From 14c1c2b9351f16d43ba4e6b6c9062edad44a6bec Mon Sep 17 00:00:00 2001 From: Alexandre Simard Date: Wed, 19 Oct 2022 13:53:52 -0400 Subject: Show PB texts at same time and earlier For big tasks (1000+ steps), waiting 1 minute to see ETA is long and this changes it so the number of steps done plays a role in showing the text as well. --- modules/ui.py | 11 +++++++---- 1 file changed, 7 insertions(+), 4 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/ui.py b/modules/ui.py index a2dbd41e..0abd177a 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -261,14 +261,14 @@ def wrap_gradio_call(func, extra_outputs=None): return f -def calc_time_left(progress, threshold, label, force_display): +def calc_time_left(progress, threshold, label, force_display, showTime): if progress == 0: return "" else: time_since_start = time.time() - shared.state.time_start eta = (time_since_start/progress) eta_relative = eta-time_since_start - if (eta_relative > threshold and progress > 0.02) or force_display: + if (eta_relative > threshold and showTime) or force_display: if eta_relative > 3600: return label + time.strftime('%H:%M:%S', time.gmtime(eta_relative)) elif eta_relative > 60: @@ -290,7 +290,10 @@ def check_progress_call(id_part): if shared.state.sampling_steps > 0: progress += 1 / shared.state.job_count * shared.state.sampling_step / shared.state.sampling_steps - time_left = calc_time_left( progress, 1, " ETA: ", shared.state.time_left_force_display ) + # Show progress percentage and time left at the same moment, and base it also on steps done + showPBText = progress >= 0.01 or shared.state.sampling_step >= 10 + + time_left = calc_time_left( progress, 1, " ETA: ", shared.state.time_left_force_display, showPBText ) if time_left != "": shared.state.time_left_force_display = True @@ -298,7 +301,7 @@ def check_progress_call(id_part): progressbar = "" if opts.show_progressbar: - progressbar = f"""
{" " * 2 + str(int(progress*100))+"%" + time_left if progress > 0.01 else ""}
""" + progressbar = f"""
{" " * 2 + str(int(progress*100))+"%" + time_left if showPBText else ""}
""" image = gr_show(False) preview_visibility = gr_show(False) -- cgit v1.2.3 From 2a25729623717cc499e873752d9f4ebebd1e1078 Mon Sep 17 00:00:00 2001 From: Muhammad Rizqi Nur Date: Fri, 28 Oct 2022 09:44:56 +0700 Subject: Gradient clipping in train tab --- modules/hypernetworks/hypernetwork.py | 10 +++++++++- modules/ui.py | 7 +++++++ 2 files changed, 16 insertions(+), 1 deletion(-) (limited to 'modules/ui.py') diff --git a/modules/hypernetworks/hypernetwork.py b/modules/hypernetworks/hypernetwork.py index 8113b35b..c5d60654 100644 --- a/modules/hypernetworks/hypernetwork.py +++ b/modules/hypernetworks/hypernetwork.py @@ -327,7 +327,7 @@ def report_statistics(loss_info:dict): -def train_hypernetwork(hypernetwork_name, learn_rate, batch_size, data_root, log_directory, training_width, training_height, 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 train_hypernetwork(hypernetwork_name, learn_rate, batch_size, data_root, log_directory, training_width, training_height, steps, clip_grad_mode, clip_grad_value, 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): # images allows training previews to have infotext. Importing it at the top causes a circular import problem. from modules import images @@ -384,6 +384,9 @@ def train_hypernetwork(hypernetwork_name, learn_rate, batch_size, data_root, log if ititial_step > steps: return hypernetwork, filename + clip_grad_mode_value = clip_grad_mode == "value" + clip_grad_mode_norm = clip_grad_mode == "norm" + scheduler = LearnRateScheduler(learn_rate, steps, ititial_step) # if optimizer == "AdamW": or else Adam / AdamW / SGD, etc... optimizer = torch.optim.AdamW(weights, lr=scheduler.learn_rate) @@ -426,6 +429,11 @@ def train_hypernetwork(hypernetwork_name, learn_rate, batch_size, data_root, log steps_without_grad = 0 assert steps_without_grad < 10, 'no gradient found for the trained weight after backward() for 10 steps in a row; this is a bug; training cannot continue' + if clip_grad_mode_value: + torch.nn.utils.clip_grad_value_(weights, clip_value=clip_grad_value) + elif clip_grad_mode_norm: + torch.nn.utils.clip_grad_norm_(weights, max_norm=clip_grad_value) + optimizer.step() if torch.isnan(losses[hypernetwork.step % losses.shape[0]]): diff --git a/modules/ui.py b/modules/ui.py index 0a63e357..97de7da2 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1313,6 +1313,9 @@ 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) + with gr.Row(): + clip_grad_mode = gr.Dropdown(value="disabled", label="Gradient Clipping", choices=["disabled", "value", "norm"]) + clip_grad_value = gr.Number(value=1.0, show_label=False) 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) @@ -1406,6 +1409,8 @@ def create_ui(wrap_gradio_gpu_call): training_width, training_height, steps, + clip_grad_mode, + clip_grad_value, create_image_every, save_embedding_every, template_file, @@ -1431,6 +1436,8 @@ def create_ui(wrap_gradio_gpu_call): training_width, training_height, steps, + clip_grad_mode, + clip_grad_value, create_image_every, save_embedding_every, template_file, -- cgit v1.2.3 From a133042c669f666763f5da0f4440abdc839db653 Mon Sep 17 00:00:00 2001 From: Muhammad Rizqi Nur Date: Fri, 28 Oct 2022 10:01:46 +0700 Subject: Forgot to remove this from train_embedding --- modules/ui.py | 2 -- 1 file changed, 2 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/ui.py b/modules/ui.py index 97de7da2..ba5e92a7 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1409,8 +1409,6 @@ def create_ui(wrap_gradio_gpu_call): training_width, training_height, steps, - clip_grad_mode, - clip_grad_value, create_image_every, save_embedding_every, template_file, -- cgit v1.2.3 From 1618df41bad092e068c61bf510b1e20856821ad5 Mon Sep 17 00:00:00 2001 From: Muhammad Rizqi Nur Date: Fri, 28 Oct 2022 10:31:27 +0700 Subject: Gradient clipping for textual embedding --- modules/textual_inversion/textual_inversion.py | 11 ++++++++++- modules/ui.py | 2 ++ 2 files changed, 12 insertions(+), 1 deletion(-) (limited to 'modules/ui.py') diff --git a/modules/textual_inversion/textual_inversion.py b/modules/textual_inversion/textual_inversion.py index ff002d3e..7bad73a6 100644 --- a/modules/textual_inversion/textual_inversion.py +++ b/modules/textual_inversion/textual_inversion.py @@ -206,7 +206,7 @@ def write_loss(log_directory, filename, step, epoch_len, values): }) -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): +def train_embedding(embedding_name, learn_rate, batch_size, data_root, log_directory, training_width, training_height, steps, clip_grad_mode, clip_grad_value, 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..." @@ -256,6 +256,9 @@ def train_embedding(embedding_name, learn_rate, batch_size, data_root, log_direc if ititial_step > steps: return embedding, filename + clip_grad_mode_value = clip_grad_mode == "value" + clip_grad_mode_norm = clip_grad_mode == "norm" + scheduler = LearnRateScheduler(learn_rate, steps, ititial_step) optimizer = torch.optim.AdamW([embedding.vec], lr=scheduler.learn_rate) @@ -280,6 +283,12 @@ def train_embedding(embedding_name, learn_rate, batch_size, data_root, log_direc optimizer.zero_grad() loss.backward() + + if clip_grad_mode_value: + torch.nn.utils.clip_grad_value_(embedding.vec, clip_value=clip_grad_value) + elif clip_grad_mode_norm: + torch.nn.utils.clip_grad_norm_(embedding.vec, max_norm=clip_grad_value) + optimizer.step() diff --git a/modules/ui.py b/modules/ui.py index ba5e92a7..97de7da2 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1409,6 +1409,8 @@ def create_ui(wrap_gradio_gpu_call): training_width, training_height, steps, + clip_grad_mode, + clip_grad_value, create_image_every, save_embedding_every, template_file, -- cgit v1.2.3 From 16451ca573220e49f2eaaab97580b6b91287c8c4 Mon Sep 17 00:00:00 2001 From: Muhammad Rizqi Nur Date: Fri, 28 Oct 2022 17:16:23 +0700 Subject: Learning rate sched syntax support for grad clipping --- modules/hypernetworks/hypernetwork.py | 13 ++++++++++--- modules/textual_inversion/learn_schedule.py | 11 ++++++++--- modules/textual_inversion/textual_inversion.py | 12 +++++++++--- modules/ui.py | 7 +++---- 4 files changed, 30 insertions(+), 13 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/hypernetworks/hypernetwork.py b/modules/hypernetworks/hypernetwork.py index c5d60654..86532063 100644 --- a/modules/hypernetworks/hypernetwork.py +++ b/modules/hypernetworks/hypernetwork.py @@ -383,11 +383,15 @@ def train_hypernetwork(hypernetwork_name, learn_rate, batch_size, data_root, log ititial_step = hypernetwork.step or 0 if ititial_step > steps: return hypernetwork, filename - + clip_grad_mode_value = clip_grad_mode == "value" clip_grad_mode_norm = clip_grad_mode == "norm" + clip_grad_enabled = clip_grad_mode_value or clip_grad_mode_norm + if clip_grad_enabled: + clip_grad_sched = LearnRateScheduler(clip_grad_value, steps, ititial_step, verbose=False) scheduler = LearnRateScheduler(learn_rate, steps, ititial_step) + # if optimizer == "AdamW": or else Adam / AdamW / SGD, etc... optimizer = torch.optim.AdamW(weights, lr=scheduler.learn_rate) @@ -407,6 +411,9 @@ def train_hypernetwork(hypernetwork_name, learn_rate, batch_size, data_root, log if