=[ + train_hypernetwork_name, + learn_rate, + dataset_directory, + log_directory, + steps, + create_image_every, + save_embedding_every, + template_file, + preview_image_prompt, + ], + outputs=[ + ti_output, + ti_outcome, + ] + ) + interrupt_training.click( fn=lambda: shared.state.interrupt(), inputs=[], outputs=[], ) + def create_setting_component(key): def fun(): return opts.data[key] if key in opts.data else opts.data_labels[key].default diff --git a/scripts/xy_grid.py b/scripts/xy_grid.py index c0c364df..5b504de6 100644 --- a/scripts/xy_grid.py +++ b/scripts/xy_grid.py @@ -78,8 +78,7 @@ def apply_checkpoint(p, x, xs): def apply_hypernetwork(p, x, xs): - hn = shared.hypernetworks.get(x, None) - opts.data["sd_hypernetwork"] = hn.name if hn is not None else 'None' + shared.hypernetwork = shared.hypernetworks.get(x, None) def format_value_add_label(p, opt, x): @@ -199,7 +198,7 @@ class Script(scripts.Script): modules.processing.fix_seed(p) p.batch_size = 1 - initial_hn = opts.sd_hypernetwork + initial_hn = shared.hypernetwork def process_axis(opt, vals): if opt.label == 'Nothing': @@ -308,6 +307,6 @@ class Script(scripts.Script): # restore checkpoint in case it was changed by axes modules.sd_models.reload_model_weights(shared.sd_model) - opts.data["sd_hypernetwork"] = initial_hn + shared.hypernetwork = initial_hn return processed diff --git a/textual_inversion_templates/hypernetwork.txt b/textual_inversion_templates/hypernetwork.txt new file mode 100644 index 00000000..91e06890 --- /dev/null +++ b/textual_inversion_templates/hypernetwork.txt @@ -0,0 +1,27 @@ +a photo of a [filewords] +a rendering of a [filewords] +a cropped photo of the [filewords] +the photo of a [filewords] +a photo of a clean [filewords] +a photo of a dirty [filewords] +a dark photo of the [filewords] +a photo of my [filewords] +a photo of the cool [filewords] +a close-up photo of a [filewords] +a bright photo of the [filewords] +a cropped photo of a [filewords] +a photo of the [filewords] +a good photo of the [filewords] +a photo of one [filewords] +a close-up photo of the [filewords] +a rendition of the [filewords] +a photo of the clean [filewords] +a rendition of a [filewords] +a photo of a nice [filewords] +a good photo of a [filewords] +a photo of the nice [filewords] +a photo of the small [filewords] +a photo of the weird [filewords] +a photo of the large [filewords] +a photo of a cool [filewords] +a photo of a small [filewords] diff --git a/textual_inversion_templates/none.txt b/textual_inversion_templates/none.txt new file mode 100644 index 00000000..f77af461 --- /dev/null +++ b/textual_inversion_templates/none.txt @@ -0,0 +1 @@ +picture diff --git a/webui.py b/webui.py index 480360fe..60f9061f 100644 --- a/webui.py +++ b/webui.py @@ -74,6 +74,15 @@ def wrap_gradio_gpu_call(func, extra_outputs=None): return modules.ui.wrap_gradio_call(f, extra_outputs=extra_outputs) +def set_hypernetwork(): + shared.hypernetwork = shared.hypernetworks.get(shared.opts.sd_hypernetwork, None) + + +shared.reload_hypernetworks() +shared.opts.onchange("sd_hypernetwork", set_hypernetwork) +set_hypernetwork() + + modules.scripts.load_scripts(os.path.join(script_path, "scripts")) shared.sd_model = modules.sd_models.load_model() -- cgit v1.2.3 From 786d9f63aaa4515df82eb2cf357ea92f3dae1e29 Mon Sep 17 00:00:00 2001 From: Trung Ngo Date: Tue, 4 Oct 2022 22:56:30 -0500 Subject: Add button to skip the current iteration --- javascript/hints.js | 1 + javascript/progressbar.js | 20 ++++++++++++++------ modules/img2img.py | 4 ++++ modules/processing.py | 4 ++++ modules/shared.py | 5 +++++ modules/ui.py | 8 ++++++++ style.css | 14 ++++++++++++-- webui.py | 1 + 8 files changed, 49 insertions(+), 8 deletions(-) (limited to 'modules/ui.py') diff --git a/javascript/hints.js b/javascript/hints.js index 8adcd983..8e352e94 100644 --- a/javascript/hints.js +++ b/javascript/hints.js @@ -35,6 +35,7 @@ titles = { "Denoising strength": "Determines how little respect the algorithm should have for image's content. At 0, nothing will change, and at 1 you'll get an unrelated image. With values below 1.0, processing will take less steps than the Sampling Steps slider specifies.", "Denoising strength change factor": "In loopback mode, on each loop the denoising strength is multiplied by this value. <1 means decreasing variety so your sequence will converge on a fixed picture. >1 means increasing variety so your sequence will become more and more chaotic.", + "Skip": "Stop processing current image and continue processing.", "Interrupt": "Stop processing images and return any results accumulated so far.", "Save": "Write image to a directory (default - log/images) and generation parameters into csv file.", diff --git a/javascript/progressbar.js b/javascript/progressbar.js index f9e9290e..4395a215 100644 --- a/javascript/progressbar.js +++ b/javascript/progressbar.js @@ -1,8 +1,9 @@ // code related to showing and updating progressbar shown as the image is being made global_progressbars = {} -function check_progressbar(id_part, id_progressbar, id_progressbar_span, id_interrupt, id_preview, id_gallery){ +function check_progressbar(id_part, id_progressbar, id_progressbar_span, id_skip, id_interrupt, id_preview, id_gallery){ var progressbar = gradioApp().getElementById(id_progressbar) + var skip = id_skip ? gradioApp().getElementById(id_skip) : null var interrupt = gradioApp().getElementById(id_interrupt) if(opts.show_progress_in_title && progressbar && progressbar.offsetParent){ @@ -32,30 +33,37 @@ function check_progressbar(id_part, id_progressbar, id_progressbar_span, id_inte var progressDiv = gradioApp().querySelectorAll('#' + id_progressbar_span).length > 0; if(!progressDiv){ + if (skip) { + skip.style.display = "none" + } interrupt.style.display = "none" } } - window.setTimeout(function(){ requestMoreProgress(id_part, id_progressbar_span, id_interrupt) }, 500) + window.setTimeout(function() { requestMoreProgress(id_part, id_progressbar_span, id_skip, id_interrupt) }, 500) }); mutationObserver.observe( progressbar, { childList:true, subtree:true }) } } onUiUpdate(function(){ - check_progressbar('txt2img', 'txt2img_progressbar', 'txt2img_progress_span', 'txt2img_interrupt', 'txt2img_preview', 'txt2img_gallery') - check_progressbar('img2img', 'img2img_progressbar', 'img2img_progress_span', 'img2img_interrupt', 'img2img_preview', 'img2img_gallery') - check_progressbar('ti', 'ti_progressbar', 'ti_progress_span', 'ti_interrupt', 'ti_preview', 'ti_gallery') + check_progressbar('txt2img', 'txt2img_progressbar', 'txt2img_progress_span', 'txt2img_skip', 'txt2img_interrupt', 'txt2img_preview', 'txt2img_gallery') + check_progressbar('img2img', 'img2img_progressbar', 'img2img_progress_span', 'img2img_skip', 'img2img_interrupt', 'img2img_preview', 'img2img_gallery') + check_progressbar('ti', 'ti_progressbar', 'ti_progress_span', '', 'ti_interrupt', 'ti_preview', 'ti_gallery') }) -function requestMoreProgress(id_part, id_progressbar_span, id_interrupt){ +function requestMoreProgress(id_part, id_progressbar_span, id_skip, id_interrupt){ btn = gradioApp().getElementById(id_part+"_check_progress"); if(btn==null) return; btn.click(); var progressDiv = gradioApp().querySelectorAll('#' + id_progressbar_span).length > 0; + var skip = id_skip ? gradioApp().getElementById(id_skip) : null var interrupt = gradioApp().getElementById(id_interrupt) if(progressDiv && interrupt){ + if (skip) { + skip.style.display = "block" + } interrupt.style.display = "block" } } diff --git a/modules/img2img.py b/modules/img2img.py index da212d72..e60b7e0f 100644 --- a/modules/img2img.py +++ b/modules/img2img.py @@ -32,6 +32,10 @@ def process_batch(p, input_dir, output_dir, args): for i, image in enumerate(images): state.job = f"{i+1} out of {len(images)}" + if state.skipped: + state.skipped = False + state.interrupted = False + continue if state.interrupted: break diff --git a/modules/processing.py b/modules/processing.py index d814d5ac..6805039c 100644 --- a/modules/processing.py +++ b/modules/processing.py @@ -355,6 +355,10 @@ def process_images(p: StableDiffusionProcessing) -> Processed: state.job_count = p.n_iter for n in range(p.n_iter): + if state.skipped: + state.skipped = False + state.interrupted = False + if state.interrupted: break diff --git a/modules/shared.py b/modules/shared.py index 864e772c..7f802bd9 100644 --- a/modules/shared.py +++ b/modules/shared.py @@ -84,6 +84,7 @@ def selected_hypernetwork(): class State: + skipped = False interrupted = False job = "" job_no = 0 @@ -96,6 +97,10 @@ class State: current_image_sampling_step = 0 textinfo = None + def skip(self): + self.skipped = True + self.interrupted = True + def interrupt(self): self.interrupted = True diff --git a/modules/ui.py b/modules/ui.py index 4f18126f..e3e62fdd 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -191,6 +191,7 @@ def wrap_gradio_call(func, extra_outputs=None): # last item is always HTML res[-1] += f"

Time taken: {elapsed_text}

{vram_html}
" + shared.state.skipped = False shared.state.interrupted = False shared.state.job_count = 0 @@ -411,9 +412,16 @@ def create_toprow(is_img2img): with gr.Column(scale=1): with gr.Row(): + skip = gr.Button('Skip', elem_id=f"{id_part}_skip") interrupt = gr.Button('Interrupt', elem_id=f"{id_part}_interrupt") submit = gr.Button('Generate', elem_id=f"{id_part}_generate", variant='primary') + skip.click( + fn=lambda: shared.state.skip(), + inputs=[], + outputs=[], + ) + interrupt.click( fn=lambda: shared.state.interrupt(), inputs=[], diff --git a/style.css b/style.css index 50c5e557..6904fc50 100644 --- a/style.css +++ b/style.css @@ -393,10 +393,20 @@ input[type="range"]{ #txt2img_interrupt, #img2img_interrupt{ position: absolute; - width: 100%; + width: 50%; height: 72px; background: #b4c0cc; - border-radius: 8px; + border-radius: 0px; + display: none; +} + +#txt2img_skip, #img2img_skip{ + position: absolute; + width: 50%; + right: 0px; + height: 72px; + background: #b4c0cc; + border-radius: 0px; display: none; } diff --git a/webui.py b/webui.py index 480360fe..3b4cf5e9 100644 --- a/webui.py +++ b/webui.py @@ -58,6 +58,7 @@ def wrap_gradio_gpu_call(func, extra_outputs=None): shared.state.current_latent = None shared.state.current_image = None shared.state.current_image_sampling_step = 0 + shared.state.skipped = False shared.state.interrupted = False shared.state.textinfo = None -- cgit v1.2.3 From 4f33289d0fc5aa3a197f4a4c926d03d44f0d597e Mon Sep 17 00:00:00 2001 From: Milly Date: Sat, 8 Oct 2022 22:48:15 +0900 Subject: Fixed typo --- modules/ui.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'modules/ui.py') diff --git a/modules/ui.py b/modules/ui.py index e3e62fdd..ffd75f6a 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -946,7 +946,7 @@ def create_ui(wrap_gradio_gpu_call): custom_name = gr.Textbox(label="Custom Name (Optional)") interp_amount = gr.Slider(minimum=0.0, maximum=1.0, step=0.05, label='Interpolation Amount', value=0.3) interp_method = gr.Radio(choices=["Weighted Sum", "Sigmoid", "Inverse Sigmoid"], value="Weighted Sum", label="Interpolation Method") - save_as_half = gr.Checkbox(value=False, label="Safe as float16") + save_as_half = gr.Checkbox(value=False, label="Save as float16") modelmerger_merge = gr.Button(elem_id="modelmerger_merge", label="Merge", variant='primary') with gr.Column(variant='panel'): -- cgit v1.2.3 From 01f8cb44474e454903c11718e6a4f33dbde34bb8 Mon Sep 17 00:00:00 2001 From: Greendayle Date: Sat, 8 Oct 2022 18:02:56 +0200 Subject: made deepdanbooru optional, added to readme, automatic download of deepbooru model --- README.md | 2 ++ launch.py | 4 ++++ modules/deepbooru.py | 20 ++++++++++---------- modules/shared.py | 1 + modules/ui.py | 19 ++++++++++++------- requirements.txt | 3 --- requirements_versions.txt | 3 --- 7 files changed, 29 insertions(+), 23 deletions(-) (limited to 'modules/ui.py') diff --git a/README.md b/README.md index ef9b5e31..6cd7a1f9 100644 --- a/README.md +++ b/README.md @@ -66,6 +66,7 @@ Check the [custom scripts](https://github.com/AUTOMATIC1111/stable-diffusion-web - separate prompts using uppercase `AND` - also supports weights for prompts: `a cat :1.2 AND a dog AND a penguin :2.2` - No token limit for prompts (original stable diffusion lets you use up to 75 tokens) +- DeepDanbooru integration, creates danbooru style tags for anime prompts (add --deepdanbooru to commandline args) ## Installation and Running Make sure the required [dependencies](https://github.com/AUTOMATIC1111/stable-diffusion-webui/wiki/Dependencies) are met and follow the instructions available for both [NVidia](https://github.com/AUTOMATIC1111/stable-diffusion-webui/wiki/Install-and-Run-on-NVidia-GPUs) (recommended) and [AMD](https://github.com/AUTOMATIC1111/stable-diffusion-webui/wiki/Install-and-Run-on-AMD-GPUs) GPUs. @@ -123,4 +124,5 @@ The documentation was moved from this README over to the project's [wiki](https: - Noise generation for outpainting mk2 - https://github.com/parlance-zz/g-diffuser-bot - CLIP interrogator idea and borrowing some code - https://github.com/pharmapsychotic/clip-interrogator - Initial Gradio script - posted on 4chan by an Anonymous user. Thank you Anonymous user. +- DeepDanbooru - interrogator for anime diffusors https://github.com/KichangKim/DeepDanbooru - (You) diff --git a/launch.py b/launch.py index 61f62096..d46426eb 100644 --- a/launch.py +++ b/launch.py @@ -33,6 +33,7 @@ def extract_arg(args, name): args, skip_torch_cuda_test = extract_arg(args, '--skip-torch-cuda-test') xformers = '--xformers' in args +deepdanbooru = '--deepdanbooru' in args def repo_dir(name): @@ -132,6 +133,9 @@ if not is_installed("xformers") and xformers: elif platform.system() == "Linux": run_pip("install xformers", "xformers") +if not is_installed("deepdanbooru") and deepdanbooru: + run_pip("install git+https://github.com/KichangKim/DeepDanbooru.git@edf73df4cdaeea2cf00e9ac08bd8a9026b7a7b26#egg=deepdanbooru[tensorflow] tensorflow==2.10.0 tensorflow-io==0.27.0", "deepdanbooru") + os.makedirs(dir_repos, exist_ok=True) git_clone("https://github.com/CompVis/stable-diffusion.git", repo_dir('stable-diffusion'), "Stable Diffusion", stable_diffusion_commit_hash) diff --git a/modules/deepbooru.py b/modules/deepbooru.py index 781b2249..7e3c0618 100644 --- a/modules/deepbooru.py +++ b/modules/deepbooru.py @@ -9,16 +9,16 @@ def _load_tf_and_return_tags(pil_image, threshold): import numpy as np this_folder = os.path.dirname(__file__) - model_path = os.path.join(this_folder, '..', 'models', 'deepbooru', 'deepdanbooru-v3-20211112-sgd-e28') - - model_good = False - for path_candidate in [model_path, os.path.dirname(model_path)]: - if os.path.exists(os.path.join(path_candidate, 'project.json')): - model_path = path_candidate - model_good = True - if not model_good: - return ("Download https://github.com/KichangKim/DeepDanbooru/releases/download/v3-20211112-sgd-e28/" - "deepdanbooru-v3-20211112-sgd-e28.zip unpack and put into models/deepbooru") + model_path = os.path.abspath(os.path.join(this_folder, '..', 'models', 'deepbooru')) + if not os.path.exists(os.path.join(model_path, 'project.json')): + # there is no point importing these every time + import zipfile + from basicsr.utils.download_util import load_file_from_url + load_file_from_url(r"https://github.com/KichangKim/DeepDanbooru/releases/download/v3-20211112-sgd-e28/deepdanbooru-v3-20211112-sgd-e28.zip", + model_path) + with zipfile.ZipFile(os.path.join(model_path, "deepdanbooru-v3-20211112-sgd-e28.zip"), "r") as zip_ref: + zip_ref.extractall(model_path) + os.remove(os.path.join(model_path, "deepdanbooru-v3-20211112-sgd-e28.zip")) tags = dd.project.load_tags_from_project(model_path) model = dd.project.load_model_from_project( diff --git a/modules/shared.py b/modules/shared.py index 02cb2722..c87b726e 100644 --- a/modules/shared.py +++ b/modules/shared.py @@ -44,6 +44,7 @@ parser.add_argument("--scunet-models-path", type=str, help="Path to directory wi parser.add_argument("--swinir-models-path", type=str, help="Path to directory with SwinIR model file(s).", default=os.path.join(models_path, 'SwinIR')) parser.add_argument("--ldsr-models-path", type=str, help="Path to directory with LDSR model file(s).", default=os.path.join(models_path, 'LDSR')) parser.add_argument("--xformers", action='store_true', help="enable xformers for cross attention layers") +parser.add_argument("--deepdanbooru", action='store_true', help="enable deepdanbooru interrogator") parser.add_argument("--opt-split-attention", action='store_true', help="force-enables cross-attention layer optimization. By default, it's on for torch.cuda and off for other torch devices.") parser.add_argument("--disable-opt-split-attention", action='store_true', help="force-disables cross-attention layer optimization") parser.add_argument("--opt-split-attention-v1", action='store_true', help="enable older version of split attention optimization that does not consume all the VRAM it can find") diff --git a/modules/ui.py b/modules/ui.py index 30583fe9..c5c11c3c 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -23,9 +23,10 @@ import gradio.utils import gradio.routes from modules import sd_hijack -from modules.deepbooru import get_deepbooru_tags from modules.paths import script_path from modules.shared import opts, cmd_opts +if cmd_opts.deepdanbooru: + from modules.deepbooru import get_deepbooru_tags import modules.shared as shared from modules.sd_samplers import samplers, samplers_for_img2img from modules.sd_hijack import model_hijack @@ -437,7 +438,10 @@ def create_toprow(is_img2img): with gr.Row(scale=1): if is_img2img: interrogate = gr.Button('Interrogate\nCLIP', elem_id="interrogate") - deepbooru = gr.Button('Interrogate\nDeepBooru', elem_id="deepbooru") + if cmd_opts.deepdanbooru: + deepbooru = gr.Button('Interrogate\nDeepBooru', elem_id="deepbooru") + else: + deepbooru = None else: interrogate = None deepbooru = None @@ -782,11 +786,12 @@ def create_ui(wrap_gradio_gpu_call): outputs=[img2img_prompt], ) - img2img_deepbooru.click( - fn=interrogate_deepbooru, - inputs=[init_img], - outputs=[img2img_prompt], - ) + if cmd_opts.deepdanbooru: + img2img_deepbooru.click( + fn=interrogate_deepbooru, + inputs=[init_img], + outputs=[img2img_prompt], + ) save.click( fn=wrap_gradio_call(save_files), diff --git a/requirements.txt b/requirements.txt index cd3953c6..81641d68 100644 --- a/requirements.txt +++ b/requirements.txt @@ -23,7 +23,4 @@ resize-right torchdiffeq kornia lark -deepdanbooru -tensorflow -tensorflow-io functorch diff --git a/requirements_versions.txt b/requirements_versions.txt index 2d256a54..fec3e9d5 100644 --- a/requirements_versions.txt +++ b/requirements_versions.txt @@ -22,7 +22,4 @@ resize-right==0.0.2 torchdiffeq==0.2.3 kornia==0.6.7 lark==1.1.2 -git+https://github.com/KichangKim/DeepDanbooru.git@edf73df4cdaeea2cf00e9ac08bd8a9026b7a7b26#egg=deepdanbooru[tensorflow] -tensorflow==2.10.0 -tensorflow-io==0.27.0 functorch==0.2.1 -- cgit v1.2.3 From 15c4278f1a18b8104e135dd82690d10cff39a2e7 Mon Sep 17 00:00:00 2001 From: DepFA <35278260+dfaker@users.noreply.github.com> Date: Sat, 8 Oct 2022 17:50:01 +0100 Subject: TI preprocess wording MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit I had to check the code to work out what splitting was 🤷🏿 --- modules/ui.py | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/ui.py b/modules/ui.py index ffd75f6a..d52d74c6 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -980,9 +980,9 @@ def create_ui(wrap_gradio_gpu_call): process_dst = gr.Textbox(label='Destination directory') with gr.Row(): - process_flip = gr.Checkbox(label='Flip') - process_split = gr.Checkbox(label='Split into two') - process_caption = gr.Checkbox(label='Add caption') + process_flip = gr.Checkbox(label='Create flipped copies') + process_split = gr.Checkbox(label='Split oversized images into two') + process_caption = gr.Checkbox(label='Use CLIP caption as filename') with gr.Row(): with gr.Column(scale=3): -- cgit v1.2.3 From b458fa48fe5734a872bca83061d702609cb52940 Mon Sep 17 00:00:00 2001 From: DepFA <35278260+dfaker@users.noreply.github.com> Date: Sat, 8 Oct 2022 17:56:28 +0100 Subject: Update ui.py --- modules/ui.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'modules/ui.py') diff --git a/modules/ui.py b/modules/ui.py index d52d74c6..b09359aa 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -982,7 +982,7 @@ def create_ui(wrap_gradio_gpu_call): with gr.Row(): process_flip = gr.Checkbox(label='Create flipped copies') process_split = gr.Checkbox(label='Split oversized images into two') - process_caption = gr.Checkbox(label='Use CLIP caption as filename') + process_caption = gr.Checkbox(label='Use BLIP caption as filename') with gr.Row(): with gr.Column(scale=3): -- cgit v1.2.3 From 432782163ae53e605470bcefc9a6f796c4556912 Mon Sep 17 00:00:00 2001 From: Aidan Holland Date: Sat, 8 Oct 2022 15:12:24 -0400 Subject: chore: Fix typos --- README.md | 2 +- javascript/imageviewer.js | 2 +- modules/interrogate.py | 4 ++-- modules/processing.py | 2 +- modules/scunet_model_arch.py | 4 ++-- modules/sd_models.py | 4 ++-- modules/sd_samplers.py | 4 ++-- modules/shared.py | 6 +++--- modules/swinir_model_arch.py | 2 +- modules/ui.py | 4 ++-- 10 files changed, 17 insertions(+), 17 deletions(-) (limited to 'modules/ui.py') diff --git a/README.md b/README.md index ef9b5e31..63dd0c18 100644 --- a/README.md +++ b/README.md @@ -34,7 +34,7 @@ Check the [custom scripts](https://github.com/AUTOMATIC1111/stable-diffusion-web - Sampling method selection - Interrupt processing at any time - 4GB video card support (also reports of 2GB working) -- Correct seeds for batches +- Correct seeds for batches - Prompt length validation - get length of prompt in tokens as you type - get a warning after generation if some text was truncated diff --git a/javascript/imageviewer.js b/javascript/imageviewer.js index 4c0e8f4b..6a00c0da 100644 --- a/javascript/imageviewer.js +++ b/javascript/imageviewer.js @@ -95,7 +95,7 @@ function showGalleryImage(){ e.addEventListener('click', function (evt) { if(!opts.js_modal_lightbox) return; - modalZoomSet(gradioApp().getElementById('modalImage'), opts.js_modal_lightbox_initialy_zoomed) + modalZoomSet(gradioApp().getElementById('modalImage'), opts.js_modal_lightbox_initially_zoomed) showModal(evt) },true); } diff --git a/modules/interrogate.py b/modules/interrogate.py index eed87144..635e266e 100644 --- a/modules/interrogate.py +++ b/modules/interrogate.py @@ -140,11 +140,11 @@ class InterrogateModels: res = caption - cilp_image = self.clip_preprocess(pil_image).unsqueeze(0).type(self.dtype).to(shared.device) + clip_image = self.clip_preprocess(pil_image).unsqueeze(0).type(self.dtype).to(shared.device) precision_scope = torch.autocast if shared.cmd_opts.precision == "autocast" else contextlib.nullcontext with torch.no_grad(), precision_scope("cuda"): - image_features = self.clip_model.encode_image(cilp_image).type(self.dtype) + image_features = self.clip_model.encode_image(clip_image).type(self.dtype) image_features /= image_features.norm(dim=-1, keepdim=True) diff --git a/modules/processing.py b/modules/processing.py index 515fc91a..31220881 100644 --- a/modules/processing.py +++ b/modules/processing.py @@ -386,7 +386,7 @@ def process_images(p: StableDiffusionProcessing) -> Processed: if state.interrupted or state.skipped: - # if we are interruped, sample returns just noise + # if we are interrupted, sample returns just noise # use the image collected previously in sampler loop samples_ddim = shared.state.current_latent diff --git a/modules/scunet_model_arch.py b/modules/scunet_model_arch.py index 972a2639..43ca8d36 100644 --- a/modules/scunet_model_arch.py +++ b/modules/scunet_model_arch.py @@ -40,7 +40,7 @@ class WMSA(nn.Module): Returns: attn_mask: should be (1 1 w p p), """ - # supporting sqaure. + # supporting square. attn_mask = torch.zeros(h, w, p, p, p, p, dtype=torch.bool, device=self.relative_position_params.device) if self.type == 'W': return attn_mask @@ -65,7 +65,7 @@ class WMSA(nn.Module): x = rearrange(x, 'b (w1 p1) (w2 p2) c -> b w1 w2 p1 p2 c', p1=self.window_size, p2=self.window_size) h_windows = x.size(1) w_windows = x.size(2) - # sqaure validation + # square validation # assert h_windows == w_windows x = rearrange(x, 'b w1 w2 p1 p2 c -> b (w1 w2) (p1 p2) c', p1=self.window_size, p2=self.window_size) diff --git a/modules/sd_models.py b/modules/sd_models.py index 9409d070..a09866ce 100644 --- a/modules/sd_models.py +++ b/modules/sd_models.py @@ -147,7 +147,7 @@ def load_model_weights(model, checkpoint_file, sd_model_hash): model.first_stage_model.load_state_dict(vae_dict) model.sd_model_hash = sd_model_hash - model.sd_model_checkpint = checkpoint_file + model.sd_model_checkpoint = checkpoint_file def load_model(): @@ -175,7 +175,7 @@ def reload_model_weights(sd_model, info=None): from modules import lowvram, devices, sd_hijack checkpoint_info = info or select_checkpoint() - if sd_model.sd_model_checkpint == checkpoint_info.filename: + if sd_model.sd_model_checkpoint == checkpoint_info.filename: return if shared.cmd_opts.lowvram or shared.cmd_opts.medvram: diff --git a/modules/sd_samplers.py b/modules/sd_samplers.py index eade0dbb..6e743f7e 100644 --- a/modules/sd_samplers.py +++ b/modules/sd_samplers.py @@ -181,7 +181,7 @@ class VanillaStableDiffusionSampler: self.initialize(p) - # existing code fails with cetain step counts, like 9 + # existing code fails with certain step counts, like 9 try: self.sampler.make_schedule(ddim_num_steps=steps, ddim_eta=self.eta, ddim_discretize=p.ddim_discretize, verbose=False) except Exception: @@ -204,7 +204,7 @@ class VanillaStableDiffusionSampler: steps = steps or p.steps - # existing code fails with cetin step counts, like 9 + # existing code fails with certain step counts, like 9 try: samples_ddim, _ = self.sampler.sample(S=steps, conditioning=conditioning, batch_size=int(x.shape[0]), shape=x[0].shape, verbose=False, unconditional_guidance_scale=p.cfg_scale, unconditional_conditioning=unconditional_conditioning, x_T=x, eta=self.eta) except Exception: diff --git a/modules/shared.py b/modules/shared.py index af8dc744..2dc092d6 100644 --- a/modules/shared.py +++ b/modules/shared.py @@ -141,9 +141,9 @@ class OptionInfo: self.section = None -def options_section(section_identifer, options_dict): +def options_section(section_identifier, options_dict): for k, v in options_dict.items(): - v.section = section_identifer + v.section = section_identifier return options_dict @@ -246,7 +246,7 @@ options_templates.update(options_section(('ui', "User interface"), { "add_model_hash_to_info": OptionInfo(True, "Add model hash to generation information"), "font": OptionInfo("", "Font for image grids that have text"), "js_modal_lightbox": OptionInfo(True, "Enable full page image viewer"), - "js_modal_lightbox_initialy_zoomed": OptionInfo(True, "Show images zoomed in by default in full page image viewer"), + "js_modal_lightbox_initially_zoomed": OptionInfo(True, "Show images zoomed in by default in full page image viewer"), "show_progress_in_title": OptionInfo(True, "Show generation progress in window title."), })) diff --git a/modules/swinir_model_arch.py b/modules/swinir_model_arch.py index 461fb354..863f42db 100644 --- a/modules/swinir_model_arch.py +++ b/modules/swinir_model_arch.py @@ -166,7 +166,7 @@ class SwinTransformerBlock(nn.Module): Args: dim (int): Number of input channels. - input_resolution (tuple[int]): Input resulotion. + input_resolution (tuple[int]): Input resolution. num_heads (int): Number of attention heads. window_size (int): Window size. shift_size (int): Shift size for SW-MSA. diff --git a/modules/ui.py b/modules/ui.py index b09359aa..b51af121 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -38,7 +38,7 @@ from modules import prompt_parser from modules.images import save_image import modules.textual_inversion.ui -# this is a fix for Windows users. Without it, javascript files will be served with text/html content-type and the bowser will not show any UI +# this is a fix for Windows users. Without it, javascript files will be served with text/html content-type and the browser will not show any UI mimetypes.init() mimetypes.add_type('application/javascript', '.js') @@ -102,7 +102,7 @@ def save_files(js_data, images, index): import csv filenames = [] - #quick dictionary to class object conversion. Its neccesary due apply_filename_pattern requiring it + #quick dictionary to class object conversion. Its necessary due apply_filename_pattern requiring it class MyObject: def __init__(self, d=None): if d is not None: -- cgit v1.2.3 From cd8673bd9b2e59bddefee8d307340d643695fe11 Mon Sep 17 00:00:00 2001 From: DepFA <35278260+dfaker@users.noreply.github.com> Date: Sun, 9 Oct 2022 05:40:57 +0100 Subject: add embed embedding to ui --- modules/ui.py | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'modules/ui.py') diff --git a/modules/ui.py b/modules/ui.py index b51af121..a5983204 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1001,7 +1001,8 @@ def create_ui(wrap_gradio_gpu_call): steps = gr.Number(label='Max steps', value=100000, precision=0) create_image_every = gr.Number(label='Save an image to log directory every N steps, 0 to disable', value=500, precision=0) save_embedding_every = gr.Number(label='Save a copy of embedding to log directory every N steps, 0 to disable', value=500, precision=0) - + save_image_with_stored_embedding = gr.Checkbox(label='Save images with embedding in PNG chunks', value=True) + with gr.Row(): with gr.Column(scale=2): gr.HTML(value="") @@ -1063,6 +1064,7 @@ def create_ui(wrap_gradio_gpu_call): create_image_every, save_embedding_every, template_file, + save_image_with_stored_embedding, ], outputs=[ ti_output, -- cgit v1.2.3 From 5ab7e88d9b0bb0125af9f7237242a00a93360ce5 Mon Sep 17 00:00:00 2001 From: aoirusann <82883326+aoirusann@users.noreply.github.com> Date: Sat, 8 Oct 2022 13:09:29 +0800 Subject: Add `Download` & `Download as zip` --- modules/ui.py | 39 ++++++++++++++++++++++++++++++++++----- 1 file changed, 34 insertions(+), 5 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/ui.py b/modules/ui.py index b51af121..fe7f10a7 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -98,9 +98,10 @@ def send_gradio_gallery_to_image(x): return image_from_url_text(x[0]) -def save_files(js_data, images, index): +def save_files(js_data, images, do_make_zip, index): import csv filenames = [] + fullfns = [] #quick dictionary to class object conversion. Its necessary due apply_filename_pattern requiring it class MyObject: @@ -141,10 +142,22 @@ def save_files(js_data, images, index): filename = os.path.relpath(fullfn, path) filenames.append(filename) + fullfns.append(fullfn) writer.writerow([data["prompt"], data["seed"], data["width"], data["height"], data["sampler"], data["cfg_scale"], data["steps"], filenames[0], data["negative_prompt"]]) - return '', '', plaintext_to_html(f"Saved: {filenames[0]}") + # Make Zip + if do_make_zip: + zip_filepath = os.path.join(path, "images.zip") + + from zipfile import ZipFile + with ZipFile(zip_filepath, "w") as zip_file: + for i in range(len(fullfns)): + with open(fullfns[i], mode="rb") as f: + zip_file.writestr(filenames[i], f.read()) + fullfns.insert(0, zip_filepath) + + return fullfns, '', '', plaintext_to_html(f"Saved: {filenames[0]}") def wrap_gradio_call(func, extra_outputs=None): @@ -521,6 +534,12 @@ def create_ui(wrap_gradio_gpu_call): button_id = "hidden_element" if shared.cmd_opts.hide_ui_dir_config else 'open_folder' open_txt2img_folder = gr.Button(folder_symbol, elem_id=button_id) + with gr.Row(): + do_make_zip = gr.Checkbox(label="Make Zip when Save?", value=False) + + with gr.Row(): + download_files = gr.File(None, file_count="multiple", interactive=False, show_label=False) + with gr.Group(): html_info = gr.HTML() generation_info = gr.Textbox(visible=False) @@ -570,13 +589,15 @@ def create_ui(wrap_gradio_gpu_call): save.click( fn=wrap_gradio_call(save_files), - _js="(x, y, z) => [x, y, selected_gallery_index()]", + _js="(x, y, z, w) => [x, y, z, selected_gallery_index()]", inputs=[ generation_info, txt2img_gallery, + do_make_zip, html_info, ], outputs=[ + download_files, html_info, html_info, html_info, @@ -701,6 +722,12 @@ def create_ui(wrap_gradio_gpu_call): button_id = "hidden_element" if shared.cmd_opts.hide_ui_dir_config else 'open_folder' open_img2img_folder = gr.Button(folder_symbol, elem_id=button_id) + with gr.Row(): + do_make_zip = gr.Checkbox(label="Make Zip when Save?", value=False) + + with gr.Row(): + download_files = gr.File(None, file_count="multiple", interactive=False, show_label=False) + with gr.Group(): html_info = gr.HTML() generation_info = gr.Textbox(visible=False) @@ -776,13 +803,15 @@ def create_ui(wrap_gradio_gpu_call): save.click( fn=wrap_gradio_call(save_files), - _js="(x, y, z) => [x, y, selected_gallery_index()]", + _js="(x, y, z, w) => [x, y, z, selected_gallery_index()]", inputs=[ generation_info, img2img_gallery, - html_info + do_make_zip, + html_info, ], outputs=[ + download_files, html_info, html_info, html_info, -- cgit v1.2.3 From 14192c5b207b16b1ec7a4c9c4ea538d1a6811a4d Mon Sep 17 00:00:00 2001 From: aoirusann Date: Sun, 9 Oct 2022 13:01:10 +0800 Subject: Support `Download` for txt files. --- modules/images.py | 39 +++++++++++++++++++++++++++++++++++++-- modules/ui.py | 5 ++++- 2 files changed, 41 insertions(+), 3 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/images.py b/modules/images.py index 29c5ee24..c0a90676 100644 --- a/modules/images.py +++ b/modules/images.py @@ -349,6 +349,38 @@ def get_next_sequence_number(path, basename): def save_image(image, path, basename, seed=None, prompt=None, extension='png', info=None, short_filename=False, no_prompt=False, grid=False, pnginfo_section_name='parameters', p=None, existing_info=None, forced_filename=None, suffix="", save_to_dirs=None): + '''Save an image. + + Args: + image (`PIL.Image`): + The image to be saved. + path (`str`): + The directory to save the image. Note, the option `save_to_dirs` will make the image to be saved into a sub directory. + basename (`str`): + The base filename which will be applied to `filename pattern`. + seed, prompt, short_filename, + extension (`str`): + Image file extension, default is `png`. + pngsectionname (`str`): + Specify the name of the section which `info` will be saved in. + info (`str` or `PngImagePlugin.iTXt`): + PNG info chunks. + existing_info (`dict`): + Additional PNG info. `existing_info == {pngsectionname: info, ...}` + no_prompt: + TODO I don't know its meaning. + p (`StableDiffusionProcessing`) + forced_filename (`str`): + If specified, `basename` and filename pattern will be ignored. + save_to_dirs (bool): + If true, the image will be saved into a subdirectory of `path`. + + Returns: (fullfn, txt_fullfn) + fullfn (`str`): + The full path of the saved imaged. + txt_fullfn (`str` or None): + If a text file is saved for this image, this will be its full path. Otherwise None. + ''' if short_filename or prompt is None or seed is None: file_decoration = "" elif opts.save_to_dirs: @@ -424,7 +456,10 @@ def save_image(image, path, basename, seed=None, prompt=None, extension='png', i piexif.insert(exif_bytes(), fullfn_without_extension + ".jpg") if opts.save_txt and info is not None: - with open(f"{fullfn_without_extension}.txt", "w", encoding="utf8") as file: + txt_fullfn = f"{fullfn_without_extension}.txt" + with open(txt_fullfn, "w", encoding="utf8") as file: file.write(info + "\n") + else: + txt_fullfn = None - return fullfn + return fullfn, txt_fullfn diff --git a/modules/ui.py b/modules/ui.py index fe7f10a7..debd8873 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -138,11 +138,14 @@ def save_files(js_data, images, do_make_zip, index): is_grid = image_index < p.index_of_first_image i = 0 if is_grid else (image_index - p.index_of_first_image) - fullfn = save_image(image, path, "", seed=p.all_seeds[i], prompt=p.all_prompts[i], extension=extension, info=p.infotexts[image_index], grid=is_grid, p=p, save_to_dirs=save_to_dirs) + fullfn, txt_fullfn = save_image(image, path, "", seed=p.all_seeds[i], prompt=p.all_prompts[i], extension=extension, info=p.infotexts[image_index], grid=is_grid, p=p, save_to_dirs=save_to_dirs) filename = os.path.relpath(fullfn, path) filenames.append(filename) fullfns.append(fullfn) + if txt_fullfn: + filenames.append(os.path.basename(txt_fullfn)) + fullfns.append(txt_fullfn) writer.writerow([data["prompt"], data["seed"], data["width"], data["height"], data["sampler"], data["cfg_scale"], data["steps"], filenames[0], data["negative_prompt"]]) -- cgit v1.2.3 From ab4fe4f44c3d2675a351269fe2ff1ddeac557aa6 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Sun, 9 Oct 2022 11:59:41 +0300 Subject: hide filenames for save button by default --- modules/ui.py | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/ui.py b/modules/ui.py index 8071b1cb..e1ab2665 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -162,7 +162,7 @@ def save_files(js_data, images, do_make_zip, index): zip_file.writestr(filenames[i], f.read()) fullfns.insert(0, zip_filepath) - return fullfns, '', '', plaintext_to_html(f"Saved: {filenames[0]}") + return gr.File.update(value=fullfns, visible=True), '', '', plaintext_to_html(f"Saved: {filenames[0]}") def wrap_gradio_call(func, extra_outputs=None): @@ -553,7 +553,7 @@ def create_ui(wrap_gradio_gpu_call): do_make_zip = gr.Checkbox(label="Make Zip when Save?", value=False) with gr.Row(): - download_files = gr.File(None, file_count="multiple", interactive=False, show_label=False) + download_files = gr.File(None, file_count="multiple", interactive=False, show_label=False, visible=False) with gr.Group(): html_info = gr.HTML() @@ -741,7 +741,7 @@ def create_ui(wrap_gradio_gpu_call): do_make_zip = gr.Checkbox(label="Make Zip when Save?", value=False) with gr.Row(): - download_files = gr.File(None, file_count="multiple", interactive=False, show_label=False) + download_files = gr.File(None, file_count="multiple", interactive=False, show_label=False, visible=False) with gr.Group(): html_info = gr.HTML() -- cgit v1.2.3 From 6f6798ddabe10d320fe8ea05edf0fdcef0c51a8e Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Sun, 9 Oct 2022 12:33:37 +0300 Subject: prevent a possible code execution error (thanks, RyotaK) --- modules/ui.py | 9 +++++++++ 1 file changed, 9 insertions(+) (limited to 'modules/ui.py') diff --git a/modules/ui.py b/modules/ui.py index e1ab2665..dad509f3 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1153,6 +1153,15 @@ def create_ui(wrap_gradio_gpu_call): component_dict = {} def open_folder(f): + if not os.path.isdir(f): + print(f""" +WARNING +An open_folder request was made with an argument that is not a folder. +This could be an error or a malicious attempt to run code on your computer. +Requested path was: {f} +""", file=sys.stderr) + return + if not shared.cmd_opts.hide_ui_dir_config: path = os.path.normpath(f) if platform.system() == "Windows": -- cgit v1.2.3 From 6c383d2e82045fc4475d665f83bdeeac8fd844d9 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Sun, 9 Oct 2022 22:24:07 +0300 Subject: show model selection setting on top of page --- modules/shared.py | 5 +++-- modules/ui.py | 54 +++++++++++++++++++++++++++++++++++++++++++++--------- style.css | 9 +++++++++ 3 files changed, 57 insertions(+), 11 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/shared.py b/modules/shared.py index 3d7f08e1..270fa402 100644 --- a/modules/shared.py +++ b/modules/shared.py @@ -131,13 +131,14 @@ def realesrgan_models_names(): class OptionInfo: - def __init__(self, default=None, label="", component=None, component_args=None, onchange=None): + def __init__(self, default=None, label="", component=None, component_args=None, onchange=None, show_on_main_page=False): self.default = default self.label = label self.component = component self.component_args = component_args self.onchange = onchange self.section = None + self.show_on_main_page = show_on_main_page def options_section(section_identifier, options_dict): @@ -214,7 +215,7 @@ options_templates.update(options_section(('system', "System"), { })) options_templates.update(options_section(('sd', "Stable Diffusion"), { - "sd_model_checkpoint": OptionInfo(None, "Stable Diffusion checkpoint", gr.Dropdown, lambda: {"choices": modules.sd_models.checkpoint_tiles()}), + "sd_model_checkpoint": OptionInfo(None, "Stable Diffusion checkpoint", gr.Dropdown, lambda: {"choices": modules.sd_models.checkpoint_tiles()}, show_on_main_page=True), "sd_hypernetwork": OptionInfo("None", "Stable Diffusion finetune hypernetwork", gr.Dropdown, lambda: {"choices": ["None"] + [x for x in hypernetworks.keys()]}), "img2img_color_correction": OptionInfo(False, "Apply color correction to img2img results to match original colors."), "save_images_before_color_correction": OptionInfo(False, "Save a copy of image before applying color correction to img2img results"), diff --git a/modules/ui.py b/modules/ui.py index dad509f3..2231a8ed 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1175,10 +1175,13 @@ Requested path was: {f} changed = 0 for key, value, comp in zip(opts.data_labels.keys(), args, components): - if not opts.same_type(value, opts.data_labels[key].default): - return f"Bad value for setting {key}: {value}; expecting {type(opts.data_labels[key].default).__name__}" + if comp != dummy_component and not opts.same_type(value, opts.data_labels[key].default): + return f"Bad value for setting {key}: {value}; expecting {type(opts.data_labels[key].default).__name__}", opts.dumpjson() for key, value, comp in zip(opts.data_labels.keys(), args, components): + if comp == dummy_component: + continue + comp_args = opts.data_labels[key].component_args if comp_args and isinstance(comp_args, dict) and comp_args.get('visible') is False: continue @@ -1196,6 +1199,21 @@ Requested path was: {f} return f'{changed} settings changed.', opts.dumpjson() + def run_settings_single(value, key): + if not opts.same_type(value, opts.data_labels[key].default): + return gr.update(visible=True), opts.dumpjson() + + oldval = opts.data.get(key, None) + opts.data[key] = value + + if oldval != value: + if opts.data_labels[key].onchange is not None: + opts.data_labels[key].onchange() + + opts.save(shared.config_filename) + + return gr.update(value=value), opts.dumpjson() + with gr.Blocks(analytics_enabled=False) as settings_interface: settings_submit = gr.Button(value="Apply settings", variant='primary') result = gr.HTML() @@ -1203,6 +1221,8 @@ Requested path was: {f} settings_cols = 3 items_per_col = int(len(opts.data_labels) * 0.9 / settings_cols) + quicksettings_list = [] + cols_displayed = 0 items_displayed = 0 previous_section = None @@ -1225,10 +1245,14 @@ Requested path was: {f} gr.HTML(elem_id="settings_header_text_{}".format(item.section[0]), value='

{}

'.format(item.section[1])) - component = create_setting_component(k) - component_dict[k] = component - components.append(component) - items_displayed += 1 + if item.show_on_main_page: + quicksettings_list.append((i, k, item)) + components.append(dummy_component) + else: + component = create_setting_component(k) + component_dict[k] = component + components.append(component) + items_displayed += 1 request_notifications = gr.Button(value='Request browser notifications', elem_id="request_notifications") request_notifications.click( @@ -1242,7 +1266,6 @@ Requested path was: {f} reload_script_bodies = gr.Button(value='Reload custom script bodies (No ui updates, No restart)', variant='secondary') restart_gradio = gr.Button(value='Restart Gradio and Refresh components (Custom Scripts, ui.py, js and css only)', variant='primary') - def reload_scripts(): modules.scripts.reload_script_body_only() @@ -1289,7 +1312,11 @@ Requested path was: {f} css += css_hide_progressbar with gr.Blocks(css=css, analytics_enabled=False, title="Stable Diffusion") as demo: - + with gr.Row(elem_id="quicksettings"): + for i, k, item in quicksettings_list: + component = create_setting_component(k) + component_dict[k] = component + settings_interface.gradio_ref = demo with gr.Tabs() as tabs: @@ -1306,7 +1333,16 @@ Requested path was: {f} inputs=components, outputs=[result, text_settings], ) - + + for i, k, item in quicksettings_list: + component = component_dict[k] + + component.change( + fn=lambda value, k=k: run_settings_single(value, key=k), + inputs=[component], + outputs=[component, text_settings], + ) + def modelmerger(*args): try: results = modules.extras.run_modelmerger(*args) diff --git a/style.css b/style.css index 101d2052..28160bdf 100644 --- a/style.css +++ b/style.css @@ -453,3 +453,12 @@ input[type="range"]{ .context-menu-items a:hover{ background: #a55000; } + +#quicksettings > div{ + border: none; +} + +#quicksettings > div > div{ + max-width: 32em; + padding: 0; +} -- cgit v1.2.3 From 6435691bb11c5a35703720bfd2a875f24c066f86 Mon Sep 17 00:00:00 2001 From: Justin Maier Date: Sun, 9 Oct 2022 19:26:52 -0600 Subject: Add "Scale to" option to Extras --- javascript/ui.js | 3 ++- modules/extras.py | 28 +++++++++++++++++++++++----- modules/ui.py | 38 +++++++++++++++++++++++++------------- 3 files changed, 50 insertions(+), 19 deletions(-) (limited to 'modules/ui.py') diff --git a/javascript/ui.js b/javascript/ui.js index b1053201..4100944e 100644 --- a/javascript/ui.js +++ b/javascript/ui.js @@ -101,7 +101,8 @@ function create_tab_index_args(tabId, args){ } function get_extras_tab_index(){ - return create_tab_index_args('mode_extras', arguments) + const [,,...args] = [...arguments] + return [get_tab_index('mode_extras'), get_tab_index('extras_resize_mode'), ...args] } function create_submit_args(args){ diff --git a/modules/extras.py b/modules/extras.py index 41e8612c..83ca7049 100644 --- a/modules/extras.py +++ b/modules/extras.py @@ -1,3 +1,4 @@ +import math import os import numpy as np @@ -19,7 +20,7 @@ import gradio as gr cached_images = {} -def run_extras(extras_mode, image, image_folder, gfpgan_visibility, codeformer_visibility, codeformer_weight, upscaling_resize, extras_upscaler_1, extras_upscaler_2, extras_upscaler_2_visibility): +def run_extras(extras_mode, resize_mode, image, image_folder, gfpgan_visibility, codeformer_visibility, codeformer_weight, upscaling_resize, upscaling_resize_w, upscaling_resize_h, upscaling_crop, extras_upscaler_1, extras_upscaler_2, extras_upscaler_2_visibility): devices.torch_gc() imageArr = [] @@ -67,8 +68,23 @@ def run_extras(extras_mode, image, image_folder, gfpgan_visibility, codeformer_v info += f"CodeFormer w: {round(codeformer_weight, 2)}, CodeFormer visibility:{round(codeformer_visibility, 2)}\n" image = res + if resize_mode == 1: + upscaling_resize = max(upscaling_resize_w/image.width, upscaling_resize_h/image.height) + crop_info = " (crop)" if upscaling_crop else "" + info += f"Resize to: {upscaling_resize_w:g}x{upscaling_resize_h:g}{crop_info}\n" + + def crop_upscaled_center(image, resize_w, resize_h): + left = int(math.ceil((image.width - resize_w) / 2)) + right = image.width - int(math.floor((image.width - resize_w) / 2)) + top = int(math.ceil((image.height - resize_h) / 2)) + bottom = image.height - int(math.floor((image.height - resize_h) / 2)) + + image = image.crop((left, top, right, bottom)) + return image + + if upscaling_resize != 1.0: - def upscale(image, scaler_index, resize): + def upscale(image, scaler_index, resize, mode, resize_w, resize_h, crop): small = image.crop((image.width // 2, image.height // 2, image.width // 2 + 10, image.height // 2 + 10)) pixels = tuple(np.array(small).flatten().tolist()) key = (resize, scaler_index, image.width, image.height, gfpgan_visibility, codeformer_visibility, codeformer_weight) + pixels @@ -77,15 +93,17 @@ def run_extras(extras_mode, image, image_folder, gfpgan_visibility, codeformer_v if c is None: upscaler = shared.sd_upscalers[scaler_index] c = upscaler.scaler.upscale(image, resize, upscaler.data_path) + if mode == 1 and crop: + c = crop_upscaled_center(c, resize_w, resize_h) cached_images[key] = c return c info += f"Upscale: {round(upscaling_resize, 3)}, model:{shared.sd_upscalers[extras_upscaler_1].name}\n" - res = upscale(image, extras_upscaler_1, upscaling_resize) + res = upscale(image, extras_upscaler_1, upscaling_resize, resize_mode, upscaling_resize_w, upscaling_resize_h, upscaling_crop) if extras_upscaler_2 != 0 and extras_upscaler_2_visibility > 0: - res2 = upscale(image, extras_upscaler_2, upscaling_resize) + res2 = upscale(image, extras_upscaler_2, upscaling_resize, resize_mode, upscaling_resize_w, upscaling_resize_h, upscaling_crop) info += f"Upscale: {round(upscaling_resize, 3)}, visibility: {round(extras_upscaler_2_visibility, 3)}, model:{shared.sd_upscalers[extras_upscaler_2].name}\n" res = Image.blend(res, res2, extras_upscaler_2_visibility) @@ -190,7 +208,7 @@ def run_modelmerger(primary_model_name, secondary_model_name, interp_method, int theta_0[key] = theta_func(theta_0[key], theta_1[key], (float(1.0) - interp_amount)) # Need to reverse the interp_amount to match the desired mix ration in the merged checkpoint if save_as_half: theta_0[key] = theta_0[key].half() - + for key in theta_1.keys(): if 'model' in key and key not in theta_0: theta_0[key] = theta_1[key] diff --git a/modules/ui.py b/modules/ui.py index 2231a8ed..4bb2892b 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -101,7 +101,7 @@ def send_gradio_gallery_to_image(x): def save_files(js_data, images, do_make_zip, index): - import csv + import csv filenames = [] fullfns = [] @@ -551,7 +551,7 @@ def create_ui(wrap_gradio_gpu_call): with gr.Row(): do_make_zip = gr.Checkbox(label="Make Zip when Save?", value=False) - + with gr.Row(): download_files = gr.File(None, file_count="multiple", interactive=False, show_label=False, visible=False) @@ -739,7 +739,7 @@ def create_ui(wrap_gradio_gpu_call): with gr.Row(): do_make_zip = gr.Checkbox(label="Make Zip when Save?", value=False) - + with gr.Row(): download_files = gr.File(None, file_count="multiple", interactive=False, show_label=False, visible=False) @@ -903,7 +903,15 @@ def create_ui(wrap_gradio_gpu_call): with gr.TabItem('Batch Process'): image_batch = gr.File(label="Batch Process", file_count="multiple", interactive=True, type="file") - upscaling_resize = gr.Slider(minimum=1.0, maximum=4.0, step=0.05, label="Resize", value=2) + with gr.Tabs(elem_id="extras_resize_mode"): + with gr.TabItem('Scale by'): + upscaling_resize = gr.Slider(minimum=1.0, maximum=4.0, step=0.05, label="Resize", value=2) + with gr.TabItem('Scale to'): + with gr.Group(): + with gr.Row(): + upscaling_resize_w = gr.Number(label="Width", value=512) + upscaling_resize_h = gr.Number(label="Height", value=512) + upscaling_crop = gr.Checkbox(label='Crop to fit', value=True) with gr.Group(): extras_upscaler_1 = gr.Radio(label='Upscaler 1', choices=[x.name for x in shared.sd_upscalers], value=shared.sd_upscalers[0].name, type="index") @@ -934,6 +942,7 @@ def create_ui(wrap_gradio_gpu_call): fn=wrap_gradio_gpu_call(modules.extras.run_extras), _js="get_extras_tab_index", inputs=[ + dummy_component, dummy_component, extras_image, image_batch, @@ -941,6 +950,9 @@ def create_ui(wrap_gradio_gpu_call): codeformer_visibility, codeformer_weight, upscaling_resize, + upscaling_resize_w, + upscaling_resize_h, + upscaling_crop, extras_upscaler_1, extras_upscaler_2, extras_upscaler_2_visibility, @@ -951,14 +963,14 @@ def create_ui(wrap_gradio_gpu_call): html_info, ] ) - + extras_send_to_img2img.click( fn=lambda x: image_from_url_text(x), _js="extract_image_from_gallery_img2img", inputs=[result_images], outputs=[init_img], ) - + extras_send_to_inpaint.click( fn=lambda x: image_from_url_text(x), _js="extract_image_from_gallery_img2img", @@ -1286,7 +1298,7 @@ Requested path was: {f} outputs=[], _js='function(){restart_reload()}' ) - + if column is not None: column.__exit__() @@ -1318,12 +1330,12 @@ Requested path was: {f} component_dict[k] = component settings_interface.gradio_ref = demo - + with gr.Tabs() as tabs: for interface, label, ifid in interfaces: with gr.TabItem(label, id=ifid): interface.render() - + if os.path.exists(os.path.join(script_path, "notification.mp3")): audio_notification = gr.Audio(interactive=False, value=os.path.join(script_path, "notification.mp3"), elem_id="audio_notification", visible=False) @@ -1456,10 +1468,10 @@ Requested path was: {f} if getattr(obj,'custom_script_source',None) is not None: key = 'customscript/' + obj.custom_script_source + '/' + key - + if getattr(obj, 'do_not_save_to_config', False): return - + saved_value = ui_settings.get(key, None) if saved_value is None: ui_settings[key] = getattr(obj, field) @@ -1483,10 +1495,10 @@ Requested path was: {f} if type(x) == gr.Textbox: apply_field(x, 'value') - + if type(x) == gr.Number: apply_field(x, 'value') - + visit(txt2img_interface, loadsave, "txt2img") visit(img2img_interface, loadsave, "img2img") visit(extras_interface, loadsave, "extras") -- cgit v1.2.3 From 1f92336be768d235c18a82acb2195b7135101ae7 Mon Sep 17 00:00:00 2001 From: JC_Array Date: Sun, 9 Oct 2022 23:58:18 -0500 Subject: refactored the deepbooru module to improve speed on running multiple interogations in a row. Added the option to generate deepbooru tags for textual inversion preproccessing. --- modules/deepbooru.py | 84 +++++++++++++++++++++++++-------- modules/textual_inversion/preprocess.py | 22 ++++++++- modules/ui.py | 52 ++++++++++++++------ 3 files changed, 122 insertions(+), 36 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/deepbooru.py b/modules/deepbooru.py index 7e3c0618..cee4a3b4 100644 --- a/modules/deepbooru.py +++ b/modules/deepbooru.py @@ -1,21 +1,74 @@ import os.path from concurrent.futures import ProcessPoolExecutor -from multiprocessing import get_context +import multiprocessing -def _load_tf_and_return_tags(pil_image, threshold): +def get_deepbooru_tags(pil_image, threshold=0.5): + """ + This method is for running only one image at a time for simple use. Used to the img2img interrogate. + """ + from modules import shared # prevents circular reference + create_deepbooru_process(threshold) + shared.deepbooru_process_return["value"] = -1 + shared.deepbooru_process_queue.put(pil_image) + while shared.deepbooru_process_return["value"] == -1: + time.sleep(0.2) + release_process() + return ret + + +def deepbooru_process(queue, deepbooru_process_return, threshold): + model, tags = get_deepbooru_tags_model() + while True: # while process is running, keep monitoring queue for new image + pil_image = queue.get() + if pil_image == "QUIT": + break + else: + deepbooru_process_return["value"] = get_deepbooru_tags_from_model(model, tags, pil_image, threshold) + + +def create_deepbooru_process(threshold=0.5): + """ + Creates deepbooru process. A queue is created to send images into the process. This enables multiple images + to be processed in a row without reloading the model or creating a new process. To return the data, a shared + dictionary is created to hold the tags created. To wait for tags to be returned, a value of -1 is assigned + to the dictionary and the method adding the image to the queue should wait for this value to be updated with + the tags. + """ + from modules import shared # prevents circular reference + shared.deepbooru_process_manager = multiprocessing.Manager() + shared.deepbooru_process_queue = shared.deepbooru_process_manager.Queue() + shared.deepbooru_process_return = shared.deepbooru_process_manager.dict() + shared.deepbooru_process_return["value"] = -1 + shared.deepbooru_process = multiprocessing.Process(target=deepbooru_process, args=(shared.deepbooru_process_queue, shared.deepbooru_process_return, threshold)) + shared.deepbooru_process.start() + + +def release_process(): + """ + Stops the deepbooru process to return used memory + """ + from modules import shared # prevents circular reference + shared.deepbooru_process_queue.put("QUIT") + shared.deepbooru_process.join() + shared.deepbooru_process_queue = None + shared.deepbooru_process = None + shared.deepbooru_process_return = None + shared.deepbooru_process_manager = None + +def get_deepbooru_tags_model(): import deepdanbooru as dd import tensorflow as tf import numpy as np - this_folder = os.path.dirname(__file__) model_path = os.path.abspath(os.path.join(this_folder, '..', 'models', 'deepbooru')) if not os.path.exists(os.path.join(model_path, 'project.json')): # there is no point importing these every time import zipfile from basicsr.utils.download_util import load_file_from_url - load_file_from_url(r"https://github.com/KichangKim/DeepDanbooru/releases/download/v3-20211112-sgd-e28/deepdanbooru-v3-20211112-sgd-e28.zip", - model_path) + load_file_from_url( + r"https://github.com/KichangKim/DeepDanbooru/releases/download/v3-20211112-sgd-e28/deepdanbooru-v3-20211112-sgd-e28.zip", + model_path) with zipfile.ZipFile(os.path.join(model_path, "deepdanbooru-v3-20211112-sgd-e28.zip"), "r") as zip_ref: zip_ref.extractall(model_path) os.remove(os.path.join(model_path, "deepdanbooru-v3-20211112-sgd-e28.zip")) @@ -24,7 +77,13 @@ def _load_tf_and_return_tags(pil_image, threshold): model = dd.project.load_model_from_project( model_path, compile_model=True ) + return model, tags + +def get_deepbooru_tags_from_model(model, tags, pil_image, threshold=0.5): + import deepdanbooru as dd + import tensorflow as tf + import numpy as np width = model.input_shape[2] height = model.input_shape[1] image = np.array(pil_image) @@ -57,17 +116,4 @@ def _load_tf_and_return_tags(pil_image, threshold): print('\n'.join(sorted(result_tags_print, reverse=True))) - return ', '.join(result_tags_out).replace('_', ' ').replace(':', ' ') - - -def subprocess_init_no_cuda(): - import os - os.environ["CUDA_VISIBLE_DEVICES"] = "-1" - - -def get_deepbooru_tags(pil_image, threshold=0.5): - context = get_context('spawn') - with ProcessPoolExecutor(initializer=subprocess_init_no_cuda, mp_context=context) as executor: - f = executor.submit(_load_tf_and_return_tags, pil_image, threshold, ) - ret = f.result() # will rethrow any exceptions - return ret \ No newline at end of file + return ', '.join(result_tags_out).replace('_', ' ').replace(':', ' ') \ No newline at end of file diff --git a/modules/textual_inversion/preprocess.py b/modules/textual_inversion/preprocess.py index f1c002a2..9f63c9a4 100644 --- a/modules/textual_inversion/preprocess.py +++ b/modules/textual_inversion/preprocess.py @@ -3,11 +3,14 @@ from PIL import Image, ImageOps import platform import sys import tqdm +import time from modules import shared, images +from modules.shared import opts, cmd_opts +if cmd_opts.deepdanbooru: + import modules.deepbooru as deepbooru - -def preprocess(process_src, process_dst, process_flip, process_split, process_caption): +def preprocess(process_src, process_dst, process_flip, process_split, process_caption, process_caption_deepbooru=False): size = 512 src = os.path.abspath(process_src) dst = os.path.abspath(process_dst) @@ -24,10 +27,21 @@ def preprocess(process_src, process_dst, process_flip, process_split, process_ca if process_caption: shared.interrogator.load() + if process_caption_deepbooru: + deepbooru.create_deepbooru_process() + def save_pic_with_caption(image, index): if process_caption: caption = "-" + shared.interrogator.generate_caption(image) caption = sanitize_caption(os.path.join(dst, f"{index:05}-{subindex[0]}"), caption, ".png") + elif process_caption_deepbooru: + shared.deepbooru_process_return["value"] = -1 + shared.deepbooru_process_queue.put(image) + while shared.deepbooru_process_return["value"] == -1: + time.sleep(0.2) + caption = "-" + shared.deepbooru_process_return["value"] + caption = sanitize_caption(os.path.join(dst, f"{index:05}-{subindex[0]}"), caption, ".png") + shared.deepbooru_process_return["value"] = -1 else: caption = filename caption = os.path.splitext(caption)[0] @@ -79,6 +93,10 @@ def preprocess(process_src, process_dst, process_flip, process_split, process_ca if process_caption: shared.interrogator.send_blip_to_ram() + if process_caption_deepbooru: + deepbooru.release_process() + + def sanitize_caption(base_path, original_caption, suffix): operating_system = platform.system().lower() if (operating_system == "windows"): diff --git a/modules/ui.py b/modules/ui.py index 2231a8ed..179e3a83 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1034,6 +1034,9 @@ def create_ui(wrap_gradio_gpu_call): process_flip = gr.Checkbox(label='Create flipped copies') process_split = gr.Checkbox(label='Split oversized images into two') process_caption = gr.Checkbox(label='Use BLIP caption as filename') + if cmd_opts.deepdanbooru: + process_caption_deepbooru = gr.Checkbox(label='Use deepbooru caption as filename') + with gr.Row(): with gr.Column(scale=3): @@ -1086,21 +1089,40 @@ def create_ui(wrap_gradio_gpu_call): ] ) - run_preprocess.click( - fn=wrap_gradio_gpu_call(modules.textual_inversion.ui.preprocess, extra_outputs=[gr.update()]), - _js="start_training_textual_inversion", - inputs=[ - process_src, - process_dst, - process_flip, - process_split, - process_caption, - ], - outputs=[ - ti_output, - ti_outcome, - ], - ) + if cmd_opts.deepdanbooru: + # if process_caption_deepbooru is None, it will cause an error, as a result only include it if it is enabled + run_preprocess.click( + fn=wrap_gradio_gpu_call(modules.textual_inversion.ui.preprocess, extra_outputs=[gr.update()]), + _js="start_training_textual_inversion", + inputs=[ + process_src, + process_dst, + process_flip, + process_split, + process_caption, + process_caption_deepbooru, + ], + outputs=[ + ti_output, + ti_outcome, + ], + ) + else: + run_preprocess.click( + fn=wrap_gradio_gpu_call(modules.textual_inversion.ui.preprocess, extra_outputs=[gr.update()]), + _js="start_training_textual_inversion", + inputs=[ + process_src, + process_dst, + process_flip, + process_split, + process_caption, + ], + outputs=[ + ti_output, + ti_outcome, + ], + ) train_embedding.click( fn=wrap_gradio_gpu_call(modules.textual_inversion.ui.train_embedding, extra_outputs=[gr.update()]), -- cgit v1.2.3 From 8a7c07a2140c98bceca858087525d77fd0352fda Mon Sep 17 00:00:00 2001 From: yfszzx Date: Mon, 10 Oct 2022 15:39:39 +0800 Subject: show image history --- javascript/images_history.js | 66 ++++++++++++++++++++++ javascript/jquery-3.6.0.min.js | 2 + modules/images_history.py | 90 ++++++++++++++++++++++++++++++ modules/ui.py | 11 +++- repositorieslatent-diffusion | 1 + testui.py | 124 +++++++++++++++++++++++++++++++++++++++++ 6 files changed, 292 insertions(+), 2 deletions(-) create mode 100644 javascript/images_history.js create mode 100644 javascript/jquery-3.6.0.min.js create mode 100644 modules/images_history.py create mode 160000 repositorieslatent-diffusion create mode 100644 testui.py (limited to 'modules/ui.py') diff --git a/javascript/images_history.js b/javascript/images_history.js new file mode 100644 index 00000000..f30b7eff --- /dev/null +++ b/javascript/images_history.js @@ -0,0 +1,66 @@ +function init_images_history(){ + if (gradioApp().getElementById('txt2img_images_history_first_page') == null) { + setTimeout(init_images_history, 1000) + } else { + tab_list = ["txt2img", "img2img"] + for (i in tab_list){ + tab = tab_list[i] + gradioApp().getElementById(tab + "_images_history_first_page").click() + $(gradioApp().getElementById(tab + '_images_history')).addClass("images_history_gallery") + item = $(gradioApp().getElementById(tab + '_images_history_set_index')) + item.addClass("images_history_set_index") + item.hide() + } + } + +} +setTimeout(init_images_history, 1000) +onUiUpdate(function(){ + fullImg_preview = gradioApp().querySelectorAll('#txt2img_images_history img.w-full') + if(fullImg_preview.length > 0){ + fullImg_preview.forEach(set_history_index_from_img); + } + fullImg_preview = gradioApp().querySelectorAll('#img2img_images_history img.w-full') + if(fullImg_preview.length > 0){ + fullImg_preview.forEach(set_history_index_from_img); + } +}) + +function set_history_gallery_index(item){ + buttons = item.find(".gallery-item") + // alert(item.attr("id") + " " + buttons.length) + index = -1 + i = 0 + buttons.each(function(){ + if($(this).hasClass("!ring-2")){ index = i } + i += 1 + }) + if (index == -1){ + setTimeout(set_history_gallery_index, 10, item) + } else { + item = item.find(".images_history_set_index").first() + item.attr("img_index", index) + item.click() + } +} +function set_history_index_from_img(e){ + if(e && e.parentElement.tagName == 'BUTTON'){ + bnt = $(e).parent() + if (bnt.hasClass("transform")){ + bnt.off("click").on("click",function(){ + set_history_gallery_index($(this).parents(".images_history_gallery").first()) + }) + } else { + bnt.off("mousedown").on("mousedown", function(){ + set_history_gallery_index($(this).parents(".images_history_gallery").first()) + }) + + } + } +} +function images_history_get_current_img(is_image2image){ + head = is_image2image?"img2img":"txt2img" + s = $(gradioApp().getElementById(head + '_images_history_set_index')).attr("img_index") + return s +} + diff --git a/javascript/jquery-3.6.0.min.js b/javascript/jquery-3.6.0.min.js new file mode 100644 index 00000000..c4c6022f --- /dev/null +++ b/javascript/jquery-3.6.0.min.js @@ -0,0 +1,2 @@ +/*! jQuery v3.6.0 | (c) OpenJS Foundation and other contributors | jquery.org/license */ +!function(e,t){"use strict";"object"==typeof module&&"object"==typeof module.exports?module.exports=e.document?t(e,!0):function(e){if(!e.document)throw new Error("jQuery requires a window with a document");return t(e)}:t(e)}("undefined"!=typeof window?window:this,function(C,e){"use strict";var t=[],r=Object.getPrototypeOf,s=t.slice,g=t.flat?function(e){return t.flat.call(e)}:function(e){return t.concat.apply([],e)},u=t.push,i=t.indexOf,n={},o=n.toString,v=n.hasOwnProperty,a=v.toString,l=a.call(Object),y={},m=function(e){return"function"==typeof e&&"number"!=typeof e.nodeType&&"function"!=typeof e.item},x=function(e){return null!=e&&e===e.window},E=C.document,c={type:!0,src:!0,nonce:!0,noModule:!0};function b(e,t,n){var r,i,o=(n=n||E).createElement("script");if(o.text=e,t)for(r in c)(i=t[r]||t.getAttribute&&t.getAttribute(r))&&o.setAttribute(r,i);n.head.appendChild(o).parentNode.removeChild(o)}function w(e){return null==e?e+"":"object"==typeof e||"function"==typeof e?n[o.call(e)]||"object":typeof e}var f="3.6.0",S=function(e,t){return new S.fn.init(e,t)};function p(e){var t=!!e&&"length"in e&&e.length,n=w(e);return!m(e)&&!x(e)&&("array"===n||0===t||"number"==typeof t&&0+~]|"+M+")"+M+"*"),U=new RegExp(M+"|>"),X=new RegExp(F),V=new RegExp("^"+I+"$"),G={ID:new RegExp("^#("+I+")"),CLASS:new RegExp("^\\.("+I+")"),TAG:new RegExp("^("+I+"|[*])"),ATTR:new RegExp("^"+W),PSEUDO:new RegExp("^"+F),CHILD:new RegExp("^:(only|first|last|nth|nth-last)-(child|of-type)(?:\\("+M+"*(even|odd|(([+-]|)(\\d*)n|)"+M+"*(?:([+-]|)"+M+"*(\\d+)|))"+M+"*\\)|)","i"),bool:new RegExp("^(?:"+R+")$","i"),needsContext:new RegExp("^"+M+"*[>+~]|:(even|odd|eq|gt|lt|nth|first|last)(?:\\("+M+"*((?:-\\d)?\\d*)"+M+"*\\)|)(?=[^-]|$)","i")},Y=/HTML$/i,Q=/^(?:input|select|textarea|button)$/i,J=/^h\d$/i,K=/^[^{]+\{\s*\[native \w/,Z=/^(?:#([\w-]+)|(\w+)|\.([\w-]+))$/,ee=/[+~]/,te=new RegExp("\\\\[\\da-fA-F]{1,6}"+M+"?|\\\\([^\\r\\n\\f])","g"),ne=function(e,t){var n="0x"+e.slice(1)-65536;return t||(n<0?String.fromCharCode(n+65536):String.fromCharCode(n>>10|55296,1023&n|56320))},re=/([\0-\x1f\x7f]|^-?\d)|^-$|[^\0-\x1f\x7f-\uFFFF\w-]/g,ie=function(e,t){return t?"\0"===e?"\ufffd":e.slice(0,-1)+"\\"+e.charCodeAt(e.length-1).toString(16)+" ":"\\"+e},oe=function(){T()},ae=be(function(e){return!0===e.disabled&&"fieldset"===e.nodeName.toLowerCase()},{dir:"parentNode",next:"legend"});try{H.apply(t=O.call(p.childNodes),p.childNodes),t[p.childNodes.length].nodeType}catch(e){H={apply:t.length?function(e,t){L.apply(e,O.call(t))}:function(e,t){var n=e.length,r=0;while(e[n++]=t[r++]);e.length=n-1}}}function se(t,e,n,r){var i,o,a,s,u,l,c,f=e&&e.ownerDocument,p=e?e.nodeType:9;if(n=n||[],"string"!=typeof t||!t||1!==p&&9!==p&&11!==p)return n;if(!r&&(T(e),e=e||C,E)){if(11!==p&&(u=Z.exec(t)))if(i=u[1]){if(9===p){if(!(a=e.getElementById(i)))return n;if(a.id===i)return n.push(a),n}else if(f&&(a=f.getElementById(i))&&y(e,a)&&a.id===i)return n.push(a),n}else{if(u[2])return H.apply(n,e.getElementsByTagName(t)),n;if((i=u[3])&&d.getElementsByClassName&&e.getElementsByClassName)return H.apply(n,e.getElementsByClassName(i)),n}if(d.qsa&&!N[t+" "]&&(!v||!v.test(t))&&(1!==p||"object"!==e.nodeName.toLowerCase())){if(c=t,f=e,1===p&&(U.test(t)||z.test(t))){(f=ee.test(t)&&ye(e.parentNode)||e)===e&&d.scope||((s=e.getAttribute("id"))?s=s.replace(re,ie):e.setAttribute("id",s=S)),o=(l=h(t)).length;while(o--)l[o]=(s?"#"+s:":scope")+" "+xe(l[o]);c=l.join(",")}try{return H.apply(n,f.querySelectorAll(c)),n}catch(e){N(t,!0)}finally{s===S&&e.removeAttribute("id")}}}return g(t.replace($,"$1"),e,n,r)}function ue(){var r=[];return function e(t,n){return r.push(t+" ")>b.cacheLength&&delete e[r.shift()],e[t+" "]=n}}function le(e){return e[S]=!0,e}function ce(e){var t=C.createElement("fieldset");try{return!!e(t)}catch(e){return!1}finally{t.parentNode&&t.parentNode.removeChild(t),t=null}}function fe(e,t){var n=e.split("|"),r=n.length;while(r--)b.attrHandle[n[r]]=t}function pe(e,t){var n=t&&e,r=n&&1===e.nodeType&&1===t.nodeType&&e.sourceIndex-t.sourceIndex;if(r)return r;if(n)while(n=n.nextSibling)if(n===t)return-1;return e?1:-1}function de(t){return function(e){return"input"===e.nodeName.toLowerCase()&&e.type===t}}function he(n){return function(e){var t=e.nodeName.toLowerCase();return("input"===t||"button"===t)&&e.type===n}}function ge(t){return function(e){return"form"in e?e.parentNode&&!1===e.disabled?"label"in e?"label"in e.parentNode?e.parentNode.disabled===t:e.disabled===t:e.isDisabled===t||e.isDisabled!==!t&&ae(e)===t:e.disabled===t:"label"in e&&e.disabled===t}}function ve(a){return le(function(o){return o=+o,le(function(e,t){var n,r=a([],e.length,o),i=r.length;while(i--)e[n=r[i]]&&(e[n]=!(t[n]=e[n]))})})}function ye(e){return e&&"undefined"!=typeof e.getElementsByTagName&&e}for(e in d=se.support={},i=se.isXML=function(e){var t=e&&e.namespaceURI,n=e&&(e.ownerDocument||e).documentElement;return!Y.test(t||n&&n.nodeName||"HTML")},T=se.setDocument=function(e){var t,n,r=e?e.ownerDocument||e:p;return r!=C&&9===r.nodeType&&r.documentElement&&(a=(C=r).documentElement,E=!i(C),p!=C&&(n=C.defaultView)&&n.top!==n&&(n.addEventListener?n.addEventListener("unload",oe,!1):n.attachEvent&&n.attachEvent("onunload",oe)),d.scope=ce(function(e){return a.appendChild(e).appendChild(C.createElement("div")),"undefined"!=typeof e.querySelectorAll&&!e.querySelectorAll(":scope fieldset div").length}),d.attributes=ce(function(e){return e.className="i",!e.getAttribute("className")}),d.getElementsByTagName=ce(function(e){return e.appendChild(C.createComment("")),!e.getElementsByTagName("*").length}),d.getElementsByClassName=K.test(C.getElementsByClassName),d.getById=ce(function(e){return a.appendChild(e).id=S,!C.getElementsByName||!C.getElementsByName(S).length}),d.getById?(b.filter.ID=function(e){var t=e.replace(te,ne);return function(e){return e.getAttribute("id")===t}},b.find.ID=function(e,t){if("undefined"!=typeof t.getElementById&&E){var n=t.getElementById(e);return n?[n]:[]}}):(b.filter.ID=function(e){var n=e.replace(te,ne);return function(e){var t="undefined"!=typeof e.getAttributeNode&&e.getAttributeNode("id");return t&&t.value===n}},b.find.ID=function(e,t){if("undefined"!=typeof t.getElementById&&E){var n,r,i,o=t.getElementById(e);if(o){if((n=o.getAttributeNode("id"))&&n.value===e)return[o];i=t.getElementsByName(e),r=0;while(o=i[r++])if((n=o.getAttributeNode("id"))&&n.value===e)return[o]}return[]}}),b.find.TAG=d.getElementsByTagName?function(e,t){return"undefined"!=typeof t.getElementsByTagName?t.getElementsByTagName(e):d.qsa?t.querySelectorAll(e):void 0}:function(e,t){var n,r=[],i=0,o=t.getElementsByTagName(e);if("*"===e){while(n=o[i++])1===n.nodeType&&r.push(n);return r}return o},b.find.CLASS=d.getElementsByClassName&&function(e,t){if("undefined"!=typeof t.getElementsByClassName&&E)return t.getElementsByClassName(e)},s=[],v=[],(d.qsa=K.test(C.querySelectorAll))&&(ce(function(e){var t;a.appendChild(e).innerHTML="",e.querySelectorAll("[msallowcapture^='']").length&&v.push("[*^$]="+M+"*(?:''|\"\")"),e.querySelectorAll("[selected]").length||v.push("\\["+M+"*(?:value|"+R+")"),e.querySelectorAll("[id~="+S+"-]").length||v.push("~="),(t=C.createElement("input")).setAttribute("name",""),e.appendChild(t),e.querySelectorAll("[name='']").length||v.push("\\["+M+"*name"+M+"*="+M+"*(?:''|\"\")"),e.querySelectorAll(":checked").length||v.push(":checked"),e.querySelectorAll("a#"+S+"+*").length||v.push(".#.+[+~]"),e.querySelectorAll("\\\f"),v.push("[\\r\\n\\f]")}),ce(function(e){e.innerHTML="";var t=C.createElement("input");t.setAttribute("type","hidden"),e.appendChild(t).setAttribute("name","D"),e.querySelectorAll("[name=d]").length&&v.push("name"+M+"*[*^$|!~]?="),2!==e.querySelectorAll(":enabled").length&&v.push(":enabled",":disabled"),a.appendChild(e).disabled=!0,2!==e.querySelectorAll(":disabled").length&&v.push(":enabled",":disabled"),e.querySelectorAll("*,:x"),v.push(",.*:")})),(d.matchesSelector=K.test(c=a.matches||a.webkitMatchesSelector||a.mozMatchesSelector||a.oMatchesSelector||a.msMatchesSelector))&&ce(function(e){d.disconnectedMatch=c.call(e,"*"),c.call(e,"[s!='']:x"),s.push("!=",F)}),v=v.length&&new RegExp(v.join("|")),s=s.length&&new RegExp(s.join("|")),t=K.test(a.compareDocumentPosition),y=t||K.test(a.contains)?function(e,t){var n=9===e.nodeType?e.documentElement:e,r=t&&t.parentNode;return e===r||!(!r||1!==r.nodeType||!(n.contains?n.contains(r):e.compareDocumentPosition&&16&e.compareDocumentPosition(r)))}:function(e,t){if(t)while(t=t.parentNode)if(t===e)return!0;return!1},j=t?function(e,t){if(e===t)return l=!0,0;var n=!e.compareDocumentPosition-!t.compareDocumentPosition;return n||(1&(n=(e.ownerDocument||e)==(t.ownerDocument||t)?e.compareDocumentPosition(t):1)||!d.sortDetached&&t.compareDocumentPosition(e)===n?e==C||e.ownerDocument==p&&y(p,e)?-1:t==C||t.ownerDocument==p&&y(p,t)?1:u?P(u,e)-P(u,t):0:4&n?-1:1)}:function(e,t){if(e===t)return l=!0,0;var n,r=0,i=e.parentNode,o=t.parentNode,a=[e],s=[t];if(!i||!o)return e==C?-1:t==C?1:i?-1:o?1:u?P(u,e)-P(u,t):0;if(i===o)return pe(e,t);n=e;while(n=n.parentNode)a.unshift(n);n=t;while(n=n.parentNode)s.unshift(n);while(a[r]===s[r])r++;return r?pe(a[r],s[r]):a[r]==p?-1:s[r]==p?1:0}),C},se.matches=function(e,t){return se(e,null,null,t)},se.matchesSelector=function(e,t){if(T(e),d.matchesSelector&&E&&!N[t+" "]&&(!s||!s.test(t))&&(!v||!v.test(t)))try{var n=c.call(e,t);if(n||d.disconnectedMatch||e.document&&11!==e.document.nodeType)return n}catch(e){N(t,!0)}return 0":{dir:"parentNode",first:!0}," ":{dir:"parentNode"},"+":{dir:"previousSibling",first:!0},"~":{dir:"previousSibling"}},preFilter:{ATTR:function(e){return e[1]=e[1].replace(te,ne),e[3]=(e[3]||e[4]||e[5]||"").replace(te,ne),"~="===e[2]&&(e[3]=" "+e[3]+" "),e.slice(0,4)},CHILD:function(e){return e[1]=e[1].toLowerCase(),"nth"===e[1].slice(0,3)?(e[3]||se.error(e[0]),e[4]=+(e[4]?e[5]+(e[6]||1):2*("even"===e[3]||"odd"===e[3])),e[5]=+(e[7]+e[8]||"odd"===e[3])):e[3]&&se.error(e[0]),e},PSEUDO:function(e){var t,n=!e[6]&&e[2];return G.CHILD.test(e[0])?null:(e[3]?e[2]=e[4]||e[5]||"":n&&X.test(n)&&(t=h(n,!0))&&(t=n.indexOf(")",n.length-t)-n.length)&&(e[0]=e[0].slice(0,t),e[2]=n.slice(0,t)),e.slice(0,3))}},filter:{TAG:function(e){var t=e.replace(te,ne).toLowerCase();return"*"===e?function(){return!0}:function(e){return e.nodeName&&e.nodeName.toLowerCase()===t}},CLASS:function(e){var t=m[e+" "];return t||(t=new RegExp("(^|"+M+")"+e+"("+M+"|$)"))&&m(e,function(e){return t.test("string"==typeof e.className&&e.className||"undefined"!=typeof e.getAttribute&&e.getAttribute("class")||"")})},ATTR:function(n,r,i){return function(e){var t=se.attr(e,n);return null==t?"!="===r:!r||(t+="","="===r?t===i:"!="===r?t!==i:"^="===r?i&&0===t.indexOf(i):"*="===r?i&&-1:\x20\t\r\n\f]*)[\x20\t\r\n\f]*\/?>(?:<\/\1>|)$/i;function j(e,n,r){return m(n)?S.grep(e,function(e,t){return!!n.call(e,t,e)!==r}):n.nodeType?S.grep(e,function(e){return e===n!==r}):"string"!=typeof n?S.grep(e,function(e){return-1)[^>]*|#([\w-]+))$/;(S.fn.init=function(e,t,n){var r,i;if(!e)return this;if(n=n||D,"string"==typeof e){if(!(r="<"===e[0]&&">"===e[e.length-1]&&3<=e.length?[null,e,null]:q.exec(e))||!r[1]&&t)return!t||t.jquery?(t||n).find(e):this.constructor(t).find(e);if(r[1]){if(t=t instanceof S?t[0]:t,S.merge(this,S.parseHTML(r[1],t&&t.nodeType?t.ownerDocument||t:E,!0)),N.test(r[1])&&S.isPlainObject(t))for(r in t)m(this[r])?this[r](t[r]):this.attr(r,t[r]);return this}return(i=E.getElementById(r[2]))&&(this[0]=i,this.length=1),this}return e.nodeType?(this[0]=e,this.length=1,this):m(e)?void 0!==n.ready?n.ready(e):e(S):S.makeArray(e,this)}).prototype=S.fn,D=S(E);var L=/^(?:parents|prev(?:Until|All))/,H={children:!0,contents:!0,next:!0,prev:!0};function O(e,t){while((e=e[t])&&1!==e.nodeType);return e}S.fn.extend({has:function(e){var t=S(e,this),n=t.length;return this.filter(function(){for(var e=0;e\x20\t\r\n\f]*)/i,he=/^$|^module$|\/(?:java|ecma)script/i;ce=E.createDocumentFragment().appendChild(E.createElement("div")),(fe=E.createElement("input")).setAttribute("type","radio"),fe.setAttribute("checked","checked"),fe.setAttribute("name","t"),ce.appendChild(fe),y.checkClone=ce.cloneNode(!0).cloneNode(!0).lastChild.checked,ce.innerHTML="",y.noCloneChecked=!!ce.cloneNode(!0).lastChild.defaultValue,ce.innerHTML="",y.option=!!ce.lastChild;var ge={thead:[1,"","
"],col:[2,"","
"],tr:[2,"","
"],td:[3,"","
"],_default:[0,"",""]};function ve(e,t){var n;return n="undefined"!=typeof e.getElementsByTagName?e.getElementsByTagName(t||"*"):"undefined"!=typeof e.querySelectorAll?e.querySelectorAll(t||"*"):[],void 0===t||t&&A(e,t)?S.merge([e],n):n}function ye(e,t){for(var n=0,r=e.length;n",""]);var me=/<|&#?\w+;/;function xe(e,t,n,r,i){for(var o,a,s,u,l,c,f=t.createDocumentFragment(),p=[],d=0,h=e.length;d\s*$/g;function je(e,t){return A(e,"table")&&A(11!==t.nodeType?t:t.firstChild,"tr")&&S(e).children("tbody")[0]||e}function De(e){return e.type=(null!==e.getAttribute("type"))+"/"+e.type,e}function qe(e){return"true/"===(e.type||"").slice(0,5)?e.type=e.type.slice(5):e.removeAttribute("type"),e}function Le(e,t){var n,r,i,o,a,s;if(1===t.nodeType){if(Y.hasData(e)&&(s=Y.get(e).events))for(i in Y.remove(t,"handle events"),s)for(n=0,r=s[i].length;n").attr(n.scriptAttrs||{}).prop({charset:n.scriptCharset,src:n.url}).on("load error",i=function(e){r.remove(),i=null,e&&t("error"===e.type?404:200,e.type)}),E.head.appendChild(r[0])},abort:function(){i&&i()}}});var _t,zt=[],Ut=/(=)\?(?=&|$)|\?\?/;S.ajaxSetup({jsonp:"callback",jsonpCallback:function(){var e=zt.pop()||S.expando+"_"+wt.guid++;return this[e]=!0,e}}),S.ajaxPrefilter("json jsonp",function(e,t,n){var r,i,o,a=!1!==e.jsonp&&(Ut.test(e.url)?"url":"string"==typeof e.data&&0===(e.contentType||"").indexOf("application/x-www-form-urlencoded")&&Ut.test(e.data)&&"data");if(a||"jsonp"===e.dataTypes[0])return r=e.jsonpCallback=m(e.jsonpCallback)?e.jsonpCallback():e.jsonpCallback,a?e[a]=e[a].replace(Ut,"$1"+r):!1!==e.jsonp&&(e.url+=(Tt.test(e.url)?"&":"?")+e.jsonp+"="+r),e.converters["script json"]=function(){return o||S.error(r+" was not called"),o[0]},e.dataTypes[0]="json",i=C[r],C[r]=function(){o=arguments},n.always(function(){void 0===i?S(C).removeProp(r):C[r]=i,e[r]&&(e.jsonpCallback=t.jsonpCallback,zt.push(r)),o&&m(i)&&i(o[0]),o=i=void 0}),"script"}),y.createHTMLDocument=((_t=E.implementation.createHTMLDocument("").body).innerHTML="
",2===_t.childNodes.length),S.parseHTML=function(e,t,n){return"string"!=typeof e?[]:("boolean"==typeof t&&(n=t,t=!1),t||(y.createHTMLDocument?((r=(t=E.implementation.createHTMLDocument("")).createElement("base")).href=E.location.href,t.head.appendChild(r)):t=E),o=!n&&[],(i=N.exec(e))?[t.createElement(i[1])]:(i=xe([e],t,o),o&&o.length&&S(o).remove(),S.merge([],i.childNodes)));var r,i,o},S.fn.load=function(e,t,n){var r,i,o,a=this,s=e.indexOf(" ");return-1").append(S.parseHTML(e)).find(r):e)}).always(n&&function(e,t){a.each(function(){n.apply(this,o||[e.responseText,t,e])})}),this},S.expr.pseudos.animated=function(t){return S.grep(S.timers,function(e){return t===e.elem}).length},S.offset={setOffset:function(e,t,n){var r,i,o,a,s,u,l=S.css(e,"position"),c=S(e),f={};"static"===l&&(e.style.position="relative"),s=c.offset(),o=S.css(e,"top"),u=S.css(e,"left"),("absolute"===l||"fixed"===l)&&-1<(o+u).indexOf("auto")?(a=(r=c.position()).top,i=r.left):(a=parseFloat(o)||0,i=parseFloat(u)||0),m(t)&&(t=t.call(e,n,S.extend({},s))),null!=t.top&&(f.top=t.top-s.top+a),null!=t.left&&(f.left=t.left-s.left+i),"using"in t?t.using.call(e,f):c.css(f)}},S.fn.extend({offset:function(t){if(arguments.length)return void 0===t?this:this.each(function(e){S.offset.setOffset(this,t,e)});var e,n,r=this[0];return r?r.getClientRects().length?(e=r.getBoundingClientRect(),n=r.ownerDocument.defaultView,{top:e.top+n.pageYOffset,left:e.left+n.pageXOffset}):{top:0,left:0}:void 0},position:function(){if(this[0]){var e,t,n,r=this[0],i={top:0,left:0};if("fixed"===S.css(r,"position"))t=r.getBoundingClientRect();else{t=this.offset(),n=r.ownerDocument,e=r.offsetParent||n.documentElement;while(e&&(e===n.body||e===n.documentElement)&&"static"===S.css(e,"position"))e=e.parentNode;e&&e!==r&&1===e.nodeType&&((i=S(e).offset()).top+=S.css(e,"borderTopWidth",!0),i.left+=S.css(e,"borderLeftWidth",!0))}return{top:t.top-i.top-S.css(r,"marginTop",!0),left:t.left-i.left-S.css(r,"marginLeft",!0)}}},offsetParent:function(){return this.map(function(){var e=this.offsetParent;while(e&&"static"===S.css(e,"position"))e=e.offsetParent;return e||re})}}),S.each({scrollLeft:"pageXOffset",scrollTop:"pageYOffset"},function(t,i){var o="pageYOffset"===i;S.fn[t]=function(e){return $(this,function(e,t,n){var r;if(x(e)?r=e:9===e.nodeType&&(r=e.defaultView),void 0===n)return r?r[i]:e[t];r?r.scrollTo(o?r.pageXOffset:n,o?n:r.pageYOffset):e[t]=n},t,e,arguments.length)}}),S.each(["top","left"],function(e,n){S.cssHooks[n]=Fe(y.pixelPosition,function(e,t){if(t)return t=We(e,n),Pe.test(t)?S(e).position()[n]+"px":t})}),S.each({Height:"height",Width:"width"},function(a,s){S.each({padding:"inner"+a,content:s,"":"outer"+a},function(r,o){S.fn[o]=function(e,t){var n=arguments.length&&(r||"boolean"!=typeof e),i=r||(!0===e||!0===t?"margin":"border");return $(this,function(e,t,n){var r;return x(e)?0===o.indexOf("outer")?e["inner"+a]:e.document.documentElement["client"+a]:9===e.nodeType?(r=e.documentElement,Math.max(e.body["scroll"+a],r["scroll"+a],e.body["offset"+a],r["offset"+a],r["client"+a])):void 0===n?S.css(e,t,i):S.style(e,t,n,i)},s,n?e:void 0,n)}})}),S.each(["ajaxStart","ajaxStop","ajaxComplete","ajaxError","ajaxSuccess","ajaxSend"],function(e,t){S.fn[t]=function(e){return this.on(t,e)}}),S.fn.extend({bind:function(e,t,n){return this.on(e,null,t,n)},unbind:function(e,t){return this.off(e,null,t)},delegate:function(e,t,n,r){return this.on(t,e,n,r)},undelegate:function(e,t,n){return 1===arguments.length?this.off(e,"**"):this.off(t,e||"**",n)},hover:function(e,t){return this.mouseenter(e).mouseleave(t||e)}}),S.each("blur focus focusin focusout resize scroll click dblclick mousedown mouseup mousemove mouseover mouseout mouseenter mouseleave change select submit keydown keypress keyup contextmenu".split(" "),function(e,n){S.fn[n]=function(e,t){return 0 max_page_index else page_index + idx_frm = (page_index - 1) * num + file_list = file_list[idx_frm:idx_frm + num] + print(f"Loading history page {page_index}") + return [os.path.join(dir_name, file) for file in file_list], page_index, file_list +def first_page_click(is_img2img, dir_name): + return get_recent_images(is_img2img, dir_name, 1, 0) +def end_page_click(is_img2img, dir_name): + return get_recent_images(is_img2img, dir_name, -1, 0) +def prev_page_click(is_img2img, dir_name, page_index): + return get_recent_images(is_img2img, dir_name, page_index, -1) +def next_page_click(is_img2img, dir_name, page_index): + return get_recent_images(is_img2img, dir_name, page_index, 1) +def page_index_change(is_img2img, dir_name, page_index): + return get_recent_images(is_img2img, dir_name, page_index, 0) +def show_image_info(num, filenames): + return filenames[int(num)] +def delete_image(is_img2img, dir_name, name, page_index, filenames): + path = os.path.join(dir_name, name) + if os.path.exists(path): + print(f"Delete file {path}") + os.remove(path) + i = 0 + for f in filenames: + if f == name: + break + i += 1 + images, page_index, file_list = get_recent_images(is_img2img, dir_name, page_index, 0) + current_file = file_list[i] if i < len(file_list) else None + return images, page_index, file_list, current_file + + +def show_images_history(gr, opts, is_img2img): + def id_name(is_img2img, name): + return ("img2img" if is_img2img else "txt2img") + "_" + name + with gr.Row(): + if is_img2img: + dir_name = opts.outdir_img2img_samples + else: + dir_name = opts.outdir_txt2img_samples + first_page = gr.Button('First Page', elem_id=id_name(is_img2img,"images_history_first_page")) + prev_page = gr.Button('Prev Page') + page_index = gr.Number(value=1) + next_page = gr.Button('Next Page') + end_page = gr.Button('End Page') + with gr.Row(): + delete = gr.Button('Delete') + Send = gr.Button('Send') + with gr.Row(): + with gr.Column(elem_id=id_name(is_img2img,"images_history")): + history_gallery = gr.Gallery(label="Images history").style(grid=6) + img_file_name = gr.Textbox() + img_file_info = gr.Textbox(dir_name) + img_path = gr.Textbox(dir_name, visible=False) + set_index = gr.Button('set_index', elem_id=id_name(is_img2img,"images_history_set_index")) + is_img2img_flag = gr.Checkbox(is_img2img, visible=False) + filenames = gr.State() + first_page.click(first_page_click, inputs=[is_img2img_flag, img_path], outputs=[history_gallery, page_index, filenames]) + next_page.click(next_page_click, inputs=[is_img2img_flag, img_path, page_index], outputs=[history_gallery, page_index, filenames]) + prev_page.click(prev_page_click, inputs=[is_img2img_flag, img_path, page_index], outputs=[history_gallery, page_index, filenames]) + end_page.click(end_page_click, inputs=[is_img2img_flag, img_path], outputs=[history_gallery, page_index, filenames]) + page_index.submit(page_index_change, inputs=[is_img2img_flag, img_path, page_index], outputs=[history_gallery, page_index, filenames]) + set_index.click(show_image_info, _js="images_history_get_current_img",inputs=[is_img2img_flag, filenames], outputs=img_file_name) + delete.click(delete_image, inputs=[is_img2img_flag, img_path, img_file_name, page_index, filenames], outputs=[history_gallery, page_index, filenames,img_file_name]) + #page_index.change(page_index_change, inputs=[is_img2img_flag, img_path, page_index], outputs=[history_gallery, page_index]) + +def create_history_tabs(gr, opts): + with gr.Blocks(analytics_enabled=False) as images_history: + with gr.Tabs() as tabs: + with gr.Tab("txt2img history", id="images_history_txt2img"): + with gr.Blocks(analytics_enabled=False) as images_history_txt2img: + show_images_history(gr, opts, is_img2img=False) + with gr.Tab("img2img history", id="images_history_img2img"): + with gr.Blocks(analytics_enabled=False) as images_history_img2img: + show_images_history(gr, opts, is_img2img=True) + return images_history diff --git a/modules/ui.py b/modules/ui.py index 4f18126f..8762fcf5 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -37,6 +37,7 @@ import modules.generation_parameters_copypaste from modules import prompt_parser from modules.images import save_image import modules.textual_inversion.ui +import modules.images_history as img_his # this is a fix for Windows users. Without it, javascript files will be served with text/html content-type and the bowser will not show any UI mimetypes.init() @@ -499,7 +500,6 @@ def create_ui(wrap_gradio_gpu_call): custom_inputs = modules.scripts.scripts_txt2img.setup_ui(is_img2img=False) with gr.Column(variant='panel'): - with gr.Group(): txt2img_preview = gr.Image(elem_id='txt2img_preview', visible=False) txt2img_gallery = gr.Gallery(label='Output', show_label=False, elem_id='txt2img_gallery').style(grid=4) @@ -516,6 +516,7 @@ def create_ui(wrap_gradio_gpu_call): with gr.Group(): html_info = gr.HTML() generation_info = gr.Textbox(visible=False) + connect_reuse_seed(seed, reuse_seed, generation_info, dummy_component, is_subseed=False) connect_reuse_seed(subseed, reuse_subseed, generation_info, dummy_component, is_subseed=True) @@ -607,6 +608,7 @@ def create_ui(wrap_gradio_gpu_call): ] modules.generation_parameters_copypaste.connect_paste(paste, txt2img_paste_fields, txt2img_prompt) token_button.click(fn=update_token_counter, inputs=[txt2img_prompt, steps], outputs=[token_counter]) + with gr.Blocks(analytics_enabled=False) as img2img_interface: img2img_prompt, roll, img2img_prompt_style, img2img_negative_prompt, img2img_prompt_style2, submit, img2img_interrogate, img2img_prompt_style_apply, img2img_save_style, paste, token_counter, token_button = create_toprow(is_img2img=True) @@ -696,6 +698,7 @@ def create_ui(wrap_gradio_gpu_call): with gr.Group(): html_info = gr.HTML() generation_info = gr.Textbox(visible=False) + connect_reuse_seed(seed, reuse_seed, generation_info, dummy_component, is_subseed=False) connect_reuse_seed(subseed, reuse_subseed, generation_info, dummy_component, is_subseed=True) @@ -1126,8 +1129,10 @@ def create_ui(wrap_gradio_gpu_call): opts.save(shared.config_filename) - return f'{changed} settings changed.', opts.dumpjson() + return f'{changed} settings changed.', opts.dumpjson() + + images_history = img_his.create_history_tabs(gr, opts) with gr.Blocks(analytics_enabled=False) as settings_interface: settings_submit = gr.Button(value="Apply settings", variant='primary') result = gr.HTML() @@ -1206,7 +1211,9 @@ def create_ui(wrap_gradio_gpu_call): (pnginfo_interface, "PNG Info", "pnginfo"), (modelmerger_interface, "Checkpoint Merger", "modelmerger"), (textual_inversion_interface, "Textual inversion", "ti"), + (images_history, "History", "images_history"), (settings_interface, "Settings", "settings"), + ] with open(os.path.join(script_path, "style.css"), "r", encoding="utf8") as file: diff --git a/repositorieslatent-diffusion b/repositorieslatent-diffusion new file mode 160000 index 00000000..abf33e70 --- /dev/null +++ b/repositorieslatent-diffusion @@ -0,0 +1 @@ +Subproject commit abf33e7002d59d9085081bce93ec798dcabd49af diff --git a/testui.py b/testui.py new file mode 100644 index 00000000..f54e4a62 --- /dev/null +++ b/testui.py @@ -0,0 +1,124 @@ +import os +import threading +import time +import importlib +import signal +import threading + +from modules.paths import script_path + +from modules import devices, sd_samplers +import modules.codeformer_model as codeformer +import modules.extras +import modules.face_restoration +import modules.gfpgan_model as gfpgan +import modules.img2img + +import modules.lowvram +import modules.paths +import modules.scripts +import modules.sd_hijack +import modules.sd_models +import modules.shared as shared +import modules.txt2img + +import modules.ui +from modules import devices +from modules import modelloader +from modules.paths import script_path +from modules.shared import cmd_opts + +modelloader.cleanup_models() +modules.sd_models.setup_model() +codeformer.setup_model(cmd_opts.codeformer_models_path) +gfpgan.setup_model(cmd_opts.gfpgan_models_path) +shared.face_restorers.append(modules.face_restoration.FaceRestoration()) +modelloader.load_upscalers() +queue_lock = threading.Lock() + + +def wrap_queued_call(func): + def f(*args, **kwargs): + with queue_lock: + res = func(*args, **kwargs) + + return res + + return f + + +def wrap_gradio_gpu_call(func, extra_outputs=None): + def f(*args, **kwargs): + devices.torch_gc() + + shared.state.sampling_step = 0 + shared.state.job_count = -1 + shared.state.job_no = 0 + shared.state.job_timestamp = shared.state.get_job_timestamp() + shared.state.current_latent = None + shared.state.current_image = None + shared.state.current_image_sampling_step = 0 + shared.state.interrupted = False + shared.state.textinfo = None + + with queue_lock: + res = func(*args, **kwargs) + + shared.state.job = "" + shared.state.job_count = 0 + + devices.torch_gc() + + return res + + return modules.ui.wrap_gradio_call(f, extra_outputs=extra_outputs) + + +modules.scripts.load_scripts(os.path.join(script_path, "scripts")) + +shared.sd_model = None #modules.sd_models.load_model() +#shared.opts.onchange("sd_model_checkpoint", wrap_queued_call(lambda: modules.sd_models.reload_model_weights(shared.sd_model))) + + +def webui(): + # make the program just exit at ctrl+c without waiting for anything + def sigint_handler(sig, frame): + print(f'Interrupted with signal {sig} in {frame}') + os._exit(0) + + signal.signal(signal.SIGINT, sigint_handler) + + while 1: + + demo = modules.ui.create_ui(wrap_gradio_gpu_call=wrap_gradio_gpu_call) + + demo.launch( + share=cmd_opts.share, + server_name="0.0.0.0" if cmd_opts.listen else None, + server_port=cmd_opts.port, + debug=cmd_opts.gradio_debug, + auth=[tuple(cred.split(':')) for cred in cmd_opts.gradio_auth.strip('"').split(',')] if cmd_opts.gradio_auth else None, + inbrowser=cmd_opts.autolaunch, + prevent_thread_lock=True + ) + + while 1: + time.sleep(0.5) + if getattr(demo, 'do_restart', False): + time.sleep(0.5) + demo.close() + time.sleep(0.5) + break + + sd_samplers.set_samplers() + + print('Reloading Custom Scripts') + modules.scripts.reload_scripts(os.path.join(script_path, "scripts")) + print('Reloading modules: modules.ui') + importlib.reload(modules.ui) + print('Restarting Gradio') + + + +if __name__ == "__main__": + webui() -- cgit v1.2.3 From 8ec069e64df48f8f202f8b93a08e91b69448eb39 Mon Sep 17 00:00:00 2001 From: JC_Array Date: Mon, 10 Oct 2022 03:23:24 -0500 Subject: removed duplicate run_preprocess.click by creating run_preprocess_inputs list and appending deepbooru variable to input list if in scope --- modules/ui.py | 49 +++++++++++++++++-------------------------------- 1 file changed, 17 insertions(+), 32 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/ui.py b/modules/ui.py index 179e3a83..22ca74c2 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1089,40 +1089,25 @@ def create_ui(wrap_gradio_gpu_call): ] ) + run_preprocess_inputs = [ + process_src, + process_dst, + process_flip, + process_split, + process_caption, + ] if cmd_opts.deepdanbooru: # if process_caption_deepbooru is None, it will cause an error, as a result only include it if it is enabled - run_preprocess.click( - fn=wrap_gradio_gpu_call(modules.textual_inversion.ui.preprocess, extra_outputs=[gr.update()]), - _js="start_training_textual_inversion", - inputs=[ - process_src, - process_dst, - process_flip, - process_split, - process_caption, - process_caption_deepbooru, - ], - outputs=[ - ti_output, - ti_outcome, - ], - ) - else: - run_preprocess.click( - fn=wrap_gradio_gpu_call(modules.textual_inversion.ui.preprocess, extra_outputs=[gr.update()]), - _js="start_training_textual_inversion", - inputs=[ - process_src, - process_dst, - process_flip, - process_split, - process_caption, - ], - outputs=[ - ti_output, - ti_outcome, - ], - ) + run_preprocess_inputs.append(process_caption_deepbooru) + run_preprocess.click( + fn=wrap_gradio_gpu_call(modules.textual_inversion.ui.preprocess, extra_outputs=[gr.update()]), + _js="start_training_textual_inversion", + inputs=run_preprocess_inputs, + outputs=[ + ti_output, + ti_outcome, + ], + ) train_embedding.click( fn=wrap_gradio_gpu_call(modules.textual_inversion.ui.train_embedding, extra_outputs=[gr.update()]), -- cgit v1.2.3 From 2f94331df2cb1181439adecc28cfd758049f6501 Mon Sep 17 00:00:00 2001 From: JC_Array Date: Mon, 10 Oct 2022 03:34:00 -0500 Subject: removed change in last commit, simplified to adding the visible argument to process_caption_deepbooru and it set to False if deepdanbooru argument is not set --- modules/ui.py | 22 ++++++++++------------ 1 file changed, 10 insertions(+), 12 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/ui.py b/modules/ui.py index 22ca74c2..f8adafb3 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1036,7 +1036,8 @@ def create_ui(wrap_gradio_gpu_call): process_caption = gr.Checkbox(label='Use BLIP caption as filename') if cmd_opts.deepdanbooru: process_caption_deepbooru = gr.Checkbox(label='Use deepbooru caption as filename') - + else: + process_caption_deepbooru = gr.Checkbox(label='Use deepbooru caption as filename', visible=False) with gr.Row(): with gr.Column(scale=3): @@ -1089,20 +1090,17 @@ def create_ui(wrap_gradio_gpu_call): ] ) - run_preprocess_inputs = [ - process_src, - process_dst, - process_flip, - process_split, - process_caption, - ] - if cmd_opts.deepdanbooru: - # if process_caption_deepbooru is None, it will cause an error, as a result only include it if it is enabled - run_preprocess_inputs.append(process_caption_deepbooru) run_preprocess.click( fn=wrap_gradio_gpu_call(modules.textual_inversion.ui.preprocess, extra_outputs=[gr.update()]), _js="start_training_textual_inversion", - inputs=run_preprocess_inputs, + inputs=[ + process_src, + process_dst, + process_flip, + process_split, + process_caption, + process_caption_deepbooru + ], outputs=[ ti_output, ti_outcome, -- cgit v1.2.3 From 23f2989799ee3911d2959cfceb74b921f20c9a51 Mon Sep 17 00:00:00 2001 From: yfszzx Date: Mon, 10 Oct 2022 18:33:49 +0800 Subject: images history over --- javascript/images_history.js | 4 +- modules/images_history.py | 141 ++++++++++++++++++++++++++----------------- modules/ui.py | 9 ++- testui.py | 124 ------------------------------------- 4 files changed, 94 insertions(+), 184 deletions(-) delete mode 100644 testui.py (limited to 'modules/ui.py') diff --git a/javascript/images_history.js b/javascript/images_history.js index f30b7eff..93d2b89a 100644 --- a/javascript/images_history.js +++ b/javascript/images_history.js @@ -58,9 +58,9 @@ function set_history_index_from_img(e){ } } } -function images_history_get_current_img(is_image2image){ +function images_history_get_current_img(is_image2image, image_path, files){ head = is_image2image?"img2img":"txt2img" s = $(gradioApp().getElementById(head + '_images_history_set_index')).attr("img_index") - return s + return [s, image_path, files] } diff --git a/modules/images_history.py b/modules/images_history.py index 23d83557..0e0a48f3 100644 --- a/modules/images_history.py +++ b/modules/images_history.py @@ -1,5 +1,6 @@ import os -def get_recent_images(is_img2img, dir_name, page_index, step): +def get_recent_images(dir_name, page_index, step, image_index): + print(image_index) page_index = int(page_index) f_list = os.listdir(dir_name) file_list = [] @@ -8,7 +9,7 @@ def get_recent_images(is_img2img, dir_name, page_index, step): continue file_list.append(file) file_list = sorted(file_list, key=lambda file: -os.path.getctime(os.path.join(dir_name, file))) - num = 24 + num = 48 max_page_index = len(file_list) // num + 1 page_index = max_page_index if page_index == -1 else page_index + step page_index = 1 if page_index < 1 else page_index @@ -16,75 +17,101 @@ def get_recent_images(is_img2img, dir_name, page_index, step): idx_frm = (page_index - 1) * num file_list = file_list[idx_frm:idx_frm + num] print(f"Loading history page {page_index}") - return [os.path.join(dir_name, file) for file in file_list], page_index, file_list -def first_page_click(is_img2img, dir_name): - return get_recent_images(is_img2img, dir_name, 1, 0) -def end_page_click(is_img2img, dir_name): - return get_recent_images(is_img2img, dir_name, -1, 0) -def prev_page_click(is_img2img, dir_name, page_index): - return get_recent_images(is_img2img, dir_name, page_index, -1) -def next_page_click(is_img2img, dir_name, page_index): - return get_recent_images(is_img2img, dir_name, page_index, 1) -def page_index_change(is_img2img, dir_name, page_index): - return get_recent_images(is_img2img, dir_name, page_index, 0) -def show_image_info(num, filenames): - return filenames[int(num)] -def delete_image(is_img2img, dir_name, name, page_index, filenames): + image_index = int(image_index) + if image_index < 0 or image_index > len(file_list) - 1: + current_file = None + hide_image = None + else: + current_file = file_list[int(image_index)] + hide_image = os.path.join(dir_name, current_file) + return [os.path.join(dir_name, file) for file in file_list], page_index, file_list, current_file, hide_image +def first_page_click(dir_name, page_index, image_index): + return get_recent_images(dir_name, 1, 0, image_index) +def end_page_click(dir_name, page_index, image_index): + return get_recent_images(dir_name, -1, 0, image_index) +def prev_page_click(dir_name, page_index, image_index): + return get_recent_images(dir_name, page_index, -1, image_index) +def next_page_click(dir_name, page_index, image_index): + return get_recent_images(dir_name, page_index, 1, image_index) +def page_index_change(dir_name, page_index, image_index): + return get_recent_images(dir_name, page_index, 0, image_index) + +def show_image_info(num, image_path, filenames): + file = filenames[int(num)] + return file, num, os.path.join(image_path, file) +def delete_image(is_img2img, dir_name, name, page_index, filenames, image_index): + print("filename", name) path = os.path.join(dir_name, name) if os.path.exists(path): print(f"Delete file {path}") os.remove(path) - i = 0 - for f in filenames: - if f == name: - break - i += 1 - images, page_index, file_list = get_recent_images(is_img2img, dir_name, page_index, 0) - current_file = file_list[i] if i < len(file_list) else None - return images, page_index, file_list, current_file + images, page_index, file_list, current_file, hide_image = get_recent_images(dir_name, page_index, 0, image_index) + return images, page_index, file_list, current_file, hide_image -def show_images_history(gr, opts, is_img2img): +def show_images_history(gr, opts, is_img2img, run_pnginfo, switch_dict): def id_name(is_img2img, name): return ("img2img" if is_img2img else "txt2img") + "_" + name - with gr.Row(): - if is_img2img: - dir_name = opts.outdir_img2img_samples - else: - dir_name = opts.outdir_txt2img_samples - first_page = gr.Button('First Page', elem_id=id_name(is_img2img,"images_history_first_page")) - prev_page = gr.Button('Prev Page') - page_index = gr.Number(value=1) - next_page = gr.Button('Next Page') - end_page = gr.Button('End Page') - with gr.Row(): - delete = gr.Button('Delete') - Send = gr.Button('Send') - with gr.Row(): - with gr.Column(elem_id=id_name(is_img2img,"images_history")): - history_gallery = gr.Gallery(label="Images history").style(grid=6) - img_file_name = gr.Textbox() - img_file_info = gr.Textbox(dir_name) - img_path = gr.Textbox(dir_name, visible=False) - set_index = gr.Button('set_index', elem_id=id_name(is_img2img,"images_history_set_index")) - is_img2img_flag = gr.Checkbox(is_img2img, visible=False) - filenames = gr.State() - first_page.click(first_page_click, inputs=[is_img2img_flag, img_path], outputs=[history_gallery, page_index, filenames]) - next_page.click(next_page_click, inputs=[is_img2img_flag, img_path, page_index], outputs=[history_gallery, page_index, filenames]) - prev_page.click(prev_page_click, inputs=[is_img2img_flag, img_path, page_index], outputs=[history_gallery, page_index, filenames]) - end_page.click(end_page_click, inputs=[is_img2img_flag, img_path], outputs=[history_gallery, page_index, filenames]) - page_index.submit(page_index_change, inputs=[is_img2img_flag, img_path, page_index], outputs=[history_gallery, page_index, filenames]) - set_index.click(show_image_info, _js="images_history_get_current_img",inputs=[is_img2img_flag, filenames], outputs=img_file_name) - delete.click(delete_image, inputs=[is_img2img_flag, img_path, img_file_name, page_index, filenames], outputs=[history_gallery, page_index, filenames,img_file_name]) + if is_img2img: + dir_name = opts.outdir_img2img_samples + else: + dir_name = opts.outdir_txt2img_samples + with gr.Row(): + first_page = gr.Button('First', elem_id=id_name(is_img2img,"images_history_first_page")) + prev_page = gr.Button('Prev') + page_index = gr.Number(value=1, label="Page Index") + next_page = gr.Button('Next') + end_page = gr.Button('End') + with gr.Row(elem_id=id_name(is_img2img,"images_history")): + with gr.Row(): + with gr.Column(): + history_gallery = gr.Gallery(show_label=False).style(grid=6) + with gr.Column(): + with gr.Row(): + delete = gr.Button('Delete') + pnginfo_send_to_txt2img = gr.Button('Send to txt2img') + pnginfo_send_to_img2img = gr.Button('Send to img2img') + with gr.Row(): + with gr.Column(): + img_file_info = gr.Textbox(dir_name, label="Generate Info") + img_file_name = gr.Textbox(label="File Name") + with gr.Row(): + # hiden items + img_path = gr.Textbox(dir_name, visible=False) + is_img2img_flag = gr.Checkbox(is_img2img, visible=False) + image_index = gr.Textbox(value=-1, visible=False) + set_index = gr.Button('set_index', elem_id=id_name(is_img2img,"images_history_set_index")) + filenames = gr.State() + hide_image = gr.Image(visible=False, type="pil") + info1 = gr.Textbox(visible=False) + info2 = gr.Textbox(visible=False) + + + # turn pages + gallery_inputs = [img_path, page_index, image_index] + gallery_outputs = [history_gallery, page_index, filenames, img_file_name, hide_image] + first_page.click(first_page_click, inputs=gallery_inputs, outputs=gallery_outputs) + next_page.click(next_page_click, inputs=gallery_inputs, outputs=gallery_outputs) + prev_page.click(prev_page_click, inputs=gallery_inputs, outputs=gallery_outputs) + end_page.click(end_page_click, inputs=gallery_inputs, outputs=gallery_outputs) + page_index.submit(page_index_change, inputs=gallery_inputs, outputs=gallery_outputs) #page_index.change(page_index_change, inputs=[is_img2img_flag, img_path, page_index], outputs=[history_gallery, page_index]) + + #other funcitons + set_index.click(show_image_info, _js="images_history_get_current_img", inputs=[is_img2img_flag, img_path, filenames], outputs=[img_file_name, image_index, hide_image]) + delete.click(delete_image, inputs=[is_img2img_flag, img_path, img_file_name, page_index, filenames, image_index], outputs=gallery_outputs) + hide_image.change(fn=run_pnginfo, inputs=[hide_image], outputs=[info1, img_file_info, info2]) + switch_dict["fn"](pnginfo_send_to_txt2img, switch_dict["t2i"], img_file_info, 'switch_to_txt2img') + switch_dict["fn"](pnginfo_send_to_img2img, switch_dict["i2i"], img_file_info, 'switch_to_img2img_img2img') + -def create_history_tabs(gr, opts): +def create_history_tabs(gr, opts, run_pnginfo, switch_dict): with gr.Blocks(analytics_enabled=False) as images_history: with gr.Tabs() as tabs: with gr.Tab("txt2img history", id="images_history_txt2img"): with gr.Blocks(analytics_enabled=False) as images_history_txt2img: - show_images_history(gr, opts, is_img2img=False) + show_images_history(gr, opts, False, run_pnginfo, switch_dict) with gr.Tab("img2img history", id="images_history_img2img"): with gr.Blocks(analytics_enabled=False) as images_history_img2img: - show_images_history(gr, opts, is_img2img=True) + show_images_history(gr, opts, True, run_pnginfo, switch_dict) return images_history diff --git a/modules/ui.py b/modules/ui.py index 8762fcf5..21c9236b 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1131,8 +1131,15 @@ def create_ui(wrap_gradio_gpu_call): return f'{changed} settings changed.', opts.dumpjson() + #images history + images_history_switch_dict = { + "fn":modules.generation_parameters_copypaste.connect_paste, + "t2i":txt2img_paste_fields, + "i2i":img2img_paste_fields + } + images_history = img_his.create_history_tabs(gr, opts, wrap_gradio_call(modules.extras.run_pnginfo), images_history_switch_dict) - images_history = img_his.create_history_tabs(gr, opts) + with gr.Blocks(analytics_enabled=False) as settings_interface: settings_submit = gr.Button(value="Apply settings", variant='primary') result = gr.HTML() diff --git a/testui.py b/testui.py deleted file mode 100644 index f54e4a62..00000000 --- a/testui.py +++ /dev/null @@ -1,124 +0,0 @@ -import os -import threading -import time -import importlib -import signal -import threading - -from modules.paths import script_path - -from modules import devices, sd_samplers -import modules.codeformer_model as codeformer -import modules.extras -import modules.face_restoration -import modules.gfpgan_model as gfpgan -import modules.img2img - -import modules.lowvram -import modules.paths -import modules.scripts -import modules.sd_hijack -import modules.sd_models -import modules.shared as shared -import modules.txt2img - -import modules.ui -from modules import devices -from modules import modelloader -from modules.paths import script_path -from modules.shared import cmd_opts - -modelloader.cleanup_models() -modules.sd_models.setup_model() -codeformer.setup_model(cmd_opts.codeformer_models_path) -gfpgan.setup_model(cmd_opts.gfpgan_models_path) -shared.face_restorers.append(modules.face_restoration.FaceRestoration()) -modelloader.load_upscalers() -queue_lock = threading.Lock() - - -def wrap_queued_call(func): - def f(*args, **kwargs): - with queue_lock: - res = func(*args, **kwargs) - - return res - - return f - - -def wrap_gradio_gpu_call(func, extra_outputs=None): - def f(*args, **kwargs): - devices.torch_gc() - - shared.state.sampling_step = 0 - shared.state.job_count = -1 - shared.state.job_no = 0 - shared.state.job_timestamp = shared.state.get_job_timestamp() - shared.state.current_latent = None - shared.state.current_image = None - shared.state.current_image_sampling_step = 0 - shared.state.interrupted = False - shared.state.textinfo = None - - with queue_lock: - res = func(*args, **kwargs) - - shared.state.job = "" - shared.state.job_count = 0 - - devices.torch_gc() - - return res - - return modules.ui.wrap_gradio_call(f, extra_outputs=extra_outputs) - - -modules.scripts.load_scripts(os.path.join(script_path, "scripts")) - -shared.sd_model = None #modules.sd_models.load_model() -#shared.opts.onchange("sd_model_checkpoint", wrap_queued_call(lambda: modules.sd_models.reload_model_weights(shared.sd_model))) - - -def webui(): - # make the program just exit at ctrl+c without waiting for anything - def sigint_handler(sig, frame): - print(f'Interrupted with signal {sig} in {frame}') - os._exit(0) - - signal.signal(signal.SIGINT, sigint_handler) - - while 1: - - demo = modules.ui.create_ui(wrap_gradio_gpu_call=wrap_gradio_gpu_call) - - demo.launch( - share=cmd_opts.share, - server_name="0.0.0.0" if cmd_opts.listen else None, - server_port=cmd_opts.port, - debug=cmd_opts.gradio_debug, - auth=[tuple(cred.split(':')) for cred in cmd_opts.gradio_auth.strip('"').split(',')] if cmd_opts.gradio_auth else None, - inbrowser=cmd_opts.autolaunch, - prevent_thread_lock=True - ) - - while 1: - time.sleep(0.5) - if getattr(demo, 'do_restart', False): - time.sleep(0.5) - demo.close() - time.sleep(0.5) - break - - sd_samplers.set_samplers() - - print('Reloading Custom Scripts') - modules.scripts.reload_scripts(os.path.join(script_path, "scripts")) - print('Reloading modules: modules.ui') - importlib.reload(modules.ui) - print('Restarting Gradio') - - - -if __name__ == "__main__": - webui() -- cgit v1.2.3 From ea00c1624bbb0dcb5be07f59c9509061baddf5b1 Mon Sep 17 00:00:00 2001 From: alg-wiki Date: Mon, 10 Oct 2022 17:07:46 +0900 Subject: Textual Inversion: Added custom training image size and number of repeats per input image in a single epoch --- modules/textual_inversion/dataset.py | 6 +++--- modules/textual_inversion/preprocess.py | 4 ++-- modules/textual_inversion/textual_inversion.py | 15 ++++++++++++--- modules/ui.py | 8 +++++++- 4 files changed, 24 insertions(+), 9 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/textual_inversion/dataset.py b/modules/textual_inversion/dataset.py index 7c44ea5b..acc4ce59 100644 --- a/modules/textual_inversion/dataset.py +++ b/modules/textual_inversion/dataset.py @@ -15,13 +15,13 @@ re_tag = re.compile(r"[a-zA-Z][_\w\d()]+") class PersonalizedBase(Dataset): - def __init__(self, data_root, size=None, repeats=100, flip_p=0.5, placeholder_token="*", width=512, height=512, model=None, device=None, template_file=None): + def __init__(self, data_root, size, repeats, flip_p=0.5, placeholder_token="*", model=None, device=None, template_file=None): self.placeholder_token = placeholder_token self.size = size - self.width = width - self.height = height + self.width = size + self.height = size self.flip = transforms.RandomHorizontalFlip(p=flip_p) self.dataset = [] diff --git a/modules/textual_inversion/preprocess.py b/modules/textual_inversion/preprocess.py index f1c002a2..b3de6fd7 100644 --- a/modules/textual_inversion/preprocess.py +++ b/modules/textual_inversion/preprocess.py @@ -7,8 +7,8 @@ import tqdm from modules import shared, images -def preprocess(process_src, process_dst, process_flip, process_split, process_caption): - size = 512 +def preprocess(process_src, process_dst, process_size, process_flip, process_split, process_caption): + size = process_size src = os.path.abspath(process_src) dst = os.path.abspath(process_dst) diff --git a/modules/textual_inversion/textual_inversion.py b/modules/textual_inversion/textual_inversion.py index cd9f3498..e34dc2e8 100644 --- a/modules/textual_inversion/textual_inversion.py +++ b/modules/textual_inversion/textual_inversion.py @@ -6,6 +6,7 @@ import torch import tqdm import html import datetime +import math from modules import shared, devices, sd_hijack, processing, sd_models @@ -156,7 +157,7 @@ def create_embedding(name, num_vectors_per_token, init_text='*'): return fn -def train_embedding(embedding_name, learn_rate, data_root, log_directory, steps, create_image_every, save_embedding_every, template_file): +def train_embedding(embedding_name, learn_rate, data_root, log_directory, training_size, steps, num_repeats, create_image_every, save_embedding_every, template_file): assert embedding_name, 'embedding not selected' shared.state.textinfo = "Initializing textual inversion training..." @@ -182,7 +183,7 @@ def train_embedding(embedding_name, learn_rate, data_root, log_directory, steps, shared.state.textinfo = f"Preparing dataset from {html.escape(data_root)}..." with torch.autocast("cuda"): - ds = modules.textual_inversion.dataset.PersonalizedBase(data_root=data_root, size=512, placeholder_token=embedding_name, model=shared.sd_model, device=devices.device, template_file=template_file) + ds = modules.textual_inversion.dataset.PersonalizedBase(data_root=data_root, size=training_size, repeats=num_repeats, placeholder_token=embedding_name, model=shared.sd_model, device=devices.device, template_file=template_file) hijack = sd_hijack.model_hijack @@ -200,6 +201,9 @@ def train_embedding(embedding_name, learn_rate, data_root, log_directory, steps, if ititial_step > steps: return embedding, filename + tr_img_len = len([os.path.join(data_root, file_path) for file_path in os.listdir(data_root)]) + epoch_len = (tr_img_len * num_repeats) + tr_img_len + pbar = tqdm.tqdm(enumerate(ds), total=steps-ititial_step) for i, (x, text) in pbar: embedding.step = i + ititial_step @@ -223,7 +227,10 @@ def train_embedding(embedding_name, learn_rate, data_root, log_directory, steps, loss.backward() optimizer.step() - pbar.set_description(f"loss: {losses.mean():.7f}") + epoch_num = math.floor(embedding.step / epoch_len) + epoch_step = embedding.step - (epoch_num * epoch_len) + + pbar.set_description(f"[Epoch {epoch_num}: {epoch_step}/{epoch_len}]loss: {losses.mean():.7f}") if embedding.step > 0 and embedding_dir is not None and embedding.step % save_embedding_every == 0: last_saved_file = os.path.join(embedding_dir, f'{embedding_name}-{embedding.step}.pt') @@ -236,6 +243,8 @@ def train_embedding(embedding_name, learn_rate, data_root, log_directory, steps, sd_model=shared.sd_model, prompt=text, steps=20, + height=training_size, + width=training_size, do_not_save_grid=True, do_not_save_samples=True, ) diff --git a/modules/ui.py b/modules/ui.py index 2231a8ed..f821fd8d 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1029,6 +1029,7 @@ def create_ui(wrap_gradio_gpu_call): process_src = gr.Textbox(label='Source directory') process_dst = gr.Textbox(label='Destination directory') + process_size = gr.Slider(minimum=64, maximum=2048, step=64, label="Size (width and height)", value=512) with gr.Row(): process_flip = gr.Checkbox(label='Create flipped copies') @@ -1043,13 +1044,15 @@ def create_ui(wrap_gradio_gpu_call): run_preprocess = gr.Button(value="Preprocess", variant='primary') with gr.Group(): - gr.HTML(value="

Train an embedding; must specify a directory with a set of 512x512 images

") + gr.HTML(value="

Train an embedding; must specify a directory with a set of 1:1 ratio images

") train_embedding_name = gr.Dropdown(label='Embedding', choices=sorted(sd_hijack.model_hijack.embedding_db.word_embeddings.keys())) learn_rate = gr.Number(label='Learning rate', value=5.0e-03) dataset_directory = gr.Textbox(label='Dataset directory', placeholder="Path to directory with input images") log_directory = gr.Textbox(label='Log directory', placeholder="Path to directory where to write outputs", value="textual_inversion") template_file = gr.Textbox(label='Prompt template file', value=os.path.join(script_path, "textual_inversion_templates", "style_filewords.txt")) + training_size = gr.Slider(minimum=64, maximum=2048, step=64, label="Size (width and height)", value=512) steps = gr.Number(label='Max steps', value=100000, precision=0) + num_repeats = gr.Number(label='Number of repeats for a single input image per epoch', value=100, precision=0) create_image_every = gr.Number(label='Save an image to log directory every N steps, 0 to disable', value=500, precision=0) save_embedding_every = gr.Number(label='Save a copy of embedding to log directory every N steps, 0 to disable', value=500, precision=0) @@ -1092,6 +1095,7 @@ def create_ui(wrap_gradio_gpu_call): inputs=[ process_src, process_dst, + process_size, process_flip, process_split, process_caption, @@ -1110,7 +1114,9 @@ def create_ui(wrap_gradio_gpu_call): learn_rate, dataset_directory, log_directory, + training_size, steps, + num_repeats, create_image_every, save_embedding_every, template_file, -- cgit v1.2.3 From 7a20f914eddfdf09c0ccced157ec108205bc3d0f Mon Sep 17 00:00:00 2001 From: alg-wiki Date: Mon, 10 Oct 2022 22:35:35 +0900 Subject: Custom Width and Height --- modules/textual_inversion/dataset.py | 7 +++---- modules/textual_inversion/preprocess.py | 19 ++++++++++--------- modules/textual_inversion/textual_inversion.py | 11 +++++------ modules/ui.py | 12 ++++++++---- 4 files changed, 26 insertions(+), 23 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/textual_inversion/dataset.py b/modules/textual_inversion/dataset.py index acc4ce59..bcf772d2 100644 --- a/modules/textual_inversion/dataset.py +++ b/modules/textual_inversion/dataset.py @@ -15,13 +15,12 @@ re_tag = re.compile(r"[a-zA-Z][_\w\d()]+") class PersonalizedBase(Dataset): - def __init__(self, data_root, size, repeats, flip_p=0.5, placeholder_token="*", model=None, device=None, template_file=None): + def __init__(self, data_root, width, height, repeats, flip_p=0.5, placeholder_token="*", model=None, device=None, template_file=None): self.placeholder_token = placeholder_token - self.size = size - self.width = size - self.height = size + self.width = width + self.height = height self.flip = transforms.RandomHorizontalFlip(p=flip_p) self.dataset = [] diff --git a/modules/textual_inversion/preprocess.py b/modules/textual_inversion/preprocess.py index b3de6fd7..d7efdef2 100644 --- a/modules/textual_inversion/preprocess.py +++ b/modules/textual_inversion/preprocess.py @@ -7,8 +7,9 @@ import tqdm from modules import shared, images -def preprocess(process_src, process_dst, process_size, process_flip, process_split, process_caption): - size = process_size +def preprocess(process_src, process_dst, process_width, process_height, process_flip, process_split, process_caption): + width = process_width + height = process_height src = os.path.abspath(process_src) dst = os.path.abspath(process_dst) @@ -55,23 +56,23 @@ def preprocess(process_src, process_dst, process_size, process_flip, process_spl is_wide = ratio < 1 / 1.35 if process_split and is_tall: - img = img.resize((size, size * img.height // img.width)) + img = img.resize((width, height * img.height // img.width)) - top = img.crop((0, 0, size, size)) + top = img.crop((0, 0, width, height)) save_pic(top, index) - bot = img.crop((0, img.height - size, size, img.height)) + bot = img.crop((0, img.height - height, width, img.height)) save_pic(bot, index) elif process_split and is_wide: - img = img.resize((size * img.width // img.height, size)) + img = img.resize((width * img.width // img.height, height)) - left = img.crop((0, 0, size, size)) + left = img.crop((0, 0, width, height)) save_pic(left, index) - right = img.crop((img.width - size, 0, img.width, size)) + right = img.crop((img.width - width, 0, img.width, height)) save_pic(right, index) else: - img = images.resize_image(1, img, size, size) + img = images.resize_image(1, img, width, height) save_pic(img, index) shared.state.nextjob() diff --git a/modules/textual_inversion/textual_inversion.py b/modules/textual_inversion/textual_inversion.py index 769682ea..5965c5a0 100644 --- a/modules/textual_inversion/textual_inversion.py +++ b/modules/textual_inversion/textual_inversion.py @@ -6,7 +6,6 @@ import torch import tqdm import html import datetime -import math from modules import shared, devices, sd_hijack, processing, sd_models @@ -157,7 +156,7 @@ def create_embedding(name, num_vectors_per_token, init_text='*'): return fn -def train_embedding(embedding_name, learn_rate, data_root, log_directory, training_size, steps, num_repeats, create_image_every, save_embedding_every, template_file): +def train_embedding(embedding_name, learn_rate, data_root, log_directory, training_width, training_height, steps, num_repeats, create_image_every, save_embedding_every, template_file): assert embedding_name, 'embedding not selected' shared.state.textinfo = "Initializing textual inversion training..." @@ -183,7 +182,7 @@ def train_embedding(embedding_name, learn_rate, data_root, log_directory, traini shared.state.textinfo = f"Preparing dataset from {html.escape(data_root)}..." with torch.autocast("cuda"): - ds = modules.textual_inversion.dataset.PersonalizedBase(data_root=data_root, size=training_size, repeats=num_repeats, placeholder_token=embedding_name, model=shared.sd_model, device=devices.device, template_file=template_file) + ds = modules.textual_inversion.dataset.PersonalizedBase(data_root=data_root, width=training_width, height=training_height, repeats=num_repeats, placeholder_token=embedding_name, model=shared.sd_model, device=devices.device, template_file=template_file) hijack = sd_hijack.model_hijack @@ -227,7 +226,7 @@ def train_embedding(embedding_name, learn_rate, data_root, log_directory, traini loss.backward() optimizer.step() - epoch_num = math.floor(embedding.step / epoch_len) + epoch_num = embedding.step // epoch_len epoch_step = embedding.step - (epoch_num * epoch_len) + 1 pbar.set_description(f"[Epoch {epoch_num}: {epoch_step}/{epoch_len}]loss: {losses.mean():.7f}") @@ -243,8 +242,8 @@ def train_embedding(embedding_name, learn_rate, data_root, log_directory, traini sd_model=shared.sd_model, prompt=text, steps=20, - height=training_size, - width=training_size, + height=training_height, + width=training_width, do_not_save_grid=True, do_not_save_samples=True, ) diff --git a/modules/ui.py b/modules/ui.py index f821fd8d..8c06ad7c 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1029,7 +1029,8 @@ def create_ui(wrap_gradio_gpu_call): process_src = gr.Textbox(label='Source directory') process_dst = gr.Textbox(label='Destination directory') - process_size = gr.Slider(minimum=64, maximum=2048, step=64, label="Size (width and height)", value=512) + process_width = gr.Slider(minimum=64, maximum=2048, step=64, label="Width", value=512) + process_height = gr.Slider(minimum=64, maximum=2048, step=64, label="Height", value=512) with gr.Row(): process_flip = gr.Checkbox(label='Create flipped copies') @@ -1050,7 +1051,8 @@ def create_ui(wrap_gradio_gpu_call): dataset_directory = gr.Textbox(label='Dataset directory', placeholder="Path to directory with input images") log_directory = gr.Textbox(label='Log directory', placeholder="Path to directory where to write outputs", value="textual_inversion") template_file = gr.Textbox(label='Prompt template file', value=os.path.join(script_path, "textual_inversion_templates", "style_filewords.txt")) - training_size = gr.Slider(minimum=64, maximum=2048, step=64, label="Size (width and height)", value=512) + training_width = gr.Slider(minimum=64, maximum=2048, step=64, label="Width", value=512) + training_height = gr.Slider(minimum=64, maximum=2048, step=64, label="Height", value=512) steps = gr.Number(label='Max steps', value=100000, precision=0) num_repeats = gr.Number(label='Number of repeats for a single input image per epoch', value=100, precision=0) create_image_every = gr.Number(label='Save an image to log directory every N steps, 0 to disable', value=500, precision=0) @@ -1095,7 +1097,8 @@ def create_ui(wrap_gradio_gpu_call): inputs=[ process_src, process_dst, - process_size, + process_width, + process_height, process_flip, process_split, process_caption, @@ -1114,7 +1117,8 @@ def create_ui(wrap_gradio_gpu_call): learn_rate, dataset_directory, log_directory, - training_size, + training_width, + training_height, steps, num_repeats, create_image_every, -- cgit v1.2.3 From f347ddfd808c56bb1bacdec0c4bedf826ff85cd8 Mon Sep 17 00:00:00 2001 From: RW21 Date: Mon, 10 Oct 2022 10:44:11 +0900 Subject: Remove max_batch_count from ui.py --- modules/ui.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/ui.py b/modules/ui.py index 8c06ad7c..8ba84911 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -524,7 +524,7 @@ def create_ui(wrap_gradio_gpu_call): denoising_strength = gr.Slider(minimum=0.0, maximum=1.0, step=0.01, label='Denoising strength', value=0.7) with gr.Row(): - batch_count = gr.Slider(minimum=1, maximum=cmd_opts.max_batch_count, step=1, label='Batch count', value=1) + batch_count = gr.Slider(minimum=1, step=1, label='Batch count', value=1) batch_size = gr.Slider(minimum=1, maximum=8, step=1, label='Batch size', value=1) cfg_scale = gr.Slider(minimum=1.0, maximum=30.0, step=0.5, label='CFG Scale', value=7.0) @@ -710,7 +710,7 @@ def create_ui(wrap_gradio_gpu_call): tiling = gr.Checkbox(label='Tiling', value=False) with gr.Row(): - batch_count = gr.Slider(minimum=1, maximum=cmd_opts.max_batch_count, step=1, label='Batch count', value=1) + batch_count = gr.Slider(minimum=1, step=1, label='Batch count', value=1) batch_size = gr.Slider(minimum=1, maximum=8, step=1, label='Batch size', value=1) with gr.Group(): -- cgit v1.2.3 From 9d33baba587637815d818e5e641d8f8b74c4900d Mon Sep 17 00:00:00 2001 From: Vladimir Repin <32306715+mezotaken@users.noreply.github.com> Date: Mon, 10 Oct 2022 18:46:48 +0300 Subject: Always show previous mask and fix extras_send dest --- modules/ui.py | 2 +- style.css | 7 +++++++ 2 files changed, 8 insertions(+), 1 deletion(-) (limited to 'modules/ui.py') diff --git a/modules/ui.py b/modules/ui.py index 8ba84911..e8039d76 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -961,7 +961,7 @@ def create_ui(wrap_gradio_gpu_call): extras_send_to_inpaint.click( fn=lambda x: image_from_url_text(x), - _js="extract_image_from_gallery_img2img", + _js="extract_image_from_gallery_inpaint", inputs=[result_images], outputs=[init_img_with_mask], ) diff --git a/style.css b/style.css index 04bb9576..00a3d07f 100644 --- a/style.css +++ b/style.css @@ -467,3 +467,10 @@ input[type="range"]{ max-width: 32em; padding: 0; } + +canvas[key="mask"] { + z-index: 12 !important; + filter: invert(); + mix-blend-mode: multiply; + pointer-events: none; +} -- cgit v1.2.3 From 1d64976dbc5a0f3124567b91fadd5014a9d93c5f Mon Sep 17 00:00:00 2001 From: Justin Maier Date: Mon, 10 Oct 2022 12:04:21 -0600 Subject: Simplify crop logic --- modules/extras.py | 14 +++----------- modules/ui.py | 4 ++-- 2 files changed, 5 insertions(+), 13 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/extras.py b/modules/extras.py index 83ca7049..b24d7de3 100644 --- a/modules/extras.py +++ b/modules/extras.py @@ -73,16 +73,6 @@ def run_extras(extras_mode, resize_mode, image, image_folder, gfpgan_visibility, crop_info = " (crop)" if upscaling_crop else "" info += f"Resize to: {upscaling_resize_w:g}x{upscaling_resize_h:g}{crop_info}\n" - def crop_upscaled_center(image, resize_w, resize_h): - left = int(math.ceil((image.width - resize_w) / 2)) - right = image.width - int(math.floor((image.width - resize_w) / 2)) - top = int(math.ceil((image.height - resize_h) / 2)) - bottom = image.height - int(math.floor((image.height - resize_h) / 2)) - - image = image.crop((left, top, right, bottom)) - return image - - if upscaling_resize != 1.0: def upscale(image, scaler_index, resize, mode, resize_w, resize_h, crop): small = image.crop((image.width // 2, image.height // 2, image.width // 2 + 10, image.height // 2 + 10)) @@ -94,7 +84,9 @@ def run_extras(extras_mode, resize_mode, image, image_folder, gfpgan_visibility, upscaler = shared.sd_upscalers[scaler_index] c = upscaler.scaler.upscale(image, resize, upscaler.data_path) if mode == 1 and crop: - c = crop_upscaled_center(c, resize_w, resize_h) + cropped = Image.new("RGB", (resize_w, resize_h)) + cropped.paste(c, box=(resize_w // 2 - c.width // 2, resize_h // 2 - c.height // 2)) + c = cropped cached_images[key] = c return c diff --git a/modules/ui.py b/modules/ui.py index 4bb2892b..1aabe18d 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -909,8 +909,8 @@ def create_ui(wrap_gradio_gpu_call): with gr.TabItem('Scale to'): with gr.Group(): with gr.Row(): - upscaling_resize_w = gr.Number(label="Width", value=512) - upscaling_resize_h = gr.Number(label="Height", value=512) + upscaling_resize_w = gr.Number(label="Width", value=512, precision=0) + upscaling_resize_h = gr.Number(label="Height", value=512, precision=0) upscaling_crop = gr.Checkbox(label='Crop to fit', value=True) with gr.Group(): -- cgit v1.2.3 From 2536ecbb1790da2af0d61b6a26f38732cba665cd Mon Sep 17 00:00:00 2001 From: Fampai <> Date: Mon, 10 Oct 2022 17:10:29 -0400 Subject: Refactored learning rate code --- modules/textual_inversion/textual_inversion.py | 51 ++++++++++++++++++++++++-- modules/ui.py | 2 +- 2 files changed, 48 insertions(+), 5 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/textual_inversion/textual_inversion.py b/modules/textual_inversion/textual_inversion.py index 5965c5a0..c64a4598 100644 --- a/modules/textual_inversion/textual_inversion.py +++ b/modules/textual_inversion/textual_inversion.py @@ -189,8 +189,6 @@ def train_embedding(embedding_name, learn_rate, data_root, log_directory, traini embedding = hijack.embedding_db.word_embeddings[embedding_name] embedding.vec.requires_grad = True - optimizer = torch.optim.AdamW([embedding.vec], lr=learn_rate) - losses = torch.zeros((32,)) last_saved_file = "" @@ -203,12 +201,24 @@ def train_embedding(embedding_name, learn_rate, data_root, log_directory, traini tr_img_len = len([os.path.join(data_root, file_path) for file_path in os.listdir(data_root)]) epoch_len = (tr_img_len * num_repeats) + tr_img_len + scheduleIter = iter(LearnSchedule(learn_rate, steps, ititial_step)) + (learn_rate, end_step) = next(scheduleIter) + print(f'Training at rate of {learn_rate} until step {end_step}') + + optimizer = torch.optim.AdamW([embedding.vec], lr=learn_rate) + pbar = tqdm.tqdm(enumerate(ds), total=steps-ititial_step) for i, (x, text) in pbar: embedding.step = i + ititial_step - if embedding.step > steps: - break + if embedding.step > end_step: + try: + (learn_rate, end_step) = next(scheduleIter) + except: + break + tqdm.tqdm.write(f'Training at rate of {learn_rate} until step {end_step}') + for pg in optimizer.param_groups: + pg['lr'] = learn_rate if shared.state.interrupted: break @@ -277,3 +287,36 @@ Last saved image: {html.escape(last_saved_image)}
return embedding, filename +class LearnSchedule: + def __init__(self, learn_rate, max_steps, cur_step=0): + pairs = learn_rate.split(',') + self.rates = [] + self.it = 0 + self.maxit = 0 + for i, pair in enumerate(pairs): + tmp = pair.split(':') + if len(tmp) == 2: + step = int(tmp[1]) + if step > cur_step: + self.rates.append((float(tmp[0]), min(step, max_steps))) + self.maxit += 1 + if step > max_steps: + return + elif step == -1: + self.rates.append((float(tmp[0]), max_steps)) + self.maxit += 1 + return + else: + self.rates.append((float(tmp[0]), max_steps)) + self.maxit += 1 + return + + def __iter__(self): + return self + + def __next__(self): + if self.it < self.maxit: + self.it += 1 + return self.rates[self.it - 1] + else: + raise StopIteration diff --git a/modules/ui.py b/modules/ui.py index 8c06ad7c..c9e8355b 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1047,7 +1047,7 @@ def create_ui(wrap_gradio_gpu_call): with gr.Group(): gr.HTML(value="

Train an embedding; must specify a directory with a set of 1:1 ratio images

") train_embedding_name = gr.Dropdown(label='Embedding', choices=sorted(sd_hijack.model_hijack.embedding_db.word_embeddings.keys())) - learn_rate = gr.Number(label='Learning rate', value=5.0e-03) + learn_rate = gr.Textbox(label='Learning rate', placeholder="Learning rate", value = "5.0e-03") dataset_directory = gr.Textbox(label='Dataset directory', placeholder="Path to directory with input images") log_directory = gr.Textbox(label='Log directory', placeholder="Path to directory where to write outputs", value="textual_inversion") template_file = gr.Textbox(label='Prompt template file', value=os.path.join(script_path, "textual_inversion_templates", "style_filewords.txt")) -- cgit v1.2.3 From 1add3cff84b7e2436d69b1e97ae689281e4a7c33 Mon Sep 17 00:00:00 2001 From: papuSpartan Date: Mon, 10 Oct 2022 19:57:43 -0500 Subject: Refresh list of models/ckpts upon hitting restart gradio in the settings pane --- modules/ui.py | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'modules/ui.py') diff --git a/modules/ui.py b/modules/ui.py index e8039d76..06ff118f 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -39,6 +39,7 @@ import modules.generation_parameters_copypaste from modules import prompt_parser from modules.images import save_image import modules.textual_inversion.ui +from modules.sd_models import list_models # this is a fix for Windows users. Without it, javascript files will be served with text/html content-type and the browser will not show any UI mimetypes.init() @@ -1290,6 +1291,9 @@ Requested path was: {f} shared.state.interrupt() settings_interface.gradio_ref.do_restart = True + # refresh models so that new models/.ckpt's show up on reload + list_models() + restart_gradio.click( fn=request_restart, inputs=[], -- cgit v1.2.3 From 8b7d3f1bef47bbe048f644ed0d8dd3ad46554045 Mon Sep 17 00:00:00 2001 From: Jairo Correa Date: Tue, 11 Oct 2022 02:22:46 -0300 Subject: Make the ctrl+enter shortcut use the generate button on the current tab --- modules/ui.py | 2 +- script.js | 11 +++++++++-- 2 files changed, 10 insertions(+), 3 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/ui.py b/modules/ui.py index e8039d76..cafda884 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1331,7 +1331,7 @@ Requested path was: {f} with gr.Tabs() as tabs: for interface, label, ifid in interfaces: - with gr.TabItem(label, id=ifid): + with gr.TabItem(label, id=ifid, elem_id='tab_' + ifid): interface.render() if os.path.exists(os.path.join(script_path, "notification.mp3")): diff --git a/script.js b/script.js index a92c0f77..9543cbe6 100644 --- a/script.js +++ b/script.js @@ -6,6 +6,10 @@ function get_uiCurrentTab() { return gradioApp().querySelector('.tabs button:not(.border-transparent)') } +function get_uiCurrentTabContent() { + return gradioApp().querySelector('.tabitem[id^=tab_]:not([style*="display: none"])') +} + uiUpdateCallbacks = [] uiTabChangeCallbacks = [] let uiCurrentTab = null @@ -50,8 +54,11 @@ document.addEventListener("DOMContentLoaded", function() { } else if (e.keyCode !== undefined) { if((e.keyCode == 13 && (e.metaKey || e.ctrlKey))) handled = true; } - if (handled) { - gradioApp().querySelector("#txt2img_generate").click(); + if (handled) { + button = get_uiCurrentTabContent().querySelector('button[id$=_generate]'); + if (button) { + button.click(); + } e.preventDefault(); } }) -- cgit v1.2.3 From 8617396c6df71074c7fd3d39419802026874712a Mon Sep 17 00:00:00 2001 From: Kenneth Date: Mon, 10 Oct 2022 17:23:07 -0600 Subject: Added slider for deepbooru score threshold in settings --- modules/shared.py | 1 + modules/ui.py | 2 +- 2 files changed, 2 insertions(+), 1 deletion(-) (limited to 'modules/ui.py') diff --git a/modules/shared.py b/modules/shared.py index ecd15ef5..e0830e28 100644 --- a/modules/shared.py +++ b/modules/shared.py @@ -239,6 +239,7 @@ options_templates.update(options_section(('interrogate', "Interrogate Options"), "interrogate_clip_min_length": OptionInfo(24, "Interrogate: minimum description length (excluding artists, etc..)", gr.Slider, {"minimum": 1, "maximum": 128, "step": 1}), "interrogate_clip_max_length": OptionInfo(48, "Interrogate: maximum description length", gr.Slider, {"minimum": 1, "maximum": 256, "step": 1}), "interrogate_clip_dict_limit": OptionInfo(1500, "Interrogate: maximum number of lines in text file (0 = No limit)"), + "interrogate_deepbooru_score_threshold": OptionInfo(0.5, "Interrogate: deepbooru score threshold", gr.Slider, {"minimum": 0, "maximum": 1, "step": 0.01}), })) options_templates.update(options_section(('ui', "User interface"), { diff --git a/modules/ui.py b/modules/ui.py index cafda884..ca3151c4 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -311,7 +311,7 @@ def interrogate(image): def interrogate_deepbooru(image): - prompt = get_deepbooru_tags(image) + prompt = get_deepbooru_tags(image, opts.interrogate_deepbooru_score_threshold) return gr_show(True) if prompt is None else prompt -- cgit v1.2.3 From 530103b586109c11fd068eb70ef09503ec6a4caf Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Tue, 11 Oct 2022 14:53:02 +0300 Subject: fixes related to merge --- modules/hypernetwork.py | 103 ------------------------- modules/hypernetwork/hypernetwork.py | 74 +++++++++++------- modules/hypernetwork/ui.py | 10 +-- modules/sd_hijack_optimizations.py | 3 +- modules/shared.py | 13 +++- modules/textual_inversion/textual_inversion.py | 12 +-- modules/ui.py | 5 +- scripts/xy_grid.py | 3 +- webui.py | 15 +--- 9 files changed, 78 insertions(+), 160 deletions(-) delete mode 100644 modules/hypernetwork.py (limited to 'modules/ui.py') diff --git a/modules/hypernetwork.py b/modules/hypernetwork.py deleted file mode 100644 index 7bbc443e..00000000 --- a/modules/hypernetwork.py +++ /dev/null @@ -1,103 +0,0 @@ -import glob -import os -import sys -import traceback - -import torch - -from ldm.util import default -from modules import devices, shared -import torch -from torch import einsum -from einops import rearrange, repeat - - -class HypernetworkModule(torch.nn.Module): - def __init__(self, dim, state_dict): - super().__init__() - - self.linear1 = torch.nn.Linear(dim, dim * 2) - self.linear2 = torch.nn.Linear(dim * 2, dim) - - self.load_state_dict(state_dict, strict=True) - self.to(devices.device) - - def forward(self, x): - return x + (self.linear2(self.linear1(x))) - - -class Hypernetwork: - filename = None - name = None - - def __init__(self, filename): - self.filename = filename - self.name = os.path.splitext(os.path.basename(filename))[0] - self.layers = {} - - state_dict = torch.load(filename, map_location='cpu') - for size, sd in state_dict.items(): - self.layers[size] = (HypernetworkModule(size, sd[0]), HypernetworkModule(size, sd[1])) - - -def list_hypernetworks(path): - res = {} - for filename in glob.iglob(os.path.join(path, '**/*.pt'), recursive=True): - name = os.path.splitext(os.path.basename(filename))[0] - res[name] = filename - return res - - -def load_hypernetwork(filename): - path = shared.hypernetworks.get(filename, None) - if path is not None: - print(f"Loading hypernetwork {filename}") - try: - shared.loaded_hypernetwork = Hypernetwork(path) - except Exception: - print(f"Error loading hypernetwork {path}", file=sys.stderr) - print(traceback.format_exc(), file=sys.stderr) - else: - if shared.loaded_hypernetwork is not None: - print(f"Unloading hypernetwork") - - shared.loaded_hypernetwork = None - - -def apply_hypernetwork(hypernetwork, context): - hypernetwork_layers = (hypernetwork.layers if hypernetwork is not None else {}).get(context.shape[2], None) - - if hypernetwork_layers is None: - return context, context - - context_k = hypernetwork_layers[0](context) - context_v = hypernetwork_layers[1](context) - return context_k, context_v - - -def attention_CrossAttention_forward(self, x, context=None, mask=None): - h = self.heads - - q = self.to_q(x) - context = default(context, x) - - context_k, context_v = apply_hypernetwork(shared.loaded_hypernetwork, context) - k = self.to_k(context_k) - v = self.to_v(context_v) - - q, k, v = map(lambda t: rearrange(t, 'b n (h d) -> (b h) n d', h=h), (q, k, v)) - - sim = einsum('b i d, b j d -> b i j', q, k) * self.scale - - if mask is not None: - mask = rearrange(mask, 'b ... -> b (...)') - max_neg_value = -torch.finfo(sim.dtype).max - mask = repeat(mask, 'b j -> (b h) () j', h=h) - sim.masked_fill_(~mask, max_neg_value) - - # attention, what we cannot get enough of - attn = sim.softmax(dim=-1) - - out = einsum('b i j, b j d -> b i d', attn, v) - out = rearrange(out, '(b h) n d -> b n (h d)', h=h) - return self.to_out(out) diff --git a/modules/hypernetwork/hypernetwork.py b/modules/hypernetwork/hypernetwork.py index a3d6a47e..aa701bda 100644 --- a/modules/hypernetwork/hypernetwork.py +++ b/modules/hypernetwork/hypernetwork.py @@ -26,10 +26,11 @@ class HypernetworkModule(torch.nn.Module): if state_dict is not None: self.load_state_dict(state_dict, strict=True) else: - self.linear1.weight.data.fill_(0.0001) - self.linear1.bias.data.fill_(0.0001) - self.linear2.weight.data.fill_(0.0001) - self.linear2.bias.data.fill_(0.0001) + + self.linear1.weight.data.normal_(mean=0.0, std=0.01) + self.linear1.bias.data.zero_() + self.linear2.weight.data.normal_(mean=0.0, std=0.01) + self.linear2.bias.data.zero_() self.to(devices.device) @@ -92,41 +93,54 @@ class Hypernetwork: self.sd_checkpoint_name = state_dict.get('sd_checkpoint_name', None) -def load_hypernetworks(path): +def list_hypernetworks(path): res = {} + for filename in glob.iglob(os.path.join(path, '**/*.pt'), recursive=True): + name = os.path.splitext(os.path.basename(filename))[0] + res[name] = filename + return res - for filename in glob.iglob(path + '**/*.pt', recursive=True): + +def load_hypernetwork(filename): + path = shared.hypernetworks.get(filename, None) + if path is not None: + print(f"Loading hypernetwork {filename}") try: - hn = Hypernetwork() - hn.load(filename) - res[hn.name] = hn + shared.loaded_hypernetwork = Hypernetwork() + shared.loaded_hypernetwork.load(path) + except Exception: - print(f"Error loading hypernetwork {filename}", file=sys.stderr) + print(f"Error loading hypernetwork {path}", file=sys.stderr) print(traceback.format_exc(), file=sys.stderr) + else: + if shared.loaded_hypernetwork is not None: + print(f"Unloading hypernetwork") - return res + shared.loaded_hypernetwork = None -def attention_CrossAttention_forward(self, x, context=None, mask=None): - h = self.heads +def apply_hypernetwork(hypernetwork, context, layer=None): + hypernetwork_layers = (hypernetwork.layers if hypernetwork is not None else {}).get(context.shape[2], None) - q = self.to_q(x) - context = default(context, x) + if hypernetwork_layers is None: + return context, context - hypernetwork_layers = (shared.hypernetwork.layers if shared.hypernetwork is not None else {}).get(context.shape[2], None) + if layer is not None: + layer.hyper_k = hypernetwork_layers[0] + layer.hyper_v = hypernetwork_layers[1] - if hypernetwork_layers is not None: - hypernetwork_k, hypernetwork_v = hypernetwork_layers + context_k = hypernetwork_layers[0](context) + context_v = hypernetwork_layers[1](context) + return context_k, context_v - self.hypernetwork_k = hypernetwork_k - self.hypernetwork_v = hypernetwork_v - context_k = hypernetwork_k(context) - context_v = hypernetwork_v(context) - else: - context_k = context - context_v = context +def attention_CrossAttention_forward(self, x, context=None, mask=None): + h = self.heads + + q = self.to_q(x) + context = default(context, x) + context_k, context_v = apply_hypernetwork(shared.loaded_hypernetwork, context, self) k = self.to_k(context_k) v = self.to_v(context_v) @@ -151,7 +165,9 @@ def attention_CrossAttention_forward(self, x, context=None, mask=None): def train_hypernetwork(hypernetwork_name, learn_rate, data_root, log_directory, steps, create_image_every, save_hypernetwork_every, template_file, preview_image_prompt): assert hypernetwork_name, 'embedding not selected' - shared.hypernetwork = shared.hypernetworks[hypernetwork_name] + path = shared.hypernetworks.get(hypernetwork_name, None) + shared.loaded_hypernetwork = Hypernetwork() + shared.loaded_hypernetwork.load(path) shared.state.textinfo = "Initializing hypernetwork training..." shared.state.job_count = steps @@ -176,9 +192,9 @@ def train_hypernetwork(hypernetwork_name, learn_rate, data_root, log_directory, shared.state.textinfo = f"Preparing dataset from {html.escape(data_root)}..." with torch.autocast("cuda"): - ds = modules.textual_inversion.dataset.PersonalizedBase(data_root=data_root, size=512, placeholder_token=hypernetwork_name, model=shared.sd_model, device=devices.device, template_file=template_file) + ds = modules.textual_inversion.dataset.PersonalizedBase(data_root=data_root, width=512, height=512, repeats=1, placeholder_token=hypernetwork_name, model=shared.sd_model, device=devices.device, template_file=template_file) - hypernetwork = shared.hypernetworks[hypernetwork_name] + hypernetwork = shared.loaded_hypernetwork weights = hypernetwork.weights() for weight in weights: weight.requires_grad = True @@ -194,7 +210,7 @@ def train_hypernetwork(hypernetwork_name, learn_rate, data_root, log_directory, if ititial_step > steps: return hypernetwork, filename - pbar = tqdm.tqdm(enumerate(ds), total=steps-ititial_step) + pbar = tqdm.tqdm(enumerate(ds), total=steps - ititial_step) for i, (x, text) in pbar: hypernetwork.step = i + ititial_step diff --git a/modules/hypernetwork/ui.py b/modules/hypernetwork/ui.py index 525f978c..f6d1d0a3 100644 --- a/modules/hypernetwork/ui.py +++ b/modules/hypernetwork/ui.py @@ -6,24 +6,24 @@ import gradio as gr import modules.textual_inversion.textual_inversion import modules.textual_inversion.preprocess from modules import sd_hijack, shared +from modules.hypernetwork import hypernetwork def create_hypernetwork(name): fn = os.path.join(shared.cmd_opts.hypernetwork_dir, f"{name}.pt") assert not os.path.exists(fn), f"file {fn} already exists" - hypernetwork = modules.hypernetwork.hypernetwork.Hypernetwork(name=name) - hypernetwork.save(fn) + hypernet = modules.hypernetwork.hypernetwork.Hypernetwork(name=name) + hypernet.save(fn) shared.reload_hypernetworks() - shared.hypernetwork = shared.hypernetworks.get(shared.opts.sd_hypernetwork, None) return gr.Dropdown.update(choices=sorted([x for x in shared.hypernetworks.keys()])), f"Created: {fn}", "" def train_hypernetwork(*args): - initial_hypernetwork = shared.hypernetwork + initial_hypernetwork = shared.loaded_hypernetwork try: sd_hijack.undo_optimizations() @@ -38,6 +38,6 @@ Hypernetwork saved to {html.escape(filename)} except Exception: raise finally: - shared.hypernetwork = initial_hypernetwork + shared.loaded_hypernetwork = initial_hypernetwork sd_hijack.apply_optimizations() diff --git a/modules/sd_hijack_optimizations.py b/modules/sd_hijack_optimizations.py index 25cb67a4..27e571fc 100644 --- a/modules/sd_hijack_optimizations.py +++ b/modules/sd_hijack_optimizations.py @@ -8,7 +8,8 @@ from torch import einsum from ldm.util import default from einops import rearrange -from modules import shared, hypernetwork +from modules import shared +from modules.hypernetwork import hypernetwork if shared.cmd_opts.xformers or shared.cmd_opts.force_enable_xformers: diff --git a/modules/shared.py b/modules/shared.py index 14b40d70..8753015e 100644 --- a/modules/shared.py +++ b/modules/shared.py @@ -13,7 +13,8 @@ import modules.memmon import modules.sd_models import modules.styles import modules.devices as devices -from modules import sd_samplers, hypernetwork +from modules import sd_samplers +from modules.hypernetwork import hypernetwork from modules.paths import models_path, script_path, sd_path sd_model_file = os.path.join(script_path, 'model.ckpt') @@ -29,6 +30,7 @@ parser.add_argument("--no-half-vae", action='store_true', help="do not switch th parser.add_argument("--no-progressbar-hiding", action='store_true', help="do not hide progressbar in gradio UI (we hide it because it slows down ML if you have hardware acceleration in browser)") parser.add_argument("--max-batch-count", type=int, default=16, help="maximum batch count value for the UI") parser.add_argument("--embeddings-dir", type=str, default=os.path.join(script_path, 'embeddings'), help="embeddings directory for textual inversion (default: embeddings)") +parser.add_argument("--hypernetwork-dir", type=str, default=os.path.join(models_path, 'hypernetworks'), help="hypernetwork directory") parser.add_argument("--allow-code", action='store_true', help="allow custom script execution from webui") parser.add_argument("--medvram", action='store_true', help="enable stable diffusion model optimizations for sacrificing a little speed for low VRM usage") parser.add_argument("--lowvram", action='store_true', help="enable stable diffusion model optimizations for sacrificing a lot of speed for very low VRM usage") @@ -82,10 +84,17 @@ parallel_processing_allowed = not cmd_opts.lowvram and not cmd_opts.medvram xformers_available = False config_filename = cmd_opts.ui_settings_file -hypernetworks = hypernetwork.list_hypernetworks(os.path.join(models_path, 'hypernetworks')) +hypernetworks = hypernetwork.list_hypernetworks(cmd_opts.hypernetwork_dir) loaded_hypernetwork = None +def reload_hypernetworks(): + global hypernetworks + + hypernetworks = hypernetwork.list_hypernetworks(cmd_opts.hypernetwork_dir) + hypernetwork.load_hypernetwork(opts.sd_hypernetwork) + + class State: skipped = False interrupted = False diff --git a/modules/textual_inversion/textual_inversion.py b/modules/textual_inversion/textual_inversion.py index 5965c5a0..d6977950 100644 --- a/modules/textual_inversion/textual_inversion.py +++ b/modules/textual_inversion/textual_inversion.py @@ -156,7 +156,7 @@ def create_embedding(name, num_vectors_per_token, init_text='*'): return fn -def train_embedding(embedding_name, learn_rate, data_root, log_directory, training_width, training_height, steps, num_repeats, create_image_every, save_embedding_every, template_file): +def train_embedding(embedding_name, learn_rate, data_root, log_directory, training_width, training_height, steps, num_repeats, create_image_every, save_embedding_every, template_file, preview_image_prompt): assert embedding_name, 'embedding not selected' shared.state.textinfo = "Initializing textual inversion training..." @@ -238,12 +238,14 @@ def train_embedding(embedding_name, learn_rate, data_root, log_directory, traini if embedding.step > 0 and images_dir is not None and embedding.step % create_image_every == 0: last_saved_image = os.path.join(images_dir, f'{embedding_name}-{embedding.step}.png') + preview_text = text if preview_image_prompt == "" else preview_image_prompt + p = processing.StableDiffusionProcessingTxt2Img( sd_model=shared.sd_model, - prompt=text, + prompt=preview_text, steps=20, - height=training_height, - width=training_width, + height=training_height, + width=training_width, do_not_save_grid=True, do_not_save_samples=True, ) @@ -254,7 +256,7 @@ def train_embedding(embedding_name, learn_rate, data_root, log_directory, traini shared.state.current_image = image image.save(last_saved_image) - last_saved_image += f", prompt: {text}" + last_saved_image += f", prompt: {preview_text}" shared.state.job_no = embedding.step diff --git a/modules/ui.py b/modules/ui.py index 10b1ee3a..df653059 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1023,7 +1023,7 @@ def create_ui(wrap_gradio_gpu_call): gr.HTML(value="") with gr.Column(): - create_embedding = gr.Button(value="Create", variant='primary') + create_embedding = gr.Button(value="Create embedding", variant='primary') with gr.Group(): gr.HTML(value="

Create a new hypernetwork

") @@ -1035,7 +1035,7 @@ def create_ui(wrap_gradio_gpu_call): gr.HTML(value="") with gr.Column(): - create_hypernetwork = gr.Button(value="Create", variant='primary') + create_hypernetwork = gr.Button(value="Create hypernetwork", variant='primary') with gr.Group(): gr.HTML(value="

Preprocess images

") @@ -1147,6 +1147,7 @@ def create_ui(wrap_gradio_gpu_call): create_image_every, save_embedding_every, template_file, + preview_image_prompt, ], outputs=[ ti_output, diff --git a/scripts/xy_grid.py b/scripts/xy_grid.py index 42e1489c..0af5993c 100644 --- a/scripts/xy_grid.py +++ b/scripts/xy_grid.py @@ -10,7 +10,8 @@ import numpy as np import modules.scripts as scripts import gradio as gr -from modules import images, hypernetwork +from modules import images +from modules.hypernetwork import hypernetwork from modules.processing import process_images, Processed, get_correct_sampler from modules.shared import opts, cmd_opts, state import modules.shared as shared diff --git a/webui.py b/webui.py index 7c200551..ba2156c8 100644 --- a/webui.py +++ b/webui.py @@ -29,6 +29,7 @@ from modules import devices from modules import modelloader from modules.paths import script_path from modules.shared import cmd_opts +import modules.hypernetwork.hypernetwork modelloader.cleanup_models() modules.sd_models.setup_model() @@ -77,22 +78,12 @@ def wrap_gradio_gpu_call(func, extra_outputs=None): return modules.ui.wrap_gradio_call(f, extra_outputs=extra_outputs) -def set_hypernetwork(): - shared.hypernetwork = shared.hypernetworks.get(shared.opts.sd_hypernetwork, None) - - -shared.reload_hypernetworks() -shared.opts.onchange("sd_hypernetwork", set_hypernetwork) -set_hypernetwork() - - modules.scripts.load_scripts(os.path.join(script_path, "scripts")) shared.sd_model = modules.sd_models.load_model() shared.opts.onchange("sd_model_checkpoint", wrap_queued_call(lambda: modules.sd_models.reload_model_weights(shared.sd_model))) -loaded_hypernetwork = modules.hypernetwork.load_hypernetwork(shared.opts.sd_hypernetwork) -shared.opts.onchange("sd_hypernetwork", wrap_queued_call(lambda: modules.hypernetwork.load_hypernetwork(shared.opts.sd_hypernetwork))) +shared.opts.onchange("sd_hypernetwork", wrap_queued_call(lambda: modules.hypernetwork.hypernetwork.load_hypernetwork(shared.opts.sd_hypernetwork))) def webui(): @@ -117,7 +108,7 @@ def webui(): prevent_thread_lock=True ) - app.add_middleware(GZipMiddleware,minimum_size=1000) + app.add_middleware(GZipMiddleware, minimum_size=1000) while 1: time.sleep(0.5) -- cgit v1.2.3 From 87b77cad5f3017c952a7dfec0e7904a9df5b72fd Mon Sep 17 00:00:00 2001 From: Ben <110583491+TheLastBen@users.noreply.github.com> Date: Mon, 10 Oct 2022 19:37:16 +0100 Subject: Layout fix --- modules/ui.py | 28 ++++++++++++++-------------- 1 file changed, 14 insertions(+), 14 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/ui.py b/modules/ui.py index df653059..de4cd7f2 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -550,15 +550,15 @@ def create_ui(wrap_gradio_gpu_call): button_id = "hidden_element" if shared.cmd_opts.hide_ui_dir_config else 'open_folder' open_txt2img_folder = gr.Button(folder_symbol, elem_id=button_id) - with gr.Row(): - do_make_zip = gr.Checkbox(label="Make Zip when Save?", value=False) + with gr.Row(): + do_make_zip = gr.Checkbox(label="Make Zip when Save?", value=False) - with gr.Row(): - download_files = gr.File(None, file_count="multiple", interactive=False, show_label=False, visible=False) + with gr.Row(): + download_files = gr.File(None, file_count="multiple", interactive=False, show_label=False, visible=False) - with gr.Group(): - html_info = gr.HTML() - generation_info = gr.Textbox(visible=False) + with gr.Group(): + html_info = gr.HTML() + generation_info = gr.Textbox(visible=False) connect_reuse_seed(seed, reuse_seed, generation_info, dummy_component, is_subseed=False) connect_reuse_seed(subseed, reuse_subseed, generation_info, dummy_component, is_subseed=True) @@ -738,15 +738,15 @@ def create_ui(wrap_gradio_gpu_call): button_id = "hidden_element" if shared.cmd_opts.hide_ui_dir_config else 'open_folder' open_img2img_folder = gr.Button(folder_symbol, elem_id=button_id) - with gr.Row(): - do_make_zip = gr.Checkbox(label="Make Zip when Save?", value=False) + with gr.Row(): + do_make_zip = gr.Checkbox(label="Make Zip when Save?", value=False) - with gr.Row(): - download_files = gr.File(None, file_count="multiple", interactive=False, show_label=False, visible=False) + with gr.Row(): + download_files = gr.File(None, file_count="multiple", interactive=False, show_label=False, visible=False) - with gr.Group(): - html_info = gr.HTML() - generation_info = gr.Textbox(visible=False) + with gr.Group(): + html_info = gr.HTML() + generation_info = gr.Textbox(visible=False) connect_reuse_seed(seed, reuse_seed, generation_info, dummy_component, is_subseed=False) connect_reuse_seed(subseed, reuse_subseed, generation_info, dummy_component, is_subseed=True) -- cgit v1.2.3 From 861297cefe2bb663f4e09dd4778a4cb93ebe8ff1 Mon Sep 17 00:00:00 2001 From: Ben <110583491+TheLastBen@users.noreply.github.com> Date: Tue, 11 Oct 2022 08:08:45 +0100 Subject: add a space holder --- modules/ui.py | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) (limited to 'modules/ui.py') diff --git a/modules/ui.py b/modules/ui.py index de4cd7f2..fc0f3d3c 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -429,7 +429,10 @@ def create_toprow(is_img2img): with gr.Row(): with gr.Column(scale=8): - negative_prompt = gr.Textbox(label="Negative prompt", elem_id="negative_prompt", show_label=False, placeholder="Negative prompt", lines=2) + with gr.Row(): + negative_prompt = gr.Textbox(label="Negative prompt", elem_id="negative_prompt", show_label=False, placeholder="Negative prompt", lines=2) + with gr.Column(scale=1, elem_id="roll_col"): + sh = gr.Button(elem_id="sh", visible=True) with gr.Column(scale=1, elem_id="style_neg_col"): prompt_style2 = gr.Dropdown(label="Style 2", elem_id=f"{id_part}_style2_index", choices=[k for k, v in shared.prompt_styles.styles.items()], value=next(iter(shared.prompt_styles.styles.keys())), visible=len(shared.prompt_styles.styles) > 1) -- cgit v1.2.3 From 59925644480b6fd84f6bb84b4df7d4fbc6a0cce8 Mon Sep 17 00:00:00 2001 From: JamnedZ Date: Tue, 11 Oct 2022 16:40:27 +0700 Subject: Cleaned ngrok integration --- modules/ngrok.py | 15 +++++++++++++++ modules/shared.py | 1 + modules/ui.py | 5 +++++ 3 files changed, 21 insertions(+) create mode 100644 modules/ngrok.py (limited to 'modules/ui.py') diff --git a/modules/ngrok.py b/modules/ngrok.py new file mode 100644 index 00000000..17e6976f --- /dev/null +++ b/modules/ngrok.py @@ -0,0 +1,15 @@ +from pyngrok import ngrok, conf, exception + + +def connect(token, port): + if token == None: + token = 'None' + conf.get_default().auth_token = token + try: + public_url = ngrok.connect(port).public_url + except exception.PyngrokNgrokError: + print(f'Invalid ngrok authtoken, ngrok connection aborted.\n' + f'Your token: {token}, get the right one on https://dashboard.ngrok.com/get-started/your-authtoken') + else: + print(f'ngrok connected to localhost:{port}! URL: {public_url}\n' + 'You can use this link after the launch is complete.') \ No newline at end of file diff --git a/modules/shared.py b/modules/shared.py index 8753015e..375e3afb 100644 --- a/modules/shared.py +++ b/modules/shared.py @@ -38,6 +38,7 @@ parser.add_argument("--always-batch-cond-uncond", action='store_true', help="dis parser.add_argument("--unload-gfpgan", action='store_true', help="does not do anything.") parser.add_argument("--precision", type=str, help="evaluate at this precision", choices=["full", "autocast"], default="autocast") parser.add_argument("--share", action='store_true', help="use share=True for gradio and make the UI accessible through their site (doesn't work for me but you might have better luck)") +parser.add_argument("--ngrok", type=str, help="ngrok authtoken, alternative to gradio --share", default=None) parser.add_argument("--codeformer-models-path", type=str, help="Path to directory with codeformer model file(s).", default=os.path.join(models_path, 'Codeformer')) parser.add_argument("--gfpgan-models-path", type=str, help="Path to directory with GFPGAN model file(s).", default=os.path.join(models_path, 'GFPGAN')) parser.add_argument("--esrgan-models-path", type=str, help="Path to directory with ESRGAN model file(s).", default=os.path.join(models_path, 'ESRGAN')) diff --git a/modules/ui.py b/modules/ui.py index fc0f3d3c..f57f32db 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -51,6 +51,11 @@ if not cmd_opts.share and not cmd_opts.listen: gradio.utils.version_check = lambda: None gradio.utils.get_local_ip_address = lambda: '127.0.0.1' +if cmd_opts.ngrok != None: + import modules.ngrok as ngrok + print('ngrok authtoken detected, trying to connect...') + ngrok.connect(cmd_opts.ngrok, cmd_opts.port if cmd_opts.port != None else 7860) + def gr_show(visible=True): return {"visible": visible, "__type__": "update"} -- cgit v1.2.3 From 873efeed49bb5197a42da18272115b326c5d68f3 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Tue, 11 Oct 2022 15:51:22 +0300 Subject: rename hypernetwork dir to hypernetworks to prevent clash with an old filename that people who use zip instead of git clone will have --- modules/hypernetwork/hypernetwork.py | 283 ---------------------------------- modules/hypernetwork/ui.py | 43 ------ modules/hypernetworks/hypernetwork.py | 283 ++++++++++++++++++++++++++++++++++ modules/hypernetworks/ui.py | 43 ++++++ modules/sd_hijack.py | 2 +- modules/sd_hijack_optimizations.py | 2 +- modules/shared.py | 2 +- modules/ui.py | 2 +- scripts/xy_grid.py | 2 +- webui.py | 2 +- 10 files changed, 332 insertions(+), 332 deletions(-) delete mode 100644 modules/hypernetwork/hypernetwork.py delete mode 100644 modules/hypernetwork/ui.py create mode 100644 modules/hypernetworks/hypernetwork.py create mode 100644 modules/hypernetworks/ui.py (limited to 'modules/ui.py') diff --git a/modules/hypernetwork/hypernetwork.py b/modules/hypernetwork/hypernetwork.py deleted file mode 100644 index aa701bda..00000000 --- a/modules/hypernetwork/hypernetwork.py +++ /dev/null @@ -1,283 +0,0 @@ -import datetime -import glob -import html -import os -import sys -import traceback -import tqdm - -import torch - -from ldm.util import default -from modules import devices, shared, processing, sd_models -import torch -from torch import einsum -from einops import rearrange, repeat -import modules.textual_inversion.dataset - - -class HypernetworkModule(torch.nn.Module): - def __init__(self, dim, state_dict=None): - super().__init__() - - self.linear1 = torch.nn.Linear(dim, dim * 2) - self.linear2 = torch.nn.Linear(dim * 2, dim) - - if state_dict is not None: - self.load_state_dict(state_dict, strict=True) - else: - - self.linear1.weight.data.normal_(mean=0.0, std=0.01) - self.linear1.bias.data.zero_() - self.linear2.weight.data.normal_(mean=0.0, std=0.01) - self.linear2.bias.data.zero_() - - self.to(devices.device) - - def forward(self, x): - return x + (self.linear2(self.linear1(x))) - - -class Hypernetwork: - filename = None - name = None - - def __init__(self, name=None): - self.filename = None - self.name = name - self.layers = {} - self.step = 0 - self.sd_checkpoint = None - self.sd_checkpoint_name = None - - for size in [320, 640, 768, 1280]: - self.layers[size] = (HypernetworkModule(size), HypernetworkModule(size)) - - def weights(self): - res = [] - - for k, layers in self.layers.items(): - for layer in layers: - layer.train() - res += [layer.linear1.weight, layer.linear1.bias, layer.linear2.weight, layer.linear2.bias] - - return res - - def save(self, filename): - state_dict = {} - - for k, v in self.layers.items(): - state_dict[k] = (v[0].state_dict(), v[1].state_dict()) - - state_dict['step'] = self.step - state_dict['name'] = self.name - state_dict['sd_checkpoint'] = self.sd_checkpoint - state_dict['sd_checkpoint_name'] = self.sd_checkpoint_name - - torch.save(state_dict, filename) - - def load(self, filename): - self.filename = filename - if self.name is None: - self.name = os.path.splitext(os.path.basename(filename))[0] - - state_dict = torch.load(filename, map_location='cpu') - - for size, sd in state_dict.items(): - if type(size) == int: - self.layers[size] = (HypernetworkModule(size, sd[0]), HypernetworkModule(size, sd[1])) - - self.name = state_dict.get('name', self.name) - self.step = state_dict.get('step', 0) - self.sd_checkpoint = state_dict.get('sd_checkpoint', None) - self.sd_checkpoint_name = state_dict.get('sd_checkpoint_name', None) - - -def list_hypernetworks(path): - res = {} - for filename in glob.iglob(os.path.join(path, '**/*.pt'), recursive=True): - name = os.path.splitext(os.path.basename(filename))[0] - res[name] = filename - return res - - -def load_hypernetwork(filename): - path = shared.hypernetworks.get(filename, None) - if path is not None: - print(f"Loading hypernetwork {filename}") - try: - shared.loaded_hypernetwork = Hypernetwork() - shared.loaded_hypernetwork.load(path) - - except Exception: - print(f"Error loading hypernetwork {path}", file=sys.stderr) - print(traceback.format_exc(), file=sys.stderr) - else: - if shared.loaded_hypernetwork is not None: - print(f"Unloading hypernetwork") - - shared.loaded_hypernetwork = None - - -def apply_hypernetwork(hypernetwork, context, layer=None): - hypernetwork_layers = (hypernetwork.layers if hypernetwork is not None else {}).get(context.shape[2], None) - - if hypernetwork_layers is None: - return context, context - - if layer is not None: - layer.hyper_k = hypernetwork_layers[0] - layer.hyper_v = hypernetwork_layers[1] - - context_k = hypernetwork_layers[0](context) - context_v = hypernetwork_layers[1](context) - return context_k, context_v - - -def attention_CrossAttention_forward(self, x, context=None, mask=None): - h = self.heads - - q = self.to_q(x) - context = default(context, x) - - context_k, context_v = apply_hypernetwork(shared.loaded_hypernetwork, context, self) - k = self.to_k(context_k) - v = self.to_v(context_v) - - q, k, v = map(lambda t: rearrange(t, 'b n (h d) -> (b h) n d', h=h), (q, k, v)) - - sim = einsum('b i d, b j d -> b i j', q, k) * self.scale - - if mask is not None: - mask = rearrange(mask, 'b ... -> b (...)') - max_neg_value = -torch.finfo(sim.dtype).max - mask = repeat(mask, 'b j -> (b h) () j', h=h) - sim.masked_fill_(~mask, max_neg_value) - - # attention, what we cannot get enough of - attn = sim.softmax(dim=-1) - - out = einsum('b i j, b j d -> b i d', attn, v) - out = rearrange(out, '(b h) n d -> b n (h d)', h=h) - return self.to_out(out) - - -def train_hypernetwork(hypernetwork_name, learn_rate, data_root, log_directory, steps, create_image_every, save_hypernetwork_every, template_file, preview_image_prompt): - assert hypernetwork_name, 'embedding not selected' - - path = shared.hypernetworks.get(hypernetwork_name, None) - shared.loaded_hypernetwork = Hypernetwork() - shared.loaded_hypernetwork.load(path) - - shared.state.textinfo = "Initializing hypernetwork training..." - shared.state.job_count = steps - - filename = os.path.join(shared.cmd_opts.hypernetwork_dir, f'{hypernetwork_name}.pt') - - log_directory = os.path.join(log_directory, datetime.datetime.now().strftime("%Y-%m-%d"), hypernetwork_name) - - if save_hypernetwork_every > 0: - hypernetwork_dir = os.path.join(log_directory, "hypernetworks") - os.makedirs(hypernetwork_dir, exist_ok=True) - else: - hypernetwork_dir = None - - if create_image_every > 0: - images_dir = os.path.join(log_directory, "images") - os.makedirs(images_dir, exist_ok=True) - else: - images_dir = None - - cond_model = shared.sd_model.cond_stage_model - - shared.state.textinfo = f"Preparing dataset from {html.escape(data_root)}..." - with torch.autocast("cuda"): - ds = modules.textual_inversion.dataset.PersonalizedBase(data_root=data_root, width=512, height=512, repeats=1, placeholder_token=hypernetwork_name, model=shared.sd_model, device=devices.device, template_file=template_file) - - hypernetwork = shared.loaded_hypernetwork - weights = hypernetwork.weights() - for weight in weights: - weight.requires_grad = True - - optimizer = torch.optim.AdamW(weights, lr=learn_rate) - - losses = torch.zeros((32,)) - - last_saved_file = "" - last_saved_image = "" - - ititial_step = hypernetwork.step or 0 - if ititial_step > steps: - return hypernetwork, filename - - pbar = tqdm.tqdm(enumerate(ds), total=steps - ititial_step) - for i, (x, text) in pbar: - hypernetwork.step = i + ititial_step - - if hypernetwork.step > steps: - break - - if shared.state.interrupted: - break - - with torch.autocast("cuda"): - c = cond_model([text]) - - x = x.to(devices.device) - loss = shared.sd_model(x.unsqueeze(0), c)[0] - del x - - losses[hypernetwork.step % losses.shape[0]] = loss.item() - - optimizer.zero_grad() - loss.backward() - optimizer.step() - - pbar.set_description(f"loss: {losses.mean():.7f}") - - if hypernetwork.step > 0 and hypernetwork_dir is not None and hypernetwork.step % save_hypernetwork_every == 0: - last_saved_file = os.path.join(hypernetwork_dir, f'{hypernetwork_name}-{hypernetwork.step}.pt') - hypernetwork.save(last_saved_file) - - if hypernetwork.step > 0 and images_dir is not None and hypernetwork.step % create_image_every == 0: - last_saved_image = os.path.join(images_dir, f'{hypernetwork_name}-{hypernetwork.step}.png') - - preview_text = text if preview_image_prompt == "" else preview_image_prompt - - p = processing.StableDiffusionProcessingTxt2Img( - sd_model=shared.sd_model, - prompt=preview_text, - steps=20, - do_not_save_grid=True, - do_not_save_samples=True, - ) - - processed = processing.process_images(p) - image = processed.images[0] - - shared.state.current_image = image - image.save(last_saved_image) - - last_saved_image += f", prompt: {preview_text}" - - shared.state.job_no = hypernetwork.step - - shared.state.textinfo = f""" -

-Loss: {losses.mean():.7f}
-Step: {hypernetwork.step}
-Last prompt: {html.escape(text)}
-Last saved embedding: {html.escape(last_saved_file)}
-Last saved image: {html.escape(last_saved_image)}
-

-""" - - checkpoint = sd_models.select_checkpoint() - - hypernetwork.sd_checkpoint = checkpoint.hash - hypernetwork.sd_checkpoint_name = checkpoint.model_name - hypernetwork.save(filename) - - return hypernetwork, filename - - diff --git a/modules/hypernetwork/ui.py b/modules/hypernetwork/ui.py deleted file mode 100644 index f6d1d0a3..00000000 --- a/modules/hypernetwork/ui.py +++ /dev/null @@ -1,43 +0,0 @@ -import html -import os - -import gradio as gr - -import modules.textual_inversion.textual_inversion -import modules.textual_inversion.preprocess -from modules import sd_hijack, shared -from modules.hypernetwork import hypernetwork - - -def create_hypernetwork(name): - fn = os.path.join(shared.cmd_opts.hypernetwork_dir, f"{name}.pt") - assert not os.path.exists(fn), f"file {fn} already exists" - - hypernet = modules.hypernetwork.hypernetwork.Hypernetwork(name=name) - hypernet.save(fn) - - shared.reload_hypernetworks() - - return gr.Dropdown.update(choices=sorted([x for x in shared.hypernetworks.keys()])), f"Created: {fn}", "" - - -def train_hypernetwork(*args): - - initial_hypernetwork = shared.loaded_hypernetwork - - try: - sd_hijack.undo_optimizations() - - hypernetwork, filename = modules.hypernetwork.hypernetwork.train_hypernetwork(*args) - - res = f""" -Training {'interrupted' if shared.state.interrupted else 'finished'} at {hypernetwork.step} steps. -Hypernetwork saved to {html.escape(filename)} -""" - return res, "" - except Exception: - raise - finally: - shared.loaded_hypernetwork = initial_hypernetwork - sd_hijack.apply_optimizations() - diff --git a/modules/hypernetworks/hypernetwork.py b/modules/hypernetworks/hypernetwork.py new file mode 100644 index 00000000..aa701bda --- /dev/null +++ b/modules/hypernetworks/hypernetwork.py @@ -0,0 +1,283 @@ +import datetime +import glob +import html +import os +import sys +import traceback +import tqdm + +import torch + +from ldm.util import default +from modules import devices, shared, processing, sd_models +import torch +from torch import einsum +from einops import rearrange, repeat +import modules.textual_inversion.dataset + + +class HypernetworkModule(torch.nn.Module): + def __init__(self, dim, state_dict=None): + super().__init__() + + self.linear1 = torch.nn.Linear(dim, dim * 2) + self.linear2 = torch.nn.Linear(dim * 2, dim) + + if state_dict is not None: + self.load_state_dict(state_dict, strict=True) + else: + + self.linear1.weight.data.normal_(mean=0.0, std=0.01) + self.linear1.bias.data.zero_() + self.linear2.weight.data.normal_(mean=0.0, std=0.01) + self.linear2.bias.data.zero_() + + self.to(devices.device) + + def forward(self, x): + return x + (self.linear2(self.linear1(x))) + + +class Hypernetwork: + filename = None + name = None + + def __init__(self, name=None): + self.filename = None + self.name = name + self.layers = {} + self.step = 0 + self.sd_checkpoint = None + self.sd_checkpoint_name = None + + for size in [320, 640, 768, 1280]: + self.layers[size] = (HypernetworkModule(size), HypernetworkModule(size)) + + def weights(self): + res = [] + + for k, layers in self.layers.items(): + for layer in layers: + layer.train() + res += [layer.linear1.weight, layer.linear1.bias, layer.linear2.weight, layer.linear2.bias] + + return res + + def save(self, filename): + state_dict = {} + + for k, v in self.layers.items(): + state_dict[k] = (v[0].state_dict(), v[1].state_dict()) + + state_dict['step'] = self.step + state_dict['name'] = self.name + state_dict['sd_checkpoint'] = self.sd_checkpoint + state_dict['sd_checkpoint_name'] = self.sd_checkpoint_name + + torch.save(state_dict, filename) + + def load(self, filename): + self.filename = filename + if self.name is None: + self.name = os.path.splitext(os.path.basename(filename))[0] + + state_dict = torch.load(filename, map_location='cpu') + + for size, sd in state_dict.items(): + if type(size) == int: + self.layers[size] = (HypernetworkModule(size, sd[0]), HypernetworkModule(size, sd[1])) + + self.name = state_dict.get('name', self.name) + self.step = state_dict.get('step', 0) + self.sd_checkpoint = state_dict.get('sd_checkpoint', None) + self.sd_checkpoint_name = state_dict.get('sd_checkpoint_name', None) + + +def list_hypernetworks(path): + res = {} + for filename in glob.iglob(os.path.join(path, '**/*.pt'), recursive=True): + name = os.path.splitext(os.path.basename(filename))[0] + res[name] = filename + return res + + +def load_hypernetwork(filename): + path = shared.hypernetworks.get(filename, None) + if path is not None: + print(f"Loading hypernetwork {filename}") + try: + shared.loaded_hypernetwork = Hypernetwork() + shared.loaded_hypernetwork.load(path) + + except Exception: + print(f"Error loading hypernetwork {path}", file=sys.stderr) + print(traceback.format_exc(), file=sys.stderr) + else: + if shared.loaded_hypernetwork is not None: + print(f"Unloading hypernetwork") + + shared.loaded_hypernetwork = None + + +def apply_hypernetwork(hypernetwork, context, layer=None): + hypernetwork_layers = (hypernetwork.layers if hypernetwork is not None else {}).get(context.shape[2], None) + + if hypernetwork_layers is None: + return context, context + + if layer is not None: + layer.hyper_k = hypernetwork_layers[0] + layer.hyper_v = hypernetwork_layers[1] + + context_k = hypernetwork_layers[0](context) + context_v = hypernetwork_layers[1](context) + return context_k, context_v + + +def attention_CrossAttention_forward(self, x, context=None, mask=None): + h = self.heads + + q = self.to_q(x) + context = default(context, x) + + context_k, context_v = apply_hypernetwork(shared.loaded_hypernetwork, context, self) + k = self.to_k(context_k) + v = self.to_v(context_v) + + q, k, v = map(lambda t: rearrange(t, 'b n (h d) -> (b h) n d', h=h), (q, k, v)) + + sim = einsum('b i d, b j d -> b i j', q, k) * self.scale + + if mask is not None: + mask = rearrange(mask, 'b ... -> b (...)') + max_neg_value = -torch.finfo(sim.dtype).max + mask = repeat(mask, 'b j -> (b h) () j', h=h) + sim.masked_fill_(~mask, max_neg_value) + + # attention, what we cannot get enough of + attn = sim.softmax(dim=-1) + + out = einsum('b i j, b j d -> b i d', attn, v) + out = rearrange(out, '(b h) n d -> b n (h d)', h=h) + return self.to_out(out) + + +def train_hypernetwork(hypernetwork_name, learn_rate, data_root, log_directory, steps, create_image_every, save_hypernetwork_every, template_file, preview_image_prompt): + assert hypernetwork_name, 'embedding not selected' + + path = shared.hypernetworks.get(hypernetwork_name, None) + shared.loaded_hypernetwork = Hypernetwork() + shared.loaded_hypernetwork.load(path) + + shared.state.textinfo = "Initializing hypernetwork training..." + shared.state.job_count = steps + + filename = os.path.join(shared.cmd_opts.hypernetwork_dir, f'{hypernetwork_name}.pt') + + log_directory = os.path.join(log_directory, datetime.datetime.now().strftime("%Y-%m-%d"), hypernetwork_name) + + if save_hypernetwork_every > 0: + hypernetwork_dir = os.path.join(log_directory, "hypernetworks") + os.makedirs(hypernetwork_dir, exist_ok=True) + else: + hypernetwork_dir = None + + if create_image_every > 0: + images_dir = os.path.join(log_directory, "images") + os.makedirs(images_dir, exist_ok=True) + else: + images_dir = None + + cond_model = shared.sd_model.cond_stage_model + + shared.state.textinfo = f"Preparing dataset from {html.escape(data_root)}..." + with torch.autocast("cuda"): + ds = modules.textual_inversion.dataset.PersonalizedBase(data_root=data_root, width=512, height=512, repeats=1, placeholder_token=hypernetwork_name, model=shared.sd_model, device=devices.device, template_file=template_file) + + hypernetwork = shared.loaded_hypernetwork + weights = hypernetwork.weights() + for weight in weights: + weight.requires_grad = True + + optimizer = torch.optim.AdamW(weights, lr=learn_rate) + + losses = torch.zeros((32,)) + + last_saved_file = "" + last_saved_image = "" + + ititial_step = hypernetwork.step or 0 + if ititial_step > steps: + return hypernetwork, filename + + pbar = tqdm.tqdm(enumerate(ds), total=steps - ititial_step) + for i, (x, text) in pbar: + hypernetwork.step = i + ititial_step + + if hypernetwork.step > steps: + break + + if shared.state.interrupted: + break + + with torch.autocast("cuda"): + c = cond_model([text]) + + x = x.to(devices.device) + loss = shared.sd_model(x.unsqueeze(0), c)[0] + del x + + losses[hypernetwork.step % losses.shape[0]] = loss.item() + + optimizer.zero_grad() + loss.backward() + optimizer.step() + + pbar.set_description(f"loss: {losses.mean():.7f}") + + if hypernetwork.step > 0 and hypernetwork_dir is not None and hypernetwork.step % save_hypernetwork_every == 0: + last_saved_file = os.path.join(hypernetwork_dir, f'{hypernetwork_name}-{hypernetwork.step}.pt') + hypernetwork.save(last_saved_file) + + if hypernetwork.step > 0 and images_dir is not None and hypernetwork.step % create_image_every == 0: + last_saved_image = os.path.join(images_dir, f'{hypernetwork_name}-{hypernetwork.step}.png') + + preview_text = text if preview_image_prompt == "" else preview_image_prompt + + p = processing.StableDiffusionProcessingTxt2Img( + sd_model=shared.sd_model, + prompt=preview_text, + steps=20, + do_not_save_grid=True, + do_not_save_samples=True, + ) + + processed = processing.process_images(p) + image = processed.images[0] + + shared.state.current_image = image + image.save(last_saved_image) + + last_saved_image += f", prompt: {preview_text}" + + shared.state.job_no = hypernetwork.step + + shared.state.textinfo = f""" +

+Loss: {losses.mean():.7f}
+Step: {hypernetwork.step}
+Last prompt: {html.escape(text)}
+Last saved embedding: {html.escape(last_saved_file)}
+Last saved image: {html.escape(last_saved_image)}
+

+""" + + checkpoint = sd_models.select_checkpoint() + + hypernetwork.sd_checkpoint = checkpoint.hash + hypernetwork.sd_checkpoint_name = checkpoint.model_name + hypernetwork.save(filename) + + return hypernetwork, filename + + diff --git a/modules/hypernetworks/ui.py b/modules/hypernetworks/ui.py new file mode 100644 index 00000000..811bc31e --- /dev/null +++ b/modules/hypernetworks/ui.py @@ -0,0 +1,43 @@ +import html +import os + +import gradio as gr + +import modules.textual_inversion.textual_inversion +import modules.textual_inversion.preprocess +from modules import sd_hijack, shared +from modules.hypernetworks import hypernetwork + + +def create_hypernetwork(name): + fn = os.path.join(shared.cmd_opts.hypernetwork_dir, f"{name}.pt") + assert not os.path.exists(fn), f"file {fn} already exists" + + hypernet = modules.hypernetwork.hypernetwork.Hypernetwork(name=name) + hypernet.save(fn) + + shared.reload_hypernetworks() + + return gr.Dropdown.update(choices=sorted([x for x in shared.hypernetworks.keys()])), f"Created: {fn}", "" + + +def train_hypernetwork(*args): + + initial_hypernetwork = shared.loaded_hypernetwork + + try: + sd_hijack.undo_optimizations() + + hypernetwork, filename = modules.hypernetwork.hypernetwork.train_hypernetwork(*args) + + res = f""" +Training {'interrupted' if shared.state.interrupted else 'finished'} at {hypernetwork.step} steps. +Hypernetwork saved to {html.escape(filename)} +""" + return res, "" + except Exception: + raise + finally: + shared.loaded_hypernetwork = initial_hypernetwork + sd_hijack.apply_optimizations() + diff --git a/modules/sd_hijack.py b/modules/sd_hijack.py index f873049a..f07ec041 100644 --- a/modules/sd_hijack.py +++ b/modules/sd_hijack.py @@ -37,7 +37,7 @@ def apply_optimizations(): def undo_optimizations(): - from modules.hypernetwork import hypernetwork + from modules.hypernetworks import hypernetwork ldm.modules.attention.CrossAttention.forward = hypernetwork.attention_CrossAttention_forward ldm.modules.diffusionmodules.model.nonlinearity = diffusionmodules_model_nonlinearity diff --git a/modules/sd_hijack_optimizations.py b/modules/sd_hijack_optimizations.py index 27e571fc..3349b9c3 100644 --- a/modules/sd_hijack_optimizations.py +++ b/modules/sd_hijack_optimizations.py @@ -9,7 +9,7 @@ from ldm.util import default from einops import rearrange from modules import shared -from modules.hypernetwork import hypernetwork +from modules.hypernetworks import hypernetwork if shared.cmd_opts.xformers or shared.cmd_opts.force_enable_xformers: diff --git a/modules/shared.py b/modules/shared.py index 375e3afb..1dc2ccf2 100644 --- a/modules/shared.py +++ b/modules/shared.py @@ -14,7 +14,7 @@ import modules.sd_models import modules.styles import modules.devices as devices from modules import sd_samplers -from modules.hypernetwork import hypernetwork +from modules.hypernetworks import hypernetwork from modules.paths import models_path, script_path, sd_path sd_model_file = os.path.join(script_path, 'model.ckpt') diff --git a/modules/ui.py b/modules/ui.py index f57f32db..42e5d866 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -39,7 +39,7 @@ import modules.generation_parameters_copypaste from modules import prompt_parser from modules.images import save_image import modules.textual_inversion.ui -import modules.hypernetwork.ui +import modules.hypernetworks.ui # this is a fix for Windows users. Without it, javascript files will be served with text/html content-type and the browser will not show any UI mimetypes.init() diff --git a/scripts/xy_grid.py b/scripts/xy_grid.py index 16918c99..cddb192a 100644 --- a/scripts/xy_grid.py +++ b/scripts/xy_grid.py @@ -11,7 +11,7 @@ import modules.scripts as scripts import gradio as gr from modules import images -from modules.hypernetwork import hypernetwork +from modules.hypernetworks import hypernetwork from modules.processing import process_images, Processed, get_correct_sampler from modules.shared import opts, cmd_opts, state import modules.shared as shared diff --git a/webui.py b/webui.py index ba2156c8..faa38a0d 100644 --- a/webui.py +++ b/webui.py @@ -29,7 +29,7 @@ from modules import devices from modules import modelloader from modules.paths import script_path from modules.shared import cmd_opts -import modules.hypernetwork.hypernetwork +import modules.hypernetworks.hypernetwork modelloader.cleanup_models() modules.sd_models.setup_model() -- cgit v1.2.3 From b0583be0884cd17dafb408fd79b52b2a0a972563 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Tue, 11 Oct 2022 15:54:34 +0300 Subject: more renames --- modules/hypernetworks/ui.py | 4 ++-- modules/ui.py | 4 ++-- webui.py | 2 +- 3 files changed, 5 insertions(+), 5 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/hypernetworks/ui.py b/modules/hypernetworks/ui.py index 811bc31e..e7540f41 100644 --- a/modules/hypernetworks/ui.py +++ b/modules/hypernetworks/ui.py @@ -13,7 +13,7 @@ def create_hypernetwork(name): fn = os.path.join(shared.cmd_opts.hypernetwork_dir, f"{name}.pt") assert not os.path.exists(fn), f"file {fn} already exists" - hypernet = modules.hypernetwork.hypernetwork.Hypernetwork(name=name) + hypernet = modules.hypernetworks.hypernetwork.Hypernetwork(name=name) hypernet.save(fn) shared.reload_hypernetworks() @@ -28,7 +28,7 @@ def train_hypernetwork(*args): try: sd_hijack.undo_optimizations() - hypernetwork, filename = modules.hypernetwork.hypernetwork.train_hypernetwork(*args) + hypernetwork, filename = modules.hypernetworks.hypernetwork.train_hypernetwork(*args) res = f""" Training {'interrupted' if shared.state.interrupted else 'finished'} at {hypernetwork.step} steps. diff --git a/modules/ui.py b/modules/ui.py index 42e5d866..ee333c3b 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1111,7 +1111,7 @@ def create_ui(wrap_gradio_gpu_call): ) create_hypernetwork.click( - fn=modules.hypernetwork.ui.create_hypernetwork, + fn=modules.hypernetworks.ui.create_hypernetwork, inputs=[ new_hypernetwork_name, ], @@ -1164,7 +1164,7 @@ def create_ui(wrap_gradio_gpu_call): ) train_hypernetwork.click( - fn=wrap_gradio_gpu_call(modules.hypernetwork.ui.train_hypernetwork, extra_outputs=[gr.update()]), + fn=wrap_gradio_gpu_call(modules.hypernetworks.ui.train_hypernetwork, extra_outputs=[gr.update()]), _js="start_training_textual_inversion", inputs=[ train_hypernetwork_name, diff --git a/webui.py b/webui.py index faa38a0d..338f58e1 100644 --- a/webui.py +++ b/webui.py @@ -83,7 +83,7 @@ modules.scripts.load_scripts(os.path.join(script_path, "scripts")) shared.sd_model = modules.sd_models.load_model() shared.opts.onchange("sd_model_checkpoint", wrap_queued_call(lambda: modules.sd_models.reload_model_weights(shared.sd_model))) -shared.opts.onchange("sd_hypernetwork", wrap_queued_call(lambda: modules.hypernetwork.hypernetwork.load_hypernetwork(shared.opts.sd_hypernetwork))) +shared.opts.onchange("sd_hypernetwork", wrap_queued_call(lambda: modules.hypernetworks.hypernetwork.load_hypernetwork(shared.opts.sd_hypernetwork))) def webui(): -- cgit v1.2.3 From d01a2d01560b31937df1f3433d210c18f97d32fa Mon Sep 17 00:00:00 2001 From: papuSpartan Date: Tue, 11 Oct 2022 08:03:31 -0500 Subject: move list refresh to webui.py and add stdout indicating it's doing so --- modules/ui.py | 3 --- webui.py | 2 ++ 2 files changed, 2 insertions(+), 3 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/ui.py b/modules/ui.py index 06ff118f..ae9317a3 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -39,7 +39,6 @@ import modules.generation_parameters_copypaste from modules import prompt_parser from modules.images import save_image import modules.textual_inversion.ui -from modules.sd_models import list_models # this is a fix for Windows users. Without it, javascript files will be served with text/html content-type and the browser will not show any UI mimetypes.init() @@ -1291,8 +1290,6 @@ Requested path was: {f} shared.state.interrupt() settings_interface.gradio_ref.do_restart = True - # refresh models so that new models/.ckpt's show up on reload - list_models() restart_gradio.click( fn=request_restart, diff --git a/webui.py b/webui.py index 270584f7..94098c4c 100644 --- a/webui.py +++ b/webui.py @@ -124,6 +124,8 @@ def webui(): modules.scripts.reload_scripts(os.path.join(script_path, "scripts")) print('Reloading modules: modules.ui') importlib.reload(modules.ui) + print('Refreshing Model List') + modules.sd_models.list_models() print('Restarting Gradio') -- cgit v1.2.3 From d682444ecc99319fbd2b142a12727501e2884ba7 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Tue, 11 Oct 2022 18:04:47 +0300 Subject: add option to select hypernetwork modules when creating --- modules/hypernetworks/hypernetwork.py | 4 ++-- modules/hypernetworks/ui.py | 4 ++-- modules/ui.py | 2 ++ 3 files changed, 6 insertions(+), 4 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/hypernetworks/hypernetwork.py b/modules/hypernetworks/hypernetwork.py index aa701bda..b081f14e 100644 --- a/modules/hypernetworks/hypernetwork.py +++ b/modules/hypernetworks/hypernetwork.py @@ -42,7 +42,7 @@ class Hypernetwork: filename = None name = None - def __init__(self, name=None): + def __init__(self, name=None, enable_sizes=None): self.filename = None self.name = name self.layers = {} @@ -50,7 +50,7 @@ class Hypernetwork: self.sd_checkpoint = None self.sd_checkpoint_name = None - for size in [320, 640, 768, 1280]: + for size in enable_sizes or [320, 640, 768, 1280]: self.layers[size] = (HypernetworkModule(size), HypernetworkModule(size)) def weights(self): diff --git a/modules/hypernetworks/ui.py b/modules/hypernetworks/ui.py index e7540f41..cdddcce1 100644 --- a/modules/hypernetworks/ui.py +++ b/modules/hypernetworks/ui.py @@ -9,11 +9,11 @@ from modules import sd_hijack, shared from modules.hypernetworks import hypernetwork -def create_hypernetwork(name): +def create_hypernetwork(name, enable_sizes): fn = os.path.join(shared.cmd_opts.hypernetwork_dir, f"{name}.pt") assert not os.path.exists(fn), f"file {fn} already exists" - hypernet = modules.hypernetworks.hypernetwork.Hypernetwork(name=name) + hypernet = modules.hypernetworks.hypernetwork.Hypernetwork(name=name, enable_sizes=[int(x) for x in enable_sizes]) hypernet.save(fn) shared.reload_hypernetworks() diff --git a/modules/ui.py b/modules/ui.py index f2d16b12..14b87b92 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1037,6 +1037,7 @@ def create_ui(wrap_gradio_gpu_call): gr.HTML(value="

Create a new hypernetwork

") new_hypernetwork_name = gr.Textbox(label="Name") + new_hypernetwork_sizes = gr.CheckboxGroup(label="Modules", value=["768", "320", "640", "1280"], choices=["768", "320", "640", "1280"]) with gr.Row(): with gr.Column(scale=3): @@ -1114,6 +1115,7 @@ def create_ui(wrap_gradio_gpu_call): fn=modules.hypernetworks.ui.create_hypernetwork, inputs=[ new_hypernetwork_name, + new_hypernetwork_sizes, ], outputs=[ train_hypernetwork_name, -- cgit v1.2.3 From d6fcc6b87bc00fcdecea276fe5b7c7945f7a8b14 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Tue, 11 Oct 2022 22:03:05 +0300 Subject: apply lr schedule to hypernets --- modules/hypernetworks/hypernetwork.py | 19 ++++++++--- modules/textual_inversion/learn_schedule.py | 34 ++++++++++++++++++++ modules/textual_inversion/textual_inversion.py | 44 +++----------------------- modules/ui.py | 2 +- 4 files changed, 54 insertions(+), 45 deletions(-) create mode 100644 modules/textual_inversion/learn_schedule.py (limited to 'modules/ui.py') diff --git a/modules/hypernetworks/hypernetwork.py b/modules/hypernetworks/hypernetwork.py index 5608e799..470659df 100644 --- a/modules/hypernetworks/hypernetwork.py +++ b/modules/hypernetworks/hypernetwork.py @@ -14,6 +14,7 @@ import torch from torch import einsum from einops import rearrange, repeat import modules.textual_inversion.dataset +from modules.textual_inversion.learn_schedule import LearnSchedule class HypernetworkModule(torch.nn.Module): @@ -202,8 +203,6 @@ def train_hypernetwork(hypernetwork_name, learn_rate, data_root, log_directory, for weight in weights: weight.requires_grad = True - optimizer = torch.optim.AdamW(weights, lr=learn_rate) - losses = torch.zeros((32,)) last_saved_file = "" @@ -213,12 +212,24 @@ def train_hypernetwork(hypernetwork_name, learn_rate, data_root, log_directory, if ititial_step > steps: return hypernetwork, filename + schedules = iter(LearnSchedule(learn_rate, steps, ititial_step)) + (learn_rate, end_step) = next(schedules) + print(f'Training at rate of {learn_rate} until step {end_step}') + + optimizer = torch.optim.AdamW(weights, lr=learn_rate) + pbar = tqdm.tqdm(enumerate(ds), total=steps - ititial_step) for i, (x, text, cond) in pbar: hypernetwork.step = i + ititial_step - if hypernetwork.step > steps: - break + if hypernetwork.step > end_step: + try: + (learn_rate, end_step) = next(schedules) + except Exception: + break + tqdm.tqdm.write(f'Training at rate of {learn_rate} until step {end_step}') + for pg in optimizer.param_groups: + pg['lr'] = learn_rate if shared.state.interrupted: break diff --git a/modules/textual_inversion/learn_schedule.py b/modules/textual_inversion/learn_schedule.py new file mode 100644 index 00000000..db720271 --- /dev/null +++ b/modules/textual_inversion/learn_schedule.py @@ -0,0 +1,34 @@ + +class LearnSchedule: + def __init__(self, learn_rate, max_steps, cur_step=0): + pairs = learn_rate.split(',') + self.rates = [] + self.it = 0 + self.maxit = 0 + for i, pair in enumerate(pairs): + tmp = pair.split(':') + if len(tmp) == 2: + step = int(tmp[1]) + if step > cur_step: + self.rates.append((float(tmp[0]), min(step, max_steps))) + self.maxit += 1 + if step > max_steps: + return + elif step == -1: + self.rates.append((float(tmp[0]), max_steps)) + self.maxit += 1 + return + else: + self.rates.append((float(tmp[0]), max_steps)) + self.maxit += 1 + return + + def __iter__(self): + return self + + def __next__(self): + if self.it < self.maxit: + self.it += 1 + return self.rates[self.it - 1] + else: + raise StopIteration diff --git a/modules/textual_inversion/textual_inversion.py b/modules/textual_inversion/textual_inversion.py index 47a27faf..7717837d 100644 --- a/modules/textual_inversion/textual_inversion.py +++ b/modules/textual_inversion/textual_inversion.py @@ -10,6 +10,7 @@ import datetime from modules import shared, devices, sd_hijack, processing, sd_models import modules.textual_inversion.dataset +from modules.textual_inversion.learn_schedule import LearnSchedule class Embedding: @@ -198,11 +199,8 @@ def train_embedding(embedding_name, learn_rate, data_root, log_directory, traini if ititial_step > steps: return embedding, filename - tr_img_len = len([os.path.join(data_root, file_path) for file_path in os.listdir(data_root)]) - epoch_len = (tr_img_len * num_repeats) + tr_img_len - - scheduleIter = iter(LearnSchedule(learn_rate, steps, ititial_step)) - (learn_rate, end_step) = next(scheduleIter) + schedules = iter(LearnSchedule(learn_rate, steps, ititial_step)) + (learn_rate, end_step) = next(schedules) print(f'Training at rate of {learn_rate} until step {end_step}') optimizer = torch.optim.AdamW([embedding.vec], lr=learn_rate) @@ -213,7 +211,7 @@ def train_embedding(embedding_name, learn_rate, data_root, log_directory, traini if embedding.step > end_step: try: - (learn_rate, end_step) = next(scheduleIter) + (learn_rate, end_step) = next(schedules) except: break tqdm.tqdm.write(f'Training at rate of {learn_rate} until step {end_step}') @@ -288,37 +286,3 @@ Last saved image: {html.escape(last_saved_image)}
embedding.save(filename) return embedding, filename - -class LearnSchedule: - def __init__(self, learn_rate, max_steps, cur_step=0): - pairs = learn_rate.split(',') - self.rates = [] - self.it = 0 - self.maxit = 0 - for i, pair in enumerate(pairs): - tmp = pair.split(':') - if len(tmp) == 2: - step = int(tmp[1]) - if step > cur_step: - self.rates.append((float(tmp[0]), min(step, max_steps))) - self.maxit += 1 - if step > max_steps: - return - elif step == -1: - self.rates.append((float(tmp[0]), max_steps)) - self.maxit += 1 - return - else: - self.rates.append((float(tmp[0]), max_steps)) - self.maxit += 1 - return - - def __iter__(self): - return self - - def __next__(self): - if self.it < self.maxit: - self.it += 1 - return self.rates[self.it - 1] - else: - raise StopIteration diff --git a/modules/ui.py b/modules/ui.py index 2b688e32..1204eef7 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1070,7 +1070,7 @@ def create_ui(wrap_gradio_gpu_call): gr.HTML(value="

Train an embedding; must specify a directory with a set of 1:1 ratio images

") train_embedding_name = gr.Dropdown(label='Embedding', choices=sorted(sd_hijack.model_hijack.embedding_db.word_embeddings.keys())) train_hypernetwork_name = gr.Dropdown(label='Hypernetwork', choices=[x for x in shared.hypernetworks.keys()]) - learn_rate = gr.Textbox(label='Learning rate', placeholder="Learning rate", value = "5.0e-03") + learn_rate = gr.Textbox(label='Learning rate', placeholder="Learning rate", value="0.005") dataset_directory = gr.Textbox(label='Dataset directory', placeholder="Path to directory with input images") log_directory = gr.Textbox(label='Log directory', placeholder="Path to directory where to write outputs", value="textual_inversion") template_file = gr.Textbox(label='Prompt template file', value=os.path.join(script_path, "textual_inversion_templates", "style_filewords.txt")) -- cgit v1.2.3 From f53f703aebc801c4204182d52bb1e0bef9808e1f Mon Sep 17 00:00:00 2001 From: JC_Array Date: Tue, 11 Oct 2022 18:12:12 -0500 Subject: resolved conflicts, moved settings under interrogate section, settings only show if deepbooru flag is enabled --- modules/deepbooru.py | 2 +- modules/shared.py | 19 +++++++++---------- modules/textual_inversion/preprocess.py | 2 +- modules/ui.py | 2 +- 4 files changed, 12 insertions(+), 13 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/deepbooru.py b/modules/deepbooru.py index 89dcac3c..29529949 100644 --- a/modules/deepbooru.py +++ b/modules/deepbooru.py @@ -8,7 +8,7 @@ def get_deepbooru_tags(pil_image): This method is for running only one image at a time for simple use. Used to the img2img interrogate. """ from modules import shared # prevents circular reference - create_deepbooru_process(shared.opts.deepbooru_threshold, shared.opts.deepbooru_sort_alpha) + create_deepbooru_process(shared.opts.interrogate_deepbooru_score_threshold, shared.opts.deepbooru_sort_alpha) shared.deepbooru_process_return["value"] = -1 shared.deepbooru_process_queue.put(pil_image) while shared.deepbooru_process_return["value"] == -1: diff --git a/modules/shared.py b/modules/shared.py index 817203f8..5456c477 100644 --- a/modules/shared.py +++ b/modules/shared.py @@ -248,15 +248,20 @@ options_templates.update(options_section(('sd', "Stable Diffusion"), { "random_artist_categories": OptionInfo([], "Allowed categories for random artists selection when using the Roll button", gr.CheckboxGroup, {"choices": artist_db.categories()}), })) -options_templates.update(options_section(('interrogate', "Interrogate Options"), { +interrogate_option_dictionary = { "interrogate_keep_models_in_memory": OptionInfo(False, "Interrogate: keep models in VRAM"), "interrogate_use_builtin_artists": OptionInfo(True, "Interrogate: use artists from artists.csv"), "interrogate_clip_num_beams": OptionInfo(1, "Interrogate: num_beams for BLIP", gr.Slider, {"minimum": 1, "maximum": 16, "step": 1}), "interrogate_clip_min_length": OptionInfo(24, "Interrogate: minimum description length (excluding artists, etc..)", gr.Slider, {"minimum": 1, "maximum": 128, "step": 1}), "interrogate_clip_max_length": OptionInfo(48, "Interrogate: maximum description length", gr.Slider, {"minimum": 1, "maximum": 256, "step": 1}), - "interrogate_clip_dict_limit": OptionInfo(1500, "Interrogate: maximum number of lines in text file (0 = No limit)"), - "interrogate_deepbooru_score_threshold": OptionInfo(0.5, "Interrogate: deepbooru score threshold", gr.Slider, {"minimum": 0, "maximum": 1, "step": 0.01}), -})) + "interrogate_clip_dict_limit": OptionInfo(1500, "Interrogate: maximum number of lines in text file (0 = No limit)") +} + +if cmd_opts.deepdanbooru: + interrogate_option_dictionary["interrogate_deepbooru_score_threshold"] = OptionInfo(0.5, "Interrogate: deepbooru score threshold", gr.Slider, {"minimum": 0, "maximum": 1, "step": 0.01}) + interrogate_option_dictionary["deepbooru_sort_alpha"] = OptionInfo(True, "Interrogate: deepbooru sort alphabetically", gr.Checkbox) + +options_templates.update(options_section(('interrogate', "Interrogate Options"), interrogate_option_dictionary)) options_templates.update(options_section(('ui', "User interface"), { "show_progressbar": OptionInfo(True, "Show progressbar"), @@ -282,12 +287,6 @@ options_templates.update(options_section(('sampler-params', "Sampler parameters" 'eta_noise_seed_delta': OptionInfo(0, "Eta noise seed delta", gr.Number, {"precision": 0}), })) -if cmd_opts.deepdanbooru: - options_templates.update(options_section(('deepbooru-params', "DeepBooru parameters"), { - "deepbooru_sort_alpha": OptionInfo(True, "Sort Alphabetical", gr.Checkbox), - 'deepbooru_threshold': OptionInfo(0.5, "Threshold", gr.Slider, {"minimum": 0.0, "maximum": 1.0, "step": 0.01}), - })) - class Options: data = None diff --git a/modules/textual_inversion/preprocess.py b/modules/textual_inversion/preprocess.py index a96388d6..113cecf1 100644 --- a/modules/textual_inversion/preprocess.py +++ b/modules/textual_inversion/preprocess.py @@ -29,7 +29,7 @@ def preprocess(process_src, process_dst, process_width, process_height, process_ shared.interrogator.load() if process_caption_deepbooru: - deepbooru.create_deepbooru_process(opts.deepbooru_threshold, opts.deepbooru_sort_alpha) + deepbooru.create_deepbooru_process(opts.interrogate_deepbooru_score_threshold, opts.deepbooru_sort_alpha) def save_pic_with_caption(image, index): if process_caption: diff --git a/modules/ui.py b/modules/ui.py index 2891fc8c..fa45edca 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -317,7 +317,7 @@ def interrogate(image): def interrogate_deepbooru(image): - prompt = get_deepbooru_tags(image, opts.interrogate_deepbooru_score_threshold) + prompt = get_deepbooru_tags(image) return gr_show(True) if prompt is None else prompt -- cgit v1.2.3 From fec2221eeaafb50afd26ba3e109bf6f928011e69 Mon Sep 17 00:00:00 2001 From: Greg Fuller Date: Tue, 11 Oct 2022 19:29:38 -0700 Subject: Truncate error text to fix service lockup / stall What: * Update wrap_gradio_call to add a limit to the maximum amount of text output Why: * wrap_gradio_call currently prints out a list of the arguments provided to the failing function. * if that function is save_image, this causes the entire image to be printed to stderr * If the image is large, this can cause the service to lock up while attempting to print all the text * It is easy to generate large images using the x/y plot script * it is easy to encounter image save exceptions, including if the output directory does not exist / cannot be written to, or if the file is too big * The huge amount of log spam is confusing and not particularly helpful --- modules/ui.py | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-) (limited to 'modules/ui.py') diff --git a/modules/ui.py b/modules/ui.py index 1204eef7..33a49d3b 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -181,8 +181,15 @@ def wrap_gradio_call(func, extra_outputs=None): try: res = list(func(*args, **kwargs)) except Exception as e: + # When printing out our debug argument list, do not print out more than a MB of text + max_debug_str_len = 131072 # (1024*1024)/8 + print("Error completing request", file=sys.stderr) - print("Arguments:", args, kwargs, file=sys.stderr) + argStr = f"Arguments: {str(args)} {str(kwargs)}" + print(argStr[:max_debug_str_len], file=sys.stderr) + if len(argStr) > max_debug_str_len: + print(f"(Argument list truncated at {max_debug_str_len}/{len(argStr)} characters)", file=sys.stderr) + print(traceback.format_exc(), file=sys.stderr) shared.state.job = "" -- cgit v1.2.3 From 57e03cdd244eee4e33ccab7554b3594563a3d0cd Mon Sep 17 00:00:00 2001 From: brkirch Date: Wed, 12 Oct 2022 00:54:24 -0400 Subject: Ensure the directory exists before saving to it The directory for the images saved with the Save button may still not exist, so it needs to be created prior to opening the log.csv file. --- modules/ui.py | 2 ++ 1 file changed, 2 insertions(+) (limited to 'modules/ui.py') diff --git a/modules/ui.py b/modules/ui.py index 00bf09ae..cd67b84b 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -131,6 +131,8 @@ def save_files(js_data, images, do_make_zip, index): images = [images[index]] start_index = index + os.makedirs(opts.outdir_save, exist_ok=True) + with open(os.path.join(opts.outdir_save, "log.csv"), "a", encoding="utf8", newline='') as file: at_start = file.tell() == 0 writer = csv.writer(file) -- cgit v1.2.3 From ee015a1af66a94a75c914659fa0d321e702a0a87 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Wed, 12 Oct 2022 11:05:57 +0300 Subject: change textual inversion tab to train remake train interface to use tabs --- modules/hypernetworks/hypernetwork.py | 2 +- modules/ui.py | 22 +++++++++------------- 2 files changed, 10 insertions(+), 14 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/hypernetworks/hypernetwork.py b/modules/hypernetworks/hypernetwork.py index 8f2192e2..8314450a 100644 --- a/modules/hypernetworks/hypernetwork.py +++ b/modules/hypernetworks/hypernetwork.py @@ -175,7 +175,7 @@ def attention_CrossAttention_forward(self, x, context=None, mask=None): def train_hypernetwork(hypernetwork_name, learn_rate, data_root, log_directory, steps, create_image_every, save_hypernetwork_every, template_file, preview_image_prompt): - assert hypernetwork_name, 'embedding not selected' + assert hypernetwork_name, 'hypernetwork not selected' path = shared.hypernetworks.get(hypernetwork_name, None) shared.loaded_hypernetwork = Hypernetwork() diff --git a/modules/ui.py b/modules/ui.py index 4bfdd275..86a2da6c 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1035,14 +1035,14 @@ def create_ui(wrap_gradio_gpu_call): sd_hijack.model_hijack.embedding_db.load_textual_inversion_embeddings() - with gr.Blocks() as textual_inversion_interface: + with gr.Blocks() as train_interface: with gr.Row().style(equal_height=False): - with gr.Column(): - with gr.Group(): - gr.HTML(value="

See wiki for detailed explanation.

") + gr.HTML(value="

See wiki for detailed explanation.

") - gr.HTML(value="

Create a new embedding

") + with gr.Row().style(equal_height=False): + with gr.Tabs(elem_id="train_tabs"): + with gr.Tab(label="Create embedding"): new_embedding_name = gr.Textbox(label="Name") initialization_text = gr.Textbox(label="Initialization text", value="*") nvpt = gr.Slider(label="Number of vectors per token", minimum=1, maximum=75, step=1, value=1) @@ -1054,9 +1054,7 @@ def create_ui(wrap_gradio_gpu_call): with gr.Column(): create_embedding = gr.Button(value="Create embedding", variant='primary') - with gr.Group(): - gr.HTML(value="

Create a new hypernetwork

") - + with gr.Tab(label="Create hypernetwork"): new_hypernetwork_name = gr.Textbox(label="Name") new_hypernetwork_sizes = gr.CheckboxGroup(label="Modules", value=["768", "320", "640", "1280"], choices=["768", "320", "640", "1280"]) @@ -1067,9 +1065,7 @@ def create_ui(wrap_gradio_gpu_call): with gr.Column(): create_hypernetwork = gr.Button(value="Create hypernetwork", variant='primary') - with gr.Group(): - gr.HTML(value="

Preprocess images

") - + with gr.Tab(label="Preprocess images"): process_src = gr.Textbox(label='Source directory') process_dst = gr.Textbox(label='Destination directory') process_width = gr.Slider(minimum=64, maximum=2048, step=64, label="Width", value=512) @@ -1091,7 +1087,7 @@ def create_ui(wrap_gradio_gpu_call): with gr.Column(): run_preprocess = gr.Button(value="Preprocess", variant='primary') - with gr.Group(): + with gr.Tab(label="Train"): gr.HTML(value="

Train an embedding; must specify a directory with a set of 1:1 ratio images

") train_embedding_name = gr.Dropdown(label='Embedding', choices=sorted(sd_hijack.model_hijack.embedding_db.word_embeddings.keys())) train_hypernetwork_name = gr.Dropdown(label='Hypernetwork', choices=[x for x in shared.hypernetworks.keys()]) @@ -1388,7 +1384,7 @@ Requested path was: {f} (extras_interface, "Extras", "extras"), (pnginfo_interface, "PNG Info", "pnginfo"), (modelmerger_interface, "Checkpoint Merger", "modelmerger"), - (textual_inversion_interface, "Textual inversion", "ti"), + (train_interface, "Train", "ti"), (settings_interface, "Settings", "settings"), ] -- cgit v1.2.3 From a1a94b8b5f342f467aecc53b21b80ed0227ee76a Mon Sep 17 00:00:00 2001 From: yfszzx Date: Thu, 13 Oct 2022 00:19:34 +0800 Subject: images history improvement --- javascript/images_history.js | 125 ++++++++++++++++++++++--------------------- modules/images_history.py | 7 +-- modules/ui.py | 40 +++++++------- 3 files changed, 88 insertions(+), 84 deletions(-) (limited to 'modules/ui.py') diff --git a/javascript/images_history.js b/javascript/images_history.js index c9a63166..620f242c 100644 --- a/javascript/images_history.js +++ b/javascript/images_history.js @@ -18,20 +18,26 @@ var images_history_click_image = function(){ } var images_history_click_tab = function(){ - var tabs_box = gradioApp().getElementById("images_history_tab"); - if (!tabs_box.classList.contains(this.getAttribute("tabname"))) { - gradioApp().getElementById(this.getAttribute("tabname") + "_images_history_renew_page").click(); - tabs_box.classList.add(this.getAttribute("tabname")) - } + var tabs_box = gradioApp().getElementById("images_history_tab"); + if (!tabs_box.classList.contains(this.getAttribute("tabname"))) { + gradioApp().getElementById(this.getAttribute("tabname") + "_images_history_renew_page").click(); + tabs_box.classList.add(this.getAttribute("tabname")) + } } var images_history_close_full_view = function(){ - var box = images_history_get_parent_by_class(this, "images_history_cantainor"); - box.querySelector(".images_history_del_button").setAttribute("disabled", "disabled"); + var box = images_history_get_parent_by_class(this, "images_history_cantainor"); + box.querySelector(".images_history_del_button").setAttribute("disabled", "disabled"); +} + +function images_history_disabled_del(){ + gradioApp().querySelectorAll(".images_history_del_button").forEach(function(btn){ + btn.setAttribute('disabled','disabled'); + }); } function images_history_get_parent_by_class(item, class_name){ - var parent = item.parentElement; + var parent = item.parentElement; while(!parent.classList.contains(class_name)){ parent = parent.parentElement; } @@ -39,14 +45,15 @@ function images_history_get_parent_by_class(item, class_name){ } function images_history_get_parent_by_tagname(item, tagname){ - var parent = item.parentElement; - tagname = tagname.toUpperCase() + var parent = item.parentElement; + tagname = tagname.toUpperCase() while(parent.tagName != tagname){ - console.log(parent.tagName, tagname) + console.log(parent.tagName, tagname) parent = parent.parentElement; } return parent; } + function images_history_hide_buttons(hidden_list, gallery){ var buttons = gallery.querySelectorAll(".gallery-item"); var num = 0; @@ -54,7 +61,7 @@ function images_history_hide_buttons(hidden_list, gallery){ if (e.style.display == "none"){ num += 1; } - }) + }); if (num == hidden_list.length){ setTimeout(images_history_hide_buttons, 10, hidden_list, gallery); } @@ -74,14 +81,15 @@ function images_history_set_image_info(button){ if(e.style.display != "none"){ i += 1; } - }) + }); var gallery = images_history_get_parent_by_class(button, "images_history_cantainor"); var set_btn = gallery.querySelector(".images_history_set_index"); - set_btn.setAttribute("img_index", index); + var curr_idx = set_btn.getAttribute("img_index", index); + if (curr_idx != index) { + set_btn.setAttribute("img_index", index); + images_history_disabled_del(); + } set_btn.click(); - gradioApp().querySelectorAll(".images_history_del_button").forEach(function(btn){ - btn.setAttribute('disabled','disabled'); - }) } @@ -102,24 +110,24 @@ function images_history_delete(tabname, img_path, img_file_name, page_index, fil if (e.style.display != 'none'){ buttons.push(e); } - }); - + }); var img_num = buttons.length / 2; if (img_num === 1){ setTimeout(function(tabname){ gradioApp().getElementById(tabname + '_images_history_renew_page').click(); }, 30, tabname); - } else { + } else { buttons[image_index].style.display = 'none'; buttons[image_index + img_num].style.display = 'none'; var bnt; if (image_index >= img_num - 1){ btn = buttons[img_num - 2]; - } else { + } else { btn = buttons[image_index + 1] ; - } + } setTimeout(function(btn){btn.click()}, 30, btn); - } + } + images_history_disabled_del(); return [tabname, img_path, img_file_name, page_index, filenames, image_index]; } @@ -132,61 +140,58 @@ function images_history_turnpage(img_path, page_index, image_index, tabname){ } function images_history_enable_del_buttons(){ - gradioApp().querySelectorAll(".images_history_del_button").forEach(function(btn){ - btn.removeAttribute('disabled'); + gradioApp().querySelectorAll(".images_history_del_button").forEach(function(btn){ + btn.removeAttribute('disabled'); }) } function images_history_init(){ - if (gradioApp().getElementById('txt2img_images_history_renew_page') == null) { - setTimeout(images_history_init, 500); - } else { + var load_txt2img_button = gradioApp().getElementById('txt2img_images_history_renew_page') + if (load_txt2img_button){ for (var i in images_history_tab_list ){ tab = images_history_tab_list[i]; gradioApp().getElementById(tab + '_images_history').classList.add("images_history_cantainor"); gradioApp().getElementById(tab + '_images_history_set_index').classList.add("images_history_set_index"); gradioApp().getElementById(tab + '_images_history_del_button').classList.add("images_history_del_button"); - gradioApp().getElementById(tab + '_images_history_gallery').classList.add("images_history_gallery"); - - + gradioApp().getElementById(tab + '_images_history_gallery').classList.add("images_history_gallery"); + } var tabs_box = gradioApp().getElementById("tab_images_history").querySelector("div").querySelector("div").querySelector("div"); - tabs_box.setAttribute("id", "images_history_tab"); - tabs_box.classList.add(images_history_tab_list[0]); - gradioApp().getElementById("txt2img_images_history_renew_page").click(); - } + tabs_box.setAttribute("id", "images_history_tab"); + var tab_btns = tabs_box.querySelectorAll("button"); + for (var i in images_history_tab_list){ + var tabname = images_history_tab_list[i] + tab_btns[i].setAttribute("tabname", tabname); + tab_btns[i].addEventListener('click', images_history_click_tab); + } + tabs_box.classList.add(images_history_tab_list[0]); + load_txt2img_button.click(); + } else { + setTimeout(images_history_init, 500); + } } var images_history_tab_list = ["txt2img", "img2img", "extras"]; -var images_history_start_flag = false; - -onUiUpdate(function(){ - var tab = gradioApp().getElementById("images_history_tab"); - if (tab) { - if (!images_history_start_flag){ - images_history_init(); - images_history_start_flag = true; - } - var tab_btns = gradioApp().getElementById("images_history_tab").querySelectorAll("button"); - for (var i in images_history_tab_list ){ - var buttons = gradioApp().querySelectorAll('#' + images_history_tab_list[i] + '_images_history .gallery-item'); - buttons.forEach(function(bnt){ - bnt.addEventListener('click', images_history_click_image, true); - }); - var tabname = images_history_tab_list[i] - tab_btns[i].setAttribute("tabname", tabname); - tab_btns[i].addEventListener('click', images_history_click_tab, true); +setTimeout(images_history_init, 500) +document.addEventListener("DOMContentLoaded", function() { + var mutationObserver = new MutationObserver(function(m){ + for (var i in images_history_tab_list ){ + var buttons = gradioApp().querySelectorAll('#' + images_history_tab_list[i] + '_images_history .gallery-item'); + buttons.forEach(function(bnt){ + bnt.addEventListener('click', images_history_click_image, true); + }); // var cls_btn = gradioApp().getElementById(tabname + '_images_history_gallery').querySelector("svg"); // if (cls_btn){ - // cls_btn.addEventListener('click', images_history_close_full_view, false); + // cls_btn.addEventListener('click', images_history_close_full_view, false); // } // console.log(cls_btn, cls_btn.parentElement.parentElement) // if (cls_btn) { - // cls_btn = images_history_get_parent_by_tagname(cls_btn, "BUTTON"); - // cls_btn.addEventListener('click', images_history_close_full_view, true); - // } - } + // cls_btn = images_history_get_parent_by_tagname(cls_btn, "BUTTON"); + // cls_btn.addEventListener('click', images_history_close_full_view, true); + } + }); + mutationObserver.observe( gradioApp(), { childList:true, subtree:true }); + +}); - } -}); diff --git a/modules/images_history.py b/modules/images_history.py index 2bc4b7ee..1bca0ad9 100644 --- a/modules/images_history.py +++ b/modules/images_history.py @@ -61,7 +61,7 @@ def delete_image(tabname, dir_name, name, page_index, filenames, image_index): os.remove(path) txt_file = os.path.splitext(path)[0] + ".txt" if os.path.exists(txt_file): - os.remove(txt_file) + os.remove(txt_file) new_file_list = [] for f in filenames: if f == name: @@ -88,7 +88,7 @@ def show_images_history(gr, opts, tabname, run_pnginfo, switch_dict): end_page = gr.Button('End Page') with gr.Row(elem_id=tabname + "_images_history"): with gr.Row(): - with gr.Column(scale=2): + with gr.Column(scale=2): history_gallery = gr.Gallery(show_label=False, elem_id=tabname + "_images_history_gallery").style(grid=6) delete = gr.Button('Delete', elem_id=tabname + "_images_history_del_button") with gr.Column(): @@ -126,9 +126,10 @@ def show_images_history(gr, opts, tabname, run_pnginfo, switch_dict): #other funcitons set_index.click(show_image_info, _js="images_history_get_current_img", inputs=[tabname_box, img_path, filenames], outputs=[img_file_name, image_index, hide_image]) + img_file_name.change(fn=None, _js="images_history_enable_del_buttons", inputs=None, outputs=None) delete.click(delete_image,_js="images_history_delete", inputs=[tabname_box, img_path, img_file_name, page_index, filenames, image_index], outputs=[page_index, filenames]) hide_image.change(fn=run_pnginfo, inputs=[hide_image], outputs=[info1, img_file_info, info2]) - hide_image.change(fn=None, _js="images_history_enable_del_buttons", inputs=None, outputs=None) + #pnginfo.click(fn=run_pnginfo, inputs=[hide_image], outputs=[info1, img_file_info, info2]) switch_dict["fn"](pnginfo_send_to_txt2img, switch_dict["t2i"], img_file_info, 'switch_to_txt2img') switch_dict["fn"](pnginfo_send_to_img2img, switch_dict["i2i"], img_file_info, 'switch_to_img2img_img2img') diff --git a/modules/ui.py b/modules/ui.py index 94297ba6..8cd12b51 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -39,7 +39,7 @@ import modules.generation_parameters_copypaste from modules import prompt_parser from modules.images import save_image import modules.textual_inversion.ui -import modules.hypernetwork.ui +import modules.hypernetworks.ui import modules.images_history as img_his # this is a fix for Windows users. Without it, javascript files will be served with text/html content-type and the browser will not show any UI @@ -554,6 +554,7 @@ def create_ui(wrap_gradio_gpu_call): custom_inputs = modules.scripts.scripts_txt2img.setup_ui(is_img2img=False) with gr.Column(variant='panel'): + with gr.Group(): txt2img_preview = gr.Image(elem_id='txt2img_preview', visible=False) txt2img_gallery = gr.Gallery(label='Output', show_label=False, elem_id='txt2img_gallery').style(grid=4) @@ -573,9 +574,9 @@ def create_ui(wrap_gradio_gpu_call): with gr.Row(): download_files = gr.File(None, file_count="multiple", interactive=False, show_label=False, visible=False) - with gr.Group(): - html_info = gr.HTML() - generation_info = gr.Textbox(visible=False) + with gr.Group(): + html_info = gr.HTML() + generation_info = gr.Textbox(visible=False) connect_reuse_seed(seed, reuse_seed, generation_info, dummy_component, is_subseed=False) connect_reuse_seed(subseed, reuse_subseed, generation_info, dummy_component, is_subseed=True) @@ -669,7 +670,6 @@ def create_ui(wrap_gradio_gpu_call): ] modules.generation_parameters_copypaste.connect_paste(paste, txt2img_paste_fields, txt2img_prompt) token_button.click(fn=update_token_counter, inputs=[txt2img_prompt, steps], outputs=[token_counter]) - with gr.Blocks(analytics_enabled=False) as img2img_interface: img2img_prompt, roll, img2img_prompt_style, img2img_negative_prompt, img2img_prompt_style2, submit, img2img_interrogate, img2img_deepbooru, img2img_prompt_style_apply, img2img_save_style, paste, token_counter, token_button = create_toprow(is_img2img=True) @@ -762,10 +762,10 @@ def create_ui(wrap_gradio_gpu_call): with gr.Row(): download_files = gr.File(None, file_count="multiple", interactive=False, show_label=False, visible=False) - with gr.Group(): - html_info = gr.HTML() - generation_info = gr.Textbox(visible=False) - + with gr.Group(): + html_info = gr.HTML() + generation_info = gr.Textbox(visible=False) + connect_reuse_seed(seed, reuse_seed, generation_info, dummy_component, is_subseed=False) connect_reuse_seed(subseed, reuse_subseed, generation_info, dummy_component, is_subseed=True) @@ -1016,6 +1016,13 @@ def create_ui(wrap_gradio_gpu_call): inputs=[image], outputs=[html, generation_info, html2], ) + #images history + images_history_switch_dict = { + "fn":modules.generation_parameters_copypaste.connect_paste, + "t2i":txt2img_paste_fields, + "i2i":img2img_paste_fields + } + images_history = img_his.create_history_tabs(gr, opts, wrap_gradio_call(modules.extras.run_pnginfo), images_history_switch_dict) with gr.Blocks() as modelmerger_interface: with gr.Row().style(equal_height=False): @@ -1285,16 +1292,7 @@ Requested path was: {f} opts.save(shared.config_filename) - return f'{changed} settings changed.', opts.dumpjson() - - #images history - images_history_switch_dict = { - "fn":modules.generation_parameters_copypaste.connect_paste, - "t2i":txt2img_paste_fields, - "i2i":img2img_paste_fields - } - images_history = img_his.create_history_tabs(gr, opts, wrap_gradio_call(modules.extras.run_pnginfo), images_history_switch_dict) - + return f'{changed} settings changed.', opts.dumpjson() def run_settings_single(value, key): if not opts.same_type(value, opts.data_labels[key].default): @@ -1393,11 +1391,10 @@ Requested path was: {f} (img2img_interface, "img2img", "img2img"), (extras_interface, "Extras", "extras"), (pnginfo_interface, "PNG Info", "pnginfo"), - (modelmerger_interface, "Checkpoint Merger", "modelmerger"), (images_history, "History", "images_history"), + (modelmerger_interface, "Checkpoint Merger", "modelmerger"), (train_interface, "Train", "ti"), (settings_interface, "Settings", "settings"), - ] with open(os.path.join(script_path, "style.css"), "r", encoding="utf8") as file: @@ -1616,3 +1613,4 @@ if 'gradio_routes_templates_response' not in globals(): gradio_routes_templates_response = gradio.routes.templates.TemplateResponse gradio.routes.templates.TemplateResponse = template_response + -- cgit v1.2.3 From c3c8eef9fd5a0c8b26319e32ca4a19b56204e6df Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Wed, 12 Oct 2022 20:49:47 +0300 Subject: train: change filename processing to be more simple and configurable train: make it possible to make text files with prompts train: rework scheduler so that there's less repeating code in textual inversion and hypernets train: move epochs setting to options --- javascript/hints.js | 3 ++ modules/hypernetworks/hypernetwork.py | 40 +++++++++------------- modules/shared.py | 3 ++ modules/textual_inversion/dataset.py | 47 +++++++++++++++++++------- modules/textual_inversion/learn_schedule.py | 37 +++++++++++++++++++- modules/textual_inversion/textual_inversion.py | 35 +++++++------------ modules/ui.py | 2 -- 7 files changed, 105 insertions(+), 62 deletions(-) (limited to 'modules/ui.py') diff --git a/javascript/hints.js b/javascript/hints.js index b81c181b..d51ee14c 100644 --- a/javascript/hints.js +++ b/javascript/hints.js @@ -81,6 +81,9 @@ titles = { "Eta noise seed delta": "If this values is non-zero, it will be added to seed and used to initialize RNG for noises when using samplers with Eta. You can use this to produce even more variation of images, or you can use this to match images of other software if you know what you are doing.", "Do not add watermark to images": "If this option is enabled, watermark will not be added to created images. Warning: if you do not add watermark, you may be behaving in an unethical manner.", + + "Filename word regex": "This regular expression will be used extract words from filename, and they will be joined using the option below into label text used for training. Leave empty to keep filename text as it is.", + "Filename join string": "This string will be used to hoin split words into a single line if the option above is enabled.", } diff --git a/modules/hypernetworks/hypernetwork.py b/modules/hypernetworks/hypernetwork.py index 8314450a..b6c06d49 100644 --- a/modules/hypernetworks/hypernetwork.py +++ b/modules/hypernetworks/hypernetwork.py @@ -14,7 +14,7 @@ import torch from torch import einsum from einops import rearrange, repeat import modules.textual_inversion.dataset -from modules.textual_inversion.learn_schedule import LearnSchedule +from modules.textual_inversion.learn_schedule import LearnRateScheduler class HypernetworkModule(torch.nn.Module): @@ -223,31 +223,23 @@ def train_hypernetwork(hypernetwork_name, learn_rate, data_root, log_directory, if ititial_step > steps: return hypernetwork, filename - schedules = iter(LearnSchedule(learn_rate, steps, ititial_step)) - (learn_rate, end_step) = next(schedules) - print(f'Training at rate of {learn_rate} until step {end_step}') - - optimizer = torch.optim.AdamW(weights, lr=learn_rate) + scheduler = LearnRateScheduler(learn_rate, steps, ititial_step) + optimizer = torch.optim.AdamW(weights, lr=scheduler.learn_rate) pbar = tqdm.tqdm(enumerate(ds), total=steps - ititial_step) - for i, (x, text, cond) in pbar: + for i, entry in pbar: hypernetwork.step = i + ititial_step - if hypernetwork.step > end_step: - try: - (learn_rate, end_step) = next(schedules) - except Exception: - break - tqdm.tqdm.write(f'Training at rate of {learn_rate} until step {end_step}') - for pg in optimizer.param_groups: - pg['lr'] = learn_rate + scheduler.apply(optimizer, hypernetwork.step) + if scheduler.finished: + break if shared.state.interrupted: break with torch.autocast("cuda"): - cond = cond.to(devices.device) - x = x.to(devices.device) + cond = entry.cond.to(devices.device) + x = entry.latent.to(devices.device) loss = shared.sd_model(x.unsqueeze(0), cond)[0] del x del cond @@ -267,7 +259,7 @@ def train_hypernetwork(hypernetwork_name, learn_rate, data_root, log_directory, if hypernetwork.step > 0 and images_dir is not None and hypernetwork.step % create_image_every == 0: last_saved_image = os.path.join(images_dir, f'{hypernetwork_name}-{hypernetwork.step}.png') - preview_text = text if preview_image_prompt == "" else preview_image_prompt + preview_text = entry.cond_text if preview_image_prompt == "" else preview_image_prompt optimizer.zero_grad() shared.sd_model.cond_stage_model.to(devices.device) @@ -282,16 +274,16 @@ def train_hypernetwork(hypernetwork_name, learn_rate, data_root, log_directory, ) processed = processing.process_images(p) - image = processed.images[0] + image = processed.images[0] if len(processed.images)>0 else None if unload: shared.sd_model.cond_stage_model.to(devices.cpu) shared.sd_model.first_stage_model.to(devices.cpu) - shared.state.current_image = image - image.save(last_saved_image) - - last_saved_image += f", prompt: {preview_text}" + if image is not None: + shared.state.current_image = image + image.save(last_saved_image) + last_saved_image += f", prompt: {preview_text}" shared.state.job_no = hypernetwork.step @@ -299,7 +291,7 @@ def train_hypernetwork(hypernetwork_name, learn_rate, data_root, log_directory,

Loss: {losses.mean():.7f}
Step: {hypernetwork.step}
-Last prompt: {html.escape(text)}
+Last prompt: {html.escape(entry.cond_text)}
Last saved embedding: {html.escape(last_saved_file)}
Last saved image: {html.escape(last_saved_image)}

diff --git a/modules/shared.py b/modules/shared.py index 42e99741..e64e69fc 100644 --- a/modules/shared.py +++ b/modules/shared.py @@ -231,6 +231,9 @@ options_templates.update(options_section(('system', "System"), { options_templates.update(options_section(('training', "Training"), { "unload_models_when_training": OptionInfo(False, "Unload VAE and CLIP from VRAM when training"), + "dataset_filename_word_regex": OptionInfo("", "Filename word regex"), + "dataset_filename_join_string": OptionInfo(" ", "Filename join string"), + "training_image_repeats_per_epoch": OptionInfo(100, "Number of repeats for a single input image per epoch; used only for displaying epoch number", gr.Number, {"precision": 0}), })) options_templates.update(options_section(('sd', "Stable Diffusion"), { diff --git a/modules/textual_inversion/dataset.py b/modules/textual_inversion/dataset.py index f61f40d3..67e90afe 100644 --- a/modules/textual_inversion/dataset.py +++ b/modules/textual_inversion/dataset.py @@ -11,11 +11,21 @@ import tqdm from modules import devices, shared import re -re_tag = re.compile(r"[a-zA-Z][_\w\d()]+") +re_numbers_at_start = re.compile(r"^[-\d]+\s*") + + +class DatasetEntry: + def __init__(self, filename=None, latent=None, filename_text=None): + self.filename = filename + self.latent = latent + self.filename_text = filename_text + self.cond = None + self.cond_text = None class PersonalizedBase(Dataset): def __init__(self, data_root, width, height, repeats, flip_p=0.5, placeholder_token="*", model=None, device=None, template_file=None, include_cond=False): + re_word = re.compile(shared.opts.dataset_filename_word_regex) if len(shared.opts.dataset_filename_word_regex)>0 else None self.placeholder_token = placeholder_token @@ -42,9 +52,18 @@ class PersonalizedBase(Dataset): except Exception: continue + text_filename = os.path.splitext(path)[0] + ".txt" filename = os.path.basename(path) - filename_tokens = os.path.splitext(filename)[0] - filename_tokens = re_tag.findall(filename_tokens) + + if os.path.exists(text_filename): + with open(text_filename, "r", encoding="utf8") as file: + filename_text = file.read() + else: + filename_text = os.path.splitext(filename)[0] + filename_text = re.sub(re_numbers_at_start, '', filename_text) + if re_word: + tokens = re_word.findall(filename_text) + filename_text = (shared.opts.dataset_filename_join_string or "").join(tokens) npimage = np.array(image).astype(np.uint8) npimage = (npimage / 127.5 - 1.0).astype(np.float32) @@ -55,13 +74,13 @@ class PersonalizedBase(Dataset): init_latent = model.get_first_stage_encoding(model.encode_first_stage(torchdata.unsqueeze(dim=0))).squeeze() init_latent = init_latent.to(devices.cpu) + entry = DatasetEntry(filename=path, filename_text=filename_text, latent=init_latent) + if include_cond: - text = self.create_text(filename_tokens) - cond = cond_model([text]).to(devices.cpu) - else: - cond = None + entry.cond_text = self.create_text(filename_text) + entry.cond = cond_model([entry.cond_text]).to(devices.cpu) - self.dataset.append((init_latent, filename_tokens, cond)) + self.dataset.append(entry) self.length = len(self.dataset) * repeats @@ -72,10 +91,10 @@ class PersonalizedBase(Dataset): def shuffle(self): self.indexes = self.initial_indexes[torch.randperm(self.initial_indexes.shape[0])] - def create_text(self, filename_tokens): + def create_text(self, filename_text): text = random.choice(self.lines) text = text.replace("[name]", self.placeholder_token) - text = text.replace("[filewords]", ' '.join(filename_tokens)) + text = text.replace("[filewords]", filename_text) return text def __len__(self): @@ -86,7 +105,9 @@ class PersonalizedBase(Dataset): self.shuffle() index = self.indexes[i % len(self.indexes)] - x, filename_tokens, cond = self.dataset[index] + entry = self.dataset[index] + + if entry.cond is None: + entry.cond_text = self.create_text(entry.filename_text) - text = self.create_text(filename_tokens) - return x, text, cond + return entry diff --git a/modules/textual_inversion/learn_schedule.py b/modules/textual_inversion/learn_schedule.py index db720271..2062726a 100644 --- a/modules/textual_inversion/learn_schedule.py +++ b/modules/textual_inversion/learn_schedule.py @@ -1,6 +1,12 @@ +import tqdm -class LearnSchedule: + +class LearnScheduleIterator: def __init__(self, learn_rate, max_steps, cur_step=0): + """ + specify learn_rate as "0.001:100, 0.00001:1000, 1e-5:10000" to have lr of 0.001 until step 100, 0.00001 until 1000, 1e-5:10000 until 10000 + """ + pairs = learn_rate.split(',') self.rates = [] self.it = 0 @@ -32,3 +38,32 @@ class LearnSchedule: return self.rates[self.it - 1] else: raise StopIteration + + +class LearnRateScheduler: + def __init__(self, learn_rate, max_steps, cur_step=0, verbose=True): + self.schedules = LearnScheduleIterator(learn_rate, max_steps, cur_step) + (self.learn_rate, self.end_step) = next(self.schedules) + self.verbose = verbose + + if self.verbose: + print(f'Training at rate of {self.learn_rate} until step {self.end_step}') + + self.finished = False + + def apply(self, optimizer, step_number): + if step_number <= self.end_step: + return + + try: + (self.learn_rate, self.end_step) = next(self.schedules) + except Exception: + self.finished = True + return + + if self.verbose: + tqdm.tqdm.write(f'Training at rate of {self.learn_rate} until step {self.end_step}') + + for pg in optimizer.param_groups: + pg['lr'] = self.learn_rate + diff --git a/modules/textual_inversion/textual_inversion.py b/modules/textual_inversion/textual_inversion.py index c5153e4a..fa0e33a2 100644 --- a/modules/textual_inversion/textual_inversion.py +++ b/modules/textual_inversion/textual_inversion.py @@ -11,7 +11,7 @@ from PIL import Image, PngImagePlugin from modules import shared, devices, sd_hijack, processing, sd_models import modules.textual_inversion.dataset -from modules.textual_inversion.learn_schedule import LearnSchedule +from modules.textual_inversion.learn_schedule import LearnRateScheduler from modules.textual_inversion.image_embedding import (embedding_to_b64, embedding_from_b64, insert_image_data_embed, extract_image_data_embed, @@ -172,8 +172,7 @@ def create_embedding(name, num_vectors_per_token, init_text='*'): return fn - -def train_embedding(embedding_name, learn_rate, data_root, log_directory, training_width, training_height, steps, num_repeats, create_image_every, save_embedding_every, template_file, save_image_with_stored_embedding, preview_image_prompt): +def train_embedding(embedding_name, learn_rate, data_root, log_directory, training_width, training_height, steps, create_image_every, save_embedding_every, template_file, save_image_with_stored_embedding, preview_image_prompt): assert embedding_name, 'embedding not selected' shared.state.textinfo = "Initializing textual inversion training..." @@ -205,7 +204,7 @@ def train_embedding(embedding_name, learn_rate, data_root, log_directory, traini shared.state.textinfo = f"Preparing dataset from {html.escape(data_root)}..." with torch.autocast("cuda"): - ds = modules.textual_inversion.dataset.PersonalizedBase(data_root=data_root, width=training_width, height=training_height, repeats=num_repeats, placeholder_token=embedding_name, model=shared.sd_model, device=devices.device, template_file=template_file) + ds = modules.textual_inversion.dataset.PersonalizedBase(data_root=data_root, width=training_width, height=training_height, repeats=shared.opts.training_image_repeats_per_epoch, placeholder_token=embedding_name, model=shared.sd_model, device=devices.device, template_file=template_file) hijack = sd_hijack.model_hijack @@ -221,32 +220,24 @@ def train_embedding(embedding_name, learn_rate, data_root, log_directory, traini if ititial_step > steps: return embedding, filename - schedules = iter(LearnSchedule(learn_rate, steps, ititial_step)) - (learn_rate, end_step) = next(schedules) - print(f'Training at rate of {learn_rate} until step {end_step}') - - optimizer = torch.optim.AdamW([embedding.vec], lr=learn_rate) + scheduler = LearnRateScheduler(learn_rate, steps, ititial_step) + optimizer = torch.optim.AdamW([embedding.vec], lr=scheduler.learn_rate) pbar = tqdm.tqdm(enumerate(ds), total=steps-ititial_step) - for i, (x, text, _) in pbar: + for i, entry in pbar: embedding.step = i + ititial_step - if embedding.step > end_step: - try: - (learn_rate, end_step) = next(schedules) - except: - break - tqdm.tqdm.write(f'Training at rate of {learn_rate} until step {end_step}') - for pg in optimizer.param_groups: - pg['lr'] = learn_rate + scheduler.apply(optimizer, embedding.step) + if scheduler.finished: + break if shared.state.interrupted: break with torch.autocast("cuda"): - c = cond_model([text]) + c = cond_model([entry.cond_text]) - x = x.to(devices.device) + x = entry.latent.to(devices.device) loss = shared.sd_model(x.unsqueeze(0), c)[0] del x @@ -268,7 +259,7 @@ def train_embedding(embedding_name, learn_rate, data_root, log_directory, traini if embedding.step > 0 and images_dir is not None and embedding.step % create_image_every == 0: last_saved_image = os.path.join(images_dir, f'{embedding_name}-{embedding.step}.png') - preview_text = text if preview_image_prompt == "" else preview_image_prompt + preview_text = entry.cond_text if preview_image_prompt == "" else preview_image_prompt p = processing.StableDiffusionProcessingTxt2Img( sd_model=shared.sd_model, @@ -314,7 +305,7 @@ def train_embedding(embedding_name, learn_rate, data_root, log_directory, traini

Loss: {losses.mean():.7f}
Step: {embedding.step}
-Last prompt: {html.escape(text)}
+Last prompt: {html.escape(entry.cond_text)}
Last saved embedding: {html.escape(last_saved_file)}
Last saved image: {html.escape(last_saved_image)}

diff --git a/modules/ui.py b/modules/ui.py index 2b332267..c42535c8 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1098,7 +1098,6 @@ def create_ui(wrap_gradio_gpu_call): training_width = gr.Slider(minimum=64, maximum=2048, step=64, label="Width", value=512) training_height = gr.Slider(minimum=64, maximum=2048, step=64, label="Height", value=512) steps = gr.Number(label='Max steps', value=100000, precision=0) - num_repeats = gr.Number(label='Number of repeats for a single input image per epoch', value=100, precision=0) create_image_every = gr.Number(label='Save an image to log directory every N steps, 0 to disable', value=500, precision=0) save_embedding_every = gr.Number(label='Save a copy of embedding to log directory every N steps, 0 to disable', value=500, precision=0) save_image_with_stored_embedding = gr.Checkbox(label='Save images with embedding in PNG chunks', value=True) @@ -1176,7 +1175,6 @@ def create_ui(wrap_gradio_gpu_call): training_width, training_height, steps, - num_repeats, create_image_every, save_embedding_every, template_file, -- cgit v1.2.3 From 698d303b04e293635bfb49c525409f3bcf671dce Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Wed, 12 Oct 2022 21:55:43 +0300 Subject: deepbooru: added option to use spaces or underscores deepbooru: added option to quote (\) in tags deepbooru/BLIP: write caption to file instead of image filename deepbooru/BLIP: now possible to use both for captions deepbooru: process is stopped even if an exception occurs --- modules/deepbooru.py | 65 ++++++++++++++++++----- modules/shared.py | 2 + modules/textual_inversion/preprocess.py | 92 ++++++++++++++------------------- modules/ui.py | 7 +-- 4 files changed, 95 insertions(+), 71 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/deepbooru.py b/modules/deepbooru.py index 29529949..419e6a9c 100644 --- a/modules/deepbooru.py +++ b/modules/deepbooru.py @@ -2,33 +2,44 @@ import os.path from concurrent.futures import ProcessPoolExecutor import multiprocessing import time +import re + +re_special = re.compile(r'([\\()])') def get_deepbooru_tags(pil_image): """ This method is for running only one image at a time for simple use. Used to the img2img interrogate. """ from modules import shared # prevents circular reference - create_deepbooru_process(shared.opts.interrogate_deepbooru_score_threshold, shared.opts.deepbooru_sort_alpha) - shared.deepbooru_process_return["value"] = -1 - shared.deepbooru_process_queue.put(pil_image) - while shared.deepbooru_process_return["value"] == -1: - time.sleep(0.2) - tags = shared.deepbooru_process_return["value"] - release_process() - return tags + try: + create_deepbooru_process(shared.opts.interrogate_deepbooru_score_threshold, create_deepbooru_opts()) + return get_tags_from_process(pil_image) + finally: + release_process() + + +def create_deepbooru_opts(): + from modules import shared -def deepbooru_process(queue, deepbooru_process_return, threshold, alpha_sort): + return { + "use_spaces": shared.opts.deepbooru_use_spaces, + "use_escape": shared.opts.deepbooru_escape, + "alpha_sort": shared.opts.deepbooru_sort_alpha, + } + + +def deepbooru_process(queue, deepbooru_process_return, threshold, deepbooru_opts): model, tags = get_deepbooru_tags_model() while True: # while process is running, keep monitoring queue for new image pil_image = queue.get() if pil_image == "QUIT": break else: - deepbooru_process_return["value"] = get_deepbooru_tags_from_model(model, tags, pil_image, threshold, alpha_sort) + deepbooru_process_return["value"] = get_deepbooru_tags_from_model(model, tags, pil_image, threshold, deepbooru_opts) -def create_deepbooru_process(threshold, alpha_sort): +def create_deepbooru_process(threshold, deepbooru_opts): """ Creates deepbooru process. A queue is created to send images into the process. This enables multiple images to be processed in a row without reloading the model or creating a new process. To return the data, a shared @@ -41,10 +52,23 @@ def create_deepbooru_process(threshold, alpha_sort): shared.deepbooru_process_queue = shared.deepbooru_process_manager.Queue() shared.deepbooru_process_return = shared.deepbooru_process_manager.dict() shared.deepbooru_process_return["value"] = -1 - shared.deepbooru_process = multiprocessing.Process(target=deepbooru_process, args=(shared.deepbooru_process_queue, shared.deepbooru_process_return, threshold, alpha_sort)) + shared.deepbooru_process = multiprocessing.Process(target=deepbooru_process, args=(shared.deepbooru_process_queue, shared.deepbooru_process_return, threshold, deepbooru_opts)) shared.deepbooru_process.start() +def get_tags_from_process(image): + from modules import shared + + shared.deepbooru_process_return["value"] = -1 + shared.deepbooru_process_queue.put(image) + while shared.deepbooru_process_return["value"] == -1: + time.sleep(0.2) + caption = shared.deepbooru_process_return["value"] + shared.deepbooru_process_return["value"] = -1 + + return caption + + def release_process(): """ Stops the deepbooru process to return used memory @@ -81,10 +105,15 @@ def get_deepbooru_tags_model(): return model, tags -def get_deepbooru_tags_from_model(model, tags, pil_image, threshold, alpha_sort): +def get_deepbooru_tags_from_model(model, tags, pil_image, threshold, deepbooru_opts): import deepdanbooru as dd import tensorflow as tf import numpy as np + + alpha_sort = deepbooru_opts['alpha_sort'] + use_spaces = deepbooru_opts['use_spaces'] + use_escape = deepbooru_opts['use_escape'] + width = model.input_shape[2] height = model.input_shape[1] image = np.array(pil_image) @@ -129,4 +158,12 @@ def get_deepbooru_tags_from_model(model, tags, pil_image, threshold, alpha_sort) print('\n'.join(sorted(result_tags_print, reverse=True))) - return ', '.join(result_tags_out).replace('_', ' ').replace(':', ' ') + tags_text = ', '.join(result_tags_out) + + if use_spaces: + tags_text = tags_text.replace('_', ' ') + + if use_escape: + tags_text = re.sub(re_special, r'\\\1', tags_text) + + return tags_text.replace(':', ' ') diff --git a/modules/shared.py b/modules/shared.py index e64e69fc..78b73aae 100644 --- a/modules/shared.py +++ b/modules/shared.py @@ -260,6 +260,8 @@ options_templates.update(options_section(('interrogate', "Interrogate Options"), "interrogate_clip_max_length": OptionInfo(48, "Interrogate: maximum description length", gr.Slider, {"minimum": 1, "maximum": 256, "step": 1}), "interrogate_deepbooru_score_threshold": OptionInfo(0.5, "Interrogate: deepbooru score threshold", gr.Slider, {"minimum": 0, "maximum": 1, "step": 0.01}), "deepbooru_sort_alpha": OptionInfo(True, "Interrogate: deepbooru sort alphabetically"), + "deepbooru_use_spaces": OptionInfo(False, "use spaces for tags in deepbooru"), + "deepbooru_escape": OptionInfo(True, "escape (\\) brackets in deepbooru (so they are used as literal brackets and not for emphasis)"), })) options_templates.update(options_section(('ui', "User interface"), { diff --git a/modules/textual_inversion/preprocess.py b/modules/textual_inversion/preprocess.py index 113cecf1..3047bede 100644 --- a/modules/textual_inversion/preprocess.py +++ b/modules/textual_inversion/preprocess.py @@ -10,7 +10,28 @@ from modules.shared import opts, cmd_opts if cmd_opts.deepdanbooru: import modules.deepbooru as deepbooru + def preprocess(process_src, process_dst, process_width, process_height, process_flip, process_split, process_caption, process_caption_deepbooru=False): + try: + if process_caption: + shared.interrogator.load() + + if process_caption_deepbooru: + deepbooru.create_deepbooru_process(opts.interrogate_deepbooru_score_threshold, deepbooru.create_deepbooru_opts()) + + preprocess_work(process_src, process_dst, process_width, process_height, process_flip, process_split, process_caption, process_caption_deepbooru) + + finally: + + if process_caption: + shared.interrogator.send_blip_to_ram() + + if process_caption_deepbooru: + deepbooru.release_process() + + + +def preprocess_work(process_src, process_dst, process_width, process_height, process_flip, process_split, process_caption, process_caption_deepbooru=False): width = process_width height = process_height src = os.path.abspath(process_src) @@ -25,30 +46,28 @@ def preprocess(process_src, process_dst, process_width, process_height, process_ shared.state.textinfo = "Preprocessing..." shared.state.job_count = len(files) - if process_caption: - shared.interrogator.load() - - if process_caption_deepbooru: - deepbooru.create_deepbooru_process(opts.interrogate_deepbooru_score_threshold, opts.deepbooru_sort_alpha) - def save_pic_with_caption(image, index): + caption = "" + if process_caption: - caption = "-" + shared.interrogator.generate_caption(image) - caption = sanitize_caption(os.path.join(dst, f"{index:05}-{subindex[0]}"), caption, ".png") - elif process_caption_deepbooru: - shared.deepbooru_process_return["value"] = -1 - shared.deepbooru_process_queue.put(image) - while shared.deepbooru_process_return["value"] == -1: - time.sleep(0.2) - caption = "-" + shared.deepbooru_process_return["value"] - caption = sanitize_caption(os.path.join(dst, f"{index:05}-{subindex[0]}"), caption, ".png") - shared.deepbooru_process_return["value"] = -1 - else: - caption = filename - caption = os.path.splitext(caption)[0] - caption = os.path.basename(caption) + caption += shared.interrogator.generate_caption(image) + + if process_caption_deepbooru: + if len(caption) > 0: + caption += ", " + caption += deepbooru.get_tags_from_process(image) + + filename_part = filename + filename_part = os.path.splitext(filename_part)[0] + filename_part = os.path.basename(filename_part) + + basename = f"{index:05}-{subindex[0]}-{filename_part}" + image.save(os.path.join(dst, f"{basename}.png")) + + if len(caption) > 0: + with open(os.path.join(dst, f"{basename}.txt"), "w", encoding="utf8") as file: + file.write(caption) - image.save(os.path.join(dst, f"{index:05}-{subindex[0]}{caption}.png")) subindex[0] += 1 def save_pic(image, index): @@ -93,34 +112,3 @@ def preprocess(process_src, process_dst, process_width, process_height, process_ save_pic(img, index) shared.state.nextjob() - - if process_caption: - shared.interrogator.send_blip_to_ram() - - if process_caption_deepbooru: - deepbooru.release_process() - - -def sanitize_caption(base_path, original_caption, suffix): - operating_system = platform.system().lower() - if (operating_system == "windows"): - invalid_path_characters = "\\/:*?\"<>|" - max_path_length = 259 - else: - invalid_path_characters = "/" #linux/macos - max_path_length = 1023 - caption = original_caption - for invalid_character in invalid_path_characters: - caption = caption.replace(invalid_character, "") - fixed_path_length = len(base_path) + len(suffix) - if fixed_path_length + len(caption) <= max_path_length: - return caption - caption_tokens = caption.split() - new_caption = "" - for token in caption_tokens: - last_caption = new_caption - new_caption = new_caption + token + " " - if (len(new_caption) + fixed_path_length - 1 > max_path_length): - break - print(f"\nPath will be too long. Truncated caption: {original_caption}\nto: {last_caption}", file=sys.stderr) - return last_caption.strip() diff --git a/modules/ui.py b/modules/ui.py index c42535c8..e07ee0e1 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1074,11 +1074,8 @@ def create_ui(wrap_gradio_gpu_call): with gr.Row(): process_flip = gr.Checkbox(label='Create flipped copies') process_split = gr.Checkbox(label='Split oversized images into two') - process_caption = gr.Checkbox(label='Use BLIP caption as filename') - if cmd_opts.deepdanbooru: - process_caption_deepbooru = gr.Checkbox(label='Use deepbooru caption as filename') - else: - process_caption_deepbooru = gr.Checkbox(label='Use deepbooru caption as filename', visible=False) + process_caption = gr.Checkbox(label='Use BLIP for caption') + process_caption_deepbooru = gr.Checkbox(label='Use deepbooru for caption', visible=True if cmd_opts.deepdanbooru else False) with gr.Row(): with gr.Column(scale=3): -- cgit v1.2.3 From 1cfc2a18981ee56bdb69a2de7b463a11ad05e329 Mon Sep 17 00:00:00 2001 From: Melan Date: Wed, 12 Oct 2022 23:36:29 +0200 Subject: Save a csv containing the loss while training --- modules/hypernetworks/hypernetwork.py | 17 ++++++++++++++++- modules/textual_inversion/textual_inversion.py | 17 ++++++++++++++++- modules/ui.py | 3 +++ 3 files changed, 35 insertions(+), 2 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/hypernetworks/hypernetwork.py b/modules/hypernetworks/hypernetwork.py index b6c06d49..6522078f 100644 --- a/modules/hypernetworks/hypernetwork.py +++ b/modules/hypernetworks/hypernetwork.py @@ -5,6 +5,7 @@ import os import sys import traceback import tqdm +import csv import torch @@ -174,7 +175,7 @@ def attention_CrossAttention_forward(self, x, context=None, mask=None): return self.to_out(out) -def train_hypernetwork(hypernetwork_name, learn_rate, data_root, log_directory, steps, create_image_every, save_hypernetwork_every, template_file, preview_image_prompt): +def train_hypernetwork(hypernetwork_name, learn_rate, data_root, log_directory, steps, create_image_every, save_hypernetwork_every, write_csv_every, template_file, preview_image_prompt): assert hypernetwork_name, 'hypernetwork not selected' path = shared.hypernetworks.get(hypernetwork_name, None) @@ -256,6 +257,20 @@ def train_hypernetwork(hypernetwork_name, learn_rate, data_root, log_directory, last_saved_file = os.path.join(hypernetwork_dir, f'{hypernetwork_name}-{hypernetwork.step}.pt') hypernetwork.save(last_saved_file) + print(f"{write_csv_every} > {hypernetwork.step % write_csv_every == 0}, {write_csv_every}") + if write_csv_every > 0 and hypernetwork_dir is not None and hypernetwork.step % write_csv_every == 0: + write_csv_header = False if os.path.exists(os.path.join(hypernetwork_dir, "hypernetwork_loss.csv")) else True + + with open(os.path.join(hypernetwork_dir, "hypernetwork_loss.csv"), "a+") as fout: + + csv_writer = csv.DictWriter(fout, fieldnames=["step", "loss"]) + + if write_csv_header: + csv_writer.writeheader() + + csv_writer.writerow({"step": hypernetwork.step, + "loss": f"{losses.mean():.7f}"}) + if hypernetwork.step > 0 and images_dir is not None and hypernetwork.step % create_image_every == 0: last_saved_image = os.path.join(images_dir, f'{hypernetwork_name}-{hypernetwork.step}.png') diff --git a/modules/textual_inversion/textual_inversion.py b/modules/textual_inversion/textual_inversion.py index fa0e33a2..25038a89 100644 --- a/modules/textual_inversion/textual_inversion.py +++ b/modules/textual_inversion/textual_inversion.py @@ -6,6 +6,7 @@ import torch import tqdm import html import datetime +import csv from PIL import Image, PngImagePlugin @@ -172,7 +173,7 @@ def create_embedding(name, num_vectors_per_token, init_text='*'): return fn -def train_embedding(embedding_name, learn_rate, data_root, log_directory, training_width, training_height, steps, create_image_every, save_embedding_every, template_file, save_image_with_stored_embedding, preview_image_prompt): +def train_embedding(embedding_name, learn_rate, data_root, log_directory, training_width, training_height, steps, create_image_every, save_embedding_every, write_csv_every, template_file, save_image_with_stored_embedding, preview_image_prompt): assert embedding_name, 'embedding not selected' shared.state.textinfo = "Initializing textual inversion training..." @@ -256,6 +257,20 @@ def train_embedding(embedding_name, learn_rate, data_root, log_directory, traini last_saved_file = os.path.join(embedding_dir, f'{embedding_name}-{embedding.step}.pt') embedding.save(last_saved_file) + if write_csv_every > 0 and log_directory is not None and embedding.step % write_csv_every == 0: + write_csv_header = False if os.path.exists(os.path.join(log_directory, "textual_inversion_loss.csv")) else True + + with open(os.path.join(log_directory, "textual_inversion_loss.csv"), "a+") as fout: + + csv_writer = csv.DictWriter(fout, fieldnames=["epoch", "epoch_step", "loss"]) + + if write_csv_header: + csv_writer.writeheader() + + csv_writer.writerow({"epoch": epoch_num + 1, + "epoch_step": epoch_step - 1, + "loss": f"{losses.mean():.7f}"}) + if embedding.step > 0 and images_dir is not None and embedding.step % create_image_every == 0: last_saved_image = os.path.join(images_dir, f'{embedding_name}-{embedding.step}.png') diff --git a/modules/ui.py b/modules/ui.py index e07ee0e1..1195c2f1 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1096,6 +1096,7 @@ def create_ui(wrap_gradio_gpu_call): training_height = gr.Slider(minimum=64, maximum=2048, step=64, label="Height", value=512) steps = gr.Number(label='Max steps', value=100000, precision=0) create_image_every = gr.Number(label='Save an image to log directory every N steps, 0 to disable', value=500, precision=0) + write_csv_every = gr.Number(label='Save an csv containing the loss to log directory every N steps, 0 to disable', value=500, precision=0) save_embedding_every = gr.Number(label='Save a copy of embedding to log directory every N steps, 0 to disable', value=500, precision=0) save_image_with_stored_embedding = gr.Checkbox(label='Save images with embedding in PNG chunks', value=True) preview_image_prompt = gr.Textbox(label='Preview prompt', value="") @@ -1174,6 +1175,7 @@ def create_ui(wrap_gradio_gpu_call): steps, create_image_every, save_embedding_every, + write_csv_every, template_file, save_image_with_stored_embedding, preview_image_prompt, @@ -1195,6 +1197,7 @@ def create_ui(wrap_gradio_gpu_call): steps, create_image_every, save_embedding_every, + write_csv_every, template_file, preview_image_prompt, ], -- cgit v1.2.3 From 54e0051bdd7dea7348825c09600ec61ea0771cb8 Mon Sep 17 00:00:00 2001 From: d8ahazard Date: Wed, 12 Oct 2022 18:17:26 -0500 Subject: Add drag/drop param loading. Drop an image or generational text onto the prompt bar, it loads the info for parsing. --- javascript/dragdrop.js | 3 +++ javascript/imageParams.js | 22 ++++++++++++++++++++++ modules/images.py | 20 ++++++++++++++++++++ modules/ui.py | 30 +++++++++++++++++++++++++++++- 4 files changed, 74 insertions(+), 1 deletion(-) create mode 100644 javascript/imageParams.js (limited to 'modules/ui.py') diff --git a/javascript/dragdrop.js b/javascript/dragdrop.js index 5aac57f7..cf900f50 100644 --- a/javascript/dragdrop.js +++ b/javascript/dragdrop.js @@ -53,6 +53,9 @@ window.document.addEventListener('dragover', e => { window.document.addEventListener('drop', e => { const target = e.composedPath()[0]; + if (target.placeholder === "Prompt") { + return; + } const imgWrap = target.closest('[data-testid="image"]'); if ( !imgWrap ) { return; diff --git a/javascript/imageParams.js b/javascript/imageParams.js new file mode 100644 index 00000000..f9d0c0aa --- /dev/null +++ b/javascript/imageParams.js @@ -0,0 +1,22 @@ +window.onload = (function(){ + window.addEventListener('drop', e => { + const target = e.composedPath()[0]; + const idx = selected_gallery_index(); + let prompt_target = "txt2img_prompt_image"; + if (idx === 1) { + prompt_target = "img2img_prompt_image"; + } + if (target.placeholder === "Prompt") { + e.stopPropagation(); + e.preventDefault(); + const imgParent = gradioApp().getElementById(prompt_target); + const files = e.dataTransfer.files; + const fileInput = imgParent.querySelector('input[type="file"]'); + if ( fileInput ) { + fileInput.files = files; + fileInput.dispatchEvent(new Event('change')); + } + } + }); + +}); \ No newline at end of file diff --git a/modules/images.py b/modules/images.py index c0a90676..f1155b7f 100644 --- a/modules/images.py +++ b/modules/images.py @@ -463,3 +463,23 @@ def save_image(image, path, basename, seed=None, prompt=None, extension='png', i txt_fullfn = None return fullfn, txt_fullfn + + +def image_data(image_path): + file, ext = os.path.splitext(image_path.name) + data = {} + if "png" in ext: + image = Image.open(image_path.name, "r") + print(f"Image data requested for {image_path.name} {image.format} of {type(image)}") + try: + data = image.text["parameters"] + except Exception as e: + print(f"Exception: {e}") + pass + print(f"Image data: {data}") + if "txt" in ext: + myfile = open(image_path.name, 'r') + data = myfile.read() + myfile.close() + + return data, None diff --git a/modules/ui.py b/modules/ui.py index 2b332267..dd793c39 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -431,7 +431,6 @@ def create_toprow(is_img2img): with gr.Column(scale=80): with gr.Row(): prompt = gr.Textbox(label="Prompt", elem_id=f"{id_part}_prompt", show_label=False, placeholder="Prompt", lines=2) - with gr.Column(scale=1, elem_id="roll_col"): roll = gr.Button(value=art_symbol, elem_id="roll", visible=len(shared.artist_db.artists) > 0) paste = gr.Button(value=paste_symbol, elem_id="paste") @@ -513,6 +512,7 @@ def create_ui(wrap_gradio_gpu_call): with gr.Blocks(analytics_enabled=False) as txt2img_interface: txt2img_prompt, roll, txt2img_prompt_style, txt2img_negative_prompt, txt2img_prompt_style2, submit, _, _, txt2img_prompt_style_apply, txt2img_save_style, paste, token_counter, token_button = create_toprow(is_img2img=False) dummy_component = gr.Label(visible=False) + txt_prompt_img = gr.File(label="", elem_id="txt2img_prompt_image", file_count="single", type="file", visible=False) with gr.Row(elem_id='txt2img_progress_row'): with gr.Column(scale=1): @@ -614,6 +614,18 @@ def create_ui(wrap_gradio_gpu_call): txt2img_prompt.submit(**txt2img_args) submit.click(**txt2img_args) + txt_prompt_img.change( + fn=modules.images.image_data, + # _js = "get_extras_tab_index", + inputs=[ + txt_prompt_img + ], + outputs=[ + txt2img_prompt, + txt_prompt_img + ] + ) + enable_hr.change( fn=lambda x: gr_show(x), inputs=[enable_hr], @@ -674,6 +686,9 @@ def create_ui(wrap_gradio_gpu_call): img2img_prompt, roll, img2img_prompt_style, img2img_negative_prompt, img2img_prompt_style2, submit, img2img_interrogate, img2img_deepbooru, img2img_prompt_style_apply, img2img_save_style, paste, token_counter, token_button = create_toprow(is_img2img=True) with gr.Row(elem_id='img2img_progress_row'): + img2img_prompt_img = gr.File(label="", elem_id="txt_prompt_image", file_count="single", type="file", + visible=False) + with gr.Column(scale=1): pass @@ -768,6 +783,18 @@ def create_ui(wrap_gradio_gpu_call): connect_reuse_seed(seed, reuse_seed, generation_info, dummy_component, is_subseed=False) connect_reuse_seed(subseed, reuse_subseed, generation_info, dummy_component, is_subseed=True) + img2img_prompt_img.change( + fn=modules.images.image_data, + # _js = "get_extras_tab_index", + inputs=[ + txt_prompt_img + ], + outputs=[ + img2img_prompt, + img2img_prompt_img + ] + ) + mask_mode.change( lambda mode, img: { init_img_with_mask: gr_show(mode == 0), @@ -956,6 +983,7 @@ def create_ui(wrap_gradio_gpu_call): button_id = "hidden_element" if shared.cmd_opts.hide_ui_dir_config else '' open_extras_folder = gr.Button('Open output directory', elem_id=button_id) + submit.click( fn=wrap_gradio_gpu_call(modules.extras.run_extras), _js="get_extras_tab_index", -- cgit v1.2.3 From aeacbac218c47f61f1d0d3f3b429c9038b8faf0f Mon Sep 17 00:00:00 2001 From: Greg Fuller Date: Tue, 11 Oct 2022 19:46:33 -0700 Subject: Fix save error --- modules/ui.py | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) (limited to 'modules/ui.py') diff --git a/modules/ui.py b/modules/ui.py index e07ee0e1..4fa405a9 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -148,7 +148,10 @@ def save_files(js_data, images, do_make_zip, index): is_grid = image_index < p.index_of_first_image i = 0 if is_grid else (image_index - p.index_of_first_image) - fullfn, txt_fullfn = save_image(image, path, "", seed=p.all_seeds[i], prompt=p.all_prompts[i], extension=extension, info=p.infotexts[image_index], grid=is_grid, p=p, save_to_dirs=save_to_dirs) + seed = p.all_seeds[i] if len(p.all_seeds) > 1 else p.seed + prompt = p.all_prompts[i] if len(p.all_prompts) > 1 else p.prompt + info = p.infotexts[image_index] if len(p.infotexts) > 1 else p.infotexts[0] + fullfn, txt_fullfn = save_image(image, path, "", seed=seed, prompt=prompt, extension=extension, info=info, grid=is_grid, p=p, save_to_dirs=save_to_dirs) filename = os.path.relpath(fullfn, path) filenames.append(filename) -- cgit v1.2.3 From 8711c2fe0135d5c160a57db41cb79ed1942ce7fa Mon Sep 17 00:00:00 2001 From: Greg Fuller Date: Wed, 12 Oct 2022 16:12:12 -0700 Subject: Fix metadata contents --- modules/ui.py | 5 +---- scripts/xy_grid.py | 52 ++++++++++++++++++++++++++++++++++------------------ 2 files changed, 35 insertions(+), 22 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/ui.py b/modules/ui.py index 4fa405a9..e07ee0e1 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -148,10 +148,7 @@ def save_files(js_data, images, do_make_zip, index): is_grid = image_index < p.index_of_first_image i = 0 if is_grid else (image_index - p.index_of_first_image) - seed = p.all_seeds[i] if len(p.all_seeds) > 1 else p.seed - prompt = p.all_prompts[i] if len(p.all_prompts) > 1 else p.prompt - info = p.infotexts[image_index] if len(p.infotexts) > 1 else p.infotexts[0] - fullfn, txt_fullfn = save_image(image, path, "", seed=seed, prompt=prompt, extension=extension, info=info, grid=is_grid, p=p, save_to_dirs=save_to_dirs) + fullfn, txt_fullfn = save_image(image, path, "", seed=p.all_seeds[i], prompt=p.all_prompts[i], extension=extension, info=p.infotexts[image_index], grid=is_grid, p=p, save_to_dirs=save_to_dirs) filename = os.path.relpath(fullfn, path) filenames.append(filename) diff --git a/scripts/xy_grid.py b/scripts/xy_grid.py index 14edacc1..02931ae6 100644 --- a/scripts/xy_grid.py +++ b/scripts/xy_grid.py @@ -176,13 +176,16 @@ axis_options = [ def draw_xy_grid(p, xs, ys, x_labels, y_labels, cell, draw_legend, include_lone_images): - res = [] - successful_images = [] - ver_texts = [[images.GridAnnotation(y)] for y in y_labels] hor_texts = [[images.GridAnnotation(x)] for x in x_labels] - first_processed = None + # Temporary list of all the images that are generated to be populated into the grid. + # Will be filled with empty images for any individual step that fails to process properly + image_cache = [] + + processed_result = None + cell_mode = "P" + cell_size = (1,1) state.job_count = len(xs) * len(ys) * p.n_iter @@ -190,26 +193,39 @@ def draw_xy_grid(p, xs, ys, x_labels, y_labels, cell, draw_legend, include_lone_ for ix, x in enumerate(xs): state.job = f"{ix + iy * len(xs) + 1} out of {len(xs) * len(ys)}" - processed = cell(x, y) - if first_processed is None: - first_processed = processed - + processed:Processed = cell(x, y) try: - processed_image = processed.images[0] - res.append(processed_image) - successful_images.append(processed_image) + # this dereference will throw an exception if the image was not processed + # (this happens in cases such as if the user stops the process from the UI) + processed_image = processed.images[0] + + if processed_result is None: + # Use our first valid processed result as a template container to hold our full results + processed_result = copy(processed) + cell_mode = processed_image.mode + cell_size = processed_image.size + processed_result.images = [Image.new(cell_mode, cell_size)] + + image_cache.append(processed_image) + if include_lone_images: + processed_result.images.append(processed_image) + processed_result.all_prompts.append(processed.prompt) + processed_result.all_seeds.append(processed.seed) + processed_result.infotexts.append(processed.infotexts[0]) except: - res.append(Image.new(res[0].mode, res[0].size)) + image_cache.append(Image.new(cell_mode, cell_size)) + + if not processed_result: + print("Unexpected error: draw_xy_grid failed to return even a single processed image") + return Processed() - grid = images.image_grid(res, rows=len(ys)) + grid = images.image_grid(image_cache, rows=len(ys)) if draw_legend: - grid = images.draw_grid_annotations(grid, res[0].width, res[0].height, hor_texts, ver_texts) + grid = images.draw_grid_annotations(grid, cell_size[0], cell_size[1], hor_texts, ver_texts) - first_processed.images = [grid] - if include_lone_images: - first_processed.images += successful_images + processed_result.images[0] = grid - return first_processed + return processed_result re_range = re.compile(r"\s*([+-]?\s*\d+)\s*-\s*([+-]?\s*\d+)(?:\s*\(([+-]\d+)\s*\))?\s*") -- cgit v1.2.3 From a3f02e4690844715a510b7bc857a0971dd05c4d8 Mon Sep 17 00:00:00 2001 From: Greg Fuller Date: Wed, 12 Oct 2022 16:48:53 -0700 Subject: fix prompt in log.csv --- modules/ui.py | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/ui.py b/modules/ui.py index e07ee0e1..edb4dab1 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -139,6 +139,8 @@ def save_files(js_data, images, do_make_zip, index): if at_start: writer.writerow(["prompt", "seed", "width", "height", "sampler", "cfgs", "steps", "filename", "negative_prompt"]) + log_prompt=data["prompt"] + log_seed=data["seed"] for image_index, filedata in enumerate(images, start_index): if filedata.startswith("data:image/png;base64,"): filedata = filedata[len("data:image/png;base64,"):] @@ -148,7 +150,9 @@ def save_files(js_data, images, do_make_zip, index): is_grid = image_index < p.index_of_first_image i = 0 if is_grid else (image_index - p.index_of_first_image) - fullfn, txt_fullfn = save_image(image, path, "", seed=p.all_seeds[i], prompt=p.all_prompts[i], extension=extension, info=p.infotexts[image_index], grid=is_grid, p=p, save_to_dirs=save_to_dirs) + log_seed=p.all_seeds[i] + log_prompt=p.all_prompts[i] + fullfn, txt_fullfn = save_image(image, path, "", seed=log_seed, prompt=log_prompt, extension=extension, info=p.infotexts[image_index], grid=is_grid, p=p, save_to_dirs=save_to_dirs) filename = os.path.relpath(fullfn, path) filenames.append(filename) @@ -157,7 +161,7 @@ def save_files(js_data, images, do_make_zip, index): filenames.append(os.path.basename(txt_fullfn)) fullfns.append(txt_fullfn) - writer.writerow([data["prompt"], data["seed"], data["width"], data["height"], data["sampler"], data["cfg_scale"], data["steps"], filenames[0], data["negative_prompt"]]) + writer.writerow([log_prompt, log_seed, data["width"], data["height"], data["sampler"], data["cfg_scale"], data["steps"], filenames[0], data["negative_prompt"]]) # Make Zip if do_make_zip: -- cgit v1.2.3 From fed7f0e281a42ea962bbe422e018468bafa6f1e6 Mon Sep 17 00:00:00 2001 From: Greg Fuller Date: Wed, 12 Oct 2022 23:09:30 -0700 Subject: Revert "fix prompt in log.csv" This reverts commit e4b5d1696429ab78dae9779420ce6ec4cd9c5f67. --- modules/ui.py | 8 ++------ 1 file changed, 2 insertions(+), 6 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/ui.py b/modules/ui.py index edb4dab1..e07ee0e1 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -139,8 +139,6 @@ def save_files(js_data, images, do_make_zip, index): if at_start: writer.writerow(["prompt", "seed", "width", "height", "sampler", "cfgs", "steps", "filename", "negative_prompt"]) - log_prompt=data["prompt"] - log_seed=data["seed"] for image_index, filedata in enumerate(images, start_index): if filedata.startswith("data:image/png;base64,"): filedata = filedata[len("data:image/png;base64,"):] @@ -150,9 +148,7 @@ def save_files(js_data, images, do_make_zip, index): is_grid = image_index < p.index_of_first_image i = 0 if is_grid else (image_index - p.index_of_first_image) - log_seed=p.all_seeds[i] - log_prompt=p.all_prompts[i] - fullfn, txt_fullfn = save_image(image, path, "", seed=log_seed, prompt=log_prompt, extension=extension, info=p.infotexts[image_index], grid=is_grid, p=p, save_to_dirs=save_to_dirs) + fullfn, txt_fullfn = save_image(image, path, "", seed=p.all_seeds[i], prompt=p.all_prompts[i], extension=extension, info=p.infotexts[image_index], grid=is_grid, p=p, save_to_dirs=save_to_dirs) filename = os.path.relpath(fullfn, path) filenames.append(filename) @@ -161,7 +157,7 @@ def save_files(js_data, images, do_make_zip, index): filenames.append(os.path.basename(txt_fullfn)) fullfns.append(txt_fullfn) - writer.writerow([log_prompt, log_seed, data["width"], data["height"], data["sampler"], data["cfg_scale"], data["steps"], filenames[0], data["negative_prompt"]]) + writer.writerow([data["prompt"], data["seed"], data["width"], data["height"], data["sampler"], data["cfg_scale"], data["steps"], filenames[0], data["negative_prompt"]]) # Make Zip if do_make_zip: -- cgit v1.2.3 From bb7baf6b9cb6b4b9fa09b6f07ef997db32fe6e58 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Thu, 13 Oct 2022 16:07:18 +0300 Subject: add option to change what's shown in quicksettings bar --- javascript/hints.js | 2 ++ modules/shared.py | 4 ++-- modules/ui.py | 16 +++++++++------- style.css | 1 + 4 files changed, 14 insertions(+), 9 deletions(-) (limited to 'modules/ui.py') diff --git a/javascript/hints.js b/javascript/hints.js index 32f10fde..06bbd9e2 100644 --- a/javascript/hints.js +++ b/javascript/hints.js @@ -84,6 +84,8 @@ titles = { "Filename word regex": "This regular expression will be used extract words from filename, and they will be joined using the option below into label text used for training. Leave empty to keep filename text as it is.", "Filename join string": "This string will be used to hoin split words into a single line if the option above is enabled.", + + "Quicksettings list": "List of setting names, separated by commas, for settings that should go to the quick access bar at the top, rather than the usual stetting tab. See modules/shared.py for setting names. Requires restart to apply." } diff --git a/modules/shared.py b/modules/shared.py index 5f6101a4..4d3ed625 100644 --- a/modules/shared.py +++ b/modules/shared.py @@ -152,7 +152,6 @@ class OptionInfo: self.component_args = component_args self.onchange = onchange self.section = None - self.show_on_main_page = show_on_main_page def options_section(section_identifier, options_dict): @@ -237,7 +236,7 @@ options_templates.update(options_section(('training', "Training"), { })) options_templates.update(options_section(('sd', "Stable Diffusion"), { - "sd_model_checkpoint": OptionInfo(None, "Stable Diffusion checkpoint", gr.Dropdown, lambda: {"choices": modules.sd_models.checkpoint_tiles()}, show_on_main_page=True), + "sd_model_checkpoint": OptionInfo(None, "Stable Diffusion checkpoint", gr.Dropdown, lambda: {"choices": modules.sd_models.checkpoint_tiles()}), "sd_hypernetwork": OptionInfo("None", "Stable Diffusion finetune hypernetwork", gr.Dropdown, lambda: {"choices": ["None"] + [x for x in hypernetworks.keys()]}), "img2img_color_correction": OptionInfo(False, "Apply color correction to img2img results to match original colors."), "save_images_before_color_correction": OptionInfo(False, "Save a copy of image before applying color correction to img2img results"), @@ -250,6 +249,7 @@ options_templates.update(options_section(('sd', "Stable Diffusion"), { "filter_nsfw": OptionInfo(False, "Filter NSFW content"), 'CLIP_stop_at_last_layers': OptionInfo(1, "Stop At last layers of CLIP model", gr.Slider, {"minimum": 1, "maximum": 12, "step": 1}), "random_artist_categories": OptionInfo([], "Allowed categories for random artists selection when using the Roll button", gr.CheckboxGroup, {"choices": artist_db.categories()}), + 'quicksettings': OptionInfo("sd_model_checkpoint", "Quicksettings list"), })) options_templates.update(options_section(('interrogate', "Interrogate Options"), { diff --git a/modules/ui.py b/modules/ui.py index e07ee0e1..a0529860 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1305,6 +1305,9 @@ Requested path was: {f} settings_cols = 3 items_per_col = int(len(opts.data_labels) * 0.9 / settings_cols) + quicksettings_names = [x.strip() for x in opts.quicksettings.split(",")] + quicksettings_names = set(x for x in quicksettings_names if x != 'quicksettings') + quicksettings_list = [] cols_displayed = 0 @@ -1329,7 +1332,7 @@ Requested path was: {f} gr.HTML(elem_id="settings_header_text_{}".format(item.section[0]), value='

{}

'.format(item.section[1])) - if item.show_on_main_page: + if k in quicksettings_names: quicksettings_list.append((i, k, item)) components.append(dummy_component) else: @@ -1338,7 +1341,11 @@ Requested path was: {f} components.append(component) items_displayed += 1 - request_notifications = gr.Button(value='Request browser notifications', elem_id="request_notifications") + with gr.Row(): + request_notifications = gr.Button(value='Request browser notifications', elem_id="request_notifications") + reload_script_bodies = gr.Button(value='Reload custom script bodies (No ui updates, No restart)', variant='secondary') + restart_gradio = gr.Button(value='Restart Gradio and Refresh components (Custom Scripts, ui.py, js and css only)', variant='primary') + request_notifications.click( fn=lambda: None, inputs=[], @@ -1346,10 +1353,6 @@ Requested path was: {f} _js='function(){}' ) - with gr.Row(): - reload_script_bodies = gr.Button(value='Reload custom script bodies (No ui updates, No restart)', variant='secondary') - restart_gradio = gr.Button(value='Restart Gradio and Refresh components (Custom Scripts, ui.py, js and css only)', variant='primary') - def reload_scripts(): modules.scripts.reload_script_body_only() @@ -1364,7 +1367,6 @@ Requested path was: {f} shared.state.interrupt() settings_interface.gradio_ref.do_restart = True - restart_gradio.click( fn=request_restart, inputs=[], diff --git a/style.css b/style.css index e6fa10b4..55c41971 100644 --- a/style.css +++ b/style.css @@ -488,6 +488,7 @@ input[type="range"]{ #quicksettings > div > div{ max-width: 32em; padding: 0; + margin-right: 0.75em; } canvas[key="mask"] { -- cgit v1.2.3 From a10b0e11fc22cc67b6a3664f2ddd17425d8433a8 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Thu, 13 Oct 2022 19:22:41 +0300 Subject: options to refresh list of models and hypernetworks --- modules/shared.py | 9 +++++---- modules/ui.py | 33 +++++++++++++++++++++++++++++---- style.css | 21 ++++++++++++++++++++- 3 files changed, 54 insertions(+), 9 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/shared.py b/modules/shared.py index 4d3ed625..d8e3a286 100644 --- a/modules/shared.py +++ b/modules/shared.py @@ -13,7 +13,7 @@ import modules.memmon import modules.sd_models import modules.styles import modules.devices as devices -from modules import sd_samplers +from modules import sd_samplers, sd_models from modules.hypernetworks import hypernetwork from modules.paths import models_path, script_path, sd_path @@ -145,13 +145,14 @@ def realesrgan_models_names(): class OptionInfo: - def __init__(self, default=None, label="", component=None, component_args=None, onchange=None, show_on_main_page=False): + def __init__(self, default=None, label="", component=None, component_args=None, onchange=None, show_on_main_page=False, refresh=None): self.default = default self.label = label self.component = component self.component_args = component_args self.onchange = onchange self.section = None + self.refresh = refresh def options_section(section_identifier, options_dict): @@ -236,8 +237,8 @@ options_templates.update(options_section(('training', "Training"), { })) options_templates.update(options_section(('sd', "Stable Diffusion"), { - "sd_model_checkpoint": OptionInfo(None, "Stable Diffusion checkpoint", gr.Dropdown, lambda: {"choices": modules.sd_models.checkpoint_tiles()}), - "sd_hypernetwork": OptionInfo("None", "Stable Diffusion finetune hypernetwork", gr.Dropdown, lambda: {"choices": ["None"] + [x for x in hypernetworks.keys()]}), + "sd_model_checkpoint": OptionInfo(None, "Stable Diffusion checkpoint", gr.Dropdown, lambda: {"choices": modules.sd_models.checkpoint_tiles()}, refresh=sd_models.list_models), + "sd_hypernetwork": OptionInfo("None", "Stable Diffusion finetune hypernetwork", gr.Dropdown, lambda: {"choices": ["None"] + [x for x in hypernetworks.keys()]}, refresh=reload_hypernetworks), "img2img_color_correction": OptionInfo(False, "Apply color correction to img2img results to match original colors."), "save_images_before_color_correction": OptionInfo(False, "Save a copy of image before applying color correction to img2img results"), "img2img_fix_steps": OptionInfo(False, "With img2img, do exactly the amount of steps the slider specifies (normally you'd do less with less denoising)."), diff --git a/modules/ui.py b/modules/ui.py index a0529860..0a58f6be 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -78,6 +78,8 @@ reuse_symbol = '\u267b\ufe0f' # ♻️ art_symbol = '\U0001f3a8' # 🎨 paste_symbol = '\u2199\ufe0f' # ↙ folder_symbol = '\U0001f4c2' # 📂 +refresh_symbol = '\U0001f504' # 🔄 + def plaintext_to_html(text): text = "

" + "
\n".join([f"{html.escape(x)}" for x in text.split('\n')]) + "

" @@ -1210,8 +1212,7 @@ def create_ui(wrap_gradio_gpu_call): outputs=[], ) - - def create_setting_component(key): + def create_setting_component(key, is_quicksettings=False): def fun(): return opts.data[key] if key in opts.data else opts.data_labels[key].default @@ -1231,7 +1232,31 @@ def create_ui(wrap_gradio_gpu_call): else: raise Exception(f'bad options item type: {str(t)} for key {key}') - return comp(label=info.label, value=fun, **(args or {})) + if info.refresh is not None: + if is_quicksettings: + res = comp(label=info.label, value=fun, **(args or {})) + refresh_button = gr.Button(value=refresh_symbol, elem_id="refresh_"+key) + else: + with gr.Row(variant="compact"): + res = comp(label=info.label, value=fun, **(args or {})) + refresh_button = gr.Button(value=refresh_symbol, elem_id="refresh_" + key) + + def refresh(): + info.refresh() + refreshed_args = info.component_args() if callable(info.component_args) else info.component_args + res.choices = refreshed_args["choices"] + return gr.update(**(refreshed_args or {})) + + refresh_button.click( + fn=refresh, + inputs=[], + outputs=[res], + ) + else: + res = comp(label=info.label, value=fun, **(args or {})) + + + return res components = [] component_dict = {} @@ -1401,7 +1426,7 @@ Requested path was: {f} with gr.Blocks(css=css, analytics_enabled=False, title="Stable Diffusion") as demo: with gr.Row(elem_id="quicksettings"): for i, k, item in quicksettings_list: - component = create_setting_component(k) + component = create_setting_component(k, is_quicksettings=True) component_dict[k] = component settings_interface.gradio_ref = demo diff --git a/style.css b/style.css index 55c41971..ad2a52cc 100644 --- a/style.css +++ b/style.css @@ -228,6 +228,8 @@ fieldset span.text-gray-500, .gr-block.gr-box span.text-gray-500, label.block s border-top: 1px solid #eee; border-left: 1px solid #eee; border-right: 1px solid #eee; + + z-index: 300; } .dark fieldset span.text-gray-500, .dark .gr-block.gr-box span.text-gray-500, .dark label.block span{ @@ -480,17 +482,30 @@ input[type="range"]{ background: #a55000; } +#quicksettings { + gap: 0.4em; +} + #quicksettings > div{ border: none; background: none; + flex: unset; + gap: 0.5em; } #quicksettings > div > div{ max-width: 32em; + min-width: 24em; padding: 0; - margin-right: 0.75em; } +#refresh_sd_model_checkpoint, #refresh_sd_hypernetwork{ + max-width: 2.5em; + min-width: 2.5em; + height: 2.4em; +} + + canvas[key="mask"] { z-index: 12 !important; filter: invert(); @@ -507,3 +522,7 @@ canvas[key="mask"] { z-index: 200; width: 8em; } + +.row.gr-compact{ + overflow: visible; +} -- cgit v1.2.3 From 354ef0da3b1f0fa5c113d04b6c79e3908c848d23 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Thu, 13 Oct 2022 20:12:37 +0300 Subject: add hypernetwork multipliers --- modules/hypernetworks/hypernetwork.py | 8 +++++++- modules/shared.py | 5 ++++- modules/ui.py | 5 ++++- scripts/xy_grid.py | 9 ++++++++- style.css | 3 +++ webui.py | 2 +- 6 files changed, 27 insertions(+), 5 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/hypernetworks/hypernetwork.py b/modules/hypernetworks/hypernetwork.py index b6c06d49..f1248bb7 100644 --- a/modules/hypernetworks/hypernetwork.py +++ b/modules/hypernetworks/hypernetwork.py @@ -18,6 +18,8 @@ from modules.textual_inversion.learn_schedule import LearnRateScheduler class HypernetworkModule(torch.nn.Module): + multiplier = 1.0 + def __init__(self, dim, state_dict=None): super().__init__() @@ -36,7 +38,11 @@ class HypernetworkModule(torch.nn.Module): self.to(devices.device) def forward(self, x): - return x + (self.linear2(self.linear1(x))) + return x + (self.linear2(self.linear1(x))) * self.multiplier + + +def apply_strength(value=None): + HypernetworkModule.multiplier = value if value is not None else shared.opts.sd_hypernetwork_strength class Hypernetwork: diff --git a/modules/shared.py b/modules/shared.py index d8e3a286..5901e605 100644 --- a/modules/shared.py +++ b/modules/shared.py @@ -238,7 +238,8 @@ options_templates.update(options_section(('training', "Training"), { options_templates.update(options_section(('sd', "Stable Diffusion"), { "sd_model_checkpoint": OptionInfo(None, "Stable Diffusion checkpoint", gr.Dropdown, lambda: {"choices": modules.sd_models.checkpoint_tiles()}, refresh=sd_models.list_models), - "sd_hypernetwork": OptionInfo("None", "Stable Diffusion finetune hypernetwork", gr.Dropdown, lambda: {"choices": ["None"] + [x for x in hypernetworks.keys()]}, refresh=reload_hypernetworks), + "sd_hypernetwork": OptionInfo("None", "Hypernetwork", gr.Dropdown, lambda: {"choices": ["None"] + [x for x in hypernetworks.keys()]}, refresh=reload_hypernetworks), + "sd_hypernetwork_strength": OptionInfo(1.0, "Hypernetwork strength", gr.Slider, {"minimum": 0.0, "maximum": 1.0, "step": 0.001}), "img2img_color_correction": OptionInfo(False, "Apply color correction to img2img results to match original colors."), "save_images_before_color_correction": OptionInfo(False, "Save a copy of image before applying color correction to img2img results"), "img2img_fix_steps": OptionInfo(False, "With img2img, do exactly the amount of steps the slider specifies (normally you'd do less with less denoising)."), @@ -348,6 +349,8 @@ class Options: item = self.data_labels.get(key) item.onchange = func + func() + def dumpjson(self): d = {k: self.data.get(k, self.data_labels.get(k).default) for k in self.data_labels.keys()} return json.dumps(d) diff --git a/modules/ui.py b/modules/ui.py index 0a58f6be..673014f2 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1244,7 +1244,10 @@ def create_ui(wrap_gradio_gpu_call): def refresh(): info.refresh() refreshed_args = info.component_args() if callable(info.component_args) else info.component_args - res.choices = refreshed_args["choices"] + + for k, v in refreshed_args.items(): + setattr(res, k, v) + return gr.update(**(refreshed_args or {})) refresh_button.click( diff --git a/scripts/xy_grid.py b/scripts/xy_grid.py index 02931ae6..efb63af5 100644 --- a/scripts/xy_grid.py +++ b/scripts/xy_grid.py @@ -107,6 +107,10 @@ def apply_hypernetwork(p, x, xs): hypernetwork.load_hypernetwork(name) +def apply_hypernetwork_strength(p, x, xs): + hypernetwork.apply_strength(x) + + def confirm_hypernetworks(p, xs): for x in xs: if x.lower() in ["", "none"]: @@ -165,6 +169,7 @@ axis_options = [ AxisOption("Sampler", str, apply_sampler, format_value, confirm_samplers), AxisOption("Checkpoint name", str, apply_checkpoint, format_value, confirm_checkpoints), AxisOption("Hypernetwork", str, apply_hypernetwork, format_value, confirm_hypernetworks), + AxisOption("Hypernet str.", float, apply_hypernetwork_strength, format_value_add_label, None), AxisOption("Sigma Churn", float, apply_field("s_churn"), format_value_add_label, None), AxisOption("Sigma min", float, apply_field("s_tmin"), format_value_add_label, None), AxisOption("Sigma max", float, apply_field("s_tmax"), format_value_add_label, None), @@ -250,7 +255,7 @@ class Script(scripts.Script): y_values = gr.Textbox(label="Y values", visible=False, lines=1) draw_legend = gr.Checkbox(label='Draw legend', value=True) - include_lone_images = gr.Checkbox(label='Include Separate Images', value=True) + include_lone_images = gr.Checkbox(label='Include Separate Images', value=False) no_fixed_seeds = gr.Checkbox(label='Keep -1 for seeds', value=False) return [x_type, x_values, y_type, y_values, draw_legend, include_lone_images, no_fixed_seeds] @@ -377,6 +382,8 @@ class Script(scripts.Script): modules.sd_models.reload_model_weights(shared.sd_model) hypernetwork.load_hypernetwork(opts.sd_hypernetwork) + hypernetwork.apply_strength() + opts.data["CLIP_stop_at_last_layers"] = CLIP_stop_at_last_layers diff --git a/style.css b/style.css index ad2a52cc..aa3d379c 100644 --- a/style.css +++ b/style.css @@ -522,6 +522,9 @@ canvas[key="mask"] { z-index: 200; width: 8em; } +#quicksettings .gr-box > div > div > input.gr-text-input { + top: -1.12em; +} .row.gr-compact{ overflow: visible; diff --git a/webui.py b/webui.py index 33ba7905..fe0ce321 100644 --- a/webui.py +++ b/webui.py @@ -72,7 +72,6 @@ def wrap_gradio_gpu_call(func, extra_outputs=None): return modules.ui.wrap_gradio_call(f, extra_outputs=extra_outputs) - def initialize(): modelloader.cleanup_models() modules.sd_models.setup_model() @@ -86,6 +85,7 @@ def initialize(): shared.sd_model = modules.sd_models.load_model() shared.opts.onchange("sd_model_checkpoint", wrap_queued_call(lambda: modules.sd_models.reload_model_weights(shared.sd_model))) shared.opts.onchange("sd_hypernetwork", wrap_queued_call(lambda: modules.hypernetworks.hypernetwork.load_hypernetwork(shared.opts.sd_hypernetwork))) + shared.opts.onchange("sd_hypernetwork_strength", modules.hypernetworks.hypernetwork.apply_strength) def webui(): -- cgit v1.2.3 From 08b3f7aef15f74f4d2254b1274dd66fcc7940348 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Thu, 13 Oct 2022 20:42:27 +0300 Subject: emergency fix for broken send to buttons --- javascript/ui.js | 8 ++++---- modules/ui.py | 2 +- 2 files changed, 5 insertions(+), 5 deletions(-) (limited to 'modules/ui.py') diff --git a/javascript/ui.js b/javascript/ui.js index 4100944e..0f8fe68e 100644 --- a/javascript/ui.js +++ b/javascript/ui.js @@ -33,27 +33,27 @@ function args_to_array(args){ } function switch_to_txt2img(){ - gradioApp().querySelectorAll('button')[0].click(); + gradioApp().querySelector('#tabs').querySelectorAll('button')[0].click(); return args_to_array(arguments); } function switch_to_img2img_img2img(){ - gradioApp().querySelectorAll('button')[1].click(); + gradioApp().querySelector('#tabs').querySelectorAll('button')[1].click(); gradioApp().getElementById('mode_img2img').querySelectorAll('button')[0].click(); return args_to_array(arguments); } function switch_to_img2img_inpaint(){ - gradioApp().querySelectorAll('button')[1].click(); + gradioApp().querySelector('#tabs').querySelectorAll('button')[1].click(); gradioApp().getElementById('mode_img2img').querySelectorAll('button')[1].click(); return args_to_array(arguments); } function switch_to_extras(){ - gradioApp().querySelectorAll('button')[2].click(); + gradioApp().querySelector('#tabs').querySelectorAll('button')[2].click(); return args_to_array(arguments); } diff --git a/modules/ui.py b/modules/ui.py index 673014f2..7446439d 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1434,7 +1434,7 @@ Requested path was: {f} settings_interface.gradio_ref = demo - with gr.Tabs() as tabs: + with gr.Tabs(elem_id="tabs") as tabs: for interface, label, ifid in interfaces: with gr.TabItem(label, id=ifid, elem_id='tab_' + ifid): interface.render() -- cgit v1.2.3 From fdecb636855748e03efc40c846a0043800aadfcc Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Fri, 14 Oct 2022 09:05:06 +0300 Subject: add an ability to merge three checkpoints --- javascript/hints.js | 5 ++++- modules/extras.py | 29 +++++++++++++++++++++-------- modules/ui.py | 11 +++++++---- 3 files changed, 32 insertions(+), 13 deletions(-) (limited to 'modules/ui.py') diff --git a/javascript/hints.js b/javascript/hints.js index 94438c5c..af010a59 100644 --- a/javascript/hints.js +++ b/javascript/hints.js @@ -85,7 +85,10 @@ titles = { "Filename word regex": "This regular expression will be used extract words from filename, and they will be joined using the option below into label text used for training. Leave empty to keep filename text as it is.", "Filename join string": "This string will be used to join split words into a single line if the option above is enabled.", - "Quicksettings list": "List of setting names, separated by commas, for settings that should go to the quick access bar at the top, rather than the usual setting tab. See modules/shared.py for setting names. Requires restarting to apply." + "Quicksettings list": "List of setting names, separated by commas, for settings that should go to the quick access bar at the top, rather than the usual setting tab. See modules/shared.py for setting names. Requires restarting to apply.", + + "Weighted Sum": "Result = A * (1 - M) + B * M", + "Add difference": "Result = A + (B - C) * (1 - M)", } diff --git a/modules/extras.py b/modules/extras.py index b24d7de3..532d869f 100644 --- a/modules/extras.py +++ b/modules/extras.py @@ -159,48 +159,61 @@ def run_pnginfo(image): return '', geninfo, info -def run_modelmerger(primary_model_name, secondary_model_name, interp_method, interp_amount, save_as_half, custom_name): +def run_modelmerger(primary_model_name, secondary_model_name, teritary_model_name, interp_method, interp_amount, save_as_half, custom_name): # Linear interpolation (https://en.wikipedia.org/wiki/Linear_interpolation) - def weighted_sum(theta0, theta1, alpha): + def weighted_sum(theta0, theta1, theta2, alpha): return ((1 - alpha) * theta0) + (alpha * theta1) # Smoothstep (https://en.wikipedia.org/wiki/Smoothstep) - def sigmoid(theta0, theta1, alpha): + def sigmoid(theta0, theta1, theta2, alpha): alpha = alpha * alpha * (3 - (2 * alpha)) return theta0 + ((theta1 - theta0) * alpha) # Inverse Smoothstep (https://en.wikipedia.org/wiki/Smoothstep) - def inv_sigmoid(theta0, theta1, alpha): + def inv_sigmoid(theta0, theta1, theta2, alpha): import math alpha = 0.5 - math.sin(math.asin(1.0 - 2.0 * alpha) / 3.0) return theta0 + ((theta1 - theta0) * alpha) + def add_difference(theta0, theta1, theta2, alpha): + return theta0 + (theta1 - theta2) * (1.0 - alpha) + primary_model_info = sd_models.checkpoints_list[primary_model_name] secondary_model_info = sd_models.checkpoints_list[secondary_model_name] + teritary_model_info = sd_models.checkpoints_list.get(teritary_model_name, None) print(f"Loading {primary_model_info.filename}...") primary_model = torch.load(primary_model_info.filename, map_location='cpu') + theta_0 = sd_models.get_state_dict_from_checkpoint(primary_model) print(f"Loading {secondary_model_info.filename}...") secondary_model = torch.load(secondary_model_info.filename, map_location='cpu') - - theta_0 = sd_models.get_state_dict_from_checkpoint(primary_model) theta_1 = sd_models.get_state_dict_from_checkpoint(secondary_model) + if teritary_model_info is not None: + print(f"Loading {teritary_model_info.filename}...") + teritary_model = torch.load(teritary_model_info.filename, map_location='cpu') + theta_2 = sd_models.get_state_dict_from_checkpoint(teritary_model) + else: + theta_2 = None + theta_funcs = { "Weighted Sum": weighted_sum, "Sigmoid": sigmoid, "Inverse Sigmoid": inv_sigmoid, + "Add difference": add_difference, } theta_func = theta_funcs[interp_method] print(f"Merging...") + for key in tqdm.tqdm(theta_0.keys()): if 'model' in key and key in theta_1: - theta_0[key] = theta_func(theta_0[key], theta_1[key], (float(1.0) - interp_amount)) # Need to reverse the interp_amount to match the desired mix ration in the merged checkpoint + theta_0[key] = theta_func(theta_0[key], theta_1[key], theta_2[key] if theta_2 else None, (float(1.0) - interp_amount)) # Need to reverse the interp_amount to match the desired mix ration in the merged checkpoint if save_as_half: theta_0[key] = theta_0[key].half() + # I believe this part should be discarded, but I'll leave it for now until I am sure for key in theta_1.keys(): if 'model' in key and key not in theta_0: theta_0[key] = theta_1[key] @@ -219,4 +232,4 @@ def run_modelmerger(primary_model_name, secondary_model_name, interp_method, int sd_models.list_models() print(f"Checkpoint saved.") - return ["Checkpoint saved to " + output_modelname] + [gr.Dropdown.update(choices=sd_models.checkpoint_tiles()) for _ in range(3)] + return ["Checkpoint saved to " + output_modelname] + [gr.Dropdown.update(choices=sd_models.checkpoint_tiles()) for _ in range(4)] diff --git a/modules/ui.py b/modules/ui.py index 7446439d..220fb80b 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1024,11 +1024,12 @@ def create_ui(wrap_gradio_gpu_call): gr.HTML(value="

A merger of the two checkpoints will be generated in your checkpoint directory.

") with gr.Row(): - primary_model_name = gr.Dropdown(modules.sd_models.checkpoint_tiles(), elem_id="modelmerger_primary_model_name", label="Primary Model Name") - secondary_model_name = gr.Dropdown(modules.sd_models.checkpoint_tiles(), elem_id="modelmerger_secondary_model_name", label="Secondary Model Name") + primary_model_name = gr.Dropdown(modules.sd_models.checkpoint_tiles(), elem_id="modelmerger_primary_model_name", label="Primary model (A)") + secondary_model_name = gr.Dropdown(modules.sd_models.checkpoint_tiles(), elem_id="modelmerger_secondary_model_name", label="Secondary model (B)") + tertiary_model_name = gr.Dropdown(modules.sd_models.checkpoint_tiles(), elem_id="modelmerger_tertiary_model_name", label="Tertiary model (C)") custom_name = gr.Textbox(label="Custom Name (Optional)") - interp_amount = gr.Slider(minimum=0.0, maximum=1.0, step=0.05, label='Interpolation Amount', value=0.3) - interp_method = gr.Radio(choices=["Weighted Sum", "Sigmoid", "Inverse Sigmoid"], value="Weighted Sum", label="Interpolation Method") + interp_amount = gr.Slider(minimum=0.0, maximum=1.0, step=0.05, label='Interpolation amount (1 - M)', value=0.3) + interp_method = gr.Radio(choices=["Weighted Sum", "Sigmoid", "Inverse Sigmoid", "Add difference"], value="Weighted Sum", label="Interpolation Method") save_as_half = gr.Checkbox(value=False, label="Save as float16") modelmerger_merge = gr.Button(elem_id="modelmerger_merge", label="Merge", variant='primary') @@ -1473,6 +1474,7 @@ Requested path was: {f} inputs=[ primary_model_name, secondary_model_name, + tertiary_model_name, interp_method, interp_amount, save_as_half, @@ -1482,6 +1484,7 @@ Requested path was: {f} submit_result, primary_model_name, secondary_model_name, + tertiary_model_name, component_dict['sd_model_checkpoint'], ] ) -- cgit v1.2.3 From b2261b53ae4ad01b3713bc73ff62ab7b6f479e26 Mon Sep 17 00:00:00 2001 From: Buckzor Date: Thu, 13 Oct 2022 17:07:06 +0100 Subject: Added first_pass_width and height as adjustable inputs to "High Res Fix" --- modules/processing.py | 6 ++++-- modules/txt2img.py | 5 ++++- modules/ui.py | 6 ++++++ 3 files changed, 14 insertions(+), 3 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/processing.py b/modules/processing.py index d5172f00..abbfdf98 100644 --- a/modules/processing.py +++ b/modules/processing.py @@ -506,11 +506,13 @@ class StableDiffusionProcessingTxt2Img(StableDiffusionProcessing): firstphase_width_truncated = 0 firstphase_height_truncated = 0 - def __init__(self, enable_hr=False, scale_latent=True, denoising_strength=0.75, **kwargs): + def __init__(self, enable_hr=False, scale_latent=True, denoising_strength=0.75, first_pass_width=512, first_pass_height=512, **kwargs): super().__init__(**kwargs) self.enable_hr = enable_hr self.scale_latent = scale_latent self.denoising_strength = denoising_strength + self.first_pass_width = first_pass_width + self.first_pass_height = first_pass_height def init(self, all_prompts, all_seeds, all_subseeds): if self.enable_hr: @@ -519,7 +521,7 @@ class StableDiffusionProcessingTxt2Img(StableDiffusionProcessing): else: state.job_count = state.job_count * 2 - desired_pixel_count = 512 * 512 + desired_pixel_count = self.first_pass_width * self.first_pass_height actual_pixel_count = self.width * self.height scale = math.sqrt(desired_pixel_count / actual_pixel_count) diff --git a/modules/txt2img.py b/modules/txt2img.py index e985242b..85cbece4 100644 --- a/modules/txt2img.py +++ b/modules/txt2img.py @@ -6,7 +6,7 @@ import modules.processing as processing from modules.ui import plaintext_to_html -def txt2img(prompt: str, negative_prompt: str, prompt_style: str, prompt_style2: str, steps: int, sampler_index: int, restore_faces: bool, tiling: bool, n_iter: int, batch_size: int, cfg_scale: float, seed: int, subseed: int, subseed_strength: float, seed_resize_from_h: int, seed_resize_from_w: int, seed_enable_extras: bool, height: int, width: int, enable_hr: bool, scale_latent: bool, denoising_strength: float, *args): +def txt2img(prompt: str, negative_prompt: str, prompt_style: str, prompt_style2: str, steps: int, sampler_index: int, restore_faces: bool, tiling: bool, n_iter: int, batch_size: int, cfg_scale: float, seed: int, subseed: int, subseed_strength: float, seed_resize_from_h: int, seed_resize_from_w: int, seed_enable_extras: bool, height: int, width: int, enable_hr: bool, scale_latent: bool, denoising_strength: float, first_pass_width: int, first_pass_height: int, *args): p = StableDiffusionProcessingTxt2Img( sd_model=shared.sd_model, outpath_samples=opts.outdir_samples or opts.outdir_txt2img_samples, @@ -32,6 +32,9 @@ def txt2img(prompt: str, negative_prompt: str, prompt_style: str, prompt_style2: enable_hr=enable_hr, scale_latent=scale_latent if enable_hr else None, denoising_strength=denoising_strength if enable_hr else None, + first_pass_width=first_pass_width if enable_hr else None, + first_pass_height=first_pass_height if enable_hr else None, + ) if cmd_opts.enable_console_prompts: diff --git a/modules/ui.py b/modules/ui.py index 220fb80b..544419b2 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -540,6 +540,8 @@ def create_ui(wrap_gradio_gpu_call): enable_hr = gr.Checkbox(label='Highres. fix', value=False) with gr.Row(visible=False) as hr_options: + first_pass_width = gr.Slider(minimum=64, maximum=1024, step=64, label="First pass width", value=512) + first_pass_height = gr.Slider(minimum=64, maximum=1024, step=64, label="First pass height", value=512) scale_latent = gr.Checkbox(label='Scale latent', value=False) denoising_strength = gr.Slider(minimum=0.0, maximum=1.0, step=0.01, label='Denoising strength', value=0.7) @@ -604,6 +606,8 @@ def create_ui(wrap_gradio_gpu_call): enable_hr, scale_latent, denoising_strength, + first_pass_width, + first_pass_height, ] + custom_inputs, outputs=[ txt2img_gallery, @@ -668,6 +672,8 @@ def create_ui(wrap_gradio_gpu_call): (denoising_strength, "Denoising strength"), (enable_hr, lambda d: "Denoising strength" in d), (hr_options, lambda d: gr.Row.update(visible="Denoising strength" in d)), + (first_pass_width, "First pass width"), + (first_pass_height, "First pass height"), ] modules.generation_parameters_copypaste.connect_paste(paste, txt2img_paste_fields, txt2img_prompt) token_button.click(fn=update_token_counter, inputs=[txt2img_prompt, steps], outputs=[token_counter]) -- cgit v1.2.3 From 40d1c6e423b4dc52b3bdae43d9e2442960760ced Mon Sep 17 00:00:00 2001 From: Buckzor Date: Thu, 13 Oct 2022 20:04:22 +0100 Subject: Option between stretch and crop for Highres. fix --- modules/processing.py | 34 ++++++++++++++++++++++------------ modules/txt2img.py | 7 ++++--- modules/ui.py | 25 ++++++++++++++++--------- 3 files changed, 42 insertions(+), 24 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/processing.py b/modules/processing.py index abbfdf98..0246f5dd 100644 --- a/modules/processing.py +++ b/modules/processing.py @@ -506,13 +506,14 @@ class StableDiffusionProcessingTxt2Img(StableDiffusionProcessing): firstphase_width_truncated = 0 firstphase_height_truncated = 0 - def __init__(self, enable_hr=False, scale_latent=True, denoising_strength=0.75, first_pass_width=512, first_pass_height=512, **kwargs): + def __init__(self, enable_hr=False, scale_latent=True, denoising_strength=0.75, firstphase_width=512, firstphase_height=512, crop_scale=False, **kwargs): super().__init__(**kwargs) self.enable_hr = enable_hr self.scale_latent = scale_latent self.denoising_strength = denoising_strength - self.first_pass_width = first_pass_width - self.first_pass_height = first_pass_height + self.firstphase_width = firstphase_width + self.firstphase_height = firstphase_height + self.crop_scale = crop_scale def init(self, all_prompts, all_seeds, all_subseeds): if self.enable_hr: @@ -521,14 +522,14 @@ class StableDiffusionProcessingTxt2Img(StableDiffusionProcessing): else: state.job_count = state.job_count * 2 - desired_pixel_count = self.first_pass_width * self.first_pass_height - actual_pixel_count = self.width * self.height - scale = math.sqrt(desired_pixel_count / actual_pixel_count) + #desired_pixel_count = self.firstphase_width * self.firstphase_height + #actual_pixel_count = self.width * self.height + #scale = math.sqrt(desired_pixel_count / actual_pixel_count) - self.firstphase_width = math.ceil(scale * self.width / 64) * 64 - self.firstphase_height = math.ceil(scale * self.height / 64) * 64 - self.firstphase_width_truncated = int(scale * self.width) - self.firstphase_height_truncated = int(scale * self.height) + #self.firstphase_width = math.ceil(scale * self.width / 64) * 64 + #self.firstphase_height = math.ceil(scale * self.height / 64) * 64 + #self.firstphase_width_truncated = int(scale * self.width) + #self.firstphase_height_truncated = int(scale * self.height) def sample(self, conditioning, unconditional_conditioning, seeds, subseeds, subseed_strength): self.sampler = sd_samplers.create_sampler_with_index(sd_samplers.samplers, self.sampler_index, self.sd_model) @@ -541,8 +542,17 @@ class StableDiffusionProcessingTxt2Img(StableDiffusionProcessing): x = create_random_tensors([opt_C, self.firstphase_height // opt_f, self.firstphase_width // opt_f], seeds=seeds, subseeds=subseeds, subseed_strength=self.subseed_strength, seed_resize_from_h=self.seed_resize_from_h, seed_resize_from_w=self.seed_resize_from_w, p=self) samples = self.sampler.sample(self, x, conditioning, unconditional_conditioning) - truncate_x = (self.firstphase_width - self.firstphase_width_truncated) // opt_f - truncate_y = (self.firstphase_height - self.firstphase_height_truncated) // opt_f + truncate_x = 0 + truncate_y = 0 + + if self.crop_scale: + if self.width/self.firstphase_width > self.height/self.firstphase_height: + #Crop to landscape + truncate_y = (self.width - self.firstphase_width)//2 // opt_f + + elif self.width/self.firstphase_width < self.height/self.firstphase_height: + #Crop to portrait + truncate_x = (self.height - self.firstphase_height)//2 // opt_f samples = samples[:, :, truncate_y//2:samples.shape[2]-truncate_y//2, truncate_x//2:samples.shape[3]-truncate_x//2] diff --git a/modules/txt2img.py b/modules/txt2img.py index 85cbece4..447ec3d3 100644 --- a/modules/txt2img.py +++ b/modules/txt2img.py @@ -6,7 +6,7 @@ import modules.processing as processing from modules.ui import plaintext_to_html -def txt2img(prompt: str, negative_prompt: str, prompt_style: str, prompt_style2: str, steps: int, sampler_index: int, restore_faces: bool, tiling: bool, n_iter: int, batch_size: int, cfg_scale: float, seed: int, subseed: int, subseed_strength: float, seed_resize_from_h: int, seed_resize_from_w: int, seed_enable_extras: bool, height: int, width: int, enable_hr: bool, scale_latent: bool, denoising_strength: float, first_pass_width: int, first_pass_height: int, *args): +def txt2img(prompt: str, negative_prompt: str, prompt_style: str, prompt_style2: str, steps: int, sampler_index: int, restore_faces: bool, tiling: bool, n_iter: int, batch_size: int, cfg_scale: float, seed: int, subseed: int, subseed_strength: float, seed_resize_from_h: int, seed_resize_from_w: int, seed_enable_extras: bool, height: int, width: int, enable_hr: bool, scale_latent: bool, denoising_strength: float, firstphase_width: int, firstphase_height: int, crop_scale: bool, *args): p = StableDiffusionProcessingTxt2Img( sd_model=shared.sd_model, outpath_samples=opts.outdir_samples or opts.outdir_txt2img_samples, @@ -32,8 +32,9 @@ def txt2img(prompt: str, negative_prompt: str, prompt_style: str, prompt_style2: enable_hr=enable_hr, scale_latent=scale_latent if enable_hr else None, denoising_strength=denoising_strength if enable_hr else None, - first_pass_width=first_pass_width if enable_hr else None, - first_pass_height=first_pass_height if enable_hr else None, + firstphase_width=firstphase_width if enable_hr else None, + firstphase_height=firstphase_height if enable_hr else None, + crop_scale=crop_scale if enable_hr else None, ) diff --git a/modules/ui.py b/modules/ui.py index 544419b2..f2d81f68 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -540,12 +540,18 @@ def create_ui(wrap_gradio_gpu_call): enable_hr = gr.Checkbox(label='Highres. fix', value=False) with gr.Row(visible=False) as hr_options: - first_pass_width = gr.Slider(minimum=64, maximum=1024, step=64, label="First pass width", value=512) - first_pass_height = gr.Slider(minimum=64, maximum=1024, step=64, label="First pass height", value=512) - scale_latent = gr.Checkbox(label='Scale latent', value=False) - denoising_strength = gr.Slider(minimum=0.0, maximum=1.0, step=0.01, label='Denoising strength', value=0.7) + with gr.Column(scale=1.0): + firstphase_width = gr.Slider(minimum=64, maximum=1024, step=64, label="First pass width", value=512) + firstphase_height = gr.Slider(minimum=64, maximum=1024, step=64, label="First pass height", value=512) + + with gr.Column(scale=1.0): + with gr.Row(): + crop_scale = gr.Checkbox(label='Crop when scaling', value=False) + scale_latent = gr.Checkbox(label='Scale latent', value=False) + with gr.Row(): + denoising_strength = gr.Slider(minimum=0.0, maximum=1.0, step=0.01, label='Denoising strength', value=0.7) - with gr.Row(): + with gr.Row(equal_height=True): batch_count = gr.Slider(minimum=1, step=1, label='Batch count', value=1) batch_size = gr.Slider(minimum=1, maximum=8, step=1, label='Batch size', value=1) @@ -606,8 +612,9 @@ def create_ui(wrap_gradio_gpu_call): enable_hr, scale_latent, denoising_strength, - first_pass_width, - first_pass_height, + firstphase_width, + firstphase_height, + crop_scale, ] + custom_inputs, outputs=[ txt2img_gallery, @@ -672,8 +679,8 @@ def create_ui(wrap_gradio_gpu_call): (denoising_strength, "Denoising strength"), (enable_hr, lambda d: "Denoising strength" in d), (hr_options, lambda d: gr.Row.update(visible="Denoising strength" in d)), - (first_pass_width, "First pass width"), - (first_pass_height, "First pass height"), + (firstphase_width, "First pass width"), + (firstphase_height, "First pass height"), ] modules.generation_parameters_copypaste.connect_paste(paste, txt2img_paste_fields, txt2img_prompt) token_button.click(fn=update_token_counter, inputs=[txt2img_prompt, steps], outputs=[token_counter]) -- cgit v1.2.3 From e644b5a80beb54b6df4caa63fb19d889dd4ceff6 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Fri, 14 Oct 2022 17:03:03 +0300 Subject: remove scale latent and no-crop options from hires fix support copy-pasting new parameters for hires fix --- modules/processing.py | 64 ++++++++++++++++++++++----------------------------- modules/txt2img.py | 9 +++----- modules/ui.py | 19 ++++----------- 3 files changed, 35 insertions(+), 57 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/processing.py b/modules/processing.py index d9b0e0e7..100a259f 100644 --- a/modules/processing.py +++ b/modules/processing.py @@ -506,14 +506,12 @@ class StableDiffusionProcessingTxt2Img(StableDiffusionProcessing): firstphase_width_truncated = 0 firstphase_height_truncated = 0 - def __init__(self, enable_hr=False, scale_latent=True, denoising_strength=0.75, firstphase_width=512, firstphase_height=512, crop_scale=False, **kwargs): + def __init__(self, enable_hr=False, denoising_strength=0.75, firstphase_width=512, firstphase_height=512, **kwargs): super().__init__(**kwargs) self.enable_hr = enable_hr - self.scale_latent = scale_latent self.denoising_strength = denoising_strength self.firstphase_width = firstphase_width self.firstphase_height = firstphase_height - self.crop_scale = crop_scale def init(self, all_prompts, all_seeds, all_subseeds): if self.enable_hr: @@ -530,6 +528,8 @@ class StableDiffusionProcessingTxt2Img(StableDiffusionProcessing): samples = self.sampler.sample(self, x, conditioning, unconditional_conditioning) return samples + self.extra_generation_params["First pass size"] = f"{self.firstphase_width}x{self.firstphase_height}" + x = create_random_tensors([opt_C, self.firstphase_height // opt_f, self.firstphase_width // opt_f], seeds=seeds, subseeds=subseeds, subseed_strength=self.subseed_strength, seed_resize_from_h=self.seed_resize_from_h, seed_resize_from_w=self.seed_resize_from_w, p=self) samples = self.sampler.sample(self, x, conditioning, unconditional_conditioning) @@ -538,46 +538,36 @@ class StableDiffusionProcessingTxt2Img(StableDiffusionProcessing): width_ratio = self.width/self.firstphase_width height_ratio = self.height/self.firstphase_height - if self.crop_scale: - if width_ratio > height_ratio: - #Crop to landscape - truncate_y = int((self.width - self.firstphase_width) / width_ratio / height_ratio / opt_f) + if width_ratio > height_ratio: + truncate_y = int((self.width - self.firstphase_width) / width_ratio / height_ratio / opt_f) - elif width_ratio < height_ratio: - #Crop to portrait - truncate_x = int((self.height - self.firstphase_height) / width_ratio / height_ratio / opt_f) + elif width_ratio < height_ratio: + truncate_x = int((self.height - self.firstphase_height) / width_ratio / height_ratio / opt_f) - samples = samples[:, :, truncate_y//2:samples.shape[2]-truncate_y//2, truncate_x//2:samples.shape[3]-truncate_x//2] - - + samples = samples[:, :, truncate_y//2:samples.shape[2]-truncate_y//2, truncate_x//2:samples.shape[3]-truncate_x//2] - + decoded_samples = decode_first_stage(self.sd_model, samples) - if self.scale_latent: - samples = torch.nn.functional.interpolate(samples, size=(self.height // opt_f, self.width // opt_f), mode="bilinear") + if opts.upscaler_for_img2img is None or opts.upscaler_for_img2img == "None": + decoded_samples = torch.nn.functional.interpolate(decoded_samples, size=(self.height, self.width), mode="bilinear") else: - decoded_samples = decode_first_stage(self.sd_model, samples) + lowres_samples = torch.clamp((decoded_samples + 1.0) / 2.0, min=0.0, max=1.0) - if opts.upscaler_for_img2img is None or opts.upscaler_for_img2img == "None": - decoded_samples = torch.nn.functional.interpolate(decoded_samples, size=(self.height, self.width), mode="bilinear") - else: - lowres_samples = torch.clamp((decoded_samples + 1.0) / 2.0, min=0.0, max=1.0) - - batch_images = [] - for i, x_sample in enumerate(lowres_samples): - x_sample = 255. * np.moveaxis(x_sample.cpu().numpy(), 0, 2) - x_sample = x_sample.astype(np.uint8) - image = Image.fromarray(x_sample) - image = images.resize_image(0, image, self.width, self.height) - image = np.array(image).astype(np.float32) / 255.0 - image = np.moveaxis(image, 2, 0) - batch_images.append(image) - - decoded_samples = torch.from_numpy(np.array(batch_images)) - decoded_samples = decoded_samples.to(shared.device) - decoded_samples = 2. * decoded_samples - 1. - - samples = self.sd_model.get_first_stage_encoding(self.sd_model.encode_first_stage(decoded_samples)) + batch_images = [] + for i, x_sample in enumerate(lowres_samples): + x_sample = 255. * np.moveaxis(x_sample.cpu().numpy(), 0, 2) + x_sample = x_sample.astype(np.uint8) + image = Image.fromarray(x_sample) + image = images.resize_image(0, image, self.width, self.height) + image = np.array(image).astype(np.float32) / 255.0 + image = np.moveaxis(image, 2, 0) + batch_images.append(image) + + decoded_samples = torch.from_numpy(np.array(batch_images)) + decoded_samples = decoded_samples.to(shared.device) + decoded_samples = 2. * decoded_samples - 1. + + samples = self.sd_model.get_first_stage_encoding(self.sd_model.encode_first_stage(decoded_samples)) shared.state.nextjob() diff --git a/modules/txt2img.py b/modules/txt2img.py index 447ec3d3..2381347f 100644 --- a/modules/txt2img.py +++ b/modules/txt2img.py @@ -6,7 +6,7 @@ import modules.processing as processing from modules.ui import plaintext_to_html -def txt2img(prompt: str, negative_prompt: str, prompt_style: str, prompt_style2: str, steps: int, sampler_index: int, restore_faces: bool, tiling: bool, n_iter: int, batch_size: int, cfg_scale: float, seed: int, subseed: int, subseed_strength: float, seed_resize_from_h: int, seed_resize_from_w: int, seed_enable_extras: bool, height: int, width: int, enable_hr: bool, scale_latent: bool, denoising_strength: float, firstphase_width: int, firstphase_height: int, crop_scale: bool, *args): +def txt2img(prompt: str, negative_prompt: str, prompt_style: str, prompt_style2: str, steps: int, sampler_index: int, restore_faces: bool, tiling: bool, n_iter: int, batch_size: int, cfg_scale: float, seed: int, subseed: int, subseed_strength: float, seed_resize_from_h: int, seed_resize_from_w: int, seed_enable_extras: bool, height: int, width: int, enable_hr: bool, denoising_strength: float, firstphase_width: int, firstphase_height: int, *args): p = StableDiffusionProcessingTxt2Img( sd_model=shared.sd_model, outpath_samples=opts.outdir_samples or opts.outdir_txt2img_samples, @@ -30,12 +30,9 @@ def txt2img(prompt: str, negative_prompt: str, prompt_style: str, prompt_style2: restore_faces=restore_faces, tiling=tiling, enable_hr=enable_hr, - scale_latent=scale_latent if enable_hr else None, denoising_strength=denoising_strength if enable_hr else None, - firstphase_width=firstphase_width if enable_hr else None, - firstphase_height=firstphase_height if enable_hr else None, - crop_scale=crop_scale if enable_hr else None, - + firstphase_width=firstphase_width if enable_hr else None, + firstphase_height=firstphase_height if enable_hr else None, ) if cmd_opts.enable_console_prompts: diff --git a/modules/ui.py b/modules/ui.py index f2d81f68..d66ddc14 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -540,16 +540,9 @@ def create_ui(wrap_gradio_gpu_call): enable_hr = gr.Checkbox(label='Highres. fix', value=False) with gr.Row(visible=False) as hr_options: - with gr.Column(scale=1.0): - firstphase_width = gr.Slider(minimum=64, maximum=1024, step=64, label="First pass width", value=512) - firstphase_height = gr.Slider(minimum=64, maximum=1024, step=64, label="First pass height", value=512) - - with gr.Column(scale=1.0): - with gr.Row(): - crop_scale = gr.Checkbox(label='Crop when scaling', value=False) - scale_latent = gr.Checkbox(label='Scale latent', value=False) - with gr.Row(): - denoising_strength = gr.Slider(minimum=0.0, maximum=1.0, step=0.01, label='Denoising strength', value=0.7) + firstphase_width = gr.Slider(minimum=64, maximum=1024, step=64, label="First pass width", value=512) + firstphase_height = gr.Slider(minimum=64, maximum=1024, step=64, label="First pass height", value=512) + denoising_strength = gr.Slider(minimum=0.0, maximum=1.0, step=0.01, label='Denoising strength', value=0.7) with gr.Row(equal_height=True): batch_count = gr.Slider(minimum=1, step=1, label='Batch count', value=1) @@ -610,11 +603,9 @@ def create_ui(wrap_gradio_gpu_call): height, width, enable_hr, - scale_latent, denoising_strength, firstphase_width, firstphase_height, - crop_scale, ] + custom_inputs, outputs=[ txt2img_gallery, @@ -679,8 +670,8 @@ def create_ui(wrap_gradio_gpu_call): (denoising_strength, "Denoising strength"), (enable_hr, lambda d: "Denoising strength" in d), (hr_options, lambda d: gr.Row.update(visible="Denoising strength" in d)), - (firstphase_width, "First pass width"), - (firstphase_height, "First pass height"), + (firstphase_width, "First pass size-1"), + (firstphase_height, "First pass size-2"), ] modules.generation_parameters_copypaste.connect_paste(paste, txt2img_paste_fields, txt2img_prompt) token_button.click(fn=update_token_counter, inputs=[txt2img_prompt, steps], outputs=[token_counter]) -- cgit v1.2.3 From 0aec19d7837d8564355fdb286541db7165852e41 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Fri, 14 Oct 2022 18:15:03 +0300 Subject: make pasting into img2img prompt work make image params request not use temp files --- modules/images.py | 36 ++++++++++++++++++------------------ modules/ui.py | 9 +++------ 2 files changed, 21 insertions(+), 24 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/images.py b/modules/images.py index f1155b7f..68cdbc93 100644 --- a/modules/images.py +++ b/modules/images.py @@ -1,4 +1,5 @@ import datetime +import io import math import os from collections import namedtuple @@ -465,21 +466,20 @@ def save_image(image, path, basename, seed=None, prompt=None, extension='png', i return fullfn, txt_fullfn -def image_data(image_path): - file, ext = os.path.splitext(image_path.name) - data = {} - if "png" in ext: - image = Image.open(image_path.name, "r") - print(f"Image data requested for {image_path.name} {image.format} of {type(image)}") - try: - data = image.text["parameters"] - except Exception as e: - print(f"Exception: {e}") - pass - print(f"Image data: {data}") - if "txt" in ext: - myfile = open(image_path.name, 'r') - data = myfile.read() - myfile.close() - - return data, None +def image_data(data): + try: + image = Image.open(io.BytesIO(data)) + textinfo = image.text["parameters"] + return textinfo, None + except Exception: + pass + + try: + text = data.decode('utf8') + assert len(text) < 10000 + return text, None + + except Exception: + pass + + return '', None diff --git a/modules/ui.py b/modules/ui.py index 0a3ee887..6266db49 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -514,7 +514,7 @@ def create_ui(wrap_gradio_gpu_call): with gr.Blocks(analytics_enabled=False) as txt2img_interface: txt2img_prompt, roll, txt2img_prompt_style, txt2img_negative_prompt, txt2img_prompt_style2, submit, _, _, txt2img_prompt_style_apply, txt2img_save_style, paste, token_counter, token_button = create_toprow(is_img2img=False) dummy_component = gr.Label(visible=False) - txt_prompt_img = gr.File(label="", elem_id="txt2img_prompt_image", file_count="single", type="file", visible=False) + txt_prompt_img = gr.File(label="", elem_id="txt2img_prompt_image", file_count="single", type="bytes", visible=False) with gr.Row(elem_id='txt2img_progress_row'): with gr.Column(scale=1): @@ -620,7 +620,6 @@ def create_ui(wrap_gradio_gpu_call): txt_prompt_img.change( fn=modules.images.image_data, - # _js = "get_extras_tab_index", inputs=[ txt_prompt_img ], @@ -692,8 +691,7 @@ def create_ui(wrap_gradio_gpu_call): img2img_prompt, roll, img2img_prompt_style, img2img_negative_prompt, img2img_prompt_style2, submit, img2img_interrogate, img2img_deepbooru, img2img_prompt_style_apply, img2img_save_style, paste, token_counter, token_button = create_toprow(is_img2img=True) with gr.Row(elem_id='img2img_progress_row'): - img2img_prompt_img = gr.File(label="", elem_id="txt_prompt_image", file_count="single", type="file", - visible=False) + img2img_prompt_img = gr.File(label="", elem_id="img2img_prompt_image", file_count="single", type="bytes", visible=False) with gr.Column(scale=1): pass @@ -791,9 +789,8 @@ def create_ui(wrap_gradio_gpu_call): img2img_prompt_img.change( fn=modules.images.image_data, - # _js = "get_extras_tab_index", inputs=[ - txt_prompt_img + img2img_prompt_img ], outputs=[ img2img_prompt, -- cgit v1.2.3 From 67f447ddcc8a17d11939c3801dca635dc22944c7 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Fri, 14 Oct 2022 19:30:28 +0300 Subject: possibility to load checkpoint, clip skip, and hypernet from infotext --- modules/ui.py | 52 +++++++++++++++++++++++++++++++++++++++++++++------- 1 file changed, 45 insertions(+), 7 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/ui.py b/modules/ui.py index 6266db49..a37a4e17 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -22,7 +22,7 @@ import gradio as gr import gradio.utils import gradio.routes -from modules import sd_hijack +from modules import sd_hijack, sd_models from modules.paths import script_path from modules.shared import opts, cmd_opts if cmd_opts.deepdanbooru: @@ -507,12 +507,38 @@ def setup_progressbar(progressbar, preview, id_part, textinfo=None): ) +def apply_setting(key, value): + if value is None: + return gr.update() + + if key == "sd_model_checkpoint": + ckpt_info = sd_models.get_closet_checkpoint_match(value) + + if ckpt_info is not None: + value = ckpt_info.title + else: + return gr.update() + + comp_args = opts.data_labels[key].component_args + if comp_args and isinstance(comp_args, dict) and comp_args.get('visible') is False: + return + + valtype = type(opts.data_labels[key].default) + oldval = opts.data[key] + opts.data[key] = valtype(value) if valtype != type(None) else value + if oldval != value and opts.data_labels[key].onchange is not None: + opts.data_labels[key].onchange() + + opts.save(shared.config_filename) + return value + + def create_ui(wrap_gradio_gpu_call): import modules.img2img import modules.txt2img with gr.Blocks(analytics_enabled=False) as txt2img_interface: - txt2img_prompt, roll, txt2img_prompt_style, txt2img_negative_prompt, txt2img_prompt_style2, submit, _, _, txt2img_prompt_style_apply, txt2img_save_style, paste, token_counter, token_button = create_toprow(is_img2img=False) + txt2img_prompt, roll, txt2img_prompt_style, txt2img_negative_prompt, txt2img_prompt_style2, submit, _, _, txt2img_prompt_style_apply, txt2img_save_style, txt2img_paste, token_counter, token_button = create_toprow(is_img2img=False) dummy_component = gr.Label(visible=False) txt_prompt_img = gr.File(label="", elem_id="txt2img_prompt_image", file_count="single", type="bytes", visible=False) @@ -684,11 +710,10 @@ def create_ui(wrap_gradio_gpu_call): (firstphase_width, "First pass size-1"), (firstphase_height, "First pass size-2"), ] - modules.generation_parameters_copypaste.connect_paste(paste, txt2img_paste_fields, txt2img_prompt) token_button.click(fn=update_token_counter, inputs=[txt2img_prompt, steps], outputs=[token_counter]) with gr.Blocks(analytics_enabled=False) as img2img_interface: - img2img_prompt, roll, img2img_prompt_style, img2img_negative_prompt, img2img_prompt_style2, submit, img2img_interrogate, img2img_deepbooru, img2img_prompt_style_apply, img2img_save_style, paste, token_counter, token_button = create_toprow(is_img2img=True) + img2img_prompt, roll, img2img_prompt_style, img2img_negative_prompt, img2img_prompt_style2, submit, img2img_interrogate, img2img_deepbooru, img2img_prompt_style_apply, img2img_save_style, img2img_paste, token_counter, token_button = create_toprow(is_img2img=True) with gr.Row(elem_id='img2img_progress_row'): img2img_prompt_img = gr.File(label="", elem_id="img2img_prompt_image", file_count="single", type="bytes", visible=False) @@ -938,7 +963,6 @@ def create_ui(wrap_gradio_gpu_call): (seed_resize_from_h, "Seed resize from-2"), (denoising_strength, "Denoising strength"), ] - modules.generation_parameters_copypaste.connect_paste(paste, img2img_paste_fields, img2img_prompt) token_button.click(fn=update_token_counter, inputs=[img2img_prompt, steps], outputs=[token_counter]) with gr.Blocks(analytics_enabled=False) as extras_interface: @@ -1580,8 +1604,22 @@ Requested path was: {f} outputs=[extras_image], ) - modules.generation_parameters_copypaste.connect_paste(pnginfo_send_to_txt2img, txt2img_paste_fields, generation_info, 'switch_to_txt2img') - modules.generation_parameters_copypaste.connect_paste(pnginfo_send_to_img2img, img2img_paste_fields, generation_info, 'switch_to_img2img_img2img') + settings_map = { + 'sd_hypernetwork': 'Hypernet', + 'CLIP_stop_at_last_layers': 'Clip skip', + 'sd_model_checkpoint': 'Model hash', + } + + settings_paste_fields = [ + (component_dict[k], lambda d, k=k, v=v: apply_setting(k, d.get(v, None))) + for k, v in settings_map.items() + ] + + modules.generation_parameters_copypaste.connect_paste(txt2img_paste, txt2img_paste_fields + settings_paste_fields, txt2img_prompt) + modules.generation_parameters_copypaste.connect_paste(img2img_paste, img2img_paste_fields + settings_paste_fields, img2img_prompt) + + modules.generation_parameters_copypaste.connect_paste(pnginfo_send_to_txt2img, txt2img_paste_fields + settings_paste_fields, generation_info, 'switch_to_txt2img') + modules.generation_parameters_copypaste.connect_paste(pnginfo_send_to_img2img, img2img_paste_fields + settings_paste_fields, generation_info, 'switch_to_img2img_img2img') ui_config_file = cmd_opts.ui_config_file ui_settings = {} -- cgit v1.2.3 From c344ba3b325459abbf9b0df2c1b18f7bf99805b2 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Fri, 14 Oct 2022 20:31:49 +0300 Subject: add option to read generation params for learning previews from txt2img --- modules/hypernetworks/hypernetwork.py | 21 ++++++++++++++++----- modules/textual_inversion/textual_inversion.py | 25 ++++++++++++++++++------- modules/ui.py | 20 +++++++++++++++++--- 3 files changed, 51 insertions(+), 15 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/hypernetworks/hypernetwork.py b/modules/hypernetworks/hypernetwork.py index f1248bb7..e5cb1817 100644 --- a/modules/hypernetworks/hypernetwork.py +++ b/modules/hypernetworks/hypernetwork.py @@ -180,7 +180,7 @@ def attention_CrossAttention_forward(self, x, context=None, mask=None): return self.to_out(out) -def train_hypernetwork(hypernetwork_name, learn_rate, data_root, log_directory, steps, create_image_every, save_hypernetwork_every, template_file, preview_image_prompt): +def train_hypernetwork(hypernetwork_name, learn_rate, data_root, log_directory, steps, create_image_every, save_hypernetwork_every, template_file, preview_from_txt2img, preview_prompt, preview_negative_prompt, preview_steps, preview_sampler_index, preview_cfg_scale, preview_seed, preview_width, preview_height): assert hypernetwork_name, 'hypernetwork not selected' path = shared.hypernetworks.get(hypernetwork_name, None) @@ -265,20 +265,31 @@ def train_hypernetwork(hypernetwork_name, learn_rate, data_root, log_directory, if hypernetwork.step > 0 and images_dir is not None and hypernetwork.step % create_image_every == 0: last_saved_image = os.path.join(images_dir, f'{hypernetwork_name}-{hypernetwork.step}.png') - preview_text = entry.cond_text if preview_image_prompt == "" else preview_image_prompt - optimizer.zero_grad() shared.sd_model.cond_stage_model.to(devices.device) shared.sd_model.first_stage_model.to(devices.device) p = processing.StableDiffusionProcessingTxt2Img( sd_model=shared.sd_model, - prompt=preview_text, - steps=20, do_not_save_grid=True, do_not_save_samples=True, ) + if preview_from_txt2img: + p.prompt = preview_prompt + p.negative_prompt = preview_negative_prompt + p.steps = preview_steps + p.sampler_index = preview_sampler_index + p.cfg_scale = preview_cfg_scale + p.seed = preview_seed + p.width = preview_width + p.height = preview_height + else: + p.prompt = entry.cond_text + p.steps = 20 + + preview_text = p.prompt + processed = processing.process_images(p) image = processed.images[0] if len(processed.images)>0 else None diff --git a/modules/textual_inversion/textual_inversion.py b/modules/textual_inversion/textual_inversion.py index fa0e33a2..3d835358 100644 --- a/modules/textual_inversion/textual_inversion.py +++ b/modules/textual_inversion/textual_inversion.py @@ -172,7 +172,7 @@ def create_embedding(name, num_vectors_per_token, init_text='*'): return fn -def train_embedding(embedding_name, learn_rate, data_root, log_directory, training_width, training_height, steps, create_image_every, save_embedding_every, template_file, save_image_with_stored_embedding, preview_image_prompt): +def train_embedding(embedding_name, learn_rate, data_root, log_directory, training_width, training_height, steps, create_image_every, save_embedding_every, template_file, save_image_with_stored_embedding, preview_from_txt2img, preview_prompt, preview_negative_prompt, preview_steps, preview_sampler_index, preview_cfg_scale, preview_seed, preview_width, preview_height): assert embedding_name, 'embedding not selected' shared.state.textinfo = "Initializing textual inversion training..." @@ -259,18 +259,29 @@ def train_embedding(embedding_name, learn_rate, data_root, log_directory, traini if embedding.step > 0 and images_dir is not None and embedding.step % create_image_every == 0: last_saved_image = os.path.join(images_dir, f'{embedding_name}-{embedding.step}.png') - preview_text = entry.cond_text if preview_image_prompt == "" else preview_image_prompt - p = processing.StableDiffusionProcessingTxt2Img( sd_model=shared.sd_model, - prompt=preview_text, - steps=20, - height=training_height, - width=training_width, do_not_save_grid=True, do_not_save_samples=True, ) + if preview_from_txt2img: + p.prompt = preview_prompt + p.negative_prompt = preview_negative_prompt + p.steps = preview_steps + p.sampler_index = preview_sampler_index + p.cfg_scale = preview_cfg_scale + p.seed = preview_seed + p.width = preview_width + p.height = preview_height + else: + p.prompt = entry.cond_text + p.steps = 20 + p.width = training_width + p.height = training_height + + preview_text = p.prompt + processed = processing.process_images(p) image = processed.images[0] diff --git a/modules/ui.py b/modules/ui.py index 828bfeea..4a04c2cc 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -711,6 +711,18 @@ def create_ui(wrap_gradio_gpu_call): (firstphase_width, "First pass size-1"), (firstphase_height, "First pass size-2"), ] + + txt2img_preview_params = [ + txt2img_prompt, + txt2img_negative_prompt, + steps, + sampler_index, + cfg_scale, + seed, + width, + height, + ] + token_button.click(fn=update_token_counter, inputs=[txt2img_prompt, steps], outputs=[token_counter]) with gr.Blocks(analytics_enabled=False) as img2img_interface: @@ -1162,7 +1174,7 @@ def create_ui(wrap_gradio_gpu_call): create_image_every = gr.Number(label='Save an image to log directory every N steps, 0 to disable', value=500, precision=0) save_embedding_every = gr.Number(label='Save a copy of embedding to log directory every N steps, 0 to disable', value=500, precision=0) save_image_with_stored_embedding = gr.Checkbox(label='Save images with embedding in PNG chunks', value=True) - preview_image_prompt = gr.Textbox(label='Preview prompt', value="") + preview_from_txt2img = gr.Checkbox(label='Read parameters (prompt, etc...) from txt2img tab when making previews', value=False) with gr.Row(): interrupt_training = gr.Button(value="Interrupt") @@ -1240,7 +1252,8 @@ def create_ui(wrap_gradio_gpu_call): save_embedding_every, template_file, save_image_with_stored_embedding, - preview_image_prompt, + preview_from_txt2img, + *txt2img_preview_params, ], outputs=[ ti_output, @@ -1260,7 +1273,8 @@ def create_ui(wrap_gradio_gpu_call): create_image_every, save_embedding_every, template_file, - preview_image_prompt, + preview_from_txt2img, + *txt2img_preview_params, ], outputs=[ ti_output, -- cgit v1.2.3 From c250cb289c97fe303cef69064bf45899406f6a40 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Fri, 14 Oct 2022 22:01:49 +0300 Subject: change checkpoint merger to work in a more obvious way remove sigmoid and inverse sigmoid because they just did the same thing as weighed sum only with changed multiplier --- javascript/hints.js | 4 ++-- modules/extras.py | 24 +++++------------------- modules/ui.py | 4 ++-- 3 files changed, 9 insertions(+), 23 deletions(-) (limited to 'modules/ui.py') diff --git a/javascript/hints.js b/javascript/hints.js index af010a59..8fec907d 100644 --- a/javascript/hints.js +++ b/javascript/hints.js @@ -87,8 +87,8 @@ titles = { "Quicksettings list": "List of setting names, separated by commas, for settings that should go to the quick access bar at the top, rather than the usual setting tab. See modules/shared.py for setting names. Requires restarting to apply.", - "Weighted Sum": "Result = A * (1 - M) + B * M", - "Add difference": "Result = A + (B - C) * (1 - M)", + "Weighted sum": "Result = A * (1 - M) + B * M", + "Add difference": "Result = A + (B - C) * M", } diff --git a/modules/extras.py b/modules/extras.py index 2e7b3751..f2f5a7b0 100644 --- a/modules/extras.py +++ b/modules/extras.py @@ -159,24 +159,12 @@ def run_pnginfo(image): return '', geninfo, info -def run_modelmerger(primary_model_name, secondary_model_name, teritary_model_name, interp_method, interp_amount, save_as_half, custom_name): - # Linear interpolation (https://en.wikipedia.org/wiki/Linear_interpolation) +def run_modelmerger(primary_model_name, secondary_model_name, teritary_model_name, interp_method, multiplier, save_as_half, custom_name): def weighted_sum(theta0, theta1, theta2, alpha): return ((1 - alpha) * theta0) + (alpha * theta1) - # Smoothstep (https://en.wikipedia.org/wiki/Smoothstep) - def sigmoid(theta0, theta1, theta2, alpha): - alpha = alpha * alpha * (3 - (2 * alpha)) - return theta0 + ((theta1 - theta0) * alpha) - - # Inverse Smoothstep (https://en.wikipedia.org/wiki/Smoothstep) - def inv_sigmoid(theta0, theta1, theta2, alpha): - import math - alpha = 0.5 - math.sin(math.asin(1.0 - 2.0 * alpha) / 3.0) - return theta0 + ((theta1 - theta0) * alpha) - def add_difference(theta0, theta1, theta2, alpha): - return theta0 + (theta1 - theta2) * (1.0 - alpha) + return theta0 + (theta1 - theta2) * alpha primary_model_info = sd_models.checkpoints_list[primary_model_name] secondary_model_info = sd_models.checkpoints_list[secondary_model_name] @@ -198,9 +186,7 @@ def run_modelmerger(primary_model_name, secondary_model_name, teritary_model_nam theta_2 = None theta_funcs = { - "Weighted Sum": weighted_sum, - "Sigmoid": sigmoid, - "Inverse Sigmoid": inv_sigmoid, + "Weighted sum": weighted_sum, "Add difference": add_difference, } theta_func = theta_funcs[interp_method] @@ -213,7 +199,7 @@ def run_modelmerger(primary_model_name, secondary_model_name, teritary_model_nam if t2 is None: t2 = torch.zeros_like(theta_0[key]) - theta_0[key] = theta_func(theta_0[key], theta_1[key], t2, (float(1.0) - interp_amount)) # Need to reverse the interp_amount to match the desired mix ration in the merged checkpoint + theta_0[key] = theta_func(theta_0[key], theta_1[key], t2, multiplier) if save_as_half: theta_0[key] = theta_0[key].half() @@ -227,7 +213,7 @@ def run_modelmerger(primary_model_name, secondary_model_name, teritary_model_nam ckpt_dir = shared.cmd_opts.ckpt_dir or sd_models.model_path - filename = primary_model_info.model_name + '_' + str(round(interp_amount, 2)) + '-' + secondary_model_info.model_name + '_' + str(round((float(1.0) - interp_amount), 2)) + '-' + interp_method.replace(" ", "_") + '-merged.ckpt' + filename = primary_model_info.model_name + '_' + str(round(1-multiplier, 2)) + '-' + secondary_model_info.model_name + '_' + str(round(multiplier, 2)) + '-' + interp_method.replace(" ", "_") + '-merged.ckpt' filename = filename if custom_name == '' else (custom_name + '.ckpt') output_modelname = os.path.join(ckpt_dir, filename) diff --git a/modules/ui.py b/modules/ui.py index 4a04c2cc..a08ffc9b 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1101,8 +1101,8 @@ def create_ui(wrap_gradio_gpu_call): secondary_model_name = gr.Dropdown(modules.sd_models.checkpoint_tiles(), elem_id="modelmerger_secondary_model_name", label="Secondary model (B)") tertiary_model_name = gr.Dropdown(modules.sd_models.checkpoint_tiles(), elem_id="modelmerger_tertiary_model_name", label="Tertiary model (C)") custom_name = gr.Textbox(label="Custom Name (Optional)") - interp_amount = gr.Slider(minimum=0.0, maximum=1.0, step=0.05, label='Interpolation amount (1 - M)', value=0.3) - interp_method = gr.Radio(choices=["Weighted Sum", "Sigmoid", "Inverse Sigmoid", "Add difference"], value="Weighted Sum", label="Interpolation Method") + interp_amount = gr.Slider(minimum=0.0, maximum=1.0, step=0.05, label='Multiplier (M) - set to 0 to get model A', value=0.3) + interp_method = gr.Radio(choices=["Weighted sum", "Add difference"], value="Weighted sum", label="Interpolation Method") save_as_half = gr.Checkbox(value=False, label="Save as float16") modelmerger_merge = gr.Button(elem_id="modelmerger_merge", label="Merge", variant='primary') -- cgit v1.2.3 From 03d62538aebeff51713619fe808c953bdb70193d Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Fri, 14 Oct 2022 22:43:55 +0300 Subject: remove duplicate code for log loss, add step, make it read from options rather than gradio input --- modules/hypernetworks/hypernetwork.py | 20 ++++-------- modules/shared.py | 3 +- modules/textual_inversion/textual_inversion.py | 44 ++++++++++++++++++-------- modules/ui.py | 3 -- 4 files changed, 38 insertions(+), 32 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/hypernetworks/hypernetwork.py b/modules/hypernetworks/hypernetwork.py index edb8cba1..59c7ac6e 100644 --- a/modules/hypernetworks/hypernetwork.py +++ b/modules/hypernetworks/hypernetwork.py @@ -15,6 +15,7 @@ import torch from torch import einsum from einops import rearrange, repeat import modules.textual_inversion.dataset +from modules.textual_inversion import textual_inversion from modules.textual_inversion.learn_schedule import LearnRateScheduler @@ -210,7 +211,7 @@ def train_hypernetwork(hypernetwork_name, learn_rate, data_root, log_directory, shared.state.textinfo = f"Preparing dataset from {html.escape(data_root)}..." with torch.autocast("cuda"): - ds = modules.textual_inversion.dataset.PersonalizedBase(data_root=data_root, width=512, height=512, repeats=1, placeholder_token=hypernetwork_name, model=shared.sd_model, device=devices.device, template_file=template_file, include_cond=True) + ds = modules.textual_inversion.dataset.PersonalizedBase(data_root=data_root, width=512, height=512, repeats=shared.opts.training_image_repeats_per_epoch, placeholder_token=hypernetwork_name, model=shared.sd_model, device=devices.device, template_file=template_file, include_cond=True) if unload: shared.sd_model.cond_stage_model.to(devices.cpu) @@ -263,19 +264,10 @@ def train_hypernetwork(hypernetwork_name, learn_rate, data_root, log_directory, last_saved_file = os.path.join(hypernetwork_dir, f'{hypernetwork_name}-{hypernetwork.step}.pt') hypernetwork.save(last_saved_file) - if write_csv_every > 0 and hypernetwork_dir is not None and hypernetwork.step % write_csv_every == 0: - write_csv_header = False if os.path.exists(os.path.join(hypernetwork_dir, "hypernetwork_loss.csv")) else True - - with open(os.path.join(hypernetwork_dir, "hypernetwork_loss.csv"), "a+") as fout: - - csv_writer = csv.DictWriter(fout, fieldnames=["step", "loss", "learn_rate"]) - - if write_csv_header: - csv_writer.writeheader() - - csv_writer.writerow({"step": hypernetwork.step, - "loss": f"{losses.mean():.7f}", - "learn_rate": scheduler.learn_rate}) + textual_inversion.write_loss(log_directory, "hypernetwork_loss.csv", hypernetwork.step, len(ds), { + "loss": f"{losses.mean():.7f}", + "learn_rate": scheduler.learn_rate + }) if hypernetwork.step > 0 and images_dir is not None and hypernetwork.step % create_image_every == 0: last_saved_image = os.path.join(images_dir, f'{hypernetwork_name}-{hypernetwork.step}.png') diff --git a/modules/shared.py b/modules/shared.py index 695d29b6..d41a7ab3 100644 --- a/modules/shared.py +++ b/modules/shared.py @@ -236,7 +236,8 @@ options_templates.update(options_section(('training', "Training"), { "unload_models_when_training": OptionInfo(False, "Unload VAE and CLIP from VRAM when training"), "dataset_filename_word_regex": OptionInfo("", "Filename word regex"), "dataset_filename_join_string": OptionInfo(" ", "Filename join string"), - "training_image_repeats_per_epoch": OptionInfo(100, "Number of repeats for a single input image per epoch; used only for displaying epoch number", gr.Number, {"precision": 0}), + "training_image_repeats_per_epoch": OptionInfo(1, "Number of repeats for a single input image per epoch; used only for displaying epoch number", gr.Number, {"precision": 0}), + "training_write_csv_every": OptionInfo(500, "Save an csv containing the loss to log directory every N steps, 0 to disable"), })) options_templates.update(options_section(('sd', "Stable Diffusion"), { diff --git a/modules/textual_inversion/textual_inversion.py b/modules/textual_inversion/textual_inversion.py index 1f5ace6f..da0d77a0 100644 --- a/modules/textual_inversion/textual_inversion.py +++ b/modules/textual_inversion/textual_inversion.py @@ -173,6 +173,32 @@ def create_embedding(name, num_vectors_per_token, init_text='*'): return fn +def write_loss(log_directory, filename, step, epoch_len, values): + if shared.opts.training_write_csv_every == 0: + return + + if step % shared.opts.training_write_csv_every != 0: + return + + write_csv_header = False if os.path.exists(os.path.join(log_directory, filename)) else True + + with open(os.path.join(log_directory, filename), "a+", newline='') as fout: + csv_writer = csv.DictWriter(fout, fieldnames=["step", "epoch", "epoch_step", *(values.keys())]) + + if write_csv_header: + csv_writer.writeheader() + + epoch = step // epoch_len + epoch_step = step - epoch * epoch_len + + csv_writer.writerow({ + "step": step + 1, + "epoch": epoch + 1, + "epoch_step": epoch_step + 1, + **values, + }) + + def train_embedding(embedding_name, learn_rate, data_root, log_directory, training_width, training_height, steps, create_image_every, save_embedding_every, template_file, save_image_with_stored_embedding, preview_from_txt2img, preview_prompt, preview_negative_prompt, preview_steps, preview_sampler_index, preview_cfg_scale, preview_seed, preview_width, preview_height): assert embedding_name, 'embedding not selected' @@ -257,20 +283,10 @@ def train_embedding(embedding_name, learn_rate, data_root, log_directory, traini last_saved_file = os.path.join(embedding_dir, f'{embedding_name}-{embedding.step}.pt') embedding.save(last_saved_file) - if write_csv_every > 0 and log_directory is not None and embedding.step % write_csv_every == 0: - write_csv_header = False if os.path.exists(os.path.join(log_directory, "textual_inversion_loss.csv")) else True - - with open(os.path.join(log_directory, "textual_inversion_loss.csv"), "a+") as fout: - - csv_writer = csv.DictWriter(fout, fieldnames=["epoch", "epoch_step", "loss", "learn_rate"]) - - if write_csv_header: - csv_writer.writeheader() - - csv_writer.writerow({"epoch": epoch_num + 1, - "epoch_step": epoch_step - 1, - "loss": f"{losses.mean():.7f}", - "learn_rate": scheduler.learn_rate}) + write_loss(log_directory, "textual_inversion_loss.csv", embedding.step, len(ds), { + "loss": f"{losses.mean():.7f}", + "learn_rate": scheduler.learn_rate + }) if embedding.step > 0 and images_dir is not None and embedding.step % create_image_every == 0: last_saved_image = os.path.join(images_dir, f'{embedding_name}-{embedding.step}.png') diff --git a/modules/ui.py b/modules/ui.py index be4a43a7..a08ffc9b 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1172,7 +1172,6 @@ def create_ui(wrap_gradio_gpu_call): training_height = gr.Slider(minimum=64, maximum=2048, step=64, label="Height", value=512) steps = gr.Number(label='Max steps', value=100000, precision=0) create_image_every = gr.Number(label='Save an image to log directory every N steps, 0 to disable', value=500, precision=0) - write_csv_every = gr.Number(label='Save an csv containing the loss to log directory every N steps, 0 to disable', value=500, precision=0) save_embedding_every = gr.Number(label='Save a copy of embedding to log directory every N steps, 0 to disable', value=500, precision=0) save_image_with_stored_embedding = gr.Checkbox(label='Save images with embedding in PNG chunks', value=True) preview_from_txt2img = gr.Checkbox(label='Read parameters (prompt, etc...) from txt2img tab when making previews', value=False) @@ -1251,7 +1250,6 @@ def create_ui(wrap_gradio_gpu_call): steps, create_image_every, save_embedding_every, - write_csv_every, template_file, save_image_with_stored_embedding, preview_from_txt2img, @@ -1274,7 +1272,6 @@ def create_ui(wrap_gradio_gpu_call): steps, create_image_every, save_embedding_every, - write_csv_every, template_file, preview_from_txt2img, *txt2img_preview_params, -- cgit v1.2.3 From cd58e44051f658f2efb544203a92837f43786372 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Fri, 14 Oct 2022 23:17:28 +0300 Subject: disabling history - i knew it was slow as fuck but i didn't realize it would also show galleries on launch --- modules/ui.py | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/ui.py b/modules/ui.py index a08ffc9b..6d193955 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1089,7 +1089,8 @@ def create_ui(wrap_gradio_gpu_call): "t2i":txt2img_paste_fields, "i2i":img2img_paste_fields } - images_history = img_his.create_history_tabs(gr, opts, wrap_gradio_call(modules.extras.run_pnginfo), images_history_switch_dict) + + #images_history = img_his.create_history_tabs(gr, opts, wrap_gradio_call(modules.extras.run_pnginfo), images_history_switch_dict) with gr.Blocks() as modelmerger_interface: with gr.Row().style(equal_height=False): @@ -1486,7 +1487,7 @@ Requested path was: {f} (img2img_interface, "img2img", "img2img"), (extras_interface, "Extras", "extras"), (pnginfo_interface, "PNG Info", "pnginfo"), - (images_history, "History", "images_history"), + #(images_history, "History", "images_history"), (modelmerger_interface, "Checkpoint Merger", "modelmerger"), (train_interface, "Train", "ti"), (settings_interface, "Settings", "settings"), -- cgit v1.2.3 From 368f4cc4c73509c1968cd9defe068d8bf4ff7c4f Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Fri, 14 Oct 2022 23:19:05 +0300 Subject: set firstpass w/h to 0 by default and rever to old behavior when any are 0 --- modules/processing.py | 49 ++++++++++++++++++++++++++++++------------------- modules/ui.py | 4 ++-- 2 files changed, 32 insertions(+), 21 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/processing.py b/modules/processing.py index 100a259f..a75b9f84 100644 --- a/modules/processing.py +++ b/modules/processing.py @@ -501,17 +501,15 @@ def process_images(p: StableDiffusionProcessing) -> Processed: class StableDiffusionProcessingTxt2Img(StableDiffusionProcessing): sampler = None - firstphase_width = 0 - firstphase_height = 0 - firstphase_width_truncated = 0 - firstphase_height_truncated = 0 - def __init__(self, enable_hr=False, denoising_strength=0.75, firstphase_width=512, firstphase_height=512, **kwargs): + def __init__(self, enable_hr=False, denoising_strength=0.75, firstphase_width=0, firstphase_height=0, **kwargs): super().__init__(**kwargs) self.enable_hr = enable_hr self.denoising_strength = denoising_strength self.firstphase_width = firstphase_width self.firstphase_height = firstphase_height + self.truncate_x = 0 + self.truncate_y = 0 def init(self, all_prompts, all_seeds, all_subseeds): if self.enable_hr: @@ -520,6 +518,32 @@ class StableDiffusionProcessingTxt2Img(StableDiffusionProcessing): else: state.job_count = state.job_count * 2 + if self.firstphase_width == 0 or self.firstphase_height == 0: + desired_pixel_count = 512 * 512 + actual_pixel_count = self.width * self.height + scale = math.sqrt(desired_pixel_count / actual_pixel_count) + self.firstphase_width = math.ceil(scale * self.width / 64) * 64 + self.firstphase_height = math.ceil(scale * self.height / 64) * 64 + firstphase_width_truncated = int(scale * self.width) + firstphase_height_truncated = int(scale * self.height) + + else: + self.extra_generation_params["First pass size"] = f"{self.firstphase_width}x{self.firstphase_height}" + + width_ratio = self.width / self.firstphase_width + height_ratio = self.height / self.firstphase_height + + if width_ratio > height_ratio: + firstphase_width_truncated = self.firstphase_width + firstphase_height_truncated = self.firstphase_width * self.height / self.width + else: + firstphase_width_truncated = self.firstphase_height * self.width / self.height + firstphase_height_truncated = self.firstphase_height + + self.truncate_x = int(self.firstphase_width - firstphase_width_truncated) // opt_f + self.truncate_y = int(self.firstphase_height - firstphase_height_truncated) // opt_f + + def sample(self, conditioning, unconditional_conditioning, seeds, subseeds, subseed_strength): self.sampler = sd_samplers.create_sampler_with_index(sd_samplers.samplers, self.sampler_index, self.sd_model) @@ -528,23 +552,10 @@ class StableDiffusionProcessingTxt2Img(StableDiffusionProcessing): samples = self.sampler.sample(self, x, conditioning, unconditional_conditioning) return samples - self.extra_generation_params["First pass size"] = f"{self.firstphase_width}x{self.firstphase_height}" - x = create_random_tensors([opt_C, self.firstphase_height // opt_f, self.firstphase_width // opt_f], seeds=seeds, subseeds=subseeds, subseed_strength=self.subseed_strength, seed_resize_from_h=self.seed_resize_from_h, seed_resize_from_w=self.seed_resize_from_w, p=self) samples = self.sampler.sample(self, x, conditioning, unconditional_conditioning) - truncate_x = 0 - truncate_y = 0 - width_ratio = self.width/self.firstphase_width - height_ratio = self.height/self.firstphase_height - - if width_ratio > height_ratio: - truncate_y = int((self.width - self.firstphase_width) / width_ratio / height_ratio / opt_f) - - elif width_ratio < height_ratio: - truncate_x = int((self.height - self.firstphase_height) / width_ratio / height_ratio / opt_f) - - samples = samples[:, :, truncate_y//2:samples.shape[2]-truncate_y//2, truncate_x//2:samples.shape[3]-truncate_x//2] + samples = samples[:, :, self.truncate_y//2:samples.shape[2]-self.truncate_y//2, self.truncate_x//2:samples.shape[3]-self.truncate_x//2] decoded_samples = decode_first_stage(self.sd_model, samples) diff --git a/modules/ui.py b/modules/ui.py index 6d193955..a1d18be9 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -567,8 +567,8 @@ def create_ui(wrap_gradio_gpu_call): enable_hr = gr.Checkbox(label='Highres. fix', value=False) with gr.Row(visible=False) as hr_options: - firstphase_width = gr.Slider(minimum=64, maximum=1024, step=64, label="First pass width", value=512) - firstphase_height = gr.Slider(minimum=64, maximum=1024, step=64, label="First pass height", value=512) + firstphase_width = gr.Slider(minimum=0, maximum=1024, step=64, label="First pass width", value=0) + firstphase_height = gr.Slider(minimum=0, maximum=1024, step=64, label="First pass height", value=0) denoising_strength = gr.Slider(minimum=0.0, maximum=1.0, step=0.01, label='Denoising strength', value=0.7) with gr.Row(equal_height=True): -- cgit v1.2.3 From 4dc426509918e90bf4557ecfd1f84031362360c0 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Sat, 15 Oct 2022 00:21:48 +0300 Subject: rename firstpass w/h to discard old user settings --- modules/ui.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/ui.py b/modules/ui.py index a1d18be9..c5d295ea 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -567,8 +567,8 @@ def create_ui(wrap_gradio_gpu_call): enable_hr = gr.Checkbox(label='Highres. fix', value=False) with gr.Row(visible=False) as hr_options: - firstphase_width = gr.Slider(minimum=0, maximum=1024, step=64, label="First pass width", value=0) - firstphase_height = gr.Slider(minimum=0, maximum=1024, step=64, label="First pass height", value=0) + firstphase_width = gr.Slider(minimum=0, maximum=1024, step=64, label="Firstpass width", value=0) + firstphase_height = gr.Slider(minimum=0, maximum=1024, step=64, label="Firstpass height", value=0) denoising_strength = gr.Slider(minimum=0.0, maximum=1.0, step=0.01, label='Denoising strength', value=0.7) with gr.Row(equal_height=True): -- cgit v1.2.3 From acedbe67d2b8a3af99ca3b9a2f809e7a2db285d1 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Sat, 15 Oct 2022 00:43:15 +0300 Subject: bring history tab back, make it behave; it's still slow but won't fuck anything up until you use it --- javascript/images_history.js | 16 ++++++++++++---- modules/ui.py | 4 ++-- 2 files changed, 14 insertions(+), 6 deletions(-) (limited to 'modules/ui.py') diff --git a/javascript/images_history.js b/javascript/images_history.js index 3a20056b..f7d052c3 100644 --- a/javascript/images_history.js +++ b/javascript/images_history.js @@ -163,10 +163,15 @@ function images_history_init(){ for (var i in images_history_tab_list){ var tabname = images_history_tab_list[i] tab_btns[i].setAttribute("tabname", tabname); - tab_btns[i].addEventListener('click', images_history_click_tab); + + // this refreshes history upon tab switch + // until the history is known to work well, which is not the case now, we do not do this at startup + //tab_btns[i].addEventListener('click', images_history_click_tab); } - tabs_box.classList.add(images_history_tab_list[0]); - load_txt2img_button.click(); + tabs_box.classList.add(images_history_tab_list[0]); + + // same as above, at page load + //load_txt2img_button.click(); } else { setTimeout(images_history_init, 500); } @@ -182,12 +187,15 @@ document.addEventListener("DOMContentLoaded", function() { buttons.forEach(function(bnt){ bnt.addEventListener('click', images_history_click_image, true); }); + + // same as load_txt2img_button.click() above + /* var cls_btn = gradioApp().getElementById(tabname + '_images_history_gallery').querySelector("svg"); if (cls_btn){ cls_btn.addEventListener('click', function(){ gradioApp().getElementById(tabname + '_images_history_renew_page').click(); }, false); - } + }*/ } }); diff --git a/modules/ui.py b/modules/ui.py index c5d295ea..1bc919c7 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1090,7 +1090,7 @@ def create_ui(wrap_gradio_gpu_call): "i2i":img2img_paste_fields } - #images_history = img_his.create_history_tabs(gr, opts, wrap_gradio_call(modules.extras.run_pnginfo), images_history_switch_dict) + images_history = img_his.create_history_tabs(gr, opts, wrap_gradio_call(modules.extras.run_pnginfo), images_history_switch_dict) with gr.Blocks() as modelmerger_interface: with gr.Row().style(equal_height=False): @@ -1487,7 +1487,7 @@ Requested path was: {f} (img2img_interface, "img2img", "img2img"), (extras_interface, "Extras", "extras"), (pnginfo_interface, "PNG Info", "pnginfo"), - #(images_history, "History", "images_history"), + (images_history, "History", "images_history"), (modelmerger_interface, "Checkpoint Merger", "modelmerger"), (train_interface, "Train", "ti"), (settings_interface, "Settings", "settings"), -- cgit v1.2.3 From c7a86f7fe9c0b8967a87e8d709f507d2f44400d8 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Sat, 15 Oct 2022 09:24:59 +0300 Subject: add option to use batch size for training --- modules/hypernetworks/hypernetwork.py | 33 +++++++++++++++++++------- modules/textual_inversion/dataset.py | 31 ++++++++++++++---------- modules/textual_inversion/textual_inversion.py | 17 +++++++------ modules/ui.py | 3 +++ 4 files changed, 54 insertions(+), 30 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/hypernetworks/hypernetwork.py b/modules/hypernetworks/hypernetwork.py index 59c7ac6e..a2b3bc0a 100644 --- a/modules/hypernetworks/hypernetwork.py +++ b/modules/hypernetworks/hypernetwork.py @@ -182,7 +182,21 @@ def attention_CrossAttention_forward(self, x, context=None, mask=None): return self.to_out(out) -def train_hypernetwork(hypernetwork_name, learn_rate, data_root, log_directory, steps, create_image_every, save_hypernetwork_every, template_file, preview_from_txt2img, preview_prompt, preview_negative_prompt, preview_steps, preview_sampler_index, preview_cfg_scale, preview_seed, preview_width, preview_height): +def stack_conds(conds): + if len(conds) == 1: + return torch.stack(conds) + + # same as in reconstruct_multicond_batch + token_count = max([x.shape[0] for x in conds]) + for i in range(len(conds)): + if conds[i].shape[0] != token_count: + last_vector = conds[i][-1:] + last_vector_repeated = last_vector.repeat([token_count - conds[i].shape[0], 1]) + conds[i] = torch.vstack([conds[i], last_vector_repeated]) + + return torch.stack(conds) + +def train_hypernetwork(hypernetwork_name, learn_rate, batch_size, data_root, log_directory, steps, create_image_every, save_hypernetwork_every, template_file, preview_from_txt2img, preview_prompt, preview_negative_prompt, preview_steps, preview_sampler_index, preview_cfg_scale, preview_seed, preview_width, preview_height): assert hypernetwork_name, 'hypernetwork not selected' path = shared.hypernetworks.get(hypernetwork_name, None) @@ -211,7 +225,7 @@ def train_hypernetwork(hypernetwork_name, learn_rate, data_root, log_directory, shared.state.textinfo = f"Preparing dataset from {html.escape(data_root)}..." with torch.autocast("cuda"): - ds = modules.textual_inversion.dataset.PersonalizedBase(data_root=data_root, width=512, height=512, repeats=shared.opts.training_image_repeats_per_epoch, placeholder_token=hypernetwork_name, model=shared.sd_model, device=devices.device, template_file=template_file, include_cond=True) + ds = modules.textual_inversion.dataset.PersonalizedBase(data_root=data_root, width=512, height=512, repeats=shared.opts.training_image_repeats_per_epoch, placeholder_token=hypernetwork_name, model=shared.sd_model, device=devices.device, template_file=template_file, include_cond=True, batch_size=batch_size) if unload: shared.sd_model.cond_stage_model.to(devices.cpu) @@ -235,7 +249,7 @@ def train_hypernetwork(hypernetwork_name, learn_rate, data_root, log_directory, optimizer = torch.optim.AdamW(weights, lr=scheduler.learn_rate) pbar = tqdm.tqdm(enumerate(ds), total=steps - ititial_step) - for i, entry in pbar: + for i, entries in pbar: hypernetwork.step = i + ititial_step scheduler.apply(optimizer, hypernetwork.step) @@ -246,11 +260,12 @@ def train_hypernetwork(hypernetwork_name, learn_rate, data_root, log_directory, break with torch.autocast("cuda"): - cond = entry.cond.to(devices.device) - x = entry.latent.to(devices.device) - loss = shared.sd_model(x.unsqueeze(0), cond)[0] + c = stack_conds([entry.cond for entry in entries]).to(devices.device) +# c = torch.vstack([entry.cond for entry in entries]).to(devices.device) + x = torch.stack([entry.latent for entry in entries]).to(devices.device) + loss = shared.sd_model(x, c)[0] del x - del cond + del c losses[hypernetwork.step % losses.shape[0]] = loss.item() @@ -292,7 +307,7 @@ def train_hypernetwork(hypernetwork_name, learn_rate, data_root, log_directory, p.width = preview_width p.height = preview_height else: - p.prompt = entry.cond_text + p.prompt = entries[0].cond_text p.steps = 20 preview_text = p.prompt @@ -315,7 +330,7 @@ def train_hypernetwork(hypernetwork_name, learn_rate, data_root, log_directory,

Loss: {losses.mean():.7f}
Step: {hypernetwork.step}
-Last prompt: {html.escape(entry.cond_text)}
+Last prompt: {html.escape(entries[0].cond_text)}
Last saved embedding: {html.escape(last_saved_file)}
Last saved image: {html.escape(last_saved_image)}

diff --git a/modules/textual_inversion/dataset.py b/modules/textual_inversion/dataset.py index 67e90afe..bd99c0cb 100644 --- a/modules/textual_inversion/dataset.py +++ b/modules/textual_inversion/dataset.py @@ -24,11 +24,12 @@ class DatasetEntry: class PersonalizedBase(Dataset): - def __init__(self, data_root, width, height, repeats, flip_p=0.5, placeholder_token="*", model=None, device=None, template_file=None, include_cond=False): - re_word = re.compile(shared.opts.dataset_filename_word_regex) if len(shared.opts.dataset_filename_word_regex)>0 else None + def __init__(self, data_root, width, height, repeats, flip_p=0.5, placeholder_token="*", model=None, device=None, template_file=None, include_cond=False, batch_size=1): + re_word = re.compile(shared.opts.dataset_filename_word_regex) if len(shared.opts.dataset_filename_word_regex) > 0 else None self.placeholder_token = placeholder_token + self.batch_size = batch_size self.width = width self.height = height self.flip = transforms.RandomHorizontalFlip(p=flip_p) @@ -78,13 +79,13 @@ class PersonalizedBase(Dataset): if include_cond: entry.cond_text = self.create_text(filename_text) - entry.cond = cond_model([entry.cond_text]).to(devices.cpu) + entry.cond = cond_model([entry.cond_text]).to(devices.cpu).squeeze(0) self.dataset.append(entry) - self.length = len(self.dataset) * repeats + self.length = len(self.dataset) * repeats // batch_size - self.initial_indexes = np.arange(self.length) % len(self.dataset) + self.initial_indexes = np.arange(len(self.dataset)) self.indexes = None self.shuffle() @@ -101,13 +102,19 @@ class PersonalizedBase(Dataset): return self.length def __getitem__(self, i): - if i % len(self.dataset) == 0: - self.shuffle() + res = [] - index = self.indexes[i % len(self.indexes)] - entry = self.dataset[index] + for j in range(self.batch_size): + position = i * self.batch_size + j + if position % len(self.indexes) == 0: + self.shuffle() - if entry.cond is None: - entry.cond_text = self.create_text(entry.filename_text) + index = self.indexes[position % len(self.indexes)] + entry = self.dataset[index] - return entry + if entry.cond is None: + entry.cond_text = self.create_text(entry.filename_text) + + res.append(entry) + + return res diff --git a/modules/textual_inversion/textual_inversion.py b/modules/textual_inversion/textual_inversion.py index da0d77a0..e754747e 100644 --- a/modules/textual_inversion/textual_inversion.py +++ b/modules/textual_inversion/textual_inversion.py @@ -199,7 +199,7 @@ def write_loss(log_directory, filename, step, epoch_len, values): }) -def train_embedding(embedding_name, learn_rate, data_root, log_directory, training_width, training_height, steps, create_image_every, save_embedding_every, template_file, save_image_with_stored_embedding, preview_from_txt2img, preview_prompt, preview_negative_prompt, preview_steps, preview_sampler_index, preview_cfg_scale, preview_seed, preview_width, preview_height): +def train_embedding(embedding_name, learn_rate, batch_size, data_root, log_directory, training_width, training_height, steps, create_image_every, save_embedding_every, template_file, save_image_with_stored_embedding, preview_from_txt2img, preview_prompt, preview_negative_prompt, preview_steps, preview_sampler_index, preview_cfg_scale, preview_seed, preview_width, preview_height): assert embedding_name, 'embedding not selected' shared.state.textinfo = "Initializing textual inversion training..." @@ -231,7 +231,7 @@ def train_embedding(embedding_name, learn_rate, data_root, log_directory, traini shared.state.textinfo = f"Preparing dataset from {html.escape(data_root)}..." with torch.autocast("cuda"): - ds = modules.textual_inversion.dataset.PersonalizedBase(data_root=data_root, width=training_width, height=training_height, repeats=shared.opts.training_image_repeats_per_epoch, placeholder_token=embedding_name, model=shared.sd_model, device=devices.device, template_file=template_file) + ds = modules.textual_inversion.dataset.PersonalizedBase(data_root=data_root, width=training_width, height=training_height, repeats=shared.opts.training_image_repeats_per_epoch, placeholder_token=embedding_name, model=shared.sd_model, device=devices.device, template_file=template_file, batch_size=batch_size) hijack = sd_hijack.model_hijack @@ -251,7 +251,7 @@ def train_embedding(embedding_name, learn_rate, data_root, log_directory, traini optimizer = torch.optim.AdamW([embedding.vec], lr=scheduler.learn_rate) pbar = tqdm.tqdm(enumerate(ds), total=steps-ititial_step) - for i, entry in pbar: + for i, entries in pbar: embedding.step = i + ititial_step scheduler.apply(optimizer, embedding.step) @@ -262,10 +262,9 @@ def train_embedding(embedding_name, learn_rate, data_root, log_directory, traini break with torch.autocast("cuda"): - c = cond_model([entry.cond_text]) - - x = entry.latent.to(devices.device) - loss = shared.sd_model(x.unsqueeze(0), c)[0] + c = cond_model([entry.cond_text for entry in entries]) + x = torch.stack([entry.latent for entry in entries]).to(devices.device) + loss = shared.sd_model(x, c)[0] del x losses[embedding.step % losses.shape[0]] = loss.item() @@ -307,7 +306,7 @@ def train_embedding(embedding_name, learn_rate, data_root, log_directory, traini p.width = preview_width p.height = preview_height else: - p.prompt = entry.cond_text + p.prompt = entries[0].cond_text p.steps = 20 p.width = training_width p.height = training_height @@ -348,7 +347,7 @@ def train_embedding(embedding_name, learn_rate, data_root, log_directory, traini

Loss: {losses.mean():.7f}
Step: {embedding.step}
-Last prompt: {html.escape(entry.cond_text)}
+Last prompt: {html.escape(entries[0].cond_text)}
Last saved embedding: {html.escape(last_saved_file)}
Last saved image: {html.escape(last_saved_image)}

diff --git a/modules/ui.py b/modules/ui.py index 1bc919c7..45550ea8 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1166,6 +1166,7 @@ def create_ui(wrap_gradio_gpu_call): train_embedding_name = gr.Dropdown(label='Embedding', choices=sorted(sd_hijack.model_hijack.embedding_db.word_embeddings.keys())) train_hypernetwork_name = gr.Dropdown(label='Hypernetwork', choices=[x for x in shared.hypernetworks.keys()]) learn_rate = gr.Textbox(label='Learning rate', placeholder="Learning rate", value="0.005") + batch_size = gr.Number(label='Batch size', value=1, precision=0) dataset_directory = gr.Textbox(label='Dataset directory', placeholder="Path to directory with input images") log_directory = gr.Textbox(label='Log directory', placeholder="Path to directory where to write outputs", value="textual_inversion") template_file = gr.Textbox(label='Prompt template file', value=os.path.join(script_path, "textual_inversion_templates", "style_filewords.txt")) @@ -1244,6 +1245,7 @@ def create_ui(wrap_gradio_gpu_call): inputs=[ train_embedding_name, learn_rate, + batch_size, dataset_directory, log_directory, training_width, @@ -1268,6 +1270,7 @@ def create_ui(wrap_gradio_gpu_call): inputs=[ train_hypernetwork_name, learn_rate, + batch_size, dataset_directory, log_directory, steps, -- cgit v1.2.3 From db27b987a97fc8b7894a9dd34bd7641536f9c424 Mon Sep 17 00:00:00 2001 From: aoirusann Date: Sat, 15 Oct 2022 11:48:13 +0800 Subject: Add hint for `ctrl/alt enter` And duplicate implementations are removed --- javascript/ui.js | 10 ---------- modules/ui.py | 10 ++++++++-- script.js | 4 ++-- 3 files changed, 10 insertions(+), 14 deletions(-) (limited to 'modules/ui.py') diff --git a/javascript/ui.js b/javascript/ui.js index 0f8fe68e..56f4216f 100644 --- a/javascript/ui.js +++ b/javascript/ui.js @@ -187,12 +187,10 @@ onUiUpdate(function(){ if (!txt2img_textarea) { txt2img_textarea = gradioApp().querySelector("#txt2img_prompt > label > textarea"); txt2img_textarea?.addEventListener("input", () => update_token_counter("txt2img_token_button")); - txt2img_textarea?.addEventListener("keyup", (event) => submit_prompt(event, "txt2img_generate")); } if (!img2img_textarea) { img2img_textarea = gradioApp().querySelector("#img2img_prompt > label > textarea"); img2img_textarea?.addEventListener("input", () => update_token_counter("img2img_token_button")); - img2img_textarea?.addEventListener("keyup", (event) => submit_prompt(event, "img2img_generate")); } }) @@ -220,14 +218,6 @@ function update_token_counter(button_id) { token_timeout = setTimeout(() => gradioApp().getElementById(button_id)?.click(), wait_time); } -function submit_prompt(event, generate_button_id) { - if (event.altKey && event.keyCode === 13) { - event.preventDefault(); - gradioApp().getElementById(generate_button_id).click(); - return; - } -} - function restart_reload(){ document.body.innerHTML='

Reloading...

'; setTimeout(function(){location.reload()},2000) diff --git a/modules/ui.py b/modules/ui.py index 45550ea8..baf4c397 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -433,7 +433,10 @@ def create_toprow(is_img2img): with gr.Row(): with gr.Column(scale=80): with gr.Row(): - prompt = gr.Textbox(label="Prompt", elem_id=f"{id_part}_prompt", show_label=False, placeholder="Prompt", lines=2) + prompt = gr.Textbox(label="Prompt", elem_id=f"{id_part}_prompt", show_label=False, lines=2, + placeholder="Prompt (press Ctrl+Enter or Alt+Enter to generate)" + ) + with gr.Column(scale=1, elem_id="roll_col"): roll = gr.Button(value=art_symbol, elem_id="roll", visible=len(shared.artist_db.artists) > 0) paste = gr.Button(value=paste_symbol, elem_id="paste") @@ -446,7 +449,10 @@ def create_toprow(is_img2img): with gr.Row(): with gr.Column(scale=8): with gr.Row(): - negative_prompt = gr.Textbox(label="Negative prompt", elem_id="negative_prompt", show_label=False, placeholder="Negative prompt", lines=2) + negative_prompt = gr.Textbox(label="Negative prompt", elem_id=f"{id_part}_neg_prompt", show_label=False, lines=2, + placeholder="Negative prompt (press Ctrl+Enter or Alt+Enter to generate)" + ) + with gr.Column(scale=1, elem_id="roll_col"): sh = gr.Button(elem_id="sh", visible=True) diff --git a/script.js b/script.js index 9543cbe6..88f2c839 100644 --- a/script.js +++ b/script.js @@ -50,9 +50,9 @@ document.addEventListener("DOMContentLoaded", function() { document.addEventListener('keydown', function(e) { var handled = false; if (e.key !== undefined) { - if((e.key == "Enter" && (e.metaKey || e.ctrlKey))) handled = true; + if((e.key == "Enter" && (e.metaKey || e.ctrlKey || e.altKey))) handled = true; } else if (e.keyCode !== undefined) { - if((e.keyCode == 13 && (e.metaKey || e.ctrlKey))) handled = true; + if((e.keyCode == 13 && (e.metaKey || e.ctrlKey || e.altKey))) handled = true; } if (handled) { button = get_uiCurrentTabContent().querySelector('button[id$=_generate]'); -- cgit v1.2.3