Comments (4)
I have done some test list below on 3080ti :
This processing time is only for processing the time of the first 3 keyframes.
-
Default setting
===== time: 171.90519642829895 sec -
Turn off Spatial-guided optimization: set optimize_temporal = False in apply_FRESCO_opt()
===== time: 47.287460803985596 sec -
Turn off Temporal-guided optimization: set correlation_matrix = [] in apply_FRESCO_opt()
===== time: 25.293941736221313 sec
Turn off Spatial-guided optimization: set optimize_temporal = False in apply_FRESCO_opt()
Turn off Temporal-guided optimization: set correlation_matrix = [] in apply_FRESCO_opt()
===== time: 15.881555318832397 sec
from fresco.
You can speed up the process by turning off some functions (sacrifice same performance)
modify
Lines 232 to 234 in 9fe1be7
based on
Lines 223 to 225 in 9fe1be7
from fresco.
On 3080ti the GPU usage is under 30%, most of time is only 1%
from fresco.
My run_fresco.ipynb ran on 4090 GPU,
the running time for 8 frames is
- Default setting: 26s
- Turn off all optimization: 3s
I have done some test list below on 3080ti :
This processing time is only for processing the time of the first 3 keyframes.
- Default setting
===== time: 171.90519642829895 sec- Turn off Spatial-guided optimization: set optimize_temporal = False in apply_FRESCO_opt()
===== time: 47.287460803985596 sec- Turn off Temporal-guided optimization: set correlation_matrix = [] in apply_FRESCO_opt()
===== time: 25.293941736221313 secTurn off Spatial-guided optimization: set optimize_temporal = False in apply_FRESCO_opt() Turn off Temporal-guided optimization: set correlation_matrix = [] in apply_FRESCO_opt() ===== time: 15.881555318832397 sec
from fresco.
Related Issues (20)
- I have a 1.5 model .safetensor file downloaded from civitai. How to use it? HOT 3
- Is there any way to speed it up? HOT 2
- 修改端口ip HOT 2
- 视频大于1M ,24GPU会爆吗?还是只有我是这样的? HOT 4
- TypeError: Video.__init__() got an unexpected keyword argument 'source' HOT 3
- Enable support to use lightning model
- ImportError: cannot import name 'logger' from 'diffusers.models.unet_2d_condition' (/usr/local/lib/python3.10/dist-packages/diffusers/models/unet_2d_condition.py) HOT 1
- notebook doesn't run HOT 3
- ebsynth generation fails HOT 7
- run on Colab with A100 GPU and 40G GPU RAM still got Error HOT 6
- run webUI.py with a 3 seconds video of 1280x720, got Error HOT 1
- [Run Key Frames] sucess, but [Run Propagation (Ebsynth)] failed: FileNotFoundError: [Errno 2] No such file or directory: 'output/480x480/blend.mp4' HOT 1
- Got RuntimeError:: element 0 of tensors does not require grad and does not have a grad_fn HOT 1
- python video_blend.py Cannot run HOT 3
- Does it support LCM models? HOT 6
- Diffusers HOT 2
- Video blend too slow: 18 vCPU / RTX 3090(24GB) HOT 1
- can not find target file while executing process_seq(video_sequence, i, blend_histogram, blend_gradient) in video_blend.py HOT 21
- blender not found
- Meaning of unet_chunk_size in FRESCOAttnProcessor2_0 codes? HOT 2
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from fresco.