吉良新

Key Skills

Software

Don't disclose

Top Subject Matter

Chinese text classification (NLP fine-tuning)

MLOps/AI training infrastructure for automated job execution

Top Data Types

Text

Image

Top Task Types

Fine-tuning

Data Collection

Freelancer Overview

Personal Lab Chinese Text Classification Training & Deployment Based on an Open-Source Model. Brings 1+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include Don't disclose. AI-training focus includes data types such as Text, Computer Code, and Programming and labeling workflows including Fine-tuning and Data Collection.

Labeling Experience

Personal Lab K8s Cluster Setup & CICD Pipeline (Supporting AI Scenarios)

Don't discloseData Collection

Designed a Kubernetes-based CI/CD pipeline to support automated AI training workflows and reproducible delivery of training environments. • Set up a 3-node Kubernetes cluster with Calico networking and NFS dynamic storage for datasets. • Deployed GitLab to manage training code and dataset versions. • Integrated Jenkins for automated building of training images. • Used Argo CD to watch configuration changes and automatically roll out new training jobs when training scripts were updated.

2025 - 2025

Personal Lab Chinese Text Classification Training & Deployment Based on an Open-Source Model

Don't discloseTextFine-tuning

Prepared and processed a Chinese text classification dataset of approximately 20,000 samples for model training. • Performed deduplication and tokenization using Python scripts. • Mapped raw inputs to classification labels and split the dataset into train/validation/test (70/20/10). • Used the prepared dataset to fine-tune a pre-trained BERT model and monitor performance metrics. • Saved the best model weights to a shared storage directory for subsequent evaluation and deployment.

2025 - 2025

Education

Work History