Contact person: Python Script needs TensorFlow Performance Optimization

Phone:Show

Email:Show

Location: niles, United States

Budget: Recommended by industry experts

Time to start: As soon as possible

Project description:
"I have a Python machine-learning script built around TensorFlow that is taking far too long to finish and is monopolising the CPU while it runs. I need fresh eyes on the codebase to pinpoint—and fix—the bottlenecks so the same model trains and predicts faster and with noticeably lower processor load.

You are free to introduce efficient TensorFlow APIs, vectorised NumPy operations, mixed-precision, graph or dataset optimisations, multiprocessing, or any other proven techniques. Just keep the model’s outputs identical (or statistically equivalent after seeding) so that performance gains come without changing results.

Deliverables
• Refactored Python script(s) ready to drop in
• Short report showing before-and-after runtime and CPU utilisation on the same dataset, plus a brief explanation of what changed
• Clear setup or dependency notes if additional packages, flags, or environment tweaks are required

I can provide a reproducible test dataset and the current run command so you can profile locally. Looking forward to seeing how much speed—and breathing room for my CPU—you can squeeze out of it.

*Budget is $15*

*compatibility*
pip install --upgrade pip
pip uninstall -y tensorflow tensorflow-intel keras tf-keras deepface
pip install "tensorflow==2.20.0" "tf-keras==2.20.0" "deepface>=0.0.95"" (client-provided description)