diff --git a/extensions-builtin/Lora/scripts/lora_script.py b/extensions-builtin/Lora/scripts/lora_script.py
index 49086f8cc..dc18866a9 100644
--- a/extensions-builtin/Lora/scripts/lora_script.py
+++ b/extensions-builtin/Lora/scripts/lora_script.py
@@ -1,22 +1,16 @@
 import re
-import gradio as gr
 from fastapi import FastAPI
-import network
 import networks
-import lora  # noqa:F401 # pylint: disable=unused-import
-# import lora_patches
-import extra_networks_lora
-import ui_extra_networks_lora
+from network import NetworkOnDisk
+from ui_extra_networks_lora import ExtraNetworksPageLora
+from extra_networks_lora import ExtraNetworkLora
+# import lora  # noqa:F401 # pylint: disable=unused-import
 from modules import script_callbacks, ui_extra_networks, extra_networks, shared
 
 
-# def unload():
-#     networks.originals.undo()
-
-
 def before_ui():
-    ui_extra_networks.register_page(ui_extra_networks_lora.ExtraNetworksPageLora())
-    networks.extra_network_lora = extra_networks_lora.ExtraNetworkLora()
+    ui_extra_networks.register_page(ExtraNetworksPageLora())
+    networks.extra_network_lora = ExtraNetworkLora()
     extra_networks.register_extra_network(networks.extra_network_lora)
     # extra_networks.register_extra_network_alias(networks.extra_network_lora, "lyco")
 
@@ -28,15 +22,7 @@ script_callbacks.on_before_ui(before_ui)
 script_callbacks.on_infotext_pasted(networks.infotext_pasted)
 
 
-shared.options_templates.update(shared.options_section(('extra_networks', "Extra Networks"), {
-    # "sd_lora": shared.OptionInfo("None", "Add network to prompt", gr.Dropdown, lambda: {"choices": ["None", *networks.available_networks], "visible": False}, refresh=networks.list_available_networks),
-    "sd_lora": shared.OptionInfo("None", "Add network to prompt", gr.Dropdown, {"choices": ["None"], "visible": False}),
-    # "lora_show_all": shared.OptionInfo(False, "Always show all networks on the Lora page").info("otherwise, those detected as for incompatible version of Stable Diffusion will be hidden"),
-    # "lora_hide_unknown_for_versions": shared.OptionInfo([], "Hide networks of unknown versions for model versions", gr.CheckboxGroup, {"choices": ["SD1", "SD2", "SDXL"]}),
-}))
-
-
-def create_lora_json(obj: network.NetworkOnDisk):
+def create_lora_json(obj: NetworkOnDisk):
     return {
         "name": obj.name,
         "alias": obj.alias,
@@ -45,7 +31,7 @@ def create_lora_json(obj: network.NetworkOnDisk):
     }
 
 
-def api_networks(_: gr.Blocks, app: FastAPI):
+def api_networks(_, app: FastAPI):
     @app.get("/sdapi/v1/loras")
     async def get_loras():
         return [create_lora_json(obj) for obj in networks.available_networks.values()]
diff --git a/extensions-builtin/Lora/ui_extra_networks_lora.py b/extensions-builtin/Lora/ui_extra_networks_lora.py
index f029aaf05..bd408b327 100644
--- a/extensions-builtin/Lora/ui_extra_networks_lora.py
+++ b/extensions-builtin/Lora/ui_extra_networks_lora.py
@@ -91,7 +91,7 @@ class ExtraNetworksPageLora(ui_extra_networks.ExtraNetworksPage):
             return None
 
     def list_items(self):
-        with concurrent.futures.ThreadPoolExecutor(max_workers=8) as executor:
+        with concurrent.futures.ThreadPoolExecutor(max_workers=shared.max_workers) as executor:
             future_items = {executor.submit(self.create_item, net): net for net in networks.available_networks}
             for future in concurrent.futures.as_completed(future_items):
                 item = future.result()
diff --git a/installer.py b/installer.py
index d7f069af9..413463fbb 100644
--- a/installer.py
+++ b/installer.py
@@ -6,8 +6,6 @@ import shutil
 import logging
 import platform
 import subprocess
-import io
-import pstats
 import cProfile
 import pkg_resources
 
@@ -29,6 +27,7 @@ opts = {}
 args = Dot({
     'debug': False,
     'reset': False,
+    'profile': False,
     'upgrade': False,
     'skip_extensions': False,
     'skip_requirements': False,
@@ -143,19 +142,9 @@ def print_dict(d):
     return ' '.join([f'{k}={v}' for k, v in d.items()])
 
 
-def print_profile(profile: cProfile.Profile, msg: str):
-    try:
-        from rich import print # pylint: disable=redefined-builtin
-    except Exception:
-        pass
-    profile.disable()
-    stream = io.StringIO()
-    ps = pstats.Stats(profile, stream=stream)
-    ps.sort_stats(pstats.SortKey.CUMULATIVE).print_stats(15)
-    profile = None
-    lines = stream.getvalue().split('\n')
-    lines = [line for line in lines if '<frozen' not in line and '{built-in' not in line and '/logging' not in line and '/rich' not in line]
-    print(f'Profile {msg}:', '\n'.join(lines))
+def print_profile(profiler: cProfile.Profile, msg: str):
+    from modules.errors import profile
+    profile(profiler, msg)
 
 
 # check if package is installed
@@ -768,6 +757,7 @@ def set_environment():
     os.environ.setdefault('TF_ENABLE_ONEDNN_OPTS', '0')
     os.environ.setdefault('USE_TORCH', '1')
     os.environ.setdefault('UVICORN_TIMEOUT_KEEP_ALIVE', '60')
+    os.environ.setdefault('KINETO_LOG_LEVEL', '3')
     os.environ.setdefault('HF_HUB_CACHE', opts.get('hfcache_dir', os.path.join(os.path.expanduser('~'), '.cache', 'huggingface', 'hub')))
     log.debug(f'Cache folder: {os.environ.get("HF_HUB_CACHE")}')
     if sys.platform == 'darwin':
diff --git a/modules/call_queue.py b/modules/call_queue.py
index 15d5274b8..21f94352d 100644
--- a/modules/call_queue.py
+++ b/modules/call_queue.py
@@ -2,9 +2,6 @@ import html
 import threading
 import time
 import cProfile
-import pstats
-import io
-from rich import print # pylint: disable=redefined-builtin
 from modules import shared, progress, errors
 
 queue_lock = threading.Lock()
@@ -62,10 +59,7 @@ def wrap_gradio_call(func, extra_outputs=None, add_stats=False, name=None):
             else:
                 res = list(res)
             if shared.cmd_opts.profile:
-                pr.disable()
-                s = io.StringIO()
-                pstats.Stats(pr, stream=s).sort_stats(pstats.SortKey.CUMULATIVE).print_stats(15)
-                print('Profile Exec:', s.getvalue())
+                errors.profile(pr, 'Wrap')
         except Exception as e:
             errors.display(e, 'gradio call')
             if extra_outputs_array is None:
diff --git a/modules/devices.py b/modules/devices.py
index 78a20c4e6..0211242b5 100644
--- a/modules/devices.py
+++ b/modules/devices.py
@@ -150,6 +150,25 @@ def torch_gc(force=False):
     log.debug(f'gc: collected={collected} device={torch.device(get_optimal_device_name())} {memstats.memory_stats()}')
 
 
+def set_cuda_sync_mode(mode):
+    """
+    Set the CUDA device synchronization mode: auto, spin, yield or block.
+    auto: Chooses spin or yield depending on the number of available CPU cores.
+    spin: Runs one CPU core per GPU at 100% to poll for completed operations.
+    yield: Gives control to other threads between polling, if any are waiting.
+    block: Lets the thread sleep until the GPU driver signals completion.
+    """
+    if mode == -1 or mode == 'none' or not cuda_ok:
+        return
+    try:
+        import ctypes
+        log.info(f'Set cuda synch: mode={mode}')
+        torch.cuda.set_device(torch.device(get_optimal_device_name()))
+        ctypes.CDLL('libcudart.so').cudaSetDeviceFlags({'auto': 0, 'spin': 1, 'yield': 2, 'block': 4}[mode])
+    except Exception:
+        pass
+
+
 def test_fp16():
     if shared.cmd_opts.experimental:
         return True
@@ -272,6 +291,9 @@ dtype = torch.float16
 dtype_vae = torch.float16
 dtype_unet = torch.float16
 unet_needs_upcast = False
+if args.profile:
+    log.info(f'Torch build config: {torch.__config__.show()}')
+# set_cuda_sync_mode('block') # none/auto/spin/yield/block
 
 
 def cond_cast_unet(tensor):
diff --git a/modules/errors.py b/modules/errors.py
index ecbc30a36..f14cfe291 100644
--- a/modules/errors.py
+++ b/modules/errors.py
@@ -55,3 +55,34 @@ def run(code, task):
 
 def exception(suppress=[]): # noqa: B006
     console.print_exception(show_locals=False, max_frames=10, extra_lines=2, suppress=suppress, theme="ansi_dark", word_wrap=False, width=min([console.width, 200]))
+
+
+def profile(profiler, msg: str):
+    profiler.disable()
+    import io
+    import pstats
+    stream = io.StringIO() # pylint: disable=abstract-class-instantiated
+    p = pstats.Stats(profiler, stream=stream)
+    p.sort_stats(pstats.SortKey.CUMULATIVE)
+    p.print_stats(100)
+    # p.print_title()
+    # p.print_call_heading(10, 'time')
+    # p.print_callees(10)
+    # p.print_callers(10)
+    profiler = None
+    lines = stream.getvalue().split('\n')
+    lines = [l for l in lines if '<frozen' not in l and '{built-in' not in l and '/logging' not in l and 'Ordered by' not in l and 'List reduced' not in l and '_lsprof' not in l and '/profiler' not in l and 'rich' not in l and l.strip() != '']
+    txt = '\n'.join(lines[:min(5, len(lines))])
+    log.debug(f'Profile {msg}: {txt}')
+
+
+def profile_torch(profiler, msg: str):
+    profiler.stop()
+    from rich import print # pylint: disable=redefined-builtin
+    # lines = profiler.key_averages().table(sort_by="self_cuda_time_total", row_limit=6)
+    lines = profiler.key_averages().table(sort_by="self_cpu_time_total", row_limit=12)
+    lines = lines.split('\n')
+    lines = [l for l in lines if '/profiler' not in l and '---' not in l]
+    txt = '\n'.join(lines)
+    # print(f'Torch {msg}:', txt)
+    log.debug(f'Torch profile {msg}: \n{txt}')
diff --git a/modules/loader.py b/modules/loader.py
index 3dbb0c4a8..fa9b34153 100644
--- a/modules/loader.py
+++ b/modules/loader.py
@@ -42,3 +42,16 @@ errors.install([gradio])
 import diffusers # pylint: disable=W0611,C0411
 timer.startup.record("diffusers")
 errors.log.info(f'Load packages: torch={getattr(torch, "__long_version__", torch.__version__)} diffusers={diffusers.__version__} gradio={gradio.__version__}')
+
+try:
+    import os
+    import math
+    cores = os.cpu_count()
+    affinity = len(os.sched_getaffinity(0))
+    threads = torch.get_num_threads()
+    if threads < (affinity / 2):
+        torch.set_num_threads(math.floor(affinity / 2))
+        threads = torch.get_num_threads()
+        errors.log.debug(f'Detected: cores={cores} affinity={affinity} set threads={threads}')
+except Exception as e:
+    pass
diff --git a/modules/lora b/modules/lora
index 95ae56bd2..0908c5414 160000
--- a/modules/lora
+++ b/modules/lora
@@ -1 +1 @@
-Subproject commit 95ae56bd22c285ccb2fe5fca96d92f39842bb99b
+Subproject commit 0908c5414dadfe2185669e2d657d542161647b79
diff --git a/modules/modelloader.py b/modules/modelloader.py
index ddaa6efef..0e8254024 100644
--- a/modules/modelloader.py
+++ b/modules/modelloader.py
@@ -574,6 +574,7 @@ def move_files(src_path: str, dest_path: str, ext_filter: str = None):
 
