-
Notifications
You must be signed in to change notification settings - Fork 83
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
Java api run slowerly than Python #81
Comments
I think I found the problem. when set MKL_VERBOSE=1, java show "MKL_VERBOSE Intel(R) MKL 2019.0 Update 3 Product build 20190220.." and python show "MKL_VERBOSE oneMKL 2021.0 Update 4 Product build 20210904...". So the old version of MKL may show worse performance? |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
I compared java and python performance with CPU and MKL enabled, Java(20ms) is much slower than Python(10ms).
BERT parameters:
batch_size=1, seq_length=24, num_hidden_layers=6, num_attention_heads=12
ENV variables:
export OMP_NUM_THREADS=8
export KMP_BLOCKTIME=0
export KMP_AFFINITY=granularity=fine,verbose,compact,1,0
Then I tested Java Native Api(JNI) instead of JNA, it's also the same phenomenon. I suspect that it's something about cuBERT.so (maybe MKL configuration) making java run slower than python?
The text was updated successfully, but these errors were encountered: