hypernetwork training mk1

modules/hypernetwork.py

@@ -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)

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 = "<none>"
+    last_saved_image = "<none>"
+
+    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"""
+<p>
+Loss: {losses.mean():.7f}<br/>
+Step: {hypernetwork.step}<br/>
+Last prompt: {html.escape(text)}<br/>
+Last saved embedding: {html.escape(last_saved_file)}<br/>
+Last saved image: {html.escape(last_saved_image)}<br/>
+</p>
+"""
+
+    checkpoint = sd_models.select_checkpoint()
+
+    hypernetwork.sd_checkpoint = checkpoint.hash
+    hypernetwork.sd_checkpoint_name = checkpoint.model_name
+    hypernetwork.save(filename)
+
+    return hypernetwork, filename
+
+

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()
+

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

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))

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:

modules/textual_inversion/ui.py

@@ -22,7 +22,6 @@ def preprocess(*args):
 
 
 def train_embedding(*args):
-
     try:
         sd_hijack.undo_optimizations()
 

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="<p style='margin-bottom: 0.7em'>Create a new hypernetwork</p>")
+
+                    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="<p style='margin-bottom: 0.7em'>Preprocess images</p>")
 
@@ -986,6 +999,7 @@ def create_ui(wrap_gradio_gpu_call):
                 with gr.Group():
                     gr.HTML(value="<p style='margin-bottom: 0.7em'>Train an embedding; must specify a directory with a set of 512x512 images</p>")
                     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

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

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]

textual_inversion_templates/none.txt

@@ -0,0 +1 @@
+picture

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()