Skip to content

Reformat S2V models #476

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
bubbliiiing wants to merge 10 commits into
base: main
Choose a base branch
from
Open

Reformat S2V models #476

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
10 commits
Select commit Hold shift + click to select a range
b11b085
Reformat S2V models
bubbliiiing Mar 11, 2026
3ced952
Update s2v models training code
bubbliiiing Mar 11, 2026
5f46550
Update LTX2, Update pipelines, Reformat S2V codes
bubbliiiing Mar 12, 2026
2b3b257
Update save_results
bubbliiiing Mar 12, 2026
f209cb3
Update init
bubbliiiing Mar 12, 2026
8f0768f
Update Wan s2v codes
bubbliiiing Mar 12, 2026
e5b7590
Update sh
bubbliiiing Mar 12, 2026
2bd2d2e
Add gradient checkpointing in s2v
bubbliiiing Mar 12, 2026
8fb48cb
Add gradient checkpointing in wan animate
bubbliiiing Mar 13, 2026
c7989a6
update attention and s2v validation
bubbliiiing Mar 13, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
1 change: 0 additions & 1 deletion examples/longcatvideo/predict_i2v.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -149,7 +149,6 @@

if GPU_memory_mode == "sequential_cpu_offload":
replace_parameters_by_name(transformer, ["modulation",], device=device)
transformer.freqs = transformer.freqs.to(device=device)
pipeline.enable_sequential_cpu_offload(device=device)
elif GPU_memory_mode == "model_cpu_offload_and_qfloat8":
convert_model_weight_to_float8(transformer, exclude_module_name=["modulation",], device=device)
Expand Down
17 changes: 4 additions & 13 deletions examples/longcatvideo/predict_s2v_avatar.py
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,9 +1,7 @@
import math
import os
import sys
from pathlib import Path

import librosa
import numpy as np
import torch
from audio_separator.separator import Separator
Expand All @@ -18,9 +16,9 @@

from videox_fun.dist import set_multi_gpus_devices, shard_model
from videox_fun.models import (AutoencoderKLLongCatVideo, AutoTokenizer,
LongCatVideoAudioEncoder,
LongCatVideoAvatarTransformer3DModel,
UMT5EncoderModel, Wav2Vec2FeatureExtractor,
Wav2Vec2ModelWrapper)
UMT5EncoderModel)
from videox_fun.models.cache_utils import get_teacache_coefficients
from videox_fun.pipeline import LongCatVideoAvatarPipeline
from videox_fun.utils.fm_solvers import FlowDPMSolverMultistepScheduler
Expand Down Expand Up @@ -136,15 +134,10 @@
)

# Get Audio encoder (for avatar mode)
audio_encoder = Wav2Vec2ModelWrapper(
audio_encoder = LongCatVideoAudioEncoder(
os.path.join(model_name_avatar, 'chinese-wav2vec2-base')
)
audio_encoder.feature_extractor._freeze_parameters()

wav2vec_feature_extractor = Wav2Vec2FeatureExtractor.from_pretrained(
os.path.join(model_name_avatar, 'chinese-wav2vec2-base'),
local_files_only=True
)
audio_encoder.audio_encoder.feature_extractor._freeze_parameters()

# Get Scheduler
Chosen_Scheduler = scheduler_dict = {
Expand All @@ -165,7 +158,6 @@
text_encoder=text_encoder,
scheduler=scheduler,
audio_encoder=audio_encoder,
wav2vec_feature_extractor=wav2vec_feature_extractor,
)

if compile_dit:
Expand All @@ -175,7 +167,6 @@

if GPU_memory_mode == "sequential_cpu_offload":
replace_parameters_by_name(transformer, ["modulation",], device=device)
transformer.freqs = transformer.freqs.to(device=device)
pipeline.enable_sequential_cpu_offload(device=device)
elif GPU_memory_mode == "model_cpu_offload_and_qfloat8":
convert_model_weight_to_float8(transformer, exclude_module_name=["modulation",], device=device)
Expand Down
1 change: 0 additions & 1 deletion examples/longcatvideo/predict_t2v.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -147,7 +147,6 @@

if GPU_memory_mode == "sequential_cpu_offload":
replace_parameters_by_name(transformer, ["modulation",], device=device)
transformer.freqs = transformer.freqs.to(device=device)
pipeline.enable_sequential_cpu_offload(device=device)
elif GPU_memory_mode == "model_cpu_offload_and_qfloat8":
convert_model_weight_to_float8(transformer, exclude_module_name=["modulation",], device=device)
Expand Down
Loading