Benchmarks

Benchmarks copied to clipboard

CANDLE-compliant Python API for Uno_IMPROVE

1

We need the API for HPO runs on Polaris - has anyone started on this?

j-woz

Tdey/intel xpu

6

Added callbacks to measure performance throughput more precisely. By default, it shows the throughput for each epoch. Updated the optimizer API for keras==2.12 and keras==3.13 version

newtdms

For use by IMPROVE

j-woz

model_name required

5

With the updates in candle_lib `model_name` is a required hyperparameter. Breaks UNO and might other Benchmarks also.

rajeeja

Proof of concept for containerizing benchmark models. Simple Dockerfile and standardized shell script for model execution.

wilke

Settable data location

3

Do not force big input data to reside inside a git clone. This forces code and big data to reside on the same FS (I use a soft link to...

j-woz

Testing: Split Theta tests into plain and Singularity

1

j-woz

add counterfactual code

8

shahashka

Missing parameter causing errors in NT3 (P1B4)

5

While running Pilot 1 NT3 using the command `python nt3_baseline_keras2.py --conf nt3_perf_bench_model.txt ` I ran into this error caused by a missing parameter ` Traceback (most recent call last): File...

vgutta

PyTorch Versions of TC1 and P1B2 Benchmarks

1

* Code is generic and handles the neural network model as defined in the model_pytorch.txt file (similar to the TensorFlow/Keras version). * baseline_pytorch.py is the main function of the benchmark....

knangia04