onnxruntime
7e0dd9d4
- [js/webgpu] Optimize Expand (#22752)
Go
Login via GitHub
Home
Pricing
FAQ
Install
Login
via GitHub
Commit
View On
GitHub
Commit
1 year ago
[js/webgpu] Optimize Expand (#22752) Use components = 4 if possible. llama3.2-1B becomes 20 tokens/s from 18 tokens/s on my iGPUs.
References
#22752 - [js/webgpu] Optimize Expand
Author
qjia7
Parents
05c8dc9d
Loading