Skip to content
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.

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

Add prefill tps in oga-bench #250

Merged
merged 2 commits into from
Dec 13, 2024
Merged

Add prefill tps in oga-bench #250

merged 2 commits into from
Dec 13, 2024

Conversation

jiafatom
Copy link
Contributor

@jiafatom jiafatom commented Dec 13, 2024

Want to know prefill tps, which is prompt_length/seconds_to_first_token

Tested lemonade -i cuda-fpmixed_14 oga-load --dtype fp16 --device cuda oga-bench

cuda-fpmixed_14:
(executed 1x)
Build dir: /home/jiafa/.cache/lemonade/cuda-fpmixed_14
Status: Successful build!
Dtype: fp16
Device: cuda
Oga Models Subfolder: cuda-fpmixed_14/cuda-fp16
Seconds To First Token: 0.004
Prefill Tokens Per Second: 1455.340
Token Generation Tokens Per Second: 273.599
Prompt Tokens: 6

@jiafatom jiafatom changed the title Add prompt processing tps in oga-bench Add prefill tps in oga-bench Dec 13, 2024
@ramkrishna2910 ramkrishna2910 merged commit 9dea250 into onnx:main Dec 13, 2024
8 checks passed
@jiafatom jiafatom deleted the metrics branch December 15, 2024 00:56
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants