Upload 4 files

app.py

@@ -1,154 +1,121 @@
-import gradio as gr
-import numpy as np
-import random
-
-# import spaces #[uncomment to use ZeroGPU]
-from diffusers import DiffusionPipeline
 import torch
+import random
+import datetime
+import json
+import os
+from PIL import Image
+import gradio as gr
+from model import UNet2DConditionModelEx
+from pipeline import StableDiffusionControlLoraV3Pipeline
+
+def setup_pipeline():
+    print("Loading models...")
+    unet = UNet2DConditionModelEx.from_pretrained(
+        "runwayml/stable-diffusion-v1-5",
+        subfolder="unet",
+        torch_dtype=torch.float16
+    )
+    unet = unet.add_extra_conditions("ow-gbi-control-lora")
 
-device = "cuda" if torch.cuda.is_available() else "cpu"
-model_repo_id = "stabilityai/sdxl-turbo"  # Replace to the model you would like to use
-
-if torch.cuda.is_available():
-    torch_dtype = torch.float16
-else:
-    torch_dtype = torch.float32
-
-pipe = DiffusionPipeline.from_pretrained(model_repo_id, torch_dtype=torch_dtype)
-pipe = pipe.to(device)
-
-MAX_SEED = np.iinfo(np.int32).max
-MAX_IMAGE_SIZE = 1024
-
-
-# @spaces.GPU #[uncomment to use ZeroGPU]
-def infer(
-    prompt,
-    negative_prompt,
-    seed,
-    randomize_seed,
-    width,
-    height,
-    guidance_scale,
-    num_inference_steps,
-    progress=gr.Progress(track_tqdm=True),
-):
-    if randomize_seed:
-        seed = random.randint(0, MAX_SEED)
-
-    generator = torch.Generator().manual_seed(seed)
-
-    image = pipe(
-        prompt=prompt,
-        negative_prompt=negative_prompt,
-        guidance_scale=guidance_scale,
-        num_inference_steps=num_inference_steps,
-        width=width,
-        height=height,
-        generator=generator,
-    ).images[0]
-
-    return image, seed
-
-
-examples = [
-    "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k",
-    "An astronaut riding a green horse",
-    "A delicious ceviche cheesecake slice",
-]
-
-css = """
-#col-container {
-    margin: 0 auto;
-    max-width: 640px;
-}
-"""
-
-with gr.Blocks(css=css) as demo:
-    with gr.Column(elem_id="col-container"):
-        gr.Markdown(" # Text-to-Image Gradio Template")
-
-        with gr.Row():
-            prompt = gr.Text(
-                label="Prompt",
-                show_label=False,
-                max_lines=1,
-                placeholder="Enter your prompt",
-                container=False,
-            )
-
-            run_button = gr.Button("Run", scale=0, variant="primary")
-
-        result = gr.Image(label="Result", show_label=False)
-
-        with gr.Accordion("Advanced Settings", open=False):
-            negative_prompt = gr.Text(
-                label="Negative prompt",
-                max_lines=1,
-                placeholder="Enter a negative prompt",
-                visible=False,
-            )
-
-            seed = gr.Slider(
-                label="Seed",
-                minimum=0,
-                maximum=MAX_SEED,
-                step=1,
-                value=0,
-            )
-
-            randomize_seed = gr.Checkbox(label="Randomize seed", value=True)
+    pipe = StableDiffusionControlLoraV3Pipeline.from_pretrained(
+        "runwayml/stable-diffusion-v1-5",
+        unet=unet,
+        torch_dtype=torch.float16
+    )
 
-            with gr.Row():
-                width = gr.Slider(
-                    label="Width",
-                    minimum=256,
-                    maximum=MAX_IMAGE_SIZE,
-                    step=32,
-                    value=1024,  # Replace with defaults that work for your model
-                )
+    lora_path = "owgbi-Dataset2-6000.safetensors"
+    pipe.load_lora_weights(lora_path)
 
-                height = gr.Slider(
-                    label="Height",
-                    minimum=256,
-                    maximum=MAX_IMAGE_SIZE,
-                    step=32,
-                    value=1024,  # Replace with defaults that work for your model
-                )
+    pipe.enable_xformers_memory_efficient_attention()
+    pipe.enable_model_cpu_offload()
 
-            with gr.Row():
-                guidance_scale = gr.Slider(
-                    label="Guidance scale",
-                    minimum=0.0,
-                    maximum=10.0,
-                    step=0.1,
-                    value=0.0,  # Replace with defaults that work for your model
-                )
+    print("Pipeline ready!")
+    return pipe
 
-                num_inference_steps = gr.Slider(
-                    label="Number of inference steps",
-                    minimum=1,
-                    maximum=50,
-                    step=1,
-                    value=2,  # Replace with defaults that work for your model
-                )
+pipe = setup_pipeline()
 
-        gr.Examples(examples=examples, inputs=[prompt])
-    gr.on(
-        triggers=[run_button.click, prompt.submit],
-        fn=infer,
-        inputs=[
+def generate_image_core(image, prompt, negative_prompt, seed, guidance_scale, steps, strength, num_images_per_prompt, guidance_rescale):
+    try:
+        pipe.unet.enable_gradient_checkpointing()
+        generator = torch.manual_seed(seed)
+        result = pipe(
             prompt,
-            negative_prompt,
-            seed,
-            randomize_seed,
-            width,
-            height,
-            guidance_scale,
-            num_inference_steps,
-        ],
-        outputs=[result, seed],
-    )
-
+            negative_prompt=negative_prompt,
+            num_inference_steps=steps,
+            generator=generator,
+            image=image,
+            guidance_scale=guidance_scale,
+            strength=strength,
+            num_images_per_prompt=num_images_per_prompt,
+            guidance_rescale=guidance_rescale,
+        )
+        return result.images
+    except RuntimeError as e:
+        if "CUDA out of memory" in str(e):
+            print("Error: CUDA out of memory. Skipping this request.")
+            return [Image.new("RGB", (512, 512), color="gray")]
+        else:
+            raise e
+
+def save_images(image, generated_images, prompt, negative_prompt, seed, guidance_scale, steps, strength, num_images_per_prompt, guidance_rescale):
+    timestamp = datetime.datetime.now().strftime("%Y%m%d_%H%M%S")
+    save_path = os.path.join('generated_images', timestamp)
+    os.makedirs(save_path, exist_ok=True)
+    
+    image.save(os.path.join(save_path, "reference.png"))
+    for i, gen_image in enumerate(generated_images):
+        gen_image.save(os.path.join(save_path, f"generated_{i}.png"))
+    
+    parameters = {
+        "prompt": prompt,
+        "negative_prompt": negative_prompt,
+        "seed": seed,
+        "guidance_scale": guidance_scale,
+        "steps": steps,
+        "strength": strength,
+        "num_images_per_prompt": num_images_per_prompt,
+        "guidance_rescale": guidance_rescale
+    }
+    with open(os.path.join(save_path, "parameters.json"), "w") as f:
+        json.dump(parameters, f, indent=4)
+
+def inference(image, prompt, negative_prompt, seed, guidance_scale, steps, strength, num_images_per_prompt, guidance_rescale):
+    try:
+        generated_images = generate_image_core(
+            image, prompt, negative_prompt, seed, guidance_scale, 
+            steps, strength, num_images_per_prompt, guidance_rescale
+        )
+        save_images(
+            image, generated_images, prompt, negative_prompt, seed,
+            guidance_scale, steps, strength, num_images_per_prompt, guidance_rescale
+        )
+        return generated_images
+    except RuntimeError as e:
+        if "CUDA out of memory" in str(e):
+            print("Error: CUDA out of memory. Skipping this request.")
+            return [Image.new("RGB", (512, 512), color="gray")]
+        else:
+            raise e
+
+# Create the Gradio interface
+iface = gr.Interface(
+    fn=inference,
+    inputs=[
+        gr.Image(type="pil", label="Input Image"),
+        gr.Textbox(label="Prompt", lines=2, placeholder="Enter your prompt here..."),
+        gr.Textbox(label="Negative Prompt", lines=2, placeholder="Enter your negative prompt here..."),
+        gr.Number(label="Seed", value=random.randint(1, 10000)),
+        gr.Number(label="Guidance Scale", value=7.5),
+        gr.Number(label="Steps", value=25, precision=0),
+        gr.Number(label="Strength", value=0.8),
+        gr.Number(label="Number of Images", value=1, precision=0),
+        gr.Number(label="Guidance Rescale", value=1.0)
+    ],
+    outputs=gr.Gallery(label="Generated Images"),
+    title="Terrain Generator",
+    description="Generate terrain images using Stable Diffusion with ControlNet and LoRA.",
+)
+
+# Launch the interface
 if __name__ == "__main__":
-    demo.launch()
+    iface.launch()

model.py

@@ -0,0 +1,303 @@
+from typing import Any, Dict, List, Optional, Tuple, Union
+
+import copy
+import torch
+from torch import nn, svd_lowrank
+
+from peft.tuners.lora import LoraLayer, Conv2d as PeftConv2d
+from diffusers.configuration_utils import register_to_config
+from diffusers.models.unets.unet_2d_condition import UNet2DConditionOutput, UNet2DConditionModel as UNet2DConditionModel
+
+
+class UNet2DConditionModelEx(UNet2DConditionModel):
+    @register_to_config
+    def __init__(
+        self,
+        sample_size: Optional[int] = None,
+        in_channels: int = 4,
+        out_channels: int = 4,
+        center_input_sample: bool = False,
+        flip_sin_to_cos: bool = True,
+        freq_shift: int = 0,
+        down_block_types: Tuple[str] = (
+            "CrossAttnDownBlock2D",
+            "CrossAttnDownBlock2D",
+            "CrossAttnDownBlock2D",
+            "DownBlock2D",
+        ),
+        mid_block_type: Optional[str] = "UNetMidBlock2DCrossAttn",
+        up_block_types: Tuple[str] = ("UpBlock2D", "CrossAttnUpBlock2D", "CrossAttnUpBlock2D", "CrossAttnUpBlock2D"),
+        only_cross_attention: Union[bool, Tuple[bool]] = False,
+        block_out_channels: Tuple[int] = (320, 640, 1280, 1280),
+        layers_per_block: Union[int, Tuple[int]] = 2,
+        downsample_padding: int = 1,
+        mid_block_scale_factor: float = 1,
+        dropout: float = 0.0,
+        act_fn: str = "silu",
+        norm_num_groups: Optional[int] = 32,
+        norm_eps: float = 1e-5,
+        cross_attention_dim: Union[int, Tuple[int]] = 1280,
+        transformer_layers_per_block: Union[int, Tuple[int], Tuple[Tuple]] = 1,
+        reverse_transformer_layers_per_block: Optional[Tuple[Tuple[int]]] = None,
+        encoder_hid_dim: Optional[int] = None,
+        encoder_hid_dim_type: Optional[str] = None,
+        attention_head_dim: Union[int, Tuple[int]] = 8,
+        num_attention_heads: Optional[Union[int, Tuple[int]]] = None,
+        dual_cross_attention: bool = False,
+        use_linear_projection: bool = False,
+        class_embed_type: Optional[str] = None,
+        addition_embed_type: Optional[str] = None,
+        addition_time_embed_dim: Optional[int] = None,
+        num_class_embeds: Optional[int] = None,
+        upcast_attention: bool = False,
+        resnet_time_scale_shift: str = "default",
+        resnet_skip_time_act: bool = False,
+        resnet_out_scale_factor: float = 1.0,
+        time_embedding_type: str = "positional",
+        time_embedding_dim: Optional[int] = None,
+        time_embedding_act_fn: Optional[str] = None,
+        timestep_post_act: Optional[str] = None,
+        time_cond_proj_dim: Optional[int] = None,
+        conv_in_kernel: int = 3,
+        conv_out_kernel: int = 3,
+        projection_class_embeddings_input_dim: Optional[int] = None,
+        attention_type: str = "default",
+        class_embeddings_concat: bool = False,
+        mid_block_only_cross_attention: Optional[bool] = None,
+        cross_attention_norm: Optional[str] = None,
+        addition_embed_type_num_heads: int = 64,
+        extra_condition_names: List[str] = [],
+    ):
+        num_extra_conditions = len(extra_condition_names)
+        super().__init__(
+            sample_size=sample_size,
+            in_channels=in_channels * (1 + num_extra_conditions),
+            out_channels=out_channels,
+            center_input_sample=center_input_sample,
+            flip_sin_to_cos=flip_sin_to_cos,
+            freq_shift=freq_shift,
+            down_block_types=down_block_types,
+            mid_block_type=mid_block_type,
+            up_block_types=up_block_types,
+            only_cross_attention=only_cross_attention,
+            block_out_channels=block_out_channels,
+            layers_per_block=layers_per_block,
+            downsample_padding=downsample_padding,
+            mid_block_scale_factor=mid_block_scale_factor,
+            dropout=dropout,
+            act_fn=act_fn,
+            norm_num_groups=norm_num_groups,
+            norm_eps=norm_eps,
+            cross_attention_dim=cross_attention_dim,
+            transformer_layers_per_block=transformer_layers_per_block,
+            reverse_transformer_layers_per_block=reverse_transformer_layers_per_block,
+            encoder_hid_dim=encoder_hid_dim,
+            encoder_hid_dim_type=encoder_hid_dim_type,
+            attention_head_dim=attention_head_dim,
+            num_attention_heads=num_attention_heads,
+            dual_cross_attention=dual_cross_attention,
+            use_linear_projection=use_linear_projection,
+            class_embed_type=class_embed_type,
+            addition_embed_type=addition_embed_type,
+            addition_time_embed_dim=addition_time_embed_dim,
+            num_class_embeds=num_class_embeds,
+            upcast_attention=upcast_attention,
+            resnet_time_scale_shift=resnet_time_scale_shift,
+            resnet_skip_time_act=resnet_skip_time_act,
+            resnet_out_scale_factor=resnet_out_scale_factor,
+            time_embedding_type=time_embedding_type,
+            time_embedding_dim=time_embedding_dim,
+            time_embedding_act_fn=time_embedding_act_fn,
+            timestep_post_act=timestep_post_act,
+            time_cond_proj_dim=time_cond_proj_dim,
+            conv_in_kernel=conv_in_kernel,
+            conv_out_kernel=conv_out_kernel,
+            projection_class_embeddings_input_dim=projection_class_embeddings_input_dim,
+            attention_type=attention_type,
+            class_embeddings_concat=class_embeddings_concat,
+            mid_block_only_cross_attention=mid_block_only_cross_attention,
+            cross_attention_norm=cross_attention_norm,
+            addition_embed_type_num_heads=addition_embed_type_num_heads,)
+        self._internal_dict = copy.deepcopy(self._internal_dict)
+        self.config.in_channels = in_channels
+        self.config.extra_condition_names = extra_condition_names
+    
+    @property
+    def extra_condition_names(self) -> List[str]:
+        return self.config.extra_condition_names
+
+    def add_extra_conditions(self, extra_condition_names: Union[str, List[str]]):
+        if isinstance(extra_condition_names, str):
+            extra_condition_names = [extra_condition_names]
+        conv_in_kernel = self.config.conv_in_kernel
+        conv_in_weight = self.conv_in.weight
+        self.config.extra_condition_names += extra_condition_names
+        full_in_channels = self.config.in_channels * (1 + len(self.config.extra_condition_names))
+        new_conv_in_weight = torch.zeros(
+            conv_in_weight.shape[0], full_in_channels, conv_in_kernel, conv_in_kernel,
+            dtype=conv_in_weight.dtype,
+            device=conv_in_weight.device,)
+        new_conv_in_weight[:,:conv_in_weight.shape[1]] = conv_in_weight
+        self.conv_in.weight = nn.Parameter(
+            new_conv_in_weight.data,
+            requires_grad=conv_in_weight.requires_grad,)
+        self.conv_in.in_channels = full_in_channels
+        
+        return self
+    
+    def activate_extra_condition_adapters(self):
+        lora_layers = [layer for layer in self.modules() if isinstance(layer, LoraLayer)]
+        if len(lora_layers) > 0:
+            self._hf_peft_config_loaded = True
+        for lora_layer in lora_layers:
+            adapter_names = [k for k in lora_layer.scaling.keys() if k in self.config.extra_condition_names] 
+            adapter_names += lora_layer.active_adapters
+            adapter_names = list(set(adapter_names))
+            lora_layer.set_adapter(adapter_names)
+    
+    def set_extra_condition_scale(self, scale: Union[float, List[float]] = 1.0):
+        if isinstance(scale, float):
+            scale = [scale] * len(self.config.extra_condition_names)
+
+        lora_layers = [layer for layer in self.modules() if isinstance(layer, LoraLayer)]
+        for s, n in zip(scale, self.config.extra_condition_names):
+            for lora_layer in lora_layers:
+                lora_layer.set_scale(n, s)
+    
+    @property
+    def default_half_lora_target_modules(self) -> List[str]:
+        module_names = []
+        for name, module in self.named_modules():
+            if "conv_out" in name or "up_blocks" in name:
+                continue
+            if isinstance(module, (nn.Linear, nn.Conv2d)):
+                module_names.append(name)
+        return list(set(module_names))
+    
+    @property
+    def default_full_lora_target_modules(self) -> List[str]:
+        module_names = []
+        for name, module in self.named_modules():
+            if isinstance(module, (nn.Linear, nn.Conv2d)):
+                module_names.append(name)
+        return list(set(module_names))
+    
+    @property
+    def default_half_skip_attn_lora_target_modules(self) -> List[str]:
+        return [
+            module_name
+            for module_name in self.default_half_lora_target_modules 
+            if all(
+                not module_name.endswith(attn_name) 
+                for attn_name in 
+                ["to_k", "to_q", "to_v", "to_out.0"]
+            )
+        ]
+    
+    @property
+    def default_full_skip_attn_lora_target_modules(self) -> List[str]:
+        return [
+            module_name
+            for module_name in self.default_full_lora_target_modules 
+            if all(
+                not module_name.endswith(attn_name) 
+                for attn_name in 
+                ["to_k", "to_q", "to_v", "to_out.0"]
+            )
+        ]
+
+    def forward(
+        self,
+        sample: torch.Tensor,
+        timestep: Union[torch.Tensor, float, int],
+        encoder_hidden_states: torch.Tensor,
+        class_labels: Optional[torch.Tensor] = None,
+        timestep_cond: Optional[torch.Tensor] = None,
+        attention_mask: Optional[torch.Tensor] = None,
+        cross_attention_kwargs: Optional[Dict[str, Any]] = None,
+        added_cond_kwargs: Optional[Dict[str, torch.Tensor]] = None,
+        down_block_additional_residuals: Optional[Tuple[torch.Tensor]] = None,
+        mid_block_additional_residual: Optional[torch.Tensor] = None,
+        down_intrablock_additional_residuals: Optional[Tuple[torch.Tensor]] = None,
+        encoder_attention_mask: Optional[torch.Tensor] = None,
+        extra_conditions: Optional[Union[torch.Tensor, List[torch.Tensor]]] = None,
+        return_dict: bool = True,
+    ) -> Union[UNet2DConditionOutput, Tuple]:
+        if extra_conditions is not None:
+            if isinstance(extra_conditions, list):
+                extra_conditions = torch.cat(extra_conditions, dim=1)
+            sample = torch.cat([sample, extra_conditions], dim=1)
+        return super().forward(
+            sample=sample,
+            timestep=timestep,
+            encoder_hidden_states=encoder_hidden_states,
+            class_labels=class_labels,
+            timestep_cond=timestep_cond,
+            attention_mask=attention_mask,
+            cross_attention_kwargs=cross_attention_kwargs,
+            added_cond_kwargs=added_cond_kwargs,
+            down_block_additional_residuals=down_block_additional_residuals,
+            mid_block_additional_residual=mid_block_additional_residual,
+            down_intrablock_additional_residuals=down_intrablock_additional_residuals,
+            encoder_attention_mask=encoder_attention_mask,
+            return_dict=return_dict,)
+
+
+class PeftConv2dEx(PeftConv2d):
+    def reset_lora_parameters(self, adapter_name, init_lora_weights):
+        if init_lora_weights is False:
+            return
+        
+        if isinstance(init_lora_weights, str) and "pissa" in init_lora_weights.lower():
+            if self.conv2d_pissa_init(adapter_name, init_lora_weights):
+                return
+            # Failed
+            init_lora_weights = "gaussian"
+
+        super(PeftConv2d, self).reset_lora_parameters(adapter_name, init_lora_weights)
+
+    def conv2d_pissa_init(self, adapter_name, init_lora_weights):
+        weight = weight_ori = self.get_base_layer().weight
+        weight = weight.flatten(start_dim=1)
+        if self.r[adapter_name] > weight.shape[0]:
+            return False
+        dtype = weight.dtype
+        if dtype not in [torch.float32, torch.float16, torch.bfloat16]:
+            raise TypeError(
+                "Please initialize PiSSA under float32, float16, or bfloat16. "
+                "Subsequently, re-quantize the residual model to help minimize quantization errors."
+            )
+        weight = weight.to(torch.float32)
+
+        if init_lora_weights == "pissa":
+            # USV^T = W <-> VSU^T = W^T, where W^T = weight.data in R^{out_channel, in_channel},
+            V, S, Uh = torch.linalg.svd(weight.data, full_matrices=False)
+            Vr = V[:, : self.r[adapter_name]]
+            Sr = S[: self.r[adapter_name]]
+            Sr /= self.scaling[adapter_name]
+            Uhr = Uh[: self.r[adapter_name]]
+        elif len(init_lora_weights.split("_niter_")) == 2:
+            Vr, Sr, Ur = svd_lowrank(
+                weight.data, self.r[adapter_name], niter=int(init_lora_weights.split("_niter_")[-1])
+            )
+            Sr /= self.scaling[adapter_name]
+            Uhr = Ur.t()
+        else:
+            raise ValueError(
+                f"init_lora_weights should be 'pissa' or 'pissa_niter_[number of iters]', got {init_lora_weights} instead."
+            )
+
+        lora_A = torch.diag(torch.sqrt(Sr)) @ Uhr
+        lora_B = Vr @ torch.diag(torch.sqrt(Sr))
+        self.lora_A[adapter_name].weight.data = lora_A.view([-1] + list(weight_ori.shape[1:]))
+        self.lora_B[adapter_name].weight.data = lora_B.view([-1, self.r[adapter_name]] + [1] * (weight_ori.ndim - 2))
+        weight = weight.data - self.scaling[adapter_name] * lora_B @ lora_A
+        weight = weight.to(dtype)
+        self.get_base_layer().weight.data = weight.view_as(weight_ori)
+        
+        return True
+
+
+# Patch peft conv2d
+PeftConv2d.reset_lora_parameters = PeftConv2dEx.reset_lora_parameters
+PeftConv2d.conv2d_pissa_init = PeftConv2dEx.conv2d_pissa_init

pipeline.py

@@ -0,0 +1,1378 @@
+import inspect
+from typing import Any, Callable, Dict, List, Optional, Tuple, Union
+
+import numpy as np
+import PIL.Image
+import torch
+import torch.nn.functional as F
+from transformers import CLIPImageProcessor, CLIPTextModel, CLIPTokenizer, CLIPVisionModelWithProjection
+
+from diffusers.callbacks import MultiPipelineCallbacks, PipelineCallback
+from diffusers.image_processor import PipelineImageInput, VaeImageProcessor
+from diffusers.loaders import FromSingleFileMixin, IPAdapterMixin, LoraLoaderMixin, TextualInversionLoaderMixin
+from diffusers.models import AutoencoderKL, ImageProjection
+from diffusers.models.lora import adjust_lora_scale_text_encoder
+from diffusers.schedulers import KarrasDiffusionSchedulers
+from diffusers.utils import (
+    USE_PEFT_BACKEND,
+    deprecate,
+    logging,
+    replace_example_docstring,
+    scale_lora_layers,
+    unscale_lora_layers,
+)
+from diffusers.utils.torch_utils import is_compiled_module, is_torch_version, randn_tensor
+from diffusers.pipelines.pipeline_utils import DiffusionPipeline, StableDiffusionMixin
+from diffusers.pipelines.stable_diffusion.pipeline_output import StableDiffusionPipelineOutput
+from diffusers.pipelines.stable_diffusion.safety_checker import StableDiffusionSafetyChecker
+from model import UNet2DConditionModelEx
+
+
+from huggingface_hub.utils import validate_hf_hub_args
+
+
+logger = logging.get_logger(__name__)  # pylint: disable=invalid-name
+
+
+EXAMPLE_DOC_STRING = """
+    Examples:
+        ```py
+        >>> # !pip install opencv-python transformers accelerate
+        >>> from diffusers import UniPCMultistepScheduler
+        >>> from diffusers.utils import load_image
+        >>> from model import UNet2DConditionModelEx
+        >>> from pipeline import StableDiffusionControlLoraV3Pipeline
+        >>> import numpy as np
+        >>> import torch
+
+        >>> import cv2
+        >>> from PIL import Image
+
+        >>> # download an image
+        >>> image = load_image(
+        ...     "https://hf.co/datasets/huggingface/documentation-images/resolve/main/diffusers/input_image_vermeer.png"
+        ... )
+        >>> image = np.array(image)
+
+        >>> # get canny image
+        >>> image = cv2.Canny(image, 100, 200)
+        >>> image = image[:, :, None]
+        >>> image = np.concatenate([image, image, image], axis=2)
+        >>> canny_image = Image.fromarray(image)
+
+        >>> # load stable diffusion v1-5 and control-lora-v3 
+        >>> unet: UNet2DConditionModelEx = UNet2DConditionModelEx.from_pretrained(
+        ...     "runwayml/stable-diffusion-v1-5", subfolder="unet", torch_dtype=torch.float16
+        ... )
+        >>> unet = unet.add_extra_conditions(["canny"])
+        >>> pipe = StableDiffusionControlLoraV3Pipeline.from_pretrained(
+        ...     "runwayml/stable-diffusion-v1-5", unet=unet, torch_dtype=torch.float16
+        ... )
+        >>> # load attention processors
+        >>> pipe.load_lora_weights("HighCWu/sd-control-lora-v3-canny")
+
+        >>> # speed up diffusion process with faster scheduler and memory optimization
+        >>> pipe.scheduler = UniPCMultistepScheduler.from_config(pipe.scheduler.config)
+        >>> # remove following line if xformers is not installed
+        >>> pipe.enable_xformers_memory_efficient_attention()
+
+        >>> pipe.enable_model_cpu_offload()
+
+        >>> # generate image
+        >>> generator = torch.manual_seed(0)
+        >>> image = pipe(
+        ...     "futuristic-looking woman", num_inference_steps=20, generator=generator, image=canny_image
+        ... ).images[0]
+        ```
+"""
+
+
+# Copied from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion.retrieve_timesteps
+def retrieve_timesteps(
+    scheduler,
+    num_inference_steps: Optional[int] = None,
+    device: Optional[Union[str, torch.device]] = None,
+    timesteps: Optional[List[int]] = None,
+    sigmas: Optional[List[float]] = None,
+    **kwargs,
+):
+    """
+    Calls the scheduler's `set_timesteps` method and retrieves timesteps from the scheduler after the call. Handles
+    custom timesteps. Any kwargs will be supplied to `scheduler.set_timesteps`.
+
+    Args:
+        scheduler (`SchedulerMixin`):
+            The scheduler to get timesteps from.
+        num_inference_steps (`int`):
+            The number of diffusion steps used when generating samples with a pre-trained model. If used, `timesteps`
+            must be `None`.
+        device (`str` or `torch.device`, *optional*):
+            The device to which the timesteps should be moved to. If `None`, the timesteps are not moved.
+        timesteps (`List[int]`, *optional*):
+            Custom timesteps used to override the timestep spacing strategy of the scheduler. If `timesteps` is passed,
+            `num_inference_steps` and `sigmas` must be `None`.
+        sigmas (`List[float]`, *optional*):
+            Custom sigmas used to override the timestep spacing strategy of the scheduler. If `sigmas` is passed,
+            `num_inference_steps` and `timesteps` must be `None`.
+
+    Returns:
+        `Tuple[torch.Tensor, int]`: A tuple where the first element is the timestep schedule from the scheduler and the
+        second element is the number of inference steps.
+    """
+    if timesteps is not None and sigmas is not None:
+        raise ValueError("Only one of `timesteps` or `sigmas` can be passed. Please choose one to set custom values")
+    if timesteps is not None:
+        accepts_timesteps = "timesteps" in set(inspect.signature(scheduler.set_timesteps).parameters.keys())
+        if not accepts_timesteps:
+            raise ValueError(
+                f"The current scheduler class {scheduler.__class__}'s `set_timesteps` does not support custom"
+                f" timestep schedules. Please check whether you are using the correct scheduler."
+            )
+        scheduler.set_timesteps(timesteps=timesteps, device=device, **kwargs)
+        timesteps = scheduler.timesteps
+        num_inference_steps = len(timesteps)
+    elif sigmas is not None:
+        accept_sigmas = "sigmas" in set(inspect.signature(scheduler.set_timesteps).parameters.keys())
+        if not accept_sigmas:
+            raise ValueError(
+                f"The current scheduler class {scheduler.__class__}'s `set_timesteps` does not support custom"
+                f" sigmas schedules. Please check whether you are using the correct scheduler."
+            )
+        scheduler.set_timesteps(sigmas=sigmas, device=device, **kwargs)
+        timesteps = scheduler.timesteps
+        num_inference_steps = len(timesteps)
+    else:
+        scheduler.set_timesteps(num_inference_steps, device=device, **kwargs)
+        timesteps = scheduler.timesteps
+    return timesteps, num_inference_steps
+
+
+class StableDiffusionControlLoraV3Pipeline(
+    DiffusionPipeline,
+    StableDiffusionMixin,
+    TextualInversionLoaderMixin,
+    LoraLoaderMixin,
+    IPAdapterMixin,
+    FromSingleFileMixin,
+):
+    r"""
+    Pipeline for text-to-image generation using Stable Diffusion with extra condition guidance.
+
+    This model inherits from [`DiffusionPipeline`]. Check the superclass documentation for the generic methods
+    implemented for all pipelines (downloading, saving, running on a particular device, etc.).
+
+    The pipeline also inherits the following loading methods:
+        - [`~loaders.TextualInversionLoaderMixin.load_textual_inversion`] for loading textual inversion embeddings
+        - [`~loaders.LoraLoaderMixin.load_lora_weights`] for loading LoRA weights
+        - [`~loaders.LoraLoaderMixin.save_lora_weights`] for saving LoRA weights
+        - [`~loaders.FromSingleFileMixin.from_single_file`] for loading `.ckpt` files
+        - [`~loaders.IPAdapterMixin.load_ip_adapter`] for loading IP Adapters
+
+    Args:
+        vae ([`AutoencoderKL`]):
+            Variational Auto-Encoder (VAE) model to encode and decode images to and from latent representations.
+        text_encoder ([`~transformers.CLIPTextModel`]):
+            Frozen text-encoder ([clip-vit-large-patch14](https://huggingface.co/openai/clip-vit-large-patch14)).
+        tokenizer ([`~transformers.CLIPTokenizer`]):
+            A `CLIPTokenizer` to tokenize text.
+        unet ([`UNet2DConditionModelEx`]):
+            A `UNet2DConditionModelEx` to denoise the encoded image latents with extra conditions.
+        scheduler ([`SchedulerMixin`]):
+            A scheduler to be used in combination with `unet` to denoise the encoded image latents. Can be one of
+            [`DDIMScheduler`], [`LMSDiscreteScheduler`], or [`PNDMScheduler`].
+        safety_checker ([`StableDiffusionSafetyChecker`]):
+            Classification module that estimates whether generated images could be considered offensive or harmful.
+            Please refer to the [model card](https://huggingface.co/runwayml/stable-diffusion-v1-5) for more details
+            about a model's potential harms.
+        feature_extractor ([`~transformers.CLIPImageProcessor`]):
+            A `CLIPImageProcessor` to extract features from generated images; used as inputs to the `safety_checker`.
+    """
+
+    model_cpu_offload_seq = "text_encoder->image_encoder->unet->vae"
+    _optional_components = ["safety_checker", "feature_extractor", "image_encoder"]
+    _exclude_from_cpu_offload = ["safety_checker"]
+    _callback_tensor_inputs = ["latents", "prompt_embeds", "negative_prompt_embeds"]
+
+    def __init__(
+        self,
+        vae: AutoencoderKL,
+        text_encoder: CLIPTextModel,
+        tokenizer: CLIPTokenizer,
+        unet: UNet2DConditionModelEx,
+        scheduler: KarrasDiffusionSchedulers,
+        safety_checker: StableDiffusionSafetyChecker,
+        feature_extractor: CLIPImageProcessor,
+        image_encoder: CLIPVisionModelWithProjection = None,
+        requires_safety_checker: bool = True,
+    ):
+        super().__init__()
+
+        if safety_checker is None and requires_safety_checker:
+            logger.warning(
+                f"You have disabled the safety checker for {self.__class__} by passing `safety_checker=None`. Ensure"
+                " that you abide to the conditions of the Stable Diffusion license and do not expose unfiltered"
+                " results in services or applications open to the public. Both the diffusers team and Hugging Face"
+                " strongly recommend to keep the safety filter enabled in all public facing circumstances, disabling"
+                " it only for use-cases that involve analyzing network behavior or auditing its results. For more"
+                " information, please have a look at https://github.com/huggingface/diffusers/pull/254 ."
+            )
+
+        if safety_checker is not None and feature_extractor is None:
+            raise ValueError(
+                "Make sure to define a feature extractor when loading {self.__class__} if you want to use the safety"
+                " checker. If you do not want to use the safety checker, you can pass `'safety_checker=None'` instead."
+            )
+
+        self.register_modules(
+            vae=vae,
+            text_encoder=text_encoder,
+            tokenizer=tokenizer,
+            unet=unet,
+            scheduler=scheduler,
+            safety_checker=safety_checker,
+            feature_extractor=feature_extractor,
+            image_encoder=image_encoder,
+        )
+        self.vae_scale_factor = 2 ** (len(self.vae.config.block_out_channels) - 1)
+        self.image_processor = VaeImageProcessor(vae_scale_factor=self.vae_scale_factor, do_convert_rgb=True)
+        self.register_to_config(requires_safety_checker=requires_safety_checker)
+
+    # Copied from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion.StableDiffusionPipeline._encode_prompt
+    def _encode_prompt(
+        self,
+        prompt,
+        device,
+        num_images_per_prompt,
+        do_classifier_free_guidance,
+        negative_prompt=None,
+        prompt_embeds: Optional[torch.Tensor] = None,
+        negative_prompt_embeds: Optional[torch.Tensor] = None,
+        lora_scale: Optional[float] = None,
+        **kwargs,
+    ):
+        deprecation_message = "`_encode_prompt()` is deprecated and it will be removed in a future version. Use `encode_prompt()` instead. Also, be aware that the output format changed from a concatenated tensor to a tuple."
+        deprecate("_encode_prompt()", "1.0.0", deprecation_message, standard_warn=False)
+
+        prompt_embeds_tuple = self.encode_prompt(
+            prompt=prompt,
+            device=device,
+            num_images_per_prompt=num_images_per_prompt,
+            do_classifier_free_guidance=do_classifier_free_guidance,
+            negative_prompt=negative_prompt,
+            prompt_embeds=prompt_embeds,
+            negative_prompt_embeds=negative_prompt_embeds,
+            lora_scale=lora_scale,
+            **kwargs,
+        )
+
+        # concatenate for backwards comp
+        prompt_embeds = torch.cat([prompt_embeds_tuple[1], prompt_embeds_tuple[0]])
+
+        return prompt_embeds
+
+    # Copied from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion.StableDiffusionPipeline.encode_prompt
+    def encode_prompt(
+        self,
+        prompt,
+        device,
+        num_images_per_prompt,
+        do_classifier_free_guidance,
+        negative_prompt=None,
+        prompt_embeds: Optional[torch.Tensor] = None,
+        negative_prompt_embeds: Optional[torch.Tensor] = None,
+        lora_scale: Optional[float] = None,
+        clip_skip: Optional[int] = None,
+    ):
+        r"""
+        Encodes the prompt into text encoder hidden states.
+
+        Args:
+            prompt (`str` or `List[str]`, *optional*):
+                prompt to be encoded
+            device: (`torch.device`):
+                torch device
+            num_images_per_prompt (`int`):
+                number of images that should be generated per prompt
+            do_classifier_free_guidance (`bool`):
+                whether to use classifier free guidance or not
+            negative_prompt (`str` or `List[str]`, *optional*):
+                The prompt or prompts not to guide the image generation. If not defined, one has to pass
+                `negative_prompt_embeds` instead. Ignored when not using guidance (i.e., ignored if `guidance_scale` is
+                less than `1`).
+            prompt_embeds (`torch.Tensor`, *optional*):
+                Pre-generated text embeddings. Can be used to easily tweak text inputs, *e.g.* prompt weighting. If not
+                provided, text embeddings will be generated from `prompt` input argument.
+            negative_prompt_embeds (`torch.Tensor`, *optional*):
+                Pre-generated negative text embeddings. Can be used to easily tweak text inputs, *e.g.* prompt
+                weighting. If not provided, negative_prompt_embeds will be generated from `negative_prompt` input
+                argument.
+            lora_scale (`float`, *optional*):
+                A LoRA scale that will be applied to all LoRA layers of the text encoder if LoRA layers are loaded.
+            clip_skip (`int`, *optional*):
+                Number of layers to be skipped from CLIP while computing the prompt embeddings. A value of 1 means that
+                the output of the pre-final layer will be used for computing the prompt embeddings.
+        """
+        # set lora scale so that monkey patched LoRA
+        # function of text encoder can correctly access it
+        if lora_scale is not None and isinstance(self, LoraLoaderMixin):
+            self._lora_scale = lora_scale
+
+            # dynamically adjust the LoRA scale
+            if not USE_PEFT_BACKEND:
+                adjust_lora_scale_text_encoder(self.text_encoder, lora_scale)
+            else:
+                scale_lora_layers(self.text_encoder, lora_scale)
+
+        if prompt is not None and isinstance(prompt, str):
+            batch_size = 1
+        elif prompt is not None and isinstance(prompt, list):
+            batch_size = len(prompt)
+        else:
+            batch_size = prompt_embeds.shape[0]
+
+        if prompt_embeds is None:
+            # textual inversion: process multi-vector tokens if necessary
+            if isinstance(self, TextualInversionLoaderMixin):
+                prompt = self.maybe_convert_prompt(prompt, self.tokenizer)
+
+            text_inputs = self.tokenizer(
+                prompt,
+                padding="max_length",
+                max_length=self.tokenizer.model_max_length,
+                truncation=True,
+                return_tensors="pt",
+            )
+            text_input_ids = text_inputs.input_ids
+            untruncated_ids = self.tokenizer(prompt, padding="longest", return_tensors="pt").input_ids
+
+            if untruncated_ids.shape[-1] >= text_input_ids.shape[-1] and not torch.equal(
+                text_input_ids, untruncated_ids
+            ):
+                removed_text = self.tokenizer.batch_decode(
+                    untruncated_ids[:, self.tokenizer.model_max_length - 1 : -1]
+                )
+                logger.warning(
+                    "The following part of your input was truncated because CLIP can only handle sequences up to"
+                    f" {self.tokenizer.model_max_length} tokens: {removed_text}"
+                )
+
+            if hasattr(self.text_encoder.config, "use_attention_mask") and self.text_encoder.config.use_attention_mask:
+                attention_mask = text_inputs.attention_mask.to(device)
+            else:
+                attention_mask = None
+
+            if clip_skip is None:
+                prompt_embeds = self.text_encoder(text_input_ids.to(device), attention_mask=attention_mask)
+                prompt_embeds = prompt_embeds[0]
+            else:
+                prompt_embeds = self.text_encoder(
+                    text_input_ids.to(device), attention_mask=attention_mask, output_hidden_states=True
+                )
+                # Access the `hidden_states` first, that contains a tuple of
+                # all the hidden states from the encoder layers. Then index into
+                # the tuple to access the hidden states from the desired layer.
