Citlali_wan_2_1_14_B_text2video_lora

Citlali Text-to-Video Generation

This repository contains the necessary steps and scripts to generate videos using the Citlali text-to-video model. The model leverages LoRA (Low-Rank Adaptation) weights and pre-trained components to create high-quality anime-style videos based on textual prompts.

Prerequisites

Before proceeding, ensure that you have the following installed on your system:

• Ubuntu (or a compatible Linux distribution) • Python 3.x • pip (Python package manager) • Git • Git LFS (Git Large File Storage) • FFmpeg

Installation

1. Update and Install Dependencies

   sudo apt-get update && sudo apt-get install cbm git-lfs ffmpeg
   

2. Clone the Repository

   git clone https://huggingface.co/svjack/Citlali_wan_2_1_14_B_text2video_lora
   cd Citlali_wan_2_1_14_B_text2video_lora
   

3. Install Python Dependencies

   pip install torch torchvision
   pip install -r requirements.txt
   pip install ascii-magic matplotlib tensorboard huggingface_hub datasets
   pip install moviepy==1.0.3
   pip install sageattention==1.0.6
   

4. Download Model Weights

   wget https://huggingface.co/Wan-AI/Wan2.1-T2V-14B/resolve/main/models_t5_umt5-xxl-enc-bf16.pth
   wget https://huggingface.co/DeepBeepMeep/Wan2.1/resolve/main/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth
   wget https://huggingface.co/Wan-AI/Wan2.1-T2V-14B/resolve/main/Wan2.1_VAE.pth
   wget https://huggingface.co/Comfy-Org/Wan_2.1_ComfyUI_repackaged/resolve/main/split_files/diffusion_models/wan2.1_t2v_1.3B_bf16.safetensors
   wget https://huggingface.co/Comfy-Org/Wan_2.1_ComfyUI_repackaged/resolve/main/split_files/diffusion_models/wan2.1_t2v_14B_bf16.safetensors
   

Usage

To generate a video, use the wan_generate_video.py script with the appropriate parameters. Below are examples of how to generate videos using the Citlali model.

14B usage

python wan_generate_video.py --fp8 --task t2v-14B --video_size 480 832 --infer_steps 35 --video_length 25 \
--save_path save --output_type video \
--dit wan2.1_t2v_14B_bf16.safetensors --vae Wan2.1_VAE.pth \
--t5 models_t5_umt5-xxl-enc-bf16.pth \
--attn_mode torch \
--lora_weight genshin_Citlali_w14_outputs/genshin_Citlali_w14_im_lora-step00007500.safetensors \
--lora_multiplier 1.0 \
--interactive 

"Melodic Solitude"

In the style of Citlali, this is a digital anime-style illustration of a young woman with long, pastel purple hair and blue eyes, wearing a dark blue, sleeveless, high-collared outfit with gold and purple accents. She has large, circular, teal headphones on her head. She sits cross-legged on a sunlit park bench, fingers tapping rhythmically on her knees as she loses herself in the music, her hair gently swaying in the breeze.

<video controls autoplay src="https://cdn-uploads.huggingface.co/production/uploads/634dffc49b777beec3bc6448/QoZhCy0_3lA5sxEh-pxgj.mp4"></video>

"Starlit Serenade"

In the style of Citlali, this is a digital anime-style illustration of a young woman with long, pastel purple hair and blue eyes, wearing a dark blue, sleeveless, high-collared outfit with gold and purple accents. She has large, circular, teal headphones on her head. She lies on a grassy hill under a starry sky, arms behind her head, eyes closed as the music transports her to another world, her hair fanned out around her like a lavender halo.

<video controls autoplay src="https://cdn-uploads.huggingface.co/production/uploads/634dffc49b777beec3bc6448/q9_I-ZjSvPvC9CVLzOfyu.mp4"></video>

Rainy Reverie

In the style of Citlali, this is a digital anime-style illustration of a young woman with long, pastel purple hair and blue eyes, wearing a dark blue, sleeveless, high-collared outfit with gold and purple accents. She has large, circular, teal headphones on her head. She stands by a rain-streaked window, fingertips pressed lightly against the glass, lost in thought as the pitter-patter of raindrops syncs with her playlist.

<video controls autoplay src="https://cdn-uploads.huggingface.co/production/uploads/634dffc49b777beec3bc6448/OfQKPH3NYWb8e1rtBblhv.mp4"></video>

Café Harmony

In the style of Citlali, this is a digital anime-style illustration of a young woman with long, pastel purple hair and blue eyes, wearing a dark blue, sleeveless, high-collared outfit with gold and purple accents. She has large, circular, teal headphones on her head. She leans over a small café table, sketching in a notebook while sipping a latte, her headphones slightly tilted as she occasionally nods to the music.

<video controls autoplay src="https://cdn-uploads.huggingface.co/production/uploads/634dffc49b777beec3bc6448/84UYZb6itAOQHa_wouHmN.mp4"></video>

Urban Groove

In the style of Citlali, this is a digital anime-style illustration of a young woman with long, pastel purple hair and blue eyes, wearing a dark blue, sleeveless, high-collared outfit with gold and purple accents. She has large, circular, teal headphones on her head. She walks down a neon-lit city street at night, her boots clicking against the pavement while she hums along to the beat, her headphones glowing softly in the dark.

• use wan 14b t2v <video controls autoplay src="https://cdn-uploads.huggingface.co/production/uploads/634dffc49b777beec3bc6448/oGIT8rlgd_nGE_v3jyB5r.mp4"></video>

• use wan fusionX 14b <video controls autoplay src="https://cdn-uploads.huggingface.co/production/uploads/634dffc49b777beec3bc6448/RexXpXTOhAo_WYOu4EF6-.mp4"></video>

Parameters

• --fp8: Enable FP8 precision (optional).
• --task: Specify the task (e.g., t2v-1.3B).
• --video_size: Set the resolution of the generated video (e.g., 1024 1024).
• --video_length: Define the length of the video in frames.
• --infer_steps: Number of inference steps.
• --save_path: Directory to save the generated video.
• --output_type: Output type (e.g., both for video and frames).
• --dit: Path to the diffusion model weights.
• --vae: Path to the VAE model weights.
• --t5: Path to the T5 model weights.
• --attn_mode: Attention mode (e.g., torch).
• --lora_weight: Path to the LoRA weights.
• --lora_multiplier: Multiplier for LoRA weights.
• --prompt: Textual prompt for video generation.

Output

The generated video and frames will be saved in the specified save_path directory.

Troubleshooting

• Ensure all dependencies are correctly installed. • Verify that the model weights are downloaded and placed in the correct locations. • Check for any missing Python packages and install them using pip.

License

This project is licensed under the MIT License. See the LICENSE file for details.

Acknowledgments

• Hugging Face for hosting the model weights. • Wan-AI for providing the pre-trained models. • DeepBeepMeep for contributing to the model weights.

Contact

For any questions or issues, please open an issue on the repository or contact the maintainer.

---