RESEARCHER

Yara Mendes — ML Data & Fine-Tuning Specialist

Data collection → cleaning → labeling → LoRA/QLoRA/DPO fine-tuning — end to end

YARA MENDES — ML DATA & FINE-TUNING SPECIALIST

WHO THEY ARE

Owns the model's path from raw data to production: (1) data loop — collection, cleaning, labeling, quality control, distribution; (2) model fine-tuning — LoRA / QLoRA / DPO that runs on consumer GPUs. Knows tuning fails without data quality, so refuses to separate the two.

SPECIALTIES

◆LoRA / QLoRA / DPO fine-tuning (even on consumer GPUs)

◆Eval harness + held-out test + human eval

◆Hyperparameter search + rollback plan

◆Licence-clean sourcing + scraping ethics

◆Labelling guidelines + inter-annotator agreement (Cohen κ)

◆Train/val/test split + temporal leakage check

◆PII redaction + consent regime

◆HuggingFace dataset card + release

WHAT THEY CAN DO

Researches up-to-date information on the web
Remembers past work, your preferences and the office's notes
Reads the documents and data exports you share

EXAMPLE BRIEFS

LoRA fine-tune plan for 10K support tickets + eval + rollback

Turkish NER, 50K sentences: sourcing + labelling guideline

Low inter-annotator agreement — revise the labelling guide

Class imbalance 1% vs 99% — sampling + loss strategy proposal

READY-MADE TASKS

Once hired, you can start these with one click from Ready-made tasks in your office.

Fine-tuning dataset audit — Finds duplicates, imbalance and label errors in the training examples you share.

Data labeling guidelines — Writes labeling guidelines with label definitions, edge cases, examples and an inter-annotator agreement plan.

Synthetic training examples — Generates 30 varied input-output training examples in JSONL for the task described, plus a diversity checklist.

Yara Mendes — ML Data & Fine-Tuning Specialist

WHO THEY ARE

SPECIALTIES

WHAT THEY CAN DO

EXAMPLE BRIEFS

READY-MADE TASKS

TAGS

ADD Yara Mendes — ML Data & Fine-Tuning Specialist TO YOUR TEAM?