+                prompt_embeds = prompt_embeds[-1][-(clip_skip + 1)]
+                # We also need to apply the final LayerNorm here to not mess with the
+                # representations. The `last_hidden_states` that we typically use for
+                # obtaining the final prompt representations passes through the LayerNorm
+                # layer.
+                prompt_embeds = self.text_encoder.text_model.final_layer_norm(prompt_embeds)
+
+        if self.text_encoder is not None:
+            prompt_embeds_dtype = self.text_encoder.dtype
+        elif self.unet is not None:
+            prompt_embeds_dtype = self.unet.dtype
+        else:
+            prompt_embeds_dtype = prompt_embeds.dtype
+
+        prompt_embeds = prompt_embeds.to(dtype=prompt_embeds_dtype, device=device)
+
+        bs_embed, seq_len, _ = prompt_embeds.shape
+        # duplicate text embeddings for each generation per prompt, using mps friendly method
+        prompt_embeds = prompt_embeds.repeat(1, num_images_per_prompt, 1)
+        prompt_embeds = prompt_embeds.view(bs_embed * num_images_per_prompt, seq_len, -1)
+
+        # get unconditional embeddings for classifier free guidance
+        if do_classifier_free_guidance and negative_prompt_embeds is None:
+            uncond_tokens: List[str]
+            if negative_prompt is None:
+                uncond_tokens = [""] * batch_size
+            elif prompt is not None and type(prompt) is not type(negative_prompt):
+                raise TypeError(
+                    f"`negative_prompt` should be the same type to `prompt`, but got {type(negative_prompt)} !="
+                    f" {type(prompt)}."
+                )
+            elif isinstance(negative_prompt, str):
+                uncond_tokens = [negative_prompt]
+            elif batch_size != len(negative_prompt):
+                raise ValueError(
+                    f"`negative_prompt`: {negative_prompt} has batch size {len(negative_prompt)}, but `prompt`:"
+                    f" {prompt} has batch size {batch_size}. Please make sure that passed `negative_prompt` matches"
+                    " the batch size of `prompt`."
+                )
+            else:
+                uncond_tokens = negative_prompt
+
+            # textual inversion: process multi-vector tokens if necessary
+            if isinstance(self, TextualInversionLoaderMixin):
+                uncond_tokens = self.maybe_convert_prompt(uncond_tokens, self.tokenizer)
+
+            max_length = prompt_embeds.shape[1]
+            uncond_input = self.tokenizer(
+                uncond_tokens,
+                padding="max_length",
+                max_length=max_length,
+                truncation=True,
+                return_tensors="pt",
+            )
+
+            if hasattr(self.text_encoder.config, "use_attention_mask") and self.text_encoder.config.use_attention_mask:
+                attention_mask = uncond_input.attention_mask.to(device)
+            else:
+                attention_mask = None
+
+            negative_prompt_embeds = self.text_encoder(
+                uncond_input.input_ids.to(device),
+                attention_mask=attention_mask,
+            )
+            negative_prompt_embeds = negative_prompt_embeds[0]
+
+        if do_classifier_free_guidance:
+            # duplicate unconditional embeddings for each generation per prompt, using mps friendly method
+            seq_len = negative_prompt_embeds.shape[1]
+
+            negative_prompt_embeds = negative_prompt_embeds.to(dtype=prompt_embeds_dtype, device=device)
+
+            negative_prompt_embeds = negative_prompt_embeds.repeat(1, num_images_per_prompt, 1)
+            negative_prompt_embeds = negative_prompt_embeds.view(batch_size * num_images_per_prompt, seq_len, -1)
+
+        if self.text_encoder is not None:
+            if isinstance(self, LoraLoaderMixin) and USE_PEFT_BACKEND:
+                # Retrieve the original scale by scaling back the LoRA layers
+                unscale_lora_layers(self.text_encoder, lora_scale)
+
+        return prompt_embeds, negative_prompt_embeds
+
+    # Copied from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion.StableDiffusionPipeline.encode_image
+    def encode_image(self, image, device, num_images_per_prompt, output_hidden_states=None):
+        dtype = next(self.image_encoder.parameters()).dtype
+
+        if not isinstance(image, torch.Tensor):
+            image = self.feature_extractor(image, return_tensors="pt").pixel_values
+
+        image = image.to(device=device, dtype=dtype)
+        if output_hidden_states:
+            image_enc_hidden_states = self.image_encoder(image, output_hidden_states=True).hidden_states[-2]
+            image_enc_hidden_states = image_enc_hidden_states.repeat_interleave(num_images_per_prompt, dim=0)
+            uncond_image_enc_hidden_states = self.image_encoder(
+                torch.zeros_like(image), output_hidden_states=True
+            ).hidden_states[-2]
+            uncond_image_enc_hidden_states = uncond_image_enc_hidden_states.repeat_interleave(
+                num_images_per_prompt, dim=0
+            )
+            return image_enc_hidden_states, uncond_image_enc_hidden_states
+        else:
+            image_embeds = self.image_encoder(image).image_embeds
+            image_embeds = image_embeds.repeat_interleave(num_images_per_prompt, dim=0)
+            uncond_image_embeds = torch.zeros_like(image_embeds)
+
+            return image_embeds, uncond_image_embeds
+
+    # Copied from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion.StableDiffusionPipeline.prepare_ip_adapter_image_embeds
+    def prepare_ip_adapter_image_embeds(
+        self, ip_adapter_image, ip_adapter_image_embeds, device, num_images_per_prompt, do_classifier_free_guidance
+    ):
+        if ip_adapter_image_embeds is None:
+            if not isinstance(ip_adapter_image, list):
+                ip_adapter_image = [ip_adapter_image]
+
+            if len(ip_adapter_image) != len(self.unet.encoder_hid_proj.image_projection_layers):
+                raise ValueError(
+                    f"`ip_adapter_image` must have same length as the number of IP Adapters. Got {len(ip_adapter_image)} images and {len(self.unet.encoder_hid_proj.image_projection_layers)} IP Adapters."
+                )
+
+            image_embeds = []
+            for single_ip_adapter_image, image_proj_layer in zip(
+                ip_adapter_image, self.unet.encoder_hid_proj.image_projection_layers
+            ):
+                output_hidden_state = not isinstance(image_proj_layer, ImageProjection)
+                single_image_embeds, single_negative_image_embeds = self.encode_image(
+                    single_ip_adapter_image, device, 1, output_hidden_state
+                )
+                single_image_embeds = torch.stack([single_image_embeds] * num_images_per_prompt, dim=0)
+                single_negative_image_embeds = torch.stack(
+                    [single_negative_image_embeds] * num_images_per_prompt, dim=0
+                )
+
+                if do_classifier_free_guidance:
+                    single_image_embeds = torch.cat([single_negative_image_embeds, single_image_embeds])
+                    single_image_embeds = single_image_embeds.to(device)
+
+                image_embeds.append(single_image_embeds)
+        else:
+            repeat_dims = [1]
+            image_embeds = []
+            for single_image_embeds in ip_adapter_image_embeds:
+                if do_classifier_free_guidance:
+                    single_negative_image_embeds, single_image_embeds = single_image_embeds.chunk(2)
+                    single_image_embeds = single_image_embeds.repeat(
+                        num_images_per_prompt, *(repeat_dims * len(single_image_embeds.shape[1:]))
+                    )
+                    single_negative_image_embeds = single_negative_image_embeds.repeat(
+                        num_images_per_prompt, *(repeat_dims * len(single_negative_image_embeds.shape[1:]))
+                    )
+                    single_image_embeds = torch.cat([single_negative_image_embeds, single_image_embeds])
+                else:
+                    single_image_embeds = single_image_embeds.repeat(
+                        num_images_per_prompt, *(repeat_dims * len(single_image_embeds.shape[1:]))
+                    )
+                image_embeds.append(single_image_embeds)
+
+        return image_embeds
+
+    # Copied from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion.StableDiffusionPipeline.run_safety_checker
+    def run_safety_checker(self, image, device, dtype):
+        if self.safety_checker is None:
+            has_nsfw_concept = None
+        else:
+            has_nsfw_concept = None
+            #if torch.is_tensor(image):
+            #    feature_extractor_input = self.image_processor.postprocess(image, output_type="pil")
+            #else:
+            #    feature_extractor_input = self.image_processor.numpy_to_pil(image)
+            #safety_checker_input = self.feature_extractor(feature_extractor_input, return_tensors="pt").to(device)
+            #image, has_nsfw_concept = self.safety_checker(
+            #    images=image, clip_input=safety_checker_input.pixel_values.to(dtype)
+            #)
+        return image, has_nsfw_concept
+
+    # Copied from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion.StableDiffusionPipeline.decode_latents
+    def decode_latents(self, latents):
+        deprecation_message = "The decode_latents method is deprecated and will be removed in 1.0.0. Please use VaeImageProcessor.postprocess(...) instead"
+        deprecate("decode_latents", "1.0.0", deprecation_message, standard_warn=False)
+
+        latents = 1 / self.vae.config.scaling_factor * latents
+        image = self.vae.decode(latents, return_dict=False)[0]
+        image = (image / 2 + 0.5).clamp(0, 1)
+        # we always cast to float32 as this does not cause significant overhead and is compatible with bfloat16
+        image = image.cpu().permute(0, 2, 3, 1).float().numpy()
+        return image
+
+    # Copied from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion.StableDiffusionPipeline.prepare_extra_step_kwargs
+    def prepare_extra_step_kwargs(self, generator, eta):
+        # prepare extra kwargs for the scheduler step, since not all schedulers have the same signature
+        # eta (η) is only used with the DDIMScheduler, it will be ignored for other schedulers.
+        # eta corresponds to η in DDIM paper: https://arxiv.org/abs/2010.02502
+        # and should be between [0, 1]
+
+        accepts_eta = "eta" in set(inspect.signature(self.scheduler.step).parameters.keys())
+        extra_step_kwargs = {}
+        if accepts_eta:
+            extra_step_kwargs["eta"] = eta
+
+        # check if the scheduler accepts generator
+        accepts_generator = "generator" in set(inspect.signature(self.scheduler.step).parameters.keys())
+        if accepts_generator:
+            extra_step_kwargs["generator"] = generator
+        return extra_step_kwargs
+
+    def check_inputs(
+        self,
+        prompt,
+        image,
+        callback_steps,
+        negative_prompt=None,
+        prompt_embeds=None,
+        negative_prompt_embeds=None,
+        ip_adapter_image=None,
+        ip_adapter_image_embeds=None,
+        extra_condition_scale=1.0,
+        control_guidance_start=0.0,
+        control_guidance_end=1.0,
+        callback_on_step_end_tensor_inputs=None,
+    ):
+        if callback_steps is not None and (not isinstance(callback_steps, int) or callback_steps <= 0):
+            raise ValueError(
+                f"`callback_steps` has to be a positive integer but is {callback_steps} of type"
+                f" {type(callback_steps)}."
+            )
+
+        if callback_on_step_end_tensor_inputs is not None and not all(
+            k in self._callback_tensor_inputs for k in callback_on_step_end_tensor_inputs
+        ):
+            raise ValueError(
+                f"`callback_on_step_end_tensor_inputs` has to be in {self._callback_tensor_inputs}, but found {[k for k in callback_on_step_end_tensor_inputs if k not in self._callback_tensor_inputs]}"
+            )
+
+        if prompt is not None and prompt_embeds is not None:
+            raise ValueError(
+                f"Cannot forward both `prompt`: {prompt} and `prompt_embeds`: {prompt_embeds}. Please make sure to"
+                " only forward one of the two."
+            )
+        elif prompt is None and prompt_embeds is None:
+            raise ValueError(
+                "Provide either `prompt` or `prompt_embeds`. Cannot leave both `prompt` and `prompt_embeds` undefined."
+            )
+        elif prompt is not None and (not isinstance(prompt, str) and not isinstance(prompt, list)):
+            raise ValueError(f"`prompt` has to be of type `str` or `list` but is {type(prompt)}")
+
+        if negative_prompt is not None and negative_prompt_embeds is not None:
+            raise ValueError(
+                f"Cannot forward both `negative_prompt`: {negative_prompt} and `negative_prompt_embeds`:"
+                f" {negative_prompt_embeds}. Please make sure to only forward one of the two."
+            )
+
+        if prompt_embeds is not None and negative_prompt_embeds is not None:
+            if prompt_embeds.shape != negative_prompt_embeds.shape:
+                raise ValueError(
+                    "`prompt_embeds` and `negative_prompt_embeds` must have the same shape when passed directly, but"
+                    f" got: `prompt_embeds` {prompt_embeds.shape} != `negative_prompt_embeds`"
+                    f" {negative_prompt_embeds.shape}."
+                )
+
+        # Check `image`
+        unet: UNet2DConditionModelEx = self.unet._orig_mod if is_compiled_module(self.unet) else self.unet
+        num_extra_conditions = len(unet.extra_condition_names)
+        if num_extra_conditions == 1:
+            self.check_image(image, prompt, prompt_embeds)
+        elif num_extra_conditions > 1:
+            if not isinstance(image, list):
+                raise TypeError("For multiple extra conditions: `image` must be type `list`")
+
+            # When `image` is a nested list:
+            # (e.g. [[canny_image_1, pose_image_1], [canny_image_2, pose_image_2]])
+            elif any(isinstance(i, list) for i in image):
+                transposed_image = [list(t) for t in zip(*image)]
+                if len(transposed_image) != num_extra_conditions:
+                    raise ValueError(
+                        f"For multiple extra conditions: if you pass`image` as a list of list, each sublist must have the same length as the number of extra conditions, but the sublists in `image` got {len(transposed_image)} images and {num_extra_conditions} extra conditions."
+                    )
+                for image_ in transposed_image:
+                    self.check_image(image_, prompt, prompt_embeds)
+            elif len(image) != num_extra_conditions:
+                raise ValueError(
+                    f"For multiple extra conditions: `image` must have the same length as the number of extra conditions, but got {len(image)} images and {num_extra_conditions} extra conditions."
+                )
+            else:
+                for image_ in image:
+                    self.check_image(image_, prompt, prompt_embeds)
+        else:
+            assert False
+
+        # Check `extra_condition_scale`
+        if num_extra_conditions == 1:
+            if not isinstance(extra_condition_scale, float):
+                raise TypeError("For single extra condition: `extra_condition_scale` must be type `float`.")
+        elif num_extra_conditions >= 1:
+            if isinstance(extra_condition_scale, list):
+                if any(isinstance(i, list) for i in extra_condition_scale):
+                    raise ValueError(
+                        "A single batch of varying conditioning scale settings (e.g. [[1.0, 0.5], [0.2, 0.8]]) is not supported at the moment. "
+                        "The conditioning scale must be fixed across the batch."
+                    )
+            elif isinstance(extra_condition_scale, list) and len(extra_condition_scale) != num_extra_conditions:
+                raise ValueError(
+                    "For multiple extra conditions: When `extra_condition_scale` is specified as `list`, it must have"
+                    " the same length as the number of extra conditions"
+                )
+        else:
+            assert False
+
+        if not isinstance(control_guidance_start, (tuple, list)):
+            control_guidance_start = [control_guidance_start]
+
+        if not isinstance(control_guidance_end, (tuple, list)):
+            control_guidance_end = [control_guidance_end]
+
+        if len(control_guidance_start) != len(control_guidance_end):
+            raise ValueError(
+                f"`control_guidance_start` has {len(control_guidance_start)} elements, but `control_guidance_end` has {len(control_guidance_end)} elements. Make sure to provide the same number of elements to each list."
+            )
+
+        if num_extra_conditions > 1:
+            if len(control_guidance_start) != num_extra_conditions:
+                raise ValueError(
+                    f"`control_guidance_start`: {control_guidance_start} has {len(control_guidance_start)} elements but there are {num_extra_conditions} extra conditions available. Make sure to provide {num_extra_conditions}."
+                )
+
+        for start, end in zip(control_guidance_start, control_guidance_end):
+            if start >= end:
+                raise ValueError(
+                    f"control guidance start: {start} cannot be larger or equal to control guidance end: {end}."
+                )
+            if start < 0.0:
+                raise ValueError(f"control guidance start: {start} can't be smaller than 0.")
+            if end > 1.0:
+                raise ValueError(f"control guidance end: {end} can't be larger than 1.0.")
+
+        if ip_adapter_image is not None and ip_adapter_image_embeds is not None:
+            raise ValueError(
+                "Provide either `ip_adapter_image` or `ip_adapter_image_embeds`. Cannot leave both `ip_adapter_image` and `ip_adapter_image_embeds` defined."
+            )
+
+        if ip_adapter_image_embeds is not None:
+            if not isinstance(ip_adapter_image_embeds, list):
+                raise ValueError(
+                    f"`ip_adapter_image_embeds` has to be of type `list` but is {type(ip_adapter_image_embeds)}"
+                )
+            elif ip_adapter_image_embeds[0].ndim not in [3, 4]:
+                raise ValueError(
+                    f"`ip_adapter_image_embeds` has to be a list of 3D or 4D tensors but is {ip_adapter_image_embeds[0].ndim}D"
+                )
+
+    def check_image(self, image, prompt, prompt_embeds):
+        image_is_pil = isinstance(image, PIL.Image.Image)
+        image_is_tensor = isinstance(image, torch.Tensor)
+        image_is_np = isinstance(image, np.ndarray)
+        image_is_pil_list = isinstance(image, list) and isinstance(image[0], PIL.Image.Image)
+        image_is_tensor_list = isinstance(image, list) and isinstance(image[0], torch.Tensor)
+        image_is_np_list = isinstance(image, list) and isinstance(image[0], np.ndarray)
+
+        if (
+            not image_is_pil
+            and not image_is_tensor
+            and not image_is_np
+            and not image_is_pil_list
+            and not image_is_tensor_list
+            and not image_is_np_list
+        ):
+            raise TypeError(
+                f"image must be passed and be one of PIL image, numpy array, torch tensor, list of PIL images, list of numpy arrays or list of torch tensors, but is {type(image)}"
+            )
+
+        if image_is_pil:
+            image_batch_size = 1
+        else:
+            image_batch_size = len(image)
+
+        if prompt is not None and isinstance(prompt, str):
+            prompt_batch_size = 1
+        elif prompt is not None and isinstance(prompt, list):
+            prompt_batch_size = len(prompt)
+        elif prompt_embeds is not None:
+            prompt_batch_size = prompt_embeds.shape[0]
+
+        if image_batch_size != 1 and image_batch_size != prompt_batch_size:
+            raise ValueError(
+                f"If image batch size is not 1, image batch size must be same as prompt batch size. image batch size: {image_batch_size}, prompt batch size: {prompt_batch_size}"
+            )
+
+    def prepare_image(
+        self,
+        image,
+        width,
+        height,
+        batch_size,
+        num_images_per_prompt,
+        device,
+        dtype,
+        do_classifier_free_guidance=False,
+    ):
+        image = self.image_processor.preprocess(image, height=height, width=width).to(dtype=torch.float32)
+        image_batch_size = image.shape[0]
+
+        if image_batch_size == 1:
+            repeat_by = batch_size
+        else:
+            # image batch size is the same as prompt batch size
+            repeat_by = num_images_per_prompt
+
+        image = image.repeat_interleave(repeat_by, dim=0)
+
+        image = image.to(device=device, dtype=dtype)
+
+        if do_classifier_free_guidance:
+            image = torch.cat([image] * 2)
+
+        return image
+
+    # Copied from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion.StableDiffusionPipeline.prepare_latents
+    def prepare_latents(self, batch_size, num_channels_latents, height, width, dtype, device, generator, latents=None):
+        shape = (
+            batch_size,
+            num_channels_latents,
+            int(height) // self.vae_scale_factor,
+            int(width) // self.vae_scale_factor,
+        )
+        if isinstance(generator, list) and len(generator) != batch_size:
+            raise ValueError(
+                f"You have passed a list of generators of length {len(generator)}, but requested an effective batch"
+                f" size of {batch_size}. Make sure the batch size matches the length of the generators."
+            )
+
+        if latents is None:
+            latents = randn_tensor(shape, generator=generator, device=device, dtype=dtype)
+        else:
+            latents = latents.to(device)
+
+        # scale the initial noise by the standard deviation required by the scheduler
+        latents = latents * self.scheduler.init_noise_sigma
+        return latents
+
+    # Copied from diffusers.pipelines.latent_consistency_models.pipeline_latent_consistency_text2img.LatentConsistencyModelPipeline.get_guidance_scale_embedding
+    def get_guidance_scale_embedding(
+        self, w: torch.Tensor, embedding_dim: int = 512, dtype: torch.dtype = torch.float32
+    ) -> torch.Tensor:
+        """
+        See https://github.com/google-research/vdm/blob/dc27b98a554f65cdc654b800da5aa1846545d41b/model_vdm.py#L298
+
+        Args:
+            w (`torch.Tensor`):
+                Generate embedding vectors with a specified guidance scale to subsequently enrich timestep embeddings.
+            embedding_dim (`int`, *optional*, defaults to 512):
+                Dimension of the embeddings to generate.
+            dtype (`torch.dtype`, *optional*, defaults to `torch.float32`):
+                Data type of the generated embeddings.
+
+        Returns:
+            `torch.Tensor`: Embedding vectors with shape `(len(w), embedding_dim)`.
+        """
+        assert len(w.shape) == 1
+        w = w * 1000.0
+
+        half_dim = embedding_dim // 2
+        emb = torch.log(torch.tensor(10000.0)) / (half_dim - 1)
+        emb = torch.exp(torch.arange(half_dim, dtype=dtype) * -emb)
+        emb = w.to(dtype)[:, None] * emb[None, :]
+        emb = torch.cat([torch.sin(emb), torch.cos(emb)], dim=1)
+        if embedding_dim % 2 == 1:  # zero pad
+            emb = torch.nn.functional.pad(emb, (0, 1))
+        assert emb.shape == (w.shape[0], embedding_dim)
+        return emb
+
+    @property
+    def guidance_scale(self):
+        return self._guidance_scale
+
+    @property
+    def clip_skip(self):
+        return self._clip_skip
+
+    # here `guidance_scale` is defined analog to the guidance weight `w` of equation (2)
+    # of the Imagen paper: https://arxiv.org/pdf/2205.11487.pdf . `guidance_scale = 1`
+    # corresponds to doing no classifier free guidance.
+    @property
+    def do_classifier_free_guidance(self):
+        return self._guidance_scale > 1 and self.unet.config.time_cond_proj_dim is None
+
+    @property
+    def cross_attention_kwargs(self):
+        return self._cross_attention_kwargs
+
+    @property
+    def num_timesteps(self):
+        return self._num_timesteps
+
+    @classmethod
+    @validate_hf_hub_args
+    def lora_state_dict(
+        cls,
+        pretrained_model_name_or_path_or_dict: Union[str, Dict[str, torch.Tensor]],
+        **kwargs,
+    ):
+        # Override to add support for different LoRA alphas
+        state_dict, network_alphas = super(StableDiffusionControlLoraV3Pipeline, cls).lora_state_dict(
+            pretrained_model_name_or_path_or_dict, **kwargs
+        )
+        if network_alphas is None:
+            network_alphas = {}
+            for k, v in state_dict.items():
+                if ".lora_A." in k:
+                    network_alphas[".".join(k.split(".lora_A.")[0].split(".") + ["alpha"])] = v.shape[0]
+        return state_dict, network_alphas
+    
+    def load_lora_weights(
+        self, 
+        pretrained_model_name_or_path_or_dict: Union[
+            Union[str, Dict[str, torch.Tensor]], 
+            List[Union[str, Dict[str, torch.Tensor]]]
+        ],
+        adapter_name=None,
+        **kwargs
+    ):
+        unet: UNet2DConditionModelEx = self.unet._orig_mod if is_compiled_module(self.unet) else self.unet
+        num_condition_names = len(unet.extra_condition_names)
+        in_channels = unet.config.in_channels
+
+        kwargs["weight_name"] = kwargs.pop("weight_name", "pytorch_lora_weights.safetensors")
+
+        if adapter_name is not None and adapter_name not in unet.extra_condition_names:
+            unet._hf_peft_config_loaded = True
+            super().load_lora_weights(pretrained_model_name_or_path_or_dict, adapter_name, **kwargs)
+            unet.set_adapter(adapter_name)
+            return
+
+        if not isinstance(pretrained_model_name_or_path_or_dict, list):
+            pretrained_model_name_or_path_or_dict = [pretrained_model_name_or_path_or_dict] * num_condition_names
+        pretrained_model_name_or_path_or_dict_list = pretrained_model_name_or_path_or_dict
+
+        assert len(pretrained_model_name_or_path_or_dict) == len(unet.extra_condition_names)
+
+        adapter_name_ori = adapter_name
+        for i, (pretrained_model_name_or_path_or_dict, adapter_name) in enumerate(zip(
+            pretrained_model_name_or_path_or_dict_list, 
+            unet.extra_condition_names
+        )):
+            _kwargs = {**kwargs}
+            subfolder = _kwargs.pop("subfolder", None)
+            if isinstance(subfolder, list):
+                subfolder = subfolder[i]
+
+            if not isinstance(pretrained_model_name_or_path_or_dict, dict):
+                pretrained_model_name_or_path_or_dict, _ = self.lora_state_dict(
+                    pretrained_model_name_or_path_or_dict,
+                    subfolder=subfolder,
+                    **_kwargs
+                )
+            
+            if adapter_name_ori is not None:
+                # only load lora of the input adapter name, then break the loop
+                i = unet.extra_condition_names.index(adapter_name_ori)
+                adapter_name = adapter_name_ori
+            
+            unet_conv_in_lora_A_name, old_weight = ([
+                (k, v) 
+                for k, v in pretrained_model_name_or_path_or_dict.items() 
+                if "unet." in k and ".conv_in." in k and ".lora_A." in k
+            ] + [(None, None)])[0]
+            if unet_conv_in_lora_A_name is not None:
+                in_weight = old_weight[:,:in_channels]
+                cond_weight = old_weight[:,in_channels:]
+                zero_weight = torch.zeros_like(in_weight)
+                new_weight = torch.cat(
+                    [in_weight] + 
+                    [zero_weight] * i + 
+                    [cond_weight] + 
+                    [zero_weight] * (num_condition_names - i - 1),
+                    dim=1
+                )
+                pretrained_model_name_or_path_or_dict[unet_conv_in_lora_A_name] = new_weight
+
+            super().load_lora_weights(pretrained_model_name_or_path_or_dict, adapter_name, **_kwargs)
+
+            if adapter_name_ori is not None:
+                break
+        
+        unet.activate_extra_condition_adapters()
+
+    @torch.no_grad()
+    @replace_example_docstring(EXAMPLE_DOC_STRING)
+    def __call__(
+        self,
+        prompt: Union[str, List[str]] = None,
+        image: PipelineImageInput = None,
+        height: Optional[int] = None,
+        width: Optional[int] = None,
+        num_inference_steps: int = 50,
+        timesteps: List[int] = None,
+        sigmas: List[float] = None,
+        guidance_scale: float = 7.5,
+        negative_prompt: Optional[Union[str, List[str]]] = None,
+        num_images_per_prompt: Optional[int] = 1,
+        eta: float = 0.0,
+        generator: Optional[Union[torch.Generator, List[torch.Generator]]] = None,
+        latents: Optional[torch.Tensor] = None,
+        prompt_embeds: Optional[torch.Tensor] = None,
+        negative_prompt_embeds: Optional[torch.Tensor] = None,
+        ip_adapter_image: Optional[PipelineImageInput] = None,
+        ip_adapter_image_embeds: Optional[List[torch.Tensor]] = None,
+        output_type: Optional[str] = "pil",
+        return_dict: bool = True,
+        cross_attention_kwargs: Optional[Dict[str, Any]] = None,
+        extra_condition_scale: Union[float, List[float]] = 1.0,
+        control_guidance_start: Union[float, List[float]] = 0.0,
+        control_guidance_end: Union[float, List[float]] = 1.0,
+        clip_skip: Optional[int] = None,
+        callback_on_step_end: Optional[
+            Union[Callable[[int, int, Dict], None], PipelineCallback, MultiPipelineCallbacks]
+        ] = None,
+        callback_on_step_end_tensor_inputs: List[str] = ["latents"],
+        **kwargs,
+    ):
+        r"""
+        The call function to the pipeline for generation.
+
+        Args:
+            prompt (`str` or `List[str]`, *optional*):
+                The prompt or prompts to guide image generation. If not defined, you need to pass `prompt_embeds`.
+            image (`torch.Tensor`, `PIL.Image.Image`, `np.ndarray`, `List[torch.Tensor]`, `List[PIL.Image.Image]`, `List[np.ndarray]`,:
+                    `List[List[torch.Tensor]]`, `List[List[np.ndarray]]` or `List[List[PIL.Image.Image]]`):
+                The extra input condition to provide guidance to the `unet` for generation after encoded by `vae`. If the type is
+                specified as `torch.Tensor`, its `vae` latent representation is passed to UNet. `PIL.Image.Image` can also be accepted
+                as an image. The dimensions of the output image defaults to `image`'s dimensions. If height and/or
+                width are passed, `image` is resized accordingly. If multiple extra conditions are specified in `unet`,
+                images must be passed as a list such that each element of the list can be correctly batched for input
+                to `unet`. When `prompt` is a list, and if a list of images is passed for `unet`, each will be paired with each prompt
+                in the `prompt` list. This also applies to multiple extra conditions, where a list of image lists can be
+                passed to batch for each prompt and each extra condition.
+            height (`int`, *optional*, defaults to `self.unet.config.sample_size * self.vae_scale_factor`):
+                The height in pixels of the generated image.
+            width (`int`, *optional*, defaults to `self.unet.config.sample_size * self.vae_scale_factor`):
+                The width in pixels of the generated image.
+            num_inference_steps (`int`, *optional*, defaults to 50):
+                The number of denoising steps. More denoising steps usually lead to a higher quality image at the
+                expense of slower inference.
+            timesteps (`List[int]`, *optional*):
+                Custom timesteps to use for the denoising process with schedulers which support a `timesteps` argument
+                in their `set_timesteps` method. If not defined, the default behavior when `num_inference_steps` is
+                passed will be used. Must be in descending order.
+            sigmas (`List[float]`, *optional*):
+                Custom sigmas to use for the denoising process with schedulers which support a `sigmas` argument in
+                their `set_timesteps` method. If not defined, the default behavior when `num_inference_steps` is passed
+                will be used.
+            guidance_scale (`float`, *optional*, defaults to 7.5):
+                A higher guidance scale value encourages the model to generate images closely linked to the text
+                `prompt` at the expense of lower image quality. Guidance scale is enabled when `guidance_scale > 1`.
+            negative_prompt (`str` or `List[str]`, *optional*):
+                The prompt or prompts to guide what to not include in image generation. If not defined, you need to
+                pass `negative_prompt_embeds` instead. Ignored when not using guidance (`guidance_scale < 1`).
+            num_images_per_prompt (`int`, *optional*, defaults to 1):
+                The number of images to generate per prompt.
+            eta (`float`, *optional*, defaults to 0.0):
+                Corresponds to parameter eta (η) from the [DDIM](https://arxiv.org/abs/2010.02502) paper. Only applies
+                to the [`~schedulers.DDIMScheduler`], and is ignored in other schedulers.
+            generator (`torch.Generator` or `List[torch.Generator]`, *optional*):
+                A [`torch.Generator`](https://pytorch.org/docs/stable/generated/torch.Generator.html) to make
+                generation deterministic.
+            latents (`torch.Tensor`, *optional*):
+                Pre-generated noisy latents sampled from a Gaussian distribution, to be used as inputs for image
+                generation. Can be used to tweak the same generation with different prompts. If not provided, a latents
+                tensor is generated by sampling using the supplied random `generator`.
+            prompt_embeds (`torch.Tensor`, *optional*):
+                Pre-generated text embeddings. Can be used to easily tweak text inputs (prompt weighting). If not
+                provided, text embeddings are generated from the `prompt` input argument.
+            negative_prompt_embeds (`torch.Tensor`, *optional*):
+                Pre-generated negative text embeddings. Can be used to easily tweak text inputs (prompt weighting). If
+                not provided, `negative_prompt_embeds` are generated from the `negative_prompt` input argument.
+            ip_adapter_image: (`PipelineImageInput`, *optional*): Optional image input to work with IP Adapters.
+            ip_adapter_image_embeds (`List[torch.Tensor]`, *optional*):
+                Pre-generated image embeddings for IP-Adapter. It should be a list of length same as number of
+                IP-adapters. Each element should be a tensor of shape `(batch_size, num_images, emb_dim)`. It should
+                contain the negative image embedding if `do_classifier_free_guidance` is set to `True`. If not
+                provided, embeddings are computed from the `ip_adapter_image` input argument.
+            output_type (`str`, *optional*, defaults to `"pil"`):
+                The output format of the generated image. Choose between `PIL.Image` or `np.array`.
+            return_dict (`bool`, *optional*, defaults to `True`):
+                Whether or not to return a [`~pipelines.stable_diffusion.StableDiffusionPipelineOutput`] instead of a
+                plain tuple.
+            callback (`Callable`, *optional*):
+                A function that calls every `callback_steps` steps during inference. The function is called with the
+                following arguments: `callback(step: int, timestep: int, latents: torch.Tensor)`.
+            callback_steps (`int`, *optional*, defaults to 1):
+                The frequency at which the `callback` function is called. If not specified, the callback is called at
+                every step.
+            cross_attention_kwargs (`dict`, *optional*):
+                A kwargs dictionary that if specified is passed along to the [`AttentionProcessor`] as defined in
+                [`self.processor`](https://github.com/huggingface/diffusers/blob/main/src/diffusers/models/attention_processor.py).
+            extra_condition_scale (`float` or `List[float]`, *optional*, defaults to 1.0):
+                The control lora scale of `unet`. If multiple extra conditions are specified in `unet`, you can set
+                the corresponding scale as a list.
+            control_guidance_start (`float` or `List[float]`, *optional*, defaults to 0.0):
+                The percentage of total steps at which the extra condtion starts applying.
+            control_guidance_end (`float` or `List[float]`, *optional*, defaults to 1.0):
+                The percentage of total steps at which the extra condtion stops applying.
+            clip_skip (`int`, *optional*):
+                Number of layers to be skipped from CLIP while computing the prompt embeddings. A value of 1 means that
+                the output of the pre-final layer will be used for computing the prompt embeddings.
+            callback_on_step_end (`Callable`, `PipelineCallback`, `MultiPipelineCallbacks`, *optional*):
+                A function or a subclass of `PipelineCallback` or `MultiPipelineCallbacks` that is called at the end of
+                each denoising step during the inference. with the following arguments: `callback_on_step_end(self:
+                DiffusionPipeline, step: int, timestep: int, callback_kwargs: Dict)`. `callback_kwargs` will include a
+                list of all tensors as specified by `callback_on_step_end_tensor_inputs`.
+            callback_on_step_end_tensor_inputs (`List`, *optional*):
+                The list of tensor inputs for the `callback_on_step_end` function. The tensors specified in the list
+                will be passed as `callback_kwargs` argument. You will only be able to include variables listed in the
+                `._callback_tensor_inputs` attribute of your pipeline class.
+
+        Examples:
+
+        Returns:
+            [`~pipelines.stable_diffusion.StableDiffusionPipelineOutput`] or `tuple`:
+                If `return_dict` is `True`, [`~pipelines.stable_diffusion.StableDiffusionPipelineOutput`] is returned,
+                otherwise a `tuple` is returned where the first element is a list with the generated images and the
+                second element is a list of `bool`s indicating whether the corresponding generated image contains
+                "not-safe-for-work" (nsfw) content.
+        """
+
+        callback = kwargs.pop("callback", None)
+        callback_steps = kwargs.pop("callback_steps", None)
+
+        if callback is not None:
+            deprecate(
+                "callback",
+                "1.0.0",
+                "Passing `callback` as an input argument to `__call__` is deprecated, consider using `callback_on_step_end`",
+            )
+        if callback_steps is not None:
+            deprecate(
+                "callback_steps",
+                "1.0.0",
+                "Passing `callback_steps` as an input argument to `__call__` is deprecated, consider using `callback_on_step_end`",
+            )
+
+        if isinstance(callback_on_step_end, (PipelineCallback, MultiPipelineCallbacks)):
+            callback_on_step_end_tensor_inputs = callback_on_step_end.tensor_inputs
+
+        unet: UNet2DConditionModelEx = self.unet._orig_mod if is_compiled_module(self.unet) else self.unet
+        num_extra_conditions = len(unet.extra_condition_names)
+
+        # align format for control guidance
+        if not isinstance(control_guidance_start, list) and isinstance(control_guidance_end, list):
+            control_guidance_start = len(control_guidance_end) * [control_guidance_start]
+        elif not isinstance(control_guidance_end, list) and isinstance(control_guidance_start, list):
+            control_guidance_end = len(control_guidance_start) * [control_guidance_end]
+        elif not isinstance(control_guidance_start, list) and not isinstance(control_guidance_end, list):
+            mult = num_extra_conditions
+            control_guidance_start, control_guidance_end = (
+                mult * [control_guidance_start],
+                mult * [control_guidance_end],
+            )
+
+        # 1. Check inputs. Raise error if not correct
+        self.check_inputs(
+            prompt,
+            image,
+            callback_steps,
+            negative_prompt,
+            prompt_embeds,
+            negative_prompt_embeds,
+            ip_adapter_image,
+            ip_adapter_image_embeds,
+            extra_condition_scale,
+            control_guidance_start,
+            control_guidance_end,
+            callback_on_step_end_tensor_inputs,
+        )
+
+        self._guidance_scale = guidance_scale
+        self._clip_skip = clip_skip
+        self._cross_attention_kwargs = cross_attention_kwargs
+
+        # 2. Define call parameters
+        if prompt is not None and isinstance(prompt, str):
+            batch_size = 1
+        elif prompt is not None and isinstance(prompt, list):
+            batch_size = len(prompt)
+        else:
+            batch_size = prompt_embeds.shape[0]
+
+        device = self._execution_device
+
+        if num_extra_conditions > 1 and isinstance(extra_condition_scale, float):
+            extra_condition_scale = [extra_condition_scale] * num_extra_conditions
+
+        # 3. Encode input prompt
+        text_encoder_lora_scale = (
+            self.cross_attention_kwargs.get("scale", None) if self.cross_attention_kwargs is not None else None
+        )
+        prompt_embeds, negative_prompt_embeds = self.encode_prompt(
+            prompt,
+            device,
+            num_images_per_prompt,
+            self.do_classifier_free_guidance,
+            negative_prompt,
+            prompt_embeds=prompt_embeds,
+            negative_prompt_embeds=negative_prompt_embeds,
+            lora_scale=text_encoder_lora_scale,
+            clip_skip=self.clip_skip,
+        )
+        # For classifier free guidance, we need to do two forward passes.
+        # Here we concatenate the unconditional and text embeddings into a single batch
+        # to avoid doing two forward passes
+        if self.do_classifier_free_guidance:
+            prompt_embeds = torch.cat([negative_prompt_embeds, prompt_embeds])
+
+        if ip_adapter_image is not None or ip_adapter_image_embeds is not None:
+            image_embeds = self.prepare_ip_adapter_image_embeds(
+                ip_adapter_image,
+                ip_adapter_image_embeds,
+                device,
+                batch_size * num_images_per_prompt,
+                self.do_classifier_free_guidance,
+            )
+
+        # 4. Prepare image
+        if num_extra_conditions == 1:
+            image = self.prepare_image(
+                image=image,
+                width=width,
+                height=height,
+                batch_size=batch_size * num_images_per_prompt,
+                num_images_per_prompt=num_images_per_prompt,
+                device=device,
+                dtype=unet.dtype,
+                do_classifier_free_guidance=self.do_classifier_free_guidance,
+            )
+            height, width = image.shape[-2:]
+            image = (
+                self.vae.encode(image.to(dtype=unet.dtype)).latent_dist.mode() * self.vae.config.scaling_factor
+            )
+        elif num_extra_conditions >= 1:
+            images = []
+
+            # Nested lists as extra condition
+            if isinstance(image[0], list):
+                # Transpose the nested image list
+                image = [list(t) for t in zip(*image)]
+
+            for image_ in image:
+                image_ = self.prepare_image(
+                    image=image_,
+                    width=width,
+                    height=height,
+                    batch_size=batch_size * num_images_per_prompt,
+                    num_images_per_prompt=num_images_per_prompt,
+                    device=device,
+                    dtype=unet.dtype,
+                    do_classifier_free_guidance=self.do_classifier_free_guidance,
+                )
+
+                images.append(image_)
+
+            image = images
+            height, width = image[0].shape[-2:]
+            image = [
+                self.vae.encode(image.to(dtype=unet.dtype)).latent_dist.mode() * self.vae.config.scaling_factor 
+                for image in images
+            ]
+        else:
+            assert False
+
+        # 5. Prepare timesteps
+        timesteps, num_inference_steps = retrieve_timesteps(
+            self.scheduler, num_inference_steps, device, timesteps, sigmas
+        )
+        self._num_timesteps = len(timesteps)
+
+        # 6. Prepare latent variables
+        num_channels_latents = self.unet.config.in_channels
+        latents = self.prepare_latents(
+            batch_size * num_images_per_prompt,
+            num_channels_latents,
+            height,
+            width,
+            prompt_embeds.dtype,
+            device,
+            generator,
+            latents,
+        )
+
+        # 6.5 Optionally get Guidance Scale Embedding
+        timestep_cond = None
+        if self.unet.config.time_cond_proj_dim is not None:
+            guidance_scale_tensor = torch.tensor(self.guidance_scale - 1).repeat(batch_size * num_images_per_prompt)
+            timestep_cond = self.get_guidance_scale_embedding(
+                guidance_scale_tensor, embedding_dim=self.unet.config.time_cond_proj_dim
+            ).to(device=device, dtype=latents.dtype)
+
+        # 7. Prepare extra step kwargs. TODO: Logic should ideally just be moved out of the pipeline
+        extra_step_kwargs = self.prepare_extra_step_kwargs(generator, eta)
+
+        # 7.1 Add image embeds for IP-Adapter
+        added_cond_kwargs = (
+            {"image_embeds": image_embeds}
+            if ip_adapter_image is not None or ip_adapter_image_embeds is not None
+            else None
+        )
+
+        # 7.2 Create tensor stating which extra_conditions to keep
+        extra_condition_keep = []
+        for i in range(len(timesteps)):
+            keeps = [
+                1.0 - float(i / len(timesteps) < s or (i + 1) / len(timesteps) > e)
+                for s, e in zip(control_guidance_start, control_guidance_end)
+            ]
+            extra_condition_keep.append(keeps[0] if num_extra_conditions == 1 else keeps)
+
+        # 8. Denoising loop
+        num_warmup_steps = len(timesteps) - num_inference_steps * self.scheduler.order
+        is_unet_compiled = is_compiled_module(self.unet)
+        is_torch_higher_equal_2_1 = is_torch_version(">=", "2.1")
+        with self.progress_bar(total=num_inference_steps) as progress_bar:
+            for i, t in enumerate(timesteps):
+                # Relevant thread:
+                # https://dev-discuss.pytorch.org/t/cudagraphs-in-pytorch-2-0/1428
+                if is_unet_compiled and is_torch_higher_equal_2_1:
+                    torch._inductor.cudagraph_mark_step_begin()
+                # expand the latents if we are doing classifier free guidance
+                latent_model_input = torch.cat([latents] * 2) if self.do_classifier_free_guidance else latents
+                latent_model_input = self.scheduler.scale_model_input(latent_model_input, t)
+
+                if isinstance(extra_condition_keep[i], list):
+                    cond_scale = [c * s for c, s in zip(extra_condition_scale, extra_condition_keep[i])]
+                else:
+                    extra_cond_scale = extra_condition_scale
+                    if isinstance(extra_cond_scale, list):
+                        extra_cond_scale = extra_cond_scale[0]
+                    cond_scale = extra_cond_scale * extra_condition_keep[i]
+
+                self.unet.set_extra_condition_scale(cond_scale)
+
+                # predict the noise residual
+                noise_pred = self.unet(
+                    latent_model_input,
+                    t,
+                    encoder_hidden_states=prompt_embeds,
+                    timestep_cond=timestep_cond,
+                    cross_attention_kwargs=self.cross_attention_kwargs,
+                    added_cond_kwargs=added_cond_kwargs,
+                    extra_conditions=image,
+                    return_dict=False,
+                )[0]
+
+                # perform guidance
+                if self.do_classifier_free_guidance:
+                    noise_pred_uncond, noise_pred_text = noise_pred.chunk(2)
+                    noise_pred = noise_pred_uncond + self.guidance_scale * (noise_pred_text - noise_pred_uncond)
+
+                # compute the previous noisy sample x_t -> x_t-1
+                latents = self.scheduler.step(noise_pred, t, latents, **extra_step_kwargs, return_dict=False)[0]
+
+                if callback_on_step_end is not None:
+                    callback_kwargs = {}
+                    for k in callback_on_step_end_tensor_inputs:
+                        callback_kwargs[k] = locals()[k]
+                    callback_outputs = callback_on_step_end(self, i, t, callback_kwargs)
+
+                    latents = callback_outputs.pop("latents", latents)
+                    prompt_embeds = callback_outputs.pop("prompt_embeds", prompt_embeds)
+                    negative_prompt_embeds = callback_outputs.pop("negative_prompt_embeds", negative_prompt_embeds)
+
+                # call the callback, if provided
+                if i == len(timesteps) - 1 or ((i + 1) > num_warmup_steps and (i + 1) % self.scheduler.order == 0):
+                    progress_bar.update()
+                    if callback is not None and i % callback_steps == 0:
+                        step_idx = i // getattr(self.scheduler, "order", 1)
+                        callback(step_idx, t, latents)
+
+        self.unet.set_extra_condition_scale(1.0)
+
+        # If we do sequential model offloading, let's offload unet
+        # manually for max memory savings
+        if hasattr(self, "final_offload_hook") and self.final_offload_hook is not None:
+            self.unet.to("cpu")
+            torch.cuda.empty_cache()
+        
+        if not output_type == "latent":
+            image = self.vae.decode(latents / self.vae.config.scaling_factor, return_dict=False, generator=generator)[
+                0
+            ]
+            image, has_nsfw_concept = self.run_safety_checker(image, device, prompt_embeds.dtype)
+        else:
+            image = latents
+            has_nsfw_concept = None
+
+        if has_nsfw_concept is None:
+            do_denormalize = [True] * image.shape[0]
+        else:
+            do_denormalize = [not has_nsfw for has_nsfw in has_nsfw_concept]
+
+        image = self.image_processor.postprocess(image, output_type=output_type, do_denormalize=do_denormalize)
+
+        # Offload all models
+        self.maybe_free_model_hooks()
+
+        if not return_dict:
+            return (image, has_nsfw_concept)
+
+        return StableDiffusionPipelineOutput(images=image, nsfw_content_detected=has_nsfw_concept)

requirements.txt

@@ -1,6 +1,558 @@
-accelerate
-diffusers
-invisible_watermark
-torch
-transformers
-xformers
+absl-py==1.4.0
+accelerate==1.2.1
+aiofiles==23.2.1
+aiohappyeyeballs==2.4.4
+aiohttp==3.11.11
+aiosignal==1.3.2
+alabaster==1.0.0
+albucore==0.0.19
+albumentations==1.4.20
+altair==5.5.0
+annotated-types==0.7.0
+anyio==3.7.1
+argon2-cffi==23.1.0
+argon2-cffi-bindings==21.2.0
+array_record==0.6.0
+arviz==0.20.0
+astropy==6.1.7
+astropy-iers-data==0.2025.1.6.0.33.42
+astunparse==1.6.3
+async-timeout==4.0.3
+atpublic==4.1.0
+attrs==24.3.0
+audioread==3.0.1
+autograd==1.7.0
+babel==2.16.0
+backcall==0.2.0
+beautifulsoup4==4.12.3
+bigframes==1.31.0
+bigquery-magics==0.5.0
+bleach==6.2.0
+blinker==1.9.0
+blis==0.7.11
+blosc2==2.7.1
+bokeh==3.6.2
+Bottleneck==1.4.2
+bqplot==0.12.44
+branca==0.8.1
+CacheControl==0.14.2
+cachetools==5.5.0
+catalogue==2.0.10
+certifi==2024.12.14
+cffi==1.17.1
+chardet==5.2.0
+charset-normalizer==3.4.1
+chex==0.1.88
+clarabel==0.9.0
+click==8.1.8
+cloudpathlib==0.20.0
+cloudpickle==3.1.0
+cmake==3.31.2
+cmdstanpy==1.2.5
+colorcet==3.1.0
+colorlover==0.3.0
+colour==0.1.5
+community==1.0.0b1
+confection==0.1.5
+cons==0.4.6
+contourpy==1.3.1
+cryptography==43.0.3
+cuda-python==12.2.1
+cudf-cu12 @ https://pypi.nvidia.com/cudf-cu12/cudf_cu12-24.10.1-cp310-cp310-manylinux_2_24_x86_64.manylinux_2_28_x86_64.whl
+cufflinks==0.17.3
+cupy-cuda12x==12.2.0
+cvxopt==1.3.2
+cvxpy==1.6.0
+cycler==0.12.1
+cymem==2.0.10
+Cython==3.0.11
+dask==2024.10.0
+datascience==0.17.6
