No official torrent yet β requesting it tells the flock what to host next. You cannot seed a catalog-only page.
Cataloged from lightx2v/Hy1.5-Distill-Models
All models by lightx2v β
π€ HuggingFace | GitHub | License
This repository contains 4-step distilled models for HunyuanVideo-1.5 optimized for use with LightX2V. These distilled models enable ultra-fast 4-step inference without CFG (Classifier-Free Guidance), significantly reducing generation time while maintaining high-quality video output.
hy1.5_t2v_480p_lightx2v_4step.safetensors - 480p Text-to-Video 4-step distilled model (16.7 GB)hy1.5_t2v_480p_scaled_fp8_e4m3_lightx2v_4step.safetensors - 480p Text-to-Video 4-step distilled model with FP8 quantization (8.85 GB)First, install LightX2V:
Or build from source:
Download the distilled models from this repository:
For even lower memory usage, use the FP8 quantized version:
These models use step distillation technology to compress the original 50-step inference process into just 4 steps, providing:
guidance_scale=1 (no classifier-free guidance needed)The FP8 quantized version (hy1.5_t2v_480p_scaled_fp8_e4m3_lightx2v_4step.safetensors) provides additional benefits:
For FP8 quantized models, you need to install the SGL kernel:
Alternatively, you can use VLLM kernels:
Using 4-step distilled models provides:
Critical Configuration:
infer_steps=4 (not the default 50)guidance_scale=1 (CFG is not used in distilled models)denoising_step_list=[1000, 750, 500, 250]Model Loading: All advanced configurations (including enable_quantize() and enable_offload()) must be called before create_generator(), otherwise they will not take effect.
Original Model Required: The original HunyuanVideo-1.5 model weights are still required. The distilled model is used in conjunction with the original model structure.
Attention Mode: For best performance, we recommend using SageAttention 2 (sage_attn2) as the attention mode.
Resolution: Currently supports 480p resolution. Higher resolutions may be available in future releases.
If you use these distilled models in your research, please cite:
This model is released under the Apache 2.0 License, same as the original HunyuanVideo-1.5 model.
pip install -v git+https://github.com/ModelTC/LightX2V.gitgit clone https://github.com/ModelTC/LightX2V.gitcd LightX2Vpip install -v -e .# Using git-lfsgit lfs installgit clone https://huggingface.co/lightx2v/Hy1.5-Distill-Models
# Or download individual files using huggingface-hubpip install huggingface-hubpython -c "from huggingface_hub import hf_hub_download; hf_hub_download(repo_id='lightx2v/Hy1.5-Distill-Models', filename='hy1.5_t2v_480p_lightx2v_4step.safetensors', local_dir='./models')""""HunyuanVideo-1.5 text-to-video generation example.This example demonstrates how to use LightX2V with HunyuanVideo-1.5 4-step distilled model for T2V generation."""
from lightx2v import LightX2VPipeline
# Initialize pipeline for HunyuanVideo-1.5pipe = LightX2VPipeline( model_path="/path/to/hunyuanvideo-1.5/", # Original model path model_cls="hunyuan_video_1.5", transformer_model_name="480p_t2v", task="t2v", # 4-step distilled model ckpt dit_original_ckpt="/path/to/hy1.5_t2v_480p_lightx2v_4step.safetensors")
# Alternative: create generator from config JSON file# pipe.create_generator(config_json="../configs/hunyuan_video_15/hunyuan_video_t2v_480p.json")
# Enable offloading to significantly reduce VRAM usage with minimal speed impact# Suitable for RTX 30/40/50 consumer GPUspipe.enable_offload( cpu_offload=True, offload_granularity="block", # For HunyuanVideo-1.5, only "block" is supported text_encoder_offload=True, image_encoder_offload=False, vae_offload=False,)
# Optional: Use lighttae# pipe.enable_lightvae(# use_tae=True,# tae_path="/path/to/lighttaehy1_5.safetensors",# use_lightvae=False,# vae_path=None,# )
# Create generator with specified parameters# Note: 4-step distillation requires infer_steps=4, guidance_scale=1, and denoising_step_listpipe.create_generator( attn_mode="sage_attn2", infer_steps=4, # 4-step inference num_frames=81, guidance_scale=1, # No CFG needed for distilled models sample_shift=9.0, aspect_ratio="16:9", fps=16, denoising_step_list=[1000, 750, 500, 250] # Required for 4-step distillation)
# Generation parametersseed = 123prompt = "A close-up shot captures a scene on a polished, light-colored granite kitchen counter, illuminated by soft natural light from an unseen window. Initially, the frame focuses on a tall, clear glass filled with golden, translucent apple juice standing next to a single, shiny red apple with a green leaf still attached to its stem. The camera moves horizontally to the right. As the shot progresses, a white ceramic plate smoothly enters the frame, revealing a fresh arrangement of about seven or eight more apples, a mix of vibrant reds and greens, piled neatly upon it. A shallow depth of field keeps the focus sharply on the fruit and glass, while the kitchen backsplash in the background remains softly blurred. The scene is in a realistic style."negative_prompt = ""save_result_path = "/path/to/save_results/output.mp4"
# Generate videopipe.generate( seed=seed, prompt=prompt, negative_prompt=negative_prompt, save_result_path=save_result_path,)from lightx2v import LightX2VPipeline
# Initialize pipelinepipe = LightX2VPipeline( model_path="/path/to/hunyuanvideo-1.5/", # Original model path model_cls="hunyuan_video_1.5", transformer_model_name="480p_t2v", task="t2v", # 4-step distilled model ckpt dit_original_ckpt="/path/to/hy1.5_t2v_480p_lightx2v_4step.safetensors")
# Enable FP8 quantization for the distilled modelpipe.enable_quantize( quant_scheme='fp8-sgl', dit_quantized=True, dit_quantized_ckpt="/path/to/hy1.5_t2v_480p_scaled_fp8_e4m3_lightx2v_4step.safetensors", text_encoder_quantized=False, # Optional: can also quantize text encoder text_encoder_quantized_ckpt="/path/to/hy15_qwen25vl_llm_encoder_fp8_e4m3_lightx2v.safetensors", # Optional image_encoder_quantized=False,)
# Enable offloading for lower VRAM usagepipe.enable_offload( cpu_offload=True, offload_granularity="block", text_encoder_offload=True, image_encoder_offload=False, vae_offload=False,)
# Create generatorpipe.create_generator( attn_mode="sage_attn2", infer_steps=4, num_frames=81, guidance_scale=1, sample_shift=9.0, aspect_ratio="16:9", fps=16, denoising_step_list=[1000, 750, 500, 250])
# Generate videopipe.generate( seed=123, prompt="Your prompt here", negative_prompt="", save_result_path="/path/to/output.mp4",)# Requires torch == 2.8.0pip install sgl-kernel --upgradepip install vllm@misc{lightx2v, author = {LightX2V Contributors}, title = {LightX2V: Light Video Generation Inference Framework}, year = {2025}, publisher = {GitHub}, journal = {GitHub repository}, howpublished = {\url{https://github.com/ModelTC/lightx2v}},}