LTX-Video

LTX-Video, metin ve görüntülerden yüksek çözünürlüklü videoların hızlı ve gerçek zamanlı olarak üretilmesi için tasarlanmış bir difüzyon dönüştürücüsüdür. LTX-Video'nun ana özelliği Video-VAE'dir. Video-VAE, daha yüksek piksel-temsil sıkıştırma oranına (1:192) sahiptir; bu da daha verimli video veri işleme ve daha hızlı üretim hızı sağlar. Üretim sırasında ince detayların kaybolmasını desteklemek ve önlemek için Video-VAE çözücü, temsil-piksel dönüşümünü ve son gürültü azaltma adımını gerçekleştirir.

Tüm orijinal LTX-Video checkpoint'lerini Lightricks kuruluşu altında bulabilirsiniz.

Diğer video oluşturma görevlerine ait daha fazla örnek için sağ kenar çubuğundaki LTX-Video modellerine tıklayın.

Aşağıdaki örnek, bellek veya çıkarım hızı için optimize edilmiş bir video üretmenin nasıl yapılacağını gösterir.

memory

çalıştırma hızı

Notlar

LTX-Video deposundan üretim için aşağıdaki önerilen ayarlara başvurun.
- Dönüştürücü, VAE ve metin kodlayıcı için önerilen dtype torch.bfloat16'dır. VAE ve metin kodlayıcı ayrıca torch.float32 veya torch.float16 olabilir.
- LTX-Video'nun guidance-distilled varyantları için guidance_scale değerini 1.0 olarak ayarlayın. Diğer herhangi bir model için iyi bir üretim kalitesi elde etmek amacıyla guidance_scale değeri, 5.0 gibi daha yüksek bir değere ayarlanmalıdır.
- Zaman adımı farkındalıklı VAE varyantları (LTX-Video 0.9.1 ve üzeri) için decode_timestep değerini 0.05 ve image_cond_noise_scale değerini 0.025 olarak ayarlayın.
- Birden fazla koşullandırma görseli ve videosu arasında interpolasyonu destekleyen varyantlar (LTX-Video 0.9.5 ve üzeri) için en iyi sonuçları almak benzer görseller ve videolar kullanın. Koşullandırma girdilerinden sapma, üretilen videoda ani geçişlere yol açabilir.

LTX-Video 0.9.7, uzamsal latent bir yükseltici ve 13B parametreli bir dönüştürücü içerir. Çıkarım sırasında, düşük çözünürlüklü bir video önce hızla oluşturulur ve ardından yükseltilip iyileştirilir.

Örnek kodu göster

import torch
from diffusers import LTXConditionPipeline, LTXLatentUpsamplePipeline
from diffusers.pipelines.ltx.pipeline_ltx_condition import LTXVideoCondition
from diffusers.utils import export_to_video, load_video

pipeline = LTXConditionPipeline.from_pretrained("Lightricks/LTX-Video-0.9.7-dev", dtype=torch.bfloat16)
pipeline_upsample = LTXLatentUpsamplePipeline.from_pretrained("Lightricks/ltxv-spatial-upscaler-0.9.7", vae=pipeline.vae, dtype=torch.bfloat16)
pipeline.to("cuda")  # or "mps", "xpu", "cpu"
pipe_upsample.to("cuda")
pipeline.vae.enable_tiling()

def round_to_nearest_resolution_acceptable_by_vae(height, width):
    height = height - (height % pipeline.vae_temporal_compression_ratio)
    width = width - (width % pipeline.vae_temporal_compression_ratio)
    return height, width

video = load_video(
    "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/cosmos/cosmos-video2world-input-vid.mp4"
)[:21]  # only use the first 21 frames as conditioning
condition1 = LTXVideoCondition(video=video, frame_index=0)

prompt = """
The video depicts a winding mountain road covered in snow, with a single vehicle 
traveling along it. The road is flanked by steep, rocky cliffs and sparse vegetation. 
The landscape is characterized by rugged terrain and a river visible in the distance. 
The scene captures the solitude and beauty of a winter drive through a mountainous region.
"""
negative_prompt = "worst quality, inconsistent motion, blurry, jittery, distorted"
expected_height, expected_width = 768, 1152
downscale_factor = 2 / 3
num_frames = 161

# 1. Generate video at smaller resolution
# Text-only conditioning is also supported without the need to pass `conditions`
downscaled_height, downscaled_width = int(expected_height * downscale_factor), int(expected_width * downscale_factor)
downscaled_height, downscaled_width = round_to_nearest_resolution_acceptable_by_vae(downscaled_height, downscaled_width)
latents = pipeline(
    conditions=[condition1],
    prompt=prompt,
    negative_prompt=negative_prompt,
    width=downscaled_width,
    height=downscaled_height,
    num_frames=num_frames,
    num_inference_steps=30,
    decode_timestep=0.05,
    decode_noise_scale=0.025,
    image_cond_noise_scale=0.0,
    guidance_scale=5.0,
    guidance_rescale=0.7,
    generator=torch.Generator().manual_seed(0),
    output_type="latent",
).frames

# 2. Upscale generated video using latent upsampler with fewer inference steps
# The available latent upsampler upscales the height/width by 2x
upscaled_height, upscaled_width = downscaled_height * 2, downscaled_width * 2
upscaled_latents = pipe_upsample(
    latents=latents,
    output_type="latent"
).frames

# 3. Denoise the upscaled video with few steps to improve texture (optional, but recommended)
video = pipeline(
    conditions=[condition1],
    prompt=prompt,
    negative_prompt=negative_prompt,
    width=upscaled_width,
    height=upscaled_height,
    num_frames=num_frames,
    denoise_strength=0.4,  # Effectively, 4 inference steps out of 10
    num_inference_steps=10,
    latents=upscaled_latents,
    decode_timestep=0.05,
    decode_noise_scale=0.025,
    image_cond_noise_scale=0.0,
    guidance_scale=5.0,
    guidance_rescale=0.7,
    generator=torch.Generator().manual_seed(0),
    output_type="pil",
).frames[0]

# 4. Downscale the video to the expected resolution
video = [frame.resize((expected_width, expected_height)) for frame in video]

export_to_video(video, "output.mp4", fps=24)

LTX-Video 0.9.7 damıtılmış modeli, üretimi hızlandırmak için rehberlik ve zaman adımı damıtması uygulanmıştır. İyi üretim kalitesi için guidance_scale'ın 1.0 olarak ayarlanması ve num_inference_steps'in 4 ile 10 arasında bir değere ayarlanması gerekir. En iyi sonuçlar için aşağıdaki özel zaman adımlarını da kullanmalısınız.

Upscaling için hazırlık amacıyla temel model çıkarımı: [1000, 993, 987, 981, 975, 909, 725, 0.03].
Yaklaştırma: [1000, 909, 725, 421, 0].

Örnek kodu göster

import torch
from diffusers import LTXConditionPipeline, LTXLatentUpsamplePipeline
from diffusers.pipelines.ltx.pipeline_ltx_condition import LTXVideoCondition
from diffusers.utils import export_to_video, load_video

pipeline = LTXConditionPipeline.from_pretrained("Lightricks/LTX-Video-0.9.7-distilled", dtype=torch.bfloat16)
pipe_upsample = LTXLatentUpsamplePipeline.from_pretrained("Lightricks/ltxv-spatial-upscaler-0.9.7", vae=pipeline.vae, dtype=torch.bfloat16)
pipeline.to("cuda")  # or "mps", "xpu", "cpu"
pipe_upsample.to("cuda")
pipeline.vae.enable_tiling()

def round_to_nearest_resolution_acceptable_by_vae(height, width):
    height = height - (height % pipeline.vae_spatial_compression_ratio)
    width = width - (width % pipeline.vae_spatial_compression_ratio)
    return height, width

prompt = """
artistic anatomical 3d render, utlra quality, human half full male body with transparent 
skin revealing structure instead of organs, muscular, intricate creative patterns, 
monochromatic with backlighting, lightning mesh, scientific concept art, blending biology 
with botany, surreal and ethereal quality, unreal engine 5, ray tracing, ultra realistic, 
16K UHD, rich details. camera zooms out in a rotating fashion
"""
negative_prompt = "worst quality, inconsistent motion, blurry, jittery, distorted"
expected_height, expected_width = 768, 1152
downscale_factor = 2 / 3
num_frames = 161

# 1. Generate video at smaller resolution
downscaled_height, downscaled_width = int(expected_height * downscale_factor), int(expected_width * downscale_factor)
downscaled_height, downscaled_width = round_to_nearest_resolution_acceptable_by_vae(downscaled_height, downscaled_width)
latents = pipeline(
    prompt=prompt,
    negative_prompt=negative_prompt,
    width=downscaled_width,
    height=downscaled_height,
    num_frames=num_frames,
    timesteps=[1000, 993, 987, 981, 975, 909, 725, 0.03],
    decode_timestep=0.05,
    decode_noise_scale=0.025,
    image_cond_noise_scale=0.0,
    guidance_scale=1.0,
    guidance_rescale=0.7,
    generator=torch.Generator().manual_seed(0),
    output_type="latent",
).frames

# 2. Upscale generated video using latent upsampler with fewer inference steps
# The available latent upsampler upscales the height/width by 2x
upscaled_height, upscaled_width = downscaled_height * 2, downscaled_width * 2
upscaled_latents = pipe_upsample(
    latents=latents,
    adain_factor=1.0,
    output_type="latent"
).frames

# 3. Denoise the upscaled video with few steps to improve texture (optional, but recommended)
video = pipeline(
    prompt=prompt,
    negative_prompt=negative_prompt,
    width=upscaled_width,
    height=upscaled_height,
    num_frames=num_frames,
    denoise_strength=0.999,  # Effectively, 4 inference steps out of 5
    timesteps=[1000, 909, 725, 421, 0],
    latents=upscaled_latents,
    decode_timestep=0.05,
    decode_noise_scale=0.025,
    image_cond_noise_scale=0.0,
    guidance_scale=1.0,
    guidance_rescale=0.7,
    generator=torch.Generator().manual_seed(0),
    output_type="pil",
).frames[0]

# 4. Downscale the video to the expected resolution
video = [frame.resize((expected_width, expected_height)) for frame in video]

export_to_video(video, "output.mp4", fps=24)

LTX-Video 0.9.8 damıtılmış modeli, 0.9.7 varyantına benzerdir. Rehberlik ve zaman adımı damıtılmıştır ve yukarıdaki gibi benzer çıkarım kodu kullanılabilir. Bu sürümün bir iyileştirmesi, çok uzun videoların oluşturulmasını desteklemesidir. Ayrıca, tone_map_compression_ratio parametresi kullanılarak oluşturulan videonun kalitesini artırmak için ton eşlemesinin (tone mapping) kullanılmasını destekler. 0.6 varsayılan değeri önerilmektedir.

Örnek kodu göster

import torch
from diffusers import LTXConditionPipeline, LTXLatentUpsamplePipeline
from diffusers.pipelines.ltx.pipeline_ltx_condition import LTXVideoCondition
from diffusers.models.latent_upscaler import LTXLatentUpsamplerModel
from diffusers.utils import export_to_video, load_video

pipeline = LTXConditionPipeline.from_pretrained("Lightricks/LTX-Video-0.9.8-13B-distilled", dtype=torch.bfloat16)
# TODO: Update the checkpoint here once updated in LTX org
upsampler = LTXLatentUpsamplerModel.from_pretrained("a-r-r-o-w/LTX-0.9.8-Latent-Upsampler", dtype=torch.bfloat16)
pipe_upsample = LTXLatentUpsamplePipeline(vae=pipeline.vae, latent_upsampler=upsampler).to(torch.bfloat16)
pipeline.to("cuda")  # or "mps", "xpu", "cpu"
pipe_upsample.to("cuda")
pipeline.vae.enable_tiling()

def round_to_nearest_resolution_acceptable_by_vae(height, width):
    height = height - (height % pipeline.vae_spatial_compression_ratio)
    width = width - (width % pipeline.vae_spatial_compression_ratio)
    return height, width

prompt = """The camera pans over a snow-covered mountain range, revealing a vast expanse of snow-capped peaks and valleys.The mountains are covered in a thick layer of snow, with some areas appearing almost white while others have a slightly darker, almost grayish hue. The peaks are jagged and irregular, with some rising sharply into the sky while others are more rounded. The valleys are deep and narrow, with steep slopes that are also covered in snow. The trees in the foreground are mostly bare, with only a few leaves remaining on their branches. The sky is overcast, with thick clouds obscuring the sun. The overall impression is one of peace and tranquility, with the snow-covered mountains standing as a testament to the power and beauty of nature."""
# prompt = """A woman walks away from a white Jeep parked on a city street at night, then ascends a staircase and knocks on a door. The woman, wearing a dark jacket and jeans, walks away from the Jeep parked on the left side of the street, her back to the camera; she walks at a steady pace, her arms swinging slightly by her sides; the street is dimly lit, with streetlights casting pools of light on the wet pavement; a man in a dark jacket and jeans walks past the Jeep in the opposite direction; the camera follows the woman from behind as she walks up a set of stairs towards a building with a green door; she reaches the top of the stairs and turns left, continuing to walk towards the building; she reaches the door and knocks on it with her right hand; the camera remains stationary, focused on the doorway; the scene is captured in real-life footage."""
negative_prompt = "bright colors, symbols, graffiti, watermarks, worst quality, inconsistent motion, blurry, jittery, distorted"
expected_height, expected_width = 480, 832
downscale_factor = 2 / 3
# num_frames = 161
num_frames = 361

# 1. Generate video at smaller resolution
downscaled_height, downscaled_width = int(expected_height * downscale_factor), int(expected_width * downscale_factor)
downscaled_height, downscaled_width = round_to_nearest_resolution_acceptable_by_vae(downscaled_height, downscaled_width)
latents = pipeline(
    prompt=prompt,
    negative_prompt=negative_prompt,
    width=downscaled_width,
    height=downscaled_height,
    num_frames=num_frames,
    timesteps=[1000, 993, 987, 981, 975, 909, 725, 0.03],
    decode_timestep=0.05,
    decode_noise_scale=0.025,
    image_cond_noise_scale=0.0,
    guidance_scale=1.0,
    guidance_rescale=0.7,
    generator=torch.Generator().manual_seed(0),
    output_type="latent",
).frames

# 2. Upscale generated video using latent upsampler with fewer inference steps
# The available latent upsampler upscales the height/width by 2x
upscaled_height, upscaled_width = downscaled_height * 2, downscaled_width * 2
upscaled_latents = pipe_upsample(
    latents=latents,
    adain_factor=1.0,
    tone_map_compression_ratio=0.6,
    output_type="latent"
).frames

# 3. Denoise the upscaled video with few steps to improve texture (optional, but recommended)
video = pipeline(
    prompt=prompt,
    negative_prompt=negative_prompt,
    width=upscaled_width,
    height=upscaled_height,
    num_frames=num_frames,
    denoise_strength=0.999,  # Effectively, 4 inference steps out of 5
    timesteps=[1000, 909, 725, 421, 0],
    latents=upscaled_latents,
    decode_timestep=0.05,
    decode_noise_scale=0.025,
    image_cond_noise_scale=0.0,
    guidance_scale=1.0,
    guidance_rescale=0.7,
    generator=torch.Generator().manual_seed(0),
    output_type="pil",
).frames[0]

# 4. Downscale the video to the expected resolution
video = [frame.resize((expected_width, expected_height)) for frame in video]

export_to_video(video, "output.mp4", fps=24)

LTX-Video, load_lora_weights() ile LoRA'ları destekler.

Örnek kodu göster

import torch
from diffusers import LTXConditionPipeline
from diffusers.utils import export_to_video, load_image

pipeline = LTXConditionPipeline.from_pretrained(
    "Lightricks/LTX-Video-0.9.5", dtype=torch.bfloat16
)

pipeline.load_lora_weights("Lightricks/LTX-Video-Cakeify-LoRA", adapter_name="cakeify")
pipeline.set_adapters("cakeify")

# use "CAKEIFY" to trigger the LoRA
prompt = "CAKEIFY a person using a knife to cut a cake shaped like a Pikachu plushie"
image = load_image("https://huggingface.co/Lightricks/LTX-Video-Cakeify-LoRA/resolve/main/assets/images/pikachu.png")

video = pipeline(
    prompt=prompt,
    image=image,
    width=576,
    height=576,
    num_frames=161,
    decode_timestep=0.03,
    decode_noise_scale=0.025,
    num_inference_steps=50,
).frames[0]
export_to_video(video, "output.mp4", fps=26)

LTX-Video, loaders.FromOriginalModelMixin.from_single_file() veya loaders.FromSingleFileMixin.from_single_file() ile GGUF checkpoint gibi tek dosyalardan yükleme destekler.

Örnek kodu göster

import torch
from diffusers.utils import export_to_video
from diffusers import LTXPipeline, AutoModel, GGUFQuantizationConfig

transformer = AutoModel.from_single_file(
    "https://huggingface.co/city96/LTX-Video-gguf/blob/main/ltx-video-2b-v0.9-Q3_K_S.gguf"
    quantization_config=GGUFQuantizationConfig(compute_dtype=torch.bfloat16),
    dtype=torch.bfloat16
)
pipeline = LTXPipeline.from_pretrained(
    "Lightricks/LTX-Video",
    transformer=transformer,
    dtype=torch.bfloat16
)

Diffusers

LTX-Video

Notlar

LTXI2VLongMultiPromptPipeline

sınıf diffusers.LTXI2VLongMultiPromptPipeline

__call__

encode_prompt

prepare_latents

vae_decode_tiled

LTXPipeline

sınıf diffusers.LTXPipeline

__call__

encode_prompt

LTXImageToVideoPipeline

sınıf diffusers.LTXImageToVideoPipeline

__call__

encode_prompt

LTXConditionPipeline

sınıf diffusers.LTXConditionPipeline

__call__

add_noise_to_image_conditioning_latents

encode_prompt

trim_conditioning_sequence

LTXLatentUpsamplePipeline

sınıf diffusers.LTXLatentUpsamplePipeline

__call__

adain_filter_latent

tone_map_latents

LTXPipelineOutput

sınıf diffusers.pipelines.ltx.pipeline_output.LTXPipelineOutput

call

call

call

call

call