 def load_upscalers():
     # We can only do this 'magic' method to dynamically load upscalers if they are referenced, so we'll try to import any _model.py files before looking in __subclasses__
+    t0 = time.time()
     modules_dir = os.path.join(shared.script_path, "modules", "postprocess")
     for file in os.listdir(modules_dir):
         if "_model.py" in file:
@@ -602,4 +603,5 @@ def load_upscalers():
         datas += scaler.scalers
         names.append(name[8:])
     shared.sd_upscalers = sorted(datas, key=lambda x: x.name.lower() if not isinstance(x.scaler, (UpscalerNone, UpscalerLanczos, UpscalerNearest)) else "") # Special case for UpscalerNone keeps it at the beginning of the list.
-    shared.log.debug(f"Load upscalers: total={len(shared.sd_upscalers)} downloaded={len([x for x in shared.sd_upscalers if x.data_path is not None and os.path.isfile(x.data_path)])} user={len([x for x in shared.sd_upscalers if x.custom])} {names}")
+    t1 = time.time()
+    shared.log.debug(f"Load upscalers: total={len(shared.sd_upscalers)} downloaded={len([x for x in shared.sd_upscalers if x.data_path is not None and os.path.isfile(x.data_path)])} user={len([x for x in shared.sd_upscalers if x.custom])} time={t1-t0:.2f} {names}")
diff --git a/modules/processing.py b/modules/processing.py
index 01f1dd3af..b04388d8f 100644
--- a/modules/processing.py
+++ b/modules/processing.py
@@ -18,7 +18,7 @@ from ldm.models.diffusion.ddpm import LatentDepth2ImageDiffusion
 from einops import repeat, rearrange
 from blendmodes.blend import blendLayers, BlendType
 from installer import git_commit
-from modules import shared, devices
+from modules import shared, devices, errors
 import modules.memstats
 import modules.lowvram
 import modules.masking
@@ -666,36 +666,6 @@ def create_infotext(p: StableDiffusionProcessing, all_prompts=None, all_seeds=No
     return infotext
 
 
-"""
-def print_profile(profile, msg: str):
-    try:
-        from rich import print # pylint: disable=redefined-builtin
-    except Exception:
-        pass
-    lines = profile.key_averages().table(sort_by="cuda_time_total", row_limit=20)
-    lines = lines.split('\n')
-    lines = [l for l in lines if '/profiler' not in l]
-    print(f'Profile {msg}:', '\n'.join(lines))
-"""
-
-
-def print_profile(profile, msg: str):
-    import io
-    import pstats
-    try:
-        from rich import print # pylint: disable=redefined-builtin
-    except Exception:
-        pass
-    profile.disable()
-    stream = io.StringIO() # pylint: disable=abstract-class-instantiated
-    ps = pstats.Stats(profile, stream=stream)
-    ps.sort_stats(pstats.SortKey.CUMULATIVE).print_stats(15)
-    profile = None
-    lines = stream.getvalue().split('\n')
-    lines = [line for line in lines if '<frozen' not in line and '{built-in' not in line and '/logging' not in line and '/rich' not in line]
-    print(f'Profile {msg}:', '\n'.join(lines))
-
-
 def process_images(p: StableDiffusionProcessing) -> Processed:
     if not hasattr(p.sd_model, 'sd_checkpoint_info'):
         return None
@@ -743,22 +713,26 @@ def process_images(p: StableDiffusionProcessing) -> Processed:
         modules.script_callbacks.before_process_callback(p)
 
         if shared.cmd_opts.profile:
-            """
-            import torch.profiler # pylint: disable=redefined-outer-name
-            with torch.profiler.profile(profile_memory=True, with_modules=True) as prof:
-                with torch.profiler.record_function("process_images"):
-                    res = process_images_inner(p)
-            print_profile(prof, 'process_images')
-            """
             import cProfile
-            pr = cProfile.Profile()
-            pr.enable()
+            profile_python = cProfile.Profile()
+            profile_python.enable()
             with context_hypertile_vae(p), context_hypertile_unet(p):
+                import torch.profiler # pylint: disable=redefined-outer-name
+                activities=[torch.profiler.ProfilerActivity.CPU]
+                if torch.cuda.is_available():
+                    activities.append(torch.profiler.ProfilerActivity.CUDA)
+                shared.log.debug(f'Torch profile: activities={activities}')
+                if shared.profiler is None:
+                    shared.profiler = torch.profiler.profile(activities=activities, profile_memory=True, with_modules=True)
+                shared.profiler.start()
+                shared.profiler.step()
                 res = process_images_inner(p)
-            print_profile(pr, 'Torch')
+                errors.profile_torch(shared.profiler, 'Process')
+            errors.profile(profile_python, 'Process')
         else:
             with context_hypertile_vae(p), context_hypertile_unet(p):
                 res = process_images_inner(p)
+
     finally:
         if not shared.opts.cuda_compile:
             modules.sd_models.apply_token_merging(p.sd_model, 0)
diff --git a/modules/processing_diffusers.py b/modules/processing_diffusers.py
index 2049fa881..e90cd0126 100644
--- a/modules/processing_diffusers.py
+++ b/modules/processing_diffusers.py
@@ -85,6 +85,8 @@ def process_diffusers(p: StableDiffusionProcessing, seeds, prompts, negative_pro
             return kwargs
         kwargs = correction_callback(p, timestep, kwargs)
         shared.state.current_latent = kwargs['latents']
+        if shared.cmd_opts.profile and shared.profiler is not None:
+            shared.profiler.step()
         return kwargs
 
     def full_vae_decode(latents, model):
@@ -139,6 +141,7 @@ def process_diffusers(p: StableDiffusionProcessing, seeds, prompts, negative_pro
         return encoded
 
     def vae_decode(latents, model, output_type='np', full_quality=True):
+        t0 = time.time()
         prev_job = shared.state.job
         shared.state.job = 'vae'
         if not torch.is_tensor(latents): # already decoded
@@ -163,6 +166,9 @@ def process_diffusers(p: StableDiffusionProcessing, seeds, prompts, negative_pro
         # decoded = validate_sample(decoded)
         imgs = model.image_processor.postprocess(decoded, output_type=output_type)
         shared.state.job = prev_job
+        if shared.cmd_opts.profile:
+            t1 = time.time()
+            shared.log.debug(f'Profile: VAE decode: {t1-t0:.2f}')
         return imgs
 
     def vae_encode(image, model, full_quality=True): # pylint: disable=unused-variable
@@ -269,6 +275,7 @@ def process_diffusers(p: StableDiffusionProcessing, seeds, prompts, negative_pro
         return task_args
 
     def set_pipeline_args(model, prompts: list, negative_prompts: list, prompts_2: typing.Optional[list]=None, negative_prompts_2: typing.Optional[list]=None, desc:str='', **kwargs):
+        t0 = time.time()
         if hasattr(model, "set_progress_bar_config"):
             model.set_progress_bar_config(bar_format='Progress {rate_fmt}{postfix} {bar} {percentage:3.0f}% {n_fmt}/{total_fmt} {elapsed} {remaining} ' + '\x1b[38;5;71m' + desc, ncols=80, colour='#327fba')
         args = {}
@@ -379,6 +386,9 @@ def process_diffusers(p: StableDiffusionProcessing, seeds, prompts, negative_pro
             shared.log.debug(txt)
         # components = [{ k: getattr(v, 'device', None) } for k, v in model.components.items()]
         # shared.log.debug(f'Diffuser pipeline components: {components}')
+        if shared.cmd_opts.profile:
+            t1 = time.time()
+            shared.log.debug(f'Profile: pipeline args: {t1-t0:.2f}')
         return args
 
     def recompile_model(hires=False):
@@ -502,7 +512,11 @@ def process_diffusers(p: StableDiffusionProcessing, seeds, prompts, negative_pro
     p.extra_generation_params['Pipeline'] = shared.sd_model.__class__.__name__
     p.extra_generation_params["Sampler Eta"] = shared.opts.scheduler_eta if shared.opts.scheduler_eta is not None and shared.opts.scheduler_eta > 0 and shared.opts.scheduler_eta < 1 else None
     try:
+        t0 = time.time()
         output = shared.sd_model(**base_args) # pylint: disable=not-callable
+        if shared.cmd_opts.profile:
+            t1 = time.time()
+            shared.log.debug(f'Profile: pipeline call: {t1-t0:.2f}')
         if not hasattr(output, 'images') and hasattr(output, 'frames'):
             shared.log.debug(f'Generated: frames={len(output.frames[0])}')
             output.images = output.frames[0]
diff --git a/modules/script_callbacks.py b/modules/script_callbacks.py
index e30f88dd1..663199968 100644
--- a/modules/script_callbacks.py
+++ b/modules/script_callbacks.py
@@ -329,7 +329,7 @@ def before_ui_callback():
 
 
 def add_callback(callbacks, fun):
-    stack = [x for x in inspect.stack() if x.filename != __file__]
+    stack = [x for x in inspect.stack(0) if x.filename != __file__]
     filename = stack[0].filename if len(stack) > 0 else 'unknown file'
     callbacks.append(ScriptCallback(filename, fun))
 
diff --git a/modules/script_loading.py b/modules/script_loading.py
index 5ac82f538..b20d4072e 100644
--- a/modules/script_loading.py
+++ b/modules/script_loading.py
@@ -3,7 +3,7 @@ import os
 import contextlib
 import importlib.util
 import modules.errors as errors
-from installer import setup_logging
+from installer import setup_logging, args
 
 
 preloaded = []
@@ -12,6 +12,10 @@ preloaded = []
 def load_module(path):
     module_spec = importlib.util.spec_from_file_location(os.path.basename(path), path)
     module = importlib.util.module_from_spec(module_spec)
+    if args.profile:
+        import cProfile
+        pr = cProfile.Profile()
+        pr.enable()
     try:
         if '/sd-extension-' in path: # safe extensions without stdout intercept
             module_spec.loader.exec_module(module)
@@ -25,6 +29,8 @@ def load_module(path):
                     errors.log.info(f"Extension: script='{os.path.relpath(path)}' {line.strip()}")
     except Exception as e:
         errors.display(e, f'Module load: {path}')
+    if args.profile:
+        errors.profile(pr, f'Scripts: {path}')
     return module
 
 
diff --git a/modules/scripts.py b/modules/scripts.py
index 01c501f2c..aee3545b2 100644
--- a/modules/scripts.py
+++ b/modules/scripts.py
@@ -261,6 +261,7 @@ def load_scripts():
             elif issubclass(script_class, scripts_postprocessing.ScriptPostprocessing):
                 postprocessing_scripts_data.append(ScriptClassData(script_class, scriptfile.path, scriptfile.basedir, module))
 
+    from installer import args
     for scriptfile in scripts_list:
         try:
             if scriptfile.basedir != paths.script_path:
@@ -274,12 +275,10 @@ def load_scripts():
             current_basedir = paths.script_path
             t.record(os.path.basename(scriptfile.basedir))
             sys.path = syspath
-
     global scripts_txt2img, scripts_img2img, scripts_postproc # pylint: disable=global-statement
     scripts_txt2img = ScriptRunner()
     scripts_img2img = ScriptRunner()
     scripts_postproc = scripts_postprocessing.ScriptPostprocessingRunner()
-
     return t, time.time()-t0
 
 
diff --git a/modules/sd_models.py b/modules/sd_models.py
index ee94ec574..0956353ce 100644
--- a/modules/sd_models.py
+++ b/modules/sd_models.py
@@ -760,6 +760,10 @@ def set_diffuser_options(sd_model, vae, op: str):
 
 def load_diffuser(checkpoint_info=None, already_loaded_state_dict=None, timer=None, op='model'): # pylint: disable=unused-argument
     import torch # pylint: disable=reimported,redefined-outer-name
+    if shared.cmd_opts.profile:
+        import cProfile
+        pr = cProfile.Profile()
+        pr.enable()
     if timer is None:
         timer = Timer()
     logging.getLogger("diffusers").setLevel(logging.ERROR)
@@ -976,6 +980,8 @@ def load_diffuser(checkpoint_info=None, already_loaded_state_dict=None, timer=No
 
     timer.record("load")
     devices.torch_gc(force=True)
+    if shared.cmd_opts.profile:
+        errors.profile(pr, 'Load')
     script_callbacks.model_loaded_callback(sd_model)
     shared.log.info(f"Load {op}: time={timer.summary()} native={get_native(sd_model)} {memory_stats()}")
 
diff --git a/modules/shared.py b/modules/shared.py
index 0ba936223..7c2bea684 100644
--- a/modules/shared.py
+++ b/modules/shared.py
@@ -781,6 +781,7 @@ class Options:
             value = expected_type(value)
         return value
 
+profiler = None
 opts = Options()
 config_filename = cmd_opts.config
 opts.load(config_filename)
@@ -804,6 +805,7 @@ device = devices.device
 batch_cond_uncond = opts.always_batch_cond_uncond or not (cmd_opts.lowvram or cmd_opts.medvram)
 parallel_processing_allowed = not cmd_opts.lowvram
 mem_mon = modules.memmon.MemUsageMonitor("MemMon", devices.device)
+max_workers = 2
 if devices.backend == "directml":
     directml_do_hijack()
 
diff --git a/modules/styles.py b/modules/styles.py
index be5259f2c..1f9a6fd52 100644
--- a/modules/styles.py
+++ b/modules/styles.py
@@ -4,6 +4,7 @@ import re
 import os
 import csv
 import json
+import time
 from installer import log
 
 
@@ -87,6 +88,7 @@ class StyleDatabase:
 
     def load_style(self, fn, prefix=None):
         with open(fn, 'r', encoding='utf-8') as f:
+            new_style = None
             try:
                 all_styles = json.load(f)
                 if type(all_styles) is dict:
@@ -100,7 +102,7 @@ class StyleDatabase:
                         name = os.path.join(prefix, name)
                     else:
                         name = os.path.join(os.path.dirname(os.path.relpath(fn, self.path)), name)
-                    self.styles[style["name"]] = Style(
+                    new_style = Style(
                         name=name,
                         desc=style.get('description', name),
                         prompt=style.get("prompt", ""),
@@ -110,26 +112,37 @@ class StyleDatabase:
                         filename=fn,
                         mtime=os.path.getmtime(fn),
                     )
+                    self.styles[style["name"]] = new_style
             except Exception as e:
                 log.error(f'Failed to load style: file={fn} error={e}')
+            return new_style
 
 
     def reload(self):
+        t0 = time.time()
         self.styles.clear()
+
         def list_folder(folder):
-            for filename in os.listdir(folder):
-                fn = os.path.abspath(os.path.join(folder, filename))
-                if os.path.isfile(fn) and fn.lower().endswith(".json"):
-                    self.load_style(fn)
-                elif os.path.isdir(fn) and not fn.startswith('.'):
-                    list_folder(fn)
+            import concurrent
+            future_items = {}
+            with concurrent.futures.ThreadPoolExecutor(max_workers=8) as executor:
+                for filename in os.listdir(folder):
+                    fn = os.path.abspath(os.path.join(folder, filename))
+                    if os.path.isfile(fn) and fn.lower().endswith(".json"):
+                        future_items[executor.submit(self.load_style, fn, None)] = fn
+                        # self.load_style(fn)
+                    elif os.path.isdir(fn) and not fn.startswith('.'):
+                        list_folder(fn)
+                self.styles = dict(sorted(self.styles.items(), key=lambda style: style[1].filename))
+                if self.built_in:
+                    fn = os.path.join('html', 'art-styles.json')
+                    future_items[executor.submit(self.load_style, fn, 'built-in')] = fn
+                for future in concurrent.futures.as_completed(future_items):
+                    future.result()
 
         list_folder(self.path)
-        self.styles = dict(sorted(self.styles.items(), key=lambda style: style[1].filename))
-        if self.built_in:
-            self.load_style(os.path.join('html', 'art-styles.json'), 'built-in')
-
-        log.debug(f'Load styles: folder="{self.path}" items={len(self.styles.keys())}')
+        t1 = time.time()
+        log.debug(f'Load styles: folder="{self.path}" items={len(self.styles.keys())} time={t1-t0:.2f}')
 
     def find_style(self, name):
         found = [style for style in self.styles.values() if style.name == name]
diff --git a/modules/ui_extra_networks.py b/modules/ui_extra_networks.py
index 9727b8752..f30dba2cc 100644
--- a/modules/ui_extra_networks.py
+++ b/modules/ui_extra_networks.py
@@ -15,7 +15,7 @@ from collections import OrderedDict
 import gradio as gr
 from PIL import Image
 from starlette.responses import FileResponse, JSONResponse
-from modules import paths, shared, scripts, modelloader
+from modules import paths, shared, scripts, modelloader, errors
 from modules.ui_components import ToolButton
 import modules.ui_symbols as symbols
 
@@ -270,7 +270,7 @@ class ExtraNetworksPage:
             self.html = f"<div id='{tabname}_{self_name_id}_subdirs' class='extra-network-subdirs'>{subdirs_html}</div><div id='{tabname}_{self_name_id}_cards' class='extra-network-cards'>{self.html}</div>"
         else:
             return ''
-        shared.log.debug(f"Extra networks: page='{self.name}' items={len(self.items)} subfolders={len(subdirs)} tab={tabname} folders={self.allowed_directories_for_previews()} list={self.list_time:.2f} desc={self.desc_time:.2f} info={self.info_time:.2f}")
+        shared.log.debug(f"Extra networks: page='{self.name}' items={len(self.items)} subfolders={len(subdirs)} tab={tabname} folders={self.allowed_directories_for_previews()} list={self.list_time:.2f} desc={self.desc_time:.2f} info={self.info_time:.2f} workers={shared.max_workers}")
         if len(self.missing_thumbs) > 0:
             threading.Thread(target=self.create_thumb).start()
         return self.html
@@ -463,6 +463,10 @@ def create_ui(container, button_parent, tabname, skip_indexing = False):
     ui.tabs = gr.Tabs(elem_id=tabname+"_extra_tabs")
     ui.button_details = gr.Button('Details', elem_id=tabname+"_extra_details_btn", visible=False)
     state = {}
+    if shared.cmd_opts.profile:
+        import cProfile
+        pr = cProfile.Profile()
+        pr.enable()
 
     def get_item(state, params = None):
         if params is not None and type(params) == dict:
@@ -567,6 +571,10 @@ def create_ui(container, button_parent, tabname, skip_indexing = False):
                 page_html = gr.HTML(page.html, elem_id=f'{tabname}{page.name}_extra_page', elem_classes="extra-networks-page")
                 ui.pages.append(page_html)
                 tab.select(ui_tab_change, _js="getENActivePage", inputs=[ui.button_details], outputs=[ui.button_scan, ui.button_save, ui.button_model])
+    if shared.cmd_opts.profile:
+        errors.profile(pr, 'ExtraNetworks')
+        pr.disable()
+
         # ui.tabs.change(fn=ui_tab_change, inputs=[], outputs=[ui.button_scan, ui.button_save])
 
     def fn_save_img(image):
diff --git a/modules/ui_extra_networks_checkpoints.py b/modules/ui_extra_networks_checkpoints.py
index 01902b376..692acdc18 100644
--- a/modules/ui_extra_networks_checkpoints.py
+++ b/modules/ui_extra_networks_checkpoints.py
@@ -64,7 +64,7 @@ class ExtraNetworksPageCheckpoints(ui_extra_networks.ExtraNetworksPage):
         return record
 
     def list_items(self):
-        with concurrent.futures.ThreadPoolExecutor(max_workers=8) as executor:
+        with concurrent.futures.ThreadPoolExecutor(max_workers=shared.max_workers) as executor:
             future_items = {executor.submit(self.create_item, cp): cp for cp in list(sd_models.checkpoints_list.copy())}
             for future in concurrent.futures.as_completed(future_items):
                 item = future.result()
diff --git a/modules/ui_extra_networks_styles.py b/modules/ui_extra_networks_styles.py
index 12f5f0a53..8dcccf96e 100644
--- a/modules/ui_extra_networks_styles.py
+++ b/modules/ui_extra_networks_styles.py
@@ -95,7 +95,7 @@ class ExtraNetworksPageStyles(ui_extra_networks.ExtraNetworksPage):
         return item
 
     def list_items(self):
-        with concurrent.futures.ThreadPoolExecutor(max_workers=8) as executor:
+        with concurrent.futures.ThreadPoolExecutor(max_workers=shared.max_workers) as executor:
             future_items = {executor.submit(self.create_item, style): style for style in list(shared.prompt_styles.styles)}
             for future in concurrent.futures.as_completed(future_items):
                 item = future.result()
diff --git a/modules/ui_extra_networks_textual_inversion.py b/modules/ui_extra_networks_textual_inversion.py
index f67031029..d94ce39dc 100644
--- a/modules/ui_extra_networks_textual_inversion.py
+++ b/modules/ui_extra_networks_textual_inversion.py
@@ -68,7 +68,7 @@ class ExtraNetworksPageTextualInversion(ui_extra_networks.ExtraNetworksPage):
             self.embeddings = []
         self.embeddings = sorted(self.embeddings, key=lambda emb: emb.filename)
 
-        with concurrent.futures.ThreadPoolExecutor(max_workers=8) as executor:
+        with concurrent.futures.ThreadPoolExecutor(max_workers=shared.max_workers) as executor:
             future_items = {executor.submit(self.create_item, net): net for net in self.embeddings}
             for future in concurrent.futures.as_completed(future_items):
                 item = future.result()
diff --git a/webui.py b/webui.py
index 157731f10..d352bed72 100644
--- a/webui.py
+++ b/webui.py
@@ -311,6 +311,9 @@ def webui(restart=False):
     modules.sd_models.write_metadata()
     load_model()
     shared.opts.save(shared.config_filename)
+    if cmd_opts.profile:
+        for k, v in modules.script_callbacks.callback_map.items():
+            shared.log.debug(f'Registered callbacks: {k}={len(v)} {[c.script for c in v]}')
     log.info(f"Startup time: {timer.startup.summary()}")
     debug = log.info if os.environ.get('SD_SCRIPT_DEBUG', None) is not None else lambda *args, **kwargs: None
     debug('Loaded scripts:')