From 586ac9d8a69353875c8d590427590308dba958d6 Mon Sep 17 00:00:00 2001 From: mi804 <1576993271@qq.com> Date: Wed, 25 Feb 2026 17:19:57 +0800 Subject: [PATCH 1/4] support ltx-2 training --- diffsynth/configs/model_configs.py | 56 +++++- diffsynth/diffusion/loss.py | 30 ++++ diffsynth/models/ltx2_audio_vae.py | 57 ++++++ diffsynth/models/ltx2_dit.py | 2 +- diffsynth/pipelines/ltx2_audio_video.py | 65 +++++-- .../LTX-2-I2AV-DistilledPipeline.py | 7 +- .../model_inference/LTX-2-I2AV-OneStage.py | 7 +- .../model_inference/LTX-2-I2AV-TwoStage.py | 7 +- .../LTX-2-T2AV-Camera-Control-Dolly-In.py | 6 +- .../LTX-2-T2AV-Camera-Control-Dolly-Left.py | 6 +- .../LTX-2-T2AV-Camera-Control-Dolly-Out.py | 6 +- .../LTX-2-T2AV-Camera-Control-Dolly-Right.py | 6 +- .../LTX-2-T2AV-Camera-Control-Jib-Down.py | 6 +- .../LTX-2-T2AV-Camera-Control-Jib-Up.py | 6 +- .../LTX-2-T2AV-Camera-Control-Static.py | 6 +- .../LTX-2-T2AV-DistilledPipeline.py | 6 +- .../model_inference/LTX-2-T2AV-OneStage.py | 15 +- .../model_inference/LTX-2-T2AV-TwoStage.py | 6 +- .../LTX-2-I2AV-DistilledPipeline.py | 7 +- .../LTX-2-I2AV-OneStage.py | 7 +- .../LTX-2-I2AV-TwoStage.py | 7 +- .../LTX-2-T2AV-Camera-Control-Dolly-In.py | 6 +- .../LTX-2-T2AV-Camera-Control-Dolly-Left.py | 6 +- .../LTX-2-T2AV-Camera-Control-Dolly-Out.py | 6 +- .../LTX-2-T2AV-Camera-Control-Dolly-Right.py | 6 +- .../LTX-2-T2AV-Camera-Control-Jib-Down.py | 6 +- .../LTX-2-T2AV-Camera-Control-Jib-Up.py | 6 +- .../LTX-2-T2AV-Camera-Control-Static.py | 6 +- .../LTX-2-T2AV-DistilledPipeline.py | 6 +- .../LTX-2-T2AV-OneStage.py | 6 +- .../LTX-2-T2AV-TwoStage.py | 6 +- .../model_training/full/LTX-2-T2AV-splited.sh | 35 ++++ .../model_training/lora/LTX-2-T2AV-noaudio.sh | 56 ++++++ .../model_training/lora/LTX-2-T2AV-splited.sh | 40 +++++ .../ltx2/model_training/lora/LTX-2-T2AV.sh | 19 ++ examples/ltx2/model_training/train.py | 162 ++++++++++++++++++ .../validate_full/LTX-2-T2AV.py | 47 +++++ .../validate_lora/LTX-2-T2AV.py | 49 ++++++ .../validate_lora/LTX-2-T2AV_noaudio.py | 49 ++++++ .../ltx2/scripts/split_model_statedicts.py | 104 +++++++++++ 40 files changed, 893 insertions(+), 49 deletions(-) create mode 100644 examples/ltx2/model_training/full/LTX-2-T2AV-splited.sh create mode 100644 examples/ltx2/model_training/lora/LTX-2-T2AV-noaudio.sh create mode 100644 examples/ltx2/model_training/lora/LTX-2-T2AV-splited.sh create mode 100644 examples/ltx2/model_training/lora/LTX-2-T2AV.sh create mode 100644 examples/ltx2/model_training/train.py create mode 100644 examples/ltx2/model_training/validate_full/LTX-2-T2AV.py create mode 100644 examples/ltx2/model_training/validate_lora/LTX-2-T2AV.py create mode 100644 examples/ltx2/model_training/validate_lora/LTX-2-T2AV_noaudio.py create mode 100644 examples/ltx2/scripts/split_model_statedicts.py diff --git a/diffsynth/configs/model_configs.py b/diffsynth/configs/model_configs.py index 9ff7ea6..dbad638 100644 --- a/diffsynth/configs/model_configs.py +++ b/diffsynth/configs/model_configs.py @@ -607,6 +607,12 @@ ltx2_series = [ "model_class": "diffsynth.models.ltx2_dit.LTXModel", "state_dict_converter": "diffsynth.utils.state_dict_converters.ltx2_dit.LTXModelStateDictConverter", }, + { + "model_hash": "c567aaa37d5ed7454c73aa6024458661", + "model_name": "ltx2_dit", + "model_class": "diffsynth.models.ltx2_dit.LTXModel", + "state_dict_converter": "diffsynth.utils.state_dict_converters.ltx2_dit.LTXModelStateDictConverter", + }, { # Example: ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-dev.safetensors") "model_hash": "aca7b0bbf8415e9c98360750268915fc", @@ -614,6 +620,12 @@ ltx2_series = [ "model_class": "diffsynth.models.ltx2_video_vae.LTX2VideoEncoder", "state_dict_converter": "diffsynth.utils.state_dict_converters.ltx2_video_vae.LTX2VideoEncoderStateDictConverter", }, + { + "model_hash": "7f7e904a53260ec0351b05f32153754b", + "model_name": "ltx2_video_vae_encoder", + "model_class": "diffsynth.models.ltx2_video_vae.LTX2VideoEncoder", + "state_dict_converter": "diffsynth.utils.state_dict_converters.ltx2_video_vae.LTX2VideoEncoderStateDictConverter", + }, { # Example: ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-dev.safetensors") "model_hash": "aca7b0bbf8415e9c98360750268915fc", @@ -621,6 +633,12 @@ ltx2_series = [ "model_class": "diffsynth.models.ltx2_video_vae.LTX2VideoDecoder", "state_dict_converter": "diffsynth.utils.state_dict_converters.ltx2_video_vae.LTX2VideoDecoderStateDictConverter", }, + { + "model_hash": "dc6029ca2825147872b45e35a2dc3a97", + "model_name": "ltx2_video_vae_decoder", + "model_class": "diffsynth.models.ltx2_video_vae.LTX2VideoDecoder", + "state_dict_converter": "diffsynth.utils.state_dict_converters.ltx2_video_vae.LTX2VideoDecoderStateDictConverter", + }, { # Example: ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-dev.safetensors") "model_hash": "aca7b0bbf8415e9c98360750268915fc", @@ -628,6 +646,12 @@ ltx2_series = [ "model_class": "diffsynth.models.ltx2_audio_vae.LTX2AudioDecoder", "state_dict_converter": "diffsynth.utils.state_dict_converters.ltx2_audio_vae.LTX2AudioDecoderStateDictConverter", }, + { + "model_hash": "7d7823dde8f1ea0b50fb07ac329dd4cb", + "model_name": "ltx2_audio_vae_decoder", + "model_class": "diffsynth.models.ltx2_audio_vae.LTX2AudioDecoder", + "state_dict_converter": "diffsynth.utils.state_dict_converters.ltx2_audio_vae.LTX2AudioDecoderStateDictConverter", + }, { # Example: ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-dev.safetensors") "model_hash": "aca7b0bbf8415e9c98360750268915fc", @@ -635,16 +659,34 @@ ltx2_series = [ "model_class": "diffsynth.models.ltx2_audio_vae.LTX2Vocoder", "state_dict_converter": "diffsynth.utils.state_dict_converters.ltx2_audio_vae.LTX2VocoderStateDictConverter", }, - # { # not used currently - # # Example: ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-dev.safetensors") - # "model_hash": "aca7b0bbf8415e9c98360750268915fc", - # "model_name": "ltx2_audio_vae_encoder", - # "model_class": "diffsynth.models.ltx2_audio_vae.LTX2AudioEncoder", - # "state_dict_converter": "diffsynth.utils.state_dict_converters.ltx2_audio_vae.LTX2AudioEncoderStateDictConverter", - # }, + { + "model_hash": "f471360f6b24bef702ab73133d9f8bb9", + "model_name": "ltx2_audio_vocoder", + "model_class": "diffsynth.models.ltx2_audio_vae.LTX2Vocoder", + "state_dict_converter": "diffsynth.utils.state_dict_converters.ltx2_audio_vae.LTX2VocoderStateDictConverter", + }, { # Example: ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-dev.safetensors") "model_hash": "aca7b0bbf8415e9c98360750268915fc", + "model_name": "ltx2_audio_vae_encoder", + "model_class": "diffsynth.models.ltx2_audio_vae.LTX2AudioEncoder", + "state_dict_converter": "diffsynth.utils.state_dict_converters.ltx2_audio_vae.LTX2AudioEncoderStateDictConverter", + }, + { + "model_hash": "29338f3b95e7e312a3460a482e4f4554", + "model_name": "ltx2_audio_vae_encoder", + "model_class": "diffsynth.models.ltx2_audio_vae.LTX2AudioEncoder", + "state_dict_converter": "diffsynth.utils.state_dict_converters.ltx2_audio_vae.LTX2AudioEncoderStateDictConverter", + }, + { + # Example: ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-dev.safetensors") + "model_hash": "aca7b0bbf8415e9c98360750268915fc", + "model_name": "ltx2_text_encoder_post_modules", + "model_class": "diffsynth.models.ltx2_text_encoder.LTX2TextEncoderPostModules", + "state_dict_converter": "diffsynth.utils.state_dict_converters.ltx2_text_encoder.LTX2TextEncoderPostModulesStateDictConverter", + }, + { + "model_hash": "981629689c8be92a712ab3c5eb4fc3f6", "model_name": "ltx2_text_encoder_post_modules", "model_class": "diffsynth.models.ltx2_text_encoder.LTX2TextEncoderPostModules", "state_dict_converter": "diffsynth.utils.state_dict_converters.ltx2_text_encoder.LTX2TextEncoderPostModulesStateDictConverter", diff --git a/diffsynth/diffusion/loss.py b/diffsynth/diffusion/loss.py index 14fdfd3..c9330e5 100644 --- a/diffsynth/diffusion/loss.py +++ b/diffsynth/diffusion/loss.py @@ -28,6 +28,36 @@ def FlowMatchSFTLoss(pipe: BasePipeline, **inputs): return loss +def FlowMatchSFTAudioVideoLoss(pipe: BasePipeline, **inputs): + max_timestep_boundary = int(inputs.get("max_timestep_boundary", 1) * len(pipe.scheduler.timesteps)) + min_timestep_boundary = int(inputs.get("min_timestep_boundary", 0) * len(pipe.scheduler.timesteps)) + + timestep_id = torch.randint(min_timestep_boundary, max_timestep_boundary, (1,)) + timestep = pipe.scheduler.timesteps[timestep_id].to(dtype=pipe.torch_dtype, device=pipe.device) + + # video + noise = torch.randn_like(inputs["input_latents"]) + inputs["video_latents"] = pipe.scheduler.add_noise(inputs["input_latents"], noise, timestep) + training_target = pipe.scheduler.training_target(inputs["input_latents"], noise, timestep) + + # audio + if inputs.get("audio_input_latents") is not None: + audio_noise = torch.randn_like(inputs["audio_input_latents"]) + inputs["audio_latents"] = pipe.scheduler.add_noise(inputs["audio_input_latents"], audio_noise, timestep) + training_target_audio = pipe.scheduler.training_target(inputs["audio_input_latents"], audio_noise, timestep) + + models = {name: getattr(pipe, name) for name in pipe.in_iteration_models} + noise_pred, noise_pred_audio = pipe.model_fn(**models, **inputs, timestep=timestep) + + loss = torch.nn.functional.mse_loss(noise_pred.float(), training_target.float()) + loss = loss * pipe.scheduler.training_weight(timestep) + if inputs.get("audio_input_latents") is not None: + loss_audio = torch.nn.functional.mse_loss(noise_pred_audio.float(), training_target_audio.float()) + loss_audio = loss_audio * pipe.scheduler.training_weight(timestep) + loss = loss + loss_audio + return loss + + def DirectDistillLoss(pipe: BasePipeline, **inputs): pipe.scheduler.set_timesteps(inputs["num_inference_steps"]) pipe.scheduler.training = True diff --git a/diffsynth/models/ltx2_audio_vae.py b/diffsynth/models/ltx2_audio_vae.py index 708ded7..0a12cb5 100644 --- a/diffsynth/models/ltx2_audio_vae.py +++ b/diffsynth/models/ltx2_audio_vae.py @@ -5,8 +5,65 @@ import einops import torch import torch.nn as nn import torch.nn.functional as F +import torchaudio from .ltx2_common import VideoLatentShape, AudioLatentShape, Patchifier, NormType, build_normalization_layer + +class AudioProcessor(nn.Module): + """Converts audio waveforms to log-mel spectrograms with optional resampling.""" + + def __init__( + self, + sample_rate: int = 16000, + mel_bins: int = 64, + mel_hop_length: int = 160, + n_fft: int = 1024, + ) -> None: + super().__init__() + self.sample_rate = sample_rate + self.mel_transform = torchaudio.transforms.MelSpectrogram( + sample_rate=sample_rate, + n_fft=n_fft, + win_length=n_fft, + hop_length=mel_hop_length, + f_min=0.0, + f_max=sample_rate / 2.0, + n_mels=mel_bins, + window_fn=torch.hann_window, + center=True, + pad_mode="reflect", + power=1.0, + mel_scale="slaney", + norm="slaney", + ) + + def resample_waveform( + self, + waveform: torch.Tensor, + source_rate: int, + target_rate: int, + ) -> torch.Tensor: + """Resample waveform to target sample rate if needed.""" + if source_rate == target_rate: + return waveform + resampled = torchaudio.functional.resample(waveform, source_rate, target_rate) + return resampled.to(device=waveform.device, dtype=waveform.dtype) + + def waveform_to_mel( + self, + waveform: torch.Tensor, + waveform_sample_rate: int, + ) -> torch.Tensor: + """Convert waveform to log-mel spectrogram [batch, channels, time, n_mels].""" + waveform = self.resample_waveform(waveform, waveform_sample_rate, self.sample_rate) + + mel = self.mel_transform(waveform) + mel = torch.log(torch.clamp(mel, min=1e-5)) + + mel = mel.to(device=waveform.device, dtype=waveform.dtype) + return mel.permute(0, 1, 3, 2).contiguous() + + class AudioPatchifier(Patchifier): def __init__( self, diff --git a/diffsynth/models/ltx2_dit.py b/diffsynth/models/ltx2_dit.py index 2e3c958..c12a3f3 100644 --- a/diffsynth/models/ltx2_dit.py +++ b/diffsynth/models/ltx2_dit.py @@ -1446,6 +1446,6 @@ class LTXModel(torch.nn.Module): cross_pe_max_pos = max(self.positional_embedding_max_pos[0], self.audio_positional_embedding_max_pos[0]) self._init_preprocessors(cross_pe_max_pos) video = Modality(video_latents, video_timesteps, video_positions, video_context) - audio = Modality(audio_latents, audio_timesteps, audio_positions, audio_context) + audio = Modality(audio_latents, audio_timesteps, audio_positions, audio_context) if audio_latents is not None else None vx, ax = self._forward(video=video, audio=audio, perturbations=None) return vx, ax diff --git a/diffsynth/pipelines/ltx2_audio_video.py b/diffsynth/pipelines/ltx2_audio_video.py index 9ed48aa..444e004 100644 --- a/diffsynth/pipelines/ltx2_audio_video.py +++ b/diffsynth/pipelines/ltx2_audio_video.py @@ -18,7 +18,7 @@ from ..diffusion.base_pipeline import BasePipeline, PipelineUnit from ..models.ltx2_text_encoder import LTX2TextEncoder, LTX2TextEncoderPostModules, LTXVGemmaTokenizer from ..models.ltx2_dit import LTXModel from ..models.ltx2_video_vae import LTX2VideoEncoder, LTX2VideoDecoder, VideoLatentPatchifier -from ..models.ltx2_audio_vae import LTX2AudioEncoder, LTX2AudioDecoder, LTX2Vocoder, AudioPatchifier +from ..models.ltx2_audio_vae import LTX2AudioEncoder, LTX2AudioDecoder, LTX2Vocoder, AudioPatchifier, AudioProcessor from ..models.ltx2_upsampler import LTX2LatentUpsampler from ..models.ltx2_common import VideoLatentShape, AudioLatentShape, VideoPixelShape, get_pixel_coords, VIDEO_SCALE_FACTORS from ..utils.data.media_io_ltx2 import ltx2_preprocess @@ -50,6 +50,7 @@ class LTX2AudioVideoPipeline(BasePipeline): self.video_patchifier: VideoLatentPatchifier = VideoLatentPatchifier(patch_size=1) self.audio_patchifier: AudioPatchifier = AudioPatchifier(patch_size=1) + self.audio_processor: AudioProcessor = AudioProcessor() self.in_iteration_models = ("dit",) self.units = [ @@ -57,6 +58,7 @@ class LTX2AudioVideoPipeline(BasePipeline): LTX2AudioVideoUnit_ShapeChecker(), LTX2AudioVideoUnit_PromptEmbedder(), LTX2AudioVideoUnit_NoiseInitializer(), + LTX2AudioVideoUnit_InputAudioEmbedder(), LTX2AudioVideoUnit_InputVideoEmbedder(), LTX2AudioVideoUnit_InputImagesEmbedder(), ] @@ -95,7 +97,7 @@ class LTX2AudioVideoPipeline(BasePipeline): stage2_lora_config.download_if_necessary() pipe.stage2_lora_path = stage2_lora_config.path # Optional, currently not used - # pipe.audio_vae_encoder = model_pool.fetch_model("ltx2_audio_vae_encoder") + pipe.audio_vae_encoder = model_pool.fetch_model("ltx2_audio_vae_encoder") # VRAM Management pipe.vram_management_enabled = pipe.check_vram_management_state() @@ -157,7 +159,6 @@ class LTX2AudioVideoPipeline(BasePipeline): num_frames=121, # Classifier-free guidance cfg_scale: Optional[float] = 3.0, - cfg_merge: Optional[bool] = False, # Scheduler num_inference_steps: Optional[int] = 40, # VAE tiling @@ -186,7 +187,7 @@ class LTX2AudioVideoPipeline(BasePipeline): "input_images": input_images, "input_images_indexes": input_images_indexes, "input_images_strength": input_images_strength, "seed": seed, "rand_device": rand_device, "height": height, "width": width, "num_frames": num_frames, - "cfg_scale": cfg_scale, "cfg_merge": cfg_merge, + "cfg_scale": cfg_scale, "tiled": tiled, "tile_size_in_pixels": tile_size_in_pixels, "tile_overlap_in_pixels": tile_overlap_in_pixels, "tile_size_in_frames": tile_size_in_frames, "tile_overlap_in_frames": tile_overlap_in_frames, "use_two_stage_pipeline": use_two_stage_pipeline, "use_distilled_pipeline": use_distilled_pipeline, @@ -422,7 +423,7 @@ class LTX2AudioVideoUnit_NoiseInitializer(PipelineUnit): def process_stage(self, pipe: LTX2AudioVideoPipeline, height, width, num_frames, seed, rand_device, frame_rate=24.0): video_pixel_shape = VideoPixelShape(batch=1, frames=num_frames, width=width, height=height, fps=frame_rate) - video_latent_shape = VideoLatentShape.from_pixel_shape(shape=video_pixel_shape, latent_channels=pipe.video_vae_encoder.latent_channels) + video_latent_shape = VideoLatentShape.from_pixel_shape(shape=video_pixel_shape, latent_channels=128) video_noise = pipe.generate_noise(video_latent_shape.to_torch_shape(), seed=seed, rand_device=rand_device) latent_coords = pipe.video_patchifier.get_patch_grid_bounds(output_shape=video_latent_shape, device=pipe.device) @@ -455,17 +456,48 @@ class LTX2AudioVideoUnit_NoiseInitializer(PipelineUnit): class LTX2AudioVideoUnit_InputVideoEmbedder(PipelineUnit): def __init__(self): super().__init__( - input_params=("input_video", "video_noise", "audio_noise", "tiled", "tile_size", "tile_stride"), - output_params=("video_latents", "audio_latents"), + input_params=("input_video", "video_noise", "tiled", "tile_size_in_pixels", "tile_overlap_in_pixels"), + output_params=("video_latents", "input_latents"), onload_model_names=("video_vae_encoder") ) - def process(self, pipe: LTX2AudioVideoPipeline, input_video, video_noise, audio_noise, tiled, tile_size, tile_stride): + def process(self, pipe: LTX2AudioVideoPipeline, input_video, video_noise, tiled, tile_size_in_pixels, tile_overlap_in_pixels): if input_video is None: - return {"video_latents": video_noise, "audio_latents": audio_noise} + return {"video_latents": video_noise} else: - # TODO: implement video-to-video - raise NotImplementedError("Video-to-video not implemented yet.") + pipe.load_models_to_device(self.onload_model_names) + input_video = pipe.preprocess_video(input_video) + input_latents = pipe.video_vae_encoder.encode(input_video, tiled, tile_size_in_pixels, tile_overlap_in_pixels).to(dtype=pipe.torch_dtype, device=pipe.device) + if pipe.scheduler.training: + return {"video_latents": input_latents, "input_latents": input_latents} + else: + # TODO: implement video-to-video + raise NotImplementedError("Video-to-video not implemented yet.") + +class LTX2AudioVideoUnit_InputAudioEmbedder(PipelineUnit): + def __init__(self): + super().__init__( + input_params=("input_audio", "audio_noise"), + output_params=("audio_latents", "audio_input_latents", "audio_positions", "audio_latent_shape"), + onload_model_names=("audio_vae_encoder",) + ) + + def process(self, pipe: LTX2AudioVideoPipeline, input_audio, audio_noise): + if input_audio is None: + return {"audio_latents": audio_noise} + else: + input_audio, sample_rate = input_audio + pipe.load_models_to_device(self.onload_model_names) + input_audio = pipe.audio_processor.waveform_to_mel(input_audio.unsqueeze(0), waveform_sample_rate=sample_rate).to(dtype=pipe.torch_dtype) + audio_input_latents = pipe.audio_vae_encoder(input_audio) + audio_noise = torch.randn_like(audio_input_latents) + audio_latent_shape = AudioLatentShape.from_torch_shape(audio_input_latents.shape) + audio_positions = pipe.audio_patchifier.get_patch_grid_bounds(audio_latent_shape, device=pipe.device) + if pipe.scheduler.training: + return {"audio_latents": audio_input_latents, "audio_input_latents": audio_input_latents, "audio_positions": audio_positions, "audio_latent_shape": audio_latent_shape} + else: + # TODO: implement video-to-video + raise NotImplementedError("Video-to-video not implemented yet.") class LTX2AudioVideoUnit_InputImagesEmbedder(PipelineUnit): def __init__(self): @@ -530,9 +562,12 @@ def model_fn_ltx2( video_timesteps = timestep.repeat(1, video_latents.shape[1], 1) if denoise_mask_video is not None: video_timesteps = video_patchifier.patchify(denoise_mask_video) * video_timesteps - _, c_a, _, mel_bins = audio_latents.shape - audio_latents = audio_patchifier.patchify(audio_latents) - audio_timesteps = timestep.repeat(1, audio_latents.shape[1], 1) + if audio_latents is not None: + _, c_a, _, mel_bins = audio_latents.shape + audio_latents = audio_patchifier.patchify(audio_latents) + audio_timesteps = timestep.repeat(1, audio_latents.shape[1], 1) + else: + audio_timesteps = None #TODO: support gradient checkpointing in training vx, ax = dit( video_latents=video_latents, @@ -546,5 +581,5 @@ def model_fn_ltx2( ) # unpatchify vx = video_patchifier.unpatchify_video(vx, f, h, w) - ax = audio_patchifier.unpatchify_audio(ax, c_a, mel_bins) + ax = audio_patchifier.unpatchify_audio(ax, c_a, mel_bins) if ax is not None else None return vx, ax diff --git a/examples/ltx2/model_inference/LTX-2-I2AV-DistilledPipeline.py b/examples/ltx2/model_inference/LTX-2-I2AV-DistilledPipeline.py index b8e0811..39623dd 100644 --- a/examples/ltx2/model_inference/LTX-2-I2AV-DistilledPipeline.py +++ b/examples/ltx2/model_inference/LTX-2-I2AV-DistilledPipeline.py @@ -19,7 +19,12 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( device="cuda", model_configs=[ ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized", origin_file_pattern="model-*.safetensors", **vram_config), - ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-distilled.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="transformer_distilled.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="text_encoder_post_modules.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_encoder.safetensors", **vram_config), ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-spatial-upscaler-x2-1.0.safetensors", **vram_config), ], tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), diff --git a/examples/ltx2/model_inference/LTX-2-I2AV-OneStage.py b/examples/ltx2/model_inference/LTX-2-I2AV-OneStage.py index 1614c1a..c50ff5f 100644 --- a/examples/ltx2/model_inference/LTX-2-I2AV-OneStage.py +++ b/examples/ltx2/model_inference/LTX-2-I2AV-OneStage.py @@ -19,7 +19,12 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( device="cuda", model_configs=[ ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized", origin_file_pattern="model-*.safetensors", **vram_config), - ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-dev.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="transformer.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="text_encoder_post_modules.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_encoder.safetensors", **vram_config), ], tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), ) diff --git a/examples/ltx2/model_inference/LTX-2-I2AV-TwoStage.py b/examples/ltx2/model_inference/LTX-2-I2AV-TwoStage.py index e73ef3d..bd86b34 100644 --- a/examples/ltx2/model_inference/LTX-2-I2AV-TwoStage.py +++ b/examples/ltx2/model_inference/LTX-2-I2AV-TwoStage.py @@ -19,7 +19,12 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( device="cuda", model_configs=[ ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized", origin_file_pattern="model-*.safetensors", **vram_config), - ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-dev.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="transformer.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="text_encoder_post_modules.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_encoder.safetensors", **vram_config), ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-spatial-upscaler-x2-1.0.safetensors", **vram_config), ], tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), diff --git a/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Dolly-In.py b/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Dolly-In.py index c1dc94b..da454dc 100644 --- a/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Dolly-In.py +++ b/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Dolly-In.py @@ -17,7 +17,11 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( device="cuda", model_configs=[ ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized", origin_file_pattern="model-*.safetensors", **vram_config), - ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-dev.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="transformer.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="text_encoder_post_modules.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-spatial-upscaler-x2-1.0.safetensors", **vram_config), ], tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), diff --git a/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Dolly-Left.py b/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Dolly-Left.py index f6b3f0a..e6dd9ba 100644 --- a/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Dolly-Left.py +++ b/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Dolly-Left.py @@ -17,7 +17,11 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( device="cuda", model_configs=[ ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized", origin_file_pattern="model-*.safetensors", **vram_config), - ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-dev.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="transformer.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="text_encoder_post_modules.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-spatial-upscaler-x2-1.0.safetensors", **vram_config), ], tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), diff --git a/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Dolly-Out.py b/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Dolly-Out.py index 6f8fd72..a84efe9 100644 --- a/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Dolly-Out.py +++ b/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Dolly-Out.py @@ -17,7 +17,11 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( device="cuda", model_configs=[ ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized", origin_file_pattern="model-*.safetensors", **vram_config), - ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-dev.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="transformer.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="text_encoder_post_modules.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-spatial-upscaler-x2-1.0.safetensors", **vram_config), ], tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), diff --git a/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Dolly-Right.py b/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Dolly-Right.py index 2de3233..0a2e968 100644 --- a/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Dolly-Right.py +++ b/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Dolly-Right.py @@ -17,7 +17,11 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( device="cuda", model_configs=[ ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized", origin_file_pattern="model-*.safetensors", **vram_config), - ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-dev.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="transformer.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="text_encoder_post_modules.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-spatial-upscaler-x2-1.0.safetensors", **vram_config), ], tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), diff --git a/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Jib-Down.py b/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Jib-Down.py index 571fd6b..4adfdec 100644 --- a/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Jib-Down.py +++ b/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Jib-Down.py @@ -17,7 +17,11 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( device="cuda", model_configs=[ ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized", origin_file_pattern="model-*.safetensors", **vram_config), - ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-dev.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="transformer.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="text_encoder_post_modules.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-spatial-upscaler-x2-1.0.safetensors", **vram_config), ], tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), diff --git a/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Jib-Up.py b/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Jib-Up.py index 18905fe..6dfa664 100644 --- a/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Jib-Up.py +++ b/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Jib-Up.py @@ -17,7 +17,11 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( device="cuda", model_configs=[ ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized", origin_file_pattern="model-*.safetensors", **vram_config), - ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-dev.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="transformer.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="text_encoder_post_modules.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-spatial-upscaler-x2-1.0.safetensors", **vram_config), ], tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), diff --git a/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Static.py b/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Static.py index ffa9b38..f351789 100644 --- a/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Static.py +++ b/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Static.py @@ -17,7 +17,11 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( device="cuda", model_configs=[ ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized", origin_file_pattern="model-*.safetensors", **vram_config), - ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-dev.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="transformer.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="text_encoder_post_modules.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-spatial-upscaler-x2-1.0.safetensors", **vram_config), ], tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), diff --git a/examples/ltx2/model_inference/LTX-2-T2AV-DistilledPipeline.py b/examples/ltx2/model_inference/LTX-2-T2AV-DistilledPipeline.py index 2b87dd3..5b57aea 100644 --- a/examples/ltx2/model_inference/LTX-2-T2AV-DistilledPipeline.py +++ b/examples/ltx2/model_inference/LTX-2-T2AV-DistilledPipeline.py @@ -17,7 +17,11 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( device="cuda", model_configs=[ ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized", origin_file_pattern="model-*.safetensors", **vram_config), - ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-distilled.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="transformer_distilled.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="text_encoder_post_modules.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-spatial-upscaler-x2-1.0.safetensors", **vram_config), ], tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), diff --git a/examples/ltx2/model_inference/LTX-2-T2AV-OneStage.py b/examples/ltx2/model_inference/LTX-2-T2AV-OneStage.py index ade78d0..2f73f80 100644 --- a/examples/ltx2/model_inference/LTX-2-T2AV-OneStage.py +++ b/examples/ltx2/model_inference/LTX-2-T2AV-OneStage.py @@ -17,7 +17,11 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( device="cuda", model_configs=[ ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized", origin_file_pattern="model-*.safetensors", **vram_config), - ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-dev.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="transformer.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="text_encoder_post_modules.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), ], tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), ) @@ -40,3 +44,12 @@ write_video_audio_ltx2( fps=24, audio_sample_rate=24000, ) + + + # ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="transformer.safetensors", **vram_config), + # ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="text_encoder_post_modules.safetensors", **vram_config), + # ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_encoder.safetensors", **vram_config), + # ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), + # ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), + # ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_encoder.safetensors", **vram_config), + # ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), \ No newline at end of file diff --git a/examples/ltx2/model_inference/LTX-2-T2AV-TwoStage.py b/examples/ltx2/model_inference/LTX-2-T2AV-TwoStage.py index 84bbc0c..065e92c 100644 --- a/examples/ltx2/model_inference/LTX-2-T2AV-TwoStage.py +++ b/examples/ltx2/model_inference/LTX-2-T2AV-TwoStage.py @@ -17,7 +17,11 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( device="cuda", model_configs=[ ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized", origin_file_pattern="model-*.safetensors", **vram_config), - ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-dev.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="transformer.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="text_encoder_post_modules.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-spatial-upscaler-x2-1.0.safetensors", **vram_config), ], tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), diff --git a/examples/ltx2/model_inference_low_vram/LTX-2-I2AV-DistilledPipeline.py b/examples/ltx2/model_inference_low_vram/LTX-2-I2AV-DistilledPipeline.py index 7020b40..370e371 100644 --- a/examples/ltx2/model_inference_low_vram/LTX-2-I2AV-DistilledPipeline.py +++ b/examples/ltx2/model_inference_low_vram/LTX-2-I2AV-DistilledPipeline.py @@ -19,7 +19,12 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( device="cuda", model_configs=[ ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized", origin_file_pattern="model-*.safetensors", **vram_config), - ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-distilled.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="transformer_distilled.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="text_encoder_post_modules.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_encoder.safetensors", **vram_config), ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-spatial-upscaler-x2-1.0.safetensors", **vram_config), ], tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), diff --git a/examples/ltx2/model_inference_low_vram/LTX-2-I2AV-OneStage.py b/examples/ltx2/model_inference_low_vram/LTX-2-I2AV-OneStage.py index 48ca23b..a42a0bb 100644 --- a/examples/ltx2/model_inference_low_vram/LTX-2-I2AV-OneStage.py +++ b/examples/ltx2/model_inference_low_vram/LTX-2-I2AV-OneStage.py @@ -19,7 +19,12 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( device="cuda", model_configs=[ ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized", origin_file_pattern="model-*.safetensors", **vram_config), - ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-dev.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="transformer.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="text_encoder_post_modules.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_encoder.safetensors", **vram_config), ], tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), vram_limit=torch.cuda.mem_get_info("cuda")[1] / (1024 ** 3) - 0.5, diff --git a/examples/ltx2/model_inference_low_vram/LTX-2-I2AV-TwoStage.py b/examples/ltx2/model_inference_low_vram/LTX-2-I2AV-TwoStage.py index 5411b8c..912fa93 100644 --- a/examples/ltx2/model_inference_low_vram/LTX-2-I2AV-TwoStage.py +++ b/examples/ltx2/model_inference_low_vram/LTX-2-I2AV-TwoStage.py @@ -19,7 +19,12 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( device="cuda", model_configs=[ ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized", origin_file_pattern="model-*.safetensors", **vram_config), - ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-dev.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="transformer.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="text_encoder_post_modules.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_encoder.safetensors", **vram_config), ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-spatial-upscaler-x2-1.0.safetensors", **vram_config), ], tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), diff --git a/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Dolly-In.py b/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Dolly-In.py index b15e4cf..9f6af9f 100644 --- a/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Dolly-In.py +++ b/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Dolly-In.py @@ -17,7 +17,11 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( device="cuda", model_configs=[ ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized", origin_file_pattern="model-*.safetensors", **vram_config), - ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-dev.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="transformer.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="text_encoder_post_modules.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-spatial-upscaler-x2-1.0.safetensors", **vram_config), ], tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), diff --git a/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Dolly-Left.py b/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Dolly-Left.py index 4a7a5aa..8ee5fa9 100644 --- a/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Dolly-Left.py +++ b/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Dolly-Left.py @@ -17,7 +17,11 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( device="cuda", model_configs=[ ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized", origin_file_pattern="model-*.safetensors", **vram_config), - ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-dev.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="transformer.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="text_encoder_post_modules.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-spatial-upscaler-x2-1.0.safetensors", **vram_config), ], tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), diff --git a/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Dolly-Out.py b/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Dolly-Out.py index 9ae6884..3bb9f50 100644 --- a/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Dolly-Out.py +++ b/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Dolly-Out.py @@ -17,7 +17,11 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( device="cuda", model_configs=[ ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized", origin_file_pattern="model-*.safetensors", **vram_config), - ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-dev.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="transformer.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="text_encoder_post_modules.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-spatial-upscaler-x2-1.0.safetensors", **vram_config), ], tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), diff --git a/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Dolly-Right.py b/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Dolly-Right.py index ab9f9ae..0301445 100644 --- a/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Dolly-Right.py +++ b/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Dolly-Right.py @@ -17,7 +17,11 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( device="cuda", model_configs=[ ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized", origin_file_pattern="model-*.safetensors", **vram_config), - ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-dev.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="transformer.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="text_encoder_post_modules.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-spatial-upscaler-x2-1.0.safetensors", **vram_config), ], tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), diff --git a/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Jib-Down.py b/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Jib-Down.py index 9fc6e41..d1a8db3 100644 --- a/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Jib-Down.py +++ b/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Jib-Down.py @@ -17,7 +17,11 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( device="cuda", model_configs=[ ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized", origin_file_pattern="model-*.safetensors", **vram_config), - ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-dev.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="transformer.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="text_encoder_post_modules.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-spatial-upscaler-x2-1.0.safetensors", **vram_config), ], tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), diff --git a/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Jib-Up.py b/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Jib-Up.py index 628e7c3..0ae8041 100644 --- a/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Jib-Up.py +++ b/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Jib-Up.py @@ -17,7 +17,11 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( device="cuda", model_configs=[ ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized", origin_file_pattern="model-*.safetensors", **vram_config), - ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-dev.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="transformer.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="text_encoder_post_modules.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-spatial-upscaler-x2-1.0.safetensors", **vram_config), ], tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), diff --git a/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Static.py b/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Static.py index b6394bc..f8130ae 100644 --- a/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Static.py +++ b/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Static.py @@ -17,7 +17,11 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( device="cuda", model_configs=[ ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized", origin_file_pattern="model-*.safetensors", **vram_config), - ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-dev.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="transformer.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="text_encoder_post_modules.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-spatial-upscaler-x2-1.0.safetensors", **vram_config), ], tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), diff --git a/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-DistilledPipeline.py b/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-DistilledPipeline.py index d8b6a5d..836077b 100644 --- a/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-DistilledPipeline.py +++ b/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-DistilledPipeline.py @@ -17,7 +17,11 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( device="cuda", model_configs=[ ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized", origin_file_pattern="model-*.safetensors", **vram_config), - ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-distilled.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="transformer_distilled.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="text_encoder_post_modules.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-spatial-upscaler-x2-1.0.safetensors", **vram_config), ], tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), diff --git a/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-OneStage.py b/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-OneStage.py index 894c417..fb08384 100644 --- a/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-OneStage.py +++ b/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-OneStage.py @@ -17,7 +17,11 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( device="cuda", model_configs=[ ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized", origin_file_pattern="model-*.safetensors", **vram_config), - ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-dev.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="transformer.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="text_encoder_post_modules.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), ], tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), vram_limit=torch.cuda.mem_get_info("cuda")[1] / (1024 ** 3) - 0.5, diff --git a/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-TwoStage.py b/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-TwoStage.py index 65650d0..f513f46 100644 --- a/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-TwoStage.py +++ b/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-TwoStage.py @@ -17,7 +17,11 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( device="cuda", model_configs=[ ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized", origin_file_pattern="model-*.safetensors", **vram_config), - ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-dev.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="transformer.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="text_encoder_post_modules.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-spatial-upscaler-x2-1.0.safetensors", **vram_config), ], tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), diff --git a/examples/ltx2/model_training/full/LTX-2-T2AV-splited.sh b/examples/ltx2/model_training/full/LTX-2-T2AV-splited.sh new file mode 100644 index 0000000..8dfdd4b --- /dev/null +++ b/examples/ltx2/model_training/full/LTX-2-T2AV-splited.sh @@ -0,0 +1,35 @@ +# Splited Training +accelerate launch examples/ltx2/model_training/train.py \ + --dataset_base_path data/example_video_dataset/ltx2 \ + --dataset_metadata_path data/example_video_dataset/ltx2_t2av.csv \ + --data_file_keys "video,input_audio" \ + --extra_inputs "input_audio" \ + --height 512 \ + --width 768 \ + --num_frames 49 \ + --dataset_repeat 1 \ + --model_id_with_origin_paths "DiffSynth-Studio/LTX-2-Repackage:text_encoder_post_modules.safetensors,DiffSynth-Studio/LTX-2-Repackage:video_vae_encoder.safetensors,DiffSynth-Studio/LTX-2-Repackage:audio_vae_encoder.safetensors,google/gemma-3-12b-it-qat-q4_0-unquantized:model-*.safetensors" \ + --learning_rate 1e-4 \ + --num_epochs 5 \ + --remove_prefix_in_ckpt "pipe.dit." \ + --output_path "./models/train/LTX2-T2AV-full-splited-cache" \ + --trainable_models "dit" \ + --use_gradient_checkpointing \ + --task "sft:data_process" + +accelerate launch examples/ltx2/model_training/train.py \ + --dataset_base_path ./models/train/LTX2-T2AV-full-splited-cache \ + --data_file_keys "video,input_audio" \ + --extra_inputs "input_audio" \ + --height 512 \ + --width 768 \ + --num_frames 49 \ + --dataset_repeat 100 \ + --model_id_with_origin_paths "DiffSynth-Studio/LTX-2-Repackage:transformer.safetensors" \ + --learning_rate 1e-4 \ + --num_epochs 5 \ + --remove_prefix_in_ckpt "pipe.dit." \ + --output_path "./models/train/LTX2-T2AV-full" \ + --trainable_models "dit" \ + --use_gradient_checkpointing \ + --task "sft:train" diff --git a/examples/ltx2/model_training/lora/LTX-2-T2AV-noaudio.sh b/examples/ltx2/model_training/lora/LTX-2-T2AV-noaudio.sh new file mode 100644 index 0000000..4f1e754 --- /dev/null +++ b/examples/ltx2/model_training/lora/LTX-2-T2AV-noaudio.sh @@ -0,0 +1,56 @@ +# single stage training +# accelerate launch examples/ltx2/model_training/train.py \ +# --dataset_base_path data/example_video_dataset/ltx2 \ +# --dataset_metadata_path data/example_video_dataset/ltx2_t2v.csv \ +# --height 256 \ +# --width 384 \ +# --num_frames 25\ +# --dataset_repeat 100 \ +# --model_id_with_origin_paths "DiffSynth-Studio/LTX-2-Repackage:transformer.safetensors,DiffSynth-Studio/LTX-2-Repackage:text_encoder_post_modules.safetensors,DiffSynth-Studio/LTX-2-Repackage:video_vae_encoder.safetensors,DiffSynth-Studio/LTX-2-Repackage:audio_vae_encoder.safetensors,google/gemma-3-12b-it-qat-q4_0-unquantized:model-*.safetensors" \ +# --learning_rate 1e-4 \ +# --num_epochs 5 \ +# --remove_prefix_in_ckpt "pipe.dit." \ +# --output_path "./models/train/LTX2-T2AV-noaudio_lora" \ +# --lora_base_model "dit" \ +# --lora_target_modules "to_k,to_q,to_v,to_out.0" \ +# --lora_rank 32 \ +# --use_gradient_checkpointing \ +# --find_unused_parameters + + +# Splited Training +accelerate launch examples/ltx2/model_training/train.py \ + --dataset_base_path data/example_video_dataset/ltx2 \ + --dataset_metadata_path data/example_video_dataset/ltx2_t2av.csv \ + --height 256 \ + --width 384 \ + --num_frames 49\ + --dataset_repeat 1 \ + --model_id_with_origin_paths "DiffSynth-Studio/LTX-2-Repackage:text_encoder_post_modules.safetensors,DiffSynth-Studio/LTX-2-Repackage:video_vae_encoder.safetensors,DiffSynth-Studio/LTX-2-Repackage:audio_vae_encoder.safetensors,google/gemma-3-12b-it-qat-q4_0-unquantized:model-*.safetensors" \ + --learning_rate 1e-4 \ + --num_epochs 5 \ + --remove_prefix_in_ckpt "pipe.dit." \ + --output_path "./models/train/LTX2-T2AV-noaudio_lora-splited-cache" \ + --lora_base_model "dit" \ + --lora_target_modules "to_k,to_q,to_v,to_out.0" \ + --lora_rank 32 \ + --use_gradient_checkpointing \ + --task "sft:data_process" + + +accelerate launch examples/ltx2/model_training/train.py \ + --dataset_base_path ./models/train/LTX2-T2AV-noaudio_lora-splited-cache \ + --height 256 \ + --width 384 \ + --num_frames 49\ + --dataset_repeat 100 \ + --model_id_with_origin_paths "DiffSynth-Studio/LTX-2-Repackage:transformer.safetensors" \ + --learning_rate 1e-4 \ + --num_epochs 5 \ + --remove_prefix_in_ckpt "pipe.dit." \ + --output_path "./models/train/LTX2-T2AV-noaudio_lora" \ + --lora_base_model "dit" \ + --lora_target_modules "to_k,to_q,to_v,to_out.0" \ + --lora_rank 32 \ + --use_gradient_checkpointing \ + --task "sft:train" diff --git a/examples/ltx2/model_training/lora/LTX-2-T2AV-splited.sh b/examples/ltx2/model_training/lora/LTX-2-T2AV-splited.sh new file mode 100644 index 0000000..e71494a --- /dev/null +++ b/examples/ltx2/model_training/lora/LTX-2-T2AV-splited.sh @@ -0,0 +1,40 @@ +# Splited Training +accelerate launch examples/ltx2/model_training/train.py \ + --dataset_base_path data/example_video_dataset/ltx2 \ + --dataset_metadata_path data/example_video_dataset/ltx2_t2av.csv \ + --data_file_keys "video,input_audio" \ + --extra_inputs "input_audio" \ + --height 512 \ + --width 768 \ + --num_frames 49 \ + --dataset_repeat 1 \ + --model_id_with_origin_paths "DiffSynth-Studio/LTX-2-Repackage:text_encoder_post_modules.safetensors,DiffSynth-Studio/LTX-2-Repackage:video_vae_encoder.safetensors,DiffSynth-Studio/LTX-2-Repackage:audio_vae_encoder.safetensors,google/gemma-3-12b-it-qat-q4_0-unquantized:model-*.safetensors" \ + --learning_rate 1e-4 \ + --num_epochs 5 \ + --remove_prefix_in_ckpt "pipe.dit." \ + --output_path "./models/train/LTX2-T2AV_lora-splited-cache" \ + --lora_base_model "dit" \ + --lora_target_modules "to_k,to_q,to_v,to_out.0" \ + --lora_rank 32 \ + --use_gradient_checkpointing \ + --task "sft:data_process" + + +accelerate launch examples/ltx2/model_training/train.py \ + --dataset_base_path ./models/train/LTX2-T2AV_lora-splited-cache \ + --data_file_keys "video,input_audio" \ + --extra_inputs "input_audio" \ + --height 512 \ + --width 768 \ + --num_frames 49 \ + --dataset_repeat 100 \ + --model_id_with_origin_paths "DiffSynth-Studio/LTX-2-Repackage:transformer.safetensors" \ + --learning_rate 1e-4 \ + --num_epochs 5 \ + --remove_prefix_in_ckpt "pipe.dit." \ + --output_path "./models/train/LTX2-T2AV_lora" \ + --lora_base_model "dit" \ + --lora_target_modules "to_k,to_q,to_v,to_out.0" \ + --lora_rank 32 \ + --use_gradient_checkpointing \ + --task "sft:train" diff --git a/examples/ltx2/model_training/lora/LTX-2-T2AV.sh b/examples/ltx2/model_training/lora/LTX-2-T2AV.sh new file mode 100644 index 0000000..1d8aaaf --- /dev/null +++ b/examples/ltx2/model_training/lora/LTX-2-T2AV.sh @@ -0,0 +1,19 @@ +accelerate launch examples/ltx2/model_training/train.py \ + --dataset_base_path data/example_video_dataset/ltx2 \ + --dataset_metadata_path data/example_video_dataset/ltx2_t2av.csv \ + --data_file_keys "video,input_audio" \ + --extra_inputs "input_audio" \ + --height 256 \ + --width 384 \ + --num_frames 25\ + --dataset_repeat 100 \ + --model_id_with_origin_paths "DiffSynth-Studio/LTX-2-Repackage:transformer.safetensors,DiffSynth-Studio/LTX-2-Repackage:text_encoder_post_modules.safetensors,DiffSynth-Studio/LTX-2-Repackage:video_vae_encoder.safetensors,DiffSynth-Studio/LTX-2-Repackage:audio_vae_encoder.safetensors,google/gemma-3-12b-it-qat-q4_0-unquantized:model-*.safetensors" \ + --learning_rate 1e-4 \ + --num_epochs 5 \ + --remove_prefix_in_ckpt "pipe.dit." \ + --output_path "./models/train/LTX2-T2AV_lora" \ + --lora_base_model "dit" \ + --lora_target_modules "to_k,to_q,to_v,to_out.0" \ + --lora_rank 32 \ + --use_gradient_checkpointing \ + --find_unused_parameters diff --git a/examples/ltx2/model_training/train.py b/examples/ltx2/model_training/train.py new file mode 100644 index 0000000..729b4ed --- /dev/null +++ b/examples/ltx2/model_training/train.py @@ -0,0 +1,162 @@ +import torch, os, argparse, accelerate, warnings +from diffsynth.core import UnifiedDataset +from diffsynth.core.data.operators import LoadAudioWithTorchaudio, ToAbsolutePath +from diffsynth.pipelines.wan_video import WanVideoPipeline, ModelConfig +from diffsynth.pipelines.ltx2_audio_video import LTX2AudioVideoPipeline, ModelConfig +from diffsynth.diffusion import * +os.environ["TOKENIZERS_PARALLELISM"] = "false" + + +class LTX2TrainingModule(DiffusionTrainingModule): + def __init__( + self, + model_paths=None, model_id_with_origin_paths=None, + tokenizer_path=None, + trainable_models=None, + lora_base_model=None, lora_target_modules="", lora_rank=32, lora_checkpoint=None, + preset_lora_path=None, preset_lora_model=None, + use_gradient_checkpointing=True, + use_gradient_checkpointing_offload=False, + extra_inputs=None, + fp8_models=None, + offload_models=None, + device="cpu", + task="sft", + ): + super().__init__() + # Warning + if not use_gradient_checkpointing: + warnings.warn("Gradient checkpointing is detected as disabled. To prevent out-of-memory errors, the training framework will forcibly enable gradient checkpointing.") + use_gradient_checkpointing = True + + # Load models + model_configs = self.parse_model_configs(model_paths, model_id_with_origin_paths, fp8_models=fp8_models, offload_models=offload_models, device=device) + tokenizer_config = ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized") if tokenizer_path is None else ModelConfig(tokenizer_path) + self.pipe = LTX2AudioVideoPipeline.from_pretrained(torch_dtype=torch.bfloat16, device=device, model_configs=model_configs, tokenizer_config=tokenizer_config) + self.pipe = self.split_pipeline_units(task, self.pipe, trainable_models, lora_base_model) + # Training mode + self.switch_pipe_to_training_mode( + self.pipe, trainable_models, + lora_base_model, lora_target_modules, lora_rank, lora_checkpoint, + preset_lora_path, preset_lora_model, + task=task, + ) + + # Store other configs + self.use_gradient_checkpointing = use_gradient_checkpointing + self.use_gradient_checkpointing_offload = use_gradient_checkpointing_offload + self.extra_inputs = extra_inputs.split(",") if extra_inputs is not None else [] + self.fp8_models = fp8_models + self.task = task + self.task_to_loss = { + "sft:data_process": lambda pipe, *args: args, + "sft": lambda pipe, inputs_shared, inputs_posi, inputs_nega: FlowMatchSFTAudioVideoLoss(pipe, **inputs_shared, **inputs_posi), + "sft:train": lambda pipe, inputs_shared, inputs_posi, inputs_nega: FlowMatchSFTAudioVideoLoss(pipe, **inputs_shared, **inputs_posi), + } + + def parse_extra_inputs(self, data, extra_inputs, inputs_shared): + for extra_input in extra_inputs: + inputs_shared[extra_input] = data[extra_input] + return inputs_shared + + def get_pipeline_inputs(self, data): + inputs_posi = {"prompt": data["prompt"]} + inputs_nega = {} + inputs_shared = { + # Assume you are using this pipeline for inference, + # please fill in the input parameters. + "input_video": data["video"], + "height": data["video"][0].size[1], + "width": data["video"][0].size[0], + "num_frames": len(data["video"]), + # Please do not modify the following parameters + # unless you clearly know what this will cause. + "cfg_scale": 1, + "tiled": False, + "rand_device": self.pipe.device, + "use_gradient_checkpointing": self.use_gradient_checkpointing, + "use_gradient_checkpointing_offload": self.use_gradient_checkpointing_offload, + "video_patchifier": self.pipe.video_patchifier, + "audio_patchifier": self.pipe.audio_patchifier, + } + inputs_shared = self.parse_extra_inputs(data, self.extra_inputs, inputs_shared) + return inputs_shared, inputs_posi, inputs_nega + + def forward(self, data, inputs=None): + if inputs is None: inputs = self.get_pipeline_inputs(data) + inputs = self.transfer_data_to_device(inputs, self.pipe.device, self.pipe.torch_dtype) + for unit in self.pipe.units: + inputs = self.pipe.unit_runner(unit, self.pipe, *inputs) + loss = self.task_to_loss[self.task](self.pipe, *inputs) + return loss + + +def ltx2_parser(): + parser = argparse.ArgumentParser(description="Simple example of a training script.") + parser = add_general_config(parser) + parser = add_video_size_config(parser) + parser.add_argument("--tokenizer_path", type=str, default=None, help="Path to tokenizer.") + parser.add_argument("--frame_rate", type=float, default=24, help="frame rate of the training videos. If not specified, it will be determined by the dataset.") + parser.add_argument("--initialize_model_on_cpu", default=False, action="store_true", help="Whether to initialize models on CPU.") + return parser + + +if __name__ == "__main__": + parser = ltx2_parser() + args = parser.parse_args() + accelerator = accelerate.Accelerator( + gradient_accumulation_steps=args.gradient_accumulation_steps, + kwargs_handlers=[accelerate.DistributedDataParallelKwargs(find_unused_parameters=args.find_unused_parameters)], + ) + dataset = UnifiedDataset( + base_path=args.dataset_base_path, + metadata_path=args.dataset_metadata_path, + repeat=args.dataset_repeat, + data_file_keys=args.data_file_keys.split(","), + main_data_operator=UnifiedDataset.default_video_operator( + base_path=args.dataset_base_path, + max_pixels=args.max_pixels, + height=args.height, + width=args.width, + height_division_factor=16, + width_division_factor=16, + num_frames=args.num_frames, + time_division_factor=4, + time_division_remainder=1, + ), + special_operator_map={ + "input_audio": ToAbsolutePath(args.dataset_base_path) >> LoadAudioWithTorchaudio(duration=float(args.num_frames) / float(args.frame_rate)), + } + ) + model = LTX2TrainingModule( + model_paths=args.model_paths, + model_id_with_origin_paths=args.model_id_with_origin_paths, + tokenizer_path=args.tokenizer_path, + trainable_models=args.trainable_models, + lora_base_model=args.lora_base_model, + lora_target_modules=args.lora_target_modules, + lora_rank=args.lora_rank, + lora_checkpoint=args.lora_checkpoint, + preset_lora_path=args.preset_lora_path, + preset_lora_model=args.preset_lora_model, + use_gradient_checkpointing=args.use_gradient_checkpointing, + use_gradient_checkpointing_offload=args.use_gradient_checkpointing_offload, + extra_inputs=args.extra_inputs, + fp8_models=args.fp8_models, + offload_models=args.offload_models, + task=args.task, + device="cpu" if args.initialize_model_on_cpu else accelerator.device, + ) + model_logger = ModelLogger( + args.output_path, + remove_prefix_in_ckpt=args.remove_prefix_in_ckpt, + ) + launcher_map = { + "sft:data_process": launch_data_process_task, + "direct_distill:data_process": launch_data_process_task, + "sft": launch_training_task, + "sft:train": launch_training_task, + "direct_distill": launch_training_task, + "direct_distill:train": launch_training_task, + } + launcher_map[args.task](accelerator, dataset, model, model_logger, args=args) diff --git a/examples/ltx2/model_training/validate_full/LTX-2-T2AV.py b/examples/ltx2/model_training/validate_full/LTX-2-T2AV.py new file mode 100644 index 0000000..6201ec1 --- /dev/null +++ b/examples/ltx2/model_training/validate_full/LTX-2-T2AV.py @@ -0,0 +1,47 @@ +import torch +from diffsynth.pipelines.ltx2_audio_video import LTX2AudioVideoPipeline, ModelConfig +from diffsynth.utils.data.media_io_ltx2 import write_video_audio_ltx2 + +vram_config = { + "offload_dtype": torch.bfloat16, + "offload_device": "cpu", + "onload_dtype": torch.bfloat16, + "onload_device": "cuda", + "preparing_dtype": torch.bfloat16, + "preparing_device": "cuda", + "computation_dtype": torch.bfloat16, + "computation_device": "cuda", +} +pipe = LTX2AudioVideoPipeline.from_pretrained( + torch_dtype=torch.bfloat16, + device="cuda", + model_configs=[ + ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized", origin_file_pattern="model-*.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="text_encoder_post_modules.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), + ModelConfig(path="./models/train/LTX2-T2AV-full/epoch-4.safetensors", **vram_config), + ], + tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), +) +prompt = "A beautiful sunset over the ocean." +negative_prompt = "blurry, out of focus, overexposed, underexposed, low contrast, washed out colors, excessive noise, grainy texture, poor lighting, flickering, motion blur, distorted proportions, unnatural skin tones, deformed facial features, asymmetrical face, missing facial features, extra limbs, disfigured hands, wrong hand count, artifacts around text, inconsistent perspective, camera shake, incorrect depth of field, background too sharp, background clutter, distracting reflections, harsh shadows, inconsistent lighting direction, color banding, cartoonish rendering, 3D CGI look, unrealistic materials, uncanny valley effect, incorrect ethnicity, wrong gender, exaggerated expressions, wrong gaze direction, mismatched lip sync, silent or muted audio, distorted voice, robotic voice, echo, background noise, off-sync audio, incorrect dialogue, added dialogue, repetitive speech, jittery movement, awkward pauses, incorrect timing, unnatural transitions, inconsistent framing, tilted camera, flat lighting, inconsistent tone, cinematic oversaturation, stylized filters, or AI artifacts." +height, width, num_frames = 512, 768, 121 +video, audio = pipe( + prompt=prompt, + negative_prompt=negative_prompt, + seed=43, + height=height, + width=width, + num_frames=num_frames, + tiled=True, + cfg_scale=4.0 +) +write_video_audio_ltx2( + video=video, + audio=audio, + output_path='ltx2_onestage.mp4', + fps=24, + audio_sample_rate=24000, +) diff --git a/examples/ltx2/model_training/validate_lora/LTX-2-T2AV.py b/examples/ltx2/model_training/validate_lora/LTX-2-T2AV.py new file mode 100644 index 0000000..4372b6c --- /dev/null +++ b/examples/ltx2/model_training/validate_lora/LTX-2-T2AV.py @@ -0,0 +1,49 @@ +import torch +from diffsynth.pipelines.ltx2_audio_video import LTX2AudioVideoPipeline, ModelConfig +from diffsynth.utils.data.media_io_ltx2 import write_video_audio_ltx2 + +vram_config = { + "offload_dtype": torch.bfloat16, + "offload_device": "cpu", + "onload_dtype": torch.bfloat16, + "onload_device": "cuda", + "preparing_dtype": torch.bfloat16, + "preparing_device": "cuda", + "computation_dtype": torch.bfloat16, + "computation_device": "cuda", +} +pipe = LTX2AudioVideoPipeline.from_pretrained( + torch_dtype=torch.bfloat16, + device="cuda", + model_configs=[ + ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized", origin_file_pattern="model-*.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="transformer.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="text_encoder_post_modules.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), + ], + tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), +) +pipe.load_lora(pipe.dit, "models/train/LTX2-T2AV_lora/epoch-4.safetensors") +prompt = "A beautiful sunset over the ocean." +negative_prompt = "blurry, out of focus, overexposed, underexposed, low contrast, washed out colors, excessive noise, grainy texture, poor lighting, flickering, motion blur, distorted proportions, unnatural skin tones, deformed facial features, asymmetrical face, missing facial features, extra limbs, disfigured hands, wrong hand count, artifacts around text, inconsistent perspective, camera shake, incorrect depth of field, background too sharp, background clutter, distracting reflections, harsh shadows, inconsistent lighting direction, color banding, cartoonish rendering, 3D CGI look, unrealistic materials, uncanny valley effect, incorrect ethnicity, wrong gender, exaggerated expressions, wrong gaze direction, mismatched lip sync, silent or muted audio, distorted voice, robotic voice, echo, background noise, off-sync audio, incorrect dialogue, added dialogue, repetitive speech, jittery movement, awkward pauses, incorrect timing, unnatural transitions, inconsistent framing, tilted camera, flat lighting, inconsistent tone, cinematic oversaturation, stylized filters, or AI artifacts." +height, width, num_frames = 512, 768, 121 +height, width, num_frames = 256, 384, 25 +video, audio = pipe( + prompt=prompt, + negative_prompt=negative_prompt, + seed=43, + height=height, + width=width, + num_frames=num_frames, + tiled=True, + cfg_scale=4.0 +) +write_video_audio_ltx2( + video=video, + audio=audio, + output_path='ltx2_onestage.mp4', + fps=24, + audio_sample_rate=24000, +) diff --git a/examples/ltx2/model_training/validate_lora/LTX-2-T2AV_noaudio.py b/examples/ltx2/model_training/validate_lora/LTX-2-T2AV_noaudio.py new file mode 100644 index 0000000..b35c23c --- /dev/null +++ b/examples/ltx2/model_training/validate_lora/LTX-2-T2AV_noaudio.py @@ -0,0 +1,49 @@ +import torch +from diffsynth.pipelines.ltx2_audio_video import LTX2AudioVideoPipeline, ModelConfig +from diffsynth.utils.data.media_io_ltx2 import write_video_audio_ltx2 + +vram_config = { + "offload_dtype": torch.bfloat16, + "offload_device": "cpu", + "onload_dtype": torch.bfloat16, + "onload_device": "cuda", + "preparing_dtype": torch.bfloat16, + "preparing_device": "cuda", + "computation_dtype": torch.bfloat16, + "computation_device": "cuda", +} +pipe = LTX2AudioVideoPipeline.from_pretrained( + torch_dtype=torch.bfloat16, + device="cuda", + model_configs=[ + ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized", origin_file_pattern="model-*.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="transformer.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="text_encoder_post_modules.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), + ], + tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), +) +pipe.load_lora(pipe.dit, "models/train/LTX2-T2AV-noaudio_lora/epoch-4.safetensors") +prompt = "A beautiful sunset over the ocean." +negative_prompt = "blurry, out of focus, overexposed, underexposed, low contrast, washed out colors, excessive noise, grainy texture, poor lighting, flickering, motion blur, distorted proportions, unnatural skin tones, deformed facial features, asymmetrical face, missing facial features, extra limbs, disfigured hands, wrong hand count, artifacts around text, inconsistent perspective, camera shake, incorrect depth of field, background too sharp, background clutter, distracting reflections, harsh shadows, inconsistent lighting direction, color banding, cartoonish rendering, 3D CGI look, unrealistic materials, uncanny valley effect, incorrect ethnicity, wrong gender, exaggerated expressions, wrong gaze direction, mismatched lip sync, silent or muted audio, distorted voice, robotic voice, echo, background noise, off-sync audio, incorrect dialogue, added dialogue, repetitive speech, jittery movement, awkward pauses, incorrect timing, unnatural transitions, inconsistent framing, tilted camera, flat lighting, inconsistent tone, cinematic oversaturation, stylized filters, or AI artifacts." +height, width, num_frames = 512, 768, 121 +height, width, num_frames = 256, 384, 25 +video, audio = pipe( + prompt=prompt, + negative_prompt=negative_prompt, + seed=43, + height=height, + width=width, + num_frames=num_frames, + tiled=True, + cfg_scale=4.0 +) +write_video_audio_ltx2( + video=video, + audio=audio, + output_path='ltx2_onestage.mp4', + fps=24, + audio_sample_rate=24000, +) diff --git a/examples/ltx2/scripts/split_model_statedicts.py b/examples/ltx2/scripts/split_model_statedicts.py new file mode 100644 index 0000000..1ca2ff3 --- /dev/null +++ b/examples/ltx2/scripts/split_model_statedicts.py @@ -0,0 +1,104 @@ +from safetensors.torch import save_file +from diffsynth import hash_state_dict_keys +from diffsynth.core import load_state_dict +from diffsynth.models.model_loader import ModelPool + +model_pool = ModelPool() +state_dict = load_state_dict("models/Lightricks/LTX-2/ltx-2-19b-dev.safetensors") + +dit_state_dict = {} +for name in state_dict: + if name.startswith("model.diffusion_model."): + new_name = name.replace("model.diffusion_model.", "") + if new_name.startswith("audio_embeddings_connector.") or new_name.startswith("video_embeddings_connector."): + continue + dit_state_dict[name] = state_dict[name] + +print(f"dit_state_dict keys hash: {hash_state_dict_keys(dit_state_dict)}") +save_file(dit_state_dict, "models/DiffSynth-Studio/LTX-2-Repackage/transformer.safetensors") +model_pool.auto_load_model( + "models/DiffSynth-Studio/LTX-2-Repackage/transformer.safetensors", +) + + +video_vae_encoder_state_dict = {} +for name in state_dict: + if name.startswith("vae.encoder."): + video_vae_encoder_state_dict[name] = state_dict[name] + elif name.startswith("vae.per_channel_statistics."): + video_vae_encoder_state_dict[name] = state_dict[name] + +save_file(video_vae_encoder_state_dict, "models/DiffSynth-Studio/LTX-2-Repackage/video_vae_encoder.safetensors") +print(f"video_vae_encoder keys hash: {hash_state_dict_keys(video_vae_encoder_state_dict)}") +model_pool.auto_load_model("models/DiffSynth-Studio/LTX-2-Repackage/video_vae_encoder.safetensors") + + +video_vae_decoder_state_dict = {} +for name in state_dict: + if name.startswith("vae.decoder."): + video_vae_decoder_state_dict[name] = state_dict[name] + elif name.startswith("vae.per_channel_statistics."): + video_vae_decoder_state_dict[name] = state_dict[name] +save_file(video_vae_decoder_state_dict, "models/DiffSynth-Studio/LTX-2-Repackage/video_vae_decoder.safetensors") +print(f"video_vae_decoder keys hash: {hash_state_dict_keys(video_vae_decoder_state_dict)}") +model_pool.auto_load_model("models/DiffSynth-Studio/LTX-2-Repackage/video_vae_decoder.safetensors") + + +audio_vae_decoder_state_dict = {} +for name in state_dict: + if name.startswith("audio_vae.decoder."): + audio_vae_decoder_state_dict[name] = state_dict[name] + elif name.startswith("audio_vae.per_channel_statistics."): + audio_vae_decoder_state_dict[name] = state_dict[name] +save_file(audio_vae_decoder_state_dict, "models/DiffSynth-Studio/LTX-2-Repackage/audio_vae_decoder.safetensors") +print(f"audio_vae_decoder keys hash: {hash_state_dict_keys(audio_vae_decoder_state_dict)}") +model_pool.auto_load_model("models/DiffSynth-Studio/LTX-2-Repackage/audio_vae_decoder.safetensors") + + +audio_vae_encoder_state_dict = {} +for name in state_dict: + if name.startswith("audio_vae.encoder."): + audio_vae_encoder_state_dict[name] = state_dict[name] + elif name.startswith("audio_vae.per_channel_statistics."): + audio_vae_encoder_state_dict[name] = state_dict[name] +save_file(audio_vae_encoder_state_dict, "models/DiffSynth-Studio/LTX-2-Repackage/audio_vae_encoder.safetensors") +print(f"audio_vae_encoder keys hash: {hash_state_dict_keys(audio_vae_encoder_state_dict)}") +model_pool.auto_load_model("models/DiffSynth-Studio/LTX-2-Repackage/audio_vae_encoder.safetensors") + + +audio_vocoder_state_dict = {} +for name in state_dict: + if name.startswith("vocoder."): + audio_vocoder_state_dict[name] = state_dict[name] +save_file(audio_vocoder_state_dict, "models/DiffSynth-Studio/LTX-2-Repackage/audio_vocoder.safetensors") +print(f"audio_vocoder keys hash: {hash_state_dict_keys(audio_vocoder_state_dict)}") +model_pool.auto_load_model("models/DiffSynth-Studio/LTX-2-Repackage/audio_vocoder.safetensors") + + +text_encoder_post_modules_state_dict = {} +for name in state_dict: + if name.startswith("text_embedding_projection."): + text_encoder_post_modules_state_dict[name] = state_dict[name] + elif name.startswith("model.diffusion_model.video_embeddings_connector."): + text_encoder_post_modules_state_dict[name] = state_dict[name] + elif name.startswith("model.diffusion_model.audio_embeddings_connector."): + text_encoder_post_modules_state_dict[name] = state_dict[name] +save_file(text_encoder_post_modules_state_dict, "models/DiffSynth-Studio/LTX-2-Repackage/text_encoder_post_modules.safetensors") +print(f"text_encoder_post_modules keys hash: {hash_state_dict_keys(text_encoder_post_modules_state_dict)}") +model_pool.auto_load_model("models/DiffSynth-Studio/LTX-2-Repackage/text_encoder_post_modules.safetensors") + + +state_dict = load_state_dict("models/Lightricks/LTX-2/ltx-2-19b-distilled.safetensors") +dit_state_dict = {} +for name in state_dict: + if name.startswith("model.diffusion_model."): + new_name = name.replace("model.diffusion_model.", "") + if new_name.startswith("audio_embeddings_connector.") or new_name.startswith("video_embeddings_connector."): + continue + dit_state_dict[name] = state_dict[name] + +print(f"dit_state_dict keys hash: {hash_state_dict_keys(dit_state_dict)}") +save_file(dit_state_dict, "models/DiffSynth-Studio/LTX-2-Repackage/transformer_distilled.safetensors") +model_pool.auto_load_model( + "models/DiffSynth-Studio/LTX-2-Repackage/transformer_distilled.safetensors", +) \ No newline at end of file From 8e15dcd2894d6acbecb72ed717009f89b9ac7c7e Mon Sep 17 00:00:00 2001 From: mi804 <1576993271@qq.com> Date: Wed, 25 Feb 2026 18:06:02 +0800 Subject: [PATCH 2/4] support ltx2 train -2 --- README.md | 4 +- README_zh.md | 4 +- diffsynth/core/data/operators.py | 17 ++++++ docs/en/Model_Details/LTX-2.md | 55 ++++++++++++++++++- docs/zh/Model_Details/LTX-2.md | 55 ++++++++++++++++++- .../LTX-2-T2AV-Camera-Control-Dolly-In.py | 1 + .../LTX-2-T2AV-Camera-Control-Dolly-Left.py | 1 + .../LTX-2-T2AV-Camera-Control-Dolly-Out.py | 1 + .../LTX-2-T2AV-Camera-Control-Dolly-Right.py | 1 + .../LTX-2-T2AV-Camera-Control-Jib-Down.py | 1 + .../LTX-2-T2AV-Camera-Control-Jib-Up.py | 1 + .../LTX-2-T2AV-Camera-Control-Static.py | 1 + .../LTX-2-T2AV-DistilledPipeline.py | 1 + .../model_inference/LTX-2-T2AV-TwoStage.py | 1 + .../LTX-2-T2AV-Camera-Control-Dolly-In.py | 1 + .../LTX-2-T2AV-Camera-Control-Dolly-Left.py | 1 + .../LTX-2-T2AV-Camera-Control-Dolly-Out.py | 1 + .../LTX-2-T2AV-Camera-Control-Dolly-Right.py | 1 + .../LTX-2-T2AV-Camera-Control-Jib-Down.py | 1 + .../LTX-2-T2AV-Camera-Control-Jib-Up.py | 1 + .../LTX-2-T2AV-Camera-Control-Static.py | 1 + .../LTX-2-T2AV-DistilledPipeline.py | 1 + .../LTX-2-T2AV-TwoStage.py | 1 + .../model_training/full/LTX-2-T2AV-splited.sh | 2 +- .../model_training/lora/LTX-2-T2AV-noaudio.sh | 8 +-- .../model_training/lora/LTX-2-T2AV-splited.sh | 22 +++++++- .../ltx2/model_training/lora/LTX-2-T2AV.sh | 19 ------- .../scripts/split_model_statedicts.py | 0 examples/ltx2/model_training/train.py | 2 +- .../validate_full/LTX-2-T2AV.py | 2 +- .../validate_lora/LTX-2-T2AV.py | 3 +- .../validate_lora/LTX-2-T2AV_noaudio.py | 3 +- 32 files changed, 175 insertions(+), 39 deletions(-) delete mode 100644 examples/ltx2/model_training/lora/LTX-2-T2AV.sh rename examples/ltx2/{ => model_training}/scripts/split_model_statedicts.py (100%) diff --git a/README.md b/README.md index 2bc45e6..f266af1 100644 --- a/README.md +++ b/README.md @@ -32,7 +32,7 @@ We believe that a well-developed open-source code framework can lower the thresh > DiffSynth-Studio has undergone major version updates, and some old features are no longer maintained. If you need to use old features, please switch to the [last historical version](https://github.com/modelscope/DiffSynth-Studio/tree/afd101f3452c9ecae0c87b79adfa2e22d65ffdc3) before the major version update. > Currently, the development personnel of this project are limited, with most of the work handled by [Artiprocher](https://github.com/Artiprocher). Therefore, the progress of new feature development will be relatively slow, and the speed of responding to and resolving issues is limited. We apologize for this and ask developers to understand. -- **February 10, 2026** Added inference support for the LTX-2 audio-video generation model. See the documentation for details. Support for model training will be implemented in the future. +- **February 10, 2026** Added inference and training support for the LTX-2 audio-video generation model. See the documentation for details. - **February 2, 2026** The first document of the Research Tutorial series is now available, guiding you through training a small 0.1B text-to-image model from scratch. For details, see the [documentation](/docs/en/Research_Tutorial/train_from_scratch.md) and [model](https://modelscope.cn/models/DiffSynth-Studio/AAAMyModel). We hope DiffSynth-Studio can evolve into a more powerful training framework for Diffusion models. @@ -614,7 +614,7 @@ Example code for LTX-2 is available at: [/examples/ltx2/](/examples/ltx2/) | Model ID | Extra Args | Inference | Low-VRAM Inference | Full Training | Full Training Validation | LoRA Training | LoRA Training Validation | |-|-|-|-|-|-|-|-| -|[Lightricks/LTX-2: OneStagePipeline-T2AV](https://www.modelscope.cn/models/Lightricks/LTX-2)||[code](/examples/ltx2/model_inference/LTX-2-T2AV-OneStage.py)|[code](/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-OneStage.py)|-|-|-|-| +|[Lightricks/LTX-2: OneStagePipeline-T2AV](https://www.modelscope.cn/models/Lightricks/LTX-2)||[code](/examples/ltx2/model_inference/LTX-2-T2AV-OneStage.py)|[code](/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-OneStage.py)|[code](https://github.com/modelscope/DiffSynth-Studio/blob/main/examples/ltx2/model_training/full/LTX-2-T2AV-splited.sh)|[code](https://github.com/modelscope/DiffSynth-Studio/blob/main/examples/ltx2/model_training/validate_full/LTX-2-T2AV.py)|[code](https://github.com/modelscope/DiffSynth-Studio/blob/main/examples/ltx2/model_training/lora/LTX-2-T2AV-splited.sh)|[code](https://github.com/modelscope/DiffSynth-Studio/blob/main/examples/ltx2/model_training/validate_lora/LTX-2-T2AV.py)| |[Lightricks/LTX-2: TwoStagePipeline-T2AV](https://www.modelscope.cn/models/Lightricks/LTX-2)||[code](/examples/ltx2/model_inference/LTX-2-T2AV-TwoStage.py)|[code](/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-TwoStage.py)|-|-|-|-| |[Lightricks/LTX-2: DistilledPipeline-T2AV](https://www.modelscope.cn/models/Lightricks/LTX-2)||[code](/examples/ltx2/model_inference/LTX-2-T2AV-DistilledPipeline.py)|[code](/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-DistilledPipeline.py)|-|-|-|-| |[Lightricks/LTX-2: OneStagePipeline-I2AV](https://www.modelscope.cn/models/Lightricks/LTX-2)|`input_images`|[code](/examples/ltx2/model_inference/LTX-2-I2AV-OneStage.py)|[code](/examples/ltx2/model_inference_low_vram/LTX-2-I2AV-OneStage.py)|-|-|-|-| diff --git a/README_zh.md b/README_zh.md index 72c80d5..93ec2cf 100644 --- a/README_zh.md +++ b/README_zh.md @@ -32,7 +32,7 @@ DiffSynth 目前包括两个开源项目： > DiffSynth-Studio 经历了大版本更新，部分旧功能已停止维护，如需使用旧版功能，请切换到大版本更新前的[最后一个历史版本](https://github.com/modelscope/DiffSynth-Studio/tree/afd101f3452c9ecae0c87b79adfa2e22d65ffdc3)。 > 目前本项目的开发人员有限，大部分工作由 [Artiprocher](https://github.com/Artiprocher) 负责，因此新功能的开发进展会比较缓慢，issue 的回复和解决速度有限，我们对此感到非常抱歉，请各位开发者理解。 -- **2026年2月10日** 新增对[LTX-2](https://www.modelscope.cn/models/Lightricks/LTX-2)音视频生成模型的推理支持，详见[文档](docs/zh/Model_Details/LTX-2.md)，后续将推进模型训练的支持。 +- **2026年2月10日** 新增对[LTX-2](https://www.modelscope.cn/models/Lightricks/LTX-2)音视频生成模型的推理和训练支持，详见[文档](docs/zh/Model_Details/LTX-2.md)。 - **2026年2月2日** Research Tutorial 的第一篇文档上线，带你从零开始训练一个 0.1B 的小型文生图模型，详见[文档](/docs/zh/Research_Tutorial/train_from_scratch.md)、[模型](https://modelscope.cn/models/DiffSynth-Studio/AAAMyModel)，我们希望 DiffSynth-Studio 能够成为一个更强大的 Diffusion 模型训练框架。 @@ -614,7 +614,7 @@ LTX-2 的示例代码位于：[/examples/ltx2/](/examples/ltx2/) |模型 ID|额外参数|推理|低显存推理|全量训练|全量训练后验证|LoRA 训练|LoRA 训练后验证| |-|-|-|-|-|-|-|-| -|[Lightricks/LTX-2: OneStagePipeline-T2AV](https://www.modelscope.cn/models/Lightricks/LTX-2)||[code](/examples/ltx2/model_inference/LTX-2-T2AV-OneStage.py)|[code](/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-OneStage.py)|-|-|-|-| +|[Lightricks/LTX-2: OneStagePipeline-T2AV](https://www.modelscope.cn/models/Lightricks/LTX-2)||[code](/examples/ltx2/model_inference/LTX-2-T2AV-OneStage.py)|[code](/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-OneStage.py)|[code](https://github.com/modelscope/DiffSynth-Studio/blob/main/examples/ltx2/model_training/full/LTX-2-T2AV-splited.sh)|[code](https://github.com/modelscope/DiffSynth-Studio/blob/main/examples/ltx2/model_training/validate_full/LTX-2-T2AV.py)|[code](https://github.com/modelscope/DiffSynth-Studio/blob/main/examples/ltx2/model_training/lora/LTX-2-T2AV-splited.sh)|[code](https://github.com/modelscope/DiffSynth-Studio/blob/main/examples/ltx2/model_training/validate_lora/LTX-2-T2AV.py)| |[Lightricks/LTX-2: TwoStagePipeline-T2AV](https://www.modelscope.cn/models/Lightricks/LTX-2)||[code](/examples/ltx2/model_inference/LTX-2-T2AV-TwoStage.py)|[code](/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-TwoStage.py)|-|-|-|-| |[Lightricks/LTX-2: DistilledPipeline-T2AV](https://www.modelscope.cn/models/Lightricks/LTX-2)||[code](/examples/ltx2/model_inference/LTX-2-T2AV-DistilledPipeline.py)|[code](/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-DistilledPipeline.py)|-|-|-|-| |[Lightricks/LTX-2: OneStagePipeline-I2AV](https://www.modelscope.cn/models/Lightricks/LTX-2)|`input_images`|[code](/examples/ltx2/model_inference/LTX-2-I2AV-OneStage.py)|[code](/examples/ltx2/model_inference_low_vram/LTX-2-I2AV-OneStage.py)|-|-|-|-| diff --git a/diffsynth/core/data/operators.py b/diffsynth/core/data/operators.py index b7a0e7e..36756d3 100644 --- a/diffsynth/core/data/operators.py +++ b/diffsynth/core/data/operators.py @@ -218,3 +218,20 @@ class LoadAudio(DataProcessingOperator): import librosa input_audio, sample_rate = librosa.load(data, sr=self.sr) return input_audio + + +class LoadAudioWithTorchaudio(DataProcessingOperator): + def __init__(self, duration=5): + self.duration = duration + + def __call__(self, data: str): + import torchaudio + waveform, sample_rate = torchaudio.load(data) + target_samples = int(self.duration * sample_rate) + current_samples = waveform.shape[-1] + if current_samples > target_samples: + waveform = waveform[..., :target_samples] + elif current_samples < target_samples: + padding = target_samples - current_samples + waveform = torch.nn.functional.pad(waveform, (0, padding)) + return waveform, sample_rate diff --git a/docs/en/Model_Details/LTX-2.md b/docs/en/Model_Details/LTX-2.md index c285a7f..e3f4ca8 100644 --- a/docs/en/Model_Details/LTX-2.md +++ b/docs/en/Model_Details/LTX-2.md @@ -67,7 +67,7 @@ write_video_audio_ltx2( ## Model Overview |Model ID|Additional Parameters|Inference|Low VRAM Inference|Full Training|Validation After Full Training|LoRA Training|Validation After LoRA Training| |-|-|-|-|-|-|-|-| -|[Lightricks/LTX-2: OneStagePipeline-T2AV](https://www.modelscope.cn/models/Lightricks/LTX-2)||[code](https://github.com/modelscope/DiffSynth-Studio/blob/main/examples/ltx2/model_inference/LTX-2-T2AV-OneStage.py)|[code](https://github.com/modelscope/DiffSynth-Studio/blob/main/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-OneStage.py)|-|-|-|-| +|[Lightricks/LTX-2: OneStagePipeline-T2AV](https://www.modelscope.cn/models/Lightricks/LTX-2)||[code](https://github.com/modelscope/DiffSynth-Studio/blob/main/examples/ltx2/model_inference/LTX-2-T2AV-OneStage.py)|[code](https://github.com/modelscope/DiffSynth-Studio/blob/main/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-OneStage.py)|[code](https://github.com/modelscope/DiffSynth-Studio/blob/main/examples/ltx2/model_training/full/LTX-2-T2AV-splited.sh)|[code](https://github.com/modelscope/DiffSynth-Studio/blob/main/examples/ltx2/model_training/validate_full/LTX-2-T2AV.py)|[code](https://github.com/modelscope/DiffSynth-Studio/blob/main/examples/ltx2/model_training/lora/LTX-2-T2AV-splited.sh)|[code](https://github.com/modelscope/DiffSynth-Studio/blob/main/examples/ltx2/model_training/validate_lora/LTX-2-T2AV.py)| |[Lightricks/LTX-2: TwoStagePipeline-T2AV](https://www.modelscope.cn/models/Lightricks/LTX-2)||[code](https://github.com/modelscope/DiffSynth-Studio/blob/main/examples/ltx2/model_inference/LTX-2-T2AV-TwoStage.py)|[code](https://github.com/modelscope/DiffSynth-Studio/blob/main/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-TwoStage.py)|-|-|-|-| |[Lightricks/LTX-2: DistilledPipeline-T2AV](https://www.modelscope.cn/models/Lightricks/LTX-2)||[code](https://github.com/modelscope/DiffSynth-Studio/blob/main/examples/ltx2/model_inference/LTX-2-T2AV-DistilledPipeline.py)|[code](https://github.com/modelscope/DiffSynth-Studio/blob/main/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-DistilledPipeline.py)|-|-|-|-| |[Lightricks/LTX-2: OneStagePipeline-I2AV](https://www.modelscope.cn/models/Lightricks/LTX-2)|`input_images`|[code](https://github.com/modelscope/DiffSynth-Studio/blob/main/examples/ltx2/model_inference/LTX-2-I2AV-OneStage.py)|[code](https://github.com/modelscope/DiffSynth-Studio/blob/main/examples/ltx2/model_inference_low_vram/LTX-2-I2AV-OneStage.py)|-|-|-|-| @@ -113,4 +113,55 @@ If VRAM is insufficient, please enable [VRAM Management](../Pipeline_Usage/VRAM_ ## Model Training -The LTX-2 series models currently do not support training functionality. We will add related support as soon as possible. +LTX-2 series models are uniformly trained through [`examples/ltx2/model_training/train.py`](https://github.com/modelscope/DiffSynth-Studio/blob/main/examples/ltx2/model_training/train.py), and the script parameters include: + +* General Training Parameters + * Dataset Basic Configuration + * `--dataset_base_path`: Root directory of the dataset. + * `--dataset_metadata_path`: Metadata file path of the dataset. + * `--dataset_repeat`: Number of times the dataset is repeated in each epoch. + * `--dataset_num_workers`: Number of processes for each DataLoader. + * `--data_file_keys`: Field names to be loaded from metadata, usually image or video file paths, separated by `,`. + * Model Loading Configuration + * `--model_paths`: Paths of models to be loaded. JSON format. + * `--model_id_with_origin_paths`: Model IDs with original paths, e.g., `"Wan-AI/Wan2.1-T2V-1.3B:diffusion_pytorch_model*.safetensors"`. Separated by commas. + * `--extra_inputs`: Extra input parameters required by the model Pipeline, e.g., extra parameters when training image editing models, separated by `,`. + * `--fp8_models`: Models loaded in FP8 format, consistent with `--model_paths` or `--model_id_with_origin_paths` format. Currently only supports models whose parameters are not updated by gradients (no gradient backpropagation, or gradients only update their LoRA). + * Training Basic Configuration + * `--learning_rate`: Learning rate. + * `--num_epochs`: Number of epochs. + * `--trainable_models`: Trainable models, e.g., `dit`, `vae`, `text_encoder`. + * `--find_unused_parameters`: Whether there are unused parameters in DDP training. Some models contain redundant parameters that do not participate in gradient calculation, and this setting needs to be enabled to avoid errors in multi-GPU training. + * `--weight_decay`: Weight decay size, see [torch.optim.AdamW](https://docs.pytorch.org/docs/stable/generated/torch.optim.AdamW.html). + * `--task`: Training task, default is `sft`. Some models support more training modes, please refer to the documentation of each specific model. + * Output Configuration + * `--output_path`: Model saving path. + * `--remove_prefix_in_ckpt`: Remove prefix in the state dict of the model file. + * `--save_steps`: Interval of training steps to save the model. If this parameter is left blank, the model is saved once per epoch. + * LoRA Configuration + * `--lora_base_model`: Which model to add LoRA to. + * `--lora_target_modules`: Which layers to add LoRA to. + * `--lora_rank`: Rank of LoRA. + * `--lora_checkpoint`: Path of the LoRA checkpoint. If this path is provided, LoRA will be loaded from this checkpoint. + * `--preset_lora_path`: Preset LoRA checkpoint path. If this path is provided, this LoRA will be loaded in the form of being merged into the base model. This parameter is used for LoRA differential training. + * `--preset_lora_model`: Model that the preset LoRA is merged into, e.g., `dit`. + * Gradient Configuration + * `--use_gradient_checkpointing`: Whether to enable gradient checkpointing. + * `--use_gradient_checkpointing_offload`: Whether to offload gradient checkpointing to memory. + * `--gradient_accumulation_steps`: Number of gradient accumulation steps. + * Video Width/Height Configuration + * `--height`: Height of the video. Leave `height` and `width` blank to enable dynamic resolution. + * `--width`: Width of the video. Leave `height` and `width` blank to enable dynamic resolution. + * `--max_pixels`: Maximum pixel area of video frames. When dynamic resolution is enabled, video frames with resolution larger than this value will be downscaled, and video frames with resolution smaller than this value will remain unchanged. + * `--num_frames`: Number of frames in the video. +* LTX-2 Series Specific Parameters + * `--tokenizer_path`: Path of the tokenizer, applicable to text-to-video models, leave blank to automatically download from remote. + * `--frame_rate`: frame rate of the training videos. + +We have built a sample video dataset for your testing. You can download this dataset with the following command: + +```shell +modelscope download --dataset DiffSynth-Studio/example_video_dataset --local_dir ./data/example_video_dataset +``` + +We have written recommended training scripts for each model, please refer to the table in the "Model Overview" section above. For how to write model training scripts, please refer to [Model Training](../Pipeline_Usage/Model_Training.md); for more advanced training algorithms, please refer to [Training Framework Detailed Explanation](https://github.com/modelscope/DiffSynth-Studio/tree/main/docs/en/Training/). diff --git a/docs/zh/Model_Details/LTX-2.md b/docs/zh/Model_Details/LTX-2.md index 6961931..86abbcd 100644 --- a/docs/zh/Model_Details/LTX-2.md +++ b/docs/zh/Model_Details/LTX-2.md @@ -67,7 +67,7 @@ write_video_audio_ltx2( ## 模型总览 |模型 ID|额外参数|推理|低显存推理|全量训练|全量训练后验证|LoRA 训练|LoRA 训练后验证| |-|-|-|-|-|-|-|-| -|[Lightricks/LTX-2: OneStagePipeline-T2AV](https://www.modelscope.cn/models/Lightricks/LTX-2)||[code](https://github.com/modelscope/DiffSynth-Studio/blob/main/examples/ltx2/model_inference/LTX-2-T2AV-OneStage.py)|[code](https://github.com/modelscope/DiffSynth-Studio/blob/main/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-OneStage.py)|-|-|-|-| +|[Lightricks/LTX-2: OneStagePipeline-T2AV](https://www.modelscope.cn/models/Lightricks/LTX-2)||[code](https://github.com/modelscope/DiffSynth-Studio/blob/main/examples/ltx2/model_inference/LTX-2-T2AV-OneStage.py)|[code](https://github.com/modelscope/DiffSynth-Studio/blob/main/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-OneStage.py)|[code](https://github.com/modelscope/DiffSynth-Studio/blob/main/examples/ltx2/model_training/full/LTX-2-T2AV-splited.sh)|[code](https://github.com/modelscope/DiffSynth-Studio/blob/main/examples/ltx2/model_training/validate_full/LTX-2-T2AV.py)|[code](https://github.com/modelscope/DiffSynth-Studio/blob/main/examples/ltx2/model_training/lora/LTX-2-T2AV-splited.sh)|[code](https://github.com/modelscope/DiffSynth-Studio/blob/main/examples/ltx2/model_training/validate_lora/LTX-2-T2AV.py)| |[Lightricks/LTX-2: TwoStagePipeline-T2AV](https://www.modelscope.cn/models/Lightricks/LTX-2)||[code](https://github.com/modelscope/DiffSynth-Studio/blob/main/examples/ltx2/model_inference/LTX-2-T2AV-TwoStage.py)|[code](https://github.com/modelscope/DiffSynth-Studio/blob/main/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-TwoStage.py)|-|-|-|-| |[Lightricks/LTX-2: DistilledPipeline-T2AV](https://www.modelscope.cn/models/Lightricks/LTX-2)||[code](https://github.com/modelscope/DiffSynth-Studio/blob/main/examples/ltx2/model_inference/LTX-2-T2AV-DistilledPipeline.py)|[code](https://github.com/modelscope/DiffSynth-Studio/blob/main/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-DistilledPipeline.py)|-|-|-|-| |[Lightricks/LTX-2: OneStagePipeline-I2AV](https://www.modelscope.cn/models/Lightricks/LTX-2)|`input_images`|[code](https://github.com/modelscope/DiffSynth-Studio/blob/main/examples/ltx2/model_inference/LTX-2-I2AV-OneStage.py)|[code](https://github.com/modelscope/DiffSynth-Studio/blob/main/examples/ltx2/model_inference_low_vram/LTX-2-I2AV-OneStage.py)|-|-|-|-| @@ -113,4 +113,55 @@ write_video_audio_ltx2( ## 模型训练 -LTX-2 系列模型目前暂不支持训练功能。我们将尽快添加相关支持。 +LTX-2 系列模型统一通过 [`examples/ltx2/model_training/train.py`](https://github.com/modelscope/DiffSynth-Studio/blob/main/examples/ltx2/model_training/train.py) 进行训练，脚本的参数包括： + +* 通用训练参数 + * 数据集基础配置 + * `--dataset_base_path`: 数据集的根目录。 + * `--dataset_metadata_path`: 数据集的元数据文件路径。 + * `--dataset_repeat`: 每个 epoch 中数据集重复的次数。 + * `--dataset_num_workers`: 每个 Dataloder 的进程数量。 + * `--data_file_keys`: 元数据中需要加载的字段名称，通常是图像或视频文件的路径，以 `,` 分隔。 + * 模型加载配置 + * `--model_paths`: 要加载的模型路径。JSON 格式。 + * `--model_id_with_origin_paths`: 带原始路径的模型 ID，例如 `"Wan-AI/Wan2.1-T2V-1.3B:diffusion_pytorch_model*.safetensors"`。用逗号分隔。 + * `--extra_inputs`: 模型 Pipeline 所需的额外输入参数，例如训练图像编辑模型时需要额外参数，以 `,` 分隔。 + * `--fp8_models`：以 FP8 格式加载的模型，格式与 `--model_paths` 或 `--model_id_with_origin_paths` 一致，目前仅支持参数不被梯度更新的模型（不需要梯度回传，或梯度仅更新其 LoRA）。 + * 训练基础配置 + * `--learning_rate`: 学习率。 + * `--num_epochs`: 轮数（Epoch）。 + * `--trainable_models`: 可训练的模型，例如 `dit`、`vae`、`text_encoder`。 + * `--find_unused_parameters`: DDP 训练中是否存在未使用的参数，少数模型包含不参与梯度计算的冗余参数，需开启这一设置避免在多 GPU 训练中报错。 + * `--weight_decay`：权重衰减大小，详见 [torch.optim.AdamW](https://docs.pytorch.org/docs/stable/generated/torch.optim.AdamW.html)。 + * `--task`: 训练任务，默认为 `sft`，部分模型支持更多训练模式，请参考每个特定模型的文档。 + * 输出配置 + * `--output_path`: 模型保存路径。 + * `--remove_prefix_in_ckpt`: 在模型文件的 state dict 中移除前缀。 + * `--save_steps`: 保存模型的训练步数间隔，若此参数留空，则每个 epoch 保存一次。 + * LoRA 配置 + * `--lora_base_model`: LoRA 添加到哪个模型上。 + * `--lora_target_modules`: LoRA 添加到哪些层上。 + * `--lora_rank`: LoRA 的秩（Rank）。 + * `--lora_checkpoint`: LoRA 检查点的路径。如果提供此路径，LoRA 将从此检查点加载。 + * `--preset_lora_path`: 预置 LoRA 检查点路径，如果提供此路径，这一 LoRA 将会以融入基础模型的形式加载。此参数用于 LoRA 差分训练。 + * `--preset_lora_model`: 预置 LoRA 融入的模型，例如 `dit`。 + * 梯度配置 + * `--use_gradient_checkpointing`: 是否启用 gradient checkpointing。 + * `--use_gradient_checkpointing_offload`: 是否将 gradient checkpointing 卸载到内存中。 + * `--gradient_accumulation_steps`: 梯度累积步数。 + * 视频宽高配置 + * `--height`: 视频的高度。将 `height` 和 `width` 留空以启用动态分辨率。 + * `--width`: 视频的宽度。将 `height` 和 `width` 留空以启用动态分辨率。 + * `--max_pixels`: 视频帧的最大像素面积，当启用动态分辨率时，分辨率大于这个数值的视频帧都会被缩小，分辨率小于这个数值的视频帧保持不变。 + * `--num_frames`: 视频的帧数。 +* LTX-2 系列特定参数 + * `--tokenizer_path`: 分词器路径，适用于文生视频模型，留空则从远程自动下载。 + * `--frame_rate`: 训练视频的帧率。 + +我们构建了一个样例视频数据集，以方便您进行测试，通过以下命令可以下载这个数据集： + +```shell +modelscope download --dataset DiffSynth-Studio/example_video_dataset --local_dir ./data/example_video_dataset +``` + +我们为每个模型编写了推荐的训练脚本，请参考前文"模型总览"中的表格。关于如何编写模型训练脚本，请参考[模型训练](../Pipeline_Usage/Model_Training.md)；更多高阶训练算法，请参考[训练框架详解](https://github.com/modelscope/DiffSynth-Studio/tree/main/docs/zh/Training/)。 diff --git a/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Dolly-In.py b/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Dolly-In.py index da454dc..9932804 100644 --- a/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Dolly-In.py +++ b/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Dolly-In.py @@ -22,6 +22,7 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_encoder.safetensors", **vram_config), ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-spatial-upscaler-x2-1.0.safetensors", **vram_config), ], tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), diff --git a/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Dolly-Left.py b/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Dolly-Left.py index e6dd9ba..4f0f273 100644 --- a/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Dolly-Left.py +++ b/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Dolly-Left.py @@ -22,6 +22,7 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_encoder.safetensors", **vram_config), ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-spatial-upscaler-x2-1.0.safetensors", **vram_config), ], tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), diff --git a/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Dolly-Out.py b/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Dolly-Out.py index a84efe9..3fdd2c3 100644 --- a/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Dolly-Out.py +++ b/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Dolly-Out.py @@ -22,6 +22,7 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_encoder.safetensors", **vram_config), ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-spatial-upscaler-x2-1.0.safetensors", **vram_config), ], tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), diff --git a/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Dolly-Right.py b/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Dolly-Right.py index 0a2e968..b4fd099 100644 --- a/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Dolly-Right.py +++ b/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Dolly-Right.py @@ -22,6 +22,7 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_encoder.safetensors", **vram_config), ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-spatial-upscaler-x2-1.0.safetensors", **vram_config), ], tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), diff --git a/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Jib-Down.py b/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Jib-Down.py index 4adfdec..1eb2ace 100644 --- a/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Jib-Down.py +++ b/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Jib-Down.py @@ -22,6 +22,7 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_encoder.safetensors", **vram_config), ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-spatial-upscaler-x2-1.0.safetensors", **vram_config), ], tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), diff --git a/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Jib-Up.py b/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Jib-Up.py index 6dfa664..cc6ac39 100644 --- a/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Jib-Up.py +++ b/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Jib-Up.py @@ -22,6 +22,7 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_encoder.safetensors", **vram_config), ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-spatial-upscaler-x2-1.0.safetensors", **vram_config), ], tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), diff --git a/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Static.py b/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Static.py index f351789..2299bf7 100644 --- a/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Static.py +++ b/examples/ltx2/model_inference/LTX-2-T2AV-Camera-Control-Static.py @@ -22,6 +22,7 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_encoder.safetensors", **vram_config), ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-spatial-upscaler-x2-1.0.safetensors", **vram_config), ], tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), diff --git a/examples/ltx2/model_inference/LTX-2-T2AV-DistilledPipeline.py b/examples/ltx2/model_inference/LTX-2-T2AV-DistilledPipeline.py index 5b57aea..78b537b 100644 --- a/examples/ltx2/model_inference/LTX-2-T2AV-DistilledPipeline.py +++ b/examples/ltx2/model_inference/LTX-2-T2AV-DistilledPipeline.py @@ -22,6 +22,7 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_encoder.safetensors", **vram_config), ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-spatial-upscaler-x2-1.0.safetensors", **vram_config), ], tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), diff --git a/examples/ltx2/model_inference/LTX-2-T2AV-TwoStage.py b/examples/ltx2/model_inference/LTX-2-T2AV-TwoStage.py index 065e92c..9a85a9f 100644 --- a/examples/ltx2/model_inference/LTX-2-T2AV-TwoStage.py +++ b/examples/ltx2/model_inference/LTX-2-T2AV-TwoStage.py @@ -22,6 +22,7 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_encoder.safetensors", **vram_config), ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-spatial-upscaler-x2-1.0.safetensors", **vram_config), ], tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), diff --git a/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Dolly-In.py b/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Dolly-In.py index 9f6af9f..0454dfd 100644 --- a/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Dolly-In.py +++ b/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Dolly-In.py @@ -22,6 +22,7 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_encoder.safetensors", **vram_config), ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-spatial-upscaler-x2-1.0.safetensors", **vram_config), ], tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), diff --git a/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Dolly-Left.py b/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Dolly-Left.py index 8ee5fa9..e66a203 100644 --- a/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Dolly-Left.py +++ b/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Dolly-Left.py @@ -22,6 +22,7 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_encoder.safetensors", **vram_config), ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-spatial-upscaler-x2-1.0.safetensors", **vram_config), ], tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), diff --git a/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Dolly-Out.py b/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Dolly-Out.py index 3bb9f50..94af241 100644 --- a/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Dolly-Out.py +++ b/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Dolly-Out.py @@ -22,6 +22,7 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_encoder.safetensors", **vram_config), ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-spatial-upscaler-x2-1.0.safetensors", **vram_config), ], tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), diff --git a/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Dolly-Right.py b/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Dolly-Right.py index 0301445..b713170 100644 --- a/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Dolly-Right.py +++ b/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Dolly-Right.py @@ -22,6 +22,7 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_encoder.safetensors", **vram_config), ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-spatial-upscaler-x2-1.0.safetensors", **vram_config), ], tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), diff --git a/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Jib-Down.py b/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Jib-Down.py index d1a8db3..fe1423c 100644 --- a/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Jib-Down.py +++ b/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Jib-Down.py @@ -22,6 +22,7 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_encoder.safetensors", **vram_config), ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-spatial-upscaler-x2-1.0.safetensors", **vram_config), ], tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), diff --git a/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Jib-Up.py b/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Jib-Up.py index 0ae8041..b6229c1 100644 --- a/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Jib-Up.py +++ b/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Jib-Up.py @@ -22,6 +22,7 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_encoder.safetensors", **vram_config), ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-spatial-upscaler-x2-1.0.safetensors", **vram_config), ], tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), diff --git a/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Static.py b/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Static.py index f8130ae..13d8ea9 100644 --- a/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Static.py +++ b/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-Camera-Control-Static.py @@ -22,6 +22,7 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_encoder.safetensors", **vram_config), ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-spatial-upscaler-x2-1.0.safetensors", **vram_config), ], tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), diff --git a/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-DistilledPipeline.py b/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-DistilledPipeline.py index 836077b..daf82f2 100644 --- a/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-DistilledPipeline.py +++ b/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-DistilledPipeline.py @@ -22,6 +22,7 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_encoder.safetensors", **vram_config), ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-spatial-upscaler-x2-1.0.safetensors", **vram_config), ], tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), diff --git a/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-TwoStage.py b/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-TwoStage.py index f513f46..a10b6ef 100644 --- a/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-TwoStage.py +++ b/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-TwoStage.py @@ -22,6 +22,7 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_encoder.safetensors", **vram_config), ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-spatial-upscaler-x2-1.0.safetensors", **vram_config), ], tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), diff --git a/examples/ltx2/model_training/full/LTX-2-T2AV-splited.sh b/examples/ltx2/model_training/full/LTX-2-T2AV-splited.sh index 8dfdd4b..4aec5f8 100644 --- a/examples/ltx2/model_training/full/LTX-2-T2AV-splited.sh +++ b/examples/ltx2/model_training/full/LTX-2-T2AV-splited.sh @@ -17,7 +17,7 @@ accelerate launch examples/ltx2/model_training/train.py \ --use_gradient_checkpointing \ --task "sft:data_process" -accelerate launch examples/ltx2/model_training/train.py \ +accelerate launch --config_file examples/qwen_image/model_training/full/accelerate_config_zero2offload.yaml examples/ltx2/model_training/train.py \ --dataset_base_path ./models/train/LTX2-T2AV-full-splited-cache \ --data_file_keys "video,input_audio" \ --extra_inputs "input_audio" \ diff --git a/examples/ltx2/model_training/lora/LTX-2-T2AV-noaudio.sh b/examples/ltx2/model_training/lora/LTX-2-T2AV-noaudio.sh index 4f1e754..b2a5609 100644 --- a/examples/ltx2/model_training/lora/LTX-2-T2AV-noaudio.sh +++ b/examples/ltx2/model_training/lora/LTX-2-T2AV-noaudio.sh @@ -22,8 +22,8 @@ accelerate launch examples/ltx2/model_training/train.py \ --dataset_base_path data/example_video_dataset/ltx2 \ --dataset_metadata_path data/example_video_dataset/ltx2_t2av.csv \ - --height 256 \ - --width 384 \ + --height 512 \ + --width 768 \ --num_frames 49\ --dataset_repeat 1 \ --model_id_with_origin_paths "DiffSynth-Studio/LTX-2-Repackage:text_encoder_post_modules.safetensors,DiffSynth-Studio/LTX-2-Repackage:video_vae_encoder.safetensors,DiffSynth-Studio/LTX-2-Repackage:audio_vae_encoder.safetensors,google/gemma-3-12b-it-qat-q4_0-unquantized:model-*.safetensors" \ @@ -40,8 +40,8 @@ accelerate launch examples/ltx2/model_training/train.py \ accelerate launch examples/ltx2/model_training/train.py \ --dataset_base_path ./models/train/LTX2-T2AV-noaudio_lora-splited-cache \ - --height 256 \ - --width 384 \ + --height 512 \ + --width 768 \ --num_frames 49\ --dataset_repeat 100 \ --model_id_with_origin_paths "DiffSynth-Studio/LTX-2-Repackage:transformer.safetensors" \ diff --git a/examples/ltx2/model_training/lora/LTX-2-T2AV-splited.sh b/examples/ltx2/model_training/lora/LTX-2-T2AV-splited.sh index e71494a..40dae1a 100644 --- a/examples/ltx2/model_training/lora/LTX-2-T2AV-splited.sh +++ b/examples/ltx2/model_training/lora/LTX-2-T2AV-splited.sh @@ -1,3 +1,24 @@ +# Single Stage Training not recommended for T2AV due to the large memory consumption. Please use the Splited Training instead. +# accelerate launch examples/ltx2/model_training/train.py \ +# --dataset_base_path data/example_video_dataset/ltx2 \ +# --dataset_metadata_path data/example_video_dataset/ltx2_t2av.csv \ +# --data_file_keys "video,input_audio" \ +# --extra_inputs "input_audio" \ +# --height 256 \ +# --width 384 \ +# --num_frames 25\ +# --dataset_repeat 100 \ +# --model_id_with_origin_paths "DiffSynth-Studio/LTX-2-Repackage:transformer.safetensors,DiffSynth-Studio/LTX-2-Repackage:text_encoder_post_modules.safetensors,DiffSynth-Studio/LTX-2-Repackage:video_vae_encoder.safetensors,DiffSynth-Studio/LTX-2-Repackage:audio_vae_encoder.safetensors,google/gemma-3-12b-it-qat-q4_0-unquantized:model-*.safetensors" \ +# --learning_rate 1e-4 \ +# --num_epochs 5 \ +# --remove_prefix_in_ckpt "pipe.dit." \ +# --output_path "./models/train/LTX2-T2AV_lora" \ +# --lora_base_model "dit" \ +# --lora_target_modules "to_k,to_q,to_v,to_out.0" \ +# --lora_rank 32 \ +# --use_gradient_checkpointing \ +# --find_unused_parameters + # Splited Training accelerate launch examples/ltx2/model_training/train.py \ --dataset_base_path data/example_video_dataset/ltx2 \ @@ -19,7 +40,6 @@ accelerate launch examples/ltx2/model_training/train.py \ --use_gradient_checkpointing \ --task "sft:data_process" - accelerate launch examples/ltx2/model_training/train.py \ --dataset_base_path ./models/train/LTX2-T2AV_lora-splited-cache \ --data_file_keys "video,input_audio" \ diff --git a/examples/ltx2/model_training/lora/LTX-2-T2AV.sh b/examples/ltx2/model_training/lora/LTX-2-T2AV.sh deleted file mode 100644 index 1d8aaaf..0000000 --- a/examples/ltx2/model_training/lora/LTX-2-T2AV.sh +++ /dev/null @@ -1,19 +0,0 @@ -accelerate launch examples/ltx2/model_training/train.py \ - --dataset_base_path data/example_video_dataset/ltx2 \ - --dataset_metadata_path data/example_video_dataset/ltx2_t2av.csv \ - --data_file_keys "video,input_audio" \ - --extra_inputs "input_audio" \ - --height 256 \ - --width 384 \ - --num_frames 25\ - --dataset_repeat 100 \ - --model_id_with_origin_paths "DiffSynth-Studio/LTX-2-Repackage:transformer.safetensors,DiffSynth-Studio/LTX-2-Repackage:text_encoder_post_modules.safetensors,DiffSynth-Studio/LTX-2-Repackage:video_vae_encoder.safetensors,DiffSynth-Studio/LTX-2-Repackage:audio_vae_encoder.safetensors,google/gemma-3-12b-it-qat-q4_0-unquantized:model-*.safetensors" \ - --learning_rate 1e-4 \ - --num_epochs 5 \ - --remove_prefix_in_ckpt "pipe.dit." \ - --output_path "./models/train/LTX2-T2AV_lora" \ - --lora_base_model "dit" \ - --lora_target_modules "to_k,to_q,to_v,to_out.0" \ - --lora_rank 32 \ - --use_gradient_checkpointing \ - --find_unused_parameters diff --git a/examples/ltx2/scripts/split_model_statedicts.py b/examples/ltx2/model_training/scripts/split_model_statedicts.py similarity index 100% rename from examples/ltx2/scripts/split_model_statedicts.py rename to examples/ltx2/model_training/scripts/split_model_statedicts.py diff --git a/examples/ltx2/model_training/train.py b/examples/ltx2/model_training/train.py index 729b4ed..a994f7a 100644 --- a/examples/ltx2/model_training/train.py +++ b/examples/ltx2/model_training/train.py @@ -96,7 +96,7 @@ def ltx2_parser(): parser = add_general_config(parser) parser = add_video_size_config(parser) parser.add_argument("--tokenizer_path", type=str, default=None, help="Path to tokenizer.") - parser.add_argument("--frame_rate", type=float, default=24, help="frame rate of the training videos. If not specified, it will be determined by the dataset.") + parser.add_argument("--frame_rate", type=float, default=24, help="frame rate of the training videos.") parser.add_argument("--initialize_model_on_cpu", default=False, action="store_true", help="Whether to initialize models on CPU.") return parser diff --git a/examples/ltx2/model_training/validate_full/LTX-2-T2AV.py b/examples/ltx2/model_training/validate_full/LTX-2-T2AV.py index 6201ec1..a5da12d 100644 --- a/examples/ltx2/model_training/validate_full/LTX-2-T2AV.py +++ b/examples/ltx2/model_training/validate_full/LTX-2-T2AV.py @@ -27,7 +27,7 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( ) prompt = "A beautiful sunset over the ocean." negative_prompt = "blurry, out of focus, overexposed, underexposed, low contrast, washed out colors, excessive noise, grainy texture, poor lighting, flickering, motion blur, distorted proportions, unnatural skin tones, deformed facial features, asymmetrical face, missing facial features, extra limbs, disfigured hands, wrong hand count, artifacts around text, inconsistent perspective, camera shake, incorrect depth of field, background too sharp, background clutter, distracting reflections, harsh shadows, inconsistent lighting direction, color banding, cartoonish rendering, 3D CGI look, unrealistic materials, uncanny valley effect, incorrect ethnicity, wrong gender, exaggerated expressions, wrong gaze direction, mismatched lip sync, silent or muted audio, distorted voice, robotic voice, echo, background noise, off-sync audio, incorrect dialogue, added dialogue, repetitive speech, jittery movement, awkward pauses, incorrect timing, unnatural transitions, inconsistent framing, tilted camera, flat lighting, inconsistent tone, cinematic oversaturation, stylized filters, or AI artifacts." -height, width, num_frames = 512, 768, 121 +height, width, num_frames = 512, 768, 49 video, audio = pipe( prompt=prompt, negative_prompt=negative_prompt, diff --git a/examples/ltx2/model_training/validate_lora/LTX-2-T2AV.py b/examples/ltx2/model_training/validate_lora/LTX-2-T2AV.py index 4372b6c..471a901 100644 --- a/examples/ltx2/model_training/validate_lora/LTX-2-T2AV.py +++ b/examples/ltx2/model_training/validate_lora/LTX-2-T2AV.py @@ -28,8 +28,7 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( pipe.load_lora(pipe.dit, "models/train/LTX2-T2AV_lora/epoch-4.safetensors") prompt = "A beautiful sunset over the ocean." negative_prompt = "blurry, out of focus, overexposed, underexposed, low contrast, washed out colors, excessive noise, grainy texture, poor lighting, flickering, motion blur, distorted proportions, unnatural skin tones, deformed facial features, asymmetrical face, missing facial features, extra limbs, disfigured hands, wrong hand count, artifacts around text, inconsistent perspective, camera shake, incorrect depth of field, background too sharp, background clutter, distracting reflections, harsh shadows, inconsistent lighting direction, color banding, cartoonish rendering, 3D CGI look, unrealistic materials, uncanny valley effect, incorrect ethnicity, wrong gender, exaggerated expressions, wrong gaze direction, mismatched lip sync, silent or muted audio, distorted voice, robotic voice, echo, background noise, off-sync audio, incorrect dialogue, added dialogue, repetitive speech, jittery movement, awkward pauses, incorrect timing, unnatural transitions, inconsistent framing, tilted camera, flat lighting, inconsistent tone, cinematic oversaturation, stylized filters, or AI artifacts." -height, width, num_frames = 512, 768, 121 -height, width, num_frames = 256, 384, 25 +height, width, num_frames = 512, 768, 49 video, audio = pipe( prompt=prompt, negative_prompt=negative_prompt, diff --git a/examples/ltx2/model_training/validate_lora/LTX-2-T2AV_noaudio.py b/examples/ltx2/model_training/validate_lora/LTX-2-T2AV_noaudio.py index b35c23c..4c2bccc 100644 --- a/examples/ltx2/model_training/validate_lora/LTX-2-T2AV_noaudio.py +++ b/examples/ltx2/model_training/validate_lora/LTX-2-T2AV_noaudio.py @@ -28,8 +28,7 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( pipe.load_lora(pipe.dit, "models/train/LTX2-T2AV-noaudio_lora/epoch-4.safetensors") prompt = "A beautiful sunset over the ocean." negative_prompt = "blurry, out of focus, overexposed, underexposed, low contrast, washed out colors, excessive noise, grainy texture, poor lighting, flickering, motion blur, distorted proportions, unnatural skin tones, deformed facial features, asymmetrical face, missing facial features, extra limbs, disfigured hands, wrong hand count, artifacts around text, inconsistent perspective, camera shake, incorrect depth of field, background too sharp, background clutter, distracting reflections, harsh shadows, inconsistent lighting direction, color banding, cartoonish rendering, 3D CGI look, unrealistic materials, uncanny valley effect, incorrect ethnicity, wrong gender, exaggerated expressions, wrong gaze direction, mismatched lip sync, silent or muted audio, distorted voice, robotic voice, echo, background noise, off-sync audio, incorrect dialogue, added dialogue, repetitive speech, jittery movement, awkward pauses, incorrect timing, unnatural transitions, inconsistent framing, tilted camera, flat lighting, inconsistent tone, cinematic oversaturation, stylized filters, or AI artifacts." -height, width, num_frames = 512, 768, 121 -height, width, num_frames = 256, 384, 25 +height, width, num_frames = 512, 768, 49 video, audio = pipe( prompt=prompt, negative_prompt=negative_prompt, From 8d8bfc7f540e1493f7d6d229872b7186f7648bad Mon Sep 17 00:00:00 2001 From: mi804 <1576993271@qq.com> Date: Wed, 25 Feb 2026 19:04:10 +0800 Subject: [PATCH 3/4] minor fix --- README.md | 2 +- README_zh.md | 2 +- diffsynth/pipelines/ltx2_audio_video.py | 1 - examples/ltx2/model_inference/LTX-2-T2AV-OneStage.py | 9 --------- 4 files changed, 2 insertions(+), 12 deletions(-) diff --git a/README.md b/README.md index f266af1..e055ae7 100644 --- a/README.md +++ b/README.md @@ -614,7 +614,7 @@ Example code for LTX-2 is available at: [/examples/ltx2/](/examples/ltx2/) | Model ID | Extra Args | Inference | Low-VRAM Inference | Full Training | Full Training Validation | LoRA Training | LoRA Training Validation | |-|-|-|-|-|-|-|-| -|[Lightricks/LTX-2: OneStagePipeline-T2AV](https://www.modelscope.cn/models/Lightricks/LTX-2)||[code](/examples/ltx2/model_inference/LTX-2-T2AV-OneStage.py)|[code](/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-OneStage.py)|[code](https://github.com/modelscope/DiffSynth-Studio/blob/main/examples/ltx2/model_training/full/LTX-2-T2AV-splited.sh)|[code](https://github.com/modelscope/DiffSynth-Studio/blob/main/examples/ltx2/model_training/validate_full/LTX-2-T2AV.py)|[code](https://github.com/modelscope/DiffSynth-Studio/blob/main/examples/ltx2/model_training/lora/LTX-2-T2AV-splited.sh)|[code](https://github.com/modelscope/DiffSynth-Studio/blob/main/examples/ltx2/model_training/validate_lora/LTX-2-T2AV.py)| +|[Lightricks/LTX-2: OneStagePipeline-T2AV](https://www.modelscope.cn/models/Lightricks/LTX-2)||[code](/examples/ltx2/model_inference/LTX-2-T2AV-OneStage.py)|[code](/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-OneStage.py)|[code](/examples/ltx2/model_training/full/LTX-2-T2AV-splited.sh)|[code](/examples/ltx2/model_training/validate_full/LTX-2-T2AV.py)|[code](/examples/ltx2/model_training/lora/LTX-2-T2AV-splited.sh)|[code](/examples/ltx2/model_training/validate_lora/LTX-2-T2AV.py)| |[Lightricks/LTX-2: TwoStagePipeline-T2AV](https://www.modelscope.cn/models/Lightricks/LTX-2)||[code](/examples/ltx2/model_inference/LTX-2-T2AV-TwoStage.py)|[code](/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-TwoStage.py)|-|-|-|-| |[Lightricks/LTX-2: DistilledPipeline-T2AV](https://www.modelscope.cn/models/Lightricks/LTX-2)||[code](/examples/ltx2/model_inference/LTX-2-T2AV-DistilledPipeline.py)|[code](/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-DistilledPipeline.py)|-|-|-|-| |[Lightricks/LTX-2: OneStagePipeline-I2AV](https://www.modelscope.cn/models/Lightricks/LTX-2)|`input_images`|[code](/examples/ltx2/model_inference/LTX-2-I2AV-OneStage.py)|[code](/examples/ltx2/model_inference_low_vram/LTX-2-I2AV-OneStage.py)|-|-|-|-| diff --git a/README_zh.md b/README_zh.md index 93ec2cf..519ebe6 100644 --- a/README_zh.md +++ b/README_zh.md @@ -614,7 +614,7 @@ LTX-2 的示例代码位于：[/examples/ltx2/](/examples/ltx2/) |模型 ID|额外参数|推理|低显存推理|全量训练|全量训练后验证|LoRA 训练|LoRA 训练后验证| |-|-|-|-|-|-|-|-| -|[Lightricks/LTX-2: OneStagePipeline-T2AV](https://www.modelscope.cn/models/Lightricks/LTX-2)||[code](/examples/ltx2/model_inference/LTX-2-T2AV-OneStage.py)|[code](/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-OneStage.py)|[code](https://github.com/modelscope/DiffSynth-Studio/blob/main/examples/ltx2/model_training/full/LTX-2-T2AV-splited.sh)|[code](https://github.com/modelscope/DiffSynth-Studio/blob/main/examples/ltx2/model_training/validate_full/LTX-2-T2AV.py)|[code](https://github.com/modelscope/DiffSynth-Studio/blob/main/examples/ltx2/model_training/lora/LTX-2-T2AV-splited.sh)|[code](https://github.com/modelscope/DiffSynth-Studio/blob/main/examples/ltx2/model_training/validate_lora/LTX-2-T2AV.py)| +|[Lightricks/LTX-2: OneStagePipeline-T2AV](https://www.modelscope.cn/models/Lightricks/LTX-2)||[code](/examples/ltx2/model_inference/LTX-2-T2AV-OneStage.py)|[code](/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-OneStage.py)|[code](/examples/ltx2/model_training/full/LTX-2-T2AV-splited.sh)|[code](/examples/ltx2/model_training/validate_full/LTX-2-T2AV.py)|[code](/examples/ltx2/model_training/lora/LTX-2-T2AV-splited.sh)|[code](/examples/ltx2/model_training/validate_lora/LTX-2-T2AV.py)| |[Lightricks/LTX-2: TwoStagePipeline-T2AV](https://www.modelscope.cn/models/Lightricks/LTX-2)||[code](/examples/ltx2/model_inference/LTX-2-T2AV-TwoStage.py)|[code](/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-TwoStage.py)|-|-|-|-| |[Lightricks/LTX-2: DistilledPipeline-T2AV](https://www.modelscope.cn/models/Lightricks/LTX-2)||[code](/examples/ltx2/model_inference/LTX-2-T2AV-DistilledPipeline.py)|[code](/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-DistilledPipeline.py)|-|-|-|-| |[Lightricks/LTX-2: OneStagePipeline-I2AV](https://www.modelscope.cn/models/Lightricks/LTX-2)|`input_images`|[code](/examples/ltx2/model_inference/LTX-2-I2AV-OneStage.py)|[code](/examples/ltx2/model_inference_low_vram/LTX-2-I2AV-OneStage.py)|-|-|-|-| diff --git a/diffsynth/pipelines/ltx2_audio_video.py b/diffsynth/pipelines/ltx2_audio_video.py index 444e004..4587449 100644 --- a/diffsynth/pipelines/ltx2_audio_video.py +++ b/diffsynth/pipelines/ltx2_audio_video.py @@ -490,7 +490,6 @@ class LTX2AudioVideoUnit_InputAudioEmbedder(PipelineUnit): pipe.load_models_to_device(self.onload_model_names) input_audio = pipe.audio_processor.waveform_to_mel(input_audio.unsqueeze(0), waveform_sample_rate=sample_rate).to(dtype=pipe.torch_dtype) audio_input_latents = pipe.audio_vae_encoder(input_audio) - audio_noise = torch.randn_like(audio_input_latents) audio_latent_shape = AudioLatentShape.from_torch_shape(audio_input_latents.shape) audio_positions = pipe.audio_patchifier.get_patch_grid_bounds(audio_latent_shape, device=pipe.device) if pipe.scheduler.training: diff --git a/examples/ltx2/model_inference/LTX-2-T2AV-OneStage.py b/examples/ltx2/model_inference/LTX-2-T2AV-OneStage.py index 2f73f80..295d73b 100644 --- a/examples/ltx2/model_inference/LTX-2-T2AV-OneStage.py +++ b/examples/ltx2/model_inference/LTX-2-T2AV-OneStage.py @@ -44,12 +44,3 @@ write_video_audio_ltx2( fps=24, audio_sample_rate=24000, ) - - - # ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="transformer.safetensors", **vram_config), - # ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="text_encoder_post_modules.safetensors", **vram_config), - # ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_encoder.safetensors", **vram_config), - # ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), - # ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), - # ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_encoder.safetensors", **vram_config), - # ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), \ No newline at end of file From f48662e8637f29147e593e2ff6b6f1d31a1b1b87 Mon Sep 17 00:00:00 2001 From: mi804 <1576993271@qq.com> Date: Thu, 26 Feb 2026 11:10:00 +0800 Subject: [PATCH 4/4] update docs --- README.md | 34 +++++++++++- README_zh.md | 34 +++++++++++- diffsynth/configs/model_configs.py | 16 +++++- docs/en/Model_Details/LTX-2.md | 52 ++++++++++++++++-- docs/zh/Model_Details/LTX-2.md | 54 +++++++++++++++++-- .../model_inference/LTX-2-T2AV-OneStage.py | 19 +++++++ .../model_inference/LTX-2-T2AV-TwoStage.py | 22 +++++++- .../LTX-2-T2AV-OneStage.py | 20 +++++++ .../LTX-2-T2AV-TwoStage.py | 22 ++++++++ 9 files changed, 258 insertions(+), 15 deletions(-) diff --git a/README.md b/README.md index e055ae7..5fe2017 100644 --- a/README.md +++ b/README.md @@ -32,7 +32,9 @@ We believe that a well-developed open-source code framework can lower the thresh > DiffSynth-Studio has undergone major version updates, and some old features are no longer maintained. If you need to use old features, please switch to the [last historical version](https://github.com/modelscope/DiffSynth-Studio/tree/afd101f3452c9ecae0c87b79adfa2e22d65ffdc3) before the major version update. > Currently, the development personnel of this project are limited, with most of the work handled by [Artiprocher](https://github.com/Artiprocher). Therefore, the progress of new feature development will be relatively slow, and the speed of responding to and resolving issues is limited. We apologize for this and ask developers to understand. -- **February 10, 2026** Added inference and training support for the LTX-2 audio-video generation model. See the documentation for details. +- **February 26, 2026** Added full and lora training support for the LTX-2 audio-video generation model. See the [documentation](/docs/en/Model_Details/LTX-2.md) for details. + +- **February 10, 2026** Added inference support for the LTX-2 audio-video generation model. See the [documentation](/docs/en/Model_Details/LTX-2.md) for details. Support for model training will be implemented in the future. - **February 2, 2026** The first document of the Research Tutorial series is now available, guiding you through training a small 0.1B text-to-image model from scratch. For details, see the [documentation](/docs/en/Research_Tutorial/train_from_scratch.md) and [model](https://modelscope.cn/models/DiffSynth-Studio/AAAMyModel). We hope DiffSynth-Studio can evolve into a more powerful training framework for Diffusion models. @@ -557,12 +559,26 @@ vram_config = { "computation_dtype": torch.bfloat16, "computation_device": "cuda", } +""" +Offical model repo: https://www.modelscope.cn/models/Lightricks/LTX-2 +Repackaged model repo: https://www.modelscope.cn/models/DiffSynth-Studio/LTX-2-Repackage +For base models of LTX-2, offical checkpoint (with model config ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-dev.safetensors")) +and repackaged checkpoints (with model config ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="*.safetensors")) are both supported. +We have repackeged the official checkpoints in DiffSynth-Studio/LTX-2-Repackage repo to support separate loading of different submodules, +and avoid redundant memory usage when users only want to use part of the model. +""" +# use the repackaged modelconfig from "DiffSynth-Studio/LTX-2-Repackage" to avoid redundant model loading pipe = LTX2AudioVideoPipeline.from_pretrained( torch_dtype=torch.bfloat16, device="cuda", model_configs=[ ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized", origin_file_pattern="model-*.safetensors", **vram_config), - ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-dev.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="transformer.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="text_encoder_post_modules.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_encoder.safetensors", **vram_config), ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-spatial-upscaler-x2-1.0.safetensors", **vram_config), ], tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), @@ -570,6 +586,20 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( vram_limit=torch.cuda.mem_get_info("cuda")[1] / (1024 ** 3) - 0.5, ) +# use the following modelconfig if you want to initialize model from offical checkpoints from "Lightricks/LTX-2" +# pipe = LTX2AudioVideoPipeline.from_pretrained( +# torch_dtype=torch.bfloat16, +# device="cuda", +# model_configs=[ +# ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized", origin_file_pattern="model-*.safetensors", **vram_config), +# ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-dev.safetensors", **vram_config), +# ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-spatial-upscaler-x2-1.0.safetensors", **vram_config), +# ], +# tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), +# stage2_lora_config=ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-distilled-lora-384.safetensors"), +# vram_limit=torch.cuda.mem_get_info("cuda")[1] / (1024 ** 3) - 0.5, +# ) + prompt = "A girl is very happy, she is speaking: \"I enjoy working with Diffsynth-Studio, it's a perfect framework.\"" negative_prompt = ( "blurry, out of focus, overexposed, underexposed, low contrast, washed out colors, excessive noise, " diff --git a/README_zh.md b/README_zh.md index 519ebe6..74843db 100644 --- a/README_zh.md +++ b/README_zh.md @@ -32,7 +32,9 @@ DiffSynth 目前包括两个开源项目： > DiffSynth-Studio 经历了大版本更新，部分旧功能已停止维护，如需使用旧版功能，请切换到大版本更新前的[最后一个历史版本](https://github.com/modelscope/DiffSynth-Studio/tree/afd101f3452c9ecae0c87b79adfa2e22d65ffdc3)。 > 目前本项目的开发人员有限，大部分工作由 [Artiprocher](https://github.com/Artiprocher) 负责，因此新功能的开发进展会比较缓慢，issue 的回复和解决速度有限，我们对此感到非常抱歉，请各位开发者理解。 -- **2026年2月10日** 新增对[LTX-2](https://www.modelscope.cn/models/Lightricks/LTX-2)音视频生成模型的推理和训练支持，详见[文档](docs/zh/Model_Details/LTX-2.md)。 +- **2026年2月26日** 新增对[LTX-2](https://www.modelscope.cn/models/Lightricks/LTX-2)音视频生成模型全量微调与LoRA训练支持，详见[文档](docs/zh/Model_Details/LTX-2.md)。 + +- **2026年2月10日** 新增对[LTX-2](https://www.modelscope.cn/models/Lightricks/LTX-2)音视频生成模型的推理支持，详见[文档](docs/zh/Model_Details/LTX-2.md)，后续将推进模型训练的支持。 - **2026年2月2日** Research Tutorial 的第一篇文档上线，带你从零开始训练一个 0.1B 的小型文生图模型，详见[文档](/docs/zh/Research_Tutorial/train_from_scratch.md)、[模型](https://modelscope.cn/models/DiffSynth-Studio/AAAMyModel)，我们希望 DiffSynth-Studio 能够成为一个更强大的 Diffusion 模型训练框架。 @@ -557,12 +559,26 @@ vram_config = { "computation_dtype": torch.bfloat16, "computation_device": "cuda", } +""" +Offical model repo: https://www.modelscope.cn/models/Lightricks/LTX-2 +Repackaged model repo: https://www.modelscope.cn/models/DiffSynth-Studio/LTX-2-Repackage +For base models of LTX-2, offical checkpoint (with model config ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-dev.safetensors")) +and repackaged checkpoints (with model config ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="*.safetensors")) are both supported. +We have repackeged the official checkpoints in DiffSynth-Studio/LTX-2-Repackage repo to support separate loading of different submodules, +and avoid redundant memory usage when users only want to use part of the model. +""" +# use the repackaged modelconfig from "DiffSynth-Studio/LTX-2-Repackage" to avoid redundant model loading pipe = LTX2AudioVideoPipeline.from_pretrained( torch_dtype=torch.bfloat16, device="cuda", model_configs=[ ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized", origin_file_pattern="model-*.safetensors", **vram_config), - ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-dev.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="transformer.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="text_encoder_post_modules.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_encoder.safetensors", **vram_config), ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-spatial-upscaler-x2-1.0.safetensors", **vram_config), ], tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), @@ -570,6 +586,20 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( vram_limit=torch.cuda.mem_get_info("cuda")[1] / (1024 ** 3) - 0.5, ) +# use the following modelconfig if you want to initialize model from offical checkpoints from "Lightricks/LTX-2" +# pipe = LTX2AudioVideoPipeline.from_pretrained( +# torch_dtype=torch.bfloat16, +# device="cuda", +# model_configs=[ +# ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized", origin_file_pattern="model-*.safetensors", **vram_config), +# ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-dev.safetensors", **vram_config), +# ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-spatial-upscaler-x2-1.0.safetensors", **vram_config), +# ], +# tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), +# stage2_lora_config=ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-distilled-lora-384.safetensors"), +# vram_limit=torch.cuda.mem_get_info("cuda")[1] / (1024 ** 3) - 0.5, +# ) + prompt = "A girl is very happy, she is speaking: \"I enjoy working with Diffsynth-Studio, it's a perfect framework.\"" negative_prompt = ( "blurry, out of focus, overexposed, underexposed, low contrast, washed out colors, excessive noise, " diff --git a/diffsynth/configs/model_configs.py b/diffsynth/configs/model_configs.py index dbad638..fbca133 100644 --- a/diffsynth/configs/model_configs.py +++ b/diffsynth/configs/model_configs.py @@ -598,7 +598,14 @@ z_image_series = [ "state_dict_converter": "diffsynth.utils.state_dict_converters.z_image_text_encoder.ZImageTextEncoderStateDictConverter", }, ] - +""" +Offical model repo: https://www.modelscope.cn/models/Lightricks/LTX-2 +Repackaged model repo: https://www.modelscope.cn/models/DiffSynth-Studio/LTX-2-Repackage +For base models of LTX-2, offical checkpoint (with model config ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-dev.safetensors")) +and repackaged checkpoints (with model config ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="*.safetensors")) are both supported. +We have repackeged the official checkpoints in DiffSynth-Studio/LTX-2-Repackage repo to support separate loading of different submodules, +and avoid redundant memory usage when users only want to use part of the model. +""" ltx2_series = [ { # Example: ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-dev.safetensors") @@ -608,6 +615,7 @@ ltx2_series = [ "state_dict_converter": "diffsynth.utils.state_dict_converters.ltx2_dit.LTXModelStateDictConverter", }, { + # Example: ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="transformer.safetensors") "model_hash": "c567aaa37d5ed7454c73aa6024458661", "model_name": "ltx2_dit", "model_class": "diffsynth.models.ltx2_dit.LTXModel", @@ -621,6 +629,7 @@ ltx2_series = [ "state_dict_converter": "diffsynth.utils.state_dict_converters.ltx2_video_vae.LTX2VideoEncoderStateDictConverter", }, { + # Example: ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_encoder.safetensors") "model_hash": "7f7e904a53260ec0351b05f32153754b", "model_name": "ltx2_video_vae_encoder", "model_class": "diffsynth.models.ltx2_video_vae.LTX2VideoEncoder", @@ -634,6 +643,7 @@ ltx2_series = [ "state_dict_converter": "diffsynth.utils.state_dict_converters.ltx2_video_vae.LTX2VideoDecoderStateDictConverter", }, { + # Example: ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors") "model_hash": "dc6029ca2825147872b45e35a2dc3a97", "model_name": "ltx2_video_vae_decoder", "model_class": "diffsynth.models.ltx2_video_vae.LTX2VideoDecoder", @@ -647,6 +657,7 @@ ltx2_series = [ "state_dict_converter": "diffsynth.utils.state_dict_converters.ltx2_audio_vae.LTX2AudioDecoderStateDictConverter", }, { + # Example: ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors") "model_hash": "7d7823dde8f1ea0b50fb07ac329dd4cb", "model_name": "ltx2_audio_vae_decoder", "model_class": "diffsynth.models.ltx2_audio_vae.LTX2AudioDecoder", @@ -660,6 +671,7 @@ ltx2_series = [ "state_dict_converter": "diffsynth.utils.state_dict_converters.ltx2_audio_vae.LTX2VocoderStateDictConverter", }, { + # Example: ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors") "model_hash": "f471360f6b24bef702ab73133d9f8bb9", "model_name": "ltx2_audio_vocoder", "model_class": "diffsynth.models.ltx2_audio_vae.LTX2Vocoder", @@ -673,6 +685,7 @@ ltx2_series = [ "state_dict_converter": "diffsynth.utils.state_dict_converters.ltx2_audio_vae.LTX2AudioEncoderStateDictConverter", }, { + # Example: ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_encoder.safetensors") "model_hash": "29338f3b95e7e312a3460a482e4f4554", "model_name": "ltx2_audio_vae_encoder", "model_class": "diffsynth.models.ltx2_audio_vae.LTX2AudioEncoder", @@ -686,6 +699,7 @@ ltx2_series = [ "state_dict_converter": "diffsynth.utils.state_dict_converters.ltx2_text_encoder.LTX2TextEncoderPostModulesStateDictConverter", }, { + # Example: ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="text_encoder_post_modules.safetensors") "model_hash": "981629689c8be92a712ab3c5eb4fc3f6", "model_name": "ltx2_text_encoder_post_modules", "model_class": "diffsynth.models.ltx2_text_encoder.LTX2TextEncoderPostModules", diff --git a/docs/en/Model_Details/LTX-2.md b/docs/en/Model_Details/LTX-2.md index e3f4ca8..68ab351 100644 --- a/docs/en/Model_Details/LTX-2.md +++ b/docs/en/Model_Details/LTX-2.md @@ -33,19 +33,62 @@ vram_config = { "computation_dtype": torch.bfloat16, "computation_device": "cuda", } +""" +Offical model repo: https://www.modelscope.cn/models/Lightricks/LTX-2 +Repackaged model repo: https://www.modelscope.cn/models/DiffSynth-Studio/LTX-2-Repackage +For base models of LTX-2, offical checkpoint (with model config ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-dev.safetensors")) +and repackaged checkpoints (with model config ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="*.safetensors")) are both supported. +We have repackeged the official checkpoints in DiffSynth-Studio/LTX-2-Repackage repo to support separate loading of different submodules, +and avoid redundant memory usage when users only want to use part of the model. +""" +# use the repackaged modelconfig from "DiffSynth-Studio/LTX-2-Repackage" to avoid redundant model loading pipe = LTX2AudioVideoPipeline.from_pretrained( torch_dtype=torch.bfloat16, device="cuda", model_configs=[ ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized", origin_file_pattern="model-*.safetensors", **vram_config), - ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-dev.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="transformer.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="text_encoder_post_modules.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_encoder.safetensors", **vram_config), + ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-spatial-upscaler-x2-1.0.safetensors", **vram_config), ], tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), + stage2_lora_config=ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-distilled-lora-384.safetensors"), vram_limit=torch.cuda.mem_get_info("cuda")[1] / (1024 ** 3) - 0.5, ) + +# use the following modelconfig if you want to initialize model from offical checkpoints from "Lightricks/LTX-2" +# pipe = LTX2AudioVideoPipeline.from_pretrained( +# torch_dtype=torch.bfloat16, +# device="cuda", +# model_configs=[ +# ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized", origin_file_pattern="model-*.safetensors", **vram_config), +# ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-dev.safetensors", **vram_config), +# ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-spatial-upscaler-x2-1.0.safetensors", **vram_config), +# ], +# tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), +# stage2_lora_config=ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-distilled-lora-384.safetensors"), +# vram_limit=torch.cuda.mem_get_info("cuda")[1] / (1024 ** 3) - 0.5, +# ) + prompt = "A girl is very happy, she is speaking: \"I enjoy working with Diffsynth-Studio, it's a perfect framework.\"" -negative_prompt = "blurry, out of focus, overexposed, underexposed, low contrast, washed out colors, excessive noise, grainy texture, poor lighting, flickering, motion blur, distorted proportions, unnatural skin tones, deformed facial features, asymmetrical face, missing facial features, extra limbs, disfigured hands, wrong hand count, artifacts around text, inconsistent perspective, camera shake, incorrect depth of field, background too sharp, background clutter, distracting reflections, harsh shadows, inconsistent lighting direction, color banding, cartoonish rendering, 3D CGI look, unrealistic materials, uncanny valley effect, incorrect ethnicity, wrong gender, exaggerated expressions, wrong gaze direction, mismatched lip sync, silent or muted audio, distorted voice, robotic voice, echo, background noise, off-sync audio, incorrect dialogue, added dialogue, repetitive speech, jittery movement, awkward pauses, incorrect timing, unnatural transitions, inconsistent framing, tilted camera, flat lighting, inconsistent tone, cinematic oversaturation, stylized filters, or AI artifacts." -height, width, num_frames = 512, 768, 121 +negative_prompt = ( + "blurry, out of focus, overexposed, underexposed, low contrast, washed out colors, excessive noise, " + "grainy texture, poor lighting, flickering, motion blur, distorted proportions, unnatural skin tones, " + "deformed facial features, asymmetrical face, missing facial features, extra limbs, disfigured hands, " + "wrong hand count, artifacts around text, inconsistent perspective, camera shake, incorrect depth of " + "field, background too sharp, background clutter, distracting reflections, harsh shadows, inconsistent " + "lighting direction, color banding, cartoonish rendering, 3D CGI look, unrealistic materials, uncanny " + "valley effect, incorrect ethnicity, wrong gender, exaggerated expressions, wrong gaze direction, " + "mismatched lip sync, silent or muted audio, distorted voice, robotic voice, echo, background noise, " + "off-sync audio, incorrect dialogue, added dialogue, repetitive speech, jittery movement, awkward " + "pauses, incorrect timing, unnatural transitions, inconsistent framing, tilted camera, flat lighting, " + "inconsistent tone, cinematic oversaturation, stylized filters, or AI artifacts." +) +height, width, num_frames = 512 * 2, 768 * 2, 121 video, audio = pipe( prompt=prompt, negative_prompt=negative_prompt, @@ -54,11 +97,12 @@ video, audio = pipe( width=width, num_frames=num_frames, tiled=True, + use_two_stage_pipeline=True, ) write_video_audio_ltx2( video=video, audio=audio, - output_path='ltx2_onestage.mp4', + output_path='ltx2_twostage.mp4', fps=24, audio_sample_rate=24000, ) diff --git a/docs/zh/Model_Details/LTX-2.md b/docs/zh/Model_Details/LTX-2.md index 86abbcd..558de9d 100644 --- a/docs/zh/Model_Details/LTX-2.md +++ b/docs/zh/Model_Details/LTX-2.md @@ -33,19 +33,62 @@ vram_config = { "computation_dtype": torch.bfloat16, "computation_device": "cuda", } +""" +Offical model repo: https://www.modelscope.cn/models/Lightricks/LTX-2 +Repackaged model repo: https://www.modelscope.cn/models/DiffSynth-Studio/LTX-2-Repackage +For base models of LTX-2, offical checkpoint (with model config ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-dev.safetensors")) +and repackaged checkpoints (with model config ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="*.safetensors")) are both supported. +We have repackeged the official checkpoints in DiffSynth-Studio/LTX-2-Repackage repo to support separate loading of different submodules, +and avoid redundant memory usage when users only want to use part of the model. +""" +# use the repackaged modelconfig from "DiffSynth-Studio/LTX-2-Repackage" to avoid redundant model loading pipe = LTX2AudioVideoPipeline.from_pretrained( torch_dtype=torch.bfloat16, device="cuda", model_configs=[ ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized", origin_file_pattern="model-*.safetensors", **vram_config), - ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-dev.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="transformer.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="text_encoder_post_modules.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vae_decoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="audio_vocoder.safetensors", **vram_config), + ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="video_vae_encoder.safetensors", **vram_config), + ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-spatial-upscaler-x2-1.0.safetensors", **vram_config), ], tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), + stage2_lora_config=ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-distilled-lora-384.safetensors"), vram_limit=torch.cuda.mem_get_info("cuda")[1] / (1024 ** 3) - 0.5, ) -prompt = "A girl is very happy, she is speaking: “I enjoy working with Diffsynth-Studio, it's a perfect framework.”" -negative_prompt = "blurry, out of focus, overexposed, underexposed, low contrast, washed out colors, excessive noise, grainy texture, poor lighting, flickering, motion blur, distorted proportions, unnatural skin tones, deformed facial features, asymmetrical face, missing facial features, extra limbs, disfigured hands, wrong hand count, artifacts around text, inconsistent perspective, camera shake, incorrect depth of field, background too sharp, background clutter, distracting reflections, harsh shadows, inconsistent lighting direction, color banding, cartoonish rendering, 3D CGI look, unrealistic materials, uncanny valley effect, incorrect ethnicity, wrong gender, exaggerated expressions, wrong gaze direction, mismatched lip sync, silent or muted audio, distorted voice, robotic voice, echo, background noise, off-sync audio, incorrect dialogue, added dialogue, repetitive speech, jittery movement, awkward pauses, incorrect timing, unnatural transitions, inconsistent framing, tilted camera, flat lighting, inconsistent tone, cinematic oversaturation, stylized filters, or AI artifacts." -height, width, num_frames = 512, 768, 121 + +# use the following modelconfig if you want to initialize model from offical checkpoints from "Lightricks/LTX-2" +# pipe = LTX2AudioVideoPipeline.from_pretrained( +# torch_dtype=torch.bfloat16, +# device="cuda", +# model_configs=[ +# ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized", origin_file_pattern="model-*.safetensors", **vram_config), +# ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-dev.safetensors", **vram_config), +# ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-spatial-upscaler-x2-1.0.safetensors", **vram_config), +# ], +# tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), +# stage2_lora_config=ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-distilled-lora-384.safetensors"), +# vram_limit=torch.cuda.mem_get_info("cuda")[1] / (1024 ** 3) - 0.5, +# ) + +prompt = "A girl is very happy, she is speaking: \"I enjoy working with Diffsynth-Studio, it's a perfect framework.\"" +negative_prompt = ( + "blurry, out of focus, overexposed, underexposed, low contrast, washed out colors, excessive noise, " + "grainy texture, poor lighting, flickering, motion blur, distorted proportions, unnatural skin tones, " + "deformed facial features, asymmetrical face, missing facial features, extra limbs, disfigured hands, " + "wrong hand count, artifacts around text, inconsistent perspective, camera shake, incorrect depth of " + "field, background too sharp, background clutter, distracting reflections, harsh shadows, inconsistent " + "lighting direction, color banding, cartoonish rendering, 3D CGI look, unrealistic materials, uncanny " + "valley effect, incorrect ethnicity, wrong gender, exaggerated expressions, wrong gaze direction, " + "mismatched lip sync, silent or muted audio, distorted voice, robotic voice, echo, background noise, " + "off-sync audio, incorrect dialogue, added dialogue, repetitive speech, jittery movement, awkward " + "pauses, incorrect timing, unnatural transitions, inconsistent framing, tilted camera, flat lighting, " + "inconsistent tone, cinematic oversaturation, stylized filters, or AI artifacts." +) +height, width, num_frames = 512 * 2, 768 * 2, 121 video, audio = pipe( prompt=prompt, negative_prompt=negative_prompt, @@ -54,11 +97,12 @@ video, audio = pipe( width=width, num_frames=num_frames, tiled=True, + use_two_stage_pipeline=True, ) write_video_audio_ltx2( video=video, audio=audio, - output_path='ltx2_onestage.mp4', + output_path='ltx2_twostage.mp4', fps=24, audio_sample_rate=24000, ) diff --git a/examples/ltx2/model_inference/LTX-2-T2AV-OneStage.py b/examples/ltx2/model_inference/LTX-2-T2AV-OneStage.py index 295d73b..9e56209 100644 --- a/examples/ltx2/model_inference/LTX-2-T2AV-OneStage.py +++ b/examples/ltx2/model_inference/LTX-2-T2AV-OneStage.py @@ -12,6 +12,15 @@ vram_config = { "computation_dtype": torch.bfloat16, "computation_device": "cuda", } +""" +Offical model repo: https://www.modelscope.cn/models/Lightricks/LTX-2 +Repackaged model repo: https://www.modelscope.cn/models/DiffSynth-Studio/LTX-2-Repackage +For base models of LTX-2, offical checkpoint (with model config ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-dev.safetensors")) +and repackaged checkpoints (with model config ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="*.safetensors")) are both supported. +We have repackeged the official checkpoints in DiffSynth-Studio/LTX-2-Repackage repo to support separate loading of different submodules, +and avoid redundant memory usage when users only want to use part of the model. +""" +# use the repackaged modelconfig from "DiffSynth-Studio/LTX-2-Repackage" to avoid redundant model loading pipe = LTX2AudioVideoPipeline.from_pretrained( torch_dtype=torch.bfloat16, device="cuda", @@ -25,6 +34,16 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( ], tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), ) +# use the following modelconfig if you want to initialize model from offical checkpoints from "Lightricks/LTX-2" +# pipe = LTX2AudioVideoPipeline.from_pretrained( +# torch_dtype=torch.bfloat16, +# device="cuda", +# model_configs=[ +# ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized", origin_file_pattern="model-*.safetensors", **vram_config), +# ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-dev.safetensors", **vram_config), +# ], +# tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), +# ) prompt = "A girl is very happy, she is speaking: “I enjoy working with Diffsynth-Studio, it's a perfect framework.”" negative_prompt = "blurry, out of focus, overexposed, underexposed, low contrast, washed out colors, excessive noise, grainy texture, poor lighting, flickering, motion blur, distorted proportions, unnatural skin tones, deformed facial features, asymmetrical face, missing facial features, extra limbs, disfigured hands, wrong hand count, artifacts around text, inconsistent perspective, camera shake, incorrect depth of field, background too sharp, background clutter, distracting reflections, harsh shadows, inconsistent lighting direction, color banding, cartoonish rendering, 3D CGI look, unrealistic materials, uncanny valley effect, incorrect ethnicity, wrong gender, exaggerated expressions, wrong gaze direction, mismatched lip sync, silent or muted audio, distorted voice, robotic voice, echo, background noise, off-sync audio, incorrect dialogue, added dialogue, repetitive speech, jittery movement, awkward pauses, incorrect timing, unnatural transitions, inconsistent framing, tilted camera, flat lighting, inconsistent tone, cinematic oversaturation, stylized filters, or AI artifacts." height, width, num_frames = 512, 768, 121 diff --git a/examples/ltx2/model_inference/LTX-2-T2AV-TwoStage.py b/examples/ltx2/model_inference/LTX-2-T2AV-TwoStage.py index 9a85a9f..e08ed50 100644 --- a/examples/ltx2/model_inference/LTX-2-T2AV-TwoStage.py +++ b/examples/ltx2/model_inference/LTX-2-T2AV-TwoStage.py @@ -12,6 +12,15 @@ vram_config = { "computation_dtype": torch.bfloat16, "computation_device": "cuda", } +""" +Offical model repo: https://www.modelscope.cn/models/Lightricks/LTX-2 +Repackaged model repo: https://www.modelscope.cn/models/DiffSynth-Studio/LTX-2-Repackage +For base models of LTX-2, offical checkpoint (with model config ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-dev.safetensors")) +and repackaged checkpoints (with model config ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="*.safetensors")) are both supported. +We have repackeged the official checkpoints in DiffSynth-Studio/LTX-2-Repackage repo to support separate loading of different submodules, +and avoid redundant memory usage when users only want to use part of the model. +""" +# use the repackaged modelconfig from "DiffSynth-Studio/LTX-2-Repackage" to avoid redundant model loading pipe = LTX2AudioVideoPipeline.from_pretrained( torch_dtype=torch.bfloat16, device="cuda", @@ -28,7 +37,18 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), stage2_lora_config=ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-distilled-lora-384.safetensors"), ) - +# use the following modelconfig if you want to initialize model from offical checkpoints from "Lightricks/LTX-2" +# pipe = LTX2AudioVideoPipeline.from_pretrained( +# torch_dtype=torch.bfloat16, +# device="cuda", +# model_configs=[ +# ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized", origin_file_pattern="model-*.safetensors", **vram_config), +# ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-dev.safetensors", **vram_config), +# ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-spatial-upscaler-x2-1.0.safetensors", **vram_config), +# ], +# tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), +# stage2_lora_config=ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-distilled-lora-384.safetensors"), +# ) prompt = "A girl is very happy, she is speaking: “I enjoy working with Diffsynth-Studio, it's a perfect framework.”" negative_prompt = ( "blurry, out of focus, overexposed, underexposed, low contrast, washed out colors, excessive noise, " diff --git a/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-OneStage.py b/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-OneStage.py index fb08384..c08332f 100644 --- a/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-OneStage.py +++ b/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-OneStage.py @@ -12,6 +12,15 @@ vram_config = { "computation_dtype": torch.bfloat16, "computation_device": "cuda", } +""" +Offical model repo: https://www.modelscope.cn/models/Lightricks/LTX-2 +Repackaged model repo: https://www.modelscope.cn/models/DiffSynth-Studio/LTX-2-Repackage +For base models of LTX-2, offical checkpoint (with model config ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-dev.safetensors")) +and repackaged checkpoints (with model config ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="*.safetensors")) are both supported. +We have repackeged the official checkpoints in DiffSynth-Studio/LTX-2-Repackage repo to support separate loading of different submodules, +and avoid redundant memory usage when users only want to use part of the model. +""" +# use the repackaged modelconfig from "DiffSynth-Studio/LTX-2-Repackage" to avoid redundant model loading pipe = LTX2AudioVideoPipeline.from_pretrained( torch_dtype=torch.bfloat16, device="cuda", @@ -26,6 +35,17 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), vram_limit=torch.cuda.mem_get_info("cuda")[1] / (1024 ** 3) - 0.5, ) +# use the following modelconfig if you want to initialize model from offical checkpoints from "Lightricks/LTX-2" +# pipe = LTX2AudioVideoPipeline.from_pretrained( +# torch_dtype=torch.bfloat16, +# device="cuda", +# model_configs=[ +# ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized", origin_file_pattern="model-*.safetensors", **vram_config), +# ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-dev.safetensors", **vram_config), +# ], +# tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), +# vram_limit=torch.cuda.mem_get_info("cuda")[1] / (1024 ** 3) - 0.5, +# ) prompt = "A girl is very happy, she is speaking: “I enjoy working with Diffsynth-Studio, it's a perfect framework.”" negative_prompt = "blurry, out of focus, overexposed, underexposed, low contrast, washed out colors, excessive noise, grainy texture, poor lighting, flickering, motion blur, distorted proportions, unnatural skin tones, deformed facial features, asymmetrical face, missing facial features, extra limbs, disfigured hands, wrong hand count, artifacts around text, inconsistent perspective, camera shake, incorrect depth of field, background too sharp, background clutter, distracting reflections, harsh shadows, inconsistent lighting direction, color banding, cartoonish rendering, 3D CGI look, unrealistic materials, uncanny valley effect, incorrect ethnicity, wrong gender, exaggerated expressions, wrong gaze direction, mismatched lip sync, silent or muted audio, distorted voice, robotic voice, echo, background noise, off-sync audio, incorrect dialogue, added dialogue, repetitive speech, jittery movement, awkward pauses, incorrect timing, unnatural transitions, inconsistent framing, tilted camera, flat lighting, inconsistent tone, cinematic oversaturation, stylized filters, or AI artifacts." height, width, num_frames = 512, 768, 121 diff --git a/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-TwoStage.py b/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-TwoStage.py index a10b6ef..98ed966 100644 --- a/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-TwoStage.py +++ b/examples/ltx2/model_inference_low_vram/LTX-2-T2AV-TwoStage.py @@ -12,6 +12,15 @@ vram_config = { "computation_dtype": torch.bfloat16, "computation_device": "cuda", } +""" +Offical model repo: https://www.modelscope.cn/models/Lightricks/LTX-2 +Repackaged model repo: https://www.modelscope.cn/models/DiffSynth-Studio/LTX-2-Repackage +For base models of LTX-2, offical checkpoint (with model config ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-dev.safetensors")) +and repackaged checkpoints (with model config ModelConfig(model_id="DiffSynth-Studio/LTX-2-Repackage", origin_file_pattern="*.safetensors")) are both supported. +We have repackeged the official checkpoints in DiffSynth-Studio/LTX-2-Repackage repo to support separate loading of different submodules, +and avoid redundant memory usage when users only want to use part of the model. +""" +# use the repackaged modelconfig from "DiffSynth-Studio/LTX-2-Repackage" to avoid redundant model loading pipe = LTX2AudioVideoPipeline.from_pretrained( torch_dtype=torch.bfloat16, device="cuda", @@ -29,6 +38,19 @@ pipe = LTX2AudioVideoPipeline.from_pretrained( stage2_lora_config=ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-distilled-lora-384.safetensors"), vram_limit=torch.cuda.mem_get_info("cuda")[1] / (1024 ** 3) - 0.5, ) +# use the following modelconfig if you want to initialize model from offical checkpoints from "Lightricks/LTX-2" +# pipe = LTX2AudioVideoPipeline.from_pretrained( +# torch_dtype=torch.bfloat16, +# device="cuda", +# model_configs=[ +# ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized", origin_file_pattern="model-*.safetensors", **vram_config), +# ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-dev.safetensors", **vram_config), +# ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-spatial-upscaler-x2-1.0.safetensors", **vram_config), +# ], +# tokenizer_config=ModelConfig(model_id="google/gemma-3-12b-it-qat-q4_0-unquantized"), +# stage2_lora_config=ModelConfig(model_id="Lightricks/LTX-2", origin_file_pattern="ltx-2-19b-distilled-lora-384.safetensors"), +# vram_limit=torch.cuda.mem_get_info("cuda")[1] / (1024 ** 3) - 0.5, +# ) prompt = "A girl is very happy, she is speaking: “I enjoy working with Diffsynth-Studio, it's a perfect framework.”" negative_prompt = (