+db-dtypes==1.3.1
+dbus-python==1.2.18
+debugpy==1.8.0
+decorator==4.4.2
+defusedxml==0.7.1
+Deprecated==1.2.15
+diffusers==0.32.1
+discord.py==2.4.0
+distro==1.9.0
+dlib==19.24.2
+dm-tree==0.1.8
+docker-pycreds==0.4.0
+docstring_parser==0.16
+docutils==0.21.2
+dopamine_rl==4.1.0
+duckdb==1.1.3
+earthengine-api==1.4.4
+easydict==1.13
+editdistance==0.8.1
+eerepr==0.0.4
+einops==0.8.0
+en-core-web-sm @ https://github.com/explosion/spacy-models/releases/download/en_core_web_sm-3.7.1/en_core_web_sm-3.7.1-py3-none-any.whl#sha256=86cc141f63942d4b2c5fcee06630fd6f904788d2f0ab005cce45aadb8fb73889
+entrypoints==0.4
+et_xmlfile==2.0.0
+etils==1.11.0
+etuples==0.3.9
+eval_type_backport==0.2.2
+exceptiongroup==1.2.2
+fastai==2.7.18
+fastapi==0.115.6
+fastcore==1.7.28
+fastdownload==0.0.7
+fastjsonschema==2.21.1
+fastprogress==1.0.3
+fastrlock==0.8.3
+ffmpy==0.5.0
+filelock==3.16.1
+firebase-admin==6.6.0
+Flask==3.1.0
+flatbuffers==24.12.23
+flax==0.10.2
+folium==0.19.4
+fonttools==4.55.3
+frozendict==2.4.6
+frozenlist==1.5.0
+fsspec==2024.10.0
+future==1.0.0
+gast==0.6.0
+gcsfs==2024.10.0
+GDAL==3.6.4
+gdown==5.2.0
+geemap==0.35.1
+gensim==4.3.3
+geocoder==1.38.1
+geographiclib==2.0
+geopandas==1.0.1
+geopy==2.4.1
+gin-config==0.5.0
+gitdb==4.0.12
+GitPython==3.1.44
+glob2==0.7
+google==2.0.3
+google-ai-generativelanguage==0.6.10
+google-api-core==2.19.2
+google-api-python-client==2.155.0
+google-auth==2.27.0
+google-auth-httplib2==0.2.0
+google-auth-oauthlib==1.2.1
+google-cloud-aiplatform==1.74.0
+google-cloud-bigquery==3.25.0
+google-cloud-bigquery-connection==1.17.0
+google-cloud-bigquery-storage==2.27.0
+google-cloud-bigtable==2.27.0
+google-cloud-core==2.4.1
+google-cloud-datastore==2.20.2
+google-cloud-firestore==2.19.0
+google-cloud-functions==1.19.0
+google-cloud-iam==2.17.0
+google-cloud-language==2.16.0
+google-cloud-pubsub==2.27.2
+google-cloud-resource-manager==1.14.0
+google-cloud-storage==2.19.0
+google-cloud-translate==3.19.0
+google-colab @ file:///colabtools/dist/google_colab-1.0.0.tar.gz
+google-crc32c==1.6.0
+google-genai==0.3.0
+google-generativeai==0.8.3
+google-pasta==0.2.0
+google-resumable-media==2.7.2
+googleapis-common-protos==1.66.0
+googledrivedownloader==0.4
+gradio==5.12.0
+gradio_client==1.5.4
+graphviz==0.20.3
+greenlet==3.1.1
+grpc-google-iam-v1==0.14.0
+grpcio==1.69.0
+grpcio-status==1.62.3
+gspread==6.1.4
+gspread-dataframe==4.0.0
+gym==0.25.2
+gym-notices==0.0.8
+h11==0.14.0
+h5netcdf==1.4.1
+h5py==3.12.1
+holidays==0.64
+holoviews==1.20.0
+html5lib==1.1
+httpcore==1.0.7
+httpimport==1.4.0
+httplib2==0.22.0
+httpx==0.28.1
+huggingface-hub==0.27.1
+humanize==4.11.0
+hyperopt==0.2.7
+ibis-framework==9.2.0
+idna==3.10
+imageio==2.36.1
+imageio-ffmpeg==0.5.1
+imagesize==1.4.1
+imbalanced-learn==0.13.0
+imgaug==0.4.0
+immutabledict==4.2.1
+importlib_metadata==8.5.0
+importlib_resources==6.5.2
+imutils==0.5.4
+inflect==7.5.0
+iniconfig==2.0.0
+intel-cmplr-lib-ur==2025.0.4
+intel-openmp==2025.0.4
+ipyevents==2.0.2
+ipyfilechooser==0.6.0
+ipykernel==5.5.6
+ipyleaflet==0.19.2
+ipyparallel==8.8.0
+ipython==7.34.0
+ipython-genutils==0.2.0
+ipython-sql==0.5.0
+ipytree==0.2.2
+ipywidgets==7.7.1
+itsdangerous==2.2.0
+jax==0.4.33
+jax-cuda12-pjrt==0.4.33
+jax-cuda12-plugin==0.4.33
+jaxlib==0.4.33
+jeepney==0.7.1
+jellyfish==1.1.0
+jieba==0.42.1
+Jinja2==3.1.5
+jiter==0.8.2
+joblib==1.4.2
+jsonpatch==1.33
+jsonpickle==4.0.1
+jsonpointer==3.0.0
+jsonschema==4.23.0
+jsonschema-specifications==2024.10.1
+jupyter-client==6.1.12
+jupyter-console==6.1.0
+jupyter-leaflet==0.19.2
+jupyter-server==1.24.0
+jupyter_core==5.7.2
+jupyterlab_pygments==0.3.0
+jupyterlab_widgets==3.0.13
+kaggle==1.6.17
+kagglehub==0.3.6
+keras==3.5.0
+keyring==23.5.0
+kiwisolver==1.4.8
+langchain==0.3.14
+langchain-core==0.3.29
+langchain-text-splitters==0.3.5
+langcodes==3.5.0
+langsmith==0.2.10
+language_data==1.3.0
+launchpadlib==1.10.16
+lazr.restfulclient==0.14.4
+lazr.uri==1.0.6
+lazy_loader==0.4
+libclang==18.1.1
+libcudf-cu12 @ https://pypi.nvidia.com/libcudf-cu12/libcudf_cu12-24.10.1-py3-none-manylinux_2_28_x86_64.whl
+librosa==0.10.2.post1
+lightgbm==4.5.0
+linkify-it-py==2.0.3
+llvmlite==0.43.0
+locket==1.0.0
+logical-unification==0.4.6
+lxml==5.3.0
+marisa-trie==1.2.1
+Markdown==3.7
+markdown-it-py==3.0.0
+MarkupSafe==2.1.5
+matplotlib==3.10.0
+matplotlib-inline==0.1.7
+matplotlib-venn==1.1.1
+mdit-py-plugins==0.4.2
+mdurl==0.1.2
+miniKanren==1.0.3
+missingno==0.5.2
+mistune==3.1.0
+mizani==0.13.1
+mkl==2025.0.1
+ml-dtypes==0.4.1
+mlxtend==0.23.3
+more-itertools==10.5.0
+moviepy==1.0.3
+mpmath==1.3.0
+msgpack==1.1.0
+multidict==6.1.0
+multipledispatch==1.0.0
+multitasking==0.0.11
+murmurhash==1.0.11
+music21==9.3.0
+namex==0.0.8
+narwhals==1.21.1
+natsort==8.4.0
+nbclassic==1.1.0
+nbclient==0.10.2
+nbconvert==7.16.5
+nbformat==5.10.4
+ndindex==1.9.2
+nest-asyncio==1.6.0
+networkx==3.4.2
+nibabel==5.3.2
+nltk==3.9.1
+notebook==6.5.5
+notebook_shim==0.2.4
+numba==0.60.0
+numexpr==2.10.2
+numpy==1.26.4
+nvidia-cublas-cu12==12.6.4.1
+nvidia-cuda-cupti-cu12==12.6.80
+nvidia-cuda-nvcc-cu12==12.6.85
+nvidia-cuda-runtime-cu12==12.6.77
+nvidia-cudnn-cu12==9.6.0.74
+nvidia-cufft-cu12==11.3.0.4
+nvidia-curand-cu12==10.3.7.77
+nvidia-cusolver-cu12==11.7.1.2
+nvidia-cusparse-cu12==12.5.4.2
+nvidia-nccl-cu12==2.24.3
+nvidia-nvjitlink-cu12==12.6.85
+nvtx==0.2.10
+nx-cugraph-cu12 @ https://pypi.nvidia.com/nx-cugraph-cu12/nx_cugraph_cu12-24.10.0-py3-none-any.whl
+oauth2client==4.1.3
+oauthlib==3.2.2
+openai==1.59.4
+opencv-contrib-python==4.10.0.84
+opencv-python==4.10.0.84
+opencv-python-headless==4.10.0.84
+openpyxl==3.1.5
+opentelemetry-api==1.29.0
+opentelemetry-sdk==1.29.0
+opentelemetry-semantic-conventions==0.50b0
+opt_einsum==3.4.0
+optax==0.2.4
+optree==0.13.1
+orbax-checkpoint==0.6.4
+orjson==3.10.13
+osqp==0.6.7.post3
+packaging==24.2
+pandas==2.2.2
+pandas-datareader==0.10.0
+pandas-gbq==0.26.1
+pandas-stubs==2.2.2.240909
+pandocfilters==1.5.1
+panel==1.5.5
+param==2.2.0
+parso==0.8.4
+parsy==2.1
+partd==1.4.2
+pathlib==1.0.1
+patsy==1.0.1
+peewee==3.17.8
+peft==0.14.0
+pexpect==4.9.0
+pickleshare==0.7.5
+pillow==11.1.0
+platformdirs==4.3.6
+plotly==5.24.1
+plotnine==0.14.5
+pluggy==1.5.0
+ply==3.11
+polars==1.9.0
+pooch==1.8.2
+portpicker==1.5.2
+preshed==3.0.9
+prettytable==3.12.0
+proglog==0.1.10
+progressbar2==4.5.0
+prometheus_client==0.21.1
+promise==2.3
+prompt_toolkit==3.0.48
+propcache==0.2.1
+prophet==1.1.6
+proto-plus==1.25.0
+protobuf==4.25.5
+psutil==5.9.5
+psycopg2==2.9.10
+ptyprocess==0.7.0
+py-cpuinfo==9.0.0
+py4j==0.10.9.7
+pyarrow==17.0.0
+pyasn1==0.6.1
+pyasn1_modules==0.4.1
+pycocotools==2.0.8
+pycparser==2.22
+pydantic==2.10.4
+pydantic_core==2.27.2
+pydata-google-auth==1.9.0
+pydot==3.0.4
+pydotplus==2.0.2
+PyDrive==1.3.1
+PyDrive2==1.21.3
+pydub==0.25.1
+pyerfa==2.0.1.5
+pygame==2.6.1
+pygit2==1.16.0
+Pygments==2.18.0
+PyGObject==3.42.1
+PyJWT==2.10.1
+pylibcudf-cu12 @ https://pypi.nvidia.com/pylibcudf-cu12/pylibcudf_cu12-24.10.1-cp310-cp310-manylinux_2_24_x86_64.manylinux_2_28_x86_64.whl
+pylibcugraph-cu12==24.10.0
+pylibraft-cu12==24.10.0
+pymc==5.19.1
+pymystem3==0.2.0
+pynvjitlink-cu12==0.4.0
+pyogrio==0.10.0
+Pyomo==6.8.2
+PyOpenGL==3.1.7
+pyOpenSSL==24.2.1
+pyparsing==3.2.1
+pyperclip==1.9.0
+pyproj==3.7.0
+pyshp==2.3.1
+PySocks==1.7.1
+pyspark==3.5.4
+pytensor==2.26.4
+pytest==8.3.4
+python-apt==0.0.0
+python-box==7.3.0
+python-dateutil==2.8.2
+python-dotenv==1.0.1
+python-louvain==0.16
+python-multipart==0.0.20
+python-slugify==8.0.4
+python-utils==3.9.1
+pytz==2024.2
+pyviz_comms==3.0.3
+PyYAML==6.0.2
+pyzmq==24.0.1
+qdldl==0.1.7.post5
+ratelim==0.1.6
+referencing==0.35.1
+regex==2024.11.6
+requests==2.32.3
+requests-oauthlib==1.3.1
+requests-toolbelt==1.0.0
+requirements-parser==0.9.0
+rich==13.9.4
+rmm-cu12==24.10.0
+rpds-py==0.22.3
+rpy2==3.4.2
+rsa==4.9
+ruff==0.9.1
+safehttpx==0.1.6
+safetensors==0.5.1
+scikit-image==0.25.0
+scikit-learn==1.6.0
+scipy==1.13.1
+scooby==0.10.0
+scs==3.2.7.post2
+seaborn==0.13.2
+SecretStorage==3.3.1
+semantic-version==2.10.0
+Send2Trash==1.8.3
+sentence-transformers==3.3.1
+sentencepiece==0.2.0
+sentry-sdk==2.19.2
+setproctitle==1.3.4
+shap==0.46.0
+shapely==2.0.6
+shellingham==1.5.4
+simple-parsing==0.1.6
+six==1.17.0
+sklearn-compat==0.1.3
+sklearn-pandas==2.2.0
+slicer==0.0.8
+smart-open==7.1.0
+smmap==5.0.2
+sniffio==1.3.1
+snowballstemmer==2.2.0
+soundfile==0.13.0
+soupsieve==2.6
+soxr==0.5.0.post1
+spacy==3.7.5
+spacy-legacy==3.0.12
+spacy-loggers==1.0.5
+Sphinx==8.1.3
+sphinxcontrib-applehelp==2.0.0
+sphinxcontrib-devhelp==2.0.0
+sphinxcontrib-htmlhelp==2.1.0
+sphinxcontrib-jsmath==1.0.1
+sphinxcontrib-qthelp==2.0.0
+sphinxcontrib-serializinghtml==2.0.0
+SQLAlchemy==2.0.36
+sqlglot==25.1.0
+sqlparse==0.5.3
+srsly==2.5.0
+stanio==0.5.1
+starlette==0.41.3
+statsmodels==0.14.4
+stringzilla==3.11.3
+sympy==1.13.1
+tables==3.10.1
+tabulate==0.9.0
+tbb==2022.0.0
+tcmlib==1.2.0
+tenacity==9.0.0
+tensorboard==2.17.1
+tensorboard-data-server==0.7.2
+tensorflow==2.17.1
+tensorflow-datasets==4.9.7
+tensorflow-hub==0.16.1
+tensorflow-io-gcs-filesystem==0.37.1
+tensorflow-metadata==1.13.1
+tensorflow-probability==0.24.0
+tensorstore==0.1.71
+termcolor==2.5.0
+terminado==0.18.1
+text-unidecode==1.3
+textblob==0.17.1
+tf-slim==1.1.0
+tf_keras==2.17.0
+thinc==8.2.5
+threadpoolctl==3.5.0
+tifffile==2024.12.12
+timm==1.0.12
+tinycss2==1.4.0
+tokenizers==0.21.0
+toml==0.10.2
+tomli==2.2.1
+tomlkit==0.13.2
+toolz==0.12.1
+torch @ https://download.pytorch.org/whl/cu121_full/torch-2.5.1%2Bcu121-cp310-cp310-linux_x86_64.whl
+torchaudio @ https://download.pytorch.org/whl/cu121/torchaudio-2.5.1%2Bcu121-cp310-cp310-linux_x86_64.whl
+torchsummary==1.5.1
+torchvision @ https://download.pytorch.org/whl/cu121/torchvision-0.20.1%2Bcu121-cp310-cp310-linux_x86_64.whl
+tornado==6.3.3
+tqdm==4.67.1
+traitlets==5.7.1
+traittypes==0.2.1
+transformers==4.47.1
+tweepy==4.14.0
+typeguard==4.4.1
+typer==0.15.1
+types-pytz==2024.2.0.20241221
+types-setuptools==75.6.0.20241223
+typing_extensions==4.12.2
+tzdata==2024.2
+tzlocal==5.2
+uc-micro-py==1.0.3
+umf==0.9.1
+uritemplate==4.1.1
+urllib3==2.3.0
+uvicorn==0.34.0
+vega-datasets==0.9.0
+wadllib==1.3.6
+wandb==0.19.1
+wasabi==1.1.3
+wcwidth==0.2.13
+weasel==0.4.1
+webcolors==24.11.1
+webencodings==0.5.1
+websocket-client==1.8.0
+websockets==14.1
+Werkzeug==3.1.3
+widgetsnbextension==3.6.10
+wordcloud==1.9.4
+wrapt==1.17.0
+xarray==2025.1.0
+xarray-einstats==0.8.0
+xformers==0.0.29.post1
+xgboost==2.1.3
+xlrd==2.0.1
+xyzservices==2024.9.0
+yarl==1.18.3
+yellowbrick==1.5
+yfinance==0.2.51
+zipp==3.21.0