Design Original ML Benchmark Challenges (ML Experts ONLY)
$30-250 USD
Posted 9 minutes ago
•
Ends in 6 days
$30-250 USD
Paid on delivery
I run an agent-evaluation platform and need a steady stream of fresh, production-grade machine-learning challenges that can be dropped straight into our test harness. For each accepted task I pay USD 200 once the repository clears a brief code and licensing review. Scope of each challenge • Build or source a commercially usable dataset (public-domain, permissive-license, or self-generated). • Write deterministic [login to view URL] and [login to view URL] scripts compatible with Python 3.10; they must run headless on Linux via a single command. • Document the problem in a concise README that explains objectives, data schema, evaluation metric, and submission format. • Provide at least five clear, task-specific rubric checks the grading script can expose as JSON. • Supply a clean, well-commented reference solution in a Jupyter notebook that reaches a reasonable benchmark score. • Keep everything reproducible and free of copyrighted Kaggle material or LLM-generated filler. Technology expectations Pandas, scikit-learn, XGBoost or LightGBM are welcome for classic tabular work; PyTorch or TensorFlow may be better for vision or NLP—use whatever best serves the task while keeping dependencies lightweight. Acceptance criteria 1. Repository installs with pip and runs end-to-end on a GitHub-hosted CI test. 2. Grade script yields identical scores across repeated runs. 3. Dataset license file included and compatible with commercial redistribution. 4. Reference solution executes without manual tweaks and passes all rubric checks. If you enjoy crafting Kaggle-style problems and have a track record of well-engineered notebooks or GitHub projects, send over a link to past work and let’s line up the first challenge.
Project ID: 40558115
About the project
41 proposals
Open for bidding
Remote project
Active 15 secs ago
Place your bid
Bid amount Email addressBenefits of bidding on Freelancer
Set your budget and timeframe
It's free to sign up and bid on jobs
41 freelancers are bidding on average $143 USD for this job


Hi there, our senior engineering team at InvisionOne (Auroinfo LLC, Official Google Cloud & Shopify Partner) can architect and build your Design Original ML Benchmark Challenges (ML Expert platform across 3 structured milestones: Milestone 1 delivers the core system architecture, database schema, and working staging prototype; Milestone 2 completes end-to-end API, backend workflow, and UI integration; Milestone 3 covers automated QA, deployment, documentation, and long-term milestone support. Could you share any existing technical specs or API docs so we can finalize the Milestone 1 plan?


I build and evaluate ML pipelines regularly, and designing benchmark tasks that are actually hard for agents (not just verbose) is its own skill. I would pull from real data analysis and CI/CD scenarios so each challenge has a clean, verifiable ground truth. Can start today. These numbers are based on the post as written, we will refine scope and pricing after a quick chat. Want me to send a sample challenge outline first?


Do you have specific themes or domains in mind for the machine-learning challenges? I’m excited about the opportunity to develop fresh, production-grade tasks for your platform. With extensive experience in creating well-structured datasets and writing robust grading scripts in Python, I can ensure that each challenge meets your requirements. My track record includes crafting high-quality notebooks and CI-compatible repositories that pass thorough reviews. Each task will include a clear README, rubric checks, and a clean reference solution that adheres to licensing requirements. I’m well-versed in libraries like Pandas, scikit-learn, and TensorFlow, allowing flexibility in selecting the best tools for each challenge. Check out my previous work [insert link] and let’s get started on your first challenge!


Hi there, I read your project "Design Original ML Benchmark Challenges (ML Experts ONLY)" and it lines up well with my experience in Java, Python, Linux. I've been doing this kind of work for 9 years and focus on clean, well-tested code and clear communication. Before starting I'd like to confirm a couple of details about the scope so the result matches exactly what you need. Happy to jump on a quick chat. I can deliver in 1 days for $75 USD. Best regards, Siddharth Kasherwal


The engineering side of this is what I do daily. Deterministic Python scripts, CI pipelines, reproducible environments, clean repos that install and run headless on Linux with zero manual tweaks. I've built structured data pipelines in Python (pandas, automated processing, test harnesses) across 10+ years of production work, and I'm comfortable with the prepare/grade/JSON-rubric architecture you're describing. That said, I want to be straight with you. My ML work is applied, not research-grade. I use scikit-learn, XGBoost, and PyTorch for practical tasks like classification and NLP pipelines, but I'm not publishing novel benchmark designs at a Kaggle grandmaster level. The reproducibility engineering, dataset licensing, CI integration, and clean notebook documentation? That's squarely in my lane. The challenge design itself would lean on well-understood problem shapes (tabular classification, regression, text clustering) rather than exotic architectures. What you'd get: repos that actually install and run first try, grade scripts that are truly deterministic (I've debugged enough flaky test suites to know where randomness sneaks in), and clean documentation that doesn't need hand-holding. If you're open to classic-ML-flavored challenges from someone who ships clean engineering, send over the harness spec and I can turn around a sample challenge to prove the workflow.


I have extensive experience designing machine learning benchmarks that align closely with real-world evaluation metrics. My background includes creating robust, production-ready challenges that drive meaningful progress in ML research. I focus on developing tests that accurately reflect deployment scenarios and integrate seamlessly into existing evaluation platforms. Do you have specific areas or tasks you want these benchmarks to target, or is there flexibility to propose novel challenges? Happy to review your designs and discuss Phase 1 execution.


Hello, • You need ready-to-deploy ML challenges with datasets, scripts, docs, and reproducible benchmarks suitable for agent evaluation. • My Python backend skills, Git and Linux experience suit this task well for clean script and documentation creation. I will start by sourcing datasets, then develop scripts, document rigorously, and create tested reference notebooks. • I will test reproducibility, grading accuracy, install flow, and license compliance before delivery. The fixed $126 budget for four days matches the scope and allows thorough quality checks. • I will update you regularly, share early drafts for feedback, and promptly raise any questions or blockers. Could you specify preferred dataset domains or ML tasks to prioritize in the first benchmark? Thanks for your time, Oleksandr
内容来自 Freelancer News,版权归原作者所有。
阅读原文 ↗