Skip to content

Issues: hao-ai-lab/FastVideo

New issue

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

#

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

Already on GitHub? # to your account

24 Open 100 Closed
24 Open 100 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Assignee
Filter by who’s assigned
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Issues list

[Feature] Generalize STA kernel to work for any sequence length
#225 opened Feb 28, 2025 by alexarmbr
[Feature] Add FlexAttention version STA support
#224 opened Feb 28, 2025 by shadow150519
[Question] Question about flattening strategy in Slide Tiling Attention
#222 opened Feb 27, 2025 by botbw
4
[Feature] Any plan to support training on Wan2.1 t2v or i2v model
#221 opened Feb 26, 2025 by fangxia1997
2
[Feature] mask_strategy可以改成可学参数吗？
#219 opened Feb 26, 2025 by wangshankun
4
[Bug] Possibly incorrect CFG formula in teacher_output computation within distill.py
#217 opened Feb 25, 2025 by SidneyPrimas
4
[Bug] It looks like the config file at 'data/hunyuan_diffusers/config.json' is not a valid JSON file.
#212 opened Feb 24, 2025 by penglipan
@BrianChen1129
3
[Feature] STA是否可以支持diffusers版本呢？
#197 opened Feb 19, 2025 by yxqyyy
2
[Feature] Any plan to support STA on A100
#196 opened Feb 19, 2025 by trouble-maker007
2
[Bug] Unable to Setup STA
#193 opened Feb 19, 2025 by TemporalLabsLLC-SOL
7
[Feature] Video text alignment after finetuning Hunyuan
#180 opened Feb 13, 2025 by yjhong89
2
[Feature] the tensor in mochi_latent_utilis.py
#178 opened Feb 7, 2025 by yomi1117
1
[Feature] The vae of mochi and hunyuan are different, so why their DIT blocks can be trained on the same dataset and the same code
#177 opened Feb 7, 2025 by yomi1117
1
Questions on Benchmarks
#175 opened Feb 7, 2025 by amogkam
3
[Feature] thanks all, but why i only produce 1s time video
#173 opened Feb 5, 2025 by praymich
3
[Feature] 2080ti and how to unable flash-atte
#171 opened Feb 1, 2025 by nigelaga
@foreverpiano
1
[Bug] latents_attention_mask not used
#168 opened Jan 27, 2025 by HaoyiZhu
@foreverpiano
2
[Bug] 404 on config.json when using inference_hunyuan_hf_quantization.sh
#164 opened Jan 22, 2025 by purzbeats
@foreverpiano
4
hunyuan lora微调nccl超时
#162 opened Jan 21, 2025 by ytm-01
@foreverpiano
8
请问全量微调hunyuan脚本是支持多机并行训练的吗？
#138 opened Jan 9, 2025 by QingQingS
4
[Feature] What is the meaning of argument "multi_phased_distill_schedule" in distill.py
#130 opened Jan 4, 2025 by 1145284121
6
Does it work with Spatiotemporal Skip Guidance?
#118 opened Dec 28, 2024 by renschni
@foreverpiano
4
The generated result is black，what's wrong with it?
#107 opened Dec 23, 2024 by TZYSJTU
@rlsu9
10
any plan to support cogvideox?
#91 opened Dec 18, 2024 by trouble-maker007
@foreverpiano
3
ProTip! Exclude everything labeled bug with -label:bug.