shared.state.interrupted: break + if clip_grad_enabled: + clip_grad_sched.step(hypernetwork.step) + with torch.autocast("cuda"): c = stack_conds([entry.cond for entry in entries]).to(devices.device) # c = torch.vstack([entry.cond for entry in entries]).to(devices.device) @@ -430,9 +437,9 @@ def train_hypernetwork(hypernetwork_name, learn_rate, batch_size, data_root, log assert steps_without_grad < 10, 'no gradient found for the trained weight after backward() for 10 steps in a row; this is a bug; training cannot continue' if clip_grad_mode_value: - torch.nn.utils.clip_grad_value_(weights, clip_value=clip_grad_value) + torch.nn.utils.clip_grad_value_(weights, clip_value=clip_grad_sched.learn_rate) elif clip_grad_mode_norm: - torch.nn.utils.clip_grad_norm_(weights, max_norm=clip_grad_value) + torch.nn.utils.clip_grad_norm_(weights, max_norm=clip_grad_sched.learn_rate) optimizer.step() diff --git a/modules/textual_inversion/learn_schedule.py b/modules/textual_inversion/learn_schedule.py index 2062726a..ffec3e1b 100644 --- a/modules/textual_inversion/learn_schedule.py +++ b/modules/textual_inversion/learn_schedule.py @@ -51,14 +51,19 @@ class LearnRateScheduler: self.finished = False - def apply(self, optimizer, step_number): + def step(self, step_number): if step_number <= self.end_step: - return + return False try: (self.learn_rate, self.end_step) = next(self.schedules) - except Exception: + except StopIteration: self.finished = True + return False + return True + + def apply(self, optimizer, step_number): + if not self.step(step_number): return if self.verbose: diff --git a/modules/textual_inversion/textual_inversion.py b/modules/textual_inversion/textual_inversion.py index 7bad73a6..6b00c6a1 100644 --- a/modules/textual_inversion/textual_inversion.py +++ b/modules/textual_inversion/textual_inversion.py @@ -255,9 +255,12 @@ def train_embedding(embedding_name, learn_rate, batch_size, data_root, log_direc ititial_step = embedding.step or 0 if ititial_step > steps: return embedding, filename - + clip_grad_mode_value = clip_grad_mode == "value" clip_grad_mode_norm = clip_grad_mode == "norm" + clip_grad_enabled = clip_grad_mode_value or clip_grad_mode_norm + if clip_grad_enabled: + clip_grad_sched = LearnRateScheduler(clip_grad_value, steps, ititial_step, verbose=False) scheduler = LearnRateScheduler(learn_rate, steps, ititial_step) optimizer = torch.optim.AdamW([embedding.vec], lr=scheduler.learn_rate) @@ -273,6 +276,9 @@ def train_embedding(embedding_name, learn_rate, batch_size, data_root, log_direc if shared.state.interrupted: break + if clip_grad_enabled: + clip_grad_sched.step(embedding.step) + with torch.autocast("cuda"): c = cond_model([entry.cond_text for entry in entries]) x = torch.stack([entry.latent for entry in entries]).to(devices.device) @@ -285,9 +291,9 @@ def train_embedding(embedding_name, learn_rate, batch_size, data_root, log_direc loss.backward() if clip_grad_mode_value: - torch.nn.utils.clip_grad_value_(embedding.vec, clip_value=clip_grad_value) + torch.nn.utils.clip_grad_value_(embedding.vec, clip_value=clip_grad_sched.learn_rate) elif clip_grad_mode_norm: - torch.nn.utils.clip_grad_norm_(embedding.vec, max_norm=clip_grad_value) + torch.nn.utils.clip_grad_norm_(embedding.vec, max_norm=clip_grad_sched.learn_rate) optimizer.step() diff --git a/modules/ui.py b/modules/ui.py index 97de7da2..47d16429 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1305,7 +1305,9 @@ def create_ui(wrap_gradio_gpu_call): with gr.Row(): embedding_learn_rate = gr.Textbox(label='Embedding Learning rate', placeholder="Embedding Learning rate", value="0.005") hypernetwork_learn_rate = gr.Textbox(label='Hypernetwork Learning rate', placeholder="Hypernetwork Learning rate", value="0.00001") - + with gr.Row(): + clip_grad_mode = gr.Dropdown(value="disabled", label="Gradient Clipping", choices=["disabled", "value", "norm"]) + clip_grad_value = gr.Textbox(placeholder="Gradient clip value", value="1.0", show_label=False) 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") @@ -1313,9 +1315,6 @@ 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) - with gr.Row(): - clip_grad_mode = gr.Dropdown(value="disabled", label="Gradient Clipping", choices=["disabled", "value", "norm"]) - clip_grad_value = gr.Number(value=1.0, show_label=False) 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) -- cgit v1.2.3 From 840307f23738c38f7ac3ad636e53ccec66e71f8b Mon Sep 17 00:00:00 2001 From: Muhammad Rizqi Nur Date: Mon, 31 Oct 2022 13:49:24 +0700 Subject: Change default clip grad value to 0.1 It still defaults to disabled. Ref for value: https://github.com/danielalcalde/stable-diffusion-webui/commit/732b15820a9bde9f47e075a6209c3d47d47acb08 --- 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 98f9565f..364953aa 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1256,7 +1256,7 @@ def create_ui(wrap_gradio_gpu_call): hypernetwork_learn_rate = gr.Textbox(label='Hypernetwork Learning rate', placeholder="Hypernetwork Learning rate", value="0.00001") with gr.Row(): clip_grad_mode = gr.Dropdown(value="disabled", label="Gradient Clipping", choices=["disabled", "value", "norm"]) - clip_grad_value = gr.Textbox(placeholder="Gradient clip value", value="1.0", show_label=False) + clip_grad_value = gr.Textbox(placeholder="Gradient clip value", value="0.1", show_label=False) 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") -- cgit v1.2.3 From 917b5bd8d0cd47c9dc241c1852ccd440a8c61668 Mon Sep 17 00:00:00 2001 From: Max Weber Date: Tue, 3 Jan 2023 18:19:56 -0700 Subject: ui: save dropdown sampling method to the ui-config --- 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 d941cb5f..bfc93634 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -635,6 +635,7 @@ def create_sampler_and_steps_selection(choices, tabname): if opts.samplers_in_dropdown: with FormRow(elem_id=f"sampler_selection_{tabname}"): sampler_index = gr.Dropdown(label='Sampling method', elem_id=f"{tabname}_sampling", choices=[x.name for x in choices], value=choices[0].name, type="index") + sampler_index.save_to_config = True steps = gr.Slider(minimum=1, maximum=150, step=1, elem_id=f"{tabname}_steps", label="Sampling Steps", value=20) else: with FormGroup(elem_id=f"sampler_selection_{tabname}"): -- cgit v1.2.3 From 3dae545a03f5102ba5d9c3f27bb6241824c5a916 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Wed, 4 Jan 2023 18:42:51 +0300 Subject: rename weirdly named variables from #3176 --- modules/ui.py | 12 +++++------- 1 file changed, 5 insertions(+), 7 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/ui.py b/modules/ui.py index e4859020..184af7ad 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -162,16 +162,14 @@ def save_files(js_data, images, do_make_zip, index): return gr.File.update(value=fullfns, visible=True), plaintext_to_html(f"Saved: {filenames[0]}") - - -def calc_time_left(progress, threshold, label, force_display, showTime): +def calc_time_left(progress, threshold, label, force_display, show_eta): if progress == 0: return "" else: time_since_start = time.time() - shared.state.time_start eta = (time_since_start/progress) eta_relative = eta-time_since_start - if (eta_relative > threshold and showTime) or force_display: + if (eta_relative > threshold and show_eta) or force_display: if eta_relative > 3600: return label + time.strftime('%H:%M:%S', time.gmtime(eta_relative)) elif eta_relative > 60: @@ -194,9 +192,9 @@ def check_progress_call(id_part): progress += 1 / shared.state.job_count * shared.state.sampling_step / shared.state.sampling_steps # Show progress percentage and time left at the same moment, and base it also on steps done - showPBText = progress >= 0.01 or shared.state.sampling_step >= 10 + show_eta = progress >= 0.01 or shared.state.sampling_step >= 10 - time_left = calc_time_left( progress, 1, " ETA: ", shared.state.time_left_force_display, showPBText ) + time_left = calc_time_left(progress, 1, " ETA: ", shared.state.time_left_force_display, show_eta) if time_left != "": shared.state.time_left_force_display = True @@ -204,7 +202,7 @@ def check_progress_call(id_part): progressbar = "" if opts.show_progressbar: - progressbar = f"""
{" " * 2 + str(int(progress*100))+"%" + time_left if showPBText else ""}
""" + progressbar = f"""
{" " * 2 + str(int(progress*100))+"%" + time_left if show_eta else ""}
""" image = gr_show(False) preview_visibility = gr_show(False) -- cgit v1.2.3 From 24d4a0841d3cc0e5908b098f65a9caa3fa889af8 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Wed, 4 Jan 2023 20:10:40 +0300 Subject: train tab visual updates allow setting train tab values from ui-config.json --- modules/ui.py | 35 +++++++++++++++++++++-------------- style.css | 2 +- 2 files changed, 22 insertions(+), 15 deletions(-) (limited to 'modules/ui.py') diff --git a/modules/ui.py b/modules/ui.py index 72e7b7d2..44f4f3a4 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -1281,42 +1281,48 @@ def create_ui(): with gr.Tab(label="Train"): gr.HTML(value="

Train an embedding or Hypernetwork; you must specify a directory with a set of 1:1 ratio images [wiki]

") - with gr.Row(): + with FormRow(): train_embedding_name = gr.Dropdown(label='Embedding', elem_id="train_embedding", choices=sorted(sd_hijack.model_hijack.embedding_db.word_embeddings.keys())) create_refresh_button(train_embedding_name, sd_hijack.model_hijack.embedding_db.load_textual_inversion_embeddings, lambda: {"choices": sorted(sd_hijack.model_hijack.embedding_db.word_embeddings.keys())}, "refresh_train_embedding_name") - with gr.Row(): + train_hypernetwork_name = gr.Dropdown(label='Hypernetwork', elem_id="train_hypernetwork", choices=[x for x in shared.hypernetworks.keys()]) create_refresh_button(train_hypernetwork_name, shared.reload_hypernetworks, lambda: {"choices": sorted([x for x in shared.hypernetworks.keys()])}, "refresh_train_hypernetwork_name") - with gr.Row(): + + with FormRow(): embedding_learn_rate = gr.Textbox(label='Embedding Learning rate', placeholder="Embedding Learning rate", value="0.005", elem_id="train_embedding_learn_rate") hypernetwork_learn_rate = gr.Textbox(label='Hypernetwork Learning rate', placeholder="Hypernetwork Learning rate", value="0.00001", elem_id="train_hypernetwork_learn_rate") - with gr.Row(): + with FormRow(): clip_grad_mode = gr.Dropdown(value="disabled", label="Gradient Clipping", choices=["disabled", "value", "norm"]) clip_grad_value = gr.Textbox(placeholder="Gradient clip value", value="0.1", show_label=False) - batch_size = gr.Number(label='Batch size', value=1, precision=0, elem_id="train_batch_size") - gradient_step = gr.Number(label='Gradient accumulation steps', value=1, precision=0, elem_id="train_gradient_step") + with FormRow(): + batch_size = gr.Number(label='Batch size', value=1, precision=0, elem_id="train_batch_size") + gradient_step = gr.Number(label='Gradient accumulation steps', value=1, precision=0, elem_id="train_gradient_step") + dataset_directory = gr.Textbox(label='Dataset directory', placeholder="Path to directory with input images", elem_id="train_dataset_directory") log_directory = gr.Textbox(label='Log directory', placeholder="Path to directory where to write outputs", value="textual_inversion", elem_id="train_log_directory") template_file = gr.Textbox(label='Prompt template file', value=os.path.join(script_path, "textual_inversion_templates", "style_filewords.txt"), elem_id="train_template_file") training_width = gr.Slider(minimum=64, maximum=2048, step=8, label="Width", value=512, elem_id="train_training_width") training_height = gr.Slider(minimum=64, maximum=2048, step=8, label="Height", value=512, elem_id="train_training_height") steps = gr.Number(label='Max steps', value=100000, precision=0, elem_id="train_steps") - create_image_every = gr.Number(label='Save an image to log directory every N steps, 0 to disable', value=500, precision=0, elem_id="train_create_image_every") - save_embedding_every = gr.Number(label='Save a copy of embedding to log directory every N steps, 0 to disable', value=500, precision=0, elem_id="train_save_embedding_every") + + with FormRow(): + create_image_every = gr.Number(label='Save an image to log directory every N steps, 0 to disable', value=500, precision=0, elem_id="train_create_image_every") + save_embedding_every = gr.Number(label='Save a copy of embedding to log directory every N steps, 0 to disable', value=500, precision=0, elem_id="train_save_embedding_every") + save_image_with_stored_embedding = gr.Checkbox(label='Save images with embedding in PNG chunks', value=True, elem_id="train_save_image_with_stored_embedding") preview_from_txt2img = gr.Checkbox(label='Read parameters (prompt, etc...) from txt2img tab when making previews', value=False, elem_id="train_preview_from_txt2img") - with gr.Row(): - shuffle_tags = gr.Checkbox(label="Shuffle tags by ',' when creating prompts.", value=False, elem_id="train_shuffle_tags") - tag_drop_out = gr.Slider(minimum=0, maximum=1, step=0.1, label="Drop out tags when creating prompts.", value=0, elem_id="train_tag_drop_out") - with gr.Row(): - latent_sampling_method = gr.Radio(label='Choose latent sampling method', value="once", choices=['once', 'deterministic', 'random'], elem_id="train_latent_sampling_method") + + shuffle_tags = gr.Checkbox(label="Shuffle tags by ',' when creating prompts.", value=False, elem_id="train_shuffle_tags") + tag_drop_out = gr.Slider(minimum=0, maximum=1, step=0.1, label="Drop out tags when creating prompts.", value=0, elem_id="train_tag_drop_out") + + latent_sampling_method = gr.Radio(label='Choose latent sampling method', value="once", choices=['once', 'deterministic', 'random'], elem_id="train_latent_sampling_method") with gr.Row(): + train_embedding = gr.Button(value="Train Embedding", variant='primary', elem_id="train_train_embedding") interrupt_training = gr.Button(value="Interrupt", elem_id="train_interrupt_training") train_hypernetwork = gr.Button(value="Train Hypernetwork", variant='primary', elem_id="train_train_hypernetwork") - train_embedding = gr.Button(value="Train Embedding", variant='primary', elem_id="train_train_embedding") params = script_callbacks.UiTrainTabParams(txt2img_preview_params) @@ -1803,6 +1809,7 @@ def create_ui(): visit(img2img_interface, loadsave, "img2img") visit(extras_interface, loadsave, "extras") visit(modelmerger_interface, loadsave, "modelmerger") + visit(train_interface, loadsave, "train") if not error_loading and (not os.path.exists(ui_config_file) or settings_count != len(ui_settings)): with open(ui_config_file, "w", encoding="utf8") as file: diff --git a/style.css b/style.css index 2116ec3c..09ee540b 100644 --- a/style.css +++ b/style.css @@ -611,7 +611,7 @@ img2maskimg, #img2maskimg > .h-60, #img2maskimg > .h-60 > div, #img2maskimg > .h padding-top: 0.9em; } -#img2img_settings div.gr-form .gr-form, #txt2img_settings div.gr-form .gr-form{ +#img2img_settings div.gr-form .gr-form, #txt2img_settings div.gr-form .gr-form, #train_tabs div.gr-form .gr-form{ border: none; padding-bottom: 0.5em; } -- cgit v1.2.3 From 81490780949fffed77493b4bd741e96ec737fe27 Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Wed, 4 Jan 2023 22:04:40 +0300 Subject: added the option to specify target resolution with possibility of truncating for hires fix; also sampling steps --- javascript/hints.js | 11 ++++--- modules/generation_parameters_copypaste.py | 9 ++++-- modules/processing.py | 51 +++++++++++++++++++++++++++--- modules/txt2img.py | 5 ++- modules/ui.py | 24 ++++++++++---- 5 files changed, 81 insertions(+), 19 deletions(-) (limited to 'modules/ui.py') diff --git a/javascript/hints.js b/javascript/hints.js index 63e17e05..dda66e09 100644 --- a/javascript/hints.js +++ b/javascript/hints.js @@ -81,9 +81,6 @@ titles = { "vram": "Torch active: Peak amount of VRAM used by Torch during generation, excluding cached data.\nTorch reserved: Peak amount of VRAM allocated by Torch, including all active and cached data.\nSys VRAM: Peak amount of VRAM allocation across all applications / total GPU VRAM (peak utilization%).", - "Highres. fix": "Use a two step process to partially create an image at smaller resolution, upscale, and then improve details in it without changing composition", - "Scale latent": "Uscale the image in latent space. Alternative is to produce the full image from latent representation, upscale that, and then move it back to latent space.", - "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.", @@ -100,7 +97,13 @@ titles = { "Clip skip": "Early stopping parameter for CLIP model; 1 is stop at last layer as usual, 2 is stop at penultimate layer, etc.", "Approx NN": "Cheap neural network approximation. Very fast compared to VAE, but produces pictures with 4 times smaller horizontal/vertical resoluton and lower quality.", - "Approx cheap": "Very cheap approximation. Very fast compared to VAE, but produces pictures with 8 times smaller horizontal/vertical resoluton and extremely low quality." + "Approx cheap": "Very cheap approximation. Very fast compared to VAE, but produces pictures with 8 times smaller horizontal/vertical resoluton and extremely low quality.", + + "Hires. fix": "Use a two step process to partially create an image at smaller resolution, upscale, and then improve details in it without changing composition", + "Hires steps": "Number of sampling steps for upscaled picture. If 0, uses same as for original.", + "Upscale by": "Adjusts the size of the image by multiplying the original width and height by the selected value. Ignored if either Resize width to or Resize height to are non-zero.", + "Resize width to": "Resizes image to this width. If 0, width is inferred from either of two nearby sliders.", + "Resize height to": "Resizes image to this height. If 0, height is inferred from either of two nearby sliders." } diff --git a/modules/generation_parameters_copypaste.py b/modules/generation_parameters_copypaste.py index 4baf4d9a..12a9de3d 100644 --- a/modules/generation_parameters_copypaste.py +++ b/modules/generation_parameters_copypaste.py @@ -212,11 +212,10 @@ def restore_old_hires_fix_params(res): firstpass_width = math.ceil(scale * width / 64) * 64 firstpass_height = math.ceil(scale * height / 64) * 64 - hr_scale = width / firstpass_width if firstpass_width > 0 else height / firstpass_height - res['Size-1'] = firstpass_width res['Size-2'] = firstpass_height - res['Hires upscale'] = hr_scale + res['Hires resize-1'] = width + res['Hires resize-2'] = height def parse_generation_parameters(x: str): @@ -276,6 +275,10 @@ Steps: 20, Sampler: Euler a, CFG scale: 7, Seed: 965400086, Size: 512x512, Model hypernet_hash = res.get("Hypernet hash", None) res["Hypernet"] = find_hypernetwork_key(hypernet_name, hypernet_hash) + if "Hires resize-1" not in res: + res["Hires resize-1"] = 0 + res["Hires resize-2"] = 0 + restore_old_hires_fix_params(res) return res diff --git a/modules/processing.py b/modules/processing.py index 47712159..9cad05f2 100644 --- a/modules/processing.py +++ b/modules/processing.py @@ -662,12 +662,17 @@ def process_images_inner(p: StableDiffusionProcessing) -> Processed: class StableDiffusionProcessingTxt2Img(StableDiffusionProcessing): sampler = None - def __init__(self, enable_hr: bool = False, denoising_strength: float = 0.75, firstphase_width: int = 0, firstphase_height: int = 0, hr_scale: float = 2.0, hr_upscaler: str = None, **kwargs): + def __init__(self, enable_hr: bool = False, denoising_strength: float = 0.75, firstphase_width: int = 0, firstphase_height: int = 0, hr_scale: float = 2.0, hr_upscaler: str = None, hr_second_pass_steps: int = 0, hr_resize_x: int = 0, hr_resize_y: int = 0, **kwargs): super().__init__(**kwargs) self.enable_hr = enable_hr self.denoising_strength = denoising_strength self.hr_scale = hr_scale self.hr_upscaler = hr_upscaler + self.hr_second_pass_steps = hr_second_pass_steps + self.hr_resize_x = hr_resize_x + self.hr_resize_y = hr_resize_y + self.hr_upscale_to_x = hr_resize_x + self.hr_upscale_to_y = hr_resize_y if firstphase_width != 0 or firstphase_height != 0: print("firstphase_width/firstphase_height no longer supported; use hr_scale", file=sys.stderr) @@ -675,6 +680,9 @@ class StableDiffusionProcessingTxt2Img(StableDiffusionProcessing): self.width = firstphase_width self.height = firstphase_height + self.truncate_x = 0 + self.truncate_y = 0 + def init(self, all_prompts, all_seeds, all_subseeds): if self.enable_hr: if state.job_count == -1: @@ -682,7 +690,38 @@ class StableDiffusionProcessingTxt2Img(StableDiffusionProcessing): else: state.job_count = state.job_count * 2 - self.extra_generation_params["Hires upscale"] = self.hr_scale + if self.hr_resize_x == 0 and self.hr_resize_y == 0: + self.extra_generation_params["Hires upscale"] = self.hr_scale + self.hr_upscale_to_x = int(self.width * self.hr_scale) + self.hr_upscale_to_y = int(self.height * self.hr_scale) + else: + self.extra_generation_params["Hires resize"] = f"{self.hr_resize_x}x{self.hr_resize_y}" + + if self.hr_resize_y == 0: + self.hr_upscale_to_x = self.hr_resize_x + self.hr_upscale_to_y = self.hr_resize_x * self.height // self.width + elif self.hr_resize_x == 0: + self.hr_upscale_to_x = self.hr_resize_y * self.width // self.height + self.hr_upscale_to_y = self.hr_resize_y + else: + target_w = self.hr_resize_x + target_h = self.hr_resize_y + src_ratio = self.width / self.height + dst_ratio = self.hr_resize_x / self.hr_resize_y + + if src_ratio < dst_ratio: + self.hr_upscale_to_x = self.hr_resize_x + self.hr_upscale_to_y = self.hr_resize_x * self.height // self.width + else: + self.hr_upscale_to_x = self.hr_resize_y * self.width // self.height + self.hr_upscale_to_y = self.hr_resize_y + + self.truncate_x = (self.hr_upscale_to_x - target_w) // opt_f + self.truncate_y = (self.hr_upscale_to_y - target_h) // opt_f + + if self.hr_second_pass_steps: + self.extra_generation_params["Hires steps"] = self.hr_second_pass_steps + if self.hr_upscaler is not None: self.extra_generation_params["Hires upscaler"] = self.hr_upscaler @@ -699,8 +738,8 @@ class StableDiffusionProcessingTxt2Img(StableDiffusionProcessing): if not self.enable_hr: return samples - target_width = int(self.width * self.hr_scale) - target_height = int(self.height * self.hr_scale) + target_width = self.hr_upscale_to_x + target_height = self.hr_upscale_to_y def save_intermediate(image, index): """saves image before applying hires fix, if enabled in options; takes as an argument either an image or batch with latent space images""" @@ -755,13 +794,15 @@ class StableDiffusionProcessingTxt2Img(StableDiffusionProcessing): self.sampler = sd_samplers.create_sampler(self.sampler_name, self.sd_model) + samples = samples[:, :, self.truncate_y//2:samples.shape[2]-(self.truncate_y+1)//2, self.truncate_x//2:samples.shape[3]-(self.truncate_x+1)//2] + noise = create_random_tensors(samples.shape[1:], seeds=seeds, subseeds=subseeds, subseed_strength=subseed_strength, p=self) # GC now before running the next img2img to prevent running out of memory x = None devices.torch_gc() - samples = self.sampler.sample_img2img(self, samples, noise, conditioning, unconditional_conditioning, steps=self.steps, image_conditioning=image_conditioning) + samples = self.sampler.sample_img2img(self, samples, noise, conditioning, unconditional_conditioning, steps=self.hr_second_pass_steps or self.steps, image_conditioning=image_conditioning) return samples diff --git a/modules/txt2img.py b/modules/txt2img.py index e189a899..38b5f591 100644 --- a/modules/txt2img.py +++ b/modules/txt2img.py @@ -8,7 +8,7 @@ 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, denoising_strength: float, hr_scale: float, hr_upscaler: str, *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, hr_scale: float, hr_upscaler: str, hr_second_pass_steps: int, hr_resize_x: int, hr_resize_y: int, *args): p = StableDiffusionProcessingTxt2Img( sd_model=shared.sd_model, outpath_samples=opts.outdir_samples or opts.outdir_txt2img_samples, @@ -35,6 +35,9 @@ def txt2img(prompt: str, negative_prompt: str, prompt_style: str, prompt_style2: denoising_strength=denoising_strength if enable_hr else None, hr_scale=hr_scale, hr_upscaler=hr_upscaler, + hr_second_pass_steps=hr_second_pass_steps, + hr_resize_x=hr_resize_x, + hr_resize_y=hr_resize_y, ) p.scripts = modules.scripts.scripts_txt2img diff --git a/modules/ui.py b/modules/ui.py index 44f4f3a4..04091e67 100644 --- a/modules/ui.py +++ b/modules/ui.py @@ -637,10 +637,10 @@ def create_sampler_and_steps_selection(choices, tabname): with FormRow(elem_id=f"sampler_selection_{tabname}"): sampler_index = gr.Dropdown(label='Sampling method', elem_id=f"{tabname}_sampling", choices=[x.name for x in choices], value=choices[0].name, type="index") sampler_index.save_to_config = True - steps = gr.Slider(minimum=1, maximum=150, step=1, elem_id=f"{tabname}_steps", label="Sampling Steps", value=20) + steps = gr.Slider(minimum=1, maximum=150, step=1, elem_id=f"{tabname}_steps", label="Sampling steps", value=20) else: with FormGroup(elem_id=f"sampler_selection_{tabname}"): - steps = gr.Slider(minimum=1, maximum=150, step=1, elem_id=f"{tabname}_steps", label="Sampling Steps", value=20) + steps = gr.Slider(minimum=1, maximum=150, step=1, elem_id=f"{tabname}_steps", label="Sampling steps", value=20) sampler_index = gr.Radio(label='Sampling method', elem_id=f"{tabname}_sampling", choices=[x.name for x in choices], value=choices[0].name, type="index") return steps, sampler_index @@ -709,10 +709,16 @@ def create_ui(): enable_hr = gr.Checkbox(label='Hires. fix', value=False, elem_id="txt2img_enable_hr") elif category == "hires_fix": - with FormRow(visible=False, elem_id="txt2img_hires_fix") as hr_options: - hr_upscaler = gr.Dropdown(label="Upscaler", elem_id="txt2img_hr_upscaler", choices=[*shared.latent_upscale_modes, *[x.name for x in shared.sd_upscalers]], value=shared.latent_upscale_default_mode) - hr_scale = gr.Slider(minimum=1.0, maximum=4.0, step=0.05, label="Upscale by", value=2.0, elem_id="txt2img_hr_scale") - denoising_strength = gr.Slider(minimum=0.0, maximum=1.0, step=0.01, label='Denoising strength', value=0.7, elem_id="txt2img_denoising_strength") + with FormGroup(visible=False, elem_id="txt2img_hires_fix") as hr_options: + with FormRow(elem_id="txt2img_hires_fix_row1"): + hr_upscaler = gr.Dropdown(label="Upscaler", elem_id="txt2img_hr_upscaler", choices=[*shared.latent_upscale_modes, *[x.name for x in shared.sd_upscalers]], value=shared.latent_upscale_default_mode) + hr_second_pass_steps = gr.Slider(minimum=0, maximum=150, step=1, label='Hires steps', value=0, elem_id="txt2img_hires_steps") + denoising_strength = gr.Slider(minimum=0.0, maximum=1.0, step=0.01, label='Denoising strength', value=0.7, elem_id="txt2img_denoising_strength") + + with FormRow(elem_id="txt2img_hires_fix_row2"): + hr_scale = gr.Slider(minimum=1.0, maximum=4.0, step=0.05, label="Upscale by", value=2.0, elem_id="txt2img_hr_scale") + hr_resize_x = gr.Slider(minimum=0, maximum=2048, step=8, label="Resize width to", value=0, elem_id="txt2img_hr_resize_x") + hr_resize_y = gr.Slider(minimum=0, maximum=2048, step=8, label="Resize height to", value=0, elem_id="txt2img_hr_resize_y") elif category == "batch": if not opts.dimensions_and_batch_together: @@ -753,6 +759,9 @@ def create_ui(): denoising_strength, hr_scale, hr_upscaler, + hr_second_pass_steps, + hr_resize_x, + hr_resize_y, ] + custom_inputs, outputs=[ @@ -804,6 +813,9 @@ def create_ui(): (hr_options, lambda d: gr.Row.update(visible="Denoising strength" in d)), (hr_scale, "Hires upscale"), (hr_upscaler, "Hires upscaler"), + (hr_second_pass_steps, "Hires steps"), + (hr_resize_x, "Hires resize-1"), + (hr_resize_y, "Hires resize-2"), *modules.scripts.scripts_txt2img.infotext_fields ] parameters_copypaste.add_paste_fields("txt2img", None, txt2img_paste_fields) -- cgit v1.2.3