Step-by-Step Tutorial: Diffusion-Pipe WSL Linux Install & Hunyuan LoRA Training on Windows.

Step-by-Step Tutorial: Diffusion-Pipe WSL Linux Install & Hunyuan LoRA Training on Windows.

Commands list:

#1

wsl --list --verbose

#2

wsl --set-default-version 2

#3

wsl --list --online

#4

wsl --install -d Ubuntu-22.04

#5

sudo apt update && sudo apt upgrade -y

#6

sudo apt install -y build-essential curl wget git unzip git-lfs

#7

mkdir -p ~/miniconda3

wget https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh -O ~/miniconda3/miniconda.sh

bash ~/miniconda3/miniconda.sh -b -u -p ~/miniconda3

rm ~/miniconda3/miniconda.sh

#8

source ~/miniconda3/bin/activate

#9

conda init --all

#10

conda create -n diffusion-pipe python=3.12

#11

conda activate diffusion-pipe

#12

pip install torch==2.4.1 torchvision==0.19.1 torchaudio --extra-index-url https://download.pytorch.org/whl/cu124

#13

conda install nvidia::cuda-nvcc

#14

git clone --recurse-submodules https://github.com/tdrussell/diffusion-pipe

#15

cd diffusion-pipe

#16

pip install -r requirements.txt

#17

wget -P models/hunyuan https://huggingface.co/Kijai/HunyuanVideo_comfy/resolve/main/hunyuan_video_720_cfgdistill_fp8_e4m3fn.safetensors

#18

wget -P models/hunyuan https://huggingface.co/Kijai/HunyuanVideo_comfy/resolve/main/hunyuan_video_vae_bf16.safetensors

#19

mkdir -p models/clip && git clone https://huggingface.co/openai/clip-vit-large-patch14 models/clip/clip-vit-large-patch14

#20

mkdir -p models/llm && git clone https://huggingface.co/Kijai/llava-llama-3-8b-text-encoder-tokenizer models/llm/llava-llama-3-8b-text-encoder-tokenizer

#21

NCCL_P2P_DISABLE="1" NCCL_IB_DISABLE="1" deepspeed --num_gpus=1 train.py --deepspeed --config examples/hunyuan_video.toml

#22 : Resume Training.

"NCCL_P2P_DISABLE="1" NCCL_IB_DISABLE="1" deepspeed --num_gpus=1 train.py --deepspeed --config examples/hunyuan_video.toml --resume_from_checkpoint"

If you like my content consider to Support me on Patreon.

TRY: "Social Fashion" Lora

TRY: Flow for ComfyUI


How to install Flow: