google-ai-edge/mediapipe-samples
Jun 2025 – Jun 2025
1 Month active
Languages Used
JSONJupyter NotebookPython
Technical Skills
Bug FixingColaboratoryData EngineeringDeep LearningFile ManagementFine-tuning
Arian Arfaian
PROFILE
Arian Arfaian
Arian Arfaian developed a Colab notebook for the google-ai-edge/mediapipe-samples repository to support the IO Connect 2025 workshop, focusing on enabling LoRA-based fine-tuning of the Gemma 3 1B model for text-to-SQL tasks. Using Python and Jupyter Notebooks, Arian updated generation prompts to improve accuracy and user guidance, while reorganizing the notebook structure for better maintainability and discoverability. The work included repository hygiene improvements, such as relocating files and ensuring clean, reproducible outputs by removing cell outputs and execution counts. This contribution demonstrated depth in deep learning, file management, and practical application of Hugging Face Transformers.
Overall Statistics
Feature vs Bugs
50%Features
Repository Contributions
4Total
Bugs
1
Commits
4
Features
1
Lines of code
20,762
Activity Months1
Your Network
21 people
Work History
June 2025
4 Commits • 1 Features
Jun 1, 2025In June 2025, delivered a Colab-focused contribution for the google-ai-edge/mediapipe-samples project to support the IO Connect 2025 workshop. The work centered on enabling LoRA-based fine-tuning for Gemma 3 1B to generate text-to-SQL, updating generation prompts, and reorganizing the notebook for maintainability and clarity. Also performed repository hygiene improvements by relocating the Colab notebook into the appropriate directory and ensuring clean, reproducible outputs for workshop participants.
4 Commits • 1 Features
Jun 1, 2025In June 2025, delivered a Colab-focused contribution for the google-ai-edge/mediapipe-samples project to support the IO Connect 2025 workshop. The work centered on enabling LoRA-based fine-tuning for Gemma 3 1B to generate text-to-SQL, updating generation prompts, and reorganizing the notebook for maintainability and clarity. Also performed repository hygiene improvements by relocating the Colab notebook into the appropriate directory and ensuring clean, reproducible outputs for workshop participants.
June 2025
Activity
Loading activity data...
Quality Metrics
Correctness95.0%
Maintainability95.0%
Architecture90.0%
Performance90.0%
AI Usage50.0%
Skills & Technologies
Programming Languages
JSONJupyter NotebookPython
Technical Skills
Bug FixingColaboratoryData EngineeringDeep LearningFile ManagementFine-tuningGoogle ColabHugging Face DatasetsHugging Face TRLHugging Face TransformersJupyter NotebooksLarge Language ModelsLoRAMachine LearningMediaPipe
Repositories Contributed To
Overview of all repositories you've contributed to across